From b382291281b8a4b857d7f3e181fe7edf4862a252 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Mon, 7 Oct 2024 15:35:00 +0200 Subject: [PATCH 01/41] Remove documentation duplicating the specification --- docs/domains/blas/asum.rst | 158 ----- docs/domains/blas/axpby.rst | 180 ------ docs/domains/blas/axpy.rst | 184 ------ docs/domains/blas/axpy_batch.rst | 350 ---------- docs/domains/blas/blas-level-1-routines.rst | 76 --- docs/domains/blas/blas-level-2-routines.rst | 105 --- docs/domains/blas/blas-level-3-routines.rst | 55 -- docs/domains/blas/blas-like-extensions.rst | 55 -- docs/domains/blas/blas.rst | 17 - docs/domains/blas/copy.rst | 159 ----- docs/domains/blas/copy_batch.rst | 328 ---------- docs/domains/blas/dgmm_batch.rst | 462 ------------- docs/domains/blas/dot.rst | 182 ------ docs/domains/blas/dotc.rst | 170 ----- docs/domains/blas/dotu.rst | 170 ----- docs/domains/blas/gbmv.rst | 285 -------- docs/domains/blas/gemm.rst | 455 ------------- docs/domains/blas/gemm_batch.rst | 606 ------------------ docs/domains/blas/gemm_bias.rst | 513 --------------- docs/domains/blas/gemmt.rst | 418 ------------ docs/domains/blas/gemv.rst | 261 -------- docs/domains/blas/gemv_batch.rst | 472 -------------- docs/domains/blas/ger.rst | 226 ------- docs/domains/blas/gerc.rst | 227 ------- docs/domains/blas/geru.rst | 227 ------- docs/domains/blas/hbmv.rst | 245 ------- docs/domains/blas/hemm.rst | 315 --------- docs/domains/blas/hemv.rst | 232 ------- docs/domains/blas/her.rst | 205 ------ docs/domains/blas/her2.rst | 231 ------- docs/domains/blas/her2k.rst | 397 ------------ docs/domains/blas/herk.rst | 309 --------- docs/domains/blas/hpmv.rst | 228 ------- docs/domains/blas/hpr.rst | 201 ------ docs/domains/blas/hpr2.rst | 226 ------- docs/domains/blas/iamax.rst | 167 ----- docs/domains/blas/iamin.rst | 160 ----- docs/domains/blas/nrm2.rst | 158 ----- docs/domains/blas/rot.rst | 208 ------ docs/domains/blas/rotg.rst | 175 ----- docs/domains/blas/rotm.rst | 266 -------- docs/domains/blas/rotmg.rst | 257 -------- docs/domains/blas/sbmv.rst | 244 ------- docs/domains/blas/scal.rst | 162 ----- docs/domains/blas/sdsdot.rst | 172 ----- docs/domains/blas/spmv.rst | 220 ------- docs/domains/blas/spr.rst | 193 ------ docs/domains/blas/spr2.rst | 213 ------ docs/domains/blas/swap.rst | 184 ------ docs/domains/blas/symm.rst | 311 --------- docs/domains/blas/symv.rst | 226 ------- docs/domains/blas/syr.rst | 202 ------ docs/domains/blas/syr2.rst | 228 ------- docs/domains/blas/syr2k.rst | 397 ------------ docs/domains/blas/syrk.rst | 296 --------- docs/domains/blas/syrk_batch.rst | 484 -------------- docs/domains/blas/tbmv.rst | 223 ------- docs/domains/blas/tbsv.rst | 225 ------- docs/domains/blas/tpmv.rst | 199 ------ docs/domains/blas/tpsv.rst | 207 ------ docs/domains/blas/trmm.rst | 288 --------- docs/domains/blas/trmv.rst | 210 ------ docs/domains/blas/trsm.rst | 286 --------- docs/domains/blas/trsm_batch.rst | 497 -------------- docs/domains/blas/trsv.rst | 215 ------- docs/domains/dense_linear_algebra.rst | 19 - docs/domains/lapack/gebrd.rst | 230 ------- docs/domains/lapack/gebrd_scratchpad_size.rst | 61 -- docs/domains/lapack/geqrf.rst | 157 ----- docs/domains/lapack/geqrf_batch.rst | 239 ------- .../lapack/geqrf_batch_scratchpad_size.rst | 111 ---- docs/domains/lapack/geqrf_scratchpad_size.rst | 64 -- docs/domains/lapack/gerqf.rst | 148 ----- docs/domains/lapack/gerqf_scratchpad_size.rst | 68 -- docs/domains/lapack/gesvd.rst | 344 ---------- docs/domains/lapack/gesvd_scratchpad_size.rst | 111 ---- docs/domains/lapack/getrf.rst | 144 ----- docs/domains/lapack/getrf_batch.rst | 226 ------- .../lapack/getrf_batch_scratchpad_size.rst | 117 ---- docs/domains/lapack/getrf_scratchpad_size.rst | 67 -- docs/domains/lapack/getri.rst | 138 ---- docs/domains/lapack/getri_batch.rst | 229 ------- .../lapack/getri_batch_scratchpad_size.rst | 111 ---- docs/domains/lapack/getri_scratchpad_size.rst | 66 -- docs/domains/lapack/getrs.rst | 200 ------ docs/domains/lapack/getrs_batch.rst | 286 --------- .../lapack/getrs_batch_scratchpad_size.rst | 135 ---- docs/domains/lapack/getrs_scratchpad_size.rst | 85 --- docs/domains/lapack/heevd.rst | 182 ------ docs/domains/lapack/heevd_scratchpad_size.rst | 81 --- docs/domains/lapack/hegvd.rst | 249 ------- docs/domains/lapack/hegvd_scratchpad_size.rst | 95 --- docs/domains/lapack/hetrd.rst | 206 ------ docs/domains/lapack/hetrd_scratchpad_size.rst | 74 --- docs/domains/lapack/hetrf.rst | 164 ----- docs/domains/lapack/hetrf_scratchpad_size.rst | 74 --- .../lapack/lapack-like-extensions.inc.rst | 74 --- .../lapack-linear-equation-routines.inc.rst | 121 ---- ...singular-value-eigenvalue-routines.inc.rst | 105 --- docs/domains/lapack/lapack.rst | 43 -- docs/domains/lapack/orgbr.rst | 226 ------- docs/domains/lapack/orgbr_scratchpad_size.rst | 90 --- docs/domains/lapack/orgqr.rst | 183 ------ docs/domains/lapack/orgqr_batch.rst | 262 -------- .../lapack/orgqr_batch_scratchpad_size.rst | 121 ---- docs/domains/lapack/orgqr_scratchpad_size.rst | 70 -- docs/domains/lapack/orgtr.rst | 148 ----- docs/domains/lapack/orgtr_scratchpad_size.rst | 67 -- docs/domains/lapack/ormqr.rst | 207 ------ docs/domains/lapack/ormqr_scratchpad_size.rst | 87 --- docs/domains/lapack/ormrq.rst | 208 ------ docs/domains/lapack/ormrq_scratchpad_size.rst | 81 --- docs/domains/lapack/ormtr.rst | 230 ------- docs/domains/lapack/ormtr_scratchpad_size.rst | 105 --- docs/domains/lapack/potrf.rst | 172 ----- docs/domains/lapack/potrf_batch.rst | 239 ------- .../lapack/potrf_batch_scratchpad_size.rst | 120 ---- docs/domains/lapack/potrf_scratchpad_size.rst | 77 --- docs/domains/lapack/potri.rst | 144 ----- docs/domains/lapack/potri_scratchpad_size.rst | 71 -- docs/domains/lapack/potrs.rst | 177 ----- docs/domains/lapack/potrs_batch.rst | 276 -------- .../lapack/potrs_batch_scratchpad_size.rst | 136 ---- docs/domains/lapack/potrs_scratchpad_size.rst | 77 --- docs/domains/lapack/syevd.rst | 186 ------ docs/domains/lapack/syevd_scratchpad_size.rst | 81 --- docs/domains/lapack/sygvd.rst | 249 ------- docs/domains/lapack/sygvd_scratchpad_size.rst | 92 --- docs/domains/lapack/sytrd.rst | 205 ------ docs/domains/lapack/sytrd_scratchpad_size.rst | 72 --- docs/domains/lapack/sytrf.rst | 166 ----- docs/domains/lapack/sytrf_scratchpad_size.rst | 77 --- docs/domains/lapack/trtrs.rst | 197 ------ docs/domains/lapack/trtrs_scratchpad_size.rst | 94 --- docs/domains/lapack/ungbr.rst | 231 ------- docs/domains/lapack/ungbr_scratchpad_size.rst | 90 --- docs/domains/lapack/ungqr.rst | 181 ------ docs/domains/lapack/ungqr_batch.rst | 274 -------- .../lapack/ungqr_batch_scratchpad_size.rst | 123 ---- docs/domains/lapack/ungqr_scratchpad_size.rst | 70 -- docs/domains/lapack/ungtr.rst | 153 ----- docs/domains/lapack/ungtr_scratchpad_size.rst | 67 -- docs/domains/lapack/unmqr.rst | 207 ------ docs/domains/lapack/unmqr_scratchpad_size.rst | 87 --- docs/domains/lapack/unmrq.rst | 207 ------ docs/domains/lapack/unmrq_scratchpad_size.rst | 79 --- docs/domains/lapack/unmtr.rst | 250 -------- docs/domains/lapack/unmtr_scratchpad_size.rst | 96 --- docs/domains/matrix-storage.rst | 581 ----------------- docs/index.rst | 2 - docs/onemkl-datatypes.rst | 140 ---- 151 files changed, 28858 deletions(-) delete mode 100644 docs/domains/blas/asum.rst delete mode 100644 docs/domains/blas/axpby.rst delete mode 100644 docs/domains/blas/axpy.rst delete mode 100644 docs/domains/blas/axpy_batch.rst delete mode 100644 docs/domains/blas/blas-level-1-routines.rst delete mode 100644 docs/domains/blas/blas-level-2-routines.rst delete mode 100644 docs/domains/blas/blas-level-3-routines.rst delete mode 100644 docs/domains/blas/blas-like-extensions.rst delete mode 100644 docs/domains/blas/blas.rst delete mode 100644 docs/domains/blas/copy.rst delete mode 100644 docs/domains/blas/copy_batch.rst delete mode 100644 docs/domains/blas/dgmm_batch.rst delete mode 100644 docs/domains/blas/dot.rst delete mode 100644 docs/domains/blas/dotc.rst delete mode 100644 docs/domains/blas/dotu.rst delete mode 100644 docs/domains/blas/gbmv.rst delete mode 100644 docs/domains/blas/gemm.rst delete mode 100644 docs/domains/blas/gemm_batch.rst delete mode 100644 docs/domains/blas/gemm_bias.rst delete mode 100644 docs/domains/blas/gemmt.rst delete mode 100644 docs/domains/blas/gemv.rst delete mode 100644 docs/domains/blas/gemv_batch.rst delete mode 100644 docs/domains/blas/ger.rst delete mode 100644 docs/domains/blas/gerc.rst delete mode 100644 docs/domains/blas/geru.rst delete mode 100644 docs/domains/blas/hbmv.rst delete mode 100644 docs/domains/blas/hemm.rst delete mode 100644 docs/domains/blas/hemv.rst delete mode 100644 docs/domains/blas/her.rst delete mode 100644 docs/domains/blas/her2.rst delete mode 100644 docs/domains/blas/her2k.rst delete mode 100644 docs/domains/blas/herk.rst delete mode 100644 docs/domains/blas/hpmv.rst delete mode 100644 docs/domains/blas/hpr.rst delete mode 100644 docs/domains/blas/hpr2.rst delete mode 100644 docs/domains/blas/iamax.rst delete mode 100644 docs/domains/blas/iamin.rst delete mode 100644 docs/domains/blas/nrm2.rst delete mode 100644 docs/domains/blas/rot.rst delete mode 100644 docs/domains/blas/rotg.rst delete mode 100644 docs/domains/blas/rotm.rst delete mode 100644 docs/domains/blas/rotmg.rst delete mode 100644 docs/domains/blas/sbmv.rst delete mode 100644 docs/domains/blas/scal.rst delete mode 100644 docs/domains/blas/sdsdot.rst delete mode 100644 docs/domains/blas/spmv.rst delete mode 100644 docs/domains/blas/spr.rst delete mode 100644 docs/domains/blas/spr2.rst delete mode 100644 docs/domains/blas/swap.rst delete mode 100644 docs/domains/blas/symm.rst delete mode 100644 docs/domains/blas/symv.rst delete mode 100644 docs/domains/blas/syr.rst delete mode 100644 docs/domains/blas/syr2.rst delete mode 100644 docs/domains/blas/syr2k.rst delete mode 100644 docs/domains/blas/syrk.rst delete mode 100644 docs/domains/blas/syrk_batch.rst delete mode 100644 docs/domains/blas/tbmv.rst delete mode 100644 docs/domains/blas/tbsv.rst delete mode 100644 docs/domains/blas/tpmv.rst delete mode 100644 docs/domains/blas/tpsv.rst delete mode 100644 docs/domains/blas/trmm.rst delete mode 100644 docs/domains/blas/trmv.rst delete mode 100644 docs/domains/blas/trsm.rst delete mode 100644 docs/domains/blas/trsm_batch.rst delete mode 100644 docs/domains/blas/trsv.rst delete mode 100644 docs/domains/dense_linear_algebra.rst delete mode 100644 docs/domains/lapack/gebrd.rst delete mode 100644 docs/domains/lapack/gebrd_scratchpad_size.rst delete mode 100644 docs/domains/lapack/geqrf.rst delete mode 100644 docs/domains/lapack/geqrf_batch.rst delete mode 100644 docs/domains/lapack/geqrf_batch_scratchpad_size.rst delete mode 100644 docs/domains/lapack/geqrf_scratchpad_size.rst delete mode 100644 docs/domains/lapack/gerqf.rst delete mode 100644 docs/domains/lapack/gerqf_scratchpad_size.rst delete mode 100644 docs/domains/lapack/gesvd.rst delete mode 100644 docs/domains/lapack/gesvd_scratchpad_size.rst delete mode 100644 docs/domains/lapack/getrf.rst delete mode 100644 docs/domains/lapack/getrf_batch.rst delete mode 100644 docs/domains/lapack/getrf_batch_scratchpad_size.rst delete mode 100644 docs/domains/lapack/getrf_scratchpad_size.rst delete mode 100644 docs/domains/lapack/getri.rst delete mode 100644 docs/domains/lapack/getri_batch.rst delete mode 100644 docs/domains/lapack/getri_batch_scratchpad_size.rst delete mode 100644 docs/domains/lapack/getri_scratchpad_size.rst delete mode 100644 docs/domains/lapack/getrs.rst delete mode 100644 docs/domains/lapack/getrs_batch.rst delete mode 100644 docs/domains/lapack/getrs_batch_scratchpad_size.rst delete mode 100644 docs/domains/lapack/getrs_scratchpad_size.rst delete mode 100644 docs/domains/lapack/heevd.rst delete mode 100644 docs/domains/lapack/heevd_scratchpad_size.rst delete mode 100644 docs/domains/lapack/hegvd.rst delete mode 100644 docs/domains/lapack/hegvd_scratchpad_size.rst delete mode 100644 docs/domains/lapack/hetrd.rst delete mode 100644 docs/domains/lapack/hetrd_scratchpad_size.rst delete mode 100644 docs/domains/lapack/hetrf.rst delete mode 100644 docs/domains/lapack/hetrf_scratchpad_size.rst delete mode 100644 docs/domains/lapack/lapack-like-extensions.inc.rst delete mode 100644 docs/domains/lapack/lapack-linear-equation-routines.inc.rst delete mode 100644 docs/domains/lapack/lapack-singular-value-eigenvalue-routines.inc.rst delete mode 100644 docs/domains/lapack/lapack.rst delete mode 100644 docs/domains/lapack/orgbr.rst delete mode 100644 docs/domains/lapack/orgbr_scratchpad_size.rst delete mode 100644 docs/domains/lapack/orgqr.rst delete mode 100644 docs/domains/lapack/orgqr_batch.rst delete mode 100644 docs/domains/lapack/orgqr_batch_scratchpad_size.rst delete mode 100644 docs/domains/lapack/orgqr_scratchpad_size.rst delete mode 100644 docs/domains/lapack/orgtr.rst delete mode 100644 docs/domains/lapack/orgtr_scratchpad_size.rst delete mode 100644 docs/domains/lapack/ormqr.rst delete mode 100644 docs/domains/lapack/ormqr_scratchpad_size.rst delete mode 100644 docs/domains/lapack/ormrq.rst delete mode 100644 docs/domains/lapack/ormrq_scratchpad_size.rst delete mode 100644 docs/domains/lapack/ormtr.rst delete mode 100644 docs/domains/lapack/ormtr_scratchpad_size.rst delete mode 100644 docs/domains/lapack/potrf.rst delete mode 100644 docs/domains/lapack/potrf_batch.rst delete mode 100644 docs/domains/lapack/potrf_batch_scratchpad_size.rst delete mode 100644 docs/domains/lapack/potrf_scratchpad_size.rst delete mode 100644 docs/domains/lapack/potri.rst delete mode 100644 docs/domains/lapack/potri_scratchpad_size.rst delete mode 100644 docs/domains/lapack/potrs.rst delete mode 100644 docs/domains/lapack/potrs_batch.rst delete mode 100644 docs/domains/lapack/potrs_batch_scratchpad_size.rst delete mode 100644 docs/domains/lapack/potrs_scratchpad_size.rst delete mode 100644 docs/domains/lapack/syevd.rst delete mode 100644 docs/domains/lapack/syevd_scratchpad_size.rst delete mode 100644 docs/domains/lapack/sygvd.rst delete mode 100644 docs/domains/lapack/sygvd_scratchpad_size.rst delete mode 100644 docs/domains/lapack/sytrd.rst delete mode 100644 docs/domains/lapack/sytrd_scratchpad_size.rst delete mode 100644 docs/domains/lapack/sytrf.rst delete mode 100644 docs/domains/lapack/sytrf_scratchpad_size.rst delete mode 100644 docs/domains/lapack/trtrs.rst delete mode 100644 docs/domains/lapack/trtrs_scratchpad_size.rst delete mode 100644 docs/domains/lapack/ungbr.rst delete mode 100644 docs/domains/lapack/ungbr_scratchpad_size.rst delete mode 100644 docs/domains/lapack/ungqr.rst delete mode 100644 docs/domains/lapack/ungqr_batch.rst delete mode 100644 docs/domains/lapack/ungqr_batch_scratchpad_size.rst delete mode 100644 docs/domains/lapack/ungqr_scratchpad_size.rst delete mode 100644 docs/domains/lapack/ungtr.rst delete mode 100644 docs/domains/lapack/ungtr_scratchpad_size.rst delete mode 100644 docs/domains/lapack/unmqr.rst delete mode 100644 docs/domains/lapack/unmqr_scratchpad_size.rst delete mode 100644 docs/domains/lapack/unmrq.rst delete mode 100644 docs/domains/lapack/unmrq_scratchpad_size.rst delete mode 100644 docs/domains/lapack/unmtr.rst delete mode 100644 docs/domains/lapack/unmtr_scratchpad_size.rst delete mode 100644 docs/domains/matrix-storage.rst delete mode 100644 docs/onemkl-datatypes.rst diff --git a/docs/domains/blas/asum.rst b/docs/domains/blas/asum.rst deleted file mode 100644 index 1fc02c84c..000000000 --- a/docs/domains/blas/asum.rst +++ /dev/null @@ -1,158 +0,0 @@ -.. _onemkl_blas_asum: - -asum -==== - -Computes the sum of magnitudes of the vector elements. - -.. _onemkl_blas_asum_description: - -.. rubric:: Description - -The ``asum`` routine computes the sum of the magnitudes of elements of a -real vector, or the sum of magnitudes of the real and imaginary parts -of elements of a complex vector: - -.. math:: - - result = \sum_{i=1}^{n}(|Re(x_i)| + |Im(x_i)|) - -where ``x`` is a vector with ``n`` elements. - -``asum`` supports the following precisions for data: - - .. list-table:: - :header-rows: 1 - - * - T - - T_res - * - ``float`` - - ``float`` - * - ``double`` - - ``double`` - * - ``std::complex`` - - ``float`` - * - ``std::complex`` - - ``double`` - -.. _onemkl_blas_asum_buffer: - -asum (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void asum(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &result) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void asum(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &result) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - -.. container:: section - - .. rubric:: Output Parameters - - result - Buffer where the scalar result is stored (the sum of magnitudes of - the real and imaginary parts of all elements of the vector). - - -.. _onemkl_blas_asum_usm: - -asum (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event asum(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - T_res *result, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event asum(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - T_res *result, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Pointer to input vector ``x``. The array holding the vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - result - Pointer to the output matrix where the scalar result is stored - (the sum of magnitudes of the real and imaginary parts of all - elements of the vector). - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/axpby.rst b/docs/domains/blas/axpby.rst deleted file mode 100644 index f95247fd6..000000000 --- a/docs/domains/blas/axpby.rst +++ /dev/null @@ -1,180 +0,0 @@ -.. _onemkl_blas_axpby: - -axpby -===== - -Computes a vector-scalar product added to a scaled-vector. - -.. _onemkl_blas_axpby_description: - -.. rubric:: Description - -The ``axpby`` routines compute two scalar-vector product and add them: - -.. math:: - - y \leftarrow beta * y + alpha * x - -where ``x`` and ``y`` are vectors of ``n`` elements and ``alpha`` and ``beta`` are scalars. - -``axpby`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_axpby_buffer: - -axpby (Buffer Version) ----------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void axpby(sycl::queue &queue, - std::int64_t n, - T alpha, - sycl::buffer &x, std::int64_t incx, - T beta, - sycl::buffer &y, std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void axpby(sycl::queue &queue, - std::int64_t n, - T alpha, - sycl::buffer &x, std::int64_t incx, - T beta, - sycl::buffer &y, std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x`` and ``y``. - - alpha - Specifies the scalar ``alpha``. - - x - Buffer holding input vector ``x``. The buffer must be of size at least - (1 + (``n`` – 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride between two consecutive elements of the ``x`` vector. - - beta - Specifies the scalar ``beta``. - - y - Buffer holding input vector ``y``. The buffer must be of size at least - (1 + (``n`` – 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride between two consecutive elements of the ``y`` vector. - -.. container:: section - - .. rubric:: Output Parameters - - y - Buffer holding the updated vector ``y``. - - -.. _onemkl_blas_axpby_usm: - -axpby (USM Version) -------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event axpby(sycl::queue &queue, - std::int64_t n, - T alpha, - const T *x, std::int64_t incx, - const T beta, - T *y, std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event axpby(sycl::queue &queue, - std::int64_t n, - T alpha, - const T *x, std::int64_t incx, - const T beta, - T *y, std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x`` and ``y``. - - alpha - Specifies the scalar alpha. - - beta - Specifies the scalar beta. - - x - Pointer to the input vector ``x``. The allocated memory must be - of size at least (1 + (``n`` – 1)*abs(``incx``)). See - :ref:`matrix-storage` for more details. - - incx - Stride between consecutive elements of the ``x`` vector. - - y - Pointer to the input vector ``y``. The allocated memory must be - of size at least (1 + (``n`` – 1)*abs(``incy``)). See - :ref:`matrix-storage` for more details. - - incy - Stride between consecutive elements of the ``y`` vector. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Array holding the updated vector ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-like-extensions` - diff --git a/docs/domains/blas/axpy.rst b/docs/domains/blas/axpy.rst deleted file mode 100644 index a3f5a69e3..000000000 --- a/docs/domains/blas/axpy.rst +++ /dev/null @@ -1,184 +0,0 @@ -.. _onemkl_blas_axpy: - -axpy -==== - -Computes a vector-scalar product and adds the result to a vector. - -.. _onemkl_blas_axpy_description: - -.. rubric:: Description - -The ``axpy`` routines compute a scalar-vector product and add the result -to a vector: - -.. math:: - - y \leftarrow alpha * x + y - -where: - -``x`` and ``y`` are vectors of ``n`` elements, - -``alpha`` is a scalar. - -``axpy`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_axpy_buffer: - -axpy (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void axpy(sycl::queue &queue, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void axpy(sycl::queue &queue, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - alpha - Specifies the scalar alpha. - - x - Buffer holding input vector ``x``. The buffer must be of size at least - (1 + (``n`` – 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input vector ``y``. The buffer must be of size at least - (1 + (``n`` – 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - y - Buffer holding the updated vector ``y``. - - -.. _onemkl_blas_axpy_usm: - -axpy (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event axpy(sycl::queue &queue, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event axpy(sycl::queue &queue, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - alpha - Specifies the scalar alpha. - - x - Pointer to the input vector ``x``. The array holding the vector - ``x`` must be of size at least (1 + (``n`` – 1)*abs(``incx``)). See - :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to the input vector ``y``. The array holding the vector - ``y`` must be of size at least (1 + (``n`` – 1)*abs(``incy``)). See - :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Pointer to the updated vector ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/axpy_batch.rst b/docs/domains/blas/axpy_batch.rst deleted file mode 100644 index 308ed2b29..000000000 --- a/docs/domains/blas/axpy_batch.rst +++ /dev/null @@ -1,350 +0,0 @@ -.. _onemkl_blas_axpy_batch: - -axpy_batch -========== - -Computes a group of ``axpy`` operations. - -.. _onemkl_blas_axpy_batch_description: - -.. rubric:: Description - -The ``axpy_batch`` routines are batched versions of :ref:`onemkl_blas_axpy`, performing -multiple ``axpy`` operations in a single call. Each ``axpy`` -operation adds a scalar-vector product to a vector. - -``axpy_batch`` supports the following precisions for data. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_axpy_batch_buffer: - -axpy_batch (Buffer Version) ---------------------------- - -.. rubric:: Description - -The buffer version of ``axpy_batch`` supports only the strided API. - -The strided API operation is defined as: -:: - - for i = 0 … batch_size – 1 - X and Y are vectors at offset i * stridex, i * stridey in x and y - Y := alpha * X + Y - end for - -where: - -``alpha`` is scalar, - -``X`` and ``Y`` are vectors. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void axpy_batch(sycl::queue &queue, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - std::int64_t stridex, - sycl::buffer &y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void axpy_batch(sycl::queue &queue, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - std::int64_t stridex, - sycl::buffer &y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in ``X`` and ``Y``. - - alpha - Specifies the scalar ``alpha``. - - x - Buffer holding input vectors ``X`` with size ``stridex`` * ``batch_size``. - - incx - Stride of vector ``X``. - - stridex - Stride between different ``X`` vectors. - - y - Buffer holding input/output vectors ``Y`` with size ``stridey`` * ``batch_size``. - - incy - Stride of vector ``Y``. - - stridey - Stride between different ``Y`` vectors. - - batch_size - Specifies the number of ``axpy`` operations to perform. - -.. container:: section - - .. rubric:: Output Parameters - - y - Output buffer, overwritten by ``batch_size`` ``axpy`` operations of the form - ``alpha`` * ``X`` + ``Y``. - - -.. _onemkl_blas_axpy_batch_usm: - -axpy_batch (USM Version) ------------------------- - -.. rubric:: Description - -The USM version of ``axpy_batch`` supports the group API and strided API. - -The group API operation is defined as -:: - - idx = 0 - for i = 0 … group_count – 1 - for j = 0 … group_size – 1 - X and Y are vectors in x[idx] and y[idx] - Y := alpha[i] * X + Y - idx := idx + 1 - end for - end for - -The strided API operation is defined as -:: - - for i = 0 … batch_size – 1 - X and Y are vectors at offset i * stridex, i * stridey in x and y - Y := alpha * X + Y - end for - -where: - -``alpha`` is scalar, - -``X`` and ``Y`` are vectors. - -For group API, ``x`` and ``y`` arrays contain the pointers for all the input vectors. -The total number of vectors in ``x`` and ``y`` are given by: - -.. math:: - - total\_batch\_count = \sum_{i=0}^{group\_count-1}group\_size[i] - -For strided API, ``x`` and ``y`` arrays contain all the input vectors. -The total number of vectors in ``x`` and ``y`` are given by the ``batch_size`` parameter. - -**Group API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event axpy_batch(sycl::queue &queue, - std::int64_t *n, - T *alpha, - const T **x, - std::int64_t *incx, - T **y, - std::int64_t *incy, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event axpy_batch(sycl::queue &queue, - std::int64_t *n, - T *alpha, - const T **x, - std::int64_t *incx, - T **y, - std::int64_t *incy, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Array of ``group_count`` integers. ``n[i]`` specifies the number of elements in vectors ``X`` and ``Y`` for every vector in group ``i``. - - alpha - Array of ``group_count`` scalar elements. ``alpha[i]`` specifies the scaling factor for vector ``X`` in group ``i``. - - x - Array of pointers to input vectors ``X`` with size ``total_batch_count``. - The size of array allocated for the ``X`` vector of the group ``i`` must be at least (1 + (``n[i]`` – 1)*abs(``incx[i]``)). - See :ref:`matrix-storage` for more details. - - incx - Array of ``group_count`` integers. ``incx[i]`` specifies the stride of vector ``X`` in group ``i``. - - y - Array of pointers to input/output vectors ``Y`` with size ``total_batch_count``. - The size of array allocated for the ``Y`` vector of the group ``i`` must be at least (1 + (``n[i]`` – 1)*abs(``incy[i]``)). - See :ref:`matrix-storage` for more details. - - incy - Array of ``group_count`` integers. ``incy[i]`` specifies the stride of vector ``Y`` in group ``i``. - - group_count - Number of groups. Must be at least 0. - - group_size - Array of ``group_count`` integers. ``group_size[i]`` specifies the number of ``axpy`` operations in group ``i``. - Each element in ``group_size`` must be at least 0. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Array of pointers holding the ``Y`` vectors, overwritten by ``total_batch_count`` ``axpy`` operations of the form - ``alpha`` * ``X`` + ``Y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event axpy_batch(sycl::queue &queue, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - std::int64_t stridex, - T *y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event axpy_batch(sycl::queue &queue, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - std::int64_t stridex, - T *y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in ``X`` and ``Y``. - - alpha - Specifies the scalar ``alpha``. - - x - Pointer to input vectors ``X`` with size ``stridex`` * ``batch_size``. - - incx - Stride of vector ``X``. - - stridex - Stride between different ``X`` vectors. - - y - Pointer to input/output vectors ``Y`` with size ``stridey`` * ``batch_size``. - - incy - Stride of vector ``Y``. - - stridey - Stride between different ``Y`` vectors. - - batch_size - Specifies the number of ``axpy`` operations to perform. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Output vectors, overwritten by ``batch_size`` ``axpy`` operations of the form - ``alpha`` * ``X`` + ``Y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:**:ref:`blas-like-extensions` diff --git a/docs/domains/blas/blas-level-1-routines.rst b/docs/domains/blas/blas-level-1-routines.rst deleted file mode 100644 index c96c2d54c..000000000 --- a/docs/domains/blas/blas-level-1-routines.rst +++ /dev/null @@ -1,76 +0,0 @@ -.. _blas-level-1-routines: - -BLAS Level 1 Routines -===================== - - -.. container:: - - - BLAS Level 1 includes routines which perform - vector-vector operations as described in the following table. - - - .. container:: tablenoborder - - - .. list-table:: - :header-rows: 1 - - * - Routines - - Description - * - :ref:`onemkl_blas_asum` - - Sum of vector magnitudes - * - :ref:`onemkl_blas_axpy` - - Scalar-vector product - * - :ref:`onemkl_blas_copy` - - Copy vector - * - :ref:`onemkl_blas_dot` - - Dot product - * - :ref:`onemkl_blas_sdsdot` - - Dot product with double precision - * - :ref:`onemkl_blas_dotc` - - Dot product conjugated - * - :ref:`onemkl_blas_dotu` - - Dot product unconjugated - * - :ref:`onemkl_blas_nrm2` - - Vector 2-norm (Euclidean norm) - * - :ref:`onemkl_blas_rot` - - Plane rotation of points - * - :ref:`onemkl_blas_rotg` - - Generate Givens rotation of points - * - :ref:`onemkl_blas_rotm` - - Modified Givens plane rotation of points - * - :ref:`onemkl_blas_rotmg` - - Generate modified Givens plane rotation of points - * - :ref:`onemkl_blas_scal` - - Vector-scalar product - * - :ref:`onemkl_blas_swap` - - Vector-vector swap - * - :ref:`onemkl_blas_iamax` - - Index of the maximum absolute value element of a vector - * - :ref:`onemkl_blas_iamin` - - Index of the minimum absolute value element of a vector - -.. toctree:: - :hidden: - - asum - axpy - copy - dot - sdsdot - dotc - dotu - nrm2 - rot - rotg - rotm - rotmg - scal - swap - iamax - iamin - - -**Parent topic:** :ref:`onemkl_blas` diff --git a/docs/domains/blas/blas-level-2-routines.rst b/docs/domains/blas/blas-level-2-routines.rst deleted file mode 100644 index 427acbc9b..000000000 --- a/docs/domains/blas/blas-level-2-routines.rst +++ /dev/null @@ -1,105 +0,0 @@ -.. _blas-level-2-routines: - -BLAS Level 2 Routines -===================== - - -.. container:: - - - BLAS Level 2 includes routines which perform - matrix-vector operations as described in the following table. - - - .. container:: tablenoborder - - - .. list-table:: - :header-rows: 1 - - * - Routines - - Description - * - :ref:`onemkl_blas_gbmv` - - Matrix-vector product using a general band matrix - * - :ref:`onemkl_blas_gemv` - - Matrix-vector product using a general matrix - * - :ref:`onemkl_blas_ger` - - Rank-1 update of a general matrix - * - :ref:`onemkl_blas_gerc` - - Rank-1 update of a conjugated general matrix - * - :ref:`onemkl_blas_geru` - - Rank-1 update of a general matrix, unconjugated - * - :ref:`onemkl_blas_hbmv` - - Matrix-vector product using a Hermitian band matrix - * - :ref:`onemkl_blas_hemv` - - Matrix-vector product using a Hermitian matrix - * - :ref:`onemkl_blas_her` - - Rank-1 update of a Hermitian matrix - * - :ref:`onemkl_blas_her2` - - Rank-2 update of a Hermitian matrix - * - :ref:`onemkl_blas_hpmv` - - Matrix-vector product using a Hermitian packed matrix - * - :ref:`onemkl_blas_hpr` - - Rank-1 update of a Hermitian packed matrix - * - :ref:`onemkl_blas_hpr2` - - Rank-2 update of a Hermitian packed matrix - * - :ref:`onemkl_blas_sbmv` - - Matrix-vector product using symmetric band matrix - * - :ref:`onemkl_blas_spmv` - - Matrix-vector product using a symmetric packed matrix - * - :ref:`onemkl_blas_spr` - - Rank-1 update of a symmetric packed matrix - * - :ref:`onemkl_blas_spr2` - - Rank-2 update of a symmetric packed matrix - * - :ref:`onemkl_blas_symv` - - Matrix-vector product using a symmetric matrix - * - :ref:`onemkl_blas_syr` - - Rank-1 update of a symmetric matrix - * - :ref:`onemkl_blas_syr2` - - Rank-2 update of a symmetric matrix - * - :ref:`onemkl_blas_tbmv` - - Matrix-vector product using a triangular band matrix - * - :ref:`onemkl_blas_tbsv` - - Solution of a linear system of equations with a triangular band matrix - * - :ref:`onemkl_blas_tpmv` - - Matrix-vector product using a triangular packed matrix - * - :ref:`onemkl_blas_tpsv` - - Solution of a linear system of equations with a triangular packed matrix - * - :ref:`onemkl_blas_trmv` - - Matrix-vector product using a triangular matrix - * - :ref:`onemkl_blas_trsv` - - Solution of a linear system of equations with a triangular matrix - - - - -.. toctree:: - :hidden: - - gbmv - gemv - ger - gerc - geru - hbmv - hemv - her - her2 - hpmv - hpr - hpr2 - sbmv - spmv - spr - spr2 - symv - syr - syr2 - tbmv - tbsv - tpmv - tpsv - trmv - trsv - -**Parent topic:** :ref:`onemkl_blas` diff --git a/docs/domains/blas/blas-level-3-routines.rst b/docs/domains/blas/blas-level-3-routines.rst deleted file mode 100644 index bb7f3f4d6..000000000 --- a/docs/domains/blas/blas-level-3-routines.rst +++ /dev/null @@ -1,55 +0,0 @@ -.. _blas-level-3-routines: - -BLAS Level 3 Routines -===================== - - -.. container:: - - BLAS Level 3 includes routines which perform - matrix-matrix operations as described in the following table. - - - .. container:: tablenoborder - - - .. list-table:: - :header-rows: 1 - - * - Routines - - Description - * - :ref:`onemkl_blas_gemm` - - Computes a matrix-matrix product with general matrices. - * - :ref:`onemkl_blas_hemm` - - Computes a matrix-matrix product where one input matrix is Hermitian and one is general. - * - :ref:`onemkl_blas_herk` - - Performs a Hermitian rank-k update. - * - :ref:`onemkl_blas_her2k` - - Performs a Hermitian rank-2k update. - * - :ref:`onemkl_blas_symm` - - Computes a matrix-matrix product where one input matrix is symmetric and one matrix is general. - * - :ref:`onemkl_blas_syrk` - - Performs a symmetric rank-k update. - * - :ref:`onemkl_blas_syr2k` - - Performs a symmetric rank-2k update. - * - :ref:`onemkl_blas_trmm` - - Computes a matrix-matrix product where one input matrix is triangular and one input matrix is general. - * - :ref:`onemkl_blas_trsm` - - Solves a triangular matrix equation (forward or backward solve). - - - -.. toctree:: - :hidden: - - gemm - hemm - herk - her2k - symm - syrk - syr2k - trmm - trsm - -**Parent topic:** :ref:`onemkl_blas` diff --git a/docs/domains/blas/blas-like-extensions.rst b/docs/domains/blas/blas-like-extensions.rst deleted file mode 100644 index f447e2f9e..000000000 --- a/docs/domains/blas/blas-like-extensions.rst +++ /dev/null @@ -1,55 +0,0 @@ -.. _blas-like-extensions: - -BLAS-like Extensions -==================== - - -.. container:: - - - oneAPI Math Kernel Library DPC++ provides additional routines to - extend the functionality of the BLAS routines. These include routines - to compute many independent vector-vector and matrix-matrix operations. - - The following table lists the BLAS-like extensions with their descriptions. - - - .. container:: tablenoborder - - - .. list-table:: - :header-rows: 1 - - * - Routines - - Description - * - :ref:`onemkl_blas_axpy_batch` - - Computes groups of vector-scalar products added to a vector. - * - :ref:`onemkl_blas_gemm_batch` - - Computes groups of matrix-matrix products with general matrices. - * - :ref:`onemkl_blas_trsm_batch` - - Solves a triangular matrix equation for a group of matrices. - * - :ref:`onemkl_blas_gemmt` - - Computes a matrix-matrix product with general matrices, but updates - only the upper or lower triangular part of the result matrix. - * - :ref:`onemkl_blas_gemm_bias` - - Computes a matrix-matrix product using general integer matrices with bias - - - - - -.. toctree:: - :hidden: - - axpy_batch - axpby - copy_batch - dgmm_batch - gemm_batch - gemv_batch - syrk_batch - trsm_batch - gemmt - gemm_bias - -**Parent topic:** :ref:`onemkl_blas` diff --git a/docs/domains/blas/blas.rst b/docs/domains/blas/blas.rst deleted file mode 100644 index 50411efb8..000000000 --- a/docs/domains/blas/blas.rst +++ /dev/null @@ -1,17 +0,0 @@ -.. _onemkl_blas: - -BLAS Routines -+++++++++++++ - -oneMKL provides DPC++ interfaces to the Basic Linear Algebra Subprograms (BLAS) routines (Level1, Level2, Level3), as well as several BLAS-like extension routines. - -.. toctree:: - :maxdepth: 1 - - blas-level-1-routines.rst - blas-level-2-routines.rst - blas-level-3-routines.rst - blas-like-extensions.rst - - -**Parent topic:** :ref:`onemkl_dense_linear_algebra` diff --git a/docs/domains/blas/copy.rst b/docs/domains/blas/copy.rst deleted file mode 100644 index e09db618e..000000000 --- a/docs/domains/blas/copy.rst +++ /dev/null @@ -1,159 +0,0 @@ -.. _onemkl_blas_copy: - -copy -==== - -Copies a vector to another vector. - -.. _onemkl_blas_copy_description: - -.. rubric:: Description - -The ``copy`` routines copy one vector to another: - -.. math:: - - y \leftarrow x - -where ``x`` and ``y`` are vectors of n elements. - -``copy`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - - -.. _onemkl_blas_copy_buffer: - -copy (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void copy(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void copy(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Buffer holding input vector ``x``. The buffer must be of size at least - (1 + (``n`` – 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - y - Buffer holding the updated vector ``y``. - - -.. _onemkl_blas_copy_usm: - -copy (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event copy(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event copy(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Pointer to the input vector ``x``. The array holding the vector - ``x`` must be of size at least (1 + (``n`` – 1)*abs(``incx``)). See - :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Pointer to the updated vector ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/copy_batch.rst b/docs/domains/blas/copy_batch.rst deleted file mode 100644 index b02d71c56..000000000 --- a/docs/domains/blas/copy_batch.rst +++ /dev/null @@ -1,328 +0,0 @@ -.. _onemkl_blas_copy_batch: - -copy_batch -========== - -Computes a group of ``copy`` operations. - -.. _onemkl_blas_copy_batch_description: - -.. rubric:: Description - -The ``copy_batch`` routines are batched versions of :ref:`onemkl_blas_copy`, performing -multiple ``copy`` operations in a single call. Each ``copy`` -operation copies one vector to another. - -``copy_batch`` supports the following precisions for data. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_copy_batch_buffer: - -copy_batch (Buffer Version) ---------------------------- - -.. rubric:: Description - -The buffer version of ``copy_batch`` supports only the strided API. - -The strided API operation is defined as: -:: - - for i = 0 … batch_size – 1 - X and Y are vectors at offset i * stridex, i * stridey in x and y - Y := X - end for - -where: - -``X`` and ``Y`` are vectors. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void copy_batch(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - std::int64_t stridex, - sycl::buffer &y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void copy_batch(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - std::int64_t stridex, - sycl::buffer &y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in ``X`` and ``Y``. - - x - Buffer holding input vectors ``X`` with size ``stridex`` * ``batch_size``. - - incx - Stride of vector ``X``. - - stridex - Stride between different ``X`` vectors. - - y - Buffer holding input/output vectors ``Y`` with size ``stridey`` * ``batch_size``. - - incy - Stride of vector ``Y``. - - stridey - Stride between different ``Y`` vectors. - - batch_size - Specifies the number of ``copy`` operations to perform. - -.. container:: section - - .. rubric:: Output Parameters - - y - Output buffer, overwritten by ``batch_size`` ``copy`` operations. - - -.. _onemkl_blas_copy_batch_usm: - -copy_batch (USM Version) ------------------------- - -.. rubric:: Description - -The USM version of ``copy_batch`` supports the group API and strided API. - -The group API operation is defined as -:: - - idx = 0 - for i = 0 … group_count – 1 - for j = 0 … group_size – 1 - X and Y are vectors in x[idx] and y[idx] - Y := X - idx := idx + 1 - end for - end for - -The strided API operation is defined as -:: - - for i = 0 … batch_size – 1 - X and Y are vectors at offset i * stridex, i * stridey in x and y - Y := X - end for - -where: - -``X`` and ``Y`` are vectors. - -For group API, ``x`` and ``y`` arrays contain the pointers for all the input vectors. -The total number of vectors in ``x`` and ``y`` are given by: - -.. math:: - - total\_batch\_count = \sum_{i=0}^{group\_count-1}group\_size[i] - -For strided API, ``x`` and ``y`` arrays contain all the input vectors. -The total number of vectors in ``x`` and ``y`` are given by the ``batch_size`` parameter. - -**Group API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event copy_batch(sycl::queue &queue, - std::int64_t *n, - const T **x, - std::int64_t *incx, - T **y, - std::int64_t *incy, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event copy_batch(sycl::queue &queue, - std::int64_t *n, - const T **x, - std::int64_t *incx, - T **y, - std::int64_t *incy, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Array of ``group_count`` integers. ``n[i]`` specifies the number of elements in vectors ``X`` and ``Y`` for every vector in group ``i``. - - x - Array of pointers to input vectors ``X`` with size ``total_batch_count``. - The size of array allocated for the ``X`` vector of the group ``i`` must be at least (1 + (``n[i]`` – 1)*abs(``incx[i]``)). - See :ref:`matrix-storage` for more details. - - incx - Array of ``group_count`` integers. ``incx[i]`` specifies the stride of vector ``X`` in group ``i``. - - y - Array of pointers to input/output vectors ``Y`` with size ``total_batch_count``. - The size of array allocated for the ``Y`` vector of the group ``i`` must be at least (1 + (``n[i]`` – 1)*abs(``incy[i]``)). - See :ref:`matrix-storage` for more details. - - incy - Array of ``group_count`` integers. ``incy[i]`` specifies the stride of vector ``Y`` in group ``i``. - - group_count - Number of groups. Must be at least 0. - - group_size - Array of ``group_count`` integers. ``group_size[i]`` specifies the number of ``copy`` operations in group ``i``. - Each element in ``group_size`` must be at least 0. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Array of pointers holding the ``Y`` vectors, overwritten by ``total_batch_count`` ``copy`` operations. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event copy_batch(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - std::int64_t stridex, - T *y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event copy_batch(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - std::int64_t stridex, - T *y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in ``X`` and ``Y``. - - x - Pointer to input vectors ``X`` with size ``stridex`` * ``batch_size``. - - incx - Stride of vector ``X``. - - stridex - Stride between different ``X`` vectors. - - y - Pointer to input/output vectors ``Y`` with size ``stridey`` * ``batch_size``. - - incy - Stride of vector ``Y``. - - stridey - Stride between different ``Y`` vectors. - - batch_size - Specifies the number of ``copy`` operations to perform. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Output vectors, overwritten by ``batch_size`` ``copy`` operations - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:**:ref:`blas-like-extensions` diff --git a/docs/domains/blas/dgmm_batch.rst b/docs/domains/blas/dgmm_batch.rst deleted file mode 100644 index 25eaace9f..000000000 --- a/docs/domains/blas/dgmm_batch.rst +++ /dev/null @@ -1,462 +0,0 @@ -.. _onemkl_blas_dgmm_batch: - -dgmm_batch -========== - -Computes a group of ``dgmm`` operations. - -.. _onemkl_blas_dgmm_batch_description: - -.. rubric:: Description - -The ``dgmm_batch`` routines perform -multiple diagonal matrix-matrix product operations in a single call. - -``dgmm_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_dgmm_batch_buffer: - -dgmm_batch (Buffer Version) ---------------------------- - -.. rubric:: Description - -The buffer version of ``dgmm_batch`` supports only the strided API. - -The strided API operation is defined as: -:: - - for i = 0 … batch_size – 1 - A and C are matrices at offset i * stridea in a, i * stridec in c. - X is a vector at offset i * stridex in x - C := diag(X) * A or C = A * diag(X) - end for - -where: - -``A`` is a matrix, - -``X`` is a diagonal matrix stored as a vector - -The ``a`` and ``x`` buffers contain all the input matrices. The stride -between matrices is given by the stride parameter. The total number -of matrices in ``a`` and ``x`` buffers is given by the ``batch_size`` parameter. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void dgmm_batch(sycl::queue &queue, - onemkl::mkl::side left_right, - std::int64_t m, - std::int64_t n, - sycl::buffer &a, - std::int64_t lda, - std::int64_t stridea, - sycl::buffer &x, - std::int64_t incx, - std::int64_t stridex, - sycl::buffer &c, - std::int64_t ldc, - std::int64_t stridec, - std::int64_t batch_size) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void dgmm_batch(sycl::queue &queue, - onemkl::mkl::side left_right, - std::int64_t m, - std::int64_t n, - sycl::buffer &a, - std::int64_t lda, - std::int64_t stridea, - sycl::buffer &x, - std::int64_t incx, - std::int64_t stridex, - sycl::buffer &c, - std::int64_t ldc, - std::int64_t stridec, - std::int64_t batch_size) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies the position of the diagonal matrix in the product. - See :ref:`onemkl_datatypes` for more details. - - m - Number of rows of matrices ``A`` and ``C``. Must be at least zero. - - n - Number of columns of matrices ``A`` and ``C``. Must be at least zero. - - a - - Buffer holding the input matrices ``A`` with size ``stridea`` * - ``batch_size``. Must be of at least ``lda`` * ``j`` + - ``stridea`` * (``batch_size`` - 1) where j is n if column major - layout is used or m if major layout is used. - - lda - The leading dimension of the matrices ``A``. It must be positive - and at least ``m`` if column major layout is used or at least - ``n`` if row major layout is used. - - stridea - Stride between different ``A`` matrices. - - x - Buffer holding the input matrices ``X`` with size ``stridex`` * - ``batch_size``. Must be of size at least - (1 + (``len`` - 1)*abs(``incx``)) + ``stridex`` * (``batch_size`` - 1) - where ``len`` is ``n`` if the diagonal matrix is on the right - of the product or ``m`` otherwise. - - incx - Stride between two consecutive elements of the ``x`` vectors. - - stridex - Stride between different ``X`` vectors, must be at least 0. - - c - Buffer holding input/output matrices ``C`` with size ``stridec`` * ``batch_size``. - - ldc - The leading dimension of the matrices ``C``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if column major layout is used to store matrices. - - stridec - Stride between different ``C`` matrices. Must be at least - ``ldc`` * ``n`` if column major layout is used or ``ldc`` * ``m`` if row - major layout is used. - - batch_size - Specifies the number of diagonal matrix-matrix product operations to perform. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output overwritten by ``batch_size`` diagonal matrix-matrix product - operations. - - -.. _onemkl_blas_dgmm_batch_usm: - -dgmm_batch (USM Version) ---------------------------- - -.. rubric:: Description - -The USM version of ``dgmm_batch`` supports the group API and strided API. - -The group API operation is defined as: -:: - - idx = 0 - for i = 0 … group_count – 1 - for j = 0 … group_size – 1 - a and c are matrices of size mxn at position idx in a_array and c_array - x is a vector of size m or n depending on left_right, at position idx in x_array - if (left_right == oneapi::mkl::side::left) - c := diag(x) * a - else - c := a * diag(x) - idx := idx + 1 - end for - end for - -The strided API operation is defined as -:: - - for i = 0 … batch_size – 1 - A and C are matrices at offset i * stridea in a, i * stridec in c. - X is a vector at offset i * stridex in x - C := diag(X) * A or C = A * diag(X) - end for - -where: - -``A`` is a matrix, - -``X`` is a diagonal matrix stored as a vector - -The ``a`` and ``x`` buffers contain all the input matrices. The stride -between matrices is given by the stride parameter. The total number -of matrices in ``a`` and ``x`` buffers is given by the ``batch_size`` parameter. - -For group API, ``a`` and ``x`` arrays contain the pointers for all the input matrices. -The total number of matrices in ``a`` and ``x`` are given by: - -.. math:: - - total\_batch\_count = \sum_{i=0}^{group\_count-1}group\_size[i] - -For strided API, ``a`` and ``x`` arrays contain all the input matrices. The total number of matrices -in ``a`` and ``x`` are given by the ``batch_size`` parameter. - -**Group API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event dgmm_batch(sycl::queue &queue, - onemkl::mkl::side *left_right, - std::int64_t *m, - std::int64_t *n, - const T **a, - std::int64_t *lda, - const T **x, - std::int64_t *incx, - T **c, - std::int64_t *ldc, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event dgmm_batch(sycl::queue &queue, - onemkl::mkl::side *left_right, - std::int64_t *m, - std::int64_t *n, - const T **a, - std::int64_t *lda, - const T **x, - std::int64_t *incx, - T **c, - std::int64_t *ldc, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies the position of the diagonal matrix in the product. - See :ref:`onemkl_datatypes` for more details. - - m - Array of ``group_count`` integers. ``m[i]`` specifies the - number of rows of ``A`` for every matrix in group ``i``. All entries must be at least zero. - - n - Array of ``group_count`` integers. ``n[i]`` specifies the - number of columns of ``A`` for every matrix in group ``i``. All entries must be at least zero. - - a - Array of pointers to input matrices ``A`` with size - ``total_batch_count``. Must be of size at least ``lda[i]`` * ``n[i]`` if - column major layout is used or at least ``lda[i]`` * ``m[i]`` if row major - layout is used. - See :ref:`matrix-storage` for more details. - - lda - Array of ``group_count`` integers. ``lda[i]`` specifies the - leading dimension of ``A`` for every matrix in group ``i``. All - entries must be positive and at least ``m[i]`` if column major - layout is used or at least ``n[i]`` if row major layout is used. - - x - Array of pointers to input vectors ``X`` with size - ``total_batch_count``. Must be of size at least (1 + ``len[i]`` – - 1)*abs(``incx[i]``)) where ``len[i]`` is ``n[i]`` if the diagonal matrix is on the - right of the product or ``m[i]`` otherwise. - See :ref:`matrix-storage` for more details. - - incx - Array of ``group_count`` integers. ``incx[i]`` specifies the - stride of ``x`` for every vector in group ``i``. All entries - must be positive. - c - Array of pointers to input/output matrices ``C`` with size ``total_batch_count``. - Must be of size at least - ``ldc[i]`` * ``n[i]`` - if column major layout is used or at least - ``ldc[i]`` * ``m[i]`` - if row major layout is used. - See :ref:`matrix-storage` for more details. - - ldc - Array of ``group_count`` integers. ``ldc[i]`` specifies the - leading dimension of ``C`` for every matrix in group ``i``. All - entries must be positive and ``ldc[i]`` must be at least - ``m[i]`` if column major layout is used to store matrices or at - least ``n[i]`` if row major layout is used to store matrices. - - group_count - Specifies the number of groups. Must be at least 0. - - group_size - Array of ``group_count`` integers. ``group_size[i]`` specifies the - number of diagonal matrix-matrix product operations in group ``i``. - All entries must be at least 0. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output overwritten by ``batch_size`` diagonal matrix-matrix product - operations. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event dgmm_batch(sycl::queue &queue, - onemkl::mkl::side left_right, - std::int64_t m, - std::int64_t n, - const T *a, - std::int64_t lda, - std::int64_t stridea, - const T *b, - std::int64_t incx, - std::int64_t stridex, - T *c, - std::int64_t ldc, - std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event dgmm_batch(sycl::queue &queue, - onemkl::mkl::side left_right, - std::int64_t m, - std::int64_t n, - const T *a, - std::int64_t lda, - std::int64_t stridea, - const T *b, - std::int64_t incx, - std::int64_t stridex, - T *c, - std::int64_t ldc, - std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies the position of the diagonal matrix in the product. - See :ref:`onemkl_datatypes` for more details. - - m - Number of rows of ``A``. Must be at least zero. - - n - Number of columns of ``A``. Must be at least zero. - - a - Pointer to input matrices ``A`` with size ``stridea`` * - ``batch_size``. Must be of size at least - ``lda`` * ``k`` + ``stridea`` * (``batch_size`` - 1) - where ``k`` is ``n`` if column major layout is used - or ``m`` if row major layout is used. - - lda - The leading dimension of the matrices ``A``. It must be positive - and at least ``m``. Must be positive and at least ``m`` if column - major layout is used or at least ``n`` if row major layout is used. - - stridea - Stride between different ``A`` matrices. - - x - Pointer to input matrices ``X`` with size ``stridex`` * ``batch_size``. - Must be of size at least - (1 + (``len`` - 1)*abs(``incx``)) + ``stridex`` * (``batch_size`` - 1) - where ``len`` is ``n`` if the diagonal matrix is on the right - of the product or ``m`` otherwise. - - incx - Stride between two consecutive elements of the ``x`` vector. - - stridex - Stride between different ``X`` vectors, must be at least 0. - - c - Pointer to input/output matrices ``C`` with size ``stridec`` * ``batch_size``. - - ldc - The leading dimension of the matrices ``C``. It must be positive and at least - ``ldc`` * ``m`` if column major layout is used to store matrices or at - least ``ldc`` * ``n`` if column major layout is used to store matrices. - - stridec - Stride between different ``C`` matrices. Must be at least - ``ldc`` * ``n`` if column major layout is used or - ``ldc`` * ``m`` if row major layout is used. - - batch_size - Specifies the number of diagonal matrix-matrix product operations to perform. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output overwritten by ``batch_size`` diagonal matrix-matrix product - operations. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-like-extensions` diff --git a/docs/domains/blas/dot.rst b/docs/domains/blas/dot.rst deleted file mode 100644 index 8ae352889..000000000 --- a/docs/domains/blas/dot.rst +++ /dev/null @@ -1,182 +0,0 @@ -.. _onemkl_blas_dot: - -dot -=== - -Computes the dot product of two real vectors. - -.. _onemkl_blas_dot_description: - -.. rubric:: Description - -The ``dot`` routines perform a dot product between two vectors: - -.. math:: - - result = \sum_{i=1}^{n}X_iY_i - -``dot`` supports the following precisions for data. - - .. list-table:: - :header-rows: 1 - - * - T - - T_res - * - ``float`` - - ``float`` - * - ``double`` - - ``double`` - * - ``float`` - - ``double`` - -.. container:: Note - - .. rubric:: Note - :class: NoteTipHead - - For the mixed precision version (inputs are float while result is - double), the dot product is computed with double precision. - -.. _onemkl_blas_dot_buffer: - -dot (Buffer Version) --------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void dot(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &result) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void dot(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &result) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vectors ``x`` and ``y``. - - x - Buffer holding input vector ``x``. The buffer must be of size at least - (1 + (``n`` – 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input vector ``y``. The buffer must be of size at least - (1 + (``n`` – 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - result - Buffer where the result (a scalar) will be stored. - - -.. _onemkl_blas_dot_usm: - -dot (USM Version) ------------------ - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event dot(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T_res *result, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event dot(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T_res *result, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vectors ``x`` and ``y``. - - x - Pointer to the input vector ``x``. The array holding the vector ``x`` - must be of size at least (1 + (``n`` – 1)*abs(``incx``)). See - :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to the input vector ``y``. The array holding the vector ``y`` - must be of size at least (1 + (``n`` – 1)*abs(``incy``)). See - :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - result - Pointer to where the result (a scalar) will be stored. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/dotc.rst b/docs/domains/blas/dotc.rst deleted file mode 100644 index d6746f124..000000000 --- a/docs/domains/blas/dotc.rst +++ /dev/null @@ -1,170 +0,0 @@ -.. _onemkl_blas_dotc: - -dotc -==== - -Computes the dot product of two complex vectors, conjugating the first vector. - -.. _onemkl_blas_dotc_description: - -.. rubric:: Description - -The ``dotc`` routines perform a dot product between two complex -vectors, conjugating the first of them: - -.. math:: - - result = \sum_{i=1}^{n}\overline{X_i}Y_i - -``dotc`` supports the following precisions for data. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_dotc_buffer: - -dotc (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void dotc(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &result) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void dotc(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &result) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - The number of elements in vectors ``x`` and ``y``. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - The stride of vector ``x``. - - y - Buffer holding input vector ``y``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details.. - - incy - The stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - result - The buffer where the result (a scalar) is stored. - - -.. _onemkl_blas_dotc_usm: - -dotc (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void dotc(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *result, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void dotc(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *result, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - The number of elements in vectors ``x`` and ``y``. - - x - Pointer to input vector ``x``. The array holding the input - vector ``x`` must be of size at least (1 + (``n`` - - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - The stride of vector ``x``. - - y - Pointer to input vector ``y``. The array holding the input - vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details.. - - incy - The stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - result - The pointer to where the result (a scalar) is stored. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/dotu.rst b/docs/domains/blas/dotu.rst deleted file mode 100644 index d936815ae..000000000 --- a/docs/domains/blas/dotu.rst +++ /dev/null @@ -1,170 +0,0 @@ -.. _onemkl_blas_dotu: - -dotu -==== - -Computes the dot product of two complex vectors. - -.. _onemkl_blas_dotu_description: - -.. rubric:: Description - -The ``dotu`` routines perform a dot product between two complex vectors: - -.. math:: - - result = \sum_{i=1}^{n}X_iY_i - -``dotu`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_dotu_buffer: - -dotu (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void dotu(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &result) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void dotu(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &result) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vectors ``x`` and ``y``. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input vector ``y``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - -.. container:: section - - .. rubric:: Output Parameters - - result - Buffer where the result (a scalar) is stored. - - -.. _onemkl_blas_dotu_usm: - -dotu (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event dotu(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *result, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event dotu(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *result, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vectors ``x`` and ``y``. - - x - Pointer to the input vector ``x``. The array holding input - vector ``x`` must be of size at least (1 + (``n`` - - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to input vector ``y``. The array holding input vector - ``y`` must be of size at least (1 + (``n`` - 1)*abs(``incy``)). - See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - result - Pointer to where the result (a scalar) is stored. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/gbmv.rst b/docs/domains/blas/gbmv.rst deleted file mode 100644 index 366ce69f2..000000000 --- a/docs/domains/blas/gbmv.rst +++ /dev/null @@ -1,285 +0,0 @@ -.. _onemkl_blas_gbmv: - -gbmv -==== - -Computes a matrix-vector product with a general band matrix. - -.. _onemkl_blas_gbmv_description: - -.. rubric:: Description - -The ``gbmv`` routines compute a scalar-matrix-vector product and add -the result to a scalar-vector product, with a general band matrix. -The operation is defined as - -.. math:: - - y \leftarrow alpha*op(A)*x + beta*y - -where: - -op(``A``) is one of op(``A``) = ``A``, or op(``A``) = -``A``\ :sup:`T`, or op(``A``) = ``A``\ :sup:`H`, - -``alpha`` and ``beta`` are scalars, - -``A`` is an ``m``-by-``n`` matrix with ``kl`` sub-diagonals and -``ku`` super-diagonals, - -``x`` and ``y`` are vectors. - -``gbmv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_gbmv_buffer: - -gbmv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void gbmv(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - std::int64_t kl, - std::int64_t ku, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void gbmv(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - std::int64_t kl, - std::int64_t ku, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - trans - Specifies op(``A``), the transposition operation applied to ``A``. - See - :ref:`onemkl_datatypes` for more - details. - - m - Number of rows of ``A``. Must be at least zero. - - n - Number of columns of ``A``. Must be at least zero. - - kl - Number of sub-diagonals of the matrix ``A``. Must be at least - zero. - - ku - Number of super-diagonals of the matrix ``A``. Must be at least - zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Buffer holding input matrix ``A``. Must have size at least ``lda``\ \*\ ``n`` - if column major layout is used or at least ``lda``\ \*\ ``m`` - if row major layout is used. See :ref:`matrix-storage` for more details. - - lda - Leading dimension of matrix ``A``. Must be at least (``kl`` + - ``ku`` + 1), and positive. - - x - Buffer holding input vector ``x``. The length ``len`` of vector - ``x`` is ``n`` if ``A`` is not transposed, and ``m`` if ``A`` is - transposed. The buffer must be of size at least (1 + (``len`` - - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Buffer holding input/output vector ``y``. The length ``len`` of - vector ``y`` is ``m``, if ``A`` is not transposed, and ``n`` if - ``A`` is transposed. The buffer must be of size at least (1 + - (``len`` - 1)*abs(``incy``)) where ``len`` is this length. See - :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - y - Buffer holding the updated vector ``y``. - - -.. _onemkl_blas_gbmv_usm: - -gbmv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event gbmv(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - std::int64_t kl, - std::int64_t ku, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event gbmv(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - std::int64_t kl, - std::int64_t ku, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - trans - Specifies op(``A``), the transposition operation applied to - ``A``. See - :ref:`onemkl_datatypes` for - more details. - - m - Number of rows of ``A``. Must be at least zero. - - n - Number of columns of ``A``. Must be at least zero. - - kl - Number of sub-diagonals of the matrix ``A``. Must be at least - zero. - - ku - Number of super-diagonals of the matrix ``A``. Must be at least - zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n`` if column - major layout is used or at least ``lda``\ \*\ ``m`` if row - major layout is used. See :ref:`matrix-storage` for more details. - - lda - Leading dimension of matrix ``A``. Must be at least (``kl`` + - ``ku`` + 1), and positive. - - x - Pointer to input vector ``x``. The length ``len`` of vector - ``x`` is ``n`` if ``A`` is not transposed, and ``m`` if ``A`` - is transposed. The array holding input vector ``x`` must be of - size at least (1 + (``len`` - 1)*abs(``incx``)). See - :ref:`matrix-storage` for more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Pointer to input/output vector ``y``. The length ``len`` of - vector ``y`` is ``m``, if ``A`` is not transposed, and ``n`` if - ``A`` is transposed. The array holding input/output vector - ``y`` must be of size at least (1 + (``len`` - - 1)*abs(``incy``)) where ``len`` is this length. - See :ref:`matrix-storage` for more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Pointer to the updated vector ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/gemm.rst b/docs/domains/blas/gemm.rst deleted file mode 100644 index e9e2c89ea..000000000 --- a/docs/domains/blas/gemm.rst +++ /dev/null @@ -1,455 +0,0 @@ -.. _onemkl_blas_gemm: - -gemm -==== - -Computes a matrix-matrix product with general matrices. - -.. _onemkl_blas_gemm_description: - -.. rubric:: Description - -The ``gemm`` routines compute a scalar-matrix-matrix product and add the -result to a scalar-matrix product, with general matrices. The -operation is defined as: - -.. math:: - - C \leftarrow alpha*op(A)*op(B) + beta*C - -where: - -op(``X``) is one of op(``X``) = ``X``, or op(``X``) = ``X``\ :sup:`T`, or -op(``X``) = ``X``\ :sup:`H`, - -``alpha`` and ``beta`` are scalars, - -``A``, ``B`` and ``C`` are matrices, - -``op(A)`` is an ``m``-by-``k`` matrix, - -``op(B)`` is a ``k``-by-``n`` matrix, - -``C`` is an ``m``-by-``n`` matrix. - -``gemm`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - Ts - - Ta - - Tb - - Tc - * - ``float`` - - ``half`` - - ``half`` - - ``float`` - * - ``half`` - - ``half`` - - ``half`` - - ``half`` - * - ``float`` - - ``bfloat16`` - - ``bfloat16`` - - ``float`` - * - ``float`` - - ``float`` - - ``float`` - - ``float`` - * - ``double`` - - ``double`` - - ``double`` - - ``double`` - * - ``std::complex`` - - ``std::complex`` - - ``std::complex`` - - ``std::complex`` - * - ``std::complex`` - - ``std::complex`` - - ``std::complex`` - - ``std::complex`` - -.. _onemkl_blas_gemm_buffer: - -gemm (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void gemm(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t m, - std::int64_t n, - std::int64_t k, - Ts alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - Ts beta, - sycl::buffer &c, - std::int64_t ldc) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void gemm(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t m, - std::int64_t n, - std::int64_t k, - Ts alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - Ts beta, - sycl::buffer &c, - std::int64_t ldc) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - transa - Specifies the form of op(``A``), the transposition operation - applied to ``A``. - - transb - Specifies the form of op(``B``), the transposition operation - applied to ``B``. - - m - Specifies the number of rows of the matrix op(``A``) and of the - matrix ``C``. The value of m must be at least zero. - - n - Specifies the number of columns of the matrix op(``B``) and the - number of columns of the matrix ``C``. The value of n must be at - least zero. - - k - Specifies the number of columns of the matrix op(``A``) and the - number of rows of the matrix op(``B``). The value of k must be at - least zero. - - alpha - Scaling factor for the matrix-matrix product. - - a - The buffer holding the input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``A`` is an ``m``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``m`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``m`` - * - Row major - - ``A`` is an ``m``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``m``. - - ``A`` is an ``k``-by-``m`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k`` - - See :ref:`matrix-storage` for more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda`` must be at least ``m``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``m``. - - b - The buffer holding the input matrix ``B``. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k`` - * - Row major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n`` - - See :ref:`matrix-storage` for more details. - - ldb - The leading dimension of ``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - * - Row major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - - beta - Scaling factor for matrix ``C``. - - c - The buffer holding the input/output matrix ``C``. It must have a - size of at least ``ldc``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldc``\ \*\ ``m`` if row - major layout is used to store matrices . See :ref:`matrix-storage` for more details. - - ldc - The leading dimension of ``C``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - -.. container:: section - - .. rubric:: Output Parameters - - c - The buffer, which is overwritten by - ``alpha``\ \*\ op(``A``)*op(``B``) + ``beta``\ \*\ ``C``. - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized before - calling ``gemm``. - - -.. _onemkl_blas_gemm_usm: - -gemm (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event gemm(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t m, - std::int64_t n, - std::int64_t k, - Ts alpha, - const Ta *a, - std::int64_t lda, - const Tb *b, - std::int64_t ldb, - Ts beta, - Tc *c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event gemm(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t m, - std::int64_t n, - std::int64_t k, - Ts alpha, - const Ta *a, - std::int64_t lda, - const Tb *b, - std::int64_t ldb, - Ts beta, - Tc *c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - transa - Specifies the form of op(``A``), the transposition operation - applied to ``A``. - - - transb - Specifies the form of op(``B``), the transposition operation - applied to ``B``. - - - m - Specifies the number of rows of the matrix op(``A``) and of the - matrix ``C``. The value of m must be at least zero. - - - n - Specifies the number of columns of the matrix op(``B``) and the - number of columns of the matrix ``C``. The value of n must be - at least zero. - - - k - Specifies the number of columns of the matrix op(``A``) and the - number of rows of the matrix op(``B``). The value of k must be - at least zero. - - - alpha - Scaling factor for the matrix-matrix product. - - - a - Pointer to input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``A`` is an ``m``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``m`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``m`` - * - Row major - - ``A`` is an ``m``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``m``. - - ``A`` is an ``k``-by-``m`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k`` - - See :ref:`matrix-storage` for more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda`` must be at least ``m``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``m``. - - b - Pointer to input matrix ``B``. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k`` - * - Row major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n`` - - See :ref:`matrix-storage` for more details. - - ldb - The leading dimension of ``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - * - Row major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - - beta - Scaling factor for matrix ``C``. - - c - The pointer to input/output matrix ``C``. It must have a - size of at least ``ldc``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldc``\ \*\ ``m`` if row - major layout is used to store matrices . See :ref:`matrix-storage` for more details. - - ldc - The leading dimension of ``C``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Pointer to the output matrix, overwritten by - ``alpha``\ \*\ op(``A``)*op(``B``) + ``beta``\ \*\ ``C``. - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized - before calling ``gemm``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-3-routines` diff --git a/docs/domains/blas/gemm_batch.rst b/docs/domains/blas/gemm_batch.rst deleted file mode 100644 index e63129fdd..000000000 --- a/docs/domains/blas/gemm_batch.rst +++ /dev/null @@ -1,606 +0,0 @@ -.. _onemkl_blas_gemm_batch: - -gemm_batch -========== - -Computes a group of ``gemm`` operations. - -.. _onemkl_blas_gemm_batch_description: - -.. rubric:: Description - -The ``gemm_batch`` routines are batched versions of :ref:`onemkl_blas_gemm`, performing -multiple ``gemm`` operations in a single call. Each ``gemm`` -operation perform a matrix-matrix product with general matrices. - -``gemm_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``half`` - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_gemm_batch_buffer: - -gemm_batch (Buffer Version) ---------------------------- - -.. rubric:: Description - -The buffer version of ``gemm_batch`` supports only the strided API. - -The strided API operation is defined as: -:: - - for i = 0 … batch_size – 1 - A, B and C are matrices at offset i * stridea, i * strideb, i * stridec in a, b and c. - C := alpha * op(A) * op(B) + beta * C - end for - -where: - -op(X) is one of op(X) = X, or op(X) = X\ :sup:`T`, or op(X) = X\ :sup:`H`, - -``alpha`` and ``beta`` are scalars, - -``A``, ``B``, and ``C`` are matrices, - -op(``A``) is ``m`` x ``k``, op(``B``) is -``k`` x ``n``, and ``C`` is ``m`` x ``n``. - -The ``a``, ``b`` and ``c`` buffers contain all the input matrices. The stride -between matrices is given by the stride parameter. The total number -of matrices in ``a``, ``b`` and ``c`` buffers is given by the ``batch_size`` parameter. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void gemm_batch(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t m, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - std::int64_t stridea, - sycl::buffer &b, - std::int64_t ldb, - std::int64_t strideb, - T beta, - sycl::buffer &c, - std::int64_t ldc, - std::int64_t stridec, - std::int64_t batch_size) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void gemm_batch(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t m, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - std::int64_t stridea, - sycl::buffer &b, - std::int64_t ldb, - std::int64_t strideb, - T beta, - sycl::buffer &c, - std::int64_t ldc, - std::int64_t stridec, - std::int64_t batch_size) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - transa - Specifies op(``A``) the transposition operation applied to the - matrices ``A``. See :ref:`onemkl_datatypes` for more details. - - transb - Specifies op(``B``) the transposition operation applied to the - matrices ``B``. See :ref:`onemkl_datatypes` for more details. - - m - Number of rows of op(``A``) and ``C``. Must be at least zero. - - - n - Number of columns of op(``B``) and ``C``. Must be at least zero. - - - k - Number of columns of op(``A``) and rows of op(``B``). Must be at - least zero. - - alpha - Scaling factor for the matrix-matrix products. - - a - Buffer holding the input matrices ``A`` with size ``stridea`` * ``batch_size``. - - lda - The leading dimension of the matrices ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda`` must be at least ``m``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``m``. - - stridea - Stride between different ``A`` matrices. - - b - Buffer holding the input matrices ``B`` with size ``strideb`` * ``batch_size``. - - ldb - The leading dimension of the matrices``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - * - Row major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - - strideb - Stride between different ``B`` matrices. - - beta - Scaling factor for the matrices ``C``. - - c - Buffer holding input/output matrices ``C`` with size ``stridec`` * ``batch_size``. - - ldc - The leading dimension of the matrices ``C``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - - stridec - Stride between different ``C`` matrices. Must be at least - ``ldc`` * ``n``. - - batch_size - Specifies the number of matrix multiply operations to perform. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output buffer, overwritten by ``batch_size`` matrix multiply - operations of the form ``alpha`` * op(``A``)*op(``B``) + ``beta`` * ``C``. - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized before - calling ``gemm_batch``. - - -.. _onemkl_blas_gemm_batch_usm: - -gemm_batch (USM Version) ---------------------------- - -.. rubric:: Description - -The USM version of ``gemm_batch`` supports the group API and strided API. - -The group API operation is defined as: -:: - - idx = 0 - for i = 0 … group_count – 1 - for j = 0 … group_size – 1 - A, B, and C are matrices in a[idx], b[idx] and c[idx] - C := alpha[i] * op(A) * op(B) + beta[i] * C - idx = idx + 1 - end for - end for - -The strided API operation is defined as -:: - - for i = 0 … batch_size – 1 - A, B and C are matrices at offset i * stridea, i * strideb, i * stridec in a, b and c. - C := alpha * op(A) * op(B) + beta * C - end for - -where: - -op(X) is one of op(X) = X, or op(X) = X\ :sup:`T`, or op(X) = X\ :sup:`H`, - -``alpha`` and ``beta`` are scalars, - -``A``, ``B``, and ``C`` are matrices, - -op(``A``) is ``m`` x ``k``, op(``B``) is ``k`` x ``n``, and ``C`` is ``m`` x ``n``. - - -For group API, ``a``, ``b`` and ``c`` arrays contain the pointers for all the input matrices. -The total number of matrices in ``a``, ``b`` and ``c`` are given by: - -.. math:: - - total\_batch\_count = \sum_{i=0}^{group\_count-1}group\_size[i] - -For strided API, ``a``, ``b``, ``c`` arrays contain all the input matrices. The total number of matrices -in ``a``, ``b`` and ``c`` are given by the ``batch_size`` parameter. - -**Group API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event gemm_batch(sycl::queue &queue, - onemkl::transpose *transa, - onemkl::transpose *transb, - std::int64_t *m, - std::int64_t *n, - std::int64_t *k, - T *alpha, - const T **a, - std::int64_t *lda, - const T **b, - std::int64_t *ldb, - T *beta, - T **c, - std::int64_t *ldc, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event gemm_batch(sycl::queue &queue, - onemkl::transpose *transa, - onemkl::transpose *transb, - std::int64_t *m, - std::int64_t *n, - std::int64_t *k, - T *alpha, - const T **a, - std::int64_t *lda, - const T **b, - std::int64_t *ldb, - T *beta, - T **c, - std::int64_t *ldc, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - transa - Array of ``group_count`` ``onemkl::transpose`` values. ``transa[i]`` specifies the form of op(``A``) used in - the matrix multiplication in group ``i``. See :ref:`onemkl_datatypes` for more details. - - transb - Array of ``group_count`` ``onemkl::transpose`` values. ``transb[i]`` specifies the form of op(``B``) used in - the matrix multiplication in group ``i``. See :ref:`onemkl_datatypes` for more details. - - m - Array of ``group_count`` integers. ``m[i]`` specifies the - number of rows of op(``A``) and ``C`` for every matrix in group ``i``. All entries must be at least zero. - - n - Array of ``group_count`` integers. ``n[i]`` specifies the - number of columns of op(``B``) and ``C`` for every matrix in group ``i``. All entries must be at least zero. - - k - Array of ``group_count`` integers. ``k[i]`` specifies the - number of columns of op(``A``) and rows of op(``B``) for every matrix in group ``i``. All entries must be at - least zero. - - alpha - Array of ``group_count`` scalar elements. ``alpha[i]`` specifies the scaling factor for every matrix-matrix - product in group ``i``. - - a - Array of pointers to input matrices ``A`` with size ``total_batch_count``. - - See :ref:`matrix-storage` for more details. - - lda - Array of ``group_count`` integers. ``lda[i]`` specifies the - leading dimension of ``A`` for every matrix in group ``i``. All - entries must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda[i]`` must be at least ``m[i]``. - - ``lda[i]`` must be at least ``k[i]``. - * - Row major - - ``lda[i]`` must be at least ``k[i]``. - - ``lda[i]`` must be at least ``m[i]``. - - b - Array of pointers to input matrices ``B`` with size ``total_batch_count``. - - See :ref:`matrix-storage` for more details. - - ldb - Array of ``group_count`` integers. ``ldb[i]`` specifies the - leading dimension of ``B`` for every matrix in group ``i``. All - entries must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``ldb[i]`` must be at least ``k[i]``. - - ``ldb[i]`` must be at least ``n[i]``. - * - Row major - - ``ldb[i]`` must be at least ``n[i]``. - - ``ldb[i]`` must be at least ``k[i]``. - - beta - Array of ``group_count`` scalar elements. ``beta[i]`` specifies the scaling factor for matrix ``C`` - for every matrix in group ``i``. - - c - Array of pointers to input/output matrices ``C`` with size ``total_batch_count``. - - See :ref:`matrix-storage` for more details. - - ldc - Array of ``group_count`` integers. ``ldc[i]`` specifies the - leading dimension of ``C`` for every matrix in group ``i``. All - entries must be positive and ``ldc[i]`` must be at least - ``m[i]`` if column major layout is used to store matrices or at - least ``n[i]`` if row major layout is used to store matrices. - - group_count - Specifies the number of groups. Must be at least 0. - - group_size - Array of ``group_count`` integers. ``group_size[i]`` specifies the - number of matrix multiply products in group ``i``. All entries must be at least 0. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Overwritten by the ``m[i]``-by-``n[i]`` matrix calculated by - (``alpha[i]`` * op(``A``)*op(``B``) + ``beta[i]`` * ``C``) for group ``i``. - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized - before calling ``gemm_batch``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event gemm_batch(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t m, - std::int64_t n, - std::int64_t k, - T alpha, - const T *a, - std::int64_t lda, - std::int64_t stridea, - const T *b, - std::int64_t ldb, - std::int64_t strideb, - T beta, - T *c, - std::int64_t ldc, - std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event gemm_batch(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t m, - std::int64_t n, - std::int64_t k, - T alpha, - const T *a, - std::int64_t lda, - std::int64_t stridea, - const T *b, - std::int64_t ldb, - std::int64_t strideb, - T beta, - T *c, - std::int64_t ldc, - std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - transa - Specifies op(``A``) the transposition operation applied to the - matrices ``A``. See :ref:`onemkl_datatypes` for more details. - - transb - Specifies op(``B``) the transposition operation applied to the - matrices ``B``. See :ref:`onemkl_datatypes` for more details. - - m - Number of rows of op(``A``) and ``C``. Must be at least zero. - - n - Number of columns of op(``B``) and ``C``. Must be at least zero. - - k - Number of columns of op(``A``) and rows of op(``B``). Must be at - least zero. - - alpha - Scaling factor for the matrix-matrix products. - - a - Pointer to input matrices ``A`` with size ``stridea`` * ``batch_size``. - - lda - The leading dimension of the matrices ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda`` must be at least ``m``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``m``. - - stridea - Stride between different ``A`` matrices. - - b - Pointer to input matrices ``B`` with size ``strideb`` * ``batch_size``. - - ldb - The leading dimension of the matrices``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - * - Row major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - - strideb - Stride between different ``B`` matrices. - - beta - Scaling factor for the matrices ``C``. - - c - Pointer to input/output matrices ``C`` with size ``stridec`` * ``batch_size``. - - ldc - The leading dimension of the matrices ``C``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - - stridec - Stride between different ``C`` matrices. - - batch_size - Specifies the number of matrix multiply operations to perform. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output matrices, overwritten by ``batch_size`` matrix multiply - operations of the form ``alpha`` * op(``A``)*op(``B``) + ``beta`` * ``C``. - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized before - calling ``gemm_batch``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-like-extensions` diff --git a/docs/domains/blas/gemm_bias.rst b/docs/domains/blas/gemm_bias.rst deleted file mode 100644 index dd7ce4ecc..000000000 --- a/docs/domains/blas/gemm_bias.rst +++ /dev/null @@ -1,513 +0,0 @@ -.. _onemkl_blas_gemm_bias: - -gemm_bias -========= - -Computes a matrix-matrix product using general integer matrices with bias. - -.. _onemkl_blas_gemm_bias_description: - -.. rubric:: Description - -The gemm_bias routines compute a scalar-matrix-matrix product and -add the result to a scalar-matrix product, using general integer matrices with biases/offsets. -The operation is defined as: - -.. math:: - - \scriptstyle C \leftarrow alpha*(op(A) - A\_offset)*(op(B) - B\_offset) + beta*C + C\_offset - -where: - -op(``X``) is one of op(``X``) = ``X``, or op(``X``) = ``X``\ :sup:`T`, or -op(``X``) = ``X``\ :sup:`H`, - -``alpha`` and ``beta`` are scalars, - -``A_offset`` is an ``m``-by-``k`` matrix with every element equal to the value ao, - -``B_offset`` is a ``k``-by-``n`` matrix with every element equal to the value bo, - -``C_offset`` is an ``m``-by-``n`` matrix defined by the -co buffer as described below, - -``A``, ``B``, and ``C`` are matrices, - -op(``A``) is ``m`` x ``k``, op(``B``) is ``k`` x ``n``, and -``C`` is ``m`` x ``n``. - -``gemm_bias`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - Ts - - Ta - - Tb - - Tc - * - ``float`` - - ``std::uint8_t`` - - ``std::uint8_t`` - - ``std::int32_t`` - * - ``float`` - - ``std::int8_t`` - - ``std::uint8_t`` - - ``std::int32_t`` - * - ``float`` - - ``std::uint8_t`` - - ``std::int8_t`` - - ``std::int32_t`` - * - ``float`` - - ``std::int8_t`` - - ``std::int8_t`` - - ``std::int32_t`` - -.. _onemkl_blas_gemm_bias_buffer: - -gemm_bias (Buffer Version) --------------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void gemm_bias(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - onemkl::offset offset_type, - std::int64_t m, - std::int64_t n, - std::int64_t k, - Ts alpha, - sycl::buffer &a, - std::int64_t lda, - Ta ao, - sycl::buffer &b, - std::int64_t ldb, - Tb bo, - Ts beta, - sycl::buffer &c, - std::int64_t ldc, - sycl::buffer &co) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void gemm_bias(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - onemkl::offset offset_type, - std::int64_t m, - std::int64_t n, - std::int64_t k, - Ts alpha, - sycl::buffer &a, - std::int64_t lda, - Ta ao, - sycl::buffer &b, - std::int64_t ldb, - Tb bo, - Ts beta, - sycl::buffer &c, - std::int64_t ldc, - sycl::buffer &co) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - transa - Specifies op(``A``), the transposition operation applied to - ``A``. See - :ref:`onemkl_datatypes` for - more details. - - transb - Specifies op(``B``), the transposition operation applied to - ``B``. See - :ref:`onemkl_datatypes` for - more details. - - offset_type - Specifies the form of ``C_offset`` used in the matrix - multiplication. See - :ref:`onemkl_datatypes` for - more details. - - m - Number of rows of op(``A``) and ``C``. Must be at least zero. - - n - Number of columns of op(``B``) and ``C``. Must be at least - zero. - - k - Number of columns of op(``A``) and rows of op(``B``). Must be - at least zero. - - alpha - Scaling factor for the matrix-matrix product. - - a - The buffer holding the input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``A`` is an ``m``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``m`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``m`` - * - Row major - - ``A`` is an ``m``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``m``. - - ``A`` is an ``k``-by-``m`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k`` - - See :ref:`matrix-storage` for more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda`` must be at least ``m``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``m``. - - ao - Specifies the scalar offset value for matrix ``A``. - - b - Buffer holding the input matrix ``B``. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k`` - * - Row major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n`` - - See :ref:`matrix-storage` for more details. - - ldb - The leading dimension of ``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - * - Row major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - - bo - Specifies the scalar offset value for matrix ``B``. - - beta - Scaling factor for matrix ``C``. - - c - Buffer holding the input/output matrix ``C``. It must have a - size of at least ``ldc``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldc``\ \*\ ``m`` if row - major layout is used to store matrices . - See :ref:`matrix-storage` for more details. - - ldc - The leading dimension of ``C``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - - co - Buffer holding the offset values for matrix ``C``. - - If ``offset_type`` = ``offset::fix``, the ``co`` array must have - size at least 1. - - - If ``offset_type`` = ``offset::col``, the ``co`` array must have - size at least ``max(1,m)``. - - - If ``offset_type`` = ``offset::row``, the ``co`` array must have - size at least ``max(1,n)``. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output buffer, overwritten by ``alpha`` * (op(``A``) - - ``A_offset``)*(op(``B``) - ``B_offset``) + ``beta`` * ``C`` + ``C_offset``. - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized - before calling ``gemm_bias``. - - -.. _onemkl_blas_gemm_bias_usm: - -gemm_bias (USM Version) ------------------------ - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event gemm_bias(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - onemkl::offset offset_type, - std::int64_t m, - std::int64_t n, - std::int64_t k, - Ts alpha, - const Ta *a, - std::int64_t lda, - Ta ao, - const Tb *b, - std::int64_t ldb, - Tb bo, - Ts beta, - Tc *c, - std::int64_t ldc, - const Tc *co, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event gemm_bias(sycl::queue &queue, - onemkl::transpose transa, - onemkl::transpose transb, - onemkl::offset offset_type, - std::int64_t m, - std::int64_t n, - std::int64_t k, - Ts alpha, - const Ta *a, - std::int64_t lda, - Ta ao, - const Tb *b, - std::int64_t ldb, - Tb bo, - Ts beta, - Tc *c, - std::int64_t ldc, - const Tc *co, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - transa - Specifies op(``A``), the transposition operation applied to - ``A``. See - :ref:`onemkl_datatypes` for - more details. - - transb - Specifies op(``B``), the transposition operation applied to - ``B``. See - :ref:`onemkl_datatypes` for - more details. - - offset_type - Specifies the form of ``C_offset`` used in the matrix - multiplication. See - :ref:`onemkl_datatypes` for - more details. - - m - Number of rows of op(``A``) and ``C``. Must be at least zero. - - n - Number of columns of op(``B``) and ``C``. Must be at least - zero. - - k - Number of columns of op(``A``) and rows of op(``B``). Must be - at least zero. - - alpha - Scaling factor for the matrix-matrix product. - - a - Pointer to input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``A`` is an ``m``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``m`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``m`` - * - Row major - - ``A`` is an ``m``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``m``. - - ``A`` is an ``k``-by-``m`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k`` - - See :ref:`matrix-storage` for more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda`` must be at least ``m``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``m``. - - ao - Specifies the scalar offset value for matrix ``A``. - - b - Pointer to input matrix ``B``. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k`` - * - Row major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n`` - - See :ref:`matrix-storage` for more details. - - ldb - The leading dimension of ``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - * - Row major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - - bo - Specifies the scalar offset value for matrix ``B``. - - beta - Scaling factor for matrix ``C``. - - c - Pointer to input/output matrix ``C``. It must have a - size of at least ``ldc``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldc``\ \*\ ``m`` if row - major layout is used to store matrices . See :ref:`matrix-storage` for more details. - - ldc - The leading dimension of ``C``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - - co - Pointer to offset values for matrix ``C``. - - - If ``offset_type`` = ``offset::fix``, the ``co`` array must have - size at least 1. - - - If ``offset_type`` = ``offset::col``, the ``co`` array must have - size at least ``max(1,m)``. - - - If ``offset_type`` = ``offset::row``, the ``co`` array must have - size at least ``max(1,n)``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Pointer to the output matrix, overwritten by ``alpha`` * (op(``A``) - - ``A_offset``)*(op(``B``) - ``B_offset``) + ``beta`` * ``C`` + ``C_offset``. - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized - before calling ``gemm_bias``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-like-extensions` diff --git a/docs/domains/blas/gemmt.rst b/docs/domains/blas/gemmt.rst deleted file mode 100644 index 7f224b2f7..000000000 --- a/docs/domains/blas/gemmt.rst +++ /dev/null @@ -1,418 +0,0 @@ -.. _onemkl_blas_gemmt: - -gemmt -===== - -Computes a matrix-matrix product with general matrices, but updates -only the upper or lower triangular part of the result matrix. - -.. _onemkl_blas_gemmt_description: - -.. rubric:: Description - -The gemmt routines compute a scalar-matrix-matrix product and add -the result to the upper or lower part of a scalar-matrix product, -with general matrices. The operation is defined as: - -.. math:: - - C \leftarrow alpha*op(A)*op(B) + beta*C - -where: - -op(``X``) is one of op(``X``) = ``X``, or op(``X``) = ``X``\ :sup:`T`, or -op(``X``) = ``X``\ :sup:`H`, - -``alpha`` and ``beta`` are scalars - -``A``, ``B``, and ``C`` are matrices - -op(``A``) is ``n`` x ``k``, op(``B``) is ``k`` x ``n``, and -``C`` is ``n`` x ``n``. - -``gemmt`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_gemmt_buffer: - -gemmt (Buffer Version) ----------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void gemmt(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - T beta, - sycl::buffer &c, - std::int64_t ldc) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void gemmt(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - T beta, - sycl::buffer &c, - std::int64_t ldc) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``C``\ ’s data is stored in its upper or - lower triangle. See :ref:`onemkl_datatypes` for more details. - - transa - Specifies op(``A``), the transposition operation applied to - ``A``. See :ref:`onemkl_datatypes` for more details. - - transb - Specifies op(``B``), the transposition operation applied to - ``B``. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows of op(``A``), columns of op(``B``), and - columns and rows of\ ``C``. Must be at least zero. - - k - Number of columns of op(``A``) and rows of op(``B``). Must be - at least zero. - - alpha - Scaling factor for the matrix-matrix product. - - a - Buffer holding the input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n`` - * - Row major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - See :ref:`matrix-storage` for more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - b - Buffer holding the input matrix ``B``. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k`` - * - Row major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - See :ref:`matrix-storage` for more details. - - ldb - The leading dimension of ``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - * - Row major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - - beta - Scaling factor for matrix ``C``. - - c - Buffer holding the input/output matrix ``C``. Must have size at - least ``ldc`` \* ``n``. See :ref:`matrix-storage` for - more details. - - ldc - Leading dimension of ``C``. Must be positive and at least - ``m``. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output buffer, overwritten by the upper or lower triangular - part of ``alpha`` * op(``A``)*op(``B``) + ``beta`` * ``C``. - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized - before calling gemmt. - - -.. _onemkl_blas_gemmt_usm: - -gemmt (USM Version) -------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event gemmt(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t n, - std::int64_t k, - T alpha, - const T* a, - std::int64_t lda, - const T* b, - std::int64_t ldb, - T beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event gemmt(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::transpose transb, - std::int64_t n, - std::int64_t k, - T alpha, - const T* a, - std::int64_t lda, - const T* b, - std::int64_t ldb, - T beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``C``\ ’s data is stored in its upper or - lower triangle. See - :ref:`onemkl_datatypes` for - more details. - - transa - Specifies op(``A``), the transposition operation applied to - ``A``. See - :ref:`onemkl_datatypes` for - more details. - - transb - Specifies op(``B``), the transposition operation applied to - ``B``. See - :ref:`onemkl_datatypes` for - more details. - - n - Number of columns of op(``A``), columns of op(``B``), and - columns of\ ``C``. Must be at least zero. - - k - Number of columns of op(``A``) and rows of op(``B``). Must be - at least zero. - - alpha - Scaling factor for the matrix-matrix product. - - a - Pointer to input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n`` - * - Row major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k`` - - See :ref:`matrix-storage` for more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - b - Pointer to input matrix ``B``. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k`` - * - Row major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n`` - - See :ref:`matrix-storage` for more details. - - ldb - The leading dimension of ``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``B`` not transposed - - ``B`` transposed - * - Column major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - * - Row major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - - beta - Scaling factor for matrix ``C``. - - c - Pointer to input/output matrix ``C``. Must have size at least - ``ldc`` \* ``n``. See :ref:`matrix-storage` for - more details. - - ldc - Leading dimension of ``C``. Must be positive and at least - ``m``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Pointer to the output matrix, overwritten by the upper or lower - triangular part of ``alpha`` * op(``A``)*op(``B``) + ``beta`` * ``C``. - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized - before calling gemmt. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-like-extensions` diff --git a/docs/domains/blas/gemv.rst b/docs/domains/blas/gemv.rst deleted file mode 100644 index 9577fc111..000000000 --- a/docs/domains/blas/gemv.rst +++ /dev/null @@ -1,261 +0,0 @@ -.. _onemkl_blas_gemv: - -gemv -==== - -Computes a matrix-vector product using a general matrix. - -.. _onemkl_blas_gemv_description: - -.. rubric:: Description - -The ``gemv`` routines compute a scalar-matrix-vector product and add the -result to a scalar-vector product, with a general matrix. The -operation is defined as: - -.. math:: - - y \leftarrow alpha*op(A)*x + beta*y - -where: - -op(``A``) is one of op(``A``) = ``A``, or op(``A``) = -``A``\ :sup:`T`, or op(``A``) = ``A``\ :sup:`H`, - -``alpha`` and ``beta`` are scalars, - -``A`` is an ``m``-by-``n`` matrix, and ``x``, ``y`` are vectors. - -``gemv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_gemv_buffer: - -gemv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void gemv(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void gemv(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - trans - Specifies ``op(A)``, the transposition operation applied to ``A``. - - m - Specifies the number of rows of the matrix ``A``. The value of - ``m`` must be at least zero. - - n - Specifies the number of columns of the matrix ``A``. The value of - ``n`` must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - a - The buffer holding the input matrix ``A``. Must have a size of at - least ``lda``\ \*\ ``n`` if column major layout is used or at - least ``lda``\ \*\ ``m`` if row major layout is used. See - :ref:`matrix-storage` for more details. - - lda - Leading dimension of matrix ``A``. Must be positive and at least - ``m`` if column major layout is used or at least ``n`` if row - major layout is used. - - x - Buffer holding input vector ``x``. The length ``len`` of vector - ``x`` is ``n`` if ``A`` is not transposed, and ``m`` if ``A`` is - transposed. The buffer must be of size at least (1 + (``len`` - - 1)*abs(``incx``)). See :ref:`matrix-storage` for more details. - - incx - The stride of vector ``x``. - - beta - The scaling factor for vector ``y``. - - y - Buffer holding input/output vector ``y``. The length ``len`` of - vector ``y`` is ``m``, if ``A`` is not transposed, and ``n`` if - ``A`` is transposed. The buffer must be of size at least (1 + - (``len`` - 1)*abs(``incy``)) where ``len`` is this length. See - :ref:`matrix-storage` for more details. - - incy - The stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - y - The buffer holding updated vector ``y``. - - -.. _onemkl_blas_gemv_usm: - -gemv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event gemv(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event gemv(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - trans - Specifies ``op(A)``, the transposition operation applied to - ``A``. See - :ref:`onemkl_datatypes` for - more details. - - m - Specifies the number of rows of the matrix ``A``. The value of - ``m`` must be at least zero. - - n - Specifies the number of columns of the matrix ``A``. The value - of ``n`` must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Pointer to the input matrix ``A``. Must have a size of at - least ``lda``\ \*\ ``n`` if column major layout is used or at - least ``lda``\ \*\ ``m`` if row major layout is used. See - :ref:`matrix-storage` for more details. - - lda - Leading dimension of matrix ``A``. Must be positive and at least - ``m`` if column major layout is used or at least ``n`` if row - major layout is used. - - x - Pointer to the input vector ``x``. The length ``len`` of vector - ``x`` is ``n`` if ``A`` is not transposed, and ``m`` if ``A`` - is transposed. The array holding vector ``x`` must be of size - at least (1 + (``len`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - The stride of vector ``x``. - - beta - The scaling factor for vector ``y``. - - y - Pointer to input/output vector ``y``. The length ``len`` of - vector ``y`` is ``m``, if ``A`` is not transposed, and ``n`` if - ``A`` is transposed. The array holding input/output vector - ``y`` must be of size at least (1 + (``len`` - - 1)*abs(``incy``)) where ``len`` is this length. See :ref:`matrix-storage` for - more details. - - incy - The stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - The pointer to updated vector ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/gemv_batch.rst b/docs/domains/blas/gemv_batch.rst deleted file mode 100644 index 0a56b73b6..000000000 --- a/docs/domains/blas/gemv_batch.rst +++ /dev/null @@ -1,472 +0,0 @@ -.. _onemkl_blas_gemv_batch: - -gemv_batch -========== - -Computes a group of ``gemv`` operations. - -.. _onemkl_blas_gemv_batch_description: - -.. rubric:: Description - -The ``gemv_batch`` routines are batched versions of -:ref:`onemkl_blas_gemv`, performing multiple ``gemv`` operations in a -single call. Each ``gemv`` operations perform a scalar-matrix-vector -product and add the result to a scalar-vector product. - -``gemv_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_gemv_batch_buffer: - -gemv_batch (Buffer Version) ---------------------------- - -.. rubric:: Description - -The buffer version of ``gemv_batch`` supports only the strided API. - -The strided API operation is defined as: -:: - - for i = 0 … batch_size – 1 - A is a matrix at offset i * stridea in a. - X and Y are matrices at offset i * stridex, i * stridey, in x and y. - Y := alpha * op(A) * X + beta * Y - end for - -where: - -op(A) is one of op(A) = A, or op(A) = A\ :sup:`T`, or op(A) = A\ :sup:`H`, - -``alpha`` and ``beta`` are scalars, - -``A`` is a matrix and ``X`` and ``Y`` are vectors, - -The ``x`` and ``y`` buffers contain all the input matrices. The stride -between vectors is given by the stride parameter. The total number of -vectors in ``x`` and ``y`` buffers is given by the ``batch_size`` -parameter. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void gemv_batch(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - std::int64_t stridea, - sycl::buffer &x, - std::int64_t incx, - std::int64_t stridex, - T beta, - sycl::buffer &y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void gemv_batch(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - std::int64_t stridea, - sycl::buffer &x, - std::int64_t incx, - std::int64_t stridex, - T beta, - sycl::buffer &y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - trans - Specifies op(``A``) the transposition operation applied to the - matrices ``A``. See :ref:`onemkl_datatypes` for more details. - - m - Number of rows of op(``A``). Must be at least zero. - - n - Number of columns of op(``A``). Must be at least zero. - - alpha - Scaling factor for the matrix-vector products. - - a - Buffer holding the input matrices ``A`` with size ``stridea`` * ``batch_size``. - - lda - The leading dimension of the matrices ``A``. It must be positive - and at least ``m`` if column major layout is used or at least - ``n`` if row major layout is used. - - stridea - Stride between different ``A`` matrices. - - x - Buffer holding the input vectors ``X`` with size ``stridex`` * ``batch_size``. - - incx - The stride of the vector ``X``. It must be positive. - - stridex - Stride between different consecutive ``X`` vectors, must be at least 0. - - beta - Scaling factor for the vector ``Y``. - - y - Buffer holding input/output vectors ``Y`` with size ``stridey`` * ``batch_size``. - - incy - Stride between two consecutive elements of the ``y`` vectors. - - stridey - Stride between two consecutive ``Y`` vectors. Must be at least - (1 + (len-1)*abs(incy)) where ``len`` is ``m`` if the matrix ``A`` - is non transpose or ``n`` otherwise. - - batch_size - Specifies the number of matrix-vector operations to perform. - -.. container:: section - - .. rubric:: Output Parameters - - y - Output overwritten by ``batch_size`` matrix-vector product - operations of the form ``alpha`` * op(``A``) * ``X`` + ``beta`` * ``Y``. - - -.. _onemkl_blas_gemv_batch_usm: - -gemv_batch (USM Version) ---------------------------- - -.. rubric:: Description - -The USM version of ``gemv_batch`` supports the group API and strided API. - -The group API operation is defined as: -:: - - idx = 0 - for i = 0 … group_count – 1 - for j = 0 … group_size – 1 - A is an m x n matrix in a[idx] - X and Y are vectors in x[idx] and y[idx] - Y := alpha[i] * op(A) * X + beta[i] * Y - idx = idx + 1 - end for - end for - -The strided API operation is defined as -:: - - for i = 0 … batch_size – 1 - A is a matrix at offset i * stridea in a. - X and Y are vectors at offset i * stridex, i * stridey in x and y. - Y := alpha * op(A) * X + beta * Y - end for - -where: - -op(A) is one of op(A) = A, or op(A) = A\ :sup:`T`, or op(A) = A\ :sup:`H`, - -``alpha`` and ``beta`` are scalars, - -``A`` is a matrix and ``X`` and ``Y`` are vectors, - -For group API, ``x`` and ``y`` arrays contain the pointers for all the input vectors. -``A`` array contains the pointers to all input matrices. -The total number of vectors in ``x`` and ``y`` and matrices in ``A`` are given by: - -.. math:: - - total\_batch\_count = \sum_{i=0}^{group\_count-1}group\_size[i] - -For strided API, ``x`` and ``y`` arrays contain all the input -vectors. ``A`` array contains the pointers to all input matrices. The -total number of vectors in ``x`` and ``y`` and matrices in ``A`` are given by the -``batch_size`` parameter. - -**Group API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event gemv_batch(sycl::queue &queue, - onemkl::transpose *trans, - std::int64_t *m, - std::int64_t *n, - T *alpha, - const T **a, - std::int64_t *lda, - const T **x, - std::int64_t *incx, - T *beta, - T **y, - std::int64_t *incy, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event gemv_batch(sycl::queue &queue, - onemkl::transpose *trans, - std::int64_t *m, - std::int64_t *n, - T *alpha, - const T **a, - std::int64_t *lda, - const T **x, - std::int64_t *incx, - T *beta, - T **y, - std::int64_t *incy, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - trans - Array of ``group_count`` ``onemkl::transpose`` values. ``trans[i]`` specifies the form of op(``A``) used in - the matrix-vector product in group ``i``. See :ref:`onemkl_datatypes` for more details. - - m - Array of ``group_count`` integers. ``m[i]`` specifies the - number of rows of op(``A``) for every matrix in group ``i``. All entries must be at least zero. - - n - Array of ``group_count`` integers. ``n[i]`` specifies the - number of columns of op(``A``) for every matrix in group ``i``. All entries must be at least zero. - - alpha - Array of ``group_count`` scalar elements. ``alpha[i]`` specifies - the scaling factor for every matrix-vector product in group - ``i``. - - a - Array of pointers to input matrices ``A`` with size ``total_batch_count``. - - See :ref:`matrix-storage` for more details. - - lda - Array of ``group_count`` integers. ``lda[i]`` specifies the - leading dimension of ``A`` for every matrix in group ``i``. All - entries must be positive and at least ``m`` if column major - layout is used or at least ``n`` if row major layout is used. - - x - Array of pointers to input vectors ``X`` with size ``total_batch_count``. - - See :ref:`matrix-storage` for more details. - - incx - Array of ``group_count`` integers. ``incx[i]`` specifies the - stride of ``X`` for every vector in group ``i``. All - entries must be positive. - - beta - Array of ``group_count`` scalar elements. ``beta[i]`` specifies - the scaling factor for vector ``Y`` for every vector in group - ``i``. - - y - Array of pointers to input/output vectors ``Y`` with size ``total_batch_count``. - - See :ref:`matrix-storage` for more details. - - incy - Array of ``group_count`` integers. ``incy[i]`` specifies the - leading dimension of ``Y`` for every vector in group ``i``. All - entries must be positive and ``incy[i]`` must be at least - ``m[i]`` if column major layout is used or at - least ``n[i]`` if row major layout is used. - - group_count - Specifies the number of groups. Must be at least 0. - - group_size - Array of ``group_count`` integers. ``group_size[i]`` specifies the - number of matrix-vector products in group ``i``. All entries must be at least 0. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Overwritten by vector calculated by - (``alpha[i]`` * op(``A``) * ``X`` + ``beta[i]`` * ``Y``) for group ``i``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event gemv_batch(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - T alpha, - const T *a, - std::int64_t lda, - std::int64_t stridea, - const T *x, - std::int64_t incx, - std::int64_t stridex, - T beta, - T *y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event gemv_batch(sycl::queue &queue, - onemkl::transpose trans, - std::int64_t m, - std::int64_t n, - T alpha, - const T *a, - std::int64_t lda, - std::int64_t stridea, - const T *x, - std::int64_t incx, - std::int64_t stridex, - T beta, - T *y, - std::int64_t incy, - std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } - - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - trans - Specifies op(``A``) the transposition operation applied to the - matrices ``A``. See :ref:`onemkl_datatypes` for more details. - - m - Number of rows of op(``A``). Must be at least zero. - - n - Number of columns of op(``A``). Must be at least zero. - - alpha - Scaling factor for the matrix-vector products. - - a - Pointer to the input matrices ``A`` with size ``stridea`` * ``batch_size``. - - lda - The leading dimension of the matrices ``A``. It must be positive - and at least ``m`` if column major layout is used or at least - ``n`` if row major layout is used. - - stridea - Stride between different ``A`` matrices. - - x - Pointer to the input vectors ``X`` with size ``stridex`` * ``batch_size``. - - incx - Stride of the vector ``X``. It must be positive. - - stridex - Stride between different consecutive ``X`` vectors, must be at least 0. - - beta - Scaling factor for the vector ``Y``. - - y - Pointer to the input/output vectors ``Y`` with size ``stridey`` * ``batch_size``. - - incy - Stride between two consecutive elements of the ``y`` vectors. - - stridey - Stride between two consecutive ``Y`` vectors. Must be at least - (1 + (len-1)*abs(incy)) where ``len`` is ``m`` if the matrix ``A`` - is non transpose or ``n`` otherwise. - - batch_size - Specifies the number of matrix-vector operations to perform. - -.. container:: section - - .. rubric:: Output Parameters - - y - Output overwritten by ``batch_size`` matrix-vector product - operations of the form ``alpha`` * op(``A``) * ``X`` + ``beta`` * ``Y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-like-extensions` diff --git a/docs/domains/blas/ger.rst b/docs/domains/blas/ger.rst deleted file mode 100644 index ea128414d..000000000 --- a/docs/domains/blas/ger.rst +++ /dev/null @@ -1,226 +0,0 @@ -.. _onemkl_blas_ger: - -ger -=== - -Computes a rank-1 update of a general matrix. - -.. _onemkl_blas_ger_description: - -.. rubric:: Description - -The ``ger`` routines compute a scalar-vector-vector product and add the -result to a general matrix. The operation is defined as: - -.. math:: - - A \leftarrow alpha*x*y^T + A - -where: - -``alpha`` is scalar, - -``A`` is an ``m``-by-``n`` matrix, - -``x`` is a vector of length ``m``, - -``y`` is a vector of length ``n``. - -``ger`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. _onemkl_blas_ger_buffer: - -ger (Buffer Version) --------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void ger(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a, - std::int64_t lda) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void ger(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a, - std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - m - Number of rows of ``A``. Must be at least zero. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``m`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n`` if column major layout is used or at least ``lda``\ \*\ ``m`` - if row major layout is used. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be positive and at least - ``m`` if column major layout is used or at least ``n`` if row - major layout is used. - -.. container:: section - - .. rubric:: Output Parameters - - a - Buffer holding the updated matrix ``A``. - - -.. _onemkl_blas_ger_usm: - -ger (USM Version) ------------------ - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event ger(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event ger(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - m - Number of rows of ``A``. Must be at least zero. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``m`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - a - Pointer to input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n`` if column major layout is used or at least ``lda``\ \*\ ``m`` - if row major layout is used. See :ref:`matrix-storage` for more details. - - lda - Leading dimension of matrix ``A``. Must be positive and at least - ``m`` if column major layout is used or at least ``n`` if row - major layout is used. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the updated matrix ``A``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/gerc.rst b/docs/domains/blas/gerc.rst deleted file mode 100644 index 8a8c28463..000000000 --- a/docs/domains/blas/gerc.rst +++ /dev/null @@ -1,227 +0,0 @@ -.. _onemkl_blas_gerc: - -gerc -==== - -Computes a rank-1 update (conjugated) of a general complex matrix. - -.. _onemkl_blas_gerc_description: - -.. rubric:: Description - -The ``gerc`` routines compute a scalar-vector-vector product and add the -result to a general matrix. The operation is defined as: - -.. math:: - - A \leftarrow alpha*x*y^H + A - - -where: - -``alpha`` is a scalar, - -``A`` is an ``m``-by-``n`` matrix, - -``x`` is a vector of length ``m``, - -``y`` is vector of length ``n``. - -``gerc`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_gerc_buffer: - -gerc (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void gerc(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a, - std::int64_t lda) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void gerc(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a, - std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - m - Number of rows of ``A``. Must be at least zero. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``m`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n`` if column major layout is used or at least ``lda``\ \*\ ``m`` - if row major layout is used. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be positive and at least - ``m`` if column major layout is used or at least ``n`` if row - major layout is used. - - -.. container:: section - - .. rubric:: Output Parameters - - a - Buffer holding the updated matrix ``A``. - - -.. _onemkl_blas_gerc_usm: - -gerc (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event gerc(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event gerc(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - m - Number of rows of ``A``. Must be at least zero. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Pointer to the input vector ``x``. The array holding input - vector ``x`` must be of size at least (1 + (``m`` - - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to the input/output vector ``y``. The array holding the - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A``\ must have size at least ``lda``\ \*\ ``n`` if column - major layout is used or at least ``lda``\ \*\ ``m`` if row - major layout is used. See :ref:`matrix-storage` for more details. - - lda - Leading dimension of matrix ``A``. Must be positive and at least - ``m`` if column major layout is used or at least ``n`` if row - major layout is used. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the updated matrix ``A``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/geru.rst b/docs/domains/blas/geru.rst deleted file mode 100644 index 8e22b3ba9..000000000 --- a/docs/domains/blas/geru.rst +++ /dev/null @@ -1,227 +0,0 @@ -.. _onemkl_blas_geru: - -geru -==== - -Computes a rank-1 update (unconjugated) of a general complex matrix. - -.. _onemkl_blas_geru_description: - -.. rubric:: Description - -The ``geru`` routines routines compute a scalar-vector-vector product and -add the result to a general matrix. The operation is defined as - -.. math:: - - A \leftarrow alpha*x*y^T + A - -where: - -``alpha`` is a scalar, - -``A`` is an ``m``-by-``n`` matrix, - -``x`` is a vector of length ``m``, - -``y`` is a vector of length ``n``. - -``geru`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_geru_buffer: - -geru (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void geru(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a, - std::int64_t lda) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void geru(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a, - std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - m - Number of rows of ``A``. Must be at least zero. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``m`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n`` if column major layout is used or at least ``lda``\ \*\ ``m`` - if row major layout is used. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be positive and at least - ``m`` if column major layout is used or at least ``n`` if row - major layout is used. - -.. container:: section - - .. rubric:: Output Parameters - - a - Buffer holding the updated matrix ``A``. - - -.. _onemkl_blas_geru_usm: - -geru (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event geru(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event geru(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - m - Number of rows of ``A``. Must be at least zero. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Pointer to the input vector ``x``. The array holding input - vector ``x`` must be of size at least (1 + (``m`` - - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n`` if column - major layout is used or at least ``lda``\ \*\ ``m`` if row - major layout is used. See :ref:`matrix-storage` for more details. - - lda - Leading dimension of matrix ``A``. Must be positive and at - least ``m`` if column major layout is used or at least ``n`` - if row major layout is used. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the updated matrix ``A``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/hbmv.rst b/docs/domains/blas/hbmv.rst deleted file mode 100644 index a86e7fef8..000000000 --- a/docs/domains/blas/hbmv.rst +++ /dev/null @@ -1,245 +0,0 @@ -.. _onemkl_blas_hbmv: - -hbmv -==== - -Computes a matrix-vector product using a Hermitian band matrix. - -.. _onemkl_blas_hbmv_description: - -.. rubric:: Description - -The ``hbmv`` routines compute a scalar-matrix-vector product and add the -result to a scalar-vector product, with a Hermitian band matrix. The -operation is defined as - -.. math:: - - y \leftarrow alpha*A*x + beta*y - -where: - -``alpha`` and ``beta`` are scalars, - -``A`` is an ``n``-by-``n`` Hermitian band matrix, with ``k`` -super-diagonals, - -``x`` and ``y`` are vectors of length ``n``. - -``hbmv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_hbmv_buffer: - -hbmv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void hbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void hbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - k - Number of super-diagonals of the matrix ``A``. Must be at least - zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least (``k`` + 1), - and positive. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``m`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - y - Buffer holding the updated vector ``y``. - - -.. _onemkl_blas_hbmv_usm: - -hbmv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event hbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - std::int64_t k, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event hbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - std::int64_t k, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - k - Number of super-diagonals of the matrix ``A``. Must be at least - zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Pointer to the input matrix ``A``. The array holding input - matrix ``A`` must have size at least ``lda``\ \*\ ``n``. See - :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least (``k`` + - 1), and positive. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``m`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Pointer to the updated vector ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/hemm.rst b/docs/domains/blas/hemm.rst deleted file mode 100644 index 4098cd75b..000000000 --- a/docs/domains/blas/hemm.rst +++ /dev/null @@ -1,315 +0,0 @@ -.. _onemkl_blas_hemm: - -hemm -==== - -Computes a matrix-matrix product where one input matrix is Hermitian -and one is general. - -.. _onemkl_blas_hemm_description: - -.. rubric:: Description - -The ``hemm`` routines compute a scalar-matrix-matrix product and add the -result to a scalar-matrix product, where one of the matrices in the -multiplication is Hermitian. The argument ``left_right`` determines -if the Hermitian matrix, ``A``, is on the left of the multiplication -(``left_right`` = ``side::left``) or on the right (``left_right`` = -``side::right``). Depending on ``left_right``, the operation is -defined as: - -.. math:: - - C \leftarrow alpha*A*B + beta*C - -or - -.. math:: - - C \leftarrow alpha*B*A + beta*C - -where: - -``alpha`` and ``beta`` are scalars, - -``A`` is a Hermitian matrix, either ``m``-by-``m`` or ``n``-by-``n`` -matrices, - -``B`` and ``C`` are ``m``-by-``n`` matrices. - -``hemm`` supports the following precisions: - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_hemm_buffer: - -hemm (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void hemm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - T beta, - sycl::buffer &c, - std::int64_t ldc) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void hemm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - T beta, - sycl::buffer &c, - std::int64_t ldc) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies whether ``A`` is on the left side of the multiplication - (``side::left``) or on the right side (``side::right``). See :ref:`onemkl_datatypes` for more details. - - uplo - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - m - Specifies the number of rows of the matrix ``B`` and ``C``. - - The value of ``m`` must be at least zero. - - n - Specifies the number of columns of the matrix ``B`` and ``C``. - - The value of ``n`` must be at least zero. - - alpha - Scaling factor for the matrix-matrix product. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``m`` if ``A`` is on the left of the multiplication, - or ``lda``\ \*\ ``n`` if ``A`` is on the right. See :ref:`matrix-storage` - for more details. - - lda - Leading dimension of ``A``. Must be at least ``m`` if ``A`` is on - the left of the multiplication, or at least ``n`` if ``A`` is on - the right. Must be positive. - - b - Buffer holding input matrix ``B``. Must have size at least - ``ldb``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldb``\ \*\ ``m`` if row - major layout is used to store matrices. See :ref:`matrix-storage` for - more details. - - ldb - Leading dimension of ``B``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if column major layout is used to store matrices. - - beta - Scaling factor for matrix ``C``. - - c - The buffer holding the input/output matrix ``C``. It must have a - size of at least ``ldc``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldc``\ \*\ ``m`` if row - major layout is used to store matrices . See :ref:`matrix-storage` for more details. - - ldc - The leading dimension of ``C``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if column major layout is used to store matrices. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output buffer, overwritten by ``alpha``\ \*\ ``A``\ \*\ ``B`` + - ``beta``\ \*\ ``C`` (``left_right`` = ``side::left``) or - ``alpha``\ \*\ ``B``\ \*\ ``A`` + ``beta``\ \*\ ``C`` - (``left_right`` = ``side::right``). - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized before - calling ``hemm``. - - - -.. _onemkl_blas_hemm_usm: - -hemm (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event hemm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - std::int64_t m, - std::int64_t n, - T alpha, - const T* a, - std::int64_t lda, - const T* b, - std::int64_t ldb, - T beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event hemm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - std::int64_t m, - std::int64_t n, - T alpha, - const T* a, - std::int64_t lda, - const T* b, - std::int64_t ldb, - T beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies whether ``A`` is on the left side of the - multiplication (``side::left``) or on the right side - (``side::right``). See :ref:`onemkl_datatypes` for more details. - - uplo - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - m - Specifies the number of rows of the matrix ``B`` and ``C``. - - The value of ``m`` must be at least zero. - - n - Specifies the number of columns of the matrix ``B`` and ``C``. - - The value of ``n`` must be at least zero. - - alpha - Scaling factor for the matrix-matrix product. - - a - Pointer to input matrix ``A``. Must have size at least - ``lda``\ \*\ ``m`` if ``A`` is on the left of the - multiplication, or ``lda``\ \*\ ``n`` if ``A`` is on the right. - See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of ``A``. Must be at least ``m`` if ``A`` is - on the left of the multiplication, or at least ``n`` if ``A`` - is on the right. Must be positive. - - b - Pointer to input matrix ``B``. Must have size at least - ``ldb``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldb``\ \*\ ``m`` if row - major layout is used to store matrices. See :ref:`matrix-storage` for - more details. - - ldb - Leading dimension of ``B``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if column major layout is used to store matrices. - - beta - Scaling factor for matrix ``C``. - - c - The pointer to input/output matrix ``C``. It must have a - size of at least ``ldc``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldc``\ \*\ ``m`` if row - major layout is used to store matrices . See :ref:`matrix-storage` for more details. - - ldc - The leading dimension of ``C``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if column major layout is used to store matrices. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Pointer to the output matrix, overwritten by - ``alpha``\ \*\ ``A``\ \*\ ``B`` + ``beta``\ \*\ ``C`` - (``left_right`` = ``side::left``) or - ``alpha``\ \*\ ``B``\ \*\ ``A`` + ``beta``\ \*\ ``C`` - (``left_right`` = ``side::right``). - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized - before calling ``hemm``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-3-routines` diff --git a/docs/domains/blas/hemv.rst b/docs/domains/blas/hemv.rst deleted file mode 100644 index cc959ba95..000000000 --- a/docs/domains/blas/hemv.rst +++ /dev/null @@ -1,232 +0,0 @@ -.. _onemkl_blas_hemv: - -hemv -==== - -Computes a matrix-vector product using a Hermitian matrix. - -.. _onemkl_blas_hemv_description: - -.. rubric:: Description - -The ``hemv`` routines compute a scalar-matrix-vector product and add the -result to a scalar-vector product, with a Hermitian matrix. The -operation is defined as - -.. math:: - - y \leftarrow alpha*A*x + beta*y - -where: - -``alpha`` and ``beta`` are scalars, - -``A`` is an ``n``-by-``n`` Hermitian matrix, - -``x`` and ``y`` are vectors of length ``n``. - -``hemv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_hemv_buffer: - -hemv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void hemv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void hemv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether *A* is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``m``, and - positive. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - y - Buffer holding the updated vector ``y``. - - - -.. _onemkl_blas_hemv_usm: - -hemv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event hemv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event hemv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether *A* is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``m``, and - positive. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Pointer to the updated vector ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/her.rst b/docs/domains/blas/her.rst deleted file mode 100644 index ed243f62d..000000000 --- a/docs/domains/blas/her.rst +++ /dev/null @@ -1,205 +0,0 @@ -.. _onemkl_blas_her: - -her -=== - -Computes a rank-1 update of a Hermitian matrix. - -.. _onemkl_blas_her_description: - -.. rubric:: Description - -The ``her`` routines compute a scalar-vector-vector product and add the -result to a Hermitian matrix. The operation is defined as: - -.. math:: - - A \leftarrow alpha*x*x^H + A - -where: - -``alpha`` is scalar, - -``A`` is an ``n``-by-``n`` Hermitian matrix, - -``x`` is a vector of length ``n``. - -``her`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_her_buffer: - -her (Buffer Version) --------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void her(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &a, - std::int64_t lda) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void her(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &a, - std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - -.. container:: section - - .. rubric:: Output Parameters - - a - Buffer holding the updated upper triangular part of the Hermitian - matrix ``A`` if ``upper_lower``\ \= ``upper`` or the updated - lower triangular part of the Hermitian matrix ``A`` if - ``upper_lower``\ \ =\ ``lower``. - - The imaginary parts of the diagonal elements are set to zero. - - -.. _onemkl_blas_her_usm: - -her (USM Version) ------------------ - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event her(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event her(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether *A* is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the updated upper triangular part of the Hermitian - matrix ``A`` if ``upper_lower``\ \=\ ``upper`` or the updated - lower triangular part of the Hermitian matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - - The imaginary parts of the diagonal elements are set to zero. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/her2.rst b/docs/domains/blas/her2.rst deleted file mode 100644 index f9272adae..000000000 --- a/docs/domains/blas/her2.rst +++ /dev/null @@ -1,231 +0,0 @@ -.. _onemkl_blas_her2: - -her2 -==== - -Computes a rank-2 update of a Hermitian matrix. - -.. _onemkl_blas_her2_description: - -.. rubric:: Description - -The ``her2`` routines compute two scalar-vector-vector products and add -them to a Hermitian matrix. The operation is defined as: - -.. math:: - - A \leftarrow alpha*x*y^H + conjg(alpha)*y*x^H + A - -where: - -``alpha`` is a scalar, - -``A`` is an ``n``-by-``n`` Hermitian matrix, - -``x`` and ``y`` are vectors or length ``n``. - -``her2`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_her2_buffer: - -her2 (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void her2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a, - std::int64_t lda) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void her2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a, - std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - -.. container:: section - - .. rubric:: Output Parameters - - a - Buffer holding the updated upper triangular part of the Hermitian - matrix ``A`` if ``upper_lower``\ \=\ ``upper``, or the updated - lower triangular part of the Hermitian matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - - The imaginary parts of the diagonal elements are set to zero. - - - -.. _onemkl_blas_her2_usm: - -her2 (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event her2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event her2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the updated upper triangular part of the Hermitian - matrix ``A`` if ``upper_lower``\ \=\ ``upper``, or the updated - lower triangular part of the Hermitian matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - - The imaginary parts of the diagonal elements are set to zero. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/her2k.rst b/docs/domains/blas/her2k.rst deleted file mode 100644 index 6f03052da..000000000 --- a/docs/domains/blas/her2k.rst +++ /dev/null @@ -1,397 +0,0 @@ -.. _onemkl_blas_her2k: - -her2k -===== - -Performs a Hermitian rank-2k update. - -.. _onemkl_blas_her2k_description: - -.. rubric:: Description - -The ``her2k`` routines perform a rank-2k update of an ``n`` x ``n`` -Hermitian matrix ``C`` by general matrices ``A`` and ``B``. - -If ``trans`` = ``transpose::nontrans``, the operation is defined as: - -.. math:: - - C \leftarrow alpha*A*B^H + conjg(alpha)*B*A^H + beta*C - -where ``A`` is ``n`` x ``k`` and ``B`` is ``k`` x ``n``. - -If ``trans`` = ``transpose::conjtrans``, the operation is defined as: - -.. math:: - - C \leftarrow alpha*B*A^H + conjg(alpha)*A*B^H + beta*C - -where ``A`` is ``k`` x ``n`` and ``B`` is ``n`` x ``k``. - -In both cases: - -``alpha`` is a complex scalar and ``beta`` is a real scalar. - -``C`` is a Hermitian matrix and ``A`` , ``B`` are general matrices. - -The inner dimension of both matrix multiplications is ``k``. - -``her2k`` supports the following precisions: - - .. list-table:: - :header-rows: 1 - - * - T - - T_real - * - ``std::complex`` - - ``float`` - * - ``std::complex`` - - ``double`` - -.. _onemkl_blas_her2k_buffer: - -her2k (Buffer Version) ----------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void her2k(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - T_real beta, - sycl::buffer &c, - std::int64_t ldc) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void her2k(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - T_real beta, - sycl::buffer &c, - std::int64_t ldc) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies the operation to apply, as described above. Supported - operations are ``transpose::nontrans`` and - ``transpose::conjtrans``. - - n - The number of rows and columns in ``C``. The value of ``n`` must - be at least zero. - - k - The inner dimension of matrix multiplications. The value of ``k`` - must be at least equal to zero. - - alpha - Complex scaling factor for the rank-2k update. - - a - Buffer holding input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n`` - * - Row major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - See :ref:`matrix-storage` for - more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - b - Buffer holding input matrix ``B``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k`` - * - Row major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - See :ref:`matrix-storage` - for more details. - - ldb - The leading dimension of ``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - * - Row major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - - beta - Real scaling factor for matrix ``C``. - - c - Buffer holding input/output matrix ``C``. Must have size at least - ``ldc``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - ldc - Leading dimension of ``C``. Must be positive and at least ``n``. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output buffer, overwritten by the updated ``C`` matrix. - - -.. _onemkl_blas_her2k_usm: - -her2k (USM Version) -------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event her2k(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - const T* a, - std::int64_t lda, - const T* b, - std::int64_t ldb, - T_real beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event her2k(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - const T* a, - std::int64_t lda, - const T* b, - std::int64_t ldb, - T_real beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies the operation to apply, as described above. Supported - operations are ``transpose::nontrans`` and - ``transpose::conjtrans``. - - n - The number of rows and columns in ``C``. The value of ``n`` - must be at least zero. - - k - The inner dimension of matrix multiplications. The value of - ``k`` must be at least equal to zero. - - alpha - Complex scaling factor for the rank-2k update. - - a - Pointer to input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n`` - * - Row major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - See :ref:`matrix-storage` for more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - b - Pointer to input matrix ``B``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k`` - * - Row major - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k``. - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - See :ref:`matrix-storage` for - more details. - - ldb - The leading dimension of ``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - * - Row major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - - beta - Real scaling factor for matrix ``C``. - - c - Pointer to input/output matrix ``C``. Must have size at least - ``ldc``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - ldc - Leading dimension of ``C``. Must be positive and at least - ``n``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Pointer to the output matrix, overwritten by the updated ``C`` - matrix. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - - **Parent topic:** :ref:`blas-level-3-routines` diff --git a/docs/domains/blas/herk.rst b/docs/domains/blas/herk.rst deleted file mode 100644 index e4d4d16a9..000000000 --- a/docs/domains/blas/herk.rst +++ /dev/null @@ -1,309 +0,0 @@ -.. _onemkl_blas_herk: - -herk -==== - -Performs a Hermitian rank-k update. - -.. _onemkl_blas_herk_description: - -.. rubric:: Description - -The ``herk`` routines compute a rank-k update of a Hermitian matrix -``C`` by a general matrix ``A``. The operation is defined as: - -.. math:: - - C \leftarrow alpha*op(A)*op(A)^H + beta*C - -where: - -op(``X``) is one of op(``X``) = ``X`` or op(``X``) = ``X``\ :sup:`H`, - -``alpha`` and ``beta`` are real scalars, - -``C`` is a Hermitian matrix and ``A`` is a general matrix. - -Here op(``A``) is ``n`` x ``k``, and ``C`` is ``n`` x ``n``. - -``herk`` supports the following precisions: - - .. list-table:: - :header-rows: 1 - - * - T - - T_real - * - ``std::complex`` - - ``float`` - * - ``std::complex`` - - ``double`` - -.. _onemkl_blas_herk_buffer: - -herk (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void herk(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T_real alpha, - sycl::buffer &a, - std::int64_t lda, - T_real beta, - sycl::buffer &c, - std::int64_t ldc) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void herk(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T_real alpha, - sycl::buffer &a, - std::int64_t lda, - T_real beta, - sycl::buffer &c, - std::int64_t ldc) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to ``A``. See - :ref:`onemkl_datatypes` for more - details. Supported operations are ``transpose::nontrans`` and - ``transpose::conjtrans``. - - n - The number of rows and columns in ``C``.The value of ``n`` must be - at least zero. - - k - Number of columns in op(``A``). - - The value of ``k`` must be at least zero. - - alpha - Real scaling factor for the rank-k update. - - a - Buffer holding input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n`` - * - Row major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - See :ref:`matrix-storage` for - more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - beta - Real scaling factor for matrix ``C``. - - c - Buffer holding input/output matrix ``C``. Must have size at least - ``ldc``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - ldc - Leading dimension of ``C``. Must be positive and at least ``n``. - -.. container:: section - - .. rubric:: Output Parameters - - c - The output buffer, overwritten by - ``alpha``\ \*op(``A``)*op(``A``)\ :sup:`T` + ``beta``\ \*\ ``C``. - The imaginary parts of the diagonal elements are set to zero. - - - -.. _onemkl_blas_herk_usm: - -herk (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event herk(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T_real alpha, - const T* a, - std::int64_t lda, - T_real beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event herk(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T_real alpha, - const T* a, - std::int64_t lda, - T_real beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to - ``A``. See :ref:`onemkl_datatypes` for more details. Supported operations are ``transpose::nontrans`` - and ``transpose::conjtrans``. - - n - The number of rows and columns in ``C``.The value of ``n`` must - be at least zero. - - k - Number of columns in op(``A``). - - The value of ``k`` must be at least zero. - - alpha - Real scaling factor for the rank-k update. - - a - Pointer to input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n`` - * - Row major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - See :ref:`matrix-storage` for more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - beta - Real scaling factor for matrix ``C``. - - c - Pointer to input/output matrix ``C``. Must have size at least - ``ldc``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - ldc - Leading dimension of ``C``. Must be positive and at least - ``n``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Pointer to the output matrix, overwritten by - ``alpha``\ \*op(``A``)*op(``A``)\ :sup:`T` + - ``beta``\ \*\ ``C``. The imaginary parts of the diagonal - elements are set to zero. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - - **Parent topic:** :ref:`blas-level-3-routines` diff --git a/docs/domains/blas/hpmv.rst b/docs/domains/blas/hpmv.rst deleted file mode 100644 index 17872b2b2..000000000 --- a/docs/domains/blas/hpmv.rst +++ /dev/null @@ -1,228 +0,0 @@ -.. _onemkl_blas_hpmv: - -hpmv -==== - -Computes a matrix-vector product using a Hermitian packed matrix. - -.. _onemkl_blas_hpmv_description: - -.. rubric:: Description - -The ``hpmv`` routines compute a scalar-matrix-vector product and add the -result to a scalar-vector product, with a Hermitian packed matrix. -The operation is defined as - -.. math:: - - y \leftarrow alpha*A*x + beta*y - -where: - -``alpha`` and ``beta`` are scalars, - -``A`` is an ``n``-by-``n`` Hermitian matrix supplied in packed form, - -``x`` and ``y`` are vectors of length ``n``. - -``hpmv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_hpmv_buffer: - -hpmv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void hpmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &a, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void hpmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &a, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Buffer holding input matrix ``A``. Must have size at least - (``n``\ \*(``n``\ +1))/2. See :ref:`matrix-storage` for - more details. - - The imaginary parts of the diagonal elements need not be set and - are assumed to be zero. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - y - Buffer holding the updated vector ``y``. - - - -.. _onemkl_blas_hpmv_usm: - -hpmv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event hpmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *a, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event hpmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *a, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least (``n``\ \*(``n``\ +1))/2. See - :ref:`matrix-storage` for - more details. - - The imaginary parts of the diagonal elements need not be set - and are assumed to be zero. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Pointer to the updated vector ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/hpr.rst b/docs/domains/blas/hpr.rst deleted file mode 100644 index a7fd49cde..000000000 --- a/docs/domains/blas/hpr.rst +++ /dev/null @@ -1,201 +0,0 @@ -.. _onemkl_blas_hpr: - -hpr -=== - -Computes a rank-1 update of a Hermitian packed matrix. - -.. _onemkl_blas_hpr_description: - -.. rubric:: Description - -The ``hpr`` routines compute a scalar-vector-vector product and add the -result to a Hermitian packed matrix. The operation is defined as - -.. math:: - - A \leftarrow alpha*x*x^H + A - -where: - -``alpha`` is scalar, - -``A`` is an ``n``-by-``n`` Hermitian matrix, supplied in packed form, - -``x`` is a vector of length ``n``. - -``hpr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_hpr_buffer: - -hpr (Buffer Version) --------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void hpr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &a) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void hpr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &a) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - a - Buffer holding input matrix ``A``. Must have size at least - (``n``\ \*(``n``-1))/2. See :ref:`matrix-storage` for - more details. - - The imaginary part of the diagonal elements need not be set and - are assumed to be zero. - -.. container:: section - - .. rubric:: Output Parameters - - a - Buffer holding the updated upper triangular part of the Hermitian - matrix ``A`` if ``upper_lower``\ \=\ ``upper``, or the updated lower - triangular part of the Hermitian matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - - The imaginary parts of the diagonal elements are set to zero. - - -.. _onemkl_blas_hpr_usm: - -hpr (USM Version) ------------------ - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event hpr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - T *a, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event hpr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - T *a, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least (``n``\ \*(``n``-1))/2. See - :ref:`matrix-storage` for - more details. - - The imaginary part of the diagonal elements need not be set and - are assumed to be zero. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the updated upper triangular part of the Hermitian - matrix ``A`` if ``upper_lower``\ \=\ ``upper``, or the updated lower - triangular part of the Hermitian matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - - The imaginary parts of the diagonal elements are set to zero. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/hpr2.rst b/docs/domains/blas/hpr2.rst deleted file mode 100644 index 3e3fa4a63..000000000 --- a/docs/domains/blas/hpr2.rst +++ /dev/null @@ -1,226 +0,0 @@ -.. _onemkl_blas_hpr2: - -hpr2 -==== - -Performs a rank-2 update of a Hermitian packed matrix. - -.. _onemkl_blas_hpr2_description: - -.. rubric:: Description - -The ``hpr2`` routines compute two scalar-vector-vector products and add -them to a Hermitian packed matrix. The operation is defined as - -.. math:: - - A \leftarrow alpha*x*y^H + conjg(alpha)*y*x^H + A - -where: - -``alpha`` is a scalar, - -``A`` is an ``n``-by-``n`` Hermitian matrix, supplied in packed form, - -``x`` and ``y`` are vectors of length ``n``. - -``hpr2`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_hpr2_buffer: - -hpr2 (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void hpr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void hpr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - - a - Buffer holding input matrix ``A``. Must have size at least - (``n``\ \*(``n``-1))/2. See :ref:`matrix-storage` for - more details. - - The imaginary parts of the diagonal elements need not be set and - are assumed to be zero. - -.. container:: section - - .. rubric:: Output Parameters - - a - Buffer holding the updated upper triangular part of the Hermitian - matrix ``A`` if ``upper_lower``\ \=\ ``upper``, or the updated lower - triangular part of the Hermitian matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - - The imaginary parts of the diagonal elements are set to zero. - - - -.. _onemkl_blas_hpr2_usm: - -hpr2 (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event hpr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event hpr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least (``n``\ \*(``n``-1))/2. See - :ref:`matrix-storage` for - more details. - - The imaginary parts of the diagonal elements need not be set - and are assumed to be zero. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the updated upper triangular part of the Hermitian - matrix ``A`` if ``upper_lower``\ \=\ ``upper``, or the updated lower - triangular part of the Hermitian matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - - The imaginary parts of the diagonal elements are set to zero. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/iamax.rst b/docs/domains/blas/iamax.rst deleted file mode 100644 index ce02af8de..000000000 --- a/docs/domains/blas/iamax.rst +++ /dev/null @@ -1,167 +0,0 @@ -.. _onemkl_blas_iamax: - -iamax -===== - -Finds the index of the element with the largest absolute value in a vector. - -.. _onemkl_blas_iamax_description: - -.. rubric:: Description - -The ``iamax`` routines return an index ``i`` such that ``x[i]`` -has the maximum absolute value of all elements in vector ``x`` (real -variants), or such that (\|Re(``x[i]``)\| + \|Im(``x[i]``)\|) is maximal -(complex variants). - -If either ``n`` or ``incx`` are not positive, the routine returns -``0``. - -If more than one vector element is found with the same largest -absolute value, the index of the first one encountered is returned. - -If the vector contains ``NaN`` values, then the routine returns the -index of the first ``NaN``. - -``iamax`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std:complex`` - -.. container:: Note - - .. rubric:: Note - :class: NoteTipHead - - The index is zero-based. - -.. _onemkl_blas_iamax_buffer: - -iamax (Buffer Version) ----------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void iamax(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &result) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void iamax(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &result) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - The number of elements in vector ``x``. - - x - The buffer that holds the input vector ``x``. The buffer must be - of size at least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` - for more details. - - incx - The stride of vector ``x``. - -.. container:: section - - .. rubric:: Output Parameters - - result - The buffer where the zero-based index ``i`` of the maximal element - is stored. - - -.. _onemkl_blas_iamax_usm: - -iamax (USM Version) -------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event iamax(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - T_res *result, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event iamax(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - T_res *result, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - The number of elements in vector ``x``. - - x - The pointer to the input vector ``x``. The array holding the - input vector ``x`` must be of size at least (1 + (``n`` - - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - The stride of vector ``x``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - result - The pointer to where the zero-based index ``i`` of the maximal - element is stored. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/iamin.rst b/docs/domains/blas/iamin.rst deleted file mode 100644 index fb724cc47..000000000 --- a/docs/domains/blas/iamin.rst +++ /dev/null @@ -1,160 +0,0 @@ -.. _onemkl_blas_iamin: - -iamin -===== - -Finds the index of the element with the smallest absolute value. - -.. _onemkl_blas_iamin_description: - -.. rubric:: Description - -The ``iamin`` routines return an index ``i`` such that ``x[i]`` has -the minimum absolute value of all elements in vector ``x`` (real -variants), or such that (\|Re(``x[i]``)\| + \|Im(``x[i]``)\|) is minimal -(complex variants). - -If either ``n`` or ``incx`` are not positive, the routine returns -``0``. - -If more than one vector element is found with the same smallest -absolute value, the index of the first one encountered is returned. - -If the vector contains ``NaN`` values, then the routine returns the -index of the first ``NaN``. - -``iamin`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. container:: Note - - .. rubric:: Note - :class: NoteTipHead - - The index is zero-based. - -.. _onemkl_blas_iamin_buffer: - -iamin (Buffer Version) ----------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void iamin(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &result) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void iamin(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &result) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector x. - -.. container:: section - - .. rubric:: Output Parameters - - result - Buffer where the zero-based index ``i`` of the minimum element - will be stored. - - -.. _onemkl_blas_iamin_usm: - -iamin (USM Version) -------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event iamin(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - T_res *result, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event iamin(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - T_res *result, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - The pointer to input vector ``x``. The array holding input - vector ``x`` must be of size at least (1 + (``n`` - - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector x. - -.. container:: section - - .. rubric:: Output Parameters - - result - Pointer to where the zero-based index ``i`` of the minimum - element will be stored. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/nrm2.rst b/docs/domains/blas/nrm2.rst deleted file mode 100644 index 879862c73..000000000 --- a/docs/domains/blas/nrm2.rst +++ /dev/null @@ -1,158 +0,0 @@ -.. _onemkl_blas_nrm2: - -nrm2 -==== - -Computes the Euclidean norm of a vector. - -.. _onemkl_blas_nrm2_description: - -.. rubric:: Description - -The ``nrm2`` routines computes Euclidean norm of a vector - -.. math:: - - result = \| x\| - -where: - -``x`` is a vector of ``n`` elements. - -``nrm2`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - - T_res - * - ``float`` - - ``float`` - * - ``double`` - - ``double`` - * - ``std::complex`` - - ``float`` - * - ``std::complex`` - - ``double`` - -.. _onemkl_blas_nrm2_buffer: - -nrm2 (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void nrm2(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &result) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void nrm2(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &result) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - -.. container:: section - - .. rubric:: Output Parameters - - result - Buffer where the Euclidean norm of the vector ``x`` will be - stored. - - -.. _onemkl_blas_nrm2_usm: - -nrm2 (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event nrm2(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - T_res *result, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event nrm2(sycl::queue &queue, - std::int64_t n, - const T *x, - std::int64_t incx, - T_res *result, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - result - Pointer to where the Euclidean norm of the vector ``x`` will be - stored. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/rot.rst b/docs/domains/blas/rot.rst deleted file mode 100644 index 736db3b4d..000000000 --- a/docs/domains/blas/rot.rst +++ /dev/null @@ -1,208 +0,0 @@ -.. _onemkl_blas_rot: - -rot -=== - -Performs rotation of points in the plane. - -.. _onemkl_blas_rot_description: - -.. rubric:: Description - -Given two vectors ``x`` and ``y`` of ``n`` elements, the ``rot`` routines -compute four scalar-vector products and update the input vectors with -the sum of two of these scalar-vector products as follow: - -.. math:: - - \left[\begin{array}{c} - x\\y - \end{array}\right] - \leftarrow - \left[\begin{array}{c} - \phantom{-}c*x + s*y\\ - -s*x + c*y - \end{array}\right] - -``rot`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - - T_scalar - * - ``float`` - - ``float`` - * - ``double`` - - ``double`` - * - ``std::complex`` - - ``float`` - * - ``std::complex`` - - ``double`` - -.. _onemkl_blas_rot_buffer: - -rot (Buffer Version) --------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void rot(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - T_scalar c, - T_scalar s) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void rot(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - T_scalar c, - T_scalar s) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input vector ``y``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - c - Scaling factor. - - s - Scaling factor. - -.. container:: section - - .. rubric:: Output Parameters - - x - Buffer holding updated buffer ``x``. - - y - Buffer holding updated buffer ``y``. - - - -.. _onemkl_blas_rot_usm: - -rot (USM Version) ------------------ - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event rot(sycl::queue &queue, - std::int64_t n, - T *x, - std::int64_t incx, - T *y, - std::int64_t incy, - T_scalar c, - T_scalar s, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event rot(sycl::queue &queue, - std::int64_t n, - T *x, - std::int64_t incx, - T *y, - std::int64_t incy, - T_scalar c, - T_scalar s, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to input vector ``y``. The array holding input vector - ``y`` must be of size at least (1 + (``n`` - 1)*abs(``incy``)). - See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - c - Scaling factor. - - s - Scaling factor. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - x - Pointer to the updated matrix ``x``. - - y - Pointer to the updated matrix ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/rotg.rst b/docs/domains/blas/rotg.rst deleted file mode 100644 index 18a065f0e..000000000 --- a/docs/domains/blas/rotg.rst +++ /dev/null @@ -1,175 +0,0 @@ -.. _onemkl_blas_rotg: - -rotg -==== - -Computes the parameters for a Givens rotation. - -.. _onemkl_blas_rotg_description: - -.. rubric:: Description - -Given the Cartesian coordinates ``(a, b)`` of a point, the ``rotg`` -routines return the parameters ``c``, ``s``, ``r``, and ``z`` -associated with the Givens rotation. The parameters ``c`` and ``s`` -define a unitary matrix such that: - -.. math:: - - \begin{bmatrix}c & s \\ -s & c\end{bmatrix}. - \begin{bmatrix}a \\ b\end{bmatrix} - =\begin{bmatrix}r \\ 0\end{bmatrix} - -The parameter ``z`` is defined such that if \|\ ``a``\ \| > -\|\ ``b``\ \|, ``z`` is ``s``; otherwise if ``c`` is not 0 ``z`` is -1/``c``; otherwise ``z`` is 1. - -``rotg`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - - T_res - * - ``float`` - - ``float`` - * - ``double`` - - ``double`` - * - ``std::complex`` - - ``float`` - * - ``std::complex`` - - ``double`` - -.. _onemkl_blas_rotg_buffer: - -rotg (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void rotg(sycl::queue &queue, - sycl::buffer &a, - sycl::buffer &b, - sycl::buffer &c, - sycl::buffer &s) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void rotg(sycl::queue &queue, - sycl::buffer &a, - sycl::buffer &b, - sycl::buffer &c, - sycl::buffer &s) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed - - a - Buffer holding the ``x``-coordinate of the point. - - b - Buffer holding the ``y``-coordinate of the point. - -.. container:: section - - .. rubric:: Output Parameters - - a - Buffer holding the parameter ``r`` associated with the Givens - rotation. - - b - Buffer holding the parameter ``z`` associated with the Givens - rotation. - - c - Buffer holding the parameter ``c`` associated with the Givens - rotation. - - s - Buffer holding the parameter ``s`` associated with the Givens - rotation. - - -.. _onemkl_blas_rotg_usm: - -rotg (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event rotg(sycl::queue &queue, - T *a, - T *b, - T_real *c, - T *s, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event rotg(sycl::queue &queue, - T *a, - T *b, - T_real *c, - T *s, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed - - a - Pointer to the ``x``-coordinate of the point. - - b - Pointer to the ``y``-coordinate of the point. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the parameter ``r`` associated with the Givens - rotation. - - b - Pointer to the parameter ``z`` associated with the Givens - rotation. - - c - Pointer to the parameter ``c`` associated with the Givens - rotation. - - s - Pointer to the parameter ``s`` associated with the Givens - rotation. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/rotm.rst b/docs/domains/blas/rotm.rst deleted file mode 100644 index da9a40c95..000000000 --- a/docs/domains/blas/rotm.rst +++ /dev/null @@ -1,266 +0,0 @@ -.. _onemkl_blas_rotm: - -rotm -==== - -Performs modified Givens rotation of points in the plane. - -.. _onemkl_blas_rotm_description: - -.. rubric:: Description - -Given two vectors ``x`` and ``y``, each vector element of these -vectors is replaced as follows: - -.. math:: - - \begin{bmatrix}x_i \\ y_i\end{bmatrix}= - H - \begin{bmatrix}x_i \\ y_i\end{bmatrix} - -for ``i`` from 1 to ``n``, where ``H`` is a modified Givens -transformation matrix. - -``rotm`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. _onemkl_blas_rotm_buffer: - -rotm (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void rotm(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer ¶m) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void rotm(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer ¶m) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - param - Buffer holding an array of size 5. - - The elements of the ``param`` array are: - - ``param[0]`` contains a switch, ``flag``. The other array elements - ``param[1-4]`` contain the components of the modified Givens - transformation matrix ``H``: - h\ :sub:`11`, h\ :sub:`21`, h\ :sub:`12`, and - h\ :sub:`22`, respectively. - - Depending on the values of ``flag``, the components of ``H`` - are set as follows: - - | ``flag = -1.0``: - - .. math:: - - H=\begin{bmatrix}h_{11} & h_{12} \\ h_{21} & h_{22}\end{bmatrix} - - | ``flag = 0.0``: - - .. math:: - - H=\begin{bmatrix}1.0 & h_{12} \\ h_{21} & 1.0\end{bmatrix} - - | ``flag = 1.0``: - - .. math:: - - H=\begin{bmatrix}h_{11} & 1.0 \\ -1.0 & h_{22}\end{bmatrix} - - | ``flag = -2.0``: - - .. math:: - - H=\begin{bmatrix}1.0 & 0.0 \\ 0.0 & 1.0\end{bmatrix} - - In the last three cases, the matrix entries of 1.0, -1.0, and 0.0 - are assumed based on the value of ``flag`` and are not required to - be set in the ``param`` vector. - -.. container:: section - - .. rubric:: Output Parameters - - x - Buffer holding updated buffer ``x``. - - y - Buffer holding updated buffer ``y``. - - - -.. _onemkl_blas_rotm_usm: - -rotm (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event rotm(sycl::queue &queue, - std::int64_t n, - T *x, - std::int64_t incx, - T *y, - std::int64_t incy, - T *param, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event rotm(sycl::queue &queue, - std::int64_t n, - T *x, - std::int64_t incx, - T *y, - std::int64_t incy, - T *param, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Pointer to the input vector ``x``. The array holding the vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - yparam - Pointer to the input vector ``y``. The array holding the vector - ``y`` must be of size at least (1 + (``n`` - 1)*abs(``incy``)). - See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - param - Buffer holding an array of size 5. - - The elements of the ``param`` array are: - - ``param[0]`` contains a switch, ``flag``. The other array elements - ``param[1-4]`` contain the components of the modified Givens - transformation matrix ``H``: - h\ :sub:`11`, h\ :sub:`21`, h\ :sub:`12`, and - h\ :sub:`22`, respectively. - - Depending on the values of ``flag``, the components of ``H`` - are set as follows: - - | ``flag = -1.0``: - - .. math:: - - H=\begin{bmatrix}h_{11} & h_{12} \\ h_{21} & h_{22}\end{bmatrix} - - | ``flag = 0.0``: - - .. math:: - - H=\begin{bmatrix}1.0 & h_{12} \\ h_{21} & 1.0\end{bmatrix} - - | ``flag = 1.0``: - - .. math:: - - H=\begin{bmatrix}h_{11} & 1.0 \\ -1.0 & h_{22}\end{bmatrix} - - | ``flag = -2.0``: - - .. math:: - - H=\begin{bmatrix}1.0 & 0.0 \\ 0.0 & 1.0\end{bmatrix} - - In the last three cases, the matrix entries of 1.0, -1.0, and 0.0 - are assumed based on the value of ``flag`` and are not required to - be set in the ``param`` vector. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - x - Pointer to the updated array ``x``. - - y - Pointer to the updated array ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/rotmg.rst b/docs/domains/blas/rotmg.rst deleted file mode 100644 index 49a16ff00..000000000 --- a/docs/domains/blas/rotmg.rst +++ /dev/null @@ -1,257 +0,0 @@ -.. _onemkl_blas_rotmg: - -rotmg -===== - -Computes the parameters for a modified Givens rotation. - -.. _onemkl_blas_rotmg_description: - -.. rubric:: Description - -Given Cartesian coordinates (``x1``, ``y1``) of an -input vector, the ``rotmg`` routines compute the components of a modified -Givens transformation matrix ``H`` that zeros the ``y``-component of -the resulting vector: - -.. math:: - - \begin{bmatrix}x1 \\ 0\end{bmatrix}= - H - \begin{bmatrix}x1\sqrt{d1} \\ y1\sqrt{d2}\end{bmatrix} - -``rotmg`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. _onemkl_blas_rotmg_buffer: - -rotmg (Buffer Version) ----------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void rotmg(sycl::queue &queue, - sycl::buffer &d1, - sycl::buffer &d2, - sycl::buffer &x1, - sycl::buffer &y1, - sycl::buffer ¶m) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void rotmg(sycl::queue &queue, - sycl::buffer &d1, - sycl::buffer &d2, - sycl::buffer &x1, - sycl::buffer &y1, - sycl::buffer ¶m) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - d1 - Buffer holding the scaling factor for the ``x``-coordinate of the - input vector. - - d2 - Buffer holding the scaling factor for the ``y``-coordinate of the - input vector. - - x1 - Buffer holding the ``x``-coordinate of the input vector. - - y1 - Scalar specifying the ``y``-coordinate of the input vector. - -.. container:: section - - .. rubric:: Output Parameters - - d1 - Buffer holding the first diagonal element of the updated matrix. - - d2 - Buffer holding the second diagonal element of the updated matrix. - - x1 - Buffer holding the ``x``-coordinate of the rotated vector before - scaling - - param - Buffer holding an array of size 5. - - The elements of the ``param`` array are: - - ``param[0]`` contains a switch, ``flag``. The other array elements - ``param[1-4]`` contain the components of the modified Givens - transformation matrix ``H``: - h\ :sub:`11`, h\ :sub:`21`, h\ :sub:`12`, and - h\ :sub:`22`, respectively. - - Depending on the values of ``flag``, the components of ``H`` are - set as follows: - - | ``flag = -1.0``: - - .. math:: - - H=\begin{bmatrix}h_{11} & h_{12} \\ h_{21} & h_{22}\end{bmatrix} - - | ``flag = 0.0``: - - .. math:: - - H=\begin{bmatrix}1.0 & h_{12} \\ h_{21} & 1.0\end{bmatrix} - - | ``flag = 1.0``: - - .. math:: - - H=\begin{bmatrix}h_{11} & 1.0 \\ -1.0 & h_{22}\end{bmatrix} - - | ``flag = -2.0``: - - .. math:: - - H=\begin{bmatrix}1.0 & 0.0 \\ 0.0 & 1.0\end{bmatrix} - - In the last three cases, the matrix entries of 1.0, -1.0, and 0.0 - are assumed based on the value of ``flag`` and are not required to - be set in the ``param`` vector. - - - -.. _onemkl_blas_rotmg_usm: - -rotmg (USM Version) -------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event rotmg(sycl::queue &queue, - T *d1, - T *d2, - T *x1, - T *y1, - T *param, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event rotmg(sycl::queue &queue, - T *d1, - T *d2, - T *x1, - T *y1, - T *param, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - d1 - Pointer to the scaling factor for the ``x``-coordinate of the - input vector. - - d2 - Pointer to the scaling factor for the ``y``-coordinate of the - input vector. - - x1 - Pointer to the ``x``-coordinate of the input vector. - - y1 - Scalar specifying the ``y``-coordinate of the input vector. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - d1 - Pointer to the first diagonal element of the updated matrix. - - d2 - Pointer to the second diagonal element of the updated matrix. - - x1 - Pointer to the ``x``-coordinate of the rotated vector before - scaling - - param - Buffer holding an array of size 5. - - The elements of the ``param`` array are: - - ``param[0]`` contains a switch, ``flag``. The other array elements - ``param[1-4]`` contain the components of the modified Givens - transformation matrix ``H``: - h\ :sub:`11`, h\ :sub:`21`, h\ :sub:`12`, and - h\ :sub:`22`, respectively. - - Depending on the values of ``flag``, the components of ``H`` - are set as follows: - - | ``flag = -1.0``: - - .. math:: - - H=\begin{bmatrix}h_{11} & h_{12} \\ h_{21} & h_{22}\end{bmatrix} - - | ``flag = 0.0``: - - .. math:: - - H=\begin{bmatrix}1.0 & h_{12} \\ h_{21} & 1.0\end{bmatrix} - - | ``flag = 1.0``: - - .. math:: - - H=\begin{bmatrix}h_{11} & 1.0 \\ -1.0 & h_{22}\end{bmatrix} - - | ``flag = -2.0``: - - .. math:: - - H=\begin{bmatrix}1.0 & 0.0 \\ 0.0 & 1.0\end{bmatrix} - - In the last three cases, the matrix entries of 1.0, -1.0, and 0.0 - are assumed based on the value of ``flag`` and are not required to - be set in the ``param`` vector. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/sbmv.rst b/docs/domains/blas/sbmv.rst deleted file mode 100644 index a0c071f3c..000000000 --- a/docs/domains/blas/sbmv.rst +++ /dev/null @@ -1,244 +0,0 @@ -.. _onemkl_blas_sbmv: - -sbmv -==== - -Computes a matrix-vector product with a symmetric band matrix. - -.. _onemkl_blas_sbmv_description: - -.. rubric:: Description - -The ``sbmv`` routines compute a scalar-matrix-vector product and add the -result to a scalar-vector product, with a symmetric band matrix. The -operation is defined as: - -.. math:: - - y \leftarrow alpha*A*x + beta*y - -where: - -``alpha`` and ``beta`` are scalars, - -``A`` is an ``n``-by-``n`` symmetric matrix with ``k`` -super-diagonals, - -``x`` and ``y`` are vectors of length ``n``. - -``sbmv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. _onemkl_blas_sbmv_buffer: - -sbmv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void sbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void sbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - k - Number of super-diagonals of the matrix ``A``. Must be at least - zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least (``k`` + 1), - and positive. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - y - Buffer holding the updated vector ``y``. - - -.. _onemkl_blas_sbmv_usm: - -sbmv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event sbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - std::int64_t k, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event sbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - std::int64_t k, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - k - Number of super-diagonals of the matrix ``A``. Must be at least - zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least (``k`` + - 1), and positive. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Pointer to the updated vector ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/scal.rst b/docs/domains/blas/scal.rst deleted file mode 100644 index 155b5fc49..000000000 --- a/docs/domains/blas/scal.rst +++ /dev/null @@ -1,162 +0,0 @@ -.. _onemkl_blas_scal: - -scal -==== - -Computes the product of a vector by a scalar. - -.. _onemkl_blas_scal_description: - -.. rubric:: Description - -The ``scal`` routines computes a scalar-vector product: - -.. math:: - - x \leftarrow alpha*x - -where: - -``x`` is a vector of ``n`` elements, - -``alpha`` is a scalar. - -``scal`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - - T_scalar - * - ``float`` - - ``float`` - * - ``double`` - - ``double`` - * - ``std::complex`` - - ``std::complex`` - * - ``std::complex`` - - ``std::complex`` - * - ``std::complex`` - - ``float`` - * - ``std::complex`` - - ``double`` - -.. _onemkl_blas_scal_buffer: - -scal (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void scal(sycl::queue &queue, - std::int64_t n, - T_scalar alpha, - sycl::buffer &x, - std::int64_t incx) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void scal(sycl::queue &queue, - std::int64_t n, - T_scalar alpha, - sycl::buffer &x, - std::int64_t incx) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - alpha - Specifies the scalar ``alpha``. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - -.. container:: section - - .. rubric:: Output Parameters - - x - Buffer holding updated buffer ``x``. - - -.. _onemkl_blas_scal_usm: - -scal (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event scal(sycl::queue &queue, - std::int64_t n, - T_scalar alpha, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event scal(sycl::queue &queue, - std::int64_t n, - T_scalar alpha, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - alpha - Specifies the scalar ``alpha``. - - x - Pointer to the input vector ``x``. The array must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - -.. container:: section - - .. rubric:: Output Parameters - - x - Pointer to the updated array ``x``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/sdsdot.rst b/docs/domains/blas/sdsdot.rst deleted file mode 100644 index 34d939c2f..000000000 --- a/docs/domains/blas/sdsdot.rst +++ /dev/null @@ -1,172 +0,0 @@ -.. _onemkl_blas_sdsdot: - -sdsdot -====== - -Computes a vector-vector dot product with double precision. - -.. _onemkl_blas_sdsdot_description: - -.. rubric:: Description - -The ``sdsdot`` routines perform a dot product between two vectors with -double precision: - -.. math:: - - result = sb + \sum_{i=1}^{n}X_iY_i - -.. _onemkl_blas_sdsdot_buffer: - -sdsdot (Buffer Version) ------------------------ - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void sdsdot(sycl::queue &queue, - std::int64_t n, - float sb, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &result) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void sdsdot(sycl::queue &queue, - std::int64_t n, - float sb, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &result) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vectors ``x`` and ``y``. - - sb - Single precision scalar to be added to the dot product. - - x - Buffer holding input vector ``x``. The buffer must be of size - at least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input vector ``y``. The buffer must be of size - at least (1 + (``n`` - 1)*abs(``incxy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - result - Buffer where the result (a scalar) will be stored. If ``n`` < 0 - the result is ``sb``. - - -.. _onemkl_blas_sdsdot_usm: - -sdsdot (USM Version) --------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event sdsdot(sycl::queue &queue, - std::int64_t n, - float sb, - const float *x, - std::int64_t incx, - const float *y, - std::int64_t incy, - float *result, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event sdsdot(sycl::queue &queue, - std::int64_t n, - float sb, - const float *x, - std::int64_t incx, - const float *y, - std::int64_t incy, - float *result, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vectors ``x`` and ``y``. - - sb - Single precision scalar to be added to the dot product. - - x - Pointer to the input vector ``x``. The array must be of size - at least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` - for more details. - - incx - Stride of vector ``x``. - - y - Pointer to the input vector ``y``. The array must be of size - at least (1 + (``n`` - 1)*abs(``incxy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if - any. If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - result - Pointer to where the result (a scalar) will be stored. If - ``n`` < 0 the result is ``sb``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/spmv.rst b/docs/domains/blas/spmv.rst deleted file mode 100644 index 3ae196901..000000000 --- a/docs/domains/blas/spmv.rst +++ /dev/null @@ -1,220 +0,0 @@ -.. _onemkl_blas_spmv: - -spmv -==== - -Computes a matrix-vector product with a symmetric packed matrix. - -.. _onemkl_blas_spmv_description: - -.. rubric:: Description - -The ``spmv`` routines compute a scalar-matrix-vector product and add the -result to a scalar-vector product, with a symmetric packed matrix. -The operation is defined as: - -.. math:: - - y \leftarrow alpha*A*x + beta*y - -where: - -``alpha`` and ``beta`` are scalars, - -``A`` is an ``n``-by-``n`` symmetric matrix, supplied in packed form, - -``x`` and ``y`` are vectors of length ``n``. - -``spmv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. _onemkl_blas_spmv_buffer: - -spmv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void spmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &a, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void spmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &a, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Buffer holding input matrix ``A``. Must have size at least - (``n``\ \*(``n``\ +1))/2. See :ref:`matrix-storage` for - more details. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - y - Buffer holding the updated vector ``y``. - - -.. _onemkl_blas_spmv_usm: - -spmv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event spmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *a, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event spmv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *a, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least (``n``\ \*(``n``\ +1))/2. See - :ref:`matrix-storage` for - more details. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - beta - Scaling factor for vector ``y``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Pointer to the updated vector ``y``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/spr.rst b/docs/domains/blas/spr.rst deleted file mode 100644 index 2ef91546d..000000000 --- a/docs/domains/blas/spr.rst +++ /dev/null @@ -1,193 +0,0 @@ -.. _onemkl_blas_spr: - -spr -=== - -Performs a rank-1 update of a symmetric packed matrix. - -.. _onemkl_blas_spr_description: - -.. rubric:: Description - -The ``spr`` routines compute a scalar-vector-vector product and add the -result to a symmetric packed matrix. The operation is defined as: - -.. math:: - - A \leftarrow alpha*x*x^T + A - -where: - -``alpha`` is scalar, - -``A`` is an ``n``-by-``n`` symmetric matrix, supplied in packed form, - -``x`` is a vector of length ``n``. - -``spr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. _onemkl_blas_spr_buffer: - -spr (Buffer Version) --------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void spr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &a) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void spr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &a) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - a - Buffer holding input matrix ``A``. Must have size at least - (``n``\ \*(``n`` + 1))/2. See :ref:`matrix-storage` for - more details. - -.. container:: section - - .. rubric:: Output Parameters - :class: sectiontitle - - a - Buffer holding the updated upper triangular part of the symmetric - matrix ``A`` if ``upper_lower``\ \=\ ``upper``, or the updated lower - triangular part of the symmetric matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - - - -.. _onemkl_blas_spr_usm: - -spr (USM Version) ------------------ - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event spr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - T *a, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event spr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - T *a, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least (``n``\ \*(``n`` + 1))/2. See - :ref:`matrix-storage` for - more details. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the updated upper triangular part of the symmetric - matrix ``A`` if ``upper_lower``\ \=\ ``upper``, or the updated lower - triangular part of the symmetric matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/spr2.rst b/docs/domains/blas/spr2.rst deleted file mode 100644 index 6cd195cbb..000000000 --- a/docs/domains/blas/spr2.rst +++ /dev/null @@ -1,213 +0,0 @@ -.. _onemkl_blas_spr2: - -spr2 -==== - -Computes a rank-2 update of a symmetric packed matrix. - -.. _onemkl_blas_spr2_description: - -.. rubric:: Description - -The ``spr2`` routines compute two scalar-vector-vector products and add -them to a symmetric packed matrix. The operation is defined as: - -.. math:: - - A \leftarrow alpha*x*y^T + alpha*y*x^T + A - -where: - -``alpha`` is scalar, - -``A`` is an ``n``-by-``n`` symmetric matrix, supplied in packed form, - -``x`` and ``y`` are vectors of length ``n``. - -``spr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. _onemkl_blas_spr2_buffer: - -spr2 (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void spr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void spr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - - a - Buffer holding input matrix ``A``. Must have size at least - (``n``\ \*(``n``-1))/2. See :ref:`matrix-storage` for - more details. - -.. container:: section - - .. rubric:: Output Parameters - - a - Buffer holding the updated upper triangular part of the symmetric - matrix ``A`` if ``upper_lower``\ \=\ ``upper`` or the updated lower - triangular part of the symmetric matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - - -.. _onemkl_blas_spr2_usm: - -spr2 (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event spr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event spr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least (``n``\ \*(``n``-1))/2. See - :ref:`matrix-storage` for - more details. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the updated upper triangular part of the symmetric - matrix ``A`` if ``upper_lower``\ \=\ ``upper`` or the updated lower - triangular part of the symmetric matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/swap.rst b/docs/domains/blas/swap.rst deleted file mode 100644 index 79c2d4121..000000000 --- a/docs/domains/blas/swap.rst +++ /dev/null @@ -1,184 +0,0 @@ -.. _onemkl_blas_swap: - -swap -==== - -Swaps a vector with another vector. - -.. _onemkl_blas_swap_description: - -.. rubric:: Description - -Given two vectors of ``n`` elements, ``x`` and ``y``, the ``swap`` -routines return vectors ``y`` and ``x`` swapped, each replacing the -other. - -.. math:: - - \left[\begin{array}{c} - y\\x - \end{array}\right] - \leftarrow - \left[\begin{array}{c} - x\\y - \end{array}\right] - -``swap`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_swap_buffer: - -swap (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void swap(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void swap(sycl::queue &queue, - std::int64_t n, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input vector ``y``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - x - Buffer holding updated buffer ``x``, that is, the input vector - ``y``. - - y - Buffer holding updated buffer ``y``, that is, the input vector - ``x``. - - - -.. _onemkl_blas_swap_usm: - -swap (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event swap(sycl::queue &queue, - std::int64_t n, - T *x, - std::int64_t incx, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event swap(sycl::queue &queue, - std::int64_t n, - T *x, - std::int64_t incx, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - n - Number of elements in vector ``x``. - - x - Pointer to the input vector ``x``. The array must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to the input vector ``y``. The array must be of size at - least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - x - Pointer to the updated array ``x``, that is, the input vector - ``y``. - - y - Pointer to the updated array ``y``, that is, the input vector - ``x``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-1-routines` diff --git a/docs/domains/blas/symm.rst b/docs/domains/blas/symm.rst deleted file mode 100644 index 26e16e499..000000000 --- a/docs/domains/blas/symm.rst +++ /dev/null @@ -1,311 +0,0 @@ -.. _onemkl_blas_symm: - -symm -==== - -Computes a matrix-matrix product where one input matrix is symmetric -and one matrix is general. - -.. _onemkl_blas_symm_description: - -.. rubric:: Description - -The ``symm`` routines compute a scalar-matrix-matrix product and add the -result to a scalar-matrix product, where one of the matrices in the -multiplication is symmetric. The argument ``left_right`` determines -if the symmetric matrix, ``A``, is on the left of the multiplication -(``left_right`` = ``side::left``) or on the right (``left_right`` = -``side::right``). Depending on ``left_right``, the operation is -defined as: - -.. math:: - - C \leftarrow alpha*A*B + beta*C - -or - -.. math:: - - C \leftarrow alpha*B*A + beta*C - -where: - -``alpha`` and ``beta`` are scalars, - -``A`` is a symmetric matrix, either ``m``-by-``m`` or ``n``-by-``n``, - -``B`` and ``C`` are ``m``-by-``n`` matrices. - -``symm`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_symm_buffer: - -symm (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void symm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - T beta, - sycl::buffer &c, - std::int64_t ldc) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void symm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - T beta, - sycl::buffer &c, - std::int64_t ldc) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies whether ``A`` is on the left side of the multiplication - (``side::left``) or on the right side (``side::right``). See :ref:`onemkl_datatypes` for more details. - - upper_lower - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - m - Number of rows of ``B`` and ``C``. The value of ``m`` must be at - least zero. - - n - Number of columns of ``B`` and ``C``. The value of ``n`` must be - at least zero. - - alpha - Scaling factor for the matrix-matrix product. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``m`` if ``A`` is on the left of the multiplication, - or ``lda``\ \*\ ``n`` if ``A`` is on the right. See :ref:`matrix-storage` - for more details. - - lda - Leading dimension of ``A``. Must be at least ``m`` if ``A`` is on - the left of the multiplication, or at least ``n`` if ``A`` is on - the right. Must be positive. - - b - Buffer holding input matrix ``B``. Must have size at least - ``ldb``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldb``\ \*\ ``m`` if row - major layout is used to store matrices. See :ref:`matrix-storage` for - more details. - - ldb - Leading dimension of ``B``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if column major layout is used to store matrices. - - beta - Scaling factor for matrix ``C``. - - c - The buffer holding the input/output matrix ``C``. It must have a - size of at least ``ldc``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldc``\ \*\ ``m`` if row - major layout is used to store matrices. See :ref:`matrix-storage` for more details. - - ldc - The leading dimension of ``C``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if column major layout is used to store matrices. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output buffer, overwritten by ``alpha``\ \*\ ``A``\ \*\ ``B`` + - ``beta``\ \*\ ``C`` (``left_right`` = ``side::left``) or - ``alpha``\ \*\ ``B``\ \*\ ``A`` + ``beta``\ \*\ ``C`` - (``left_right`` = ``side::right``). - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized before - calling ``symm``. - - -.. _onemkl_blas_symm_usm: - -symm (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event symm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - std::int64_t m, - std::int64_t n, - T alpha, - const T* a, - std::int64_t lda, - const T* b, - std::int64_t ldb, - T beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event symm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - std::int64_t m, - std::int64_t n, - T alpha, - const T* a, - std::int64_t lda, - const T* b, - std::int64_t ldb, - T beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies whether ``A`` is on the left side of the - multiplication (``side::left``) or on the right side - (``side::right``). See :ref:`onemkl_datatypes` for more details. - - upper_lower - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - m - Number of rows of ``B`` and ``C``. The value of ``m`` must be - at least zero. - - n - Number of columns of ``B`` and ``C``. The value of ``n`` must - be at least zero. - - alpha - Scaling factor for the matrix-matrix product. - - a - Pointer to input matrix ``A``. Must have size at least - ``lda``\ \*\ ``m`` if ``A`` is on the left of the - multiplication, or ``lda``\ \*\ ``n`` if ``A`` is on the right. - See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of ``A``. Must be at least ``m`` if ``A`` is - on the left of the multiplication, or at least ``n`` if ``A`` - is on the right. Must be positive. - - b - Pointer to input matrix ``B``. Must have size at least - ``ldb``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldb``\ \*\ ``m`` if row - major layout is used to store matrices. See :ref:`matrix-storage` for - more details. - - ldb - Leading dimension of ``B``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if column major layout is used to store matrices. - - beta - Scaling factor for matrix ``C``. - - c - The pointer to input/output matrix ``C``. It must have a - size of at least ``ldc``\ \*\ ``n`` if column major layout is - used to store matrices or at least ``ldc``\ \*\ ``m`` if row - major layout is used to store matrices . See :ref:`matrix-storage` for more details. - - ldc - The leading dimension of ``C``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if column major layout is used to store matrices. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Pointer to the output matrix, overwritten by - ``alpha``\ \*\ ``A``\ \*\ ``B`` + ``beta``\ \*\ ``C`` - (``left_right`` = ``side::left``) or - ``alpha``\ \*\ ``B``\ \*\ ``A`` + ``beta``\ \*\ ``C`` - (``left_right`` = ``side::right``). - -.. container:: section - - .. rubric:: Notes - - If ``beta`` = 0, matrix ``C`` does not need to be initialized - before calling ``symm``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-3-routines` diff --git a/docs/domains/blas/symv.rst b/docs/domains/blas/symv.rst deleted file mode 100644 index d3750ec58..000000000 --- a/docs/domains/blas/symv.rst +++ /dev/null @@ -1,226 +0,0 @@ -.. _onemkl_blas_symv: - -symv -==== - -Computes a matrix-vector product for a symmetric matrix. - -.. _onemkl_blas_symv_description: - -.. rubric:: Description - -The ``symv`` routines routines compute a scalar-matrix-vector product and -add the result to a scalar-vector product, with a symmetric matrix. -The operation is defined as: - -.. math:: - - y \leftarrow alpha*A*x + beta*y - -where: - -``alpha`` and ``beta`` are scalars, - -``A`` is an ``n``-by-``n`` symmetric matrix, - -``x`` and ``y`` are vectors of length ``n``. - -``symv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. _onemkl_blas_symv_buffer: - -symv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void symv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void symv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx, - T beta, - sycl::buffer &y, - std::int64_t incy) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``m``, and - positive. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - -.. container:: section - - .. rubric:: Output Parameters - - y - Buffer holding the updated vector ``y``. - - -.. _onemkl_blas_symv_usm: - -symv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event symv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event symv(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *a, - std::int64_t lda, - const T *x, - std::int64_t incx, - T beta, - T *y, - std::int64_t incy, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``m``, and - positive. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - y - Pointer to the updated vector ``y``. - - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/syr.rst b/docs/domains/blas/syr.rst deleted file mode 100644 index 74f692a70..000000000 --- a/docs/domains/blas/syr.rst +++ /dev/null @@ -1,202 +0,0 @@ -.. _onemkl_blas_syr: - -syr -=== - -Computes a rank-1 update of a symmetric matrix. - -.. _onemkl_blas_syr_description: - -.. rubric:: Description - -The ``syr`` routines compute a scalar-vector-vector product add them and -add the result to a matrix, with a symmetric matrix. The operation is -defined as: - -.. math:: - - A \leftarrow alpha*x*x^T + A - -where: - -``alpha`` is scalar, - -``A`` is an ``n``-by-``n`` symmetric matrix, - -``x`` is a vector of length ``n``. - -``syr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. _onemkl_blas_syr_buffer: - -syr (Buffer Version) --------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void syr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &a, - std::int64_t lda) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void syr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &a, - std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - -.. container:: section - - .. rubric:: Output Parameters - - a - Buffer holding the updated upper triangular part of the symmetric - matrix ``A`` if ``upper_lower``\ \=\ ``upper`` or the updated lower - triangular part of the symmetric matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - - -.. _onemkl_blas_syr_usm: - -syr (USM Version) ------------------ - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event syr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event syr(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the updated upper triangular part of the symmetric - matrix ``A`` if ``upper_lower``\ \=\ ``upper`` or the updated lower - triangular part of the symmetric matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/syr2.rst b/docs/domains/blas/syr2.rst deleted file mode 100644 index b7628e3f1..000000000 --- a/docs/domains/blas/syr2.rst +++ /dev/null @@ -1,228 +0,0 @@ -.. _onemkl_blas_syr2: - -syr2 -==== - -Computes a rank-2 update of a symmetric matrix. - -.. _onemkl_blas_syr2_description: - -.. rubric:: Description - -The ``syr2`` routines compute two scalar-vector-vector product add them -and add the result to a matrix, with a symmetric matrix. The -operation is defined as: - -.. math:: - - A \leftarrow alpha*x*y^T + alpha*y*x^T + A - -where: - -``alpha`` is a scalar, - -``A`` is an ``n``-by-``n`` symmetric matrix, - -``x`` and ``y`` are vectors of length ``n``. - -``syr2`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. _onemkl_blas_syr2_buffer: - -syr2 (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void syr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a, - std::int64_t lda) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void syr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - sycl::buffer &x, - std::int64_t incx, - sycl::buffer &y, - std::int64_t incy, - sycl::buffer &a, - std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Buffer holding input/output vector ``y``. The buffer must be of - size at least (1 + (``n`` - 1)*abs(``incy``)). See :ref:`matrix-storage` - for more details. - - incy - Stride of vector ``y``. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - -.. container:: section - - .. rubric:: Output Parameters - - a - Buffer holding the updated upper triangular part of the symmetric - matrix ``A`` if ``upper_lower``\ \=\ ``upper``, or the updated lower - triangular part of the symmetric matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - - - -.. _onemkl_blas_syr2_usm: - -syr2 (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event syr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event syr2(sycl::queue &queue, - onemkl::uplo upper_lower, - std::int64_t n, - T alpha, - const T *x, - std::int64_t incx, - const T *y, - std::int64_t incy, - T *a, - std::int64_t lda, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - n - Number of columns of ``A``. Must be at least zero. - - alpha - Scaling factor for the matrix-vector product. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - y - Pointer to input/output vector ``y``. The array holding - input/output vector ``y`` must be of size at least (1 + (``n`` - - 1)*abs(``incy``)). See :ref:`matrix-storage` for - more details. - - incy - Stride of vector ``y``. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - a - Pointer to the updated upper triangular part of the symmetric - matrix ``A`` if ``upper_lower``\ \=\ ``upper``, or the updated lower - triangular part of the symmetric matrix ``A`` if - ``upper_lower``\ \=\ ``lower``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/syr2k.rst b/docs/domains/blas/syr2k.rst deleted file mode 100644 index 8605779c6..000000000 --- a/docs/domains/blas/syr2k.rst +++ /dev/null @@ -1,397 +0,0 @@ -.. _onemkl_blas_syr2k: - -syr2k -===== - -Performs a symmetric rank-2k update. - -.. _onemkl_blas_syr2k_description: - -.. rubric:: Description - -The ``syr2k`` routines perform a rank-2k update of an ``n`` x ``n`` -symmetric matrix ``C`` by general matrices ``A`` and ``B``. - -If ``trans`` = ``transpose::nontrans``, the operation is defined as: - -.. math:: - - C \leftarrow alpha*(A*B^T + B*A^T) + beta*C - -where ``A`` and ``B`` are ``n`` x ``k`` matrices. - -If ``trans`` = ``transpose::trans``, the operation is defined as: - -.. math:: - - C \leftarrow alpha*(A^T*B + B^T*A) + beta * C - - -where ``A`` and ``B`` are ``k`` x ``n`` matrices. - - -In both cases: - -``alpha`` and ``beta`` are scalars, - -``C`` is a symmetric matrix and ``A``,\ ``B`` are general matrices, - -The inner dimension of both matrix multiplications is ``k``. - -``syr2k`` supports the following precisions: - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_syr2k_buffer: - -syr2k (Buffer Version) ----------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void syr2k(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - T beta, - sycl::buffer &c, - std::int64_t ldc) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void syr2k(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, - T beta, - sycl::buffer &c, - std::int64_t ldc) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies the operation to apply, as described above. Conjugation - is never performed, even if ``trans`` = ``transpose::conjtrans``. - - n - Number of rows and columns in ``C``.The value of ``n`` must be at - least zero. - - k - Inner dimension of matrix multiplications.The value of ``k`` must - be at least zero. - - alpha - Scaling factor for the rank-2k update. - - a - Buffer holding input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n`` - * - Row major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - See :ref:`matrix-storage` for - more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - b - Buffer holding input matrix ``B``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k`` - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - * - Row major - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k``. - - See :ref:`matrix-storage` - for more details. - - ldb - The leading dimension of ``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - * - Row major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - - beta - Scaling factor for matrix ``C``. - - c - Buffer holding input/output matrix ``C``. Must have size at least - ``ldc``\ \*\ ``n``. See :ref:`matrix-storage` for - more details - - ldc - Leading dimension of ``C``. Must be positive and at least ``n``. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output buffer, overwritten by the updated ``C`` matrix. - - - -.. _onemkl_blas_syr2k_usm: - -syr2k (USM Version) -------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event syr2k(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - const T* a, - std::int64_t lda, - const T* b, - std::int64_t ldb, - T beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event syr2k(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - const T* a, - std::int64_t lda, - const T* b, - std::int64_t ldb, - T beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies the operation to apply, as described above. - Conjugation is never performed, even if ``trans`` = - ``transpose::conjtrans``. - - n - Number of rows and columns in ``C``. The value of ``n`` must be - at least zero. - - k - Inner dimension of matrix multiplications.The value of ``k`` - must be at least zero. - - alpha - Scaling factor for the rank-2k update. - - a - Pointer to input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n`` - * - Row major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - See :ref:`matrix-storage` for more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - b - Pointer to input matrix ``B``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k`` - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - * - Row major - - ``B`` is an ``n``-by-``k`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``n``. - - ``B`` is an ``k``-by-``n`` matrix so the array ``b`` - must have size at least ``ldb``\ \*\ ``k``. - - See :ref:`matrix-storage` for - more details. - - ldb - The leading dimension of ``B``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``ldb`` must be at least ``n``. - - ``ldb`` must be at least ``k``. - * - Row major - - ``ldb`` must be at least ``k``. - - ``ldb`` must be at least ``n``. - - beta - Scaling factor for matrix ``C``. - - c - Pointer to input/output matrix ``C``. Must have size at least - ``ldc``\ \*\ ``n``. See :ref:`matrix-storage` for - more details - - ldc - Leading dimension of ``C``. Must be positive and at least - ``n``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Pointer to the output matrix, overwritten by the updated ``C`` - matrix. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-3-routines` diff --git a/docs/domains/blas/syrk.rst b/docs/domains/blas/syrk.rst deleted file mode 100644 index 74cf63af0..000000000 --- a/docs/domains/blas/syrk.rst +++ /dev/null @@ -1,296 +0,0 @@ -.. _onemkl_blas_syrk: - -syrk -==== - -Performs a symmetric rank-k update. - -.. _onemkl_blas_syrk_description: - -.. rubric:: Description - -The ``syrk`` routines perform a rank-k update of a symmetric matrix ``C`` -by a general matrix ``A``. The operation is defined as: - -.. math:: - - C \leftarrow alpha*op(A)*op(A)^T + beta*C - -where: - -op(``X``) is one of op(``X``) = ``X`` or op(``X``) = ``X``\ :sup:`T` -, - -``alpha`` and ``beta`` are scalars, - -``C`` is a symmetric matrix and ``A``\ is a general matrix. - -Here op(``A``) is ``n``-by-``k``, and ``C`` is ``n``-by-``n``. - -``syrk`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_syrk_buffer: - -syrk (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void syrk(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - T beta, - sycl::buffer &c, - std::int64_t ldc) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void syrk(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - T beta, - sycl::buffer &c, - std::int64_t ldc) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to ``A`` (See :ref:`onemkl_datatypes` for more details). Conjugation is never performed, even if ``trans`` = ``transpose::conjtrans``. - - n - Number of rows and columns in ``C``. The value of ``n`` must be at - least zero. - - k - Number of columns in op(``A``).The value of ``k`` must be at least - zero. - - alpha - Scaling factor for the rank-k update. - - a - Buffer holding input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n`` - * - Row major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - See :ref:`matrix-storage` for - more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - beta - Scaling factor for matrix ``C``. - - c - Buffer holding input/output matrix ``C``. Must have size at least - ``ldc``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - ldc - Leading dimension of ``C``. Must be positive and at least ``n``. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output buffer, overwritten by - ``alpha``\ \*op(``A``)*op(``A``)\ :sup:`T` + ``beta``\ \*\ ``C``. - - -.. _onemkl_blas_syrk_usm: - -syrk (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event syrk(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - const T* a, - std::int64_t lda, - T beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event syrk(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - const T* a, - std::int64_t lda, - T beta, - T* c, - std::int64_t ldc, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A``'s data is stored in its upper or lower - triangle. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to - ``A`` (See :ref:`onemkl_datatypes` for more details). Conjugation is never performed, even if - ``trans`` = ``transpose::conjtrans``. - - n - Number of rows and columns in ``C``. The value of ``n`` must be - at least zero. - - k - Number of columns in op(``A``). The value of ``k`` must be at - least zero. - - alpha - Scaling factor for the rank-k update. - - a - Pointer to input matrix ``A``. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n`` - * - Row major - - ``A`` is an ``n``-by-``k`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``n``. - - ``A`` is an ``k``-by-``n`` matrix so the array ``a`` - must have size at least ``lda``\ \*\ ``k``. - - See :ref:`matrix-storage` for more details. - - lda - The leading dimension of ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``trans`` = ``transpose::nontrans`` - - ``trans`` = ``transpose::trans`` or ``transpose::conjtrans`` - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - beta - Scaling factor for matrix ``C``. - - c - Pointer to input/output matrix ``C``. Must have size at least - ``ldc``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - ldc - Leading dimension of ``C``. Must be positive and at least - ``n``. - -.. container:: section - - .. rubric:: Output Parameters - - c - Pointer to the output matrix, overwritten by - ``alpha``\ \*op(``A``)*op(``A``)\ :sup:`T` + - ``beta``\ \*\ ``C``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-3-routines` diff --git a/docs/domains/blas/syrk_batch.rst b/docs/domains/blas/syrk_batch.rst deleted file mode 100644 index b9782041e..000000000 --- a/docs/domains/blas/syrk_batch.rst +++ /dev/null @@ -1,484 +0,0 @@ -.. _onemkl_blas_syrk_batch: - -syrk_batch -========== - -Computes a group of ``syrk`` operations. - -.. _onemkl_blas_syrk_batch_description: - -.. rubric:: Description - -The ``syrk_batch`` routines are batched versions of :ref:`onemkl_blas_syrk`, performing -multiple ``syrk`` operations in a single call. Each ``syrk`` -operation perform a rank-k update with general matrices. - -``syrk_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_syrk_batch_buffer: - -syrk_batch (Buffer Version) ---------------------------- - -.. rubric:: Description - -The buffer version of ``syrk_batch`` supports only the strided API. - -The strided API operation is defined as: -:: - - for i = 0 … batch_size – 1 - A and C are matrices at offset i * stridea, i * stridec in a and c. - C := alpha * op(A) * op(A)^T + beta * C - end for - -where: - -op(X) is one of op(X) = X, or op(X) = X\ :sup:`T`, or op(X) = X\ :sup:`H`, - -``alpha`` and ``beta`` are scalars, - -``A`` and ``C`` are matrices, - -op(``A``) is ``n`` x ``k`` and ``C`` is ``n`` x ``n``. - -The ``a`` and ``c`` buffers contain all the input matrices. The stride -between matrices is given by the stride parameter. The total number -of matrices in ``a`` and ``c`` buffers is given by the ``batch_size`` parameter. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void syrk_batch(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - std::int64_t stridea, - T beta, - sycl::buffer &c, - std::int64_t ldc, - std::int64_t stridec, - std::int64_t batch_size) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void syrk_batch(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - sycl::buffer &a, - std::int64_t lda, - std::int64_t stridea, - T beta, - sycl::buffer &c, - std::int64_t ldc, - std::int64_t stridec, - std::int64_t batch_size) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether data in ``C`` is stored in its upper or lower triangle. - For more details, see :ref:`onemkl_datatypes`. - - trans - Specifies op(``A``) the transposition operation applied to the - matrix ``A``. Conjugation is never performed, even if trans = - transpose::conjtrans. See :ref:`onemkl_datatypes` for more - details. - - n - Number of rows and columns of ``C``. - Must be at least zero. - - k - Number of columns of op(``A``). - Must be at least zero. - - alpha - Scaling factor for the rank-k update. - - a - Buffer holding the input matrices ``A`` with size ``stridea`` * ``batch_size``. - - lda - The leading dimension of the matrices ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - stridea - Stride between different ``A`` matrices. - - beta - Scaling factor for the matrices ``C``. - - c - Buffer holding input/output matrices ``C`` with size ``stridec`` * ``batch_size``. - - ldc - The leading dimension of the matrices ``C``. It must be positive - and at least ``n``. - - stridec - Stride between different ``C`` matrices. Must be at least - ``ldc`` * ``n``. - - batch_size - Specifies the number of rank-k update operations to perform. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output buffer, overwritten by ``batch_size`` rank-k update - operations of the form ``alpha`` * op(``A``)*op(``A``)^T + ``beta`` * ``C``. - - -.. _onemkl_blas_syrk_batch_usm: - -syrk_batch (USM Version) ---------------------------- - -.. rubric:: Description - -The USM version of ``syrk_batch`` supports the group API and strided API. - -The group API operation is defined as: -:: - - idx = 0 - for i = 0 … group_count – 1 - for j = 0 … group_size – 1 - A, B, and C are matrices in a[idx] and c[idx] - C := alpha[i] * op(A) * op(A)^T + beta[i] * C - idx = idx + 1 - end for - end for - -The strided API operation is defined as -:: - - for i = 0 … batch_size – 1 - A, B and C are matrices at offset i * stridea, i * stridec in a and c. - C := alpha * op(A) * op(A)^T + beta * C - end for - -where: - -op(X) is one of op(X) = X, or op(X) = X\ :sup:`T`, or op(X) = X\ :sup:`H`, - -``alpha`` and ``beta`` are scalars, - -``A`` and ``C`` are matrices, - -op(``A``) is ``n`` x ``k`` and ``C`` is ``n`` x ``n``. - - -For group API, ``a`` and ``c`` arrays contain the pointers for all the input matrices. -The total number of matrices in ``a`` and ``c`` are given by: - -.. math:: - - total\_batch\_count = \sum_{i=0}^{group\_count-1}group\_size[i] - -For strided API, ``a`` and ``c`` arrays contain all the input matrices. The total number of matrices -in ``a`` and ``c`` are given by the ``batch_size`` parameter. - -**Group API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event syrk_batch(sycl::queue &queue, - uplo *upper_lower, - transpose *trans, - std::int64_t *n, - std::int64_t *k, - T *alpha, - const T **a, - std::int64_t *lda, - T *beta, - T **c, - std::int64_t *ldc, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event syrk_batch(sycl::queue &queue, - uplo *upper_lower, - transpose *trans, - std::int64_t *n, - std::int64_t *k, - T *alpha, - const T **a, - std::int64_t *lda, - T *beta, - T **c, - std::int64_t *ldc, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Array of ``group_count`` ``onemkl::upper_lower`` - values. ``upper_lower[i]`` specifies whether data in C for every - matrix in group ``i`` is in upper or lower triangle. - - trans - Array of ``group_count`` ``onemkl::transpose`` values. ``trans[i]`` specifies the form of op(``A``) used in - the rank-k update in group ``i``. See :ref:`onemkl_datatypes` for more details. - - n - Array of ``group_count`` integers. ``n[i]`` specifies the - number of rows and columns of ``C`` for every matrix in group ``i``. All entries must be at least zero. - - k - Array of ``group_count`` integers. ``k[i]`` specifies the - number of columns of op(``A``) for every matrix in group ``i``. All entries must be at - least zero. - - alpha - Array of ``group_count`` scalar elements. ``alpha[i]`` specifies the scaling factor for every rank-k update in group ``i``. - - a - Array of pointers to input matrices ``A`` with size ``total_batch_count``. - - See :ref:`matrix-storage` for more details. - - lda - Array of ``group_count`` integers. ``lda[i]`` specifies the - leading dimension of ``A`` for every matrix in group ``i``. All - entries must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda[i]`` must be at least ``n[i]``. - - ``lda[i]`` must be at least ``k[i]``. - * - Row major - - ``lda[i]`` must be at least ``k[i]``. - - ``lda[i]`` must be at least ``n[i]``. - - beta - Array of ``group_count`` scalar elements. ``beta[i]`` specifies the scaling factor for matrix ``C`` - for every matrix in group ``i``. - - c - Array of pointers to input/output matrices ``C`` with size ``total_batch_count``. - - See :ref:`matrix-storage` for more details. - - ldc - Array of ``group_count`` integers. ``ldc[i]`` specifies the - leading dimension of ``C`` for every matrix in group ``i``. All - entries must be positive and ``ldc[i]`` must be at least ``n[i]``. - - group_count - Specifies the number of groups. Must be at least 0. - - group_size - Array of ``group_count`` integers. ``group_size[i]`` specifies the - number of rank-k update products in group ``i``. All entries must be at least 0. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Overwritten by the ``n[i]``-by-``n[i]`` matrix calculated by - (``alpha[i]`` * op(``A``)*op(``A``)^T + ``beta[i]`` * ``C``) for group ``i``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event syrk_batch(sycl::queue &queue, - uplo upper_lower, - transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - const T *a, - std::int64_t lda, - std::int64_t stride_a, - T beta, - T *c, - std::int64_t ldc, - std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event syrk_batch(sycl::queue &queue, - uplo upper_lower, - transpose trans, - std::int64_t n, - std::int64_t k, - T alpha, - const T *a, - std::int64_t lda, - std::int64_t stride_a, - T beta, - T *c, - std::int64_t ldc, - std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether data in ``C`` is stored in its upper or lower triangle. - For more details, see :ref:`onemkl_datatypes`. - - trans - Specifies op(``A``) the transposition operation applied to the - matrices ``A``. Conjugation is never performed, even if trans = - transpose::conjtrans. See :ref:`onemkl_datatypes` for more - details. - - n - Number of rows and columns of ``C``. - Must be at least zero. - - k - Number of columns of op(``A``). - Must be at least zero. - - alpha - Scaling factor for the rank-k updates. - - a - Pointer to input matrices ``A`` with size ``stridea`` * ``batch_size``. - - lda - The leading dimension of the matrices ``A``. It must be positive. - - .. list-table:: - :header-rows: 1 - - * - - - ``A`` not transposed - - ``A`` transposed - * - Column major - - ``lda`` must be at least ``n``. - - ``lda`` must be at least ``k``. - * - Row major - - ``lda`` must be at least ``k``. - - ``lda`` must be at least ``n``. - - stridea - Stride between different ``A`` matrices. - - beta - Scaling factor for the matrices ``C``. - - c - Pointer to input/output matrices ``C`` with size ``stridec`` * ``batch_size``. - - ldc - The leading dimension of the matrices ``C``. It must be positive - and at least ``n``. - - stridec - Stride between different ``C`` matrices. - - batch_size - Specifies the number of rank-k update operations to perform. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - c - Output matrices, overwritten by ``batch_size`` rank-k update - operations of the form ``alpha`` * op(``A``)*op(``A``)^T + ``beta`` * ``C``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-like-extensions` diff --git a/docs/domains/blas/tbmv.rst b/docs/domains/blas/tbmv.rst deleted file mode 100644 index 716013cb3..000000000 --- a/docs/domains/blas/tbmv.rst +++ /dev/null @@ -1,223 +0,0 @@ -.. _onemkl_blas_tbmv: - -tbmv -==== - -Computes a matrix-vector product using a triangular band matrix. - -.. _onemkl_blas_tbmv_description: - -.. rubric:: Description - -The ``tbmv`` routines compute a matrix-vector product with a triangular -band matrix. The operation is defined as: - -.. math:: - - x \leftarrow op(A)*x - -where: - -op(``A``) is one of op(``A``) = ``A``, or op(``A``) = -``A``\ :sup:`T`, or op(``A``) = ``A``\ :sup:`H`, - -``A`` is an ``n``-by-``n`` unit or non-unit, upper or lower -triangular band matrix, with (``k`` + 1) diagonals, - -``x`` is a vector of length ``n``. - -``tbmv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_tbmv_buffer: - -tbmv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void tbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void tbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Numbers of rows and columns of ``A``. Must be at least zero. - - k - Number of sub/super-diagonals of the matrix ``A``. Must be at - least zero. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least (``k`` + 1), - and positive. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - -.. container:: section - - .. rubric:: Output Parameters - - x - Buffer holding the updated vector ``x``. - - - -.. _onemkl_blas_tbmv_usm: - -tbmv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event tbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - const T *a, - std::int64_t lda, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event tbmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - const T *a, - std::int64_t lda, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to - ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Numbers of rows and columns of ``A``. Must be at least zero. - - k - Number of sub/super-diagonals of the matrix ``A``. Must be at - least zero. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least (``k`` + - 1), and positive. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - x - Pointer to the updated vector ``x``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/tbsv.rst b/docs/domains/blas/tbsv.rst deleted file mode 100644 index bce876115..000000000 --- a/docs/domains/blas/tbsv.rst +++ /dev/null @@ -1,225 +0,0 @@ -.. _onemkl_blas_tbsv: - -tbsv -==== - -Solves a system of linear equations whose coefficients are in a -triangular band matrix. - -.. _onemkl_blas_tbsv_description: - -.. rubric:: Description - -The ``tbsv`` routines solve a system of linear equations whose -coefficients are in a triangular band matrix. The operation is -defined as: - -.. math:: - - op(A)*x = b - -where: - -op(``A``) is one of op(``A``) = ``A``, or op(``A``) = -``A``\ :sup:`T`, or op(``A``) = ``A``\ :sup:`H`, - -``A`` is an ``n``-by-``n`` unit or non-unit, upper or lower -triangular band matrix, with (``k`` + 1) diagonals, - -``b`` and ``x`` are vectors of length ``n``. - -``tbsv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_tbsv_buffer: - -tbsv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void tbsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void tbsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - k - Number of sub/super-diagonals of the matrix ``A``. Must be at - least zero. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least (``k`` + 1), - and positive. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - -.. container:: section - - .. rubric:: Output Parameters - - x - Buffer holding the solution vector ``x``. - - - -.. _onemkl_blas_tbsv_usm: - -tbsv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event tbsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - const T *a, - std::int64_t lda, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event tbsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - const T *a, - std::int64_t lda, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to - ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Number of rows and columns of ``A``. Must be at least zero. - - k - Number of sub/super-diagonals of the matrix ``A``. Must be at - least zero. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least (``k`` + - 1), and positive. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - x - Pointer to the solution vector ``x``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/tpmv.rst b/docs/domains/blas/tpmv.rst deleted file mode 100644 index 736fbcf21..000000000 --- a/docs/domains/blas/tpmv.rst +++ /dev/null @@ -1,199 +0,0 @@ -.. _onemkl_blas_tpmv: - -tpmv -==== - -Computes a matrix-vector product using a triangular packed matrix. - -.. _onemkl_blas_tpmv_description: - -.. rubric:: Description - -The ``tpmv`` routines compute a matrix-vector product with a triangular -packed matrix. The operation is defined as: - -.. math:: - - x \leftarrow op(A)*x - -where: - -op(``A``) is one of op(``A``) = ``A``, or op(``A``) = -``A``\ :sup:`T`, or op(``A``) = ``A``\ :sup:`H`, - -``A`` is an ``n``-by-``n`` unit or non-unit, upper or lower -triangular band matrix, supplied in packed form, - -``x`` is a vector of length ``n``. - -``tpmv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_tpmv_buffer: - -tpmv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void tpmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - sycl::buffer &a, - sycl::buffer &x, - std::int64_t incx) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void tpmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - sycl::buffer &a, - sycl::buffer &x, - std::int64_t incx) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Numbers of rows and columns of ``A``. Must be at least zero. - - a - Buffer holding input matrix ``A``. Must have size at least - (``n``\ \*(``n``\ +1))/2. See :ref:`matrix-storage` for - more details. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - -.. container:: section - - .. rubric:: Output Parameters - - x - Buffer holding the updated vector ``x``. - - -.. _onemkl_blas_tpmv_usm: - -tpmv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event tpmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - const T *a, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event tpmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - const T *a, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to - ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Numbers of rows and columns of ``A``. Must be at least zero. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least (``n``\ \*(``n``\ +1))/2. See - :ref:`matrix-storage` for - more details. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - x - Pointer to the updated vector ``x``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/tpsv.rst b/docs/domains/blas/tpsv.rst deleted file mode 100644 index 14082a077..000000000 --- a/docs/domains/blas/tpsv.rst +++ /dev/null @@ -1,207 +0,0 @@ -.. _onemkl_blas_tpsv: - -tpsv -==== - -Solves a system of linear equations whose coefficients are in a -triangular packed matrix. - -.. _onemkl_blas_tpsv_description: - -.. rubric:: Description - -The ``tpsv`` routines solve a system of linear equations whose -coefficients are in a triangular packed matrix. The operation is -defined as: - -.. math:: - - op(A)*x = b - -where: - -op(``A``) is one of op(``A``) = ``A``, or op(``A``) = -``A``\ :sup:`T`, or op(``A``) = ``A``\ :sup:`H`, - -``A`` is an ``n``-by-``n`` unit or non-unit, upper or lower -triangular band matrix, supplied in packed form, - -``b`` and ``x`` are vectors of length ``n``. - -``tpsv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_tpsv_buffer: - -tpsv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void tpsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - sycl::buffer &a, - sycl::buffer &x, - std::int64_t incx) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void tpsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - sycl::buffer &a, - sycl::buffer &x, - std::int64_t incx) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Numbers of rows and columns of ``A``. Must be at least zero. - - a - Buffer holding input matrix ``A``. Must have size at least - (``n``\ \*(``n``\ +1))/2. See :ref:`matrix-storage` for - more details. - - x - Buffer holding the ``n``-element right-hand side vector ``b``. The - buffer must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - -.. container:: section - - .. rubric:: Output Parameters - - x - Buffer holding the solution vector ``x``. - - -.. _onemkl_blas_tpsv_usm: - -tpsv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event tpsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - const T *a, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event tpsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - const T *a, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to - ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Numbers of rows and columns of ``A``. Must be at least zero. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least (``n``\ \*(``n``\ +1))/2. See - :ref:`matrix-storage` for - more details. - - x - Pointer to the ``n``-element right-hand side vector ``b``. The - array holding the ``n``-element right-hand side vector ``b`` - must be of size at least (1 + (``n`` - 1)*abs(``incx``)). See - :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - x - Pointer to the solution vector ``x``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/trmm.rst b/docs/domains/blas/trmm.rst deleted file mode 100644 index 1a812d56c..000000000 --- a/docs/domains/blas/trmm.rst +++ /dev/null @@ -1,288 +0,0 @@ -.. _onemkl_blas_trmm: - -trmm -==== - -Computes a matrix-matrix product where one input matrix is triangular -and one input matrix is general. - -.. _onemkl_blas_trmm_description: - -.. rubric:: Description - -The ``trmm`` routines compute a scalar-matrix-matrix product where one of -the matrices in the multiplication is triangular. The argument -``left_right`` determines if the triangular matrix, ``A``, is on the -left of the multiplication (``left_right`` = ``side::left``) or on -the right (``left_right`` = ``side::right``). Depending on -``left_right``. The operation is defined as: - -.. math:: - - B \leftarrow alpha*op(A)*B - -or - -.. math:: - - B \leftarrow alpha*B*op(A) - -where: - -op(``A``) is one of op(``A``) = *A*, or op(``A``) = ``A``\ :sup:`T`, -or op(``A``) = ``A``\ :sup:`H`, - -``alpha`` is a scalar, - -``A`` is a triangular matrix, and ``B`` is a general matrix. - -Here ``B`` is ``m`` x ``n`` and ``A`` is either ``m`` x ``m`` or -``n`` x ``n``, depending on ``left_right``. - -``trmm`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_trmm_buffer: - -trmm (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void trmm(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void trmm(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies whether ``A`` is on the left side of the multiplication - (``side::left``) or on the right side (``side::right``). See :ref:`onemkl_datatypes` for more details. - - uplo - Specifies whether the matrix ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_diag - Specifies whether ``A`` is assumed to be unit triangular (all - diagonal elements are 1). See :ref:`onemkl_datatypes` for more details. - - m - Specifies the number of rows of ``B``. The value of ``m`` must be - at least zero. - - n - Specifies the number of columns of ``B``. The value of ``n`` must - be at least zero. - - alpha - Scaling factor for the matrix-matrix product. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``m`` if ``left_right`` = ``side::left``, or - ``lda``\ \*\ ``n`` if ``left_right`` = ``side::right``. See - :ref:`matrix-storage` for - more details. - - lda - Leading dimension of ``A``. Must be at least ``m`` if - ``left_right`` = ``side::left``, and at least ``n`` if - ``left_right`` = ``side::right``. Must be positive. - - b - Buffer holding input/output matrix ``B``. Must have size at - least ``ldb``\ \*\ ``n`` if column major layout is used to store - matrices or at least ``ldb``\ \*\ ``m`` if row major layout is - used to store matrices. See :ref:`matrix-storage` for more details. - - ldb - Leading dimension of ``B``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - -.. container:: section - - .. rubric:: Output Parameters - - b - Output buffer, overwritten by ``alpha``\ \*op(``A``)\*\ ``B`` or - ``alpha``\ \*\ ``B``\ \*op(``A``). - -.. container:: section - - .. rubric:: Notes - - If ``alpha`` = 0, matrix ``B`` is set to zero, and ``A`` and ``B`` do - not need to be initialized at entry. - - - -.. _onemkl_blas_trmm_usm: - -trmm (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event trmm(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - const T* a, - std::int64_t lda, - T* b, - std::int64_t ldb, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event trmm(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - const T* a, - std::int64_t lda, - T* b, - std::int64_t ldb, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies whether ``A`` is on the left side of the - multiplication (``side::left``) or on the right side - (``side::right``). See :ref:`onemkl_datatypes` for more details. - - uplo - Specifies whether the matrix ``A`` is upper or lower - triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to - ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_diag - Specifies whether ``A`` is assumed to be unit triangular (all - diagonal elements are 1). See :ref:`onemkl_datatypes` for more details. - - m - Specifies the number of rows of ``B``. The value of ``m`` must - be at least zero. - - n - Specifies the number of columns of ``B``. The value of ``n`` - must be at least zero. - - alpha - Scaling factor for the matrix-matrix product. - - a - Pointer to input matrix ``A``. Must have size at least - ``lda``\ \*\ ``m`` if ``left_right`` = ``side::left``, or - ``lda``\ \*\ ``n`` if ``left_right`` = ``side::right``. See - :ref:`matrix-storage` for - more details. - - lda - Leading dimension of ``A``. Must be at least ``m`` if - ``left_right`` = ``side::left``, and at least ``n`` if - ``left_right`` = ``side::right``. Must be positive. - - b - Pointer to input/output matrix ``B``. Must have size at - least ``ldb``\ \*\ ``n`` if column major layout is used to store - matrices or at least ``ldb``\ \*\ ``m`` if row major layout is - used to store matrices. See :ref:`matrix-storage` for more details. - - ldb - Leading dimension of ``B``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - b - Pointer to the output matrix, overwritten by - ``alpha``\ \*op(``A``)\*\ ``B`` or - ``alpha``\ \*\ ``B``\ \*op(``A``). - -.. container:: section - - .. rubric:: Notes - - If ``alpha`` = 0, matrix ``B`` is set to zero, and ``A`` and ``B`` - do not need to be initialized at entry. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-3-routines` diff --git a/docs/domains/blas/trmv.rst b/docs/domains/blas/trmv.rst deleted file mode 100644 index d779c12a9..000000000 --- a/docs/domains/blas/trmv.rst +++ /dev/null @@ -1,210 +0,0 @@ -.. _onemkl_blas_trmv: - -trmv -==== - -Computes a matrix-vector product using a triangular matrix. - -.. _onemkl_blas_trmv_description: - -.. rubric:: Description - -The ``trmv`` routines compute a matrix-vector product with a triangular -matrix. The operation is defined as: - -.. math:: - - x \leftarrow op(A)*x - -where: - -op(``A``) is one of op(``A``) = ``A``, or op(``A``) = -``A``\ :sup:`T`, or op(``A``) = ``A``\ :sup:`H`, - -``A`` is an ``n``-by-``n`` unit or non-unit, upper or lower -triangular band matrix, - -``x`` is a vector of length ``n``. - -``trmv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_trmv_buffer: - -trmv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void trmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void trmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Numbers of rows and columns of ``A``. Must be at least zero. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - - x - Buffer holding input vector ``x``. The buffer must be of size at - least (1 + (``n`` - 1)*abs(``incx``)). See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - -.. container:: section - - .. rubric:: Output Parameters - - x - Buffer holding the updated vector ``x``. - - -.. _onemkl_blas_trmv_usm: - -trmv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event trmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - const T *a, - std::int64_t lda, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event trmv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - const T *a, - std::int64_t lda, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to - ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Numbers of rows and columns of ``A``. Must be at least zero. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - - x - Pointer to input vector ``x``. The array holding input vector - ``x`` must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for - more details. - - incx - Stride of vector ``x``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - x - Pointer to the updated vector ``x``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/blas/trsm.rst b/docs/domains/blas/trsm.rst deleted file mode 100644 index 0185e69d7..000000000 --- a/docs/domains/blas/trsm.rst +++ /dev/null @@ -1,286 +0,0 @@ -.. _onemkl_blas_trsm: - -trsm -==== - -Solves a triangular matrix equation (forward or backward solve). - -.. _onemkl_blas_trsm_description: - -.. rubric:: Description - -The ``trsm`` routines solve one of the following matrix equations: - -.. math:: - - op(A)*X = alpha*B - -or - -.. math:: - - X*op(A) = alpha*B - -where: - -op(``A``) is one of op(``A``) = ``A``, or op(``A``) = -``A``\ :sup:`T`, or op(``A``) = ``A``\ :sup:`H`, - -``alpha`` is a scalar, - -``A`` is a triangular matrix, and - -``B`` and ``X`` are ``m`` x ``n`` general matrices. - -``A`` is either ``m`` x ``m`` or ``n`` x ``n``, depending on whether -it multiplies ``X`` on the left or right. On return, the matrix ``B`` -is overwritten by the solution matrix ``X``. - -``trsm`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_trsm_buffer: - -trsm (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void trsm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void trsm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies whether ``A`` multiplies ``X`` on the left - (``side::left``) or on the right (``side::right``). See :ref:`onemkl_datatypes` for more details. - - uplo - Specifies whether the matrix ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_diag - Specifies whether ``A`` is assumed to be unit triangular (all - diagonal elements are 1). See :ref:`onemkl_datatypes` for more details. - - m - Specifies the number of rows of ``B``. The value of ``m`` must be - at least zero. - - n - Specifies the number of columns of ``B``. The value of ``n`` must - be at least zero. - - alpha - Scaling factor for the solution. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``m`` if ``left_right`` = ``side::left``, or - ``lda``\ \*\ ``n`` if ``left_right`` = ``side::right``. See - :ref:`matrix-storage` for - more details. - - lda - Leading dimension of ``A``. Must be at least ``m`` if - ``left_right`` = ``side::left``, and at least ``n`` if - ``left_right`` = ``side::right``. Must be positive. - - b - Buffer holding input/output matrix ``B``. Must have size at - least ``ldb``\ \*\ ``n`` if column major layout is used to store - matrices or at least ``ldb``\ \*\ ``m`` if row major layout is - used to store matrices. See :ref:`matrix-storage` for more details. - - ldb - Leading dimension of ``B``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - -.. container:: section - - .. rubric:: Output Parameters - - b - Output buffer. Overwritten by the solution matrix ``X``. - -.. container:: section - - .. rubric:: Notes - - If ``alpha`` = 0, matrix ``B`` is set to zero, and ``A`` and ``B`` do - not need to be initialized at entry. - - - -.. _onemkl_blas_trsm_usm: - -trsm (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event trsm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - const T* a, - std::int64_t lda, - T* b, - std::int64_t ldb, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event trsm(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - onemkl::transpose transa, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - const T* a, - std::int64_t lda, - T* b, - std::int64_t ldb, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies whether ``A`` multiplies ``X`` on the left - (``side::left``) or on the right (``side::right``). See :ref:`onemkl_datatypes` for more details. - - uplo - Specifies whether the matrix ``A`` is upper or lower - triangular. See :ref:`onemkl_datatypes` for more details. - - transa - Specifies op(``A``), the transposition operation applied to - ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_diag - Specifies whether ``A`` is assumed to be unit triangular (all - diagonal elements are 1). See :ref:`onemkl_datatypes` for more details. - - m - Specifies the number of rows of ``B``. The value of ``m`` must - be at least zero. - - n - Specifies the number of columns of ``B``. The value of ``n`` - must be at least zero. - - alpha - Scaling factor for the solution. - - a - Pointer to input matrix ``A``. Must have size at least - ``lda``\ \*\ ``m`` if ``left_right`` = ``side::left``, or - ``lda``\ \*\ ``n`` if ``left_right`` = ``side::right``. See - :ref:`matrix-storage` for - more details. - - lda - Leading dimension of ``A``. Must be at least ``m`` if - ``left_right`` = ``side::left``, and at least ``n`` if - ``left_right`` = ``side::right``. Must be positive. - - b - Pointer to input/output matrix ``B``. Must have size at - least ``ldb``\ \*\ ``n`` if column major layout is used to store - matrices or at least ``ldb``\ \*\ ``m`` if row major layout is - used to store matrices. See :ref:`matrix-storage` for more details. - - ldb - Leading dimension of ``B``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - b - Pointer to the output matrix. Overwritten by the solution - matrix ``X``. - -.. container:: section - - .. rubric:: Notes - - If ``alpha`` = 0, matrix ``B`` is set to zero, and ``A`` and ``B`` - do not need to be initialized at entry. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-3-routines` diff --git a/docs/domains/blas/trsm_batch.rst b/docs/domains/blas/trsm_batch.rst deleted file mode 100644 index e68b68aa2..000000000 --- a/docs/domains/blas/trsm_batch.rst +++ /dev/null @@ -1,497 +0,0 @@ -.. _onemkl_blas_trsm_batch: - -trsm_batch -========== - -Computes a group of ``trsm`` operations. - -.. _onemkl_blas_trsm_batch_description: - -.. rubric:: Description - -The ``trsm_batch`` routines are batched versions of :ref:`onemkl_blas_trsm`, performing -multiple ``trsm`` operations in a single call. Each ``trsm`` -solves an equation of the form op(A) \* X = alpha \* B or X \* op(A) = alpha \* B. - -``trsm_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_trsm_batch_buffer: - -trsm_batch (Buffer Version) ---------------------------- - -.. rubric:: Description - -The buffer version of ``trsm_batch`` supports only the strided API. - -The strided API operation is defined as: -:: - - for i = 0 … batch_size – 1 - A and B are matrices at offset i * stridea and i * strideb in a and b. - if (left_right == onemkl::side::left) then - compute X such that op(A) * X = alpha * B - else - compute X such that X * op(A) = alpha * B - end if - B := X - end for - -where: - -op(``A``) is one of op(``A``) = ``A``, or op(A) = ``A``\ :sup:`T`, -or op(``A``) = ``A``\ :sup:`H`, - -``alpha`` is a scalar, - -``A`` is a triangular matrix, - -``B`` and ``X`` are ``m`` x ``n`` general matrices, - -``A`` is either ``m`` x ``m`` or ``n`` x ``n``,depending on whether -it multiplies ``X`` on the left or right. On return, the matrix ``B`` -is overwritten by the solution matrix ``X``. - -The ``a`` and ``b`` buffers contain all the input matrices. The stride -between matrices is given by the stride parameter. The total number -of matrices in ``a`` and ``b`` buffers are given by the ``batch_size`` parameter. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void trsm_batch(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - std::int64_t stridea, - sycl::buffer &b, - std::int64_t ldb, - std::int64_t strideb, - std::int64_t batch_size) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void trsm_batch(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - sycl::buffer &a, - std::int64_t lda, - std::int64_t stridea, - sycl::buffer &b, - std::int64_t ldb, - std::int64_t strideb, - std::int64_t batch_size) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies whether the matrices ``A`` multiply ``X`` on the left - (``side::left``) or on the right (``side::right``). See :ref:`onemkl_datatypes` for more details. - - upper_lower - Specifies whether the matrices ``A`` are upper or lower - triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to the - matrices ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_diag - Specifies whether the matrices ``A`` are assumed to be unit - triangular (all diagonal elements are 1). See :ref:`onemkl_datatypes` for more details. - - m - Number of rows of the ``B`` matrices. Must be at least zero. - - n - Number of columns of the ``B`` matrices. Must be at least zero. - - alpha - Scaling factor for the solutions. - - a - Buffer holding the input matrices ``A`` with size ``stridea`` * ``batch_size``. - - lda - Leading dimension of the matrices ``A``. Must be at least ``m`` if - ``left_right`` = ``side::left``, and at least ``n`` if ``left_right`` = - ``side::right``. Must be positive. - - stridea - Stride between different ``A`` matrices. - - b - Buffer holding the input matrices ``B`` with size ``strideb`` * ``batch_size``. - - ldb - Leading dimension of the matrices ``B``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - - strideb - Stride between different ``B`` matrices. - - batch_size - Specifies the number of triangular linear systems to solve. - -.. container:: section - - .. rubric:: Output Parameters - - b - Output buffer, overwritten by ``batch_size`` solution matrices - ``X``. - -.. container:: section - - .. rubric:: Notes - - If ``alpha`` = 0, matrix ``B`` is set to zero and the matrices ``A`` - and ``B`` do not need to be initialized before calling ``trsm_batch``. - - -.. rubric:: Description - -The USM version of ``trsm_batch`` supports the group API and strided API. - -The group API operation is defined as: -:: - - idx = 0 - for i = 0 … group_count – 1 - for j = 0 … group_size – 1 - A and B are matrices in a[idx] and b[idx] - if (left_right == onemkl::side::left) then - compute X such that op(A) * X = alpha[i] * B - else - compute X such that X * op(A) = alpha[i] * B - end if - B := X - idx = idx + 1 - end for - end for - - -The strided API operation is defined as: -:: - - for i = 0 … batch_size – 1 - A and B are matrices at offset i * stridea and i * strideb in a and b. - if (left_right == onemkl::side::left) then - compute X such that op(A) * X = alpha * B - else - compute X such that X * op(A) = alpha * B - end if - B := X - end for - - where: - -op(``A``) is one of op(``A``) = ``A``, or op(A) = ``A``\ :sup:`T`, -or op(``A``) = ``A``\ :sup:`H`, - -``alpha`` is a scalar, - -``A`` is a triangular matrix, - -``B`` and ``X`` are ``m`` x ``n`` general matrices, - -``A`` is either ``m`` x ``m`` or ``n`` x ``n``,depending on whether -it multiplies ``X`` on the left or right. On return, the matrix ``B`` -is overwritten by the solution matrix ``X``. - -For group API, ``a`` and ``b`` arrays contain the pointers for all the input matrices. -The total number of matrices in ``a`` and ``b`` are given by: - -.. math:: - - total\_batch\_count = \sum_{i=0}^{group\_count-1}group\_size[i] - -For strided API, ``a`` and ``b`` arrays contain all the input matrices. The total number of matrices -in ``a`` and ``b`` are given by the ``batch_size`` parameter. - -**Group API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event trsm_batch(sycl::queue &queue, - onemkl::side *left_right, - onemkl::uplo *upper_lower, - onemkl::transpose *trans, - onemkl::diag *unit_diag, - std::int64_t *m, - std::int64_t *n, - T *alpha, - const T **a, - std::int64_t *lda, - T **b, - std::int64_t *ldb, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event trsm_batch(sycl::queue &queue, - onemkl::side *left_right, - onemkl::uplo *upper_lower, - onemkl::transpose *trans, - onemkl::diag *unit_diag, - std::int64_t *m, - std::int64_t *n, - T *alpha, - const T **a, - std::int64_t *lda, - T **b, - std::int64_t *ldb, - std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Array of ``group_count`` ``onemkl::side`` values. ``left_right[i]`` specifies whether ``A`` multiplies - ``X`` on the left (``side::left``) or on the right - (``side::right``) for every ``trsm`` operation in group ``i``. See :ref:`onemkl_datatypes` for more details. - - upper_lower - Array of ``group_count`` ``onemkl::uplo`` values. ``upper_lower[i]`` specifies whether ``A`` is upper or lower - triangular for every matrix in group ``i``. See :ref:`onemkl_datatypes` for more details. - - trans - Array of ``group_count`` ``onemkl::transpose`` values. ``trans[i]`` specifies the form of op(``A``) used - for every ``trsm`` operation in group ``i``. See :ref:`onemkl_datatypes` for more details. - - unit_diag - Array of ``group_count`` ``onemkl::diag`` values. ``unit_diag[i]`` specifies whether ``A`` is assumed to - be unit triangular (all diagonal elements are 1) for every matrix in group ``i``. See :ref:`onemkl_datatypes` for more details. - - m - Array of ``group_count`` integers. ``m[i]`` specifies the - number of rows of ``B`` for every matrix in group ``i``. All entries must be at least zero. - - n - Array of ``group_count`` integers. ``n[i]`` specifies the - number of columns of ``B`` for every matrix in group ``i``. All entries must be at least zero. - - alpha - Array of ``group_count`` scalar elements. ``alpha[i]`` specifies the scaling factor in group ``i``. - - a - Array of pointers to input matrices ``A`` with size ``total_batch_count``. See :ref:`matrix-storage` for more details. - - lda - Array of ``group_count`` integers. ``lda[i]`` specifies the leading dimension of ``A`` for every matrix in group ``i``. - All entries must be at least ``m`` - if ``left_right`` is ``side::left``, and at least - ``n`` if ``left_right`` is ``side::right``. All entries must be positive. - - b - Array of pointers to input matrices ``B`` with size ``total_batch_count``. See :ref:`matrix-storage` for more details. - - ldb - Array of ``group_count`` integers. ``ldb[i]`` specifies the - leading dimension of ``B`` for every matrix in group ``i``. All - entries must be positive and at least ``m`` and positive if - column major layout is used to store matrices or at least ``n`` - if row major layout is used to store matrices. - - group_count - Specifies the number of groups. Must be at least 0. - - group_size - Array of ``group_count`` integers. ``group_size[i]`` specifies the - number of ``trsm`` operations in group ``i``. All entries must be at least 0. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - b - Output buffer, overwritten by the ``total_batch_count`` solution - matrices ``X``. - -.. container:: section - - .. rubric:: Notes - - If ``alpha`` = 0, matrix ``B`` is set to zero and the matrices ``A`` - and ``B`` do not need to be initialized before calling ``trsm_batch``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - -**Strided API** - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event trsm_batch(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - const T *a, - std::int64_t lda, - std::int64_t stridea, - T *b, - std::int64_t ldb, - std::int64_t strideb, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event trsm_batch(sycl::queue &queue, - onemkl::side left_right, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_diag, - std::int64_t m, - std::int64_t n, - T alpha, - const T *a, - std::int64_t lda, - std::int64_t stridea, - T *b, - std::int64_t ldb, - std::int64_t strideb, - std::int64_t batch_size, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - left_right - Specifies whether the matrices ``A`` multiply ``X`` on the left - (``side::left``) or on the right (``side::right``). See :ref:`onemkl_datatypes` for more details. - - upper_lower - Specifies whether the matrices ``A`` are upper or lower - triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to the - matrices ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_diag - Specifies whether the matrices ``A`` are assumed to be unit - triangular (all diagonal elements are 1). See :ref:`onemkl_datatypes` for more details. - - m - Number of rows of the ``B`` matrices. Must be at least zero. - - n - Number of columns of the ``B`` matrices. Must be at least zero. - - alpha - Scaling factor for the solutions. - - a - Pointer to input matrices ``A`` with size ``stridea`` * ``batch_size``. - - lda - Leading dimension of the matrices ``A``. Must be at least ``m`` if - ``left_right`` = ``side::left``, and at least ``n`` if ``left_right`` = - ``side::right``. Must be positive. - - stridea - Stride between different ``A`` matrices. - - b - Pointer to input matrices ``B`` with size ``strideb`` * ``batch_size``. - - ldb - Leading dimension of the matrices ``B``. It must be positive and at least - ``m`` if column major layout is used to store matrices or at - least ``n`` if row major layout is used to store matrices. - - strideb - Stride between different ``B`` matrices. - - batch_size - Specifies the number of triangular linear systems to solve. - -.. container:: section - - .. rubric:: Output Parameters - - b - Output matrices, overwritten by ``batch_size`` solution matrices - ``X``. - -.. container:: section - - .. rubric:: Notes - - If ``alpha`` = 0, matrix ``B`` is set to zero and the matrices ``A`` - and ``B`` do not need to be initialized before calling ``trsm_batch``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-like-extensions` diff --git a/docs/domains/blas/trsv.rst b/docs/domains/blas/trsv.rst deleted file mode 100644 index 1a30dad8d..000000000 --- a/docs/domains/blas/trsv.rst +++ /dev/null @@ -1,215 +0,0 @@ -.. _onemkl_blas_trsv: - -trsv -==== - -Solves a system of linear equations whose coefficients are in a -triangular matrix. - -.. _onemkl_blas_trsv_description: - -.. rubric:: Description - -The ``trsv`` routines compute a matrix-vector product with a triangular -band matrix. The operation is defined as: - -.. math:: - - op(A)*x = b - -where: - -op(``A``) is one of op(``A``) = ``A``, or op(``A``) = -``A``\ :sup:`T`, or op(``A``) = ``A``\ :sup:`H`, - -``A`` is an ``n``-by-``n`` unit or non-unit, upper or lower -triangular matrix, - -``b`` and ``x`` are vectors of length ``n``. - -``trsv`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_blas_trsv_buffer: - -trsv (Buffer Version) ---------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - void trsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - void trsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - sycl::buffer &a, - std::int64_t lda, - sycl::buffer &x, - std::int64_t incx) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Numbers of rows and columns of ``A``. Must be at least zero. - - a - Buffer holding input matrix ``A``. Must have size at least - ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - - x - Buffer holding the ``n``-element right-hand side vector ``b``. The - buffer must be of size at least (1 + (``n`` - 1)*abs(``incx``)). - See :ref:`matrix-storage` for more details. - - incx - Stride of vector ``x``. - -.. container:: section - - .. rubric:: Output Parameters - - x - Buffer holding the solution vector ``x``. - - - -.. _onemkl_blas_trsv_usm: - -trsv (USM Version) ------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::blas::column_major { - sycl::event trsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - const T *a, - std::int64_t lda, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } -.. code-block:: cpp - - namespace oneapi::mkl::blas::row_major { - sycl::event trsv(sycl::queue &queue, - onemkl::uplo upper_lower, - onemkl::transpose trans, - onemkl::diag unit_nonunit, - std::int64_t n, - std::int64_t k, - const T *a, - std::int64_t lda, - T *x, - std::int64_t incx, - const std::vector &dependencies = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - queue - The queue where the routine should be executed. - - upper_lower - Specifies whether ``A`` is upper or lower triangular. See :ref:`onemkl_datatypes` for more details. - - trans - Specifies op(``A``), the transposition operation applied to - ``A``. See :ref:`onemkl_datatypes` for more details. - - unit_nonunit - Specifies whether the matrix ``A`` is unit triangular or not. See :ref:`onemkl_datatypes` for more details. - - n - Numbers of rows and columns of ``A``. Must be at least zero. - - a - Pointer to input matrix ``A``. The array holding input matrix - ``A`` must have size at least ``lda``\ \*\ ``n``. See :ref:`matrix-storage` for - more details. - - lda - Leading dimension of matrix ``A``. Must be at least ``n``, and - positive. - - x - Pointer to the ``n``-element right-hand side vector ``b``. The - array holding the ``n``-element right-hand side vector ``b`` - must be of size at least (1 + (``n`` - 1)*abs(``incx``)). See - :ref:`matrix-storage` for more details. - - incx - Stride of vector ``x``. - - dependencies - List of events to wait for before starting computation, if any. - If omitted, defaults to no dependencies. - -.. container:: section - - .. rubric:: Output Parameters - - x - Pointer to the solution vector ``x``. - -.. container:: section - - .. rubric:: Return Values - - Output event to wait on to ensure computation is complete. - - - **Parent topic:** :ref:`blas-level-2-routines` diff --git a/docs/domains/dense_linear_algebra.rst b/docs/domains/dense_linear_algebra.rst deleted file mode 100644 index 6544b9074..000000000 --- a/docs/domains/dense_linear_algebra.rst +++ /dev/null @@ -1,19 +0,0 @@ -.. _onemkl_dense_linear_algebra: - -Dense Linear Algebra ---------------------- - -This section contains information about dense linear algebra routines: - -:ref:`matrix-storage` provides information about dense matrix and vector storage formats that are used by oneMKL :ref:`onemkl_blas` and :ref:`onemkl_lapack`. - -:ref:`onemkl_blas` provides vector, matrix-vector, and matrix-matrix routines for dense matrices and vector operations. - -:ref:`onemkl_lapack` provides more complex dense linear algebra routines, e.g., matrix factorization, solving dense systems of linear equations, least square problems, eigenvalue and singular value problems, and performing a number of related computational tasks. - -.. toctree:: - :hidden: - - matrix-storage.rst - blas/blas.rst - lapack/lapack.rst diff --git a/docs/domains/lapack/gebrd.rst b/docs/domains/lapack/gebrd.rst deleted file mode 100644 index 7e014f97f..000000000 --- a/docs/domains/lapack/gebrd.rst +++ /dev/null @@ -1,230 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_gebrd: - -gebrd -===== - -Reduces a general matrix to bidiagonal form. - -.. container:: section - - .. rubric:: Description - -``gebrd`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -The routine reduces a general :math:`m \times n` matrix :math:`A` to a -bidiagonal matrix :math:`B` by an orthogonal (unitary) transformation. - - -If :math:`m \ge n`, the reduction is given by :math:`A=QBP^H=\begin{pmatrix}B_1\\0\end{pmatrix}P^H=Q_1B_1P_H` - -where :math:`B_{1}` is an :math:`n \times n` upper diagonal matrix, -:math:`Q` and :math:`P` are orthogonal or, for a complex :math:`A`, unitary -matrices; :math:`Q_{1}` consists of the first :math:`n` columns of -:math:`Q`. - -If :math:`m < n`, the reduction is given by - -:math:`A = QBP^H = Q\begin{pmatrix}B_1\\0\end{pmatrix}P^H = Q_1B_1P_1^H`, - -where :math:`B_{1}` is an :math:`m \times m` lower diagonal matrix, -:math:`Q` and :math:`P` are orthogonal or, for a complex :math:`A`, unitary -matrices; :math:`P_{1}` consists of the first :math:`m` columns of -:math:`P`. - -The routine does not form the matrices :math:`Q` and :math:`P` explicitly, -but represents them as products of elementary reflectors. Routines -are provided to work with the matrices :math:`Q` and :math:`P` in this -representation: - -If the matrix :math:`A` is real, - -- to compute :math:`Q` and :math:`P` explicitly, call - :ref:`onemkl_lapack_orgbr`. - -If the matrix :math:`A` is complex, - -- to compute :math:`Q` and :math:`P` explicitly, call - :ref:`onemkl_lapack_ungbr` - -gebrd (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -a - The buffer :math:`a`, size (``lda,*``). The buffer ``a`` contains the - matrix :math:`A`. The second dimension of ``a`` must be at least - :math:`\max(1, m)`. - -lda - The leading dimension of :math:`a`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_gebrd_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - If :math:`m \ge n`, the diagonal and first super-diagonal of a are - overwritten by the upper bidiagonal matrix :math:`B`. The elements - below the diagonal, with the buffer tauq, represent the orthogonal - matrix :math:`Q` as a product of elementary reflectors, and the - elements above the first superdiagonal, with the buffer ``taup``, - represent the orthogonal matrix :math:`P` as a product of elementary - reflectors. - - If :math:`m &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -a - Pointer to matrix :math:`A`. The second dimension of ``a`` must be at least - :math:`\max(1, m)`. - -lda - The leading dimension of ``a``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type T. - Size should not be less than the value returned by :ref:`onemkl_lapack_gebrd_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - If :math:`m \ge n`, the diagonal and first super-diagonal of a are - overwritten by the upper bidiagonal matrix :math:`B`. The elements - below the diagonal, with the array tauq, represent the orthogonal - matrix :math:`Q` as a product of elementary reflectors, and the - elements above the first superdiagonal, with the array ``taup``, - represent the orthogonal matrix :math:`P` as a product of elementary - reflectors. - - If :math:`m`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_gebrd` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_gebrd` function will be performed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_gebrd` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/geqrf.rst b/docs/domains/lapack/geqrf.rst deleted file mode 100644 index 31a2c97ce..000000000 --- a/docs/domains/lapack/geqrf.rst +++ /dev/null @@ -1,157 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_geqrf: - -geqrf -===== - -Computes the QR factorization of a general :math:`m \times n` matrix. - -.. rubric:: Description - -``geqrf`` supports the following precisions: - -.. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -The routine forms the QR factorization of a general -:math:`m \times n` matrix :math:`A`. No pivoting is performed. - -The routine does not form the matrix :math:`Q` explicitly. Instead, :math:`Q` -is represented as a product of :math:`\min(m, n)` elementary -reflectors. Routines are provided to work with :math:`Q` in this -representation. - -geqrf (Buffer Version) ----------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in :math:`A` (:math:`0 \le n`). - -a - Buffer holding input matrix :math:`A`. Must have size at least - :math:`\text{lda} \cdot n`. - -lda - The leading dimension of :math:`A`; at least :math:`\max(1, m)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_geqrf_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Output buffer, overwritten by the factorization data as follows: - - The elements on and above the diagonal of the array contain the - :math:`\min(m,n) \times n` upper trapezoidal matrix :math:`R` (:math:`R` is upper - triangular if :math:`m \ge n`); the elements below the diagonal, with the - array tau, represent the orthogonal matrix :math:`Q` as a product of - :math:`\min(m,n)` elementary reflectors. - -tau - Output buffer, size at least :math:`\max(1, \min(m, n))`. Contains scalars - that define elementary reflectors for the matrix :math:`Q` in its - decomposition in a product of elementary reflectors. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -geqrf (USM Version) ----------------------- - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in :math:`A` (:math:`0 \le n`). - -a - Pointer to memory holding input matrix :math:`A`. Must have size at least - :math:`\text{lda} \cdot n`. - -lda - The leading dimension of :math:`A`; at least :math:`\max(1, m)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_geqrf_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by the factorization data as follows: - - The elements on and above the diagonal of the array contain the - :math:`\min(m,n) \times n` upper trapezoidal matrix :math:`R` (:math:`R` is upper - triangular if :math:`m \ge n`); the elements below the diagonal, with the - array tau, represent the orthogonal matrix :math:`Q` as a product of - :math:`\min(m,n)` elementary reflectors. - -tau - Array, size at least :math:`\max(1, \min(m, n))`. Contains scalars - that define elementary reflectors for the matrix :math:`Q` in its - decomposition in a product of elementary reflectors. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - - diff --git a/docs/domains/lapack/geqrf_batch.rst b/docs/domains/lapack/geqrf_batch.rst deleted file mode 100644 index 12581248c..000000000 --- a/docs/domains/lapack/geqrf_batch.rst +++ /dev/null @@ -1,239 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_geqrf_batch: - -geqrf_batch -=========== - -Computes the QR factorizations of a batch of general matrices. - -.. container:: section - - .. rubric:: Description - -``geqrf_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_lapack_geqrf_batch_buffer: - -geqrf_batch (Buffer Version) ----------------------------- - -.. container:: section - - .. rubric:: Description - -The buffer version of ``geqrf_batch`` supports only the strided API. - -**Strided API** - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Number of rows in matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in matrices :math:`A_i` (:math:`0 \le n`). - -a - Array holding input matrices :math:`A_i`. - -lda - Leading dimension of matrices :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -stride_tau - Stride between the beginnings of arrays :math:`\tau_i` inside the array ``tau``. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as the number of floating point elements of type ``T``. Size should not be less than the value returned by the Strided API of the :ref:`onemkl_lapack_geqrf_batch_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Factorization data as follows: The elements on and above the diagonal of :math:`A_i` contain the :math:`\min(m,n) \times n` upper trapezoidal matrices :math:`R_i` (:math:`R_i` is upper triangular if :math:`m \ge n`); the elements below the diagonal, with the array :math:`\tau_i`, contain the orthogonal matrix :math:`Q_i` as a product of :math:`\min(m,n)` elementary reflectors. - -tau - Array to store batch of :math:`\tau_i`, each of size :math:`\min(m,n)`, containing scalars that define elementary reflectors for the matrices :math:`Q_i` in its decomposition in a product of elementary reflectors. - -geqrf_batch (USM Version) -------------------------- - -.. container:: section - - .. rubric:: Description - -The USM version of ``geqrf_batch`` supports the group API and strided API. - -**Group API** - -The routine forms the :math:`Q_iR_i` factorizations of a general :math:`m \times n` matrices :math:`A_i`, :math:`i \in \{1...batch\_size\}`, where ``batch_size`` is the sum of all parameter group sizes as provided with ``group_sizes`` array. -No pivoting is performed during factorization. -The routine does not form the matrices :math:`Q_i` explicitly. Instead, :math:`Q_i` is represented as a product of :math:`\min(m,n)` elementary reflectors. Routines are provided to work with :math:`Q_i` in this representation. -The total number of problems to solve, ``batch_size``, is a sum of sizes of all of the groups of parameters as provided by ``group_sizes`` array. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, T **a, std::int64_t *lda, T **tau, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Array of ``group_count`` :math:`m_g` parameters. Each :math:`m_g` specifies the number of rows in matrices :math:`A_i` from array ``a``, belonging to group :math:`g`. - -n - Array of ``group_count`` :math:`n_g` parameters. - Each :math:`n_g` specifies the number of columns in matrices :math:`A_i` from array ``a``, belonging to group :math:`g`. - -a - Array of ``batch_size`` pointers to input matrices :math:`A_i`, each of size :math:`\text{lda}_g\cdot n_g` (:math:`g` is an index of group to which :math:`A_i` belongs) - -lda - Array of ``group_count`` :math:`\text{lda}_g`` parameters, each representing the leading dimensions of input matrices :math:`A_i` from array ``a``, belonging to group :math:`g`. - -group_count - Specifies the number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as the number of floating point elements of type ``T``. Size should not be less than the value returned by the Group API of the :ref:`onemkl_lapack_geqrf_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Factorization data as follows: The elements on and above the diagonal of :math:`A_i` contain the :math:`\min(m_g,n_g) \times n_g` upper trapezoidal matrices :math:`R_i` (:math:`R_i` is upper triangular if :math:`m_g \ge n_g`); the elements below the diagonal, with the array :math:`\tau_i`, contain the orthogonal matrix :math:`Q_i` as a product of :math:`\min(m_g,n_g)` elementary reflectors. Here :math:`g` is the index of the parameters group corresponding to the :math:`i`-th decomposition. - -tau - Array of pointers to store arrays :math:`\tau_i`, each of size :math:`\min(m_g,n_g)`, containing scalars that define elementary reflectors for the matrices :math:`Q_i` in its decomposition in a product of elementary reflectors. Here :math:`g` is the index of the parameters group corresponding to the :math:`i`-th decomposition. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Strided API** - -The routine forms the :math:`Q_iR_i` factorizations of general :math:`m \times n` matrices :math:`A_i`. No pivoting is performed. -The routine does not form the matrices :math:`Q_i` explicitly. Instead, :math:`Q_i` is represented as a product of :math:`\min(m,n)` elementary reflectors. Routines are provided to work with :math:`Q_i` in this representation. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, std::int64_t stride_a, T *tau, std::int64_t stride_tau, std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Number of rows in matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in matrices :math:`A_i` (:math:`0 \le n`). - -a - Array holding input matrices :math:`A_i`. - -lda - Leading dimensions of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -stride_tau - Stride between the beginnings of arrays :math:`\tau_i` inside the array ``tau``. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as the number of floating point elements of type ``T``. Size should not be less than the value returned by the Strided API of the :ref:`onemkl_lapack_geqrf_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Factorization data as follows: The elements on and above the diagonal of :math:`A_i` contain the :math:`\min(m,n) \times n` upper trapezoidal matrices :math:`R_i` (:math:`R_i` is upper triangular if :math:`m \ge n`); the elements below the diagonal, with the array :math:`\tau_i`, contain the orthogonal matrix :math:`Q_i` as a product of :math:`\min(m,n)` elementary reflectors. - -tau - Array to store batch of :math:`\tau_i`, each of size :math:`\min(m,n)`, containing scalars that define elementary reflectors for the matrices :math:`Q_i` in its decomposition in a product of elementary reflectors. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` diff --git a/docs/domains/lapack/geqrf_batch_scratchpad_size.rst b/docs/domains/lapack/geqrf_batch_scratchpad_size.rst deleted file mode 100644 index bea681f3d..000000000 --- a/docs/domains/lapack/geqrf_batch_scratchpad_size.rst +++ /dev/null @@ -1,111 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_geqrf_batch_scratchpad_size: - -geqrf_batch_scratchpad_size -=========================== - -Computes size of scratchpad memory required for the :ref:`onemkl_lapack_geqrf_batch` function. - -.. rubric:: Description - -``geqrf_batch_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -**Group API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_geqrf_batch` function. - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. -m - | Array of ``group_count`` :math:`m_g` parameters. - | Each of :math:`m_g` specifies the number of rows in the matrices :math:`A_i` belonging to group :math:`g`. - -n - | Array of ``group_count`` :math:`n_g` parameters. - | Each of :math:`n_g` specifies the number of columns in the matrices :math:`A_i` belonging to group :math:`g`. - -lda - Array of ``group_count`` :math:`lda_g` parameters, each representing the leading dimensions of input matrices belonging to group :math:`g`. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_geqrf_batch` function. - -**Strided API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_geqrf_batch` function. - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Number of rows in the matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in :math:`A_i` (:math:`0 \le n`). - -lda - Leading dimension of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -stride_tau - Stride between the beginnings of arrays :math:`\tau_i` inside the array ``tau``. - -batch_size - Number of problems in a batch. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_geqrf_batch` function. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` diff --git a/docs/domains/lapack/geqrf_scratchpad_size.rst b/docs/domains/lapack/geqrf_scratchpad_size.rst deleted file mode 100644 index 8541bc724..000000000 --- a/docs/domains/lapack/geqrf_scratchpad_size.rst +++ /dev/null @@ -1,64 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_geqrf_scratchpad_size: - -geqrf_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_geqrf` function. - -.. container:: section - - .. rubric:: Description - -``geqrf_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_geqrf` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_geqrf` function will be performed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_geqrf` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/gerqf.rst b/docs/domains/lapack/gerqf.rst deleted file mode 100644 index 7f072eba3..000000000 --- a/docs/domains/lapack/gerqf.rst +++ /dev/null @@ -1,148 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_gerqf: - -gerqf -===== - -Computes the RQ factorization of a general :math:`m \times n` matrix. - -.. container:: section - - .. rubric:: Description - -``gerqf`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -The routine forms the RQ factorization of a general :math:`m \times n` matrix :math:`A`. No pivoting is performed. -The routine does not form the matrix :math:`Q` explicitly. Instead, :math:`Q` is represented as a product of :math:`\min(m, n)` elementary reflectors. Routines are provided to work with :math:`Q` in this representation - -gerqf (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -a - Buffer holding input matrix :math:`A`. The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``, at least :math:`\max(1, m)`. - -scratchpad - Buffer holding scratchpad memory to be used by the routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less than the value returned by the :ref:`onemkl_lapack_gerqf_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Output buffer, overwritten by the factorization data as follows: - - If :math:`m \le n`, the upper triangle of the subarray ``a(1:m, n-m+1:n)`` contains the :math:`m \times m` upper triangular matrix :math:`R`; if :math:`m \ge n`, the elements on and above the :math:`(m-n)`-th subdiagonal contain the :math:`m \times n` upper trapezoidal matrix :math:`R` - - In both cases, the remaining elements, with the array ``tau``, represent the orthogonal/unitary matrix :math:`Q` as a product of :math:`\min(m,n)` elementary reflectors. - -tau - Array, size at least :math:`\min(m,n)`. - - Contains scalars that define elementary reflectors for the matrix :math:`Q` in its decomposition in a product of elementary reflectors. - -gerqf (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -a - Buffer holding input matrix :math:`A`. The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``, at least :math:`\max(1, m)`. - -scratchpad - Buffer holding scratchpad memory to be used by the routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less than the value returned by the :ref:`onemkl_lapack_gerqf_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Output buffer, overwritten by the factorization data as follows: - - If :math:`m \le n`, the upper triangle of the subarray ``a(1:m, n-m+1:n)`` contains the :math:`m \times m` upper triangular matrix :math:`R`; if :math:`m \ge n`, the elements on and above the :math:`(m-n)`-th subdiagonal contain the :math:`m \times n` upper trapezoidal matrix :math:`R` - - In both cases, the remaining elements, with the array ``tau``, represent the orthogonal/unitary matrix :math:`Q` as a product of :math:`\min(m,n)` elementary reflectors. - -tau - Array, size at least :math:`\min(m,n)`. - - Contains scalars that define elementary reflectors for the matrix :math:`Q` in its decomposition in a product of elementary reflectors. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/gerqf_scratchpad_size.rst b/docs/domains/lapack/gerqf_scratchpad_size.rst deleted file mode 100644 index f35d02ef6..000000000 --- a/docs/domains/lapack/gerqf_scratchpad_size.rst +++ /dev/null @@ -1,68 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_gerqf_scratchpad_size: - -gerqf_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_gerqf` function. - -.. container:: section - - .. rubric:: Description - -``gerqf_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_gerqf` function should be able to hold. -Calls to this routine must specify the template parameter -explicitly. - -gerqf_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by the gerqf (buffer or USM version) function will be performed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -lda - The leading dimension of ``a``; at least :math:`\max(1,m)`. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_gerqf` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/gesvd.rst b/docs/domains/lapack/gesvd.rst deleted file mode 100644 index f778d7a73..000000000 --- a/docs/domains/lapack/gesvd.rst +++ /dev/null @@ -1,344 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_gesvd: - -gesvd -===== - -Computes the singular value decomposition of a general rectangular matrix. - -.. container:: section - - .. rubric:: Description - -``gesvd`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_lapack_gesvd_batch_buffer: - -gesvd (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Description - -The routine computes the singular value decomposition (SVD) of a -real/complex :math:`m \times n` matrix :math:`A`, optionally computing the -left and/or right singular vectors. The SVD is written as - -:math:`A = U\Sigma V^T` for real routines - -:math:`A = U\Sigma V^H` for complex routines - -where :math:`\Sigma` is an :math:`m \times n` diagonal matrix, :math:`U` is an -:math:`m \times m` orthogonal/unitary matrix, and :math:`V` is an -:math:`n \times n` orthogonal/unitary matrix. The diagonal elements of :math:`\Sigma` -are the singular values of :math:`A`; they are real and non-negative, and -are returned in descending order. The first :math:`\min(m, n)` columns of -:math:`U` and :math:`V` are the left and right singular vectors of :math:`A`. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void gesvd(sycl::queue &queue, oneapi::mkl::job jobu, oneapi::mkl::job jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -jobu - Must be ``job::allvec``, ``job::somevec``, ``job::overwritevec``, - or ``job::novec``. Specifies options for computing all or part of - the matrix :math:`U`. - - If ``jobu = job::allvec``, all :math:`m` columns of :math:`U` are returned - in the buffer ``u``; - - if ``jobu = job::somevec``, the first :math:`\min(m, n)` columns of - :math:`U` (the left singular vectors) are returned in the buffer ``u``; - - if ``jobu = job::overwritevec``, the first :math:`\min(m, n)` columns - of :math:`U` (the left singular vectors) are overwritten on the buffer - a; - - if ``jobu = job::novec``, no columns of :math:`U` (no left singular - vectors) are computed. - -jobvt - Must be ``job::allvec, job::somevec``, ``job::overwritevec``, or - ``job::novec``. Specifies options for computing all or part of the - matrix :math:`V^T/V^H`. - - If ``jobvt = job::allvec``, all :math:`n` columns of :math:`V^T/V^H` are - returned in the buffer vt; - - if ``jobvt = job::somevec``, the first :math:`\min(m, n)` columns of - :math:`V^T/V^H` (the left singular vectors) are returned in the buffer - vt; - - if ``jobvt = job::overwritevec``, the first :math:`\min(m, n)` columns - of :math:`V^T/V^H` (the left singular vectors) are overwritten on the - buffer ``a``; - - if ``jobvt = job::novec``, no columns of :math:`V^T/V^H` (no left - singular vectors) are computed. - - ``jobvt`` and ``jobu`` cannot both be ``job::overwritevec``. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -a - The buffer ``a``, size ``(lda,*)``. The buffer ``a`` contains the - matrix :math:`A`. The second dimension of ``a`` must be at least - :math:`\max(1, m)`. - -lda - The leading dimension of ``a``. - -ldu - The leading dimension of ``u``. - -ldvt - The leading dimension of ``vt``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_gesvd_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - On exit, - - If ``jobu = job::overwritevec``, ``a`` is overwritten with the first - :math:`\min(m,n)` columns of :math:`U` (the left singular vectors stored - columnwise); - - If ``jobvt = job::overwritevec``, ``a`` is overwritten with the first - :math:`\min(m, n)` rows of :math:`V^{T}`/:math:`V^{H}` (the right - singular vectors stored rowwise); - - If ``jobu`` :math:`\ne` ``job::overwritevec`` and ``jobvt`` :math:`\ne` ``job::overwritevec``, - the contents of a are destroyed. - -s - Buffer containing the singular values, size at least - :math:`\max(1, \min(m,n))`. Contains the singular values of :math:`A` sorted - so that :math:`s(i) \ge s(i+1)`. - -u - Buffer containing :math:`U`; the second dimension of ``u`` must be at - least :math:`\max(1, m)` if ``jobu = job::allvec``, and at least - :math:`\max(1, \min(m, n))` if ``jobu = job::somevec``. - - If ``jobu = job::allvec``, ``u`` contains the :math:`m \times m` - orthogonal/unitary matrix :math:`U`. - - If ``jobu = job::somevec``, ``u`` contains the first :math:`\min(m, n)` - columns of :math:`U` (the left singular vectors stored column-wise). - - If ``jobu = job::novec`` or ``job::overwritevec``, ``u`` is not - referenced. - -vt - Buffer containing :math:`V^{T}`; the second dimension of ``vt`` must - be at least :math:`\max(1, n)`. - - If ``jobvt = job::allvec``, ``vt`` contains the :math:`n \times n` - orthogonal/unitary matrix :math:`V^{T}`/:math:`V^{H}`. - - If ``jobvt = job::somevec``, ``vt`` contains the first :math:`\min(m, n)` - rows of :math:`V^{T}`/:math:`V^{H}` (the right singular - vectors stored row-wise). - - If ``jobvt = job::novec`` or ``job::overwritevec``, ``vt`` is not - referenced. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -gesvd (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Description - -The routine computes the singular value decomposition (SVD) of a -real/complex :math:`m \times n` matrix :math:`A`, optionally computing the -left and/or right singular vectors. The SVD is written as - -:math:`A = U\Sigma V^T` for real routines - -:math:`A = U\Sigma V^H` for complex routines - -where :math:`\Sigma` is an :math:`m \times n` diagonal matrix, :math:`U` is an -:math:`m \times m` orthogonal/unitary matrix, and :math:`V` is an -:math:`n \times n` orthogonal/unitary matrix. The diagonal elements of :math:`\Sigma` -are the singular values of :math:`A`; they are real and non-negative, and -are returned in descending order. The first :math:`\min(m, n)` columns of -:math:`U` and :math:`V` are the left and right singular vectors of :math:`A`. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event gesvd(sycl::queue &queue, oneapi::mkl::job jobu, oneapi::mkl::job jobvt, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, RealT *s, T *u, std::int64_t ldu, T *vt, std::int64_t ldvt, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -jobu - Must be ``job::allvec``, ``job::somevec``, ``job::overwritevec``, - or ``job::novec``. Specifies options for computing all or part of - the matrix :math:`U`. - - If ``jobu = job::allvec``, all :math:`m` columns of :math:`U` are returned - in the array ``u``; - - if ``jobu = job::somevec``, the first :math:`\min(m, n)` columns of - :math:`U` (the left singular vectors) are returned in the array ``u``; - - if ``jobu = job::overwritevec``, the first :math:`\min(m, n)` columns - of :math:`U` (the left singular vectors) are overwritten on the array - a; - - if ``jobu = job::novec``, no columns of :math:`U` (no left singular - vectors) are computed. - -jobvt - Must be ``job::allvec, job::somevec``, ``job::overwritevec``, or - ``job::novec``. Specifies options for computing all or part of the - matrix :math:`V^T/V^H`. - - If ``jobvt = job::allvec``, all :math:`n` columns of :math:`V^T/V^H` are - returned in the array ``vt``; - - if ``jobvt = job::somevec``, the first :math:`\min(m, n)` columns of - :math:`V^T/V^H` (the left singular vectors) are returned in the array - vt; - - if ``jobvt = job::overwritevec``, the first :math:`\min(m, n)` columns - of :math:`V^T/V^H` (the left singular vectors) are overwritten on the - array ``a``; - - if ``jobvt = job::novec``, no columns of :math:`V^T/V^H` (no left - singular vectors) are computed. - - ``jobvt`` and ``jobu`` cannot both be ``job::overwritevec``. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -a - Pointer to array ``a``, size ``(lda,*)``, containing the - matrix :math:`A`. The second dimension of ``a`` must be at least - :math:`\max(1, m)`. - -lda - The leading dimension of ``a``. - -ldu - The leading dimension of ``u``. - -ldvt - The leading dimension of ``vt``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_gesvd_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - On exit, - - If ``jobu = job::overwritevec``, ``a`` is overwritten with the first - :math:`\min(m,n)` columns of :math:`U` (the left singular vectors stored - columnwise); - - If ``jobvt = job::overwritevec``, ``a`` is overwritten with the first - :math:`\min(m, n)` rows of :math:`V^{T}`/:math:`V^{H}` (the right - singular vectors stored rowwise); - - If ``jobu`` :math:`\ne` ``job::overwritevec`` and ``jobvt`` :math:`\ne` ``job::overwritevec``, - the contents of a are destroyed. - -s - Array containing the singular values, size at least - :math:`\max(1, \min(m,n))`. Contains the singular values of :math:`A` sorted - so that :math:`s(i) \ge s(i+1)`. - -u - Array containing :math:`U`; the second dimension of ``u`` must be at - least :math:`\max(1, m)` if ``jobu = job::allvec``, and at least - :math:`\max(1, \min(m, n))` if ``jobu = job::somevec``. - - If ``jobu = job::allvec``, ``u`` contains the :math:`m \times m` - orthogonal/unitary matrix :math:`U`. - - If ``jobu = job::somevec``, ``u`` contains the first :math:`\min(m, n)` - columns of :math:`U` (the left singular vectors stored column-wise). - - If ``jobu = job::novec`` or ``job::overwritevec``, ``u`` is not - referenced. - -vt - Array containing :math:`V^{T}`; the second dimension of ``vt`` must - be at least :math:`\max(1, n)`. - - If ``jobvt = job::allvec``, ``vt`` contains the :math:`n \times n` - orthogonal/unitary matrix :math:`V^{T}`/:math:`V^{H}`. - - If ``jobvt = job::somevec``, ``vt`` contains the first :math:`\min(m, n)` - rows of :math:`V^{T}`/:math:`V^{H}` (the right singular - vectors stored row-wise). - - If ``jobvt = job::novec`` or ``job::overwritevec``, ``vt`` is not - referenced. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` diff --git a/docs/domains/lapack/gesvd_scratchpad_size.rst b/docs/domains/lapack/gesvd_scratchpad_size.rst deleted file mode 100644 index fab43c61f..000000000 --- a/docs/domains/lapack/gesvd_scratchpad_size.rst +++ /dev/null @@ -1,111 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_gesvd_scratchpad_size: - -gesvd_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_gesvd` function. - -.. container:: section - - .. rubric:: Description - -``gesvd_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_gesvd` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -gesvd_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobu, oneapi::mkl::job jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_gesvd` function will be performed. - -jobu - Must be ``job::allvec``, ``job::somevec``, - ``job::overwritevec``, or ``job::novec``. Specifies options for - computing all or part of the matrix :math:`U`. - - If ``jobu = job::allvec``, all :math:`m` columns of :math:`U` are - returned in the buffer ``u``; - - if ``jobu = job::somevec``, the first :math:`\min(m, n)` columns of - :math:`U` (the left singular vectors) are returned in the buffer ``v``; - - if ``jobu = job::overwritevec``, the first :math:`\min(m, n)` - columns of :math:`U` (the left singular vectors) are overwritten on - the buffer ``a``; - - if ``jobu = job::novec``, no columns of :math:`U` (no left singular - vectors) are computed. - -jobvt - Must be ``job::allvec``, ``job::somevec``, - ``job::overwritevec``, or ``job::novec``. Specifies options for - computing all or part of the matrix :math:`V^T/V^H`. - - If ``jobvt = job::allvec``, all :math:`n` columns of :math:`V^T/V^H` are - returned in the buffer ``vt``; - - if ``jobvt = job::somevec``, the first :math:`\min(m, n)` columns of - :math:`V^T/V^H` (the left singular vectors) are returned in the - buffer ``vt``; - - if ``jobvt = job::overwritevec``, the first :math:`\min(m, n)` - columns of :math:`V^T/V^H` (the left singular vectors) are - overwritten on the buffer ``a``; - - if ``jobvt = job::novec``, no columns of :math:`V^T/V^H` (no left - singular vectors) are computed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -lda - The leading dimension of ``a``. - -ldu - The leading dimension of ``u``. - -ldvt - The leading dimension of ``vt``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_gesvd` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/getrf.rst b/docs/domains/lapack/getrf.rst deleted file mode 100644 index d23594843..000000000 --- a/docs/domains/lapack/getrf.rst +++ /dev/null @@ -1,144 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getrf: - -getrf -===== - -Computes the LU factorization of a general :math:`m \times n` matrix. - -.. container:: section - - .. rubric:: Description - -``getrf`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -The routine computes the LU factorization of a general -:math:`m \times n` matrix :math:`A` as :math:`A = PLU`, - -where :math:`P` is a permutation matrix, :math:`L` is lower triangular with -unit diagonal elements (lower trapezoidal if :math:`m > n`) and :math:`U` is -upper triangular (upper trapezoidal if :math:`m < n`). The routine uses -partial pivoting, with row interchanges. - -getrf (BUFFER Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in :math:`A` (:math:`0 \le n`). - -a - Buffer holding input matrix :math:`A`. The buffer a contains the matrix :math:`A`. The second dimension of a must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_getrf_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by :math:`L` and :math:`U`. The unit diagonal elements of :math:`L` are not stored. - -ipiv - Array, size at least :math:`\max(1,\min(m, n))`. Contains the pivot indices; for :math:`1 \le i \le \min(m, n)`, row :math:`i` was interchanged with row :math:`\text{ipiv}(i)`. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -getrf (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, std::int64_t *ipiv, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in :math:`A` (:math:`0 \le n`). - -a - Pointer to array holding input matrix :math:`A`. The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_getrf_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by :math:`L` and :math:`U`. The unit diagonal elements of :math:`L` are not stored. - -ipiv - Array, size at least :math:`\max(1,\min(m, n))`. Contains the pivot indices; for :math:`1 \le i \le \min(m, n)`, row :math:`i` was interchanged with row :math:`\text{ipiv}(i)`. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - - diff --git a/docs/domains/lapack/getrf_batch.rst b/docs/domains/lapack/getrf_batch.rst deleted file mode 100644 index 502707ee5..000000000 --- a/docs/domains/lapack/getrf_batch.rst +++ /dev/null @@ -1,226 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getrf_batch: - -getrf_batch -=========== - -Computes the LU factorizations of a batch of general matrices. - -.. rubric:: Description - -``getrf_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_lapack_getrf_batch_buffer: - -getrf_batch (Buffer Version) ----------------------------- - -.. rubric:: Description - -The buffer version of ``getrf_batch`` supports only the strided API. - -**Strided API** - -The routine computes the LU factorizations of general :math:`m \times n` matrices :math:`A_i` as :math:`A_i = P_iL_iU_i`, where :math:`P_i` is a permutation matrix, :math:`L_i` is lower triangular with unit diagonal elements (lower trapezoidal if :math:`m > n`) and :math:`U_i` is upper triangular (upper trapezoidal if :math:`m < n`). The routine uses partial pivoting, with row interchanges. - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Number of rows in matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in matrices :math:`A_i` (:math:`0 \le n`). - -a - Array holding input matrices :math:`A_i`. - -lda - Leading dimension of matrices :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -stride_ipiv - Stride between the beginnings of arrays :math:`ipiv_i` inside the array ``ipiv``. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less than the value returned by the Strided API of the :ref:`onemkl_lapack_getrf_batch_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - :math:`L_i` and :math:`U_i`. The unit diagonal elements of :math:`L_i` are not stored. - -ipiv - Array containing batch of the pivot indices :math:`\text{ipiv}_i` each of size at least :math:`\max(1,\min(m,n))`; for :math:`1 \le k \le \min(m,n)`, where row :math:`k` of :math:`A_i` was interchanged with row :math:`\text{ipiv}_i(k)`. - -.. _onemkl_lapack_getrf_batch_usm: - -getrf_batch (USM Version) -------------------------- - -.. rubric:: Description - -The USM version of ``getrf_batch`` supports the group API and strided API. - -**Group API** - -The routine computes the batch of LU factorizations of general :math:`m \times n` matrices :math:`A_i` (:math:`i \in \{1...batch\_size\}`) as :math:`A_i = P_iL_iU_i`, where :math:`P_i` is a permutation matrix, :math:`L_i` is lower triangular with unit diagonal elements (lower trapezoidal if :math:`m > n`) and :math:`U_i` is upper triangular (upper trapezoidal if :math:`m < n`). The routine uses partial pivoting, with row interchanges. Total number of problems to solve, ``batch_size``, is a sum of sizes of all of the groups of parameters as provided by ``group_sizes`` array. - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, T **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Array of ``group_count`` parameters :math:`m_g` specifying the number of rows in matrices :math:`A_i` (:math:`0 \le m_g`) belonging to group :math:`g`. - -n - Array of ``group_count`` parameters :math:`n_g` specifying the number of columns in matrices :math:`A_i` (:math:`0 \le n_g`) belonging to group :math:`g`. - -a - Array holding ``batch_size`` pointers to input matrices :math:`A_i`. - -lda - Array of ``group_count`` parameters :math:`lda_g` specifying the leading dimensions of :math:`A_i` belonging to group :math:`g`. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of group_count integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Group API of the :ref:`onemkl_lapack_getrf_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - :math:`L_i` and :math:`U_i`. The unit diagonal elements of :math:`L_i` are not stored. - -ipiv - Arrays of batch_size pointers to arrays containing pivot indices :math:`\text{ipiv}_i` each of size at least :math:`\max(1,\min(m_g,n_g))`; for :math:`1 \le k \le \min(m_g,n_g)`, where row :math:`k` of :math:`A_i` was interchanged with row :math:`\text{ipiv}_i(k)`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Strided API** - -The routine computes the LU factorizations of general :math:`m \times n` matrices :math:`A_i` as :math:`A_i = P_iL_iU_i`, where :math:`P_i` is a permutation matrix, :math:`L_i` is lower triangular with unit diagonal elements (lower trapezoidal if :math:`m > n`) and :math:`U_i` is upper triangular (upper trapezoidal if :math:`m < n`). The routine uses partial pivoting, with row interchanges. - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Number of rows in matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in matrices :math:`A_i` (:math:`0 \le n`). - -a - Array holding input matrices :math:`A_i`. - -lda - Leading dimension of matrices :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -stride_ipiv - Stride between the beginnings of arrays :math:`\text{ipiv}_i` inside the array ``ipiv``. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Strided API of the :ref:`onemkl_lapack_getrf_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - :math:`L_i` and :math:`U_i`. The unit diagonal elements of :math:`L_i` are not stored. - -ipiv - Array containing batch of the pivot indices :math:`\text{ipiv}_i` each of size at least :math:`\max(1,\min(m,n))`; for :math:`1 \le k \le \min(m,n)`, where row :math:`k` of :math:`A_i` was interchanged with row :math:`\text{ipiv}_i(k)`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` diff --git a/docs/domains/lapack/getrf_batch_scratchpad_size.rst b/docs/domains/lapack/getrf_batch_scratchpad_size.rst deleted file mode 100644 index 27cfdb4da..000000000 --- a/docs/domains/lapack/getrf_batch_scratchpad_size.rst +++ /dev/null @@ -1,117 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getrf_batch_scratchpad_size: - -getrf_batch_scratchpad_size -=========================== - -Computes size of scratchpad memory required for the :ref:`onemkl_lapack_getrf_batch` function. - -.. container:: section - - .. rubric:: Description - -``getrf_batch_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -**Group API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_getrf_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Array of ``group_count`` parameters :math:`m_g` specifying the number of rows in the matrices belonging to group :math:`g`. - -n - Array of ``group_count`` parameters :math:`n_g` specifying the number of columns in matrices belonging to group :math:`g`. - -lda - Array of ``group_count`` parameters :math:`\text{lda}_g` specifying the leading dimensions of matrices belonging to group :math:`g`. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_getrf_batch` function. - -**Strided API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_getrf_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Number of rows in the matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in :math:`A_i` (:math:`0 \le n`). - -lda - Leading dimension of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -stride_ipiv - Stride between the beginnings of arrays :math:`\text{ipiv}_i` inside the array ``ipiv``. - -batch_size - Number of problems in a batch. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_getrf_batch` function. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/getrf_scratchpad_size.rst b/docs/domains/lapack/getrf_scratchpad_size.rst deleted file mode 100644 index ae382d0fc..000000000 --- a/docs/domains/lapack/getrf_scratchpad_size.rst +++ /dev/null @@ -1,67 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getrf_scratchpad_size: - -getrf_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_getrf` function. - -.. container:: section - - .. rubric:: Description - -``getrf_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_getrf` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -getrf_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_getrf` function will be performed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in :math:`A` (:math:`0 \le n`). - -lda - The leading dimension of ``a`` :math:`(n \le \text{lda})`. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_getrf` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/getri.rst b/docs/domains/lapack/getri.rst deleted file mode 100644 index d0f1854e8..000000000 --- a/docs/domains/lapack/getri.rst +++ /dev/null @@ -1,138 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getri: - -getri -===== - -Computes the inverse of an LU-factored general matrix determined by -:ref:`onemkl_lapack_getrf`. - -.. container:: section - - .. rubric:: Description - -``getri`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -The routine computes the inverse :math:`A^{-1}` of a general matrix -:math:`A`. Before calling this routine, call :ref:`onemkl_lapack_getrf` -to factorize :math:`A`. - -getri (BUFFER Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -n - The order of the matrix :math:`A` :math:`(0 \le n)`. - -a - The buffer ``a`` as returned by :ref:`onemkl_lapack_getrf`. Must - be of size at least :math:`\text{lda} \cdot \max(1,n)`. - -lda - The leading dimension of ``a`` :math:`(n \le \text{lda})`. - -ipiv - The buffer as returned by :ref:`onemkl_lapack_getrf`. The - dimension of ``ipiv`` must be at least :math:`\max(1, n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_getri_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by the :math:`n \times n` matrix :math:`A`. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -getri (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event getri(sycl::queue &queue, std::int64_t n, T *a, std::int64_t lda, std::int64_t *ipiv, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -n - The order of the matrix :math:`A` :math:`(0 \le n)`. - -a - The array as returned by :ref:`onemkl_lapack_getrf`. Must - be of size at least :math:`\text{lda} \cdot \max(1,n)`. - -lda - The leading dimension of ``a`` :math:`(n \le \text{lda})`. - -ipiv - The array as returned by :ref:`onemkl_lapack_getrf`. The - dimension of ``ipiv`` must be at least :math:`\max(1, n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_getri_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by the :math:`n \times n` matrix :math:`A`. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/getri_batch.rst b/docs/domains/lapack/getri_batch.rst deleted file mode 100644 index 9112100a7..000000000 --- a/docs/domains/lapack/getri_batch.rst +++ /dev/null @@ -1,229 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getri_batch: - -getri_batch -=========== - -Computes the inverses of a batch of LU-factored matrices determined by :ref:`onemkl_lapack_getrf_batch`. - -.. container:: section - - .. rubric:: Description - -``getri_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_lapack_getri_batch_buffer: - -getri_batch (Buffer Version) ----------------------------- - -.. container:: section - - .. rubric:: Description - -The buffer version of ``getri_batch`` supports only the strided API. - -**Strided API** - -The routine computes the inverses :math:`A_i^{-1}` of general matrices :math:`A_i`. Before calling this routine, call the Strided API of the :ref:`onemkl_lapack_getrf_batch_buffer` function to factorize :math:`A_i`. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -n - Order of the matrices :math:`A_i` (:math:`0 \le n`). - -a - Result of the Strided API of the :ref:`onemkl_lapack_getrf_batch_buffer` function. - -lda - Leading dimension of :math:`A_i` (:math:`n\le \text{lda}`). - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -ipiv - Arrays returned by the Strided API of the :ref:`onemkl_lapack_getrf_batch_buffer` function. - -stride_ipiv - Stride between the beginnings of arrays :math:`\text{ipiv}_i` inside the array ``ipiv``. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less than the value returned by the Strided API of the :ref:`onemkl_lapack_getri_batch_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Inverse :math:`n \times n` matrices :math:`A_i^{-1}`. - -getri_batch (USM Version) -------------------------- - -.. container:: section - - .. rubric:: Description - -The USM version of ``getri_batch`` supports the group API and strided API. - -**Group API** - -The routine computes the inverses :math:`A_i^{-1}` of general matrices :math:`A_i`, :math:`i \in \{1...batch\_size\}`. Before calling this routine, call the Group API of the :ref:`onemkl_lapack_getrf_batch_usm` function to factorize :math:`A_i`. -Total number of problems to solve, ``batch_size``, is a sum of sizes of all of the groups of parameters as provided by ``group_sizes`` array. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, T **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -n - Array of ``group_count`` :math:`n_g` parameters specifying the order of the matrices :math:`A_i` (:math:`0 \le n_g`) belonging to group :math:`g`. - -a - Result of the Group API of the :ref:`onemkl_lapack_getrf_batch_usm` function. - -lda - Array of ``group_count`` :math:`\text{lda}_g` parameters specifying the leading dimensions of the matrices :math:`A_i` (:math:`n_g \le \text{lda}_g`) belonging to group :math:`g`. - -ipiv - Arrays returned by the Group API of the :ref:`onemkl_lapack_getrf_batch_usm` function. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less than the value returned by the Group API of the :ref:`onemkl_lapack_getri_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Inverse :math:`n_g \times n_g` matrices :math:`A_i^{-1}`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Strided API** - -The routine computes the inverses :math:`A_i^{-1}` of general matrices :math:`A_i`. Before calling this routine, call the Strided API of the :ref:`onemkl_lapack_getrf_batch_usm` function to factorize :math:`A_i`. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event getri_batch(sycl::queue &queue, std::int64_t n, T *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -n - Order of the matrices :math:`A_i` (:math:`0 \le n`). - -a - Result of the Strided API of the :ref:`onemkl_lapack_getrf_batch_usm` function. - -lda - Leading dimension of :math:`A_i` (:math:`n \le \text{lda}`). - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -ipiv - Arrays returned by the Strided API of the :ref:`onemkl_lapack_getrf_batch_usm` function. - -stride_ipiv - Stride between the beginnings of arrays :math:`\text{ipiv}_i` inside the array ``ipiv``. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less than the value returned by the Strided API of the :ref:`onemkl_lapack_getri_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Inverse :math:`n \times n` matrices :math:`A_i^{-1}`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/getri_batch_scratchpad_size.rst b/docs/domains/lapack/getri_batch_scratchpad_size.rst deleted file mode 100644 index 3e30e4400..000000000 --- a/docs/domains/lapack/getri_batch_scratchpad_size.rst +++ /dev/null @@ -1,111 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getri_batch_scratchpad_size: - -getri_batch_scratchpad_size -=========================== - -Computed size of scratchpad memory required for the :ref:`onemkl_lapack_getri_batch` function. - -.. container:: section - - .. rubric:: Description - -``getri_batch_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -**Group API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_getri_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -n - Array of ``group_count`` :math:`n_g` parameters specifying the order of the matrices belonging to group :math:`g`. - -lda - Array of ``group_count`` :math:`\text{lda}_g` parameters specifying the leading dimensions of the matrices belonging to group :math:`g`. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_getri_batch` function. - -**Strided API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_getri_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -n - The order of the matrices :math:`A_i` (0 \le n). - -lda - Leading dimension of :math:`A_i` (:math:`n \le \text{lda}`). - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -stride_ipiv - Stride between the beginnings of arrays :math:`ipiv_i` inside the array ipiv. - -batch_size - Specifies the number of problems in a batch. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_getri_batch` function. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/getri_scratchpad_size.rst b/docs/domains/lapack/getri_scratchpad_size.rst deleted file mode 100644 index 734f51a69..000000000 --- a/docs/domains/lapack/getri_scratchpad_size.rst +++ /dev/null @@ -1,66 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getri_scratchpad_size: - -getri_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_getri` function. - -.. container:: section - - .. rubric:: Description - -``getri_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_getri` function should be able to hold. -Calls to this routine must specify the template parameter -explicitly. - -getri_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_getri` function will be performed. - -n - The order of the matrix :math:`A` :math:`(0 \le n)`. - -lda - The leading dimension of ``a`` :math:`(n \le \text{lda})`. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_getri` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines`  - - diff --git a/docs/domains/lapack/getrs.rst b/docs/domains/lapack/getrs.rst deleted file mode 100644 index f96ca356a..000000000 --- a/docs/domains/lapack/getrs.rst +++ /dev/null @@ -1,200 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getrs: - -getrs -===== - -Solves a system of linear equations with an LU-factored square -coefficient matrix, with multiple right-hand sides. - -.. container:: section - - .. rubric:: Description - -``getrs`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -The routine solves for :math:`X` the following systems of linear -equations: - - .. list-table:: - :header-rows: 1 - - * - \ :math:`AX = B`\ - - if ``trans``\ =\ ``oneapi::mkl::transpose::nontrans``\ - * - \ :math:`A^TX = B`\ - - if ``trans``\ =\ ``oneapi::mkl::transpose::trans``\ - * - \ :math:`A^HX = B`\ - - if ``trans``\ =\ ``oneapi::mkl::transpose::conjtrans``\ - -Before calling this routine, you must call -:ref:`onemkl_lapack_getrf` -to compute the LU factorization of :math:`A`. - -getrs (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -trans - Indicates the form of the equations: - - If ``trans=oneapi::mkl::transpose::nontrans``, then :math:`AX = B` is solved - for :math:`X`. - - If ``trans=oneapi::mkl::transpose::trans``, then :math:`A^TX = B` is solved - for :math:`X`. - - If ``trans=oneapi::mkl::transpose::conjtrans``, then :math:`A^HX = B` is - solved for :math:`X`. - -n - The order of the matrix :math:`A` and the number of rows in matrix - :math:`B(0 \le n)`. - -nrhs - The number of right-hand sides (:math:`0 \le \text{nrhs}`). - -a - Buffer containing the factorization of the matrix :math:`A`, as - returned by :ref:`onemkl_lapack_getrf`. The second dimension of ``a`` must be at least - :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -ipiv - Array, size at least :math:`\max(1, n)`. The ``ipiv`` array, as returned by - :ref:`onemkl_lapack_getrf`. - -b - The array ``b`` contains the matrix :math:`B` whose columns are the - right-hand sides for the systems of equations. The second - dimension of ``b`` must be at least :math:`\max(1,\text{nrhs})`. - -ldb - The leading dimension of ``b``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_getrs_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -b - The buffer ``b`` is overwritten by the solution matrix :math:`X`. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -getrs (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, T *a, std::int64_t lda, std::int64_t *ipiv, T *b, std::int64_t ldb, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -trans - Indicates the form of the equations: - - If ``trans=oneapi::mkl::transpose::nontrans``, then :math:`AX = B` is solved - for :math:`X`. - - If ``trans=oneapi::mkl::transpose::trans``, then :math:`A^TX = B` is solved - for :math:`X`. - - If ``trans=oneapi::mkl::transpose::conjtrans``, then :math:`A^HX = B` is - solved for :math:`X`. - -n - The order of the matrix :math:`A` and the number of rows in matrix - :math:`B(0 \le n)`. - -nrhs - The number of right-hand sides (:math:`0 \le \text{nrhs}`). - -a - Pointer to array containing the factorization of the matrix :math:`A`, as - returned by :ref:`onemkl_lapack_getrf`. The second dimension of ``a`` must be at least - :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -ipiv - Array, size at least :math:`\max(1, n)`. The ``ipiv`` array, as returned by - :ref:`onemkl_lapack_getrf`. - -b - The array ``b`` contains the matrix :math:`B` whose columns are the - right-hand sides for the systems of equations. The second - dimension of ``b`` must be at least :math:`\max(1,\text{nrhs})`. - -ldb - The leading dimension of ``b``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_getrs_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -b - The array ``b`` is overwritten by the solution matrix :math:`X`. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` diff --git a/docs/domains/lapack/getrs_batch.rst b/docs/domains/lapack/getrs_batch.rst deleted file mode 100644 index 4c23fb04d..000000000 --- a/docs/domains/lapack/getrs_batch.rst +++ /dev/null @@ -1,286 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getrs_batch: - -getrs_batch -=========== - -Solves a system of linear equations with a batch of LU-factored square coefficient matrices, with multiple right-hand sides. - -.. container:: section - - .. rubric:: Description - -``getrs_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_lapack_getrs_batch_buffer: - -getrs_batch (Buffer Version) ----------------------------- - -.. container:: section - - .. rubric:: Description - -The buffer version of ``getrs_batch`` supports only the strided API. - -**Strided API** - - | The routine solves for the following systems of linear equations :math:`X_i`: - | :math:`A_iX_i = B_i`, if ``trans=mkl::transpose::nontrans`` - | :math:`A_i^TX_i = B_i`, if ``trans=mkl::transpose::trans`` - | :math:`A_i^HX_i = B_i`, if ``trans=mkl::transpose::conjtrans`` - | Before calling this routine, the Strided API of the :ref:`onemkl_lapack_getrf_batch_buffer` function should be called to compute the LU factorizations of :math:`A_i`. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void getrs_batch(sycl::queue &queue, mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -trans - | Form of the equations: - | If ``trans = mkl::transpose::nontrans``, then :math:`A_iX_i = B_i` is solved for :math:`Xi`. - | If ``trans = mkl::transpose::trans``, then :math:`A_i^TX_i = B_i` is solved for :math:`X_i`. - | If ``trans = mkl::transpose::conjtrans``, then :math:`A_i^HX_i = B_i` is solved for :math:`X_i`. - -n - Order of the matrices :math:`A_i` and the number of rows in matrices :math:`B_i` (:math:`0 \le n`). - -nrhs - Number of right-hand sides (:math:`0 \le \text{nrhs}`). - -a - Array containing the factorizations of the matrices :math:`A_i`, as returned the Strided API of the :ref:`onemkl_lapack_getrf_batch_buffer` function. - -lda - Leading dimension of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`B_i` inside the batch array ``b``. - -ipiv - ``ipiv`` array, as returned by the Strided API of the :ref:`onemkl_lapack_getrf_batch_buffer` function. - -stride_ipiv - Stride between the beginnings of arrays :math:`\text{ipiv}_i` inside the array ``ipiv``. - -b - Array containing the matrices :math:`B_i` whose columns are the right-hand sides for the systems of equations. - -ldb - Leading dimension of :math:`B_i`. - -batch_size - Specifies the number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Strided API of the :ref:`onemkl_lapack_getrs_batch_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -b - Solution matrices :math:`X_i`. - -.. _onemkl_lapack_getrs_batch_usm: - -getrs_batch (USM Version) -------------------------- - -.. container:: section - - .. rubric:: Description - -The USM version of ``getrs_batch`` supports the group API and strided API. - -**Group API** - - | The routine solves the following systems of linear equations for :math:`X_i` (:math:`i \in \{1...batch\_size\}`): - | :math:`A_iX_i = B_i`, if ``trans=mkl::transpose::nontrans`` - | :math:`A_i^TX_i = B_i`, if ``trans=mkl::transpose::trans`` - | :math:`A_i^HX_i = B_i`, if ``trans=mkl::transpose::conjtrans`` - | Before calling this routine, call the Group API of the :ref:`onemkl_lapack_getrf_batch_usm` function to compute the LU factorizations of :math:`A_i`. - | Total number of problems to solve, ``batch_size``, is a sum of sizes of all of the groups of parameters as provided by ``group_sizes`` array. - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event getrs_batch(sycl::queue &queue, mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, T **a, std::int64_t *lda, std::int64_t **ipiv, T **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -trans - | Array of ``group_count`` parameters :math:`trans_g` indicating the form of the equations for the group :math:`g`: - | If ``trans = mkl::transpose::nontrans``, then :math:`A_iX_i = B_i` is solved for :math:`X_i`. - | If ``trans = mkl::transpose::trans``, then :math:`A_i^TX_i = B_i` is solved for :math:`X_i`. - | If ``trans = mkl::transpose::conjtrans``, then :math:`A_i^HX_i = B_i` is solved for :math:`X_i`. - -n - Array of ``group_count`` parameters :math:`n_g` specifying the order of the matrices :math:`A_i` and the number of rows in matrices :math:`B_i` (:math:`0 \le n_g`) belonging to group :math:`g`. - -nrhs - Array of ``group_count`` parameters :math:`\text{nrhs}_g` specifying the number of right-hand sides (:math:`0 \le \text{nrhs}_g`) for group :math:`g`. - -a - Array of ``batch_size`` pointers to factorizations of the matrices :math:`A_i`, as returned by the Group API of the:ref:`onemkl_lapack_getrf_batch_usm` function. - -lda - Array of ``group_count`` parameters :math:`\text{lda}_g` specifying the leading dimensions of :math:`A_i` from group :math:`g`. - -ipiv - ``ipiv`` array, as returned by the Group API of the :ref:`onemkl_lapack_getrf_batch_usm` function. - -b - The array containing ``batch_size`` pointers to the matrices :math:`B_i` whose columns are the right-hand sides for the systems of equations. - -ldb - Array of ``group_count`` parameters :math:`\text{ldb}_g` specifying the leading dimensions of :math:`B_i` in the group :math:`g`. - -group_count - Specifies the number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Group API of the :ref:`onemkl_lapack_getrs_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -b - Solution matrices :math:`X_i`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Strided API** - - | The routine solves the following systems of linear equations for :math:`X_i`: - | :math:`A_iX_i = B_i`, if ``trans=mkl::transpose::nontrans`` - | :math:`A_i^TX_i = B_i`, if ``trans=mkl::transpose::trans`` - | :math:`A_i^HX_i = B_i`, if ``trans=mkl::transpose::conjtrans`` - | Before calling this routine, the Strided API of the :ref:`onemkl_lapack_getrf_batch` function should be called to compute the LU factorizations of :math:`A_i`. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event getrs_batch(sycl::queue &queue, mkl::transpose trans, std::int64_t n, std::int64_t nrhs, T *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, T *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -trans - | Form of the equations: - | If ``trans = mkl::transpose::nontrans``, then :math:`A_iX_i = B_i` is solved for :math:`X_i`. - | If ``trans = mkl::transpose::trans``, then :math:`A_i^TX_i = B_i` is solved for :math:`X_i`. - | If ``trans = mkl::transpose::conjtrans``, then :math:`A_i^HX_i = B_i` is solved for :math:`X_i`. - -n - Order of the matrices :math:`A_i` and the number of rows in matrices :math:`B_i` (:math:`0 \le n`). - -nrhs - Number of right-hand sides (:math:`0 \le \text{nrhs}`). - -a - Array containing the factorizations of the matrices :math:`A_i`, as returned by the Strided API of the:ref:`onemkl_lapack_getrf_batch_usm` function. - -lda - Leading dimension of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`B_i` inside the batch array ``b``. - -ipiv - ``ipiv`` array, as returned by getrf_batch (USM) function. - -stride_ipiv - Stride between the beginnings of arrays :math:`\text{ipiv}_i` inside the array ``ipiv``. - -b - Array containing the matrices :math:`B_i` whose columns are the right-hand sides for the systems of equations. - -ldb - Leading dimensions of :math:`B_i`. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Strided API of the :ref:`onemkl_lapack_getrs_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -b - Solution matrices :math:`X_i`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/getrs_batch_scratchpad_size.rst b/docs/domains/lapack/getrs_batch_scratchpad_size.rst deleted file mode 100644 index 491524482..000000000 --- a/docs/domains/lapack/getrs_batch_scratchpad_size.rst +++ /dev/null @@ -1,135 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getrs_batch_scratchpad_size: - -getrs_batch_scratchpad_size -=========================== - -Computes size of scratchpad memory required for the :ref:`onemkl_lapack_getrs_batch` function. - -.. container:: section - - .. rubric:: Description - -``getrs_batch_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -**Group API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_getrs_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -trans - | Array of ``group_count`` parameters :math:`\text{trans}_g` indicating the form of the equations for the group :math:`g`: - | If ``trans = mkl::transpose::nontrans``, then :math:`A_iX_i = B_i` is solved for :math:`X_i`. - | If ``trans = mkl::transpose::trans``, then :math:`A_i^TX_i = B_i` is solved for :math:`X_i`. - | If ``trans = mkl::transpose::conjtrans``, then :math:`A_iHX_i = B_i` is solved for :math:`X_i`. - -n - Array of ``group_count`` parameters :math:`n_g` specifying the order of the matrices :math:`A_i` and the number of rows in matrices :math:`B_i` (:math:`0 \le n_g`) belonging to group :math:`g`. - -nrhs - Array of ``group_count`` parameters nrhsg specifying the number of right-hand sides (:math:`0 \le \text{nrhs}_g`) for group :math:`g`. - -lda - Array of ``group_count`` parameters :math:`\text{lda}_g` specifying the leading dimensions of :math:`A_i` from group :math:`g`. - -ldb - Array of ``group_count`` parameters :math:`\text{ldb}_g` specifying the leading dimensions of :math:`B_i` in the group :math:`g`. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_getrs_batch` function. - -**Strided API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_getrs_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -trans - | Indicates the form of the equations: - | ``If trans = mkl::transpose::nontrans``, then :math:`A_iX_i = B_i` is solved for :math:`X_i`. - | If ``trans = mkl::transpose::trans``, then :math:`A_i^TX_i = B_i` is solved for :math:`X_i`. - | If ``trans = mkl::transpose::conjtrans``, then :math:`A_i^HX_i = B_i` is solved for :math:`X_i`. - -n - Order of the matrices :math:`A_i` and the number of rows in matrices :math:`B_i` (:math:`0 \le n`). - -nrhs - Number of right-hand sides (:math:`0 \le \text{nrhs}`). - -lda - Leading dimension of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`B_i` inside the batch array ``b``. - -stride_ipiv - Stride between the beginnings of arrays ipivi inside the array ``ipiv``. - -ldb - Leading dimension of :math:`B_i`. - -batch_size - Number of problems in a batch. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_getrs_batch` function. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/getrs_scratchpad_size.rst b/docs/domains/lapack/getrs_scratchpad_size.rst deleted file mode 100644 index 8a2741745..000000000 --- a/docs/domains/lapack/getrs_scratchpad_size.rst +++ /dev/null @@ -1,85 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_getrs_scratchpad_size: - -getrs_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_getrs` function. - -.. container:: section - - .. rubric:: Description - -``getrs_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_getrs` function should be able to hold. -Calls to this routine must specify the template parameter -explicitly. - -getrs_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_getrs` function will be performed. - -trans - Indicates the form of the equations: - - If ``trans=oneapi::mkl::transpose::nontrans``, then :math:`AX = B` is solved - for :math:`X`. - - If ``trans=oneapi::mkl::transpose::trans``, then :math:`A^TX = B` is solved - for :math:`X`. - - If ``trans=oneapi::mkl::transpose::conjtrans``, then :math:`A^HX = B` is - solved for :math:`X`. - -n - The order of the matrix :math:`A` :math:`(0 \le n)` and the number of rows in matrix - :math:`B(0 \le n)`. - -nrhs - The number of right-hand sides (:math:`0 \le \text{nrhs}`). - -lda - The leading dimension of ``a``. - -ldb - The leading dimension of ``b``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_getrs` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines`  - - diff --git a/docs/domains/lapack/heevd.rst b/docs/domains/lapack/heevd.rst deleted file mode 100644 index 0e0247dd3..000000000 --- a/docs/domains/lapack/heevd.rst +++ /dev/null @@ -1,182 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_heevd: - -heevd -===== - -Computes all eigenvalues and, optionally, all eigenvectors of a -complex Hermitian matrix using divide and conquer algorithm. - -.. container:: section - - .. rubric:: Description - -``heevd`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -The routine computes all the eigenvalues, and optionally all the -eigenvectors, of a complex Hermitian matrix :math:`A`. In other words, it -can compute the spectral factorization of :math:`A` as: :math:`A = Z\Lambda Z^H`. - -Here :math:`\Lambda` is a real diagonal matrix whose diagonal elements are the -eigenvalues :math:`\lambda_i`, and :math:`Z` is the (complex) unitary matrix -whose columns are the eigenvectors :math:`z_{i}`. Thus, - -:math:`Az_i = \lambda_i z_i` for :math:`i = 1, 2, ..., n`. - -If the eigenvectors are requested, then this routine uses a divide -and conquer algorithm to compute eigenvalues and eigenvectors. -However, if only eigenvalues are required, then it uses the -Pal-Walker-Kahan variant of the QL or QR algorithm. - -heevd (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, butter &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = job::upper``, a stores the upper triangular - part of :math:`A`. - - If ``upper_lower = job::lower``, a stores the lower triangular - part of :math:`A`. - -n - The order of the matrix :math:`A` (:math:`0 \le n`). - -a - The buffer ``a``, size (``lda,*``). The buffer ``a`` contains the matrix - :math:`A`. The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. Must be at least :math:`\max(1,n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_heevd_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - If ``jobz = job::vec``, then on exit this buffer is overwritten by - the unitary matrix :math:`Z` which contains the eigenvectors of :math:`A`. - -w - Buffer, size at least n. Contains the eigenvalues - of the matrix :math:`A` in ascending order. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -heevd (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, butter &a, std::int64_t lda, RealT *w, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = job::upper``, a stores the upper triangular - part of :math:`A`. - - If ``upper_lower = job::lower``, a stores the lower triangular - part of :math:`A`. - -n - The order of the matrix :math:`A` (:math:`0 \le n`). - -a - Pointer to array containing :math:`A`, size (``lda,*``).The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. Must be at least :math:`\max(1,n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_heevd_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - If ``jobz = job::vec``, then on exit this array is overwritten by - the unitary matrix :math:`Z` which contains the eigenvectors of :math:`A`. - -w - Pointer to array of size at least :math:`n`. Contains the eigenvalues - of the matrix :math:`A` in ascending order. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - diff --git a/docs/domains/lapack/heevd_scratchpad_size.rst b/docs/domains/lapack/heevd_scratchpad_size.rst deleted file mode 100644 index 4825f73e7..000000000 --- a/docs/domains/lapack/heevd_scratchpad_size.rst +++ /dev/null @@ -1,81 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_heevd_scratchpad_size: - -heevd_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_heevd` function. - -.. container:: section - - .. rubric:: Description - -``heevd_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_heevd` function should be able to hold. -Calls to this routine must specify the template parameter -explicitly. - -heevd_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_heevd` function will be performed. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = job::upper``, a stores the upper triangular - part of :math:`A`. - - If ``upper_lower = job::lower``, a stores the lower triangular - part of :math:`A`. - -n - The order of the matrix :math:`A` (:math:`0 \le n`). - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_heevd` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/hegvd.rst b/docs/domains/lapack/hegvd.rst deleted file mode 100644 index 266b9964d..000000000 --- a/docs/domains/lapack/hegvd.rst +++ /dev/null @@ -1,249 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_hegvd: - -hegvd -===== - -Computes all eigenvalues and, optionally, eigenvectors of a real -generalized symmetric definite eigenproblem using a divide and -conquer method. - -.. container:: section - - .. rubric:: Description - -``hegvd`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -The routine computes all the eigenvalues, and optionally, the -eigenvectors of a complex generalized Hermitian positive-definite -eigenproblem, of the form - -:math:`Ax = \lambda Bx, ABx = \lambda x`, or :math:`BAx =\lambda x`. - -Here :math:`A` and :math:`B` are assumed to be Hermitian and :math:`B` is also -positive definite. - -It uses a divide and conquer algorithm. - -hegvd (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -itype - Must be 1 or 2 or 3. Specifies the problem type to be solved: - - if :math:`\text{itype} = 1`, the problem type is :math:`Ax = \lambda Bx;` - - if :math:`\text{itype} = 2`, the problem type is :math:`ABx = \lambda x;` - - if :math:`\text{itype} = 3`, the problem type is :math:`BAx = \lambda x`. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = uplo::upper``, ``a`` and ``b`` store the upper - triangular part of :math:`A` and :math:`B`. - - If ``upper_lower = uplo::lower``, ``a`` and ``b`` stores the lower - triangular part of :math:`A` and :math:`B`. - -n - The order of the matrices :math:`A` and :math:`B` (:math:`0 \le n`). - -a - Buffer, size ``a(lda,*)`` contains the upper or lower triangle of - the Hermitian matrix :math:`A`, as specified by upper_lower. - - The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``; at least :math:`\max(1,n)`. - -b - Buffer, size ``b(ldb,*)`` contains the upper or lower triangle of - the Hermitian matrix :math:`B`, as specified by upper_lower. - - The second dimension of ``b`` must be at least :math:`\max(1, n)`. - -ldb - The leading dimension of ``b``; at least :math:`\max(1,n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_hegvd_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - On exit, if ``jobz = job::vec``, then if :math:`\text{info} = 0`, ``a`` contains - the matrix :math:`Z` of eigenvectors. The eigenvectors are normalized - as follows: - - if :math:`\text{itype} = 1` or :math:`\text{itype} = 2`, :math:`Z^{H}BZ = I`; - - if :math:`\text{itype} = 3`, :math:`Z^{H}B^{-1}Z = I`; - - If ``jobz = job::novec``, then on exit the upper triangle (if - ``upper_lower = uplo::upper``) or the lower triangle (if - ``upper_lower = uplo::lower``) of :math:`A`, including the diagonal, - is destroyed. - -b - On exit, if :math:`\text{info} \le n`, the part of ``b`` containing the matrix is - overwritten by the triangular factor :math:`U` or :math:`L` from the - Cholesky factorization :math:`B = U^{H}U`\ or :math:`B = LL^{H}`. - -w - Buffer, size at least :math:`n`. If :math:`\text{info} = 0`, contains the eigenvalues - of the matrix :math:`A` in ascending order. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -hegvd (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, T *a, std::int64_t lda, T *b, std::int64_t ldb, RealT *w, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -itype - Must be 1 or 2 or 3. Specifies the problem type to be solved: - - if :math:`\text{itype} = 1`, the problem type is :math:`Ax = \lambda Bx;` - - if :math:`\text{itype} = 2`, the problem type is :math:`ABx = \lambda x;` - - if :math:`\text{itype} = 3`, the problem type is :math:`BAx = \lambda x`. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = uplo::upper``, ``a`` and ``b`` store the upper - triangular part of :math:`A` and :math:`B`. - - If ``upper_lower = uplo::lower``, ``a`` and ``b`` stores the lower - triangular part of :math:`A` and :math:`B`. - -n - The order of the matrices :math:`A` and :math:`B` (:math:`0 \le n`). - -a - Pointer to array of size ``a(lda,*)`` containing the upper or lower triangle of - the Hermitian matrix :math:`A`, as specified by upper_lower. - The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``; at least :math:`\max(1,n)`. - -b - Pointer to array of size ``b(ldb,*)`` containing the upper or lower triangle of - the Hermitian matrix :math:`B`, as specified by upper_lower. - The second dimension of ``b`` must be at least :math:`\max(1, n)`. - -ldb - The leading dimension of ``b``; at least :math:`\max(1,n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_hegvd_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - On exit, if ``jobz = job::vec``, then if :math:`\text{info} = 0`, ``a`` contains - the matrix :math:`Z` of eigenvectors. The eigenvectors are normalized - as follows: - - if :math:`\text{itype} = 1`` or :math:`\text{itype} = 2`, :math:`Z^{H}BZ = I`; - - if :math:`\text{itype} = 3`, :math:`Z^{H} B^{-1} Z = I`; - - If ``jobz = job::novec``, then on exit the upper triangle (if - ``upper_lower = uplo::upper``) or the lower triangle (if - ``upper_lower = uplo::lower``) of :math:`A`, including the diagonal, - is destroyed. - -b - On exit, if :math:`\text{info} \le n`, the part of ``b`` containing the matrix is - overwritten by the triangular factor :math:`U` or :math:`L` from the - Cholesky factorization :math:`B = U^{H}U`\ or :math:`B` = - :math:`LL^{H}`. - -w - Pointer to array of size at least n. If :math:`\text{info} = 0`, contains the eigenvalues - of the matrix :math:`A` in ascending order. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - diff --git a/docs/domains/lapack/hegvd_scratchpad_size.rst b/docs/domains/lapack/hegvd_scratchpad_size.rst deleted file mode 100644 index 8e33d8b63..000000000 --- a/docs/domains/lapack/hegvd_scratchpad_size.rst +++ /dev/null @@ -1,95 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_hegvd_scratchpad_size: - -hegvd_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_hegvd` function. - -.. container:: section - - .. rubric:: Description - -``hegvd_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_hegvd` function should be able to hold. -Calls to this routine must specify the template parameter -explicitly. - -hegvd_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda, std::int64_t ldb) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_hegvd` function will be performed. - -itype - Must be 1 or 2 or 3. Specifies the problem type to be solved: - - if :math:`\text{itype} = 1`, the problem type is :math:`Ax = \lambda Bx`; - - if :math:`\text{itype} = 2`, the problem type is :math:`ABx = \lambda x`; - - if :math:`\text{itype} = 3`, the problem type is :math:`BAx = \lambda x`. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = uplo::upper``, ``a`` and ``b`` store the upper - triangular part of :math:`A` and :math:`B`. - - If ``upper_lower = uplo::lower``, ``a`` and ``b`` store the lower - triangular part of :math:`A` and :math:`B`. - -n - The order of the matrices :math:`A` and :math:`B` (:math:`0 \le n`). - -lda - The leading dimension of ``a``. Currently ``lda`` is not referenced in - this function. - -ldb - The leading dimension of ``b``. Currently ``ldb`` is not referenced in - this function. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_hegvd` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/hetrd.rst b/docs/domains/lapack/hetrd.rst deleted file mode 100644 index b8c855d62..000000000 --- a/docs/domains/lapack/hetrd.rst +++ /dev/null @@ -1,206 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_hetrd: - -hetrd -===== - -Reduces a complex Hermitian matrix to tridiagonal form. - -.. container:: section - - .. rubric:: Description - -``hetrd`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - Routine name - - T - * - ``chetrd`` - - ``std::complex`` - * - ``zhetrd`` - - ``std::complex`` - -The routine reduces a complex Hermitian matrix :math:`A` to symmetric -tridiagonal form :math:`T` by a unitary similarity transformation: -:math:`A = QTQ^H`. The unitary matrix :math:`Q` is not formed explicitly but -is represented as a product of :math:`n-1` elementary reflectors. -Routines are provided to work with :math:`Q` in this representation. - -hetrd (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void hetrd(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = uplo::upper``, ``a`` stores the upper triangular - part of :math:`A`. - - If ``upper_lower = uplo::lower``, ``a`` stores the lower triangular - part of :math:`A`. - -n - The order of the matrices :math:`A` :math:`(0 \le n)`. - -a - Buffer, size ``(lda,*)``. The buffer ``a`` contains either the upper - or lower triangle of the Hermitian matrix :math:`A`, as specified by - upper_lower. - - The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``; at least :math:`\max(1, n)` - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_hetrd_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - On exit, - - if ``upper_lower = uplo::upper``, the diagonal and first - superdiagonal of :math:`A` are overwritten by the corresponding - elements of the tridiagonal matrix :math:`T`, and the elements above - the first superdiagonal, with the buffer ``tau``, represent the - orthogonal matrix :math:`Q` as a product of elementary reflectors; - - if ``upper_lower = uplo::lower``, the diagonal and first - subdiagonal of :math:`A` are overwritten by the corresponding elements - of the tridiagonal matrix :math:`T`, and the elements below the first - subdiagonal, with the buffer ``tau``, represent the orthogonal matrix - :math:`Q` as a product of elementary reflectors. - -d - Buffer containing the diagonal elements of the matrix :math:`T`. The - dimension of ``d`` must be at least :math:`\max(1, n)`. - -e - Buffer containing the off diagonal elements of the matrix :math:`T`. - The dimension of ``e`` must be at least :math:`\max(1, n-1)`. - -tau - Buffer, size at least :math:`\max(1, n-1)`. Stores :math:`(n-1)` scalars - that define elementary reflectors in decomposition of the unitary - matrix :math:`Q` in a product of :math:`n-1` elementary reflectors. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -hetrd (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, T *a, std::int64_t lda, RealT *d, RealT *e, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = uplo::upper``, ``a`` stores the upper triangular - part of :math:`A`. - - If ``upper_lower = uplo::lower``, ``a`` stores the lower triangular - part of :math:`A`. - -n - The order of the matrices :math:`A` :math:`(0 \le n)`. - -a - The pointer to matrix :math:`A`, size ``(lda,*)``. Contains either the upper - or lower triangle of the Hermitian matrix :math:`A`, as specified by - ``upper_lower``. - The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``; at least :math:`\max(1, n)` - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_hetrd_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - On exit, - - if ``upper_lower = uplo::upper``, the diagonal and first - superdiagonal of :math:`A` are overwritten by the corresponding - elements of the tridiagonal matrix :math:`T`, and the elements above - the first superdiagonal, with the array ``tau``, represent the - orthogonal matrix :math:`Q` as a product of elementary reflectors; - - if ``upper_lower = uplo::lower``, the diagonal and first - subdiagonal of :math:`A` are overwritten by the corresponding elements - of the tridiagonal matrix :math:`T`, and the elements below the first - subdiagonal, with the array ``tau``, represent the orthogonal matrix - :math:`Q` as a product of elementary reflectors. - -d - Pointer to diagonal elements of the matrix :math:`T`. The - dimension of ``d`` must be at least :math:`\max(1, n)`. - -e - Pointer to off diagonal elements of the matrix :math:`T`. - The dimension of ``e`` must be at least :math:`\max(1, n-1)`. - -tau - Pointer to array of size at least :math:`\max(1, n-1)`. Stores :math:`(n-1)` scalars - that define elementary reflectors in decomposition of the unitary - matrix :math:`Q` in a product of :math:`n-1` elementary reflectors. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/hetrd_scratchpad_size.rst b/docs/domains/lapack/hetrd_scratchpad_size.rst deleted file mode 100644 index 006d50a3c..000000000 --- a/docs/domains/lapack/hetrd_scratchpad_size.rst +++ /dev/null @@ -1,74 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_hetrd_scratchpad_size: - -hetrd_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_hetrd` function. - -.. container:: section - - .. rubric:: Description - -``hetrd_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_hetrd` function should be able to hold. -Calls to this routine must specify the template parameter -explicitly. - -hetrd_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_hetrd` function will be performed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = uplo::upper``, ``a`` stores the upper triangular - part of :math:`A` and :math:`B`. - - If ``upper_lower = uplo::lower``, ``a`` stores the lower triangular - part of :math:`A`. - -n - The order of the matrices :math:`A` and :math:`B` (:math:`0 \le n`). - -lda - The leading dimension of ``a``. Currently, ``lda`` is not referenced in - this function. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_hetrd` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/hetrf.rst b/docs/domains/lapack/hetrf.rst deleted file mode 100644 index 7be2b4bbe..000000000 --- a/docs/domains/lapack/hetrf.rst +++ /dev/null @@ -1,164 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_hetrf: - -hetrf -===== - -Computes the Bunch-Kaufman factorization of a complex Hermitian matrix. - -.. container:: section - - .. rubric:: Description - -``hetrf`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -The routine computes the factorization of a complex Hermitian -matrix :math:`A` using the Bunch-Kaufman diagonal pivoting method. The -form of the factorization is: - -- if ``upper_lower=uplo::upper``, :math:`A` = :math:`UDU^{H}` - -- if ``upper_lower=uplo::lower``, :math:`A` = :math:`LDL^{H}` - -where :math:`A` is the input matrix, :math:`U` and :math:`L` are products of -permutation and triangular matrices with unit diagonal (upper -triangular for :math:`U` and lower triangular for :math:`L`), and :math:`D` is a -Hermitian block-diagonal matrix with :math:`1 \times 1` and :math:`2 \times 2` diagonal -blocks. :math:`U` and :math:`L` have :math:`2 \times 2` unit diagonal blocks -corresponding to the :math:`2 \times 2` blocks of :math:`D`. - -hetrf (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void hetrf(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates whether the upper or lower triangular part of :math:`A` is stored and how :math:`A` is factored: - - If ``upper_lower=uplo::upper``, the buffer ``a`` stores the upper triangular part of the matrix :math:`A`, and :math:`A` is factored as :math:`UDU^H`. - - If ``upper_lower=uplo::lower``, the buffer ``a`` stores the lower triangular part of the matrix :math:`A`, and :math:`A` is factored as :math:`LDL^H`. - -n - The order of matrix :math:`A` (:math:`0 \le n`). - -a - The buffer ``a``, size :math:`\max(1,\text{lda} \cdot n)`. The buffer ``a`` contains either the upper or the lower triangular part of the matrix :math:`A` (see ``upper_lower``). The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -scratchpad - Buffer holding scratchpad memory to be used by the routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_hetrf_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - The upper or lower triangular part of a is overwritten by details of the block-diagonal matrix :math:`D` and the multipliers used to obtain the factor :math:`U` (or :math:`L`). - -ipiv - Buffer, size at least :math:`\max(1, n)`. Contains details of the interchanges and the block structure of :math:`D`. If :math:`\text{ipiv}(i)=k>0`, then :math:`d_{ii}` is a :math:`1 \times 1` block, and the :math:`i`-th row and column of :math:`A` was interchanged with the :math:`k`-th row and column. - - If ``upper_lower=oneapi::mkl::uplo::upper`` and :math:`\text{ipiv}(i)=\text{ipiv}(i-1)=-m<0`, then :math:`D` has a :math:`2 \times 2` block in rows/columns :math:`i` and :math:`i`-1, and (:math:`i-1`)-th row and column of :math:`A` was interchanged with the :math:`m`-th row and column. - - If ``upper_lower=oneapi::mkl::uplo::lower`` and :math:`\text{ipiv}(i)=\text{ipiv}(i+1)=-m<0`, then :math:`D` has a :math:`2 \times 2` block in rows/columns :math:`i` and :math:`i+1`, and (:math:`i+1`)-th row and column of :math:`A` was interchanged with the :math:`m`-th row and column. - -hetrf (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, T *a, std::int64_t lda, int_64 *ipiv, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates whether the upper or lower triangular part of :math:`A` is stored and how :math:`A` is factored: - - If ``upper_lower=uplo::upper``, the array ``a`` stores the upper triangular part of the matrix :math:`A`, and :math:`A` is factored as :math:`UDU^H`. - - If ``upper_lower=uplo::lower``, the array ``a`` stores the lower triangular part of the matrix :math:`A`, and :math:`A` is factored as :math:`LDL^H`. - -n - The order of matrix :math:`A` (:math:`0 \le n`). - -a - The pointer to :math:`A`, size :math:`\max(1,\text{lda} \cdot n)`, containing either the upper or the lower triangular part of the matrix :math:`A` (see ``upper_lower``). The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -scratchpad - Pointer to scratchpad memory to be used by the routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_hetrf_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - The upper or lower triangular part of a is overwritten by details of the block-diagonal matrix :math:`D` and the multipliers used to obtain the factor :math:`U` (or :math:`L`). - -ipiv - Pointer to array of size at least :math:`\max(1, n)`. Contains details of the interchanges and the block structure of :math:`D`. If :math:`\text{ipiv}(i)=k>0`, then :math:`d_{ii}` is a :math:`1 \times 1` block, and the :math:`i`-th row and column of :math:`A` was interchanged with the :math:`k`-th row and column. - - If ``upper_lower=oneapi::mkl::uplo::upper`` and :math:`\text{ipiv}(i)=\text{ipiv}(i-1)=-m<0`, then :math:`D` has a :math:`2 \times 2` block in rows/columns :math:`i` and :math:`i-1`, and (:math:`i-1`)-th row and column of :math:`A` was interchanged with the :math:`m`-th row and column. - - If ``upper_lower=oneapi::mkl::uplo::lower`` and :math:`\text{ipiv}(i)=\text{ipiv}(i+1)=-m<0`, then :math:`D` has a :math:`2 \times 2` block in rows/columns :math:`i` and :math:`i+1`, and (:math:`i+1`)-th row and column of :math:`A` was interchanged with the :math:`m`-th row and column. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/hetrf_scratchpad_size.rst b/docs/domains/lapack/hetrf_scratchpad_size.rst deleted file mode 100644 index 38447dd6e..000000000 --- a/docs/domains/lapack/hetrf_scratchpad_size.rst +++ /dev/null @@ -1,74 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_hetrf_scratchpad_size: - -hetrf_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_hetrf` function. - -.. container:: section - - .. rubric:: Description - -``hetrf_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_hetrf` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -hetrf_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_hetrf` function will be performed. - -upper_lower - Indicates whether the upper or lower triangular part of :math:`A` is - stored and how :math:`A` is factored: - - If ``upper_lower=uplo::upper``, the buffer ``a`` stores the - upper triangular part of the matrix :math:`A`, and :math:`A` is - factored as :math:`UDU^H`. - - If ``upper_lower=uplo::lower``, the buffer ``a`` stores the - lower triangular part of the matrix :math:`A`, and :math:`A` is - factored as :math:`LDL^H` - -n - The order of the matrix :math:`A` (:math:`0 \le n`). - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_hetrf` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/lapack-like-extensions.inc.rst b/docs/domains/lapack/lapack-like-extensions.inc.rst deleted file mode 100644 index b3378b25b..000000000 --- a/docs/domains/lapack/lapack-like-extensions.inc.rst +++ /dev/null @@ -1,74 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack-like-extensions-routines: - -LAPACK-like Extensions Routines -=============================== - - -.. container:: - - - oneAPI Math Kernel Library DPC++ provides additional routines to - extend the functionality of the LAPACK routines. These include routines - to compute many independent factorizations, linear equation solutions, and similar. - The following table lists the LAPACK-like Extensions routine groups. - - - .. container:: tablenoborder - - - .. list-table:: - :header-rows: 1 - - * - Routines - - Scratchpad Size Routines - - Description - * - :ref:`onemkl_lapack_geqrf_batch` - - :ref:`onemkl_lapack_geqrf_batch_scratchpad_size` - - Computes the QR factorizations of a batch of general matrices. - * - :ref:`onemkl_lapack_getrf_batch` - - :ref:`onemkl_lapack_getrf_batch_scratchpad_size` - - Computes the LU factorizations of a batch of general matrices. - * - :ref:`onemkl_lapack_getri_batch` - - :ref:`onemkl_lapack_getri_batch_scratchpad_size` - - Computes the inverses of a batch of LU-factored general matrices. - * - :ref:`onemkl_lapack_getrs_batch` - - :ref:`onemkl_lapack_getrs_batch_scratchpad_size` - - Solves systems of linear equations with a batch of LU-factored square coefficient matrices, with multiple right-hand sides. - * - :ref:`onemkl_lapack_orgqr_batch` - - :ref:`onemkl_lapack_orgqr_batch_scratchpad_size` - - Generates the real orthogonal/complex unitary matrix :math:`Q_i` of the QR factorization formed by geqrf_batch. - * - :ref:`onemkl_lapack_potrf_batch` - - :ref:`onemkl_lapack_potrf_batch_scratchpad_size` - - Computes the Cholesky factorization of a batch of symmetric (Hermitian) positive-definite matrices. - * - :ref:`onemkl_lapack_potrs_batch` - - :ref:`onemkl_lapack_potrs_batch_scratchpad_size` - - Solves systems of linear equations with a batch of Cholesky-factored symmetric (Hermitian) positive-definite coefficient matrices, with multiple right-hand sides. - * - :ref:`onemkl_lapack_ungqr_batch` - - :ref:`onemkl_lapack_ungqr_batch_scratchpad_size` - - Generates the complex unitary matrix :math:`Q_i` with the QR factorization formed by geqrf_batch. - - - -.. toctree:: - :hidden: - - geqrf_batch - geqrf_batch_scratchpad_size - getrf_batch - getrf_batch_scratchpad_size - getri_batch - getri_batch_scratchpad_size - getrs_batch - getrs_batch_scratchpad_size - orgqr_batch - orgqr_batch_scratchpad_size - potrf_batch - potrf_batch_scratchpad_size - potrs_batch - potrs_batch_scratchpad_size - ungqr_batch - ungqr_batch_scratchpad_size diff --git a/docs/domains/lapack/lapack-linear-equation-routines.inc.rst b/docs/domains/lapack/lapack-linear-equation-routines.inc.rst deleted file mode 100644 index 6e6c25574..000000000 --- a/docs/domains/lapack/lapack-linear-equation-routines.inc.rst +++ /dev/null @@ -1,121 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack-linear-equation-routines: - -LAPACK Linear Equation Routines -=============================== - - -.. container:: - - - LAPACK Linear Equation routines are used for factoring a matrix, - solving a system of linear equations, solving linear least squares problems, - and inverting a matrix. The following table lists the LAPACK Linear Equation - routine groups. - - - .. container:: tablenoborder - - - .. list-table:: - :header-rows: 1 - - * - Routines - - Scratchpad Size Routines - - Description - * - :ref:`onemkl_lapack_geqrf` - - :ref:`onemkl_lapack_geqrf_scratchpad_size` - - Computes the QR factorization of a general m-by-n matrix. - * - :ref:`onemkl_lapack_gerqf` - - :ref:`onemkl_lapack_gerqf_scratchpad_size` - - Computes the RQ factorization of a general m-by-n matrix. - * - :ref:`onemkl_lapack_getrf` - - :ref:`onemkl_lapack_getrf_scratchpad_size` - - Computes the LU factorization of a general m-by-n matrix. - * - :ref:`onemkl_lapack_getri` - - :ref:`onemkl_lapack_getri_scratchpad_size` - - Computes the inverse of an LU-factored general matrix. - * - :ref:`onemkl_lapack_getrs` - - :ref:`onemkl_lapack_getrs_scratchpad_size` - - Solves a system of linear equations with an LU-factored square coefficient matrix, with multiple right-hand sides. - * - :ref:`onemkl_lapack_hetrf` - - :ref:`onemkl_lapack_hetrf_scratchpad_size` - - Computes the Bunch-Kaufman factorization of a complex Hermitian matrix. - * - :ref:`onemkl_lapack_orgqr` - - :ref:`onemkl_lapack_orgqr_scratchpad_size` - - Generates the real orthogonal matrix :math:`Q` of the QR factorization formed by geqrf. - * - :ref:`onemkl_lapack_ormqr` - - :ref:`onemkl_lapack_ormqr_scratchpad_size` - - Multiplies a real matrix by the orthogonal matrix :math:`Q` of the QR factorization formed by geqrf. - * - :ref:`onemkl_lapack_ormrq` - - :ref:`onemkl_lapack_ormrq_scratchpad_size` - - Multiplies a real matrix by the orthogonal matrix :math:`Q` of the RQ factorization formed by gerqf. - * - :ref:`onemkl_lapack_potrf` - - :ref:`onemkl_lapack_potrf_scratchpad_size` - - Computes the Cholesky factorization of a symmetric (Hermitian) positive-definite matrix. - * - :ref:`onemkl_lapack_potri` - - :ref:`onemkl_lapack_potri_scratchpad_size` - - Computes the inverse of a Cholesky-factored symmetric (Hermitian) positive-definite matrix. - * - :ref:`onemkl_lapack_potrs` - - :ref:`onemkl_lapack_potrs_scratchpad_size` - - Solves a system of linear equations with a Cholesky-factored symmetric (Hermitian) positive-definite coefficient matrix, with multiple right-hand sides. - * - :ref:`onemkl_lapack_sytrf` - - :ref:`onemkl_lapack_sytrf_scratchpad_size` - - Computes the Bunch-Kaufman factorization of a symmetric matrix. - * - :ref:`onemkl_lapack_trtrs` - - :ref:`onemkl_lapack_trtrs_scratchpad_size` - - Solves a system of linear equations with a triangular coefficient matrix, with multiple right-hand sides. - * - :ref:`onemkl_lapack_ungqr` - - :ref:`onemkl_lapack_ungqr_scratchpad_size` - - Generates the complex unitary matrix :math:`Q` of the QR factorization formed by geqrf. - * - :ref:`onemkl_lapack_unmqr` - - :ref:`onemkl_lapack_unmqr_scratchpad_size` - - Multiplies a complex matrix by the unitary matrix :math:`Q` of the QR factorization formed by geqrf. - * - :ref:`onemkl_lapack_unmrq` - - :ref:`onemkl_lapack_unmrq_scratchpad_size` - - Multiplies a complex matrix by the unitary matrix :math:`Q` of the RQ factorization formed by gerqf. - - - - - -.. toctree:: - :hidden: - - geqrf - geqrf_scratchpad_size - gerqf - gerqf_scratchpad_size - getrf - getrf_scratchpad_size - getri - getri_scratchpad_size - getrs - getrs_scratchpad_size - hetrf - hetrf_scratchpad_size - orgqr - orgqr_scratchpad_size - ormqr - ormqr_scratchpad_size - ormrq - ormrq_scratchpad_size - potrf - potrf_scratchpad_size - potri - potri_scratchpad_size - potrs - potrs_scratchpad_size - sytrf - sytrf_scratchpad_size - trtrs - trtrs_scratchpad_size - ungqr - ungqr_scratchpad_size - unmqr - unmqr_scratchpad_size - unmrq - unmrq_scratchpad_size diff --git a/docs/domains/lapack/lapack-singular-value-eigenvalue-routines.inc.rst b/docs/domains/lapack/lapack-singular-value-eigenvalue-routines.inc.rst deleted file mode 100644 index bcf8c1af3..000000000 --- a/docs/domains/lapack/lapack-singular-value-eigenvalue-routines.inc.rst +++ /dev/null @@ -1,105 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack-singular-value-eigenvalue-routines: - -LAPACK Singular Value and Eigenvalue Problem Routines -===================================================== - - -.. container:: - - - LAPACK Singular Value and Eigenvalue Problem routines are used for - singular value and eigenvalue problems, and for performing a number of related - computational tasks. The following table lists the LAPACK Singular Value and - Eigenvalue Problem routine groups. - - - .. container:: tablenoborder - - - .. list-table:: - :header-rows: 1 - - * - Routines - - Scratchpad Size Routines - - Description - * - :ref:`onemkl_lapack_gebrd` - - :ref:`onemkl_lapack_gebrd_scratchpad_size` - - Reduces a general matrix to bidiagonal form. - * - :ref:`onemkl_lapack_gesvd` - - :ref:`onemkl_lapack_gesvd_scratchpad_size` - - Computes the singular value decomposition of a general rectangular matrix. - * - :ref:`onemkl_lapack_heevd` - - :ref:`onemkl_lapack_heevd_scratchpad_size` - - Computes all eigenvalues and, optionally, all eigenvectors of a complex Hermitian matrix using divide and conquer algorithm. - * - :ref:`onemkl_lapack_hegvd` - - :ref:`onemkl_lapack_hegvd_scratchpad_size` - - Computes all eigenvalues and, optionally, all eigenvectors of a complex generalized Hermitian definite eigenproblem using divide and conquer algorithm. - * - :ref:`onemkl_lapack_hetrd` - - :ref:`onemkl_lapack_hetrd_scratchpad_size` - - Reduces a complex Hermitian matrix to tridiagonal form. - * - :ref:`onemkl_lapack_orgbr` - - :ref:`onemkl_lapack_orgbr_scratchpad_size` - - Generates the real orthogonal matrix :math:`Q` or :math:`P^T` determined by gebrd. - * - :ref:`onemkl_lapack_orgtr` - - :ref:`onemkl_lapack_orgtr_scratchpad_size` - - Generates the real orthogonal matrix :math:`Q` determined by sytrd. - * - :ref:`onemkl_lapack_ormtr` - - :ref:`onemkl_lapack_ormtr_scratchpad_size` - - Multiplies a real matrix by the orthogonal matrix :math:`Q` determined by sytrd. - * - :ref:`onemkl_lapack_syevd` - - :ref:`onemkl_lapack_syevd_scratchpad_size` - - Computes all eigenvalues and, optionally, all eigenvectors of a real symmetric matrix using divide and conquer algorithm. - * - :ref:`onemkl_lapack_sygvd` - - :ref:`onemkl_lapack_sygvd_scratchpad_size` - - Computes all eigenvalues and, optionally, all eigenvectors of a real generalized symmetric definite eigenproblem using divide and conquer algorithm. - * - :ref:`onemkl_lapack_sytrd` - - :ref:`onemkl_lapack_sytrd_scratchpad_size` - - Reduces a real symmetric matrix to tridiagonal form. - * - :ref:`onemkl_lapack_ungbr` - - :ref:`onemkl_lapack_ungbr_scratchpad_size` - - Generates the complex unitary matrix :math:`Q` or :math:`P^T` determined by gebrd. - * - :ref:`onemkl_lapack_ungtr` - - :ref:`onemkl_lapack_ungtr_scratchpad_size` - - Generates the complex unitary matrix :math:`Q` determined by hetrd. - * - :ref:`onemkl_lapack_unmtr` - - :ref:`onemkl_lapack_unmtr_scratchpad_size` - - Multiplies a complex matrix by the unitary matrix :math:`Q` determined by hetrd. - - - - -.. toctree:: - :hidden: - - gebrd - gebrd_scratchpad_size - gesvd - gesvd_scratchpad_size - heevd - heevd_scratchpad_size - hegvd - hegvd_scratchpad_size - hetrd - hetrd_scratchpad_size - orgbr - orgbr_scratchpad_size - orgtr - orgtr_scratchpad_size - ormtr - ormtr_scratchpad_size - syevd - syevd_scratchpad_size - sygvd - sygvd_scratchpad_size - sytrd - sytrd_scratchpad_size - ungbr - ungbr_scratchpad_size - ungtr - ungtr_scratchpad_size - unmtr - unmtr_scratchpad_size diff --git a/docs/domains/lapack/lapack.rst b/docs/domains/lapack/lapack.rst deleted file mode 100644 index bb11e72ed..000000000 --- a/docs/domains/lapack/lapack.rst +++ /dev/null @@ -1,43 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack: - -LAPACK Routines -+++++++++++++++ - -oneMKL provides a DPC++ interface to select routines from the Linear Algebra PACKage (LAPACK), as well as several LAPACK-like extension routines. - -.. include:: lapack-linear-equation-routines.inc.rst -.. include:: lapack-singular-value-eigenvalue-routines.inc.rst -.. include:: lapack-like-extensions.inc.rst - - -.. container:: - - .. container:: Note - - - .. rubric:: Note - :class: NoteTipHead - - - Different arrays used as parameters to oneMKL LAPACK routines must - not overlap. - - - .. container:: Note - - - .. rubric:: Warning - :name: warning - :class: NoteTipHead - - - LAPACK routines assume that input matrices do not contain IEEE 754 - special values such as INF or NaN values. Using these special - values may cause LAPACK to return unexpected results or become - unstable. - -**Parent topic:** :ref:`onemkl_dense_linear_algebra` diff --git a/docs/domains/lapack/orgbr.rst b/docs/domains/lapack/orgbr.rst deleted file mode 100644 index 6ff70338f..000000000 --- a/docs/domains/lapack/orgbr.rst +++ /dev/null @@ -1,226 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_orgbr: - -orgbr -===== - -Generates the real orthogonal matrix :math:`Q` or :math:`P^{T}` -determined by -:ref:`onemkl_lapack_gebrd`. - -``orgbr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. container:: section - - .. rubric:: Description - -The routine generates the whole or part of the orthogonal matrices -:math:`Q` and :math:`P^{T}` formed by the routines :ref:`onemkl_lapack_gebrd`. -All valid combinations of arguments are described in *Input parameters*. In -most cases you need the following: - -To compute the whole :math:`m \times m` matrix :math:`Q`: - -:: - - orgbr(queue, generate::q, m, m, n, a, ...) - -(note that the array ``a`` must have at least :math:`m` columns). - -To form the :math:`n` leading columns of :math:`Q` if :math:`m > n`: - -:: - - orgbr(queue, generate::q, m, n, n, a, ...) - -To compute the whole :math:`n \times n` matrix :math:`P^{T}`: - -:: - - orgbr(queue, generate::p, n, n, m, a, ...) - -(note that the array ``a`` must have at least :math:`n` rows). - -To form the :math:`m` leading rows of :math:`P^{T}` if :math:`m < n`: - -:: - - orgbr(queue, generate::p, m, n, m, a, ...) - -orgbr (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void orgbr(sycl::queue &queue, oneapi::mkl::generate gen, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -gen - Must be ``generate::q`` or ``generate::p``. - - If ``gen = generate::q``, the routine generates the matrix :math:`Q`. - - If ``gen = generate::p``, the routine generates the matrix - :math:`P^{T}`. - -m - The number of rows in the matrix :math:`Q` or :math:`P^{T}` to be - returned :math:`(0 \le m)`. - - If ``gen = generate::q``, :math:`m \le n \le \min(m, k)`. - - If ``gen = generate::p``, :math:`n \le m \le \min(n, k)`. - -n - The number of rows in the matrix :math:`Q` or :math:`P^{T}` to be - returned :math:`(0 \le n)`. See m for constraints. - -k - If ``gen = generate::q``, the number of columns in the original - :math:`m \times k` matrix reduced by - :ref:`onemkl_lapack_gebrd`. - - If ``gen = generate::p``, the number of rows in the original - :math:`k \times n` matrix reduced by - :ref:`onemkl_lapack_gebrd`. - -a - The buffer ``a`` as returned by - :ref:`onemkl_lapack_gebrd`. - -lda - The leading dimension of ``a``. - -tau - Buffer, size :math:`\min(m,k)` if ``gen = generate::q``, size - :math:`\min(n,k)` if ``gen = generate::p``. Scalar factor of the - elementary reflectors, as returned by :ref:`onemkl_lapack_gebrd` in the array tauq - or taup. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_orgbr_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by n leading columns of the :math:`m \times m` orthogonal matrix - :math:`Q` or :math:`P^{T}` (or the leading rows or columns thereof) - as specified by ``gen``, ``m``, and ``n``. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -orgbr (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate gen, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -gen - Must be ``generate::q`` or ``generate::p``. - - If ``gen = generate::q``, the routine generates the matrix :math:`Q`. - - If ``gen = generate::p``, the routine generates the matrix - :math:`P^{T}`. - -m - The number of rows in the matrix :math:`Q` or :math:`P^{T}` to be - returned :math:`(0 \le m)`. - - If ``gen = generate::q``, :math:`m \le n \le \min(m, k)`. - - If ``gen = generate::p``, :math:`n \le m \le \min(n, k)`. - -n - The number of rows in the matrix :math:`Q` or :math:`P^{T}` to be - returned :math:`(0 \le n)`. See m for constraints. - -k - If ``gen = generate::q``, the number of columns in the original - :math:`m \times k` matrix reduced by - :ref:`onemkl_lapack_gebrd`. - - If ``gen = generate::p``, the number of rows in the original - :math:`k \times n` matrix reduced by - :ref:`onemkl_lapack_gebrd`. - -a - Pointer to array ``a`` as returned by - :ref:`onemkl_lapack_gebrd`. - -lda - The leading dimension of ``a``. - -tau - Pointer to array of size :math:`\min(m,k)` if ``gen = generate::q``, size - :math:`\min(n,k)` if ``gen = generate::p``. Scalar factor of the - elementary reflectors, as returned by :ref:`onemkl_lapack_gebrd` in the array tauq - or taup. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_orgbr_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by n leading columns of the :math:`m \times m` orthogonal matrix - :math:`Q` or :math:`P^{T}` (or the leading rows or columns thereof) - as specified by ``gen``, ``m``, and ``n``. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - diff --git a/docs/domains/lapack/orgbr_scratchpad_size.rst b/docs/domains/lapack/orgbr_scratchpad_size.rst deleted file mode 100644 index 7e7804158..000000000 --- a/docs/domains/lapack/orgbr_scratchpad_size.rst +++ /dev/null @@ -1,90 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_orgbr_scratchpad_size: - -orgbr_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_orgbr` function. - -``orgbr_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. container:: section - - .. rubric:: Description - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_orgbr` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -orgbr_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate gen, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t &scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_orgbr` function will be performed. - -gen - Must be ``generate::q`` or ``generate::p``. - - If ``gen = generate::q``, the routine generates the matrix - :math:`Q`. - - If ``gen = generate::p``, the routine generates the matrix - :math:`P^{T}`. - -m - The number of rows in the matrix :math:`Q` or :math:`P^{T}` to be - returned :math:`(0 \le m)`. - - If ``gen = generate::q``, :math:`m \le n \le \min(m, k)`. - - If ``gen = generate::p``, :math:`n \le m \le \min(n, k)`. - -n - The number of rows in the matrix :math:`Q` or :math:`P^{T}` to be - returned :math:`(0 \le n)`. See ``m`` for constraints. - -k - If ``gen = generate::q``, the number of columns in the original - :math:`m \times k` matrix returned by - :ref:`onemkl_lapack_gebrd`. - - If ``gen = generate::p``, the number of rows in the original - :math:`k \times n` matrix returned by - :ref:`onemkl_lapack_gebrd`. - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_orgbr` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines`  - - diff --git a/docs/domains/lapack/orgqr.rst b/docs/domains/lapack/orgqr.rst deleted file mode 100644 index 532e7fd9c..000000000 --- a/docs/domains/lapack/orgqr.rst +++ /dev/null @@ -1,183 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_orgqr: - -orgqr -===== - -Generates the real orthogonal matrix :math:`Q` of the QR factorization formed -by :ref:`onemkl_lapack_geqrf`. - -.. container:: section - - .. rubric:: Description - -``orgqr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -The routine generates the whole or part of :math:`m \times m` orthogonal -matrix :math:`Q` of the QR factorization formed by the routine -:ref:`onemkl_lapack_geqrf`. - -Usually :math:`Q` is determined from the QR factorization of an ``m`` -by ``p`` matrix :math:`A` with :math:`m \ge p`. To compute the whole matrix -:math:`Q`, use: - -:: - - oneapi::mkl::lapack::orgqr(queue, m, m, p, a, lda, tau, scratchpad, scratchpad_size) - -To compute the leading :math:`p` columns of :math:`Q` (which form an -orthonormal basis in the space spanned by the columns of :math:`A`): - -:: - - oneapi::mkl::lapack::orgqr(queue, m, p, p, a, lda, tau, scratchpad, scratchpad_size) - -To compute the matrix :math:`Q^{k}` of the QR factorization of -leading :math:`k` columns of the matrix :math:`A`: - -:: - - oneapi::mkl::lapack::orgqr(queue, m, m, k, a, lda, tau, scratchpad, scratchpad_size) - -To compute the leading :math:`k` columns of :math:`Q^{k}` (which form -an orthonormal basis in the space spanned by leading :math:`k` columns of -the matrix :math:`A`): - -:: - - oneapi::mkl::lapack::orgqr(queue, m, k, k, a, lda, tau, scratchpad, scratchpad_size) - -orgqr (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` (:math:`0 \le k \le n`). - -a - The buffer ``a`` as returned by - :ref:`onemkl_lapack_geqrf`. - -lda - The leading dimension of ``a`` (:math:`\text{lda} \le m`). - -tau - The buffer ``tau`` as returned by - :ref:`onemkl_lapack_geqrf`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_orgqr_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by :math:`n` leading columns of the :math:`m \times m` orthogonal matrix - :math:`Q`. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -orgqr (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` (:math:`0 \le k \le n`). - -a - The pointer to ``a`` as returned by - :ref:`onemkl_lapack_geqrf`. - -lda - The leading dimension of ``a`` (:math:`\text{lda} \le m`). - -tau - The pointer to ``tau`` as returned by - :ref:`onemkl_lapack_geqrf`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_orgqr_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by :math:`n` leading columns of the :math:`m \times m` orthogonal matrix - :math:`Q`. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - - diff --git a/docs/domains/lapack/orgqr_batch.rst b/docs/domains/lapack/orgqr_batch.rst deleted file mode 100644 index 6984ebead..000000000 --- a/docs/domains/lapack/orgqr_batch.rst +++ /dev/null @@ -1,262 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_orgqr_batch: - -orgqr_batch -=========== - -Generates the orthogonal/unitary matrix :math:`Q_i` of the QR factorizations for a group of general matrices. - -.. rubric:: Description - -``orgqr_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -.. _onemkl_lapack_orgqr_batch_buffer: - -orgqr_batch (Buffer Version) ----------------------------- - -.. rubric:: Description - -The buffer version of ``orgqr_batch`` supports only the strided API. - -**Strided API** - - | The routine generates the wholes or parts of :math:`m \times n` orthogonal matrices :math:`Q_i` of the batch of QR factorizations formed by the Strided API of the :ref:`onemkl_lapack_geqrf_batch_buffer` function. - | Usually :math:`Q_i` is determined from the QR factorization of an :math:`m \times p` matrix :math:`A_i` with :math:`m \ge p`. - | To compute the whole matrices :math:`Q_i`, use: - | ``orgqr_batch(queue, m, m, p, a, ...)`` - | To compute the leading :math:`p` columns of :math:`Q_i` (which form an orthonormal basis in the space spanned by the columns of :math:`A_i`): - | ``orgqr_batch(queue, m, p, p, a, ...)`` - | To compute the matrices :math:`Q_i^k` of the QR factorizations of leading :math:`k` columns of the matrices :math:`A_i`: - | ``orgqr_batch(queue, m, m, k, a, ...)`` - | To compute the leading :math:`k` columns of :math:`Q_i^k` (which form an orthonormal basis in the space spanned by leading :math:`k` columns of the matrices :math:`A_i`): - | ``orgqr_batch(queue, m, k, k, a, ...)`` - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Number of rows in the matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in the matrices :math:`A_i` (:math:`0 \le n`). - -k - Number of elementary reflectors whose product defines the matrices :math:`Q_i` (:math:`0 \le k \le n`). - -a - Array resulting after call to the Strided API of the :ref:`onemkl_lapack_geqrf_batch_buffer` function. - -lda - Leading dimension of :math:`A_i` (:math:`\text{lda} \le m`). - -stride_a - The stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -tau - Array resulting from call to the Strided API of the :ref:`onemkl_lapack_geqrf_batch_buffer` function. - -stride_tau - Stride between the beginnings of arrays :math:`\tau_i` inside the array ``tau``. - -batch_size - Specifies the number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Strided API of the :ref:`onemkl_lapack_orgqr_batch_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Batch of :math:`n` leading columns of the :math:`m \times m` orthogonal matrices :math:`Q_i`. - -.. _onemkl_lapack_orgqr_batch_usm: - -orgqr_batch (USM Version) -------------------------- - -.. rubric:: Description - -The USM version of ``orgqr_batch`` supports the group API and strided API. - -**Group API** - - | The routine generates the wholes or parts of :math:`m \times n` orthogonal matrices :math:`Q_i` of the batch of QR factorizations formed by the Group API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - | Usually :math:`Q_i` is determined from the QR factorization of an :math:`m \times p` matrix :math:`A_i` with :math:`m \ge p`. - | To compute the whole matrices :math:`Q_i`, use: - | ``orgqr_batch(queue, m, m, p, a, ...)`` - | To compute the leading :math:`p` columns of :math:`Q_i` (which form an orthonormal basis in the space spanned by the columns of :math:`A_i`): - | ``orgqr_batch(queue, m, p, p, a, ...)`` - | To compute the matrices :math:`Q_i^k` of the QR factorizations of leading :math:`k` columns of the matrices :math:`A_i`: - | ``orgqr_batch(queue, m, m, k, a, ...)`` - | To compute the leading :math:`k` columns of :math:`Q_i^k` (which form an orthonormal basis in the space spanned by leading :math:`k` columns of the matrices :math:`A_i`): - | ``orgqr_batch(queue, m, k, k, a, ...)`` - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, T **a, std::int64_t *lda, T **tau, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - - -queue - Device queue where calculations will be performed. - -m - Array of ``group_count`` :math:`m_g` parameters as previously supplied to group version of geqrf_batch function. - -n - Array of ``group_count`` :math:`n_g` parameters as previously supplied to group version of geqrf_batch function. - -k - Array of ``group_count`` :math:`k_g` parameters as previously supplied to the Group API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. The number of elementary reflectors whose product defines the matrices :math:`Q_i` (:math:`0 \le k_g \le n_g`). - -a - Array resulting after call to the Group API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -lda - Array of leading dimensions of :math:`A_i` as previously supplied to the Group API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -tau - Array resulting after call to the Group API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by Group API of the :ref:`onemkl_lapack_orgqr_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - :math:`n_g` leading columns of the :math:`m_g \times m_g` orthogonal matrices :math:`Q_i`, where :math:`g` is an index of group of parameters corresponding to :math:`Q_i`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Strided API** - - | The routine generates the wholes or parts of :math:`m \times n` orthogonal matrices :math:`Q_i` of the batch of QR factorizations formed by the Strided API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - | Usually :math:`Q_i` is determined from the QR factorization of an :math:`m \times p` matrix :math:`A_i` with :math:`m \ge p`. - | To compute the whole matrices :math:`Q_i`, use: - | ``orgqr_batch(queue, m, m, p, a, ...)`` - | To compute the leading :math:`p` columns of :math:`Q_i` (which form an orthonormal basis in the space spanned by the columns of :math:`A_i`): - | ``orgqr_batch(queue, m, p, p, a, ...)`` - | To compute the matrices :math:`Q_i^k` of the QR factorizations of leading :math:`k` columns of the matrices :math:`A_i`: - | ``orgqr_batch(queue, m, m, k, a, ...)`` - | To compute the leading :math:`k` columns of :math:`Q_i^k` (which form an orthonormal basis in the space spanned by leading :math:`k` columns of the matrices :math:`A_i`): - | ``orgqr_batch(queue, m, k, k, a, ...)`` - -.. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, std::int64_t stride_a, T *tau, std::int64_t stride_tau, std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Number of rows in the matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in the matrices :math:`A_i` (:math:`0 \le n`). - -k - Number of elementary reflectors whose product defines the matrices :math:`Q_i` (:math:`0 \le k \le n`). - -a - Array resulting after call to the Strided API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -lda - Leading dimension of :math:`A_i` (:math:`\text{lda} \le m`). - -stride_a - The stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -tau - Array resulting from call to the Strided API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -stride_tau - Stride between the beginnings of arrays :math:`\tau_i` inside the array ``tau``. - -batch_size - Specifies the number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Strided API of the :ref:`onemkl_lapack_orgqr_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Batch of :math:`n` leading columns of the :math:`m \times m` orthogonal matrices :math:`Q_i`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/orgqr_batch_scratchpad_size.rst b/docs/domains/lapack/orgqr_batch_scratchpad_size.rst deleted file mode 100644 index 444075609..000000000 --- a/docs/domains/lapack/orgqr_batch_scratchpad_size.rst +++ /dev/null @@ -1,121 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_orgqr_batch_scratchpad_size: - -orgqr_batch_scratchpad_size -=========================== - -Computes size of scratchpad memory required for the :ref:`onemkl_lapack_orgqr_batch` function. - -.. container:: section - - .. rubric:: Description - -``orgqr_batch_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -**Group API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_orgqr_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Array of ``group_count`` :math:`m_g` parameters. - -n - Array of ``group_count`` :math:`n_g` parameters. - -k - Array of ``group_count`` kg parameters. The number of elementary reflectors whose product defines the matrices :math:`Q_i` (:math:`0 \le k_g \le n_g`). - -lda - Array of leading dimensions of :math:`A_i`. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_orgqr_batch` function. - -**Strided API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_orgqr_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Number of rows in the matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in the matrices Ai (:math:`0 \le n`). - -k - Number of elementary reflectors whose product defines the matrices :math:`Q_i` (:math:`0 \le k \le n`). - -lda - Leading dimension of :math:`A_i` (:math:`\text{lda} \le m`). - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -stride_tau - Stride between the beginnings of arrays :math:`tau_i` inside the array ``tau``. - -batch_size - Number of problems in a batch. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_orgqr_batch` function. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/orgqr_scratchpad_size.rst b/docs/domains/lapack/orgqr_scratchpad_size.rst deleted file mode 100644 index 4ca2f4d6c..000000000 --- a/docs/domains/lapack/orgqr_scratchpad_size.rst +++ /dev/null @@ -1,70 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_orgqr_scratchpad_size: - -orgqr_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_orgqr` function. - -.. container:: section - - .. rubric:: Description - -``orgqr_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_orgqr` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -orgqr_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_orgqr` function will be performed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n \le m`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` (:math:`0 \le k \le n`). - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_orgqr` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - - diff --git a/docs/domains/lapack/orgtr.rst b/docs/domains/lapack/orgtr.rst deleted file mode 100644 index 757e6b9ae..000000000 --- a/docs/domains/lapack/orgtr.rst +++ /dev/null @@ -1,148 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_orgtr: - -orgtr -===== - -Generates the real orthogonal matrix :math:`Q` determined by -:ref:`onemkl_lapack_sytrd`. - -.. container:: section - - .. rubric:: Description - -``orgtr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -The routine explicitly generates the :math:`n \times n` orthogonal matrix -:math:`Q` formed by :ref:`onemkl_lapack_sytrd` when -reducing a real symmetric matrix :math:`A` to tridiagonal form: -:math:`A = QTQ^T`. Use this routine after a call to -:ref:`onemkl_lapack_sytrd`. - -orgtr (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void orgtr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. Uses the same - ``upper_lower`` as supplied to :ref:`onemkl_lapack_sytrd`. - -n - The order of the matrix :math:`Q` :math:`(0 \le n)`. - -a - The buffer ``a`` as returned by :ref:`onemkl_lapack_sytrd`. The - second dimension of ``a`` must be at least :math:`\max(1,n)`. - -lda - The leading dimension of ``a`` :math:`(n \le \text{lda})`. - -tau - The buffer ``tau`` as returned by :ref:`onemkl_lapack_sytrd`. The - dimension of ``tau`` must be at least :math:`\max(1, n-1)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_orgtr_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by the orthogonal matrix :math:`Q`. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -orgtr (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. Uses the same - ``upper_lower`` as supplied - to :ref:`onemkl_lapack_sytrd`. - -n - The order of the matrix :math:`Q` :math:`(0 \le n)`. - -a - The pointer to ``a`` as returned by - :ref:`onemkl_lapack_sytrd`. The - second dimension of ``a`` must be at least :math:`\max(1,n)`. - -lda - The leading dimension of ``a`` :math:`(n \le \text{lda})`. - -tau - The pointer to ``tau`` as returned by :ref:`onemkl_lapack_sytrd`. The - dimension of ``tau`` must be at least :math:`\max(1, n-1)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_orgtr_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by the orthogonal matrix :math:`Q`. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - diff --git a/docs/domains/lapack/orgtr_scratchpad_size.rst b/docs/domains/lapack/orgtr_scratchpad_size.rst deleted file mode 100644 index aee5516c6..000000000 --- a/docs/domains/lapack/orgtr_scratchpad_size.rst +++ /dev/null @@ -1,67 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_orgtr_scratchpad_size: - -orgtr_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_orgtr` function. - -.. container:: section - - .. rubric:: Description - -``orgtr_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_orgtr` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -orgtr_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_orgtr` function will be performed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. Uses the same - ``upper_lower`` as supplied to :ref:`onemkl_lapack_sytrd`. - -n - The order of the matrix :math:`Q` :math:`(0 \le n)`. - -lda - The leading dimension of ``a`` :math:`(n \le \text{lda})`. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_orgtr` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/ormqr.rst b/docs/domains/lapack/ormqr.rst deleted file mode 100644 index 320bfe69d..000000000 --- a/docs/domains/lapack/ormqr.rst +++ /dev/null @@ -1,207 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ormqr: - -ormqr -===== - -Multiplies a real matrix by the orthogonal matrix :math:`Q` of the QR -factorization formed by :ref:`onemkl_lapack_geqrf`. - -.. container:: section - - .. rubric:: Description - -``ormqr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -The routine multiplies a rectangular real :math:`m \times n` matrix :math:`C` by -:math:`Q` or :math:`Q^T`, where :math:`Q` is the complex unitary matrix defined -as a product of :math:`k` elementary reflectors :math:`H(i)` of order :math:`n`: -:math:`Q = H(1)^TH(2)^T ... H(k)^T` as returned by the RQ factorization routine -:ref:`onemkl_lapack_gerqf`. - -Depending on the parameters ``side`` and ``trans``, the routine can form one of -the matrix products :math:`QC`, :math:`Q^TC`, :math:`CQ`, or :math:`CQ^T` -(overwriting the result over :math:`C`). - -ormqr (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -side - If ``side = oneapi::mkl::side::left``, :math:`Q` or :math:`Q^{T}` is applied - to :math:`C` from the left. - - If ``side = oneapi::mkl::side::right``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the right. - -trans - If ``trans = oneapi::mkl::transpose::nontrans``, the routine multiplies - :math:`C` by :math:`Q`. - - If ``trans = oneapi::mkl::transpose::trans``, the routine multiplies :math:`C` - by :math:`Q^{T}`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`C` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` - - If ``side = oneapi::mkl::side::left``, :math:`0 \le k \le m` - - If ``side = oneapi::mkl::side::right``, :math:`0 \le k \le n` - -a - The buffer ``a`` as returned by :ref:`onemkl_lapack_geqrf`. - The second dimension of ``a`` must be at least :math:`\max(1,k)`. - -lda - The leading dimension of ``a``. - -tau - The buffer ``tau`` as returned by :ref:`onemkl_lapack_geqrf`. - -c - The buffer ``c`` contains the matrix :math:`C`. The second dimension of - ``c`` must be at least :math:`\max(1,n)`. - -ldc - The leading dimension of ``c``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type - ``T``. Size should not be less than the value returned by the - :ref:`onemkl_lapack_ormqr_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^{T}C`, :math:`CQ`, or - :math:`CQ^{T}` (as specified by ``side`` and ``trans``). - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -ormqr (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -side - If ``side = oneapi::mkl::side::left``, :math:`Q` or :math:`Q^{T}` is applied - to :math:`C` from the left. - - If ``side = oneapi::mkl::side::right``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the right. - -trans - If ``trans = oneapi::mkl::transpose::nontrans``, the routine multiplies - :math:`C` by :math:`Q`. - - If ``trans = oneapi::mkl::transpose::trans``, the routine multiplies :math:`C` - by :math:`Q^{T}`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`C` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` - - If ``side = oneapi::mkl::side::left``, :math:`0 \le k \le m` - - If ``side = oneapi::mkl::side::right``, :math:`0 \le k \le n` - -a - The pointer to ``a`` as returned by :ref:`onemkl_lapack_geqrf`. - The second dimension of ``a`` must be at least :math:`\max(1,k)`. - -lda - The leading dimension of ``a``. - -tau - The pointer to ``tau`` as returned by :ref:`onemkl_lapack_geqrf`. - -c - The pointer ``c`` points to the matrix :math:`C`. The second dimension of - ``c`` must be at least :math:`\max(1,n)`. - -ldc - The leading dimension of ``c``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type - ``T``. Size should not be less than the value returned by the - :ref:`onemkl_lapack_ormqr_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^{T}C`, :math:`CQ`, or - :math:`CQ^{T}` (as specified by ``side`` and ``trans``). - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` diff --git a/docs/domains/lapack/ormqr_scratchpad_size.rst b/docs/domains/lapack/ormqr_scratchpad_size.rst deleted file mode 100644 index 66cd996d8..000000000 --- a/docs/domains/lapack/ormqr_scratchpad_size.rst +++ /dev/null @@ -1,87 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ormqr_scratchpad_size: - -ormqr_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_ormqr` function. - -.. container:: section - - .. rubric:: Description - -``ormqr_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_ormqr` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -ormqr_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc, std::int64_t &scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_ormqr` function will be performed. - -side - If ``side=oneapi::mkl::side::left``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the left. - - If ``side=oneapi::mkl::side::right``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the right. - -trans - If ``trans=oneapi::mkl::transpose::nontrans``, the routine multiplies - :math:`C` by :math:`Q`. - - If ``trans=oneapi::mkl::transpose::trans``, the routine multiplies - :math:`C` by :math:`Q^{T}`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`C` (:math:`0 \le n \le m`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` (:math:`0 \le k \le n`). - -lda - The leading dimension of ``a``. - -ldc - The leading dimension of ``c``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_ormqr` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines`  - - diff --git a/docs/domains/lapack/ormrq.rst b/docs/domains/lapack/ormrq.rst deleted file mode 100644 index d49f0e3dc..000000000 --- a/docs/domains/lapack/ormrq.rst +++ /dev/null @@ -1,208 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ormrq: - -ormrq -===== - -Multiplies a real matrix by the orthogonal matrix :math:`Q` of the RQ -factorization formed by :ref:`onemkl_lapack_gerqf`. - -.. container:: section - - .. rubric:: Description - -``ormrq`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -The routine multiplies a rectangular real :math:`m \times n` matrix :math:`C` by -:math:`Q` or :math:`Q^T`, where :math:`Q` is the complex unitary matrix defined -as a product of :math:`k` elementary reflectors :math:`H(i)` of order :math:`n`: -:math:`Q = H(1)^TH(2)^T ... H(k)^T` as returned by the RQ factorization routine -:ref:`onemkl_lapack_gerqf`. - -Depending on the parameters ``side`` and ``trans``, the routine can form one of -the matrix products :math:`QC`, :math:`Q^TC`, :math:`CQ`, or :math:`CQ^T` -(overwriting the result over :math:`C`). - -ormrq (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -side - If ``side = oneapi::mkl::side::left``, :math:`Q` or :math:`Q^{T}` is applied - to :math:`C` from the left. - - If ``side = oneapi::mkl::side::right``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the right. - -trans - If ``trans = oneapi::mkl::transpose::nontrans``, the routine multiplies - :math:`C` by :math:`Q`. - - If ``trans = oneapi::mkl::transpose::trans``, the routine multiplies :math:`C` - by :math:`Q^{T}`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`C` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` - - If ``side = oneapi::mkl::side::left``, :math:`0 \le k \le m` - - If ``side = oneapi::mkl::side::right``, :math:`0 \le k \le n` - -a - The buffer ``a`` as returned by :ref:`onemkl_lapack_gerqf`. - The second dimension of ``a`` must be at least :math:`\max(1,k)`. - -lda - The leading dimension of ``a``. - -tau - The buffer ``tau`` as returned by :ref:`onemkl_lapack_gerqf`. - -c - The buffer ``c`` contains the matrix :math:`C`. The second dimension of - ``c`` must be at least :math:`\max(1,n)`. - -ldc - The leading dimension of ``c``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type - ``T``. Size should not be less than the value returned by the - :ref:`onemkl_lapack_ormrq_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^{T}C`, :math:`CQ`, or - :math:`CQ^{T}` (as specified by ``side`` and ``trans``). - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -ormrq (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -side - If ``side = oneapi::mkl::side::left``, :math:`Q` or :math:`Q^{T}` is applied - to :math:`C` from the left. - - If ``side = oneapi::mkl::side::right``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the right. - -trans - If ``trans = oneapi::mkl::transpose::nontrans``, the routine multiplies - :math:`C` by :math:`Q`. - - If ``trans = oneapi::mkl::transpose::trans``, the routine multiplies :math:`C` - by :math:`Q^{T}`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`C` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` - - If ``side = oneapi::mkl::side::left``, :math:`0 \le k \le m` - - If ``side = oneapi::mkl::side::right``, :math:`0 \le k \le n` - -a - The pointer to ``a`` as returned by :ref:`onemkl_lapack_gerqf`. - The second dimension of ``a`` must be at least :math:`\max(1,k)`. - -lda - The leading dimension of ``a``. - -tau - The pointer to ``tau`` as returned by :ref:`onemkl_lapack_gerqf`. - -c - The pointer ``c`` points to the matrix :math:`C`. The second dimension of - ``c`` must be at least :math:`\max(1,n)`. - -ldc - The leading dimension of ``c``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type - ``T``. Size should not be less than the value returned by the - :ref:`onemkl_lapack_ormrq_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^{T}C`, :math:`CQ`, or - :math:`CQ^{T}` (as specified by ``side`` and ``trans``). - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/ormrq_scratchpad_size.rst b/docs/domains/lapack/ormrq_scratchpad_size.rst deleted file mode 100644 index 6699947de..000000000 --- a/docs/domains/lapack/ormrq_scratchpad_size.rst +++ /dev/null @@ -1,81 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ormrq_scratchpad_size: - -ormrq_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_ormrq` function. - -.. container:: section - - .. rubric:: Description - -``ormrq_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_ormrq` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -ormrq_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by the ormrq function will be performed. - -side - If ``side = oneapi::mkl::side::left``, :math:`Q` or :math:`Q^T` is applied to :math:`C` from the left. - - If ``side = oneapi::mkl::side::right``, :math:`Q` or :math:`Q^T` is applied to :math:`C` from the right. - -trans - If ``trans=oneapi::mkl::transpose::nontrans``, the routine multiplies :math:`C` by :math:`Q`. - - If ``trans=oneapi::mkl::transpose::trans``, the routine multiplies :math:`C` by :math:`Q^T`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`C` (:math:`0 \le n \le m`). - -k - The number of elementary reflectors whose product defines the matrix :math:`Q` (:math:`0 \le k \le n`). - -lda - The leading dimension of ``a``. - -ldc - The leading dimension of ``c``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_ormrq` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/ormtr.rst b/docs/domains/lapack/ormtr.rst deleted file mode 100644 index d7413a1fe..000000000 --- a/docs/domains/lapack/ormtr.rst +++ /dev/null @@ -1,230 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ormtr: - -ormtr -===== - -Multiplies a real matrix by the real orthogonal matrix :math:`Q` determined by -:ref:`onemkl_lapack_sytrd`. - -.. container:: section - - .. rubric:: Description - -``ormtr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -The routine multiplies a real matrix :math:`C` by :math:`Q` or :math:`Q^{T}`, -where :math:`Q` is the orthogonal matrix :math:`Q` formed by:ref:`onemkl_lapack_sytrd` -when reducing a real symmetric matrix :math:`A` to tridiagonal form: -:math:`A = QTQ^T`. Use this routine after a call to :ref:`onemkl_lapack_sytrd`. - -Depending on the parameters side and trans, the routine can -form one of the matrix products :math:`QC`, :math:`Q^TC`, :math:`CQ`, or -:math:`CQ^T` (overwriting the result on :math:`C`). - -ormtr (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -In the descriptions below, ``r`` denotes the order of :math:`Q`: - -.. container:: tablenoborder - - .. list-table:: - :header-rows: 1 - - * - :math:`r = m` - - if ``side = side::left`` - * - :math:`r = n` - - if ``side = side::right`` - -queue - The queue where the routine should be executed. - -side - Must be either ``side::left`` or ``side::right``. - - If ``side = side::left``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the left. - - If ``side = side::right``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the right. - -upper_lower - Must be either ``uplo::upper`` or ``uplo::lower``. Uses the same - ``upper_lower`` as supplied to - :ref:`onemkl_lapack_sytrd`. - -trans - Must be either ``transpose::nontrans`` or ``transpose::trans``. - - If ``trans = transpose::nontrans``, the routine multiplies :math:`C` - by :math:`Q`. - - If ``trans = transpose::trans``, the routine multiplies :math:`C` by - :math:`Q^{T}`. - -m - The number of rows in the matrix :math:`C` :math:`(m \ge 0)`. - -n - The number of columns in the matrix :math:`C` :math:`(n \ge 0)`. - -a - The buffer ``a`` as returned by :ref:`onemkl_lapack_sytrd`. - -lda - The leading dimension of ``a`` :math:`(\max(1, r) \le \text{lda})`. - -tau - The buffer ``tau`` as returned bya :ref:`onemkl_lapack_sytrd`. The - dimension of ``tau`` must be at least :math:`\max(1, r-1)`. - -c - The buffer ``c`` contains the matrix :math:`C`. The second dimension of ``c`` - must be at least :math:`\max(1, n)`. - -ldc - The leading dimension of ``c`` :math:`(\max(1, n) \le \text{ldc})`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_ormtr_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^TC`, :math:`CQ`, or :math:`CQ^T` - (as specified by ``side`` and ``trans``). - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -ormtr (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -In the descriptions below, ``r`` denotes the order of :math:`Q`: - -.. container:: tablenoborder - - .. list-table:: - :header-rows: 1 - - * - :math:`r = m` - - if ``side = side::left`` - * - :math:`r = n` - - if ``side = side::right`` - -queue - The queue where the routine should be executed. - -side - Must be either ``side::left`` or ``side::right``. - - If ``side = side::left``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the left. - - If ``side = side::right``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the right. - -upper_lower - Must be either ``uplo::upper`` or ``uplo::lower``. Uses the same - ``upper_lower`` as supplied to :ref:`onemkl_lapack_sytrd`. - -trans - Must be either ``transpose::nontrans`` or ``transpose::trans``. - - If ``trans = transpose::nontrans``, the routine multiplies :math:`C` - by :math:`Q`. - - If ``trans = transpose::trans``, the routine multiplies :math:`C` by - :math:`Q^{T}`. - -m - The number of rows in the matrix :math:`C` :math:`(m \ge 0)`. - -n - The number of columns in the matrix :math:`C` :math:`(n \ge 0)`. - -a - The pointer to ``a`` as returned by :ref:`onemkl_lapack_sytrd`. - -lda - The leading dimension of ``a`` :math:`(\max(1, r) \le \text{lda})`. - -tau - The buffer ``tau`` as returned by :ref:`onemkl_lapack_sytrd`. The - dimension of ``tau`` must be at least :math:`\max(1, r-1)`. - -c - The pointer to memory containing the matrix :math:`C`. The second dimension of ``c`` - must be at least :math:`\max(1, n)`. - -ldc - The leading dimension of ``c`` :math:`(\max(1, n) \le \text{ldc})`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_ormtr_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^TC`, :math:`CQ`, or :math:`CQ^T` - (as specified by ``side`` and ``trans``). - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - diff --git a/docs/domains/lapack/ormtr_scratchpad_size.rst b/docs/domains/lapack/ormtr_scratchpad_size.rst deleted file mode 100644 index a71506b8e..000000000 --- a/docs/domains/lapack/ormtr_scratchpad_size.rst +++ /dev/null @@ -1,105 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ormtr_scratchpad_size: - -ormtr_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_ormtr` function. - -.. container:: section - - .. rubric:: Description - -``ormtr_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_ormtr` function should be able to hold. -Calls to this routine must specify the template parameter -explicitly. - -ormtr_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) - } - -.. container:: section - - .. rubric:: Input Parameters - -In the descriptions below, ``r`` denotes the order of :math:`Q`: - -.. container:: tablenoborder - - .. list-table:: - :header-rows: 1 - - * - :math:`r = m` - - if ``side = side::left`` - * - :math:`r = n` - - if ``side = side::right`` - -queue - Device queue where calculations by :ref:`onemkl_lapack_ormtr` function will be performed. - -side - Must be either ``side::left`` or ``side::right``. - - If ``side = side::left``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the left. - - If ``side = side::right``, :math:`Q` or :math:`Q^{T}` is - applied to :math:`C` from the right. - -upper_lower - Must be either ``uplo::upper`` or ``uplo::lower``. Uses the - same ``upper_lower`` as supplied to - :ref:`onemkl_lapack_sytrd`. - -trans - Must be either ``transpose::nontrans`` or ``transpose::trans``. - - If ``trans = transpose::nontrans``, the routine multiplies - :math:`C` by :math:`Q`. - - If ``trans = transpose::trans``, the routine multiplies :math:`C` - by :math:`Q^{T}`. - -m - The number of rows in the matrix :math:`C` :math:`(m \ge 0)`. - -n - The number of rows in the matrix :math:`C` :math:`(n \ge 0)`. - -lda - The leading dimension of ``a`` :math:`(\max(1, r) \le \text{lda})`. - -ldc - The leading dimension of ``c`` :math:`(\max(1, n) \le \text{ldc})`. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_ormtr` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/potrf.rst b/docs/domains/lapack/potrf.rst deleted file mode 100644 index dd044e1f2..000000000 --- a/docs/domains/lapack/potrf.rst +++ /dev/null @@ -1,172 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_potrf: - -potrf -===== - -Computes the Cholesky factorization of a symmetric (Hermitian) -positive-definite matrix. - -.. container:: section - - .. rubric:: Description - -``potrf`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -The routine forms the Cholesky factorization of a symmetric -positive-definite or, for complex data, Hermitian positive-definite -matrix :math:`A`: - - .. list-table:: - :header-rows: 1 - - * - :math:`A` = :math:`U^{T}U` for real data, :math:`A = U^{H}U` for complex data - - if upper_lower=\ ``oneapi::mkl::uplo::upper`` - * - :math:`A` = :math:`LL^{T}` for real data, :math:`A = LL^{H}` for complex data - - if upper_lower=\ ``oneapi::mkl::uplo::lower`` - -where :math:`L` is a lower triangular matrix and :math:`U` is upper -triangular. - -potrf (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void potrf(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates whether the upper or lower triangular part of :math:`A` is - stored and how :math:`A` is factored: - - If upper_lower=\ ``oneapi::mkl::uplo::upper``, the array ``a`` stores the - upper triangular part of the matrix :math:`A`, and the strictly lower - triangular part of the matrix is not referenced. - - If upper_lower=\ ``oneapi::mkl::uplo::lower``, the array ``a`` stores the - lower triangular part of the matrix :math:`A`, and the strictly upper - triangular part of the matrix is not referenced. - -n - Specifies the order of the matrix :math:`A` (:math:`0 \le n`). - -a - Buffer holding input matrix :math:`A`. The buffer ``a`` contains either - the upper or the lower triangular part of the matrix :math:`A` (see - upper_lower). The second dimension of ``a`` must be at least - :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_potrf_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - The buffer ``a`` is overwritten by the Cholesky factor :math:`U` or :math:`L`, - as specified by ``upper_lower``. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -potrf (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, T *a, std::int64_t lda, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates whether the upper or lower triangular part of :math:`A` is - stored and how :math:`A` is factored: - - If upper_lower=\ ``oneapi::mkl::uplo::upper``, the array ``a`` stores the - upper triangular part of the matrix :math:`A`, and the strictly lower - triangular part of the matrix is not referenced. - - If upper_lower=\ ``oneapi::mkl::uplo::lower``, the array ``a`` stores the - lower triangular part of the matrix :math:`A`, and the strictly upper - triangular part of the matrix is not referenced. - -n - Specifies the order of the matrix :math:`A` (:math:`0 \le n`). - -a - Pointer to input matrix :math:`A`. The array ``a`` contains either - the upper or the lower triangular part of the matrix :math:`A` (see - upper_lower). The second dimension of ``a`` must be at least - :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_potrf_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - The memory pointer to by pointer ``a`` is overwritten by the Cholesky factor :math:`U` or :math:`L`, - as specified by ``upper_lower``. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - - diff --git a/docs/domains/lapack/potrf_batch.rst b/docs/domains/lapack/potrf_batch.rst deleted file mode 100644 index 872886cd7..000000000 --- a/docs/domains/lapack/potrf_batch.rst +++ /dev/null @@ -1,239 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_potrf_batch: - -potrf_batch -=========== - -Computes the LU factorizations of a batch of general matrices. - -.. container:: section - - .. rubric:: Description - -``potrf_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_lapack_potrf_batch_buffer: - -potrf_batch (Buffer Version) ----------------------------- - -.. container:: section - - .. rubric:: Description - -The buffer version of ``potrf_batch`` supports only the strided API. - -**Strided API** - - | The routine forms the Cholesky factorizations of a symmetric positive-definite or, for complex data, Hermitian positive-definite matrices :math:`A_i`, :math:`i \in \{1...batch\_size\}`: - | :math:`A_i = U_i^TU_i` for real data, :math:`A_i = U_i^HU_i` for complex data if ``uplo = mkl::uplo::upper``, - | :math:`A_i = L_iL_i^T` for real data, :math:`A_i = L_iL_i^H` for complex data if ``uplo = mkl::uplo::lower``, - | where :math:`L_i` is a lower triangular matrix and :math:`U_i` is upper triangular. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void potrf_batch(sycl::queue &queue, mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -uplo - | Indicates whether the upper or lower triangular part of :math:`A_i` is stored and how :math:`A_i` is factored: - | If ``uplo = mkl::uplo::upper``, the array ``a`` stores the upper triangular parts of the matrices :math:`A_i`, - | If ``uplo = mkl::uplo::lower``, the array ``a`` stores the lower triangular parts of the matrices :math:`A_i`. - -n - Order of the matrices :math:`A_i`, (:math:`0 \le n`). - -a - Array containing batch of input matrices :math:`A_i`, each of :math:`A_i` being of size :math:`\text{lda} \cdot n` and holding either upper or lower triangular parts of the matrices :math:`A_i` (see ``uplo``). - -lda - Leading dimension of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Strided API of the :ref:`onemkl_lapack_potrf_batch_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Cholesky factors :math:`U_i` or :math:`L_i`, as specified by ``uplo``. - -.. _onemkl_lapack_potrf_batch_usm: - -potrf_batch (USM Version) -------------------------- - -.. container:: section - - .. rubric:: Description - -The USM version of ``potrf_batch`` supports the group API and strided API. - -**Group API** - - | The routine forms the Cholesky factorizations of symmetric positive-definite or, for complex data, Hermitian positive-definite matrices :math:`A_i`, :math:`i \in \{1...batch\_size\}`: - | :math:`A_i = U_i^TU_i` for real data (:math:`A_i = U_i^HU_i` for complex), if :math:`\text{uplo}_g` is ``mkl::uplo::upper``, - | :math:`A_i = L_iL_i^T` for real data (:math:`A_i = L_iL_i^H` for complex), if :math:`\text{uplo}_g` is ``mkl::uplo::lower``, - | where :math:`L_i` is a lower triangular matrix and :math:`U_i` is upper triangular, :math:`g` is an index of group of parameters corresponding to :math:`A_i`, and total number of problems to solve, ``batch_size``, is a sum of sizes of all of the groups of parameters as provided by ``group_sizes`` array - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event potrf_batch(sycl::queue &queue, mkl::uplo *uplo, std::int64_t *n, T **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -uplo - | Array of ``group_count`` :math:`\text{uplo}_g` parameters. Each :math:`\text{uplo}_g` indicates whether the upper or lower triangular parts of the input matrices are provided: - | If :math:`\text{uplo}_g` is ``mkl::uplo::upper``, input matrices from array ``a`` belonging to group :math:`g` store the upper triangular parts, - | If :math:`\text{uplo}_g` is ``mkl::uplo::lower``, input matrices from array ``a`` belonging to group :math:`g` store the lower triangular parts. - -n - Array of ``group_count`` :math:`n_g` parameters. Each :math:`n_g` specifies the order of the input matrices from array a belonging to group :math:`g`. - -a - Array of ``batch_size`` pointers to input matrices :math:`A_i`, each being of size :math:`\text{lda}_g \cdot n_g` (:math:`g` is an index of group to which :math:`A_i` belongs to) and holding either upper or lower triangular part as specified by :math:`\text{uplo}_g`. - -lda - Array of ``group_count`` :math:`\text{lda}_g` parameters. Each :math:`\text{lda}_g` specifies the leading dimensions of the matrices from a belonging to group :math:`g`. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of group_count integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Group API of the :ref:`onemkl_lapack_potrf_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Cholesky factors :math:`U_i` or :math:`L_i`, as specified by :math:`\text{uplo}_g` from corresponding group of parameters. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Strided API** - - | The routine forms the Cholesky factorizations of a symmetric positive-definite or, for complex data, Hermitian positive-definite matrices :math:`A_i`, :math:`i \in \{1...batch\_size\}`: - | :math:`A_i = U_i^TU_i` for real data, :math:`A_i = U_i^HU_i` for complex data if ``uplo = mkl::uplo::upper``, - | :math:`A_i = L_iL_i^T` for real data, :math:`A_i = L_iL_i^H` for complex data if ``uplo = mkl::uplo::lower``, - | where :math:`L_i` is a lower triangular matrix and :math:`U_i` is upper triangular. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event potrf_batch(sycl::queue &queue, mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -uplo - | Indicates whether the upper or lower triangular part of :math:`A_i` is stored and how :math:`A_i` is factored: - | If ``uplo = mkl::uplo::upper``, the array ``a`` stores the upper triangular parts of the matrices :math:`A_i`, - | If ``uplo = mkl::uplo::lower``, the array ``a`` stores the lower triangular parts of the matrices :math:`A_i`. - -n - Order of the matrices :math:`A_i`, (:math:`0 \le n`). - -a - Array containing batch of input matrices :math:`A_i`, each of :math:`A_i` being of size :math:`\text{lda} \cdot n` and holding either upper or lower triangular parts of the matrices :math:`A_i` (see ``uplo``). - -lda - Leading dimension of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Strided API of the :ref:`onemkl_lapack_potrf_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Cholesky factors :math:`U_i` or :math:`L_i`, as specified by ``uplo``. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/potrf_batch_scratchpad_size.rst b/docs/domains/lapack/potrf_batch_scratchpad_size.rst deleted file mode 100644 index 4a1b57bf8..000000000 --- a/docs/domains/lapack/potrf_batch_scratchpad_size.rst +++ /dev/null @@ -1,120 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_potrf_batch_scratchpad_size: - -potrf_batch_scratchpad_size -=========================== - -Computes size of scratchpad memory required for the :ref:`onemkl_lapack_potrf_batch` function. - -.. container:: section - - .. rubric:: Description - -``potrf_batch_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -**Group API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_potrf_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -uplo - | Array of ``group_count`` :math:`\text{uplo}_g` parameters. - | Each of :math:`\text{uplo}_g` indicates whether the upper or lower triangular parts of the input matrices are provided: - | If :math:`\text{uplo}_g` is ``mkl::uplo::upper``, input matrices from array ``a`` belonging to group :math:`g` store the upper triangular parts, - | If :math:`\text{uplo}_g` is ``mkl::uplo::lower``, input matrices from array ``a`` belonging to group :math:`g` store the lower triangular parts. - -n - | Array of ``group_count`` :math:`n_g` parameters. - | Each ng specifies the order of the input matrices belonging to group :math:`g`. - -lda - | Array of ``group_count`` :math:`\text{lda}_g` parameters. - | Each ldag specifies the leading dimensions of the matrices belonging to group :math:`g`. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_potrf_batch` function. - -**Strided API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_potrf_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. -uplo - | Indicates whether the upper or lower triangular part of :math:`A_i` is stored and how :math:`A_i` is factored: - | If ``uplo = mkl::uplo::upper``, the array ``a`` stores the upper triangular parts of the matrices :math:`A_i`, - | If ``uplo = mkl::uplo::lower``, the array ``a`` stores the lower triangular parts of the matrices :math:`A_i`. - -n - Order of the matrices :math:`A_i`, (:math:`0 \le n`). - -lda - Leading dimension of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch. - -batch_size - Number of problems in a batch. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_potrf_batch` function. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/potrf_scratchpad_size.rst b/docs/domains/lapack/potrf_scratchpad_size.rst deleted file mode 100644 index c1e423fa2..000000000 --- a/docs/domains/lapack/potrf_scratchpad_size.rst +++ /dev/null @@ -1,77 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_potrf_scratchpad_size: - -potrf_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_potrf` function. - -.. container:: section - - .. rubric:: Description - -``potrf_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_potrf` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -potrf_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_potrf` function will be performed. - -upper_lower - Indicates whether the upper or lower triangular part of :math:`A` is - stored and how :math:`A` is factored: - - If ``upper_lower = oneapi::mkl::uplo::upper``, the array ``a`` stores the - upper triangular part of the matrix :math:`A`, and the strictly lower - triangular part of the matrix is not referenced. - - If ``upper_lower = oneapi::mkl::uplo::lower``, the array ``a`` stores the - lower triangular part of the matrix :math:`A`, and the strictly upper - triangular part of the matrix is not referenced. - -n - Specifies the order of the matrix :math:`A` (:math:`0 \le n`). - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_potrf` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - - diff --git a/docs/domains/lapack/potri.rst b/docs/domains/lapack/potri.rst deleted file mode 100644 index 3e8de09d7..000000000 --- a/docs/domains/lapack/potri.rst +++ /dev/null @@ -1,144 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_potri: - -potri -===== - -Computes the inverse of a symmetric (Hermitian) positive-definite -matrix using the Cholesky factorization. - -.. container:: section - - .. rubric:: Description - -``potri`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -The routine computes the inverse :math:`A^{-1}` of a symmetric positive -definite or, for complex flavors, Hermitian positive-definite matrix -:math:`A`. Before calling this routine, call :ref:`onemkl_lapack_potrf` -to factorize :math:`A`. - -potri (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void potri(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates how the input matrix :math:`A` has been factored: - - If ``upper_lower = oneapi::mkl::uplo::upper``, the upper triangle of :math:`A` is stored. - - If ``upper_lower = oneapi::mkl::uplo::lower``, the lower triangle of :math:`A` is stored. - -n - Specifies the order of the matrix :math:`A` (:math:`0 \le n`). - -a - Contains the factorization of the matrix :math:`A`, as returned by :ref:`onemkl_lapack_potrf`. The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_potri_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by the upper or lower triangle of the inverse of :math:`A`. Specified by ``upper_lower``. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -potri (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, T *a, std::int64_t lda, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates how the input matrix :math:`A` has been factored: - - If ``upper_lower = oneapi::mkl::uplo::upper``, the upper triangle of :math:`A` is stored. - - If ``upper_lower = oneapi::mkl::uplo::lower``, the lower triangle of :math:`A` is stored. - -n - Specifies the order of the matrix :math:`A` (:math:`0 \le n`). - -a - Contains the factorization of the matrix :math:`A`, as returned by :ref:`onemkl_lapack_potrf`. The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_potri_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by the upper or lower triangle of the inverse of :math:`A`. Specified by ``upper_lower``. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - - diff --git a/docs/domains/lapack/potri_scratchpad_size.rst b/docs/domains/lapack/potri_scratchpad_size.rst deleted file mode 100644 index 07b14a341..000000000 --- a/docs/domains/lapack/potri_scratchpad_size.rst +++ /dev/null @@ -1,71 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_potri_scratchpad_size: - -potri_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_potri` function. - -.. container:: section - - .. rubric:: Description - -``potri_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_potri` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -potri_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_potri` function will be performed. - -upper_lower - Indicates how the input matrix :math:`A` has been factored: - - If ``upper_lower = oneapi::mkl::uplo::upper``, the upper triangle of :math:`A` is stored. - - If ``upper_lower = oneapi::mkl::uplo::lower``, the lower triangle of :math:`A` is stored. - -n - Specifies the order of the matrix :math:`A` (:math:`0 \le n`). - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_potri` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/potrs.rst b/docs/domains/lapack/potrs.rst deleted file mode 100644 index 26fbc7c31..000000000 --- a/docs/domains/lapack/potrs.rst +++ /dev/null @@ -1,177 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_potrs: - -potrs -===== - -Solves a system of linear equations with a Cholesky-factored -symmetric (Hermitian) positive-definite coefficient matrix. - -.. container:: section - - .. rubric:: Description - -``potrs`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -The routine solves for :math:`X` the system of linear equations -:math:`AX = B` with a symmetric positive-definite or, for complex data, -Hermitian positive-definite matrix :math:`A`, given the Cholesky -factorization of :math:`A`: - -.. list-table:: - :header-rows: 1 - - * - :math:`A = U^TU` for real data, :math:`A = U^HU` for complex data - - if ``upper_lower=oneapi::mkl::uplo::upper`` - * - :math:`A = LL^T` for real data, :math:`A = LL^H` for complex data - - if ``upper_lower=oneapi::mkl::uplo::lower`` - -where :math:`L` is a lower triangular matrix and :math:`U` is upper -triangular. The system is solved with multiple right-hand sides -stored in the columns of the matrix :math:`B`. - -Before calling this routine, you must call :ref:`onemkl_lapack_potrf` to compute -the Cholesky factorization of :math:`A`. - -potrs (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void potrs(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates how the input matrix has been factored: - - If ``upper_lower = oneapi::mkl::uplo::upper``, the upper triangle :math:`U` of :math:`A` is stored, where :math:`A` = :math:`U^{T}`U` for real data, :math:`A` = :math:`U^{H}U` for complex data. - - If ``upper_lower = oneapi::mkl::uplo::lower``, the lower triangle :math:`L` of :math:`A` is stored, where :math:`A` = :math:`LL^{T}` for real data, :math:`A` = :math:`LL^{H}` for complex data. - -n - The order of matrix :math:`A` (:math:`0 \le n`).\ - -nrhs - The number of right-hand sides (:math:`0 \le \text{nrhs}`). - -a - Buffer containing the factorization of the matrix A, as returned by :ref:`onemkl_lapack_potrf`. The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -b - The array ``b`` contains the matrix :math:`B` whose columns are the right-hand sides for the systems of equations. The second dimension of ``b`` must be at least :math:`\max(1,\text{nrhs})`. - -ldb - The leading dimension of ``b``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_potrs_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -b - Overwritten by the solution matrix :math:`X`. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -potrs (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t nrhs, T *a, std::int64_t lda, T *b, std::int64_t ldb, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates how the input matrix has been factored: - - If ``upper_lower = oneapi::mkl::uplo::upper``, the upper triangle :math:`U` of :math:`A` is stored, where :math:`A` = :math:`U^{T}U` for real data, :math:`A` = :math:`U^{H}U` for complex data. - - If ``upper_lower = oneapi::mkl::uplo::lower``, the lower triangle :math:`L` of :math:`A` is stored, where :math:`A` = :math:`LL^{T}` for real data, :math:`A` = :math:`LL^{H}` for complex data. - -n - The order of matrix :math:`A` (:math:`0 \le n`).\ - -nrhs - The number of right-hand sides (:math:`0 \le \text{nrhs}`). - -a - Pointer to array containing the factorization of the matrix :math:`A`, as returned by :ref:`onemkl_lapack_potrf`. The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -b - The array ``b`` contains the matrix :math:`B` whose columns are the right-hand sides for the systems of equations. The second dimension of ``b`` must be at least :math:`\max(1,\text{nrhs})`. - -ldb - The leading dimension of ``b``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_potrs_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -b - Overwritten by the solution matrix :math:`X`. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - - diff --git a/docs/domains/lapack/potrs_batch.rst b/docs/domains/lapack/potrs_batch.rst deleted file mode 100644 index f9c8f6477..000000000 --- a/docs/domains/lapack/potrs_batch.rst +++ /dev/null @@ -1,276 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_potrs_batch: - -potrs_batch -=========== - -Computes the LU factorizations of a batch of general matrices. - -.. container:: section - - .. rubric:: Description - -``potrs_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_lapack_potrs_batch_buffer: - -potrs_batch (Buffer Version) ----------------------------- - -.. container:: section - - .. rubric:: Description - -The buffer version of ``potrs_batch`` supports only the strided API. - -**Strided API** - - | The routine solves for :math:`X_i` the systems of linear equations :math:`A_iX_i = B_i` with a symmetric positive-definite or, for complex data, Hermitian positive-definite matrices :math:`A_i`, given the Cholesky factorization of :math:`A_i`, :math:`i \in \{1...batch\_size\}`: - | :math:`A_i = U_i^TU_i` for real data, :math:`A_i = U_i^HU_i` for complex data if ``uplo = mkl::uplo::upper``, - | :math:`A_i = L_iL_i^T` for real data, :math:`A_i = L_iL_i^H` for complex data if ``uplo = mkl::uplo::lower``, - | where :math:`L_i` is a lower triangular matrix and :math:`U_i` is upper triangular. - | The systems are solved with multiple right-hand sides stored in the columns of the matrices :math:`B_i`. - | Before calling this routine, matrices :math:`A_i` should be factorized by call to the Strided API of the :ref:`onemkl_lapack_potrf_batch_buffer` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void potrs_batch(sycl::queue &queue, mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -uplo - | Indicates how the input matrices have been factored: - | If ``uplo = mkl::uplo::upper``, the upper triangle :math:`U_i` of :math:`A_i` is stored, where :math:`A_i = U_i^TU_i` for real data, :math:`A_i = U_i^HU_i` for complex data. - | If ``uplo = mkl::uplo::lower``, the upper triangle :math:`L_i` of :math:`A_i` is stored, where :math:`A_i = L_iL_i^T` for real data, :math:`A_i = L_iL_i^H` for complex data. - -n - The order of matrices :math:`A_i` (:math:`0 \le n`). - -nrhs - The number of right-hand sides (:math:`0 \le \text{nrhs}`). - -a - Array containing batch of factorizations of the matrices :math:`A_i`, as returned by the Strided API of the :ref:`onemkl_lapack_potrf_batch_buffer` function. - -lda - Leading dimension of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices inside the batch array ``a``. - -b - Array containing batch of matrices :math:`B_i` whose columns are the right-hand sides for the systems of equations. - -ldb - Leading dimension of :math:`B_i`. - -stride_b - Stride between the beginnings of matrices :math:`B_i` inside the batch array ``b``. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Strided API of the :ref:`onemkl_lapack_potrs_batch_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -b - Solution matrices :math:`X_i`. - -.. _onemkl_lapack_potrs_batch_usm: - -potrs_batch (USM Version) -------------------------- - -.. container:: section - - .. rubric:: Description - -The USM version of ``potrs_batch`` supports the group API and strided API. - -**Group API** - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event potrs_batch(sycl::queue &queue, mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, T **a, std::int64_t *lda, T **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -uplo - | Array of ``group_count`` :math:`\text{uplo}_g` parameters. - | Each of :math:`\text{uplo}_g` indicates whether the upper or lower triangular parts of the input matrices are provided: - | If :math:`\text{uplo}_g` is ``mkl::uplo::upper``, input matrices from array ``a`` belonging to group :math:`g` store the upper triangular parts, - | If :math:`\text{uplo}_g` is ``mkl::uplo::lower``, input matrices from array ``a`` belonging to group :math:`g` store the lower triangular parts. - -n - | Array of ``group_count`` :math:`n_g` parameters. - | Each :math:`n_g` specifies the order of the input matrices from array ``a`` belonging to group :math:`g`. - -nrhs - | Array of ``group_count`` :math:`\text{nrhs}_g` parameters. - | Each :math:`\text{nrhs}_g` specifies the number of right-hand sides supplied for group :math:`g` in corresponding part of array ``b``. - -a - Array of ``batch_size`` pointers to Cholesky factored matrices :math:`A_i` as returned by the Group API of the :ref:`onemkl_lapack_potrf_batch_usm` function. - -lda - | Array of ``group_count`` :math:`\text{lda}_g` parameters. - | Each :math:`\text{lda}_g` specifies the leading dimensions of the matrices from ``a`` belonging to group :math:`g`. - -b - Array of ``batch_size`` pointers to right-hand side matrices :math:`B_i`, each of size :math:`\text{ldb}_g \cdot \text{nrhs}_g`, where :math:`g` is an index of group corresponding to :math:`B_i`. - -ldb - | Array of ``group_count`` :math:`\text{ldb}_g` parameters. - | Each :math:`\text{ldb}_g` specifies the leading dimensions of the matrices from ``b`` belonging to group :math:`g`. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Group API of the :ref:`onemkl_lapack_potrs_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -b - Solution matrices :math:`X_i`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Strided API** - - | The routine solves for :math:`X_i` the systems of linear equations :math:`A_iX_i = B_i` with a symmetric positive-definite or, for complex data, Hermitian positive-definite matrices :math:`A_i`, given the Cholesky factorization of :math:`A_i`, :math:`i \in \{1...batch\_size\}`: - | :math:`A_i = U_i^TU_i` for real data, :math:`A_i = U_i^HU_i` for complex data if ``uplo = mkl::uplo::upper``, - | :math:`A_i = L_iL_i^T` for real data, :math:`A_i = L_iL_i^H` for complex data if ``uplo = mkl::uplo::lower``, - | where :math:`L_i` is a lower triangular matrix and :math:`U_i` is upper triangular. - | The systems are solved with multiple right-hand sides stored in the columns of the matrices :math:`B_i`. - | Before calling this routine, matrices :math:`A_i` should be factorized by call to the Strided API of the :ref:`onemkl_lapack_potrf_batch_usm` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event potrs_batch(sycl::queue &queue, mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, T *a, std::int64_t lda, std::int64_t stride_a, T *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -uplo - | Indicates how the input matrices have been factored: - | If ``uplo = mkl::uplo::upper``, the upper triangle :math:`U_i` of :math:`A_i` is stored, where :math:`A_i = U_i^TU_i` for real data, :math:`A_i = U_i^HU_i` for complex data. - | If ``uplo = mkl::uplo::lower``, the upper triangle :math:`L_i` of :math:`A_i` is stored, where :math:`A_i = L_iL_i^T` for real data, :math:`A_i = L_iL_i^H` for complex data. - -n - The order of matrices :math:`A_i` (:math:`0 \le n`). - -nrhs - The number of right-hand sides (:math:`0 \le nrhs`). - -a - Array containing batch of factorizations of the matrices :math:`A_i`, as returned by the Strided API of the :ref:`onemkl_lapack_potrf_batch_usm` function. - -lda - Leading dimension of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices inside the batch array ``a``. - -b - Array containing batch of matrices :math:`B_i` whose columns are the right-hand sides for the systems of equations. - -ldb - Leading dimension of :math:`B_i`. - -stride_b - Stride between the beginnings of matrices :math:`B_i` inside the batch array ``b``. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by the Strided API of the :ref:`onemkl_lapack_potrs_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -b - Solution matrices :math:`X_i`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/potrs_batch_scratchpad_size.rst b/docs/domains/lapack/potrs_batch_scratchpad_size.rst deleted file mode 100644 index 6136fd8b2..000000000 --- a/docs/domains/lapack/potrs_batch_scratchpad_size.rst +++ /dev/null @@ -1,136 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_potrs_batch_scratchpad_size: - -potrs_batch_scratchpad_size -=========================== - -Computes size of scratchpad memory required for the :ref:`onemkl_lapack_potrs_batch` function. - -.. container:: section - - .. rubric:: Description - -``potrs_batch_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -**Group API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_potrs_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -uplo - | Array of ``group_count`` :math:`\text{uplo}_g` parameters. - | Each of :math:`\text{uplo}_g` indicates whether the upper or lower triangular parts of the input matrices are provided: - | If :math:`\text{uplo}_g` is ``mkl::uplo::upper``, input matrices from array ``a`` belonging to group :math:`g` store the upper triangular parts, - | If :math:`\text{uplo}_g` is ``mkl::uplo::lower``, input matrices from array ``a`` belonging to group :math:`g` store the lower triangular parts. - -n - | Array of ``group_count`` :math:`n_g` parameters. - | Each :math:`n_g` specifies the order of the input matrices belonging to group :math:`g`. - -nrhs - | Array of ``group_count`` :math:`\text{nrhs}_g` parameters. - | Each :math:`rhs_g` specifies the number of right-hand sides supplied for group :math:`g`. - -lda - | Array of ``group_count`` :math:`\text{lda}_g` parameters. - | Each :math:`\text{lda}_g` specifies the leading dimensions of the matrices belonging to group :math:`g`. - -ldb - | Array of ``group_count`` :math:`\text{ldb}_g` parameters. - | Each :math:`\text{ldb}_g` specifies the leading dimensions of the matrices belonging to group :math:`g`. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes Array of group_count integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_potrs_batch` function. - -**Strided API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_potrs_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. -uplo - | Indicates how the input matrices have been factored: - | If ``uplo = mkl::uplo::upper``, the upper triangle :math:`U_i` of :math:`A_i` is stored, where :math:`A_i = U_i^TU_i` for real data, :math:`A_i = U_i^HU_i` for complex data. - | If ``uplo = mkl::uplo::lower``, the upper triangle :math:`L_i` of :math:`A_i` is stored, where :math:`A_i = L_iL_i^T` for real data, :math:`A_i = L_iL_i^H` for complex data. - -n - Order of matrices :math:`A_i` (:math:`0 \le n`). - -nrhs - Number of right-hand sides (:math:`0 \le \text{nrhs}`). - -lda - Leading dimension of :math:`A_i`. - -stride_a - Stride between the beginnings of matrices inside the batch array ``a``. - -ldb - Leading dimensions of :math:`B_i`. - -stride_b - Stride between the beginnings of matrices :math:`B_i` inside the batch array ``b``. - -batch_size - Number of problems in a batch. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_potrs_batch` function. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/potrs_scratchpad_size.rst b/docs/domains/lapack/potrs_scratchpad_size.rst deleted file mode 100644 index 125d68944..000000000 --- a/docs/domains/lapack/potrs_scratchpad_size.rst +++ /dev/null @@ -1,77 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_potrs_scratchpad_size: - -potrs_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_potrs` function. - -.. container:: section - - .. rubric:: Description - -``potrs_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_potrs` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -potrs_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_potrs` function will be performed. - -upper_lower - Indicates how the input matrix has been factored: - - If ``upper_lower = oneapi::mkl::uplo::upper``, the upper triangle :math:`U` of :math:`A` is stored, where :math:`A = U^{T}U` for real data, :math:`A = U^{H}U` for complex data. - - If ``upper_lower = oneapi::mkl::uplo::lower``, the lower triangle :math:`L` of :math:`A` is stored, where :math:`A = LL^{T}` for real data, :math:`A = LL^{H}` for complex data. - -n - The order of matrix :math:`A` (:math:`0 \le n`). - -nrhs - The number of right-hand sides (:math:`0 \le nrhs`). - -lda - The leading dimension of ``a``. - -ldb - The leading dimension of ``b``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_potrs` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/syevd.rst b/docs/domains/lapack/syevd.rst deleted file mode 100644 index 93df0ba27..000000000 --- a/docs/domains/lapack/syevd.rst +++ /dev/null @@ -1,186 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_syevd: - -syevd -===== - -Computes all eigenvalues and, optionally, all eigenvectors of a real -symmetric matrix using divide and conquer algorithm. - -.. container:: section - - .. rubric:: Description - -``syevd`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -The routine computes all the eigenvalues, and optionally all the -eigenvectors, of a real symmetric matrix :math:`A`. In other words, it -can compute the spectral factorization of :math:`A` as: :math:`A = Z\lambda Z^T`. - -Here :math:`\Lambda` is a diagonal matrix whose diagonal elements are the -eigenvalues :math:`\lambda_i`, and :math:`Z` is the orthogonal matrix whose -columns are the eigenvectors :math:`z_{i}`. Thus, - -:math:`A z_i = \lambda_i z_i` for :math:`i = 1, 2, ..., n`. - -If the eigenvectors are requested, then this routine uses a divide -and conquer algorithm to compute eigenvalues and eigenvectors. -However, if only eigenvalues are required, then it uses the -Pal-Walker-Kahan variant of the QL or QR algorithm. - -syevd (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void syevd(sycl::queue &queue, jobz jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = job::upper``, a stores the upper triangular - part of :math:`A`. - - If ``upper_lower = job::lower``, a stores the lower triangular - part of :math:`A`. - -n - The order of the matrix :math:`A` (:math:`0 \le n`). - -a - The buffer ``a``, size (``lda,*``). The buffer ``a`` contains the matrix - :math:`A`. The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. Must be at least :math:`\max(1,n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_syevd_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - If ``jobz = job::vec``, then on exit this buffer is overwritten by - the orthogonal matrix :math:`Z` which contains the eigenvectors of - :math:`A`. - -w - Buffer, size at least :math:`n`. Contains the eigenvalues - of the matrix :math:`A` in ascending order. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -syevd (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event syevd(sycl::queue &queue, jobz jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, T *a, std::int64_t lda, T *w, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = job::upper``, a stores the upper triangular - part of :math:`A`. - - If ``upper_lower = job::lower``, a stores the lower triangular - part of :math:`A`. - -n - The order of the matrix :math:`A` (:math:`0 \le n`). - -a - Pointer to array containing :math:`A`, size (``lda,*``). - The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. Must be at least :math:`\max(1,n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_syevd_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - If ``jobz = job::vec``, then on exit this array is overwritten by - the orthogonal matrix :math:`Z` which contains the eigenvectors of - :math:`A`. - -w - Pointer to array of size at least :math:`n`. Contains the eigenvalues - of the matrix :math:`A` in ascending order. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/syevd_scratchpad_size.rst b/docs/domains/lapack/syevd_scratchpad_size.rst deleted file mode 100644 index e9d08fc18..000000000 --- a/docs/domains/lapack/syevd_scratchpad_size.rst +++ /dev/null @@ -1,81 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_syevd_scratchpad_size: - -syevd_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_syevd` function. - -.. container:: section - - .. rubric:: Description - -``syevd_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_syevd` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -syevd_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_syevd` function will be performed. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = job::upper``, a stores the upper triangular - part of :math:`A`. - - If ``upper_lower = job::lower``, a stores the lower triangular - part of :math:`A`. - -n - The order of the matrix :math:`A` (:math:`0 \le n`). - -lda - The leading dimension of ``a``. Currently ``lda`` is not referenced in - this function. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_syevd` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/sygvd.rst b/docs/domains/lapack/sygvd.rst deleted file mode 100644 index 797fdf3a6..000000000 --- a/docs/domains/lapack/sygvd.rst +++ /dev/null @@ -1,249 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_sygvd: - -sygvd -===== - -Computes all eigenvalues and, optionally, eigenvectors of a real -generalized symmetric definite eigenproblem using a divide and -conquer method. - -.. container:: section - - .. rubric:: Description - -``sygvd`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -The routine computes all the eigenvalues, and optionally, the -eigenvectors of a real generalized symmetric-definite eigenproblem, -of the form - -:math:`Ax = \lambda Bx`, :math:`ABx = \lambda x`, or :math:`BAx = \lambda x` . - -Here :math:`A` and :math:`B` are assumed to be symmetric and :math:`B` is also -positive definite. - -It uses a divide and conquer algorithm. - -sygvd (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -itype - Must be 1 or 2 or 3. Specifies the problem type to be solved: - - if :math:`\text{itype} = 1`, the problem type is :math:`Ax = \lambda Bx`; - - if :math:`\text{itype} = 2`, the problem type is :math:`ABx = \lambda x`; - - if :math:`\text{itype} = 3`, the problem type is :math:`BAx = \lambda x`. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = job::upper``, ``a`` and ``b`` store the upper - triangular part of :math:`A` and :math:`B`. - - If ``upper_lower = job::lower``, ``a`` and ``b`` stores the lower - triangular part of :math:`A` and :math:`B`. - -n - The order of the matrices :math:`A` and :math:`B` :math:`(0 \le n)`. - -a - Buffer, size a\ ``(lda,*)`` contains the upper or lower triangle - of the symmetric matrix :math:`A`, as specified by ``upper_lower``. The - second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``; at least :math:`\max(1, n)`. - -b - Buffer, size ``b`` ``(ldb,*)`` contains the upper or lower triangle - of the symmetric matrix :math:`B`, as specified by ``upper_lower``. The - second dimension of ``b`` must be at least :math:`\max(1, n)`. - -ldb - The leading dimension of ``b``; at least :math:`\max(1, n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_sygvd_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - On exit, if ``jobz = job::vec``, then if :math:`\text{info} = 0`, ``a`` - contains the matrix :math:`Z` of eigenvectors. The eigenvectors are - normalized as follows: - - if :math:`\text{itype} = 1` or :math:`2` , :math:`Z^{T}BZ = I`; - - if :math:`\text{itype} = 3` , :math:`Z^{T}B^{-1}Z = I`; - - If ``jobz = job::novec``, then on exit the upper triangle (if - ``upper_lower = uplo::upper``) or the lower triangle (if - ``upper_lower = uplo::lower``) of :math:`A`, including the diagonal, - is destroyed. - -b - On exit, if :math:`\text{info} \le n`, the part of ``b`` containing the matrix is - overwritten by the triangular factor :math:`U` or :math:`L` from the - Cholesky factorization :math:`B = U^{T}U` or - :math:`B = LL^{T}`. - -w - Buffer, size at least :math:`n`. If :math:`\text{info} = 0`, contains the - eigenvalues of the matrix :math:`A` in ascending order. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -sygvd (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, T *a, std::int64_t lda, T *b, std::int64_t ldb, T *w, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -itype - Must be 1 or 2 or 3. Specifies the problem type to be solved: - - if :math:`\text{itype} = 1`, the problem type is :math:`Ax = \lambda Bx`; - - if :math:`\text{itype} = 2`, the problem type is :math:`ABx = \lambda x`; - - if :math:`\text{itype} = 3`, the problem type is :math:`BAx = \lambda x`. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = job::upper``, ``a`` and ``b`` store the upper - triangular part of :math:`A` and :math:`B`. - - If ``upper_lower = job::lower``, ``a`` and ``b`` stores the lower - triangular part of :math:`A` and :math:`B`. - -n - The order of the matrices :math:`A` and :math:`B` :math:`(0 \le n)`. - -a - Pointer to array of size a\ ``(lda,*)`` containing the upper or lower triangle - of the symmetric matrix :math:`A`, as specified by ``upper_lower``. The - second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``; at least :math:`\max(1, n)`. - -b - Pointer to array of size ``b`` ``(ldb,*)`` contains the upper or lower triangle - of the symmetric matrix :math:`B`, as specified by ``upper_lower``. The - second dimension of ``b`` must be at least :math:`\max(1, n)`. - -ldb - The leading dimension of ``b``; at least :math:`\max(1, n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_sygvd_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - On exit, if ``jobz = job::vec``, then if :math:`\text{info} = 0`, :math:`a` - contains the matrix :math:`Z` of eigenvectors. The eigenvectors are - normalized as follows: - - if :math:`\text{itype} = 1` or :math:`2`, :math:`Z^{T}BZ = I`; - - if :math:`\text{itype} = 3`, :math:`Z^{T}B^{-1}Z = I`; - - If ``jobz = job::novec``, then on exit the upper triangle (if - ``upper_lower = uplo::upper``) or the lower triangle (if - ``upper_lower = uplo::lower``) of :math:`A`, including the diagonal, - is destroyed. - -b - On exit, if :math:`\text{info} \le n`, the part of ``b`` containing the matrix is - overwritten by the triangular factor :math:`U` or :math:`L` from the - Cholesky factorization :math:`B` = :math:`U^{T}U` or - :math:`B = LL^{T}`. - -w - Pointer to array of size at least ``n``. If :math:`\text{info} = 0`, contains the - eigenvalues of the matrix :math:`A` in ascending order. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/sygvd_scratchpad_size.rst b/docs/domains/lapack/sygvd_scratchpad_size.rst deleted file mode 100644 index d14ef8f61..000000000 --- a/docs/domains/lapack/sygvd_scratchpad_size.rst +++ /dev/null @@ -1,92 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_sygvd_scratchpad_size: - -sygvd_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_sygvd` function. - -.. container:: section - - .. rubric:: Description - -`sygvd_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_sygvd` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -sygvd_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda, std::int64_t ldb) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_sygvd` function will be performed. - -itype - Must be 1 or 2 or 3. Specifies the problem type to be solved: - - if :math:`\text{itype} = 1`, the problem type is :math:`Ax = \lambda Bx`; - - if :math:`\text{itype} = 2`, the problem type is :math:`ABx = \lambda x`; - - if :math:`\text{itype} = 3`, the problem type is :math:`BAx = \lambda x`. - -jobz - Must be ``job::novec`` or ``job::vec``. - - If ``jobz = job::novec``, then only eigenvalues are computed. - - If ``jobz = job::vec``, then eigenvalues and eigenvectors are - computed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = job::upper``, ``a`` and ``b`` store the upper - triangular part of :math:`A` and :math:`B`. - - If ``upper_lower = job::lower``, ``a`` and ``b`` stores the lower - triangular part of :math:`A` and :math:`B`. - -n - The order of the matrices :math:`A` and :math:`B` :math:`(0 \le n)`. - -lda - The leading dimension of ``a``. - -ldb - The leading dimension of ``b``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_sygvd` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/sytrd.rst b/docs/domains/lapack/sytrd.rst deleted file mode 100644 index 2df61c2b6..000000000 --- a/docs/domains/lapack/sytrd.rst +++ /dev/null @@ -1,205 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_sytrd: - -sytrd -===== - -Reduces a real symmetric matrix to tridiagonal form. - -.. container:: section - - .. rubric:: Description - -``sytrd`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -The routine reduces a real symmetric matrix :math:`A` to symmetric -tridiagonal form :math:`T` by an orthogonal similarity transformation: -:math:`A = QTQ^T`. The orthogonal matrix :math:`Q` is not formed explicitly -but is represented as a product of :math:`n-1` elementary reflectors. -Routines are provided for working with :math:`Q` in this representation . - -sytrd (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void sytrd(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = uplo::upper``, ``a`` stores the upper - triangular part of :math:`A`. - - If ``upper_lower = uplo::lower``, ``a`` stores the lower - triangular part of :math:`A`. - -n - The order of the matrices :math:`A` :math:`(0 \le n)`. - -a - The buffer ``a``, size ``(lda,*)``. Contains the upper or lower - triangle of the symmetric matrix :math:`A`, as specified by - ``upper_lower``. - - The second dimension of ``a`` must be at least :math:`\max(1,n)`. - -lda - The leading dimension of ``a``; at least :math:`\max(1,n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_sytrd_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - On exit, - - if ``upper_lower = uplo::upper``, the diagonal and first - superdiagonal of :math:`A` are overwritten by the corresponding - elements of the tridiagonal matrix :math:`T`, and the elements above - the first superdiagonal, with the buffer ``tau``, represent the - orthogonal matrix :math:`Q` as a product of elementary reflectors; - - if ``upper_lower = uplo::lower``, the diagonal and first - subdiagonal of :math:`A` are overwritten by the corresponding elements - of the tridiagonal matrix :math:`T`, and the elements below the first - subdiagonal, with the buffer ``tau``, represent the orthogonal matrix - :math:`Q` as a product of elementary reflectors. - -d - Buffer containing the diagonal elements of the matrix :math:`T`. The - dimension of ``d`` must be at least :math:`\max(1, n)`. - -e - Buffer containing the off diagonal elements of the matrix :math:`T`. - The dimension of ``e`` must be at least :math:`\max(1, n-1)`. - -tau - Buffer, size at least :math:`\max(1, n)`. Stores :math:`(n-1)` scalars that - define elementary reflectors in decomposition of the unitary - matrix :math:`Q` in a product of :math:`n-1` elementary reflectors. - :math:`\tau(n)` is used as workspace. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -sytrd (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, T *a, std::int64_t lda, T *d, T *e, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = uplo::upper``, ``a`` stores the upper - triangular part of :math:`A`. - - If ``upper_lower = uplo::lower``, ``a`` stores the lower - triangular part of :math:`A`. - -n - The order of the matrices :math:`A` :math:`(0 \le n)`. - -a - The pointer to matrix :math:`A`, size ``(lda,*)``. Contains the upper or lower - triangle of the symmetric matrix :math:`A`, as specified by - ``upper_lower``. - The second dimension of ``a`` must be at least :math:`\max(1,n)`. - -lda - The leading dimension of ``a``; at least :math:`\max(1,n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_sytrd_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - On exit, - - if ``upper_lower = uplo::upper``, the diagonal and first - superdiagonal of :math:`A` are overwritten by the corresponding - elements of the tridiagonal matrix :math:`T`, and the elements above - the first superdiagonal, with the array ``tau``, represent the - orthogonal matrix :math:`Q` as a product of elementary reflectors; - - if ``upper_lower = uplo::lower``, the diagonal and first - subdiagonal of :math:`A` are overwritten by the corresponding elements - of the tridiagonal matrix :math:`T`, and the elements below the first - subdiagonal, with the array ``tau``, represent the orthogonal matrix - :math:`Q` as a product of elementary reflectors. - -d - Pointer to diagonal elements of the matrix :math:`T`. The - dimension of ``d`` must be at least :math:`\max(1, n)`. - -e - Pointer to off diagonal elements of the matrix :math:`T`. - The dimension of ``e`` must be at least :math:`\max(1, n-1)`. - -tau - Pointer to array of size at least :math:`\max(1, n)`. Stores :math:`(n-1)` scalars that - define elementary reflectors in decomposition of the unitary - matrix :math:`Q` in a product of :math:`n-1` elementary reflectors. - :math:`\tau(n)` is used as workspace. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/sytrd_scratchpad_size.rst b/docs/domains/lapack/sytrd_scratchpad_size.rst deleted file mode 100644 index b3db401f1..000000000 --- a/docs/domains/lapack/sytrd_scratchpad_size.rst +++ /dev/null @@ -1,72 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_sytrd_scratchpad_size: - -sytrd_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_sytrd` function. - -.. container:: section - - .. rubric:: Description - -``sytrd_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_sytrd` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -sytrd_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_sytrd` function will be performed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. - - If ``upper_lower = uplo::upper``, a stores the upper triangular - part of :math:`A`. - - If ``upper_lower = uplo::lower``, a stores the lower triangular - part of :math:`A`. - -n - The order of the matrices :math:`A` :math:`(0 \le n)`. - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_sytrd` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/sytrf.rst b/docs/domains/lapack/sytrf.rst deleted file mode 100644 index 0baea1476..000000000 --- a/docs/domains/lapack/sytrf.rst +++ /dev/null @@ -1,166 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_sytrf: - -sytrf -===== - -Computes the Bunch-Kaufman factorization of a symmetric matrix. - -.. container:: section - - .. rubric:: Description - -``sytrf`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -The routine computes the factorization of a real/complex symmetric -matrix :math:`A` using the Bunch-Kaufman diagonal pivoting method. The -form of the factorization is: - -- if ``upper_lower=uplo::upper``, :math:`A` = :math:`UDU^{T}` - -- if ``upper_lower=uplo::lower``, :math:`A` = :math:`LDL^{T}` - -where :math:`A` is the input matrix, :math:`U` and :math:`L` are products of -permutation and triangular matrices with unit diagonal (upper -triangular for :math:`U` and lower triangular for :math:`L`), and :math:`D` is a -symmetric block-diagonal matrix with :math:`1 \times 1` and :math:`2 \times 2` diagonal -blocks. :math:`U` and :math:`L` have :math:`2 \times 2` unit diagonal blocks -corresponding to the :math:`2 \times 2` blocks of :math:`D`. - -sytrf (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void sytrf(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates whether the upper or lower triangular part of :math:`A` is stored and how :math:`A` is factored: - - If ``upper_lower=uplo::upper``, the buffer ``a`` stores the upper triangular part of the matrix :math:`A`, and :math:`A` is factored as :math:`UDU^T`. - - If ``upper_lower=uplo::lower``, the buffer ``a`` stores the lower triangular part of the matrix :math:`A`, and :math:`A` is factored as :math:`LDL^T`. - -n - The order of matrix :math:`A` (:math:`0 \le n`). - -a - The buffer ``a``, size :math:`\max(1,lda \cdot n)`. The buffer ``a`` contains either the upper or the lower triangular part of the matrix :math:`A` (see ``upper_lower``). The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_sytrf_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - The upper or lower triangular part of a is overwritten by details of the block-diagonal matrix :math:`D` and the multipliers used to obtain the factor :math:`U` (or :math:`L`). - -ipiv - Buffer, size at least :math:`\max(1, n)`. Contains details of the interchanges and the block structure of :math:`D`. If :math:`\text{ipiv}(i)=k>0`, then :math:`d_{ii}` is a :math:`1 \times 1` block, and the :math:`i`-th row and column of :math:`A` was interchanged with the :math:`k`-th row and column. - - If ``upper_lower=oneapi::mkl::uplo::upper`` and :math:`\text{ipiv}(i)=\text{ipiv}(i-1)=-m<0`, then :math:`D` has a :math:`2 \times 2` block in rows/columns :math:`i` and :math:`i`-1, and (:math:`i-1`)-th row and column of :math:`A` was interchanged with the :math:`m`-th row and column. - - If ``upper_lower=oneapi::mkl::uplo::lower`` and :math:`\text{ipiv}(i)=\text{ipiv}(i+1)=-m<0`, then :math:`D` has a :math:`2 \times 2` block in rows/columns :math:`i` and :math:`i+1`, and (:math:`i+1`)-th row and column of :math:`A` was interchanged with the :math:`m`-th row and column. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -sytrf (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, T *a, std::int64_t lda, int_64 *ipiv, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates whether the upper or lower triangular part of :math:`A` is stored and how :math:`A` is factored: - - If ``upper_lower=uplo::upper``, the array ``a`` stores the upper triangular part of the matrix :math:`A`, and :math:`A` is factored as :math:`UDU^T`. - - If ``upper_lower=uplo::lower``, the array ``a`` stores the lower triangular part of the matrix :math:`A`, and :math:`A` is factored as :math:`LDL^T`. - -n - The order of matrix :math:`A` (:math:`0 \le n`). - -a - The pointer to :math:`A`, size :math:`\max(1,\text{lda} \cdot n)`, containing either the upper or the lower triangular part of the matrix :math:`A` (see ``upper_lower``). The second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_sytrf_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - The upper or lower triangular part of a is overwritten by details of the block-diagonal matrix :math:`D` and the multipliers used to obtain the factor :math:`U` (or :math:`L`). - -ipiv - Pointer to array of size at least :math:`\max(1, n)`. Contains details of the interchanges and the block structure of :math:`D`. If :math:`\text{ipiv}(i)=k>0`, then :math:`d_{ii}` is a :math:`1 \times 1` block, and the :math:`i`-th row and column of :math:`A` was interchanged with the :math:`k`-th row and column. - - If ``upper_lower=oneapi::mkl::uplo::upper`` and :math:`\text{ipiv}(i)=\text{ipiv}(i-1)=-m<0`, then :math:`D` has a :math:`2 \times 2` block in rows/columns :math:`i` and :math:`i-1`, and (:math:`i-1`)-th row and column of :math:`A` was interchanged with the :math:`m`-th row and column. - - If ``upper_lower=oneapi::mkl::uplo::lower`` and :math:`\text{ipiv}(i)=\text{ipiv}(i+1)=-m<0`, then :math:`D` has a :math:`2 \times 2` block in rows/columns :math:`i` and :math:`i+1`, and (:math:`i+1`)-th row and column of :math:`A` was interchanged with the :math:`m`-th row and column. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/sytrf_scratchpad_size.rst b/docs/domains/lapack/sytrf_scratchpad_size.rst deleted file mode 100644 index 5b56c6385..000000000 --- a/docs/domains/lapack/sytrf_scratchpad_size.rst +++ /dev/null @@ -1,77 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_sytrf_scratchpad_size: - -sytrf_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_sytrf` function. - -.. container:: section - - .. rubric:: Description - -``sytrf_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_sytrf` function should be able to hold. -Calls to this routine must specify the template parameter -explicitly. - -sytrf_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_sytrf` function will be performed. - -upper_lower - Indicates whether the upper or lower triangular part of :math:`A` is - stored and how :math:`A` is factored: - - If ``upper_lower=uplo::upper``, the buffer ``a`` stores the - upper triangular part of the matrix :math:`A`, and :math:`A` is - factored as :math:`UDU^T`. - - If ``upper_lower=uplo::lower``, the buffer ``a`` stores the - lower triangular part of the matrix :math:`A`, and :math:`A` is - factored as :math:`LDL^T` - -n - The order of the matrix :math:`A` (:math:`0 \le n`). - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_sytrf` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/trtrs.rst b/docs/domains/lapack/trtrs.rst deleted file mode 100644 index 8f980465b..000000000 --- a/docs/domains/lapack/trtrs.rst +++ /dev/null @@ -1,197 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_trtrs: - -trtrs -===== - -Solves a system of linear equations with a triangular coefficient -matrix, with multiple right-hand sides. - -.. container:: section - - .. rubric:: Description - -``trtrs`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -The routine solves for :math:`X` the following systems of linear -equations with a triangular matrix :math:`A`, with multiple right-hand -sides stored in :math:`B`: - - .. list-table:: - :header-rows: 1 - - * - :math:`AX = B` - - - - if ``transa`` =\ ``transpose::nontrans``, - * - \ :math:`A^TX = B`\ - - - - if ``transa`` =\ ``transpose::trans``, - * - :math:`A^HX = B` - - - - if ``transa`` =\ ``transpose::conjtrans`` (for complex matrices only). - -trtrs (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void trtrs(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates whether :math:`A` is upper or lower triangular: - - If upper_lower = ``uplo::upper``, then :math:`A` is upper triangular. - - If upper_lower = ``uplo::lower``, then :math:`A` is lower triangular. - -transa - If transa = ``transpose::nontrans``, then :math:`AX = B` is solved for :math:`X`. - - If transa = ``transpose::trans``, then :math:`A^{T}X = B` is solved for :math:`X`. - - If transa = ``transpose::conjtrans``, then :math:`A^{H}X = B` is solved for :math:`X`. - -unit_diag - If unit_diag = ``diag::nonunit``, then :math:`A` is not a unit triangular matrix. - - If unit_diag = ``diag::unit``, then :math:`A` is unit triangular: diagonal elements of :math:`A` are assumed to be 1 and not referenced in the array ``a``. - -n - The order of :math:`A`; the number of rows in :math:`B`; :math:`n \ge 0`. - -nrhs - The number of right-hand sides; :math:`\text{nrhs} \ge 0`. - -a - Buffer containing the matrix :math:`A`. The second dimension of ``a`` must be at least :math:`\max(1,n)`. - -lda - The leading dimension of ``a``; :math:`\text{lda} \ge \max(1, n)`. - -b - Buffer containing the matrix :math:`B` whose columns are the right-hand sides for the systems of equations. The second dimension of ``b`` at least :math:`\max(1,\text{nrhs})`. - -ldb - The leading dimension of ``b``; :math:`\text{ldb} \ge \max(1, n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_trtrs_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -b - Overwritten by the solution matrix :math:`X`. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -trtrs (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t nrhs, T *a, std::int64_t lda, T *b, std::int64_t ldb, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Indicates whether :math:`A` is upper or lower triangular: - - If upper_lower = ``uplo::upper``, then :math:`A` is upper triangular. - - If upper_lower = ``uplo::lower``, then :math:`A` is lower triangular. - -transa - If transa = ``transpose::nontrans``, then :math:`AX = B` is solved for :math:`X`. - - If transa = ``transpose::trans``, then :math:`A^{T}X = B` is solved for :math:`X`. - - If transa = ``transpose::conjtrans``, then :math:`A^{H}X = B` is solved for :math:`X`. - -unit_diag - If unit_diag = ``diag::nonunit``, then :math:`A` is not a unit triangular matrix. - - If unit_diag = ``diag::unit``, then :math:`A` is unit triangular: diagonal elements of :math:`A` are assumed to be 1 and not referenced in the array ``a``. - -n - The order of :math:`A`; the number of rows in :math:`B`; :math:`n \ge 0`. - -nrhs - The number of right-hand sides; :math:`\text{nrhs} \ge 0`. - -a - Array containing the matrix :math:`A`. The second dimension of ``a`` must be at least :math:`\max(1,n)`. - -lda - The leading dimension of ``a``; :math:`\text{lda} \ge \max(1, n)`. - -b - Array containing the matrix :math:`B` whose columns are the right-hand sides for the systems of equations. The second dimension of ``b`` at least :math:`\max(1,\text{nrhs})`. - -ldb - The leading dimension of ``b``; :math:`\text{ldb} \ge \max(1, n)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_trtrs_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -b - Overwritten by the solution matrix :math:`X`. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/trtrs_scratchpad_size.rst b/docs/domains/lapack/trtrs_scratchpad_size.rst deleted file mode 100644 index 150800492..000000000 --- a/docs/domains/lapack/trtrs_scratchpad_size.rst +++ /dev/null @@ -1,94 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_trtrs_scratchpad_size: - -trtrs_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_trtrs` function. - -.. container:: section - - .. rubric:: Description - -``trtrs_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``float`` - * - ``double`` - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_trtrs` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -trtrs_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_trtrs` function will be performed. - -upper_lower - Indicates whether :math:`A` is upper or lower triangular: - - If upper_lower = ``uplo::upper``, then :math:`A` is upper triangular. - - If upper_lower = ``uplo::lower``, then :math:`A` is lower triangular. - -trans - Indicates the form of the equations: - - If ``trans=oneapi::mkl::transpose::nontrans``, then :math:`AX = B` is solved - for :math:`X`. - - If ``trans=oneapi::mkl::transpose::trans``, then :math:`A^TX = B` is solved - for :math:`X`. - - If ``trans=oneapi::mkl::transpose::conjtrans``, then :math:`A^HX = B` is - solved for :math:`X`. - -diag - If diag = ``oneapi::mkl::diag::nonunit``, then :math:`A` is not a unit triangular matrix. - - If unit_diag = ``diag::unit``, then :math:`A` is unit triangular: diagonal elements of :math:`A` are assumed to be 1 and not referenced in the array ``a``. - -n - The order of :math:`A`; the number of rows in :math:`B`; :math:`n \ge 0`. - -nrhs - The number of right-hand sides (:math:`0 \le \text{nrhs}`). - -lda - The leading dimension of ``a``; :math:`\text{lda} \ge \max(1, n)`. - -ldb - The leading dimension of ``b``; :math:`\text{ldb} \ge \max(1, n)`. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_trtrs` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/ungbr.rst b/docs/domains/lapack/ungbr.rst deleted file mode 100644 index 671d169f5..000000000 --- a/docs/domains/lapack/ungbr.rst +++ /dev/null @@ -1,231 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ungbr: - -ungbr -===== - -Generates the complex unitary matrix :math:`Q` or :math:`P^{t}` determined by -:ref:`onemkl_lapack_gebrd`. - -.. container:: section - - .. rubric:: Description - -``ungbr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -The routine generates the whole or part of the unitary matrices :math:`Q` -and :math:`P^{H}` formed by the routines -:ref:`onemkl_lapack_gebrd`. -All valid combinations of arguments are described in *Input Parameters*; in -most cases you need the following: - -To compute the whole :math:`m \times m` matrix :math:`Q`, use: - -:: - - oneapi::mkl::lapack::ungbr(queue, generate::q, m, m, n, a, ...) - -(note that the buffer ``a`` must have at least :math:`m` columns). - -To form the :math:`n` leading columns of :math:`Q` if :math:`m > n`, use: - -:: - - oneapi::mkl::lapack::ungbr(queue, generate::q, m, n, n, a, ...) - -To compute the whole :math:`n \times n` matrix :math:`P^{T}`, use: - -:: - - oneapi::mkl::lapack::ungbr(queue, generate::p, n, n, m, a, ...) - -(note that the array ``a`` must have at least :math:`n` rows). - -To form the :math:`m` leading rows of :math:`P^{T}` if :math:`m < n`, use: - -:: - - oneapi::mkl::lapack::ungbr(queue, generate::p, m, n, m, a, ...) - -ungbr (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void ungbr(sycl::queue &queue, oneapi::mkl::generate gen, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -gen - Must be ``generate::q`` or ``generate::p``. - - If ``gen = generate::q``, the routine generates the matrix :math:`Q`. - - If ``gen = generate::p``, the routine generates the matrix - :math:`P^{T}`. - -m - The number of rows in the matrix :math:`Q` or :math:`P^{T}` to be - returned :math:`(0 \le m)`. - - If ``gen = generate::q``, :math:`m \ge n \ge \min(m, k)`. - - If ``gen = generate::p``, :math:`n \ge m \ge \min(n, k)`. - -n - The number of columns in the matrix :math:`Q` or :math:`P^{T}` to be - returned :math:`(0 \le n)`. See ``m`` for constraints. - -k - If ``gen = generate::q``, the number of columns in the original - :math:`m \times k` matrix returned by - :ref:`onemkl_lapack_gebrd`. - - If ``gen = generate::p``, the number of rows in the original - :math:`k \times n` matrix returned by - :ref:`onemkl_lapack_gebrd`. - -a - The buffer ``a`` as returned by - :ref:`onemkl_lapack_gebrd`. - -lda - The leading dimension of ``a``. - -tau - For ``gen = generate::q``, the array ``tauq`` as returned by :ref:`onemkl_lapack_gebrd`. - For ``gen = generate::p``, the array ``taup`` as returned by :ref:`onemkl_lapack_gebrd`. - - The dimension of ``tau`` must be at least :math:`\max(1, \min(m, k))` for - ``gen = generate::q``, or :math:`\max(1, \min(m, k))` for - ``gen = generate::p``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type :math:`T`. - Size should not be less than the value returned by :ref:`onemkl_lapack_ungbr_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by :math:`n` leading columns of the :math:`m \times m` unitary matrix - :math:`Q` or :math:`P^{T}`, (or the leading rows or columns thereof) - as specified by ``gen``, ``m``, and ``n``. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -ungbr (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate gen, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -gen - Must be ``generate::q`` or ``generate::p``. - - If ``gen = generate::q``, the routine generates the matrix :math:`Q`. - - If ``gen = generate::p``, the routine generates the matrix - :math:`P^{T}`. - -m - The number of rows in the matrix :math:`Q` or :math:`P^{T}` to be - returned :math:`(0 \ge m)`. - - If ``gen = generate::q``, :math:`m \ge n \ge \min(m, k)`. - - If ``gen = generate::p``, :math:`n \ge m \ge \min(n, k)`. - -n - The number of columns in the matrix :math:`Q` or :math:`P^{T}` to be - returned :math:`(0 \le n)`. See ``m`` for constraints. - -k - If ``gen = generate::q``, the number of columns in the original - :math:`m \times k` matrix returned by - :ref:`onemkl_lapack_gebrd`. - - If ``gen = generate::p``, the number of rows in the original - :math:`k \times n` matrix returned by - :ref:`onemkl_lapack_gebrd`. - -a - The pointer to ``a`` as returned by - :ref:`onemkl_lapack_gebrd`. - -lda - The leading dimension of ``a``. - -tau - For ``gen = generate::q``, the array ``tauq`` as returned by :ref:`onemkl_lapack_gebrd`. - For ``gen = generate::p``, the array ``taup`` as returned by :ref:`onemkl_lapack_gebrd`. - - The dimension of ``tau`` must be at least :math:`\max(1, \min(m, k))` for - ``gen = generate::q``, or :math:`\max(1, \min(m, k))` for - ``gen = generate::p``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type :math:`T`. - Size should not be less than the value returned by :ref:`onemkl_lapack_ungbr_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by :math:`n` leading columns of the :math:`m \times m` unitary matrix - :math:`Q` or :math:`P^{T}`, (or the leading rows or columns thereof) - as specified by ``gen``, ``m``, and ``n``. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/ungbr_scratchpad_size.rst b/docs/domains/lapack/ungbr_scratchpad_size.rst deleted file mode 100644 index 5a39e9cc1..000000000 --- a/docs/domains/lapack/ungbr_scratchpad_size.rst +++ /dev/null @@ -1,90 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ungbr_scratchpad_size: - -ungbr_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_ungbr` function. - -.. container:: section - - .. rubric:: Description - -``ungbr_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type :math:`T` the scratchpad memory to be passed to :ref:`onemkl_lapack_ungbr` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -ungbr_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate gen, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t &scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_ungbr` function will be performed. - -gen - Must be ``generate::q`` or ``generate::p``. - - If ``gen = generate::q``, the routine generates the matrix - :math:`Q`. - - If ``gen = generate::p``, the routine generates the matrix - :math:`P^{T}`. - -m - The number of rows in the matrix :math:`Q` or :math:`P^{T}` to be - returned :math:`(0 \le m)`. - - If ``gen = generate::q``, :math:`m \ge n \ge \min(m, k)`. - - If ``gen = generate::p``, :math:`n \ge m \ge \min(n, k)`. - -n - The number of columns in the matrix :math:`Q` or :math:`P^{T}` to - be returned :math:`(0 \le n)`. See m for constraints. - -k - If ``gen = generate::q``, the number of columns in the original - :math:`m \times k` matrix reduced by - :ref:`onemkl_lapack_gebrd`. - - If ``gen = generate::p``, the number of rows in the original - :math:`k \times n` matrix reduced by - :ref:`onemkl_lapack_gebrd`. - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_ungbr` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/ungqr.rst b/docs/domains/lapack/ungqr.rst deleted file mode 100644 index 044546adf..000000000 --- a/docs/domains/lapack/ungqr.rst +++ /dev/null @@ -1,181 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ungqr: - -ungqr -===== - -Generates the complex unitary matrix :math:`Q` of the QR factorization formed -by :ref:`onemkl_lapack_geqrf`. - -.. container:: section - - .. rubric:: Description - -``ungqr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -The routine generates the whole or part of :math:`m \times m` unitary -matrix :math:`Q` of the QR factorization formed by the routines -:ref:`onemkl_lapack_geqrf`. - -Usually :math:`Q` is determined from the QR factorization of an :math:`m \times p` matrix :math:`A` with :math:`m \ge p`. To compute the whole matrix -:math:`Q`, use: - -:: - - oneapi::mkl::lapack::ungqr(queue, m, m, p, a, lda, tau, scratchpad, scratchpad_size) - -To compute the leading :math:`p` columns of :math:`Q` (which form an -orthonormal basis in the space spanned by the columns of :math:`A`): - -:: - - oneapi::mkl::lapack::ungqr(queue, m, p, p, a, lda, tau, scratchpad, scratchpad_size) - -To compute the matrix :math:`Q^{k}` of the QR factorization of -the leading :math:`k` columns of the matrix :math:`A`: - -:: - - oneapi::mkl::lapack::ungqr(queue, m, m, k, a, lda, tau, scratchpad, scratchpad_size) - -To compute the leading :math:`k` columns of :math:`Q^{k}` (which form -an orthonormal basis in the space spanned by the leading :math:`k` -columns of the matrix :math:`A`): - -:: - - oneapi::mkl::lapack::ungqr(queue, m, k, k, a, lda, tau, scratchpad, scratchpad_size) - -ungqr (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` (:math:`0 \le k \le n`). - -a - The buffer ``a`` as returned by - :ref:`onemkl_lapack_geqrf`. - -lda - The leading dimension of ``a`` (:math:`\text{lda} \le m`). - -tau - The buffer ``tau`` as returned by - :ref:`onemkl_lapack_geqrf`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_ungqr_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by :math:`n` leading columns of the :math:`m \times m` - orthogonal matrix :math:`Q`. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -ungqr (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`A` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` (:math:`0 \le k \le n`). - -a - The pointer to ``a`` as returned by - :ref:`onemkl_lapack_geqrf`. - -lda - The leading dimension of ``a`` (:math:`\text{lda} \le m`). - -tau - The pointer to ``tau`` as returned by - :ref:`onemkl_lapack_geqrf`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_ungqr_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by :math:`n` leading columns of the :math:`m \times m` - orthogonal matrix :math:`Q`. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/ungqr_batch.rst b/docs/domains/lapack/ungqr_batch.rst deleted file mode 100644 index 0d69e33b8..000000000 --- a/docs/domains/lapack/ungqr_batch.rst +++ /dev/null @@ -1,274 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ungqr_batch: - -ungqr_batch -=========== - -Generates the complex unitary matrices :math:`Q_i` of the batch of QR factorizations formed by the :ref:`onemkl_lapack_geqrf_batch` function. - -.. container:: section - - .. rubric:: Description - -``ungqr_batch`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -.. _onemkl_lapack_ungqr_batch_buffer: - -ungqr_batch (Buffer Version) ----------------------------- - -.. container:: section - - .. rubric:: Description - -The buffer version of ``ungqr_batch`` supports only the strided API. - -**Strided API** - - | The routine generates the wholes or parts of :math`m \times m` unitary matrices :math:`Q_i` of the batch of QR factorization formed by the Strided API of the :ref:`onemkl_lapack_geqrf_batch_buffer`. - | Usually :math:`Q_i` is determined from the QR factorization of an :math:`m \times p` matrix :math:`A_i` with :math`m \ge p`. - | To compute the whole matrices :math:`Q_i`, use: - | ``ungqr_batch(queue, m, m, p, a, ...)`` - | To compute the leading :math:`p` columns of :math:`Q_i` (which form an orthonormal basis in the space spanned by the columns of :math:`A_i`): - | ``ungqr_batch(queue, m, p, p, a, ...)`` - | To compute the matrices :math:`Q_i`^k` of the QR factorizations of leading :math:`k` columns of the matrices :math:`A_i`: - | ``ungqr_batch(queue, m, m, k, a, ...)`` - | To compute the leading :math:`k` columns of :math:`Q_i^k` (which form an orthonormal basis in the space spanned by leading :math:`k` columns of the matrices :math:`A_i`): - | ``ungqr_batch(queue, m, k, k, a, ...)`` - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Number of rows in the matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in the matrices :math:`A_i` (:math:`0\le n`). - -k - Number of elementary reflectors whose product defines the matrices :math:`Q_i` (:math:`0 \le k \le n`). - -a - Array resulting after call to the Strided API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -lda - Leading dimension of :math:`A_i` (:math:`\text{lda} \le m`). - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -tau - Array resulting after call to the Strided API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -stride_tau - Stride between the beginnings of arrays :math:`tau_i` inside the array ``tau``. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by strided version of the Strided API of the :ref:`onemkl_lapack_ungqr_batch_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Array data is overwritten by ``a`` batch of n leading columns of the :math:`m \times m` unitary matrices :math:`Q_i`. - -.. _onemkl_lapack_ungqr_batch_usm: - -ungqr_batch (USM Version) -------------------------- - -.. container:: section - - .. rubric:: Description - -The USM version of ``ungqr_batch`` supports the group API and strided API. - -**Group API** - - | The routine generates the wholes or parts of :math`m \times m` unitary matrices :math:`Q_i` of the batch of QR factorization formed by the Group API of the :ref:`onemkl_lapack_geqrf_batch_buffer`. - | Usually :math:`Q_i` is determined from the QR factorization of an :math:`m \times p` matrix :math:`A_i` with :math`m \ge p`. - | To compute the whole matrices :math:`Q_i`, use: - | ``ungqr_batch(queue, m, m, p, a, ...)`` - | To compute the leading :math:`p` columns of :math:`Q_i` (which form an orthonormal basis in the space spanned by the columns of :math:`A_i`): - | ``ungqr_batch(queue, m, p, p, a, ...)`` - | To compute the matrices :math:`Q_i`^k` of the QR factorizations of leading :math:`k` columns of the matrices :math:`A_i`: - | ``ungqr_batch(queue, m, m, k, a, ...)`` - | To compute the leading :math:`k` columns of :math:`Q_i^k` (which form an orthonormal basis in the space spanned by leading :math:`k` columns of the matrices :math:`A_i`): - | ``ungqr_batch(queue, m, k, k, a, ...)`` - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, T **a, std::int64_t *lda, T **tau, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Array of ``group_count`` :math:`m_g` parameters as previously supplied to the Group API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -n - Array of ``group_count`` :math:`n_g` parameters as previously supplied to the Group API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -k - | Array of ``group_count`` :math:`k_g` parameters as previously supplied to the Group API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - | The number of elementary reflectors whose product defines the matrices :math:`Q_i` (:math:`0 \le k_g \le n_g`). - -a - Array resulting after call to the Group API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -lda - Array of leading dimensions of :math:`A_i` as previously supplied to the Group API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -tau - Array resulting after call to the Group API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by Group API of the :ref:`onemkl_lapack_ungqr_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Matrices pointed to by array ``a`` are overwritten by :math:`n_g` leading columns of the :math:`m_g \times m_g` orthogonal matrices :math:`Q_i`, where :math:`g` is an index of group of parameters corresponding to :math:`Q_i`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Strided API** - - | The routine generates the wholes or parts of :math`m \times m` unitary matrices :math:`Q_i` of the batch of QR factorization formed by the Strided API of the :ref:`onemkl_lapack_geqrf_batch_usm`. - | Usually :math:`Q_i` is determined from the QR factorization of an :math:`m \times p` matrix :math:`A_i` with :math`m \ge p`. - | To compute the whole matrices :math:`Q_i`, use: - | ``ungqr_batch(queue, m, m, p, a, ...)`` - | To compute the leading :math:`p` columns of :math:`Q_i` (which form an orthonormal basis in the space spanned by the columns of :math:`A_i`): - | ``ungqr_batch(queue, m, p, p, a, ...)`` - | To compute the matrices :math:`Q_i`^k` of the QR factorizations of leading :math:`k` columns of the matrices :math:`A_i`: - | ``ungqr_batch(queue, m, m, k, a, ...)`` - | To compute the leading :math:`k` columns of :math:`Q_i^k` (which form an orthonormal basis in the space spanned by leading :math:`k` columns of the matrices :math:`A_i`): - | ``ungqr_batch(queue, m, k, k, a, ...)`` - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, std::int64_t stride_a, T *tau, std::int64_t stride_tau, std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - }; - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Number of rows in the matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in the matrices :math:`A_i` (:math:`0\le n`). - -k - Number of elementary reflectors whose product defines the matrices :math:`Q_i` (:math:`0 \le k \le n`). - -a - Array resulting after call to the Strided API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -lda - Leading dimension of :math:`A_i` (:math:`\text{lda} \le m`). - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -tau - Array resulting after call to the Strided API of the :ref:`onemkl_lapack_geqrf_batch_usm` function. - -stride_tau - Stride between the beginnings of arrays :math:`tau_i` inside the array ``tau``. - -batch_size - Number of problems in a batch. - -scratchpad - Scratchpad memory to be used by routine for storing intermediate results. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. Size should not be less then the value returned by strided version of the Strided API of the :ref:`onemkl_lapack_ungqr_batch_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Array data is overwritten by ``a`` batch of n leading columns of the :math:`m \times m` unitary matrices :math:`Q_i`. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/ungqr_batch_scratchpad_size.rst b/docs/domains/lapack/ungqr_batch_scratchpad_size.rst deleted file mode 100644 index 6a427d7b6..000000000 --- a/docs/domains/lapack/ungqr_batch_scratchpad_size.rst +++ /dev/null @@ -1,123 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ungqr_batch_scratchpad_size: - -ungqr_batch_scratchpad_size -=========================== - -Computes size of scratchpad memory required for the :ref:`onemkl_lapack_ungqr_batch` function. - -.. container:: section - - .. rubric:: Description - -``ungqr_batch_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -**Group API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_ungqr_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations will be performed. - -m - Array of ``group_count`` :math:`m_g` parameters. - -n - Array of ``group_count`` :math:`n_g` parameters. - -k - | Array of ``group_count`` :math:`k_g` parameters. - | Number of elementary reflectors whose product defines the matrices :math:`Q_i` (:math:`0 \le k_g \le n_g`). - -lda - Array of leading dimensions of :math:`A_i`. - -group_count - Number of groups of parameters. Must be at least 0. - -group_sizes - Array of ``group_count`` integers. Array element with index :math:`g` specifies the number of problems to solve for each of the groups of parameters :math:`g`. So the total number of problems to solve, ``batch_size``, is a sum of all parameter group sizes. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Group API of the :ref:`onemkl_lapack_ungqr_batch` function. - -**Strided API** - -Computes the number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_ungqr_batch` function. - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) - }; - -.. container:: section - - .. rubric:: Input Parameters - - -queue - Device queue where calculations will be performed. - -m - Number of rows in the matrices :math:`A_i` (:math:`0 \le m`). - -n - Number of columns in the matrices :math:`A_i` (:math:`0 \le n`). - -k - Number of elementary reflectors whose product defines the matrices :math:`Q_i` (:math:`0 \le k \le n`). - -lda - Leading dimensions of :math:`A_i` (:math:`\text{lda} \le m`). - -stride_a - Stride between the beginnings of matrices :math:`A_i` inside the batch array ``a``. - -stride_tau - Stride between the beginnings of arrays :math:`\tau_i` inside the array ``tau``. - -batch_size - Number of problems in a batch. - -.. container:: section - - .. rubric:: Return Values - -Number of elements of type ``T`` the scratchpad memory should able to hold to be passed to the Strided API of the :ref:`onemkl_lapack_ungqr_batch` function. - -**Parent topic:** :ref:`onemkl_lapack-like-extensions-routines` - diff --git a/docs/domains/lapack/ungqr_scratchpad_size.rst b/docs/domains/lapack/ungqr_scratchpad_size.rst deleted file mode 100644 index 7fed35d15..000000000 --- a/docs/domains/lapack/ungqr_scratchpad_size.rst +++ /dev/null @@ -1,70 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ungqr_scratchpad_size: - -ungqr_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_ungqr` function. - -.. container:: section - - .. rubric:: Description - -``ungqr_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_ungqr` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -ungqr_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_ungqr` function will be performed. - -m - The number of rows in the matrix :math:`A` (:math:`0 \le m`). - -n - The number of columns the matrix :math:`A` (:math:`0 \le n \le m`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` (:math:`0 \le k \le n`). - -lda - The leading dimension of ``a``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_ungqr` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines`  - - diff --git a/docs/domains/lapack/ungtr.rst b/docs/domains/lapack/ungtr.rst deleted file mode 100644 index 764a10f96..000000000 --- a/docs/domains/lapack/ungtr.rst +++ /dev/null @@ -1,153 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ungtr: - -ungtr -===== - -Generates the complex unitary matrix :math:`Q` determined by -:ref:`onemkl_lapack_hetrd`. - -.. container:: section - - .. rubric:: Description - -``ungtr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -The routine explicitly generates the :math:`n \times n` unitary matrix -:math:`Q` formed by :ref:`onemkl_lapack_hetrd` when -reducing a complex Hermitian matrix :math:`A` to tridiagonal form: -:math:`A = QTQ^H`. Use this routine after a call to -:ref:`onemkl_lapack_hetrd`. - -ungtr (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void ungtr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. Uses the same - ``upper_lower`` as supplied to - :ref:`onemkl_lapack_hetrd`. - -n - The order of the matrix :math:`Q` :math:`(0 \le n)`. - -a - The buffer ``a`` as returned by - :ref:`onemkl_lapack_hetrd`. The - second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a`` :math:`(n \le \text{lda})`. - -tau - The buffer ``tau`` as returned by - :ref:`onemkl_lapack_hetrd`. The - dimension of ``tau`` must be at least :math:`\max(1, n-1)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_ungtr_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by the unitary matrix :math:`Q`. - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -ungtr (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. Uses the same - ``upper_lower`` as supplied to - :ref:`onemkl_lapack_hetrd`. - -n - The order of the matrix :math:`Q` :math:`(0 \le n)`. - -a - The pointer to ``a`` as returned by - :ref:`onemkl_lapack_hetrd`. The - second dimension of ``a`` must be at least :math:`\max(1, n)`. - -lda - The leading dimension of ``a`` :math:`(n \le \text{lda})`. - -tau - The pointer to ``tau`` as returned by - :ref:`onemkl_lapack_hetrd`. The - dimension of ``tau`` must be at least :math:`\max(1, n-1)`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_ungtr_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -a - Overwritten by the unitary matrix :math:`Q`. - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - - diff --git a/docs/domains/lapack/ungtr_scratchpad_size.rst b/docs/domains/lapack/ungtr_scratchpad_size.rst deleted file mode 100644 index 6b91bbbaa..000000000 --- a/docs/domains/lapack/ungtr_scratchpad_size.rst +++ /dev/null @@ -1,67 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_ungtr_scratchpad_size: - -ungtr_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_ungtr` function. - -.. container:: section - - .. rubric:: Description - -``ungtr_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_ungtr` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -ungtr_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t lda) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_ungtr` function will be performed. - -upper_lower - Must be ``uplo::upper`` or ``uplo::lower``. Uses the same - ``upper_lower`` as supplied to - :ref:`onemkl_lapack_hetrd`. - -n - The order of the matrix :math:`Q` :math:`(0 \le n)`. - -lda - The leading dimension of ``a`` :math:`(n \le lda)`. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type T the scratchpad memory to be passed to :ref:`onemkl_lapack_ungtr` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - diff --git a/docs/domains/lapack/unmqr.rst b/docs/domains/lapack/unmqr.rst deleted file mode 100644 index 0515be510..000000000 --- a/docs/domains/lapack/unmqr.rst +++ /dev/null @@ -1,207 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_unmqr: - -unmqr -===== - -Multiplies a complex matrix by the unitary matrix :math:`Q` of the QR -factorization formed by :ref:`onemkl_lapack_geqrf`. - -.. container:: section - - .. rubric:: Description - -``unmqr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -The routine multiplies a rectangular complex :math:`m \times n` matrix :math:`C` by -:math:`Q` or :math:`Q^H`, where :math:`Q` is the complex unitary matrix defined -as a product of :math:`k` elementary reflectors :math:`H(i)` of order :math:`n`: -:math:`Q = H(1)^HH(2)^H ... H(k)^H` as returned by the RQ factorization routine -:ref:`onemkl_lapack_gerqf`. - -Depending on the parameters ``side`` and ``trans``, the routine can form one of -the matrix products :math:`QC`, :math:`Q^HC`, :math:`CQ`, or :math:`CQ^H` -(overwriting the result over :math:`C`). - -unmqr (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -side - If ``side = oneapi::mkl::side::left``, :math:`Q` or :math:`Q^{H}` is applied - to :math:`C` from the left. - - If ``side = oneapi::mkl::side::right``, :math:`Q` or :math:`Q^{H}` is - applied to :math:`C` from the right. - -trans - If ``trans = oneapi::mkl::transpose::nontrans``, the routine multiplies - :math:`C` by :math:`Q`. - - If ``trans = oneapi::mkl::transpose::conjtrans``, the routine multiplies :math:`C` - by :math:`Q^{H}`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`C` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` - - If ``side = oneapi::mkl::side::left``, :math:`0 \le k \le m` - - If ``side = oneapi::mkl::side::right``, :math:`0 \le k \le n` - -a - The buffer ``a`` as returned by :ref:`onemkl_lapack_geqrf`. - The second dimension of ``a`` must be at least :math:`\max(1,k)`. - -lda - The leading dimension of ``a``. - -tau - The buffer ``tau`` as returned by :ref:`onemkl_lapack_geqrf`. - -c - The buffer ``c`` contains the matrix :math:`C`. The second dimension of - ``c`` must be at least :math:`\max(1,n)`. - -ldc - The leading dimension of ``c``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type - ``T``. Size should not be less than the value returned by the - :ref:`onemkl_lapack_unmqr_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^{H}C`, :math:`CQ`, or - :math:`CQ^H` (as specified by ``side`` and ``trans``). - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -unmqr (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -side - If ``side = oneapi::mkl::side::left``, :math:`Q` or :math:`Q^{H}` is applied - to :math:`C` from the left. - - If ``side = oneapi::mkl::side::right``, :math:`Q` or :math:`Q^{H}` is - applied to :math:`C` from the right. - -trans - If ``trans = oneapi::mkl::transpose::nontrans``, the routine multiplies - :math:`C` by :math:`Q`. - - If ``trans = oneapi::mkl::transpose::conjtrans``, the routine multiplies :math:`C` - by :math:`Q^{H}`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`C` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` - - If ``side = oneapi::mkl::side::left``, :math:`0 \le k \le m` - - If ``side = oneapi::mkl::side::right``, :math:`0 \le k \le n` - -a - The pointer to ``a`` as returned by :ref:`onemkl_lapack_geqrf`. - The second dimension of ``a`` must be at least :math:`\max(1,k)`. - -lda - The leading dimension of ``a``. - -tau - The pointer to ``tau`` as returned by :ref:`onemkl_lapack_geqrf`. - -c - The pointer ``c`` points to the matrix :math:`C`. The second dimension of - ``c`` must be at least :math:`\max(1,n)`. - -ldc - The leading dimension of ``c``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type - ``T``. Size should not be less than the value returned by - :ref:`onemkl_lapack_unmqr_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^{H}C`, :math:`CQ`, or - :math:`CQ^{H}` (as specified by ``side`` and ``trans``). - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` diff --git a/docs/domains/lapack/unmqr_scratchpad_size.rst b/docs/domains/lapack/unmqr_scratchpad_size.rst deleted file mode 100644 index a95127507..000000000 --- a/docs/domains/lapack/unmqr_scratchpad_size.rst +++ /dev/null @@ -1,87 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_unmqr_scratchpad_size: - -unmqr_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_unmqr` function. - -.. container:: section - - .. rubric:: Description - -``unmqr_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_unmqr` function should be able to hold. -Calls to this routine must specify the template parameter -explicitly. - -unmqr_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc, std::int64_t &scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_unmqr` function will be performed. - -side - If ``side=oneapi::mkl::side::left``, :math:`Q` or :math:`Q^{H}` is - applied to :math:`C` from the left. - - If ``side=oneapi::mkl::side::right``, :math:`Q` or :math:`Q^{H}` is - applied to :math:`C` from the right. - -trans - If ``trans=oneapi::mkl::transpose::nontrans``, the routine multiplies - :math:`C` by :math:`Q`. - - If ``trans=oneapi::mkl::transpose::conjtrans``, the routine multiplies - :math:`C` by :math:`Q^H`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns the matrix :math:`C` (:math:`0 \le n \le m`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` (:math:`0 \le k \le n`). - -lda - The leading dimension of ``a``. - -ldc - The leading dimension of ``c``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_unmqr` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/unmrq.rst b/docs/domains/lapack/unmrq.rst deleted file mode 100644 index 2b3872261..000000000 --- a/docs/domains/lapack/unmrq.rst +++ /dev/null @@ -1,207 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_unmrq: - -unmrq -===== - -Multiplies a complex matrix by the unitary matrix :math:`Q` of the RQ -factorization formed by :ref:`onemkl_lapack_gerqf`. - -.. container:: section - - .. rubric:: Description - -``unmrq`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -The routine multiplies a rectangular complex :math:`m \times n` matrix :math:`C` by -:math:`Q` or :math:`Q^H`, where :math:`Q` is the complex unitary matrix defined -as a product of :math:`k` elementary reflectors :math:`H(i)` of order :math:`n`: -:math:`Q = H(1)^HH(2)^H ... H(k)^H` as returned by the RQ factorization routine -:ref:`onemkl_lapack_gerqf`. - -Depending on the parameters ``side`` and ``trans``, the routine can form one of -the matrix products :math:`QC`, :math:`Q^HC`, :math:`CQ`, or :math:`CQ^H` -(overwriting the result over :math:`C`). - -unmrq (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -side - If ``side = oneapi::mkl::side::left``, :math:`Q` or :math:`Q^{H}` is applied - to :math:`C` from the left. - - If ``side = oneapi::mkl::side::right``, :math:`Q` or :math:`Q^{H}` is - applied to :math:`C` from the right. - -trans - If ``trans = oneapi::mkl::transpose::nontrans``, the routine multiplies - :math:`C` by :math:`Q`. - - If ``trans = oneapi::mkl::transpose::conjtrans``, the routine multiplies :math:`C` - by :math:`Q^{H}`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`C` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` - - If ``side = oneapi::mkl::side::left``, :math:`0 \le k \le m` - - If ``side = oneapi::mkl::side::right``, :math:`0 \le k \le n` - -a - The buffer ``a`` as returned by :ref:`onemkl_lapack_gerqf`. - The second dimension of ``a`` must be at least :math:`\max(1,k)`. - -lda - The leading dimension of ``a``. - -tau - The buffer ``tau`` as returned by :ref:`onemkl_lapack_gerqf`. - -c - The buffer ``c`` contains the matrix :math:`C`. The second dimension of - ``c`` must be at least :math:`\max(1,n)`. - -ldc - The leading dimension of ``c``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type - ``T``. Size should not be less than the value returned by - :ref:`onemkl_lapack_unmrq_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^{H}C`, :math:`CQ`, or - :math:`CQ^H` (as specified by ``side`` and ``trans``). - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -unmrq (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - The queue where the routine should be executed. - -side - If ``side = oneapi::mkl::side::left``, :math:`Q` or :math:`Q^{H}` is applied - to :math:`C` from the left. - - If ``side = oneapi::mkl::side::right``, :math:`Q` or :math:`Q^{H}` is - applied to :math:`C` from the right. - -trans - If ``trans = oneapi::mkl::transpose::nontrans``, the routine multiplies - :math:`C` by :math:`Q`. - - If ``trans = oneapi::mkl::transpose::conjtrans``, the routine multiplies :math:`C` - by :math:`Q^{H}`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`C` (:math:`0 \le n`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` - - If ``side = oneapi::mkl::side::left``, :math:`0 \le k \le m` - - If ``side = oneapi::mkl::side::right``, :math:`0 \le k \le n` - -a - The pointer to ``a`` as returned by :ref:`onemkl_lapack_gerqf`. - The second dimension of ``a`` must be at least :math:`\max(1,k)`. - -lda - The leading dimension of ``a``. - -tau - The pointer to ``tau`` as returned by :ref:`onemkl_lapack_gerqf`. - -c - The pointer ``c`` points to the matrix :math:`C`. The second dimension of - ``c`` must be at least :math:`\max(1,n)`. - -ldc - The leading dimension of ``c``. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type - ``T``. Size should not be less than the value returned by - :ref:`onemkl_lapack_unmrq_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^{H}C`, :math:`CQ`, or - :math:`CQ^{H}` (as specified by ``side`` and ``trans``). - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` diff --git a/docs/domains/lapack/unmrq_scratchpad_size.rst b/docs/domains/lapack/unmrq_scratchpad_size.rst deleted file mode 100644 index 8b771db75..000000000 --- a/docs/domains/lapack/unmrq_scratchpad_size.rst +++ /dev/null @@ -1,79 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_unmrq_scratchpad_size: - -unmrq_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_unmrq` function. - -.. container:: section - - .. rubric:: Description - -``unmrq_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_unmrq` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -unmrq_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by the unmrq function will be performed. - -side - If ``side = oneapi::mkl::side::left``, :math:`Q` or :math:`Q^T` is applied to :math:`C` from the left. If ``side = oneapi::mkl::side::right``, :math:`Q` or :math:`Q^T` is applied to :math:`C` from the right. - -trans - If ``trans=oneapi::mkl::transpose::nontrans``, the routine multiplies :math:`C` by :math:`Q`. - - If ``trans=oneapi::mkl::transpose::conjtrans``, the routine multiplies :math:`C` by :math:`Q^H`. - -m - The number of rows in the matrix :math:`C` (:math:`0 \le m`). - -n - The number of columns in the matrix :math:`C` (:math:`0 \le n \le m`). - -k - The number of elementary reflectors whose product defines the matrix :math:`Q` (:math:`0 \le k \le n`). - -lda - The leading dimension of ``a``. - -ldc - The leading dimension of ``c``. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_unmrq` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-linear-equation-routines` - diff --git a/docs/domains/lapack/unmtr.rst b/docs/domains/lapack/unmtr.rst deleted file mode 100644 index 0c156b4dc..000000000 --- a/docs/domains/lapack/unmtr.rst +++ /dev/null @@ -1,250 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_unmtr: - -unmtr -===== - -Multiplies a complex matrix by the complex unitary matrix Q -determined by -:ref:`onemkl_lapack_hetrd`. - -.. container:: section - - .. rubric:: Description - -``unmtr`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -The routine multiplies a complex matrix :math:`C` by :math:`Q` or -:math:`Q^{H}`, where :math:`Q` is the unitary matrix :math:`Q` formed by -:ref:`onemkl_lapack_hetrd` -when reducing a complex Hermitian matrix :math:`A` to tridiagonal form: -:math:`A = QTQ^H`. Use this routine after a call to -:ref:`onemkl_lapack_hetrd`. - -Depending on the parameters ``side`` and ``trans``, the routine can -form one of the matrix products :math:`QC`, :math:`Q^{H}C`, -:math:`CQ`, or :math:`CQ^{H}` (overwriting the result on :math:`C`). - -unmtr (Buffer Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) - } - -.. container:: section - - .. rubric:: Input Parameters - -In the descriptions below, ``r`` denotes the order of :math:`Q`: - -.. container:: tablenoborder - - .. list-table:: - :header-rows: 1 - - * - :math:`r`\ =\ :math:`m` - - if ``side = side::left`` - * - :math:`r`\ =\ :math:`n` - - if ``side = side::right`` - -queue - The queue where the routine should be executed. - -side - Must be either ``side::left`` or ``side::right``. - - If ``side=side::left``, :math:`Q` or :math:`Q^{H}` is applied - to :math:`C` from the left. - - If ``side=side::right``, :math:`Q` or :math:`Q^{H}` is applied - to :math:`C` from the right. - -upper_lower - Must be either ``uplo::upper`` or ``uplo::lower``. Uses the same - ``upper_lower`` as supplied to - :ref:`onemkl_lapack_hetrd`. - -trans - Must be either ``transpose::nontrans`` or - ``transpose::conjtrans``. - - If ``trans=transpose::nontrans``, the routine multiplies :math:`C` by - :math:`Q`. - - If ``trans=transpose::conjtrans``, the routine multiplies :math:`C` by - :math:`Q^{H}`. - -m - The number of rows in the matrix :math:`C` (:math:`m \ge 0`). - -n - The number of columns the matrix :math:`C` (:math:`n \ge 0`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` (:math:`0 \le k \le n`). - -a - The buffer ``a`` as returned by - :ref:`onemkl_lapack_hetrd`. - -lda - The leading dimension of ``a`` :math:`(\max(1,r) \le \text{lda})`. - -tau - The buffer ``tau`` as returned by - :ref:`onemkl_lapack_hetrd`. The - dimension of ``tau`` must be at least :math:`\max(1,r-1)`. - -c - The buffer ``c`` contains the matrix :math:`C`. The second dimension of ``c`` - must be at least :math:`\max(1,n)`. - -ldc - The leading dimension of ``c`` :math:`(\max(1,n) \le \text{ldc})`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_unmtr_scratchpad_size` function. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^{H}C`, - :math:`CQ`, or :math:`CQ^{H}` (as specified by ``side`` and - ``trans``). - -scratchpad - Buffer holding scratchpad memory to be used by routine for storing intermediate results. - -unmtr (USM Version) ----------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &events = {}) - } - -.. container:: section - - .. rubric:: Input Parameters - -In the descriptions below, ``r`` denotes the order of :math:`Q`: - -.. container:: tablenoborder - - .. list-table:: - :header-rows: 1 - - * - :math:`r`\ =\ :math:`m` - - if ``side = side::left`` - * - :math:`r`\ =\ :math:`n` - - if ``side = side::right`` - -queue - The queue where the routine should be executed. - -side - Must be either ``side::left`` or ``side::right``. - - If ``side=side::left``, :math:`Q` or :math:`Q^{H}` is applied - to :math:`C` from the left. - - If ``side=side::right``, :math:`Q` or :math:`Q^{H}` is applied - to :math:`C` from the right. - -upper_lower - Must be either ``uplo::upper`` or ``uplo::lower``. Uses the same - ``upper_lower`` as supplied to - :ref:`onemkl_lapack_hetrd`. - -trans - Must be either ``transpose::nontrans`` or - ``transpose::conjtrans``. - - If ``trans=transpose::nontrans``, the routine multiplies :math:`C` by - :math:`Q`. - - If ``trans=transpose::conjtrans``, the routine multiplies :math:`C` by - :math:`Q^{H}`. - -m - The number of rows in the matrix :math:`C` (:math:`m \ge 0`). - -n - The number of columns the matrix :math:`C` (:math:`n \ge 0`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` (:math:`0 \le k \le n`). - -a - The pointer to ``a`` as returned by - :ref:`onemkl_lapack_hetrd`. - -lda - The leading dimension of ``a`` :math:`(\max(1,r) \le \text{lda})`. - -tau - The pointer to ``tau`` as returned by - :ref:`onemkl_lapack_hetrd`. The - dimension of ``tau`` must be at least :math:`\max(1,r-1)`. - -c - The array ``c`` contains the matrix :math:`C`. The second dimension of ``c`` - must be at least :math:`\max(1,n)`. - -ldc - The leading dimension of ``c`` :math:`(\max(1,n) \le \text{ldc})`. - -scratchpad_size - Size of scratchpad memory as a number of floating point elements of type ``T``. - Size should not be less than the value returned by :ref:`onemkl_lapack_unmtr_scratchpad_size` function. - -events - List of events to wait for before starting computation. Defaults to empty list. - -.. container:: section - - .. rubric:: Output Parameters - -c - Overwritten by the product :math:`QC`, :math:`Q^{H}C`, - :math:`CQ`, or :math:`CQ^{H}` (as specified by ``side`` and - trans). - -scratchpad - Pointer to scratchpad memory to be used by routine for storing intermediate results. - -.. container:: section - - .. rubric:: Return Values - -Output event to wait on to ensure computation is complete. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - diff --git a/docs/domains/lapack/unmtr_scratchpad_size.rst b/docs/domains/lapack/unmtr_scratchpad_size.rst deleted file mode 100644 index 8ca39d3f0..000000000 --- a/docs/domains/lapack/unmtr_scratchpad_size.rst +++ /dev/null @@ -1,96 +0,0 @@ -.. SPDX-FileCopyrightText: 2019-2020 Intel Corporation -.. -.. SPDX-License-Identifier: CC-BY-4.0 - -.. _onemkl_lapack_unmtr_scratchpad_size: - -unmtr_scratchpad_size -===================== - -Computes size of scratchpad memory required for :ref:`onemkl_lapack_unmtr` function. - -.. container:: section - - .. rubric:: Description - -``unmtr_scratchpad_size`` supports the following precisions. - - .. list-table:: - :header-rows: 1 - - * - T - * - ``std::complex`` - * - ``std::complex`` - -Computes the number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_unmtr` function should be able to hold. -Calls to this routine must specify the template parameter explicitly. - -unmtr_scratchpad_size ---------------------- - -.. container:: section - - .. rubric:: Syntax - -.. code-block:: cpp - - namespace oneapi::mkl::lapack { - template - std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) - } - -.. container:: section - - .. rubric:: Input Parameters - -queue - Device queue where calculations by :ref:`onemkl_lapack_unmtr` function will be performed. - -side - Must be either ``side::left`` or ``side::right``. - - If ``side=side::left``, :math:`Q` or :math:`Q^{H}` is - applied to :math:`C` from the left. - - If ``side=side::right``, :math:`Q` or :math:`Q^{H}` is - applied to :math:`C` from the right. - -upper_lower - Must be either ``uplo::upper`` or ``uplo::lower``. Uses the - same ``upper_lower`` as supplied to - :ref:`onemkl_lapack_hetrd`. - -trans - Must be either ``transpose::nontrans`` or - ``transpose::conjtrans``. - - If ``trans=transpose::nontrans``, the routine multiplies :math:`C` - by :math:`Q`. - - If ``trans=transpose::conjtrans``, the routine multiplies :math:`C` - by :math:`Q^{H}`. - -m - The number of rows in the matrix :math:`C` (:math:`m \ge 0`). - -n - The number of columns the matrix :math:`C` (:math:`n \ge 0`). - -k - The number of elementary reflectors whose product defines the - matrix :math:`Q` (:math:`0 \le k \le n`). - -lda - The leading dimension of :math:`a` :math:`(\max(1,r) \le \text{lda})`. - -ldc - The leading dimension of :math:`c` :math:`(\max(1,n) \le \text{ldc})`. - -.. container:: section - - .. rubric:: Return Value - -The number of elements of type ``T`` the scratchpad memory to be passed to :ref:`onemkl_lapack_unmtr` function should be able to hold. - -**Parent topic:** :ref:`onemkl_lapack-singular-value-eigenvalue-routines` - diff --git a/docs/domains/matrix-storage.rst b/docs/domains/matrix-storage.rst deleted file mode 100644 index 656262f55..000000000 --- a/docs/domains/matrix-storage.rst +++ /dev/null @@ -1,581 +0,0 @@ -.. _matrix-storage: - -Matrix Storage -============== - - -.. container:: - - - The oneMKL BLAS and LAPACK routines for DPC++ use several matrix and - vector storage formats. These are the same formats used in - traditional Fortran BLAS/LAPACK. - - .. container:: section - - .. rubric:: General Matrix - :name: general-matrix - :class: sectiontitle - - A general matrix ``A`` of ``m`` rows and ``n`` columns with - leading dimension ``lda`` is represented as a one dimensional - array ``a`` of size of at least ``lda`` \* ``n`` if column major - layout is used and at least ``lda`` \* ``m`` if row major layout - is used. Before entry in any BLAS function using a general - matrix, the leading ``m`` by ``n`` part of the array ``a`` must - contain the matrix ``A``. For column (respectively row) major - layout, the elements of each column (respectively row) are - contiguous in memory while the elements of each row - (respectively column) are at distance ``lda`` from the element - in the same row (respectively column) and the previous column - (respectively row). - - Visually, the matrix - - .. math:: - - A = \begin{bmatrix} - A_{11} & A_{12} & A_{13} & \ldots & A_{1n}\\ - A_{21} & A_{22} & A_{23} & \ldots & A_{2n}\\ - A_{31} & A_{32} & A_{33} & \ldots & A_{3n}\\ - \vdots & \vdots & \vdots & \ddots & \vdots\\ - A_{m1} & A_{m2} & A_{m3} & \ldots & A_{mn} - \end{bmatrix} - - is stored in memory as an array - - - For column major layout, - - .. math:: - - \scriptstyle a = - [\underbrace{\underbrace{A_{11},A_{21},A_{31},...,A_{m1},*,...,*}_\text{lda}, - \underbrace{A_{12},A_{22},A_{32},...,A_{m2},*,...,*}_\text{lda}, - ..., - \underbrace{A_{1n},A_{2n},A_{3n},...,A_{mn},*,...,*}_\text{lda}} - _\text{lda x n}] - - - For row major layout, - - .. math:: - - \scriptstyle a = - [\underbrace{\underbrace{A_{11},A_{12},A_{13},...,A_{1n},*,...,*}_\text{lda}, - \underbrace{A_{21},A_{22},A_{23},...,A_{2n},*,...,*}_\text{lda}, - ..., - \underbrace{A_{m1},A_{m2},A_{m3},...,A_{mn},*,...,*}_\text{lda}} - _\text{m x lda}] - - .. container:: section - - .. rubric:: Triangular Matrix - :name: triangular-matrix - :class: sectiontitle - - A triangular matrix ``A`` of ``n`` rows and ``n`` columns with - leading dimension ``lda`` is represented as a one dimensional - array ``a``, of a size of at least ``lda`` \* ``n``. When column - (respectively row) major layout is used, the elements of each - column (respectively row) are contiguous in memory while the - elements of each row (respectively column) are at distance - ``lda`` from the element in the same row (respectively column) - and the previous column (respectively row). - - Before entry in any BLAS function using a triangular matrix, - - - If ``upper_lower = uplo::upper``, the leading ``n`` by ``n`` - upper triangular part of the array ``a`` must contain the upper - triangular part of the matrix ``A``. The strictly lower - triangular part of the array ``a`` is not referenced. In other - words, the matrix - - .. math:: - - A = \begin{bmatrix} - A_{11} & A_{12} & A_{13} & \ldots & A_{1n}\\ - * & A_{22} & A_{23} & \ldots & A_{2n}\\ - * & * & A_{33} & \ldots & A_{3n}\\ - \vdots & \vdots & \vdots & \ddots & \vdots\\ - * & * & * & \ldots & A_{nn} - \end{bmatrix} - - is stored in memory as the array - - - For column major layout, - - .. math:: - - \scriptstyle a = - [\underbrace{\underbrace{A_{11},*,...,*}_\text{lda}, - \underbrace{A_{12},A_{22},*,...,*}_\text{lda}, - ..., - \underbrace{A_{1n},A_{2n},A_{3n},...,A_{nn},*,...,*}_\text{lda}} - _\text{lda x n}] - - - For row major layout, - - .. math:: - - \scriptstyle a = - [\underbrace{\underbrace{A_{11},A_{12},A_{13},...,A_{1n},*,...,*}_\text{lda}, - \underbrace{*,A_{22},A_{23},...,A_{2n},*,...,*}_\text{lda}, - ..., - \underbrace{*,...,*,A_{nn},*,...,*}_\text{lda}} - _\text{lda x n}] - - - If ``upper_lower = uplo::lower``, the leading ``n`` by ``n`` - lower triangular part of the array ``a`` must contain the lower - triangular part of the matrix ``A``. The strictly upper - triangular part of the array ``a`` is not referenced. That is, - the matrix - - .. math:: - - A = \begin{bmatrix} - A_{11} & * & * & \ldots & * \\ - A_{21} & A_{22} & * & \ldots & * \\ - A_{31} & A_{32} & A_{33} & \ldots & * \\ - \vdots & \vdots & \vdots & \ddots & \vdots\\ - A_{n1} & A_{n2} & A_{n3} & \ldots & A_{nn} - \end{bmatrix} - - is stored in memory as the array - - - For column major layout, - - .. math:: - - \scriptstyle a = - [\underbrace{\underbrace{A_{11},A_{21},A_{31},..,A_{n1},*,...,*}_\text{lda}, - \underbrace{*,A_{22},A_{32},...,A_{n2},*,...,*}_\text{lda}, - ..., - \underbrace{*,...,*,A_{nn},*,...,*}_\text{lda}} - _\text{lda x n}] - - - For row major layout, - - .. math:: - - \scriptstyle a = - [\underbrace{\underbrace{A_{11},*,...,*}_\text{lda}, - \underbrace{A_{21},A_{22},*,...,*}_\text{lda}, - ..., - \underbrace{A_{n1},A_{n2},A_{n3},...,A_{nn},*,...,*}_\text{lda}} - _\text{lda x n}] - - .. container:: section - - .. rubric:: Band Matrix - :name: band-matrix - :class: sectiontitle - - A general band matrix ``A`` of ``m`` rows and ``n`` columns with - ``kl`` sub-diagonals, ``ku`` super-diagonals, and leading - dimension ``lda`` is represented as a one dimensional array - ``a`` of a size of at least ``lda`` \* ``n`` (respectively - ``lda`` \* ``m``) if column (respectively row) major layout is - used. - - Before entry in any BLAS function using a general band matrix, - the leading (``kl`` + ``ku`` + 1\ ``)`` by ``n`` (respectively - ``m``) part of the array ``a`` must contain the matrix - ``A``. This matrix must be supplied column-by-column - (respectively row-by-row), with the main diagonal of the matrix - in row ``ku`` (respectively ``kl``) of the array (0-based - indexing), the first super-diagonal starting at position 1 - (respectively 0) in row (``ku`` - 1) (respectively column - (``kl`` + 1)), the first sub-diagonal starting at position 0 - (respectively 1) in row (``ku`` + 1) (respectively column - (``kl`` - 1)), and so on. Elements in the array ``a`` that do - not correspond to elements in the band matrix (such as the top - left ``ku`` by ``ku`` triangle) are not referenced. - - Visually, the matrix ``A`` - - .. math:: - - A = \left[\begin{smallmatrix} - A_{11} & A_{12} & A_{13} & \ldots & A_{1,ku+1} & * & \ldots & \ldots & \ldots & \ldots & \ldots & * \\ - A_{21} & A_{22} & A_{23} & A_{24} & \ldots & A_{2,ku+2} & * & \ldots & \ldots & \ldots & \ldots & * \\ - A_{31} & A_{32} & A_{33} & A_{34} & A_{35} & \ldots & A_{3,ku+3} & * & \ldots & \ldots & \ldots & * \\ - \vdots & A_{42} & A_{43} & \ddots & \ddots & \ddots & \ddots & \ddots & * & \ldots & \ldots & \vdots \\ - A_{kl+1,1} & \vdots & A_{53} & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & * & \ldots & \vdots \\ - * & A_{kl+2,2} & \vdots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \vdots \\ - \vdots & * & A_{kl+3,3} & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & * \\ - \vdots & \vdots & * & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & A_{n-ku,n}\\ - \vdots & \vdots & \vdots & * & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \vdots \\ - \vdots & \vdots & \vdots & \vdots & * & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & A_{m-2,n} \\ - \vdots & \vdots & \vdots & \vdots & \vdots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & A_{m-1,n} \\ - * & * & * & \ldots & \ldots & \ldots & * & A_{m,m-kl} & \ldots & A_{m,n-2} & A_{m,n-1} & A_{m,n} - \end{smallmatrix}\right] - - - is stored in memory as an array - - - For column major layout, - - .. math:: - - \scriptscriptstyle a = - [\underbrace{ - \underbrace{\underbrace{*,...,*}_\text{ku},A_{11}, A_{12},...,A_{min(kl+1,m),1},*,...,*}_\text{lda}, - \underbrace{\underbrace{*,...,*}_\text{ku-1},A_{max(1,2-ku),2},...,A_{min(kl+2,m),2},*,...*}_\text{lda}, - ..., - \underbrace{\underbrace{*,...,*}_\text{max(0,ku-n+1)},A_{max(1,n-ku),n},...,A_{min(kl+n,m),n},*,...*}_\text{lda} - }_\text{lda x n}] - - - - For row major layout, - - .. math:: - - \scriptscriptstyle a = - [\underbrace{ - \underbrace{\underbrace{*,...,*}_\text{kl},A_{11}, A_{12},...,A_{1,min(ku+1,n)},*,...,*}_\text{lda}, - \underbrace{\underbrace{*,...,*}_\text{kl-1},A_{2,max(1,2-kl)},...,A_{2,min(ku+2,n)},*,...*}_\text{lda}, - ..., - \underbrace{\underbrace{*,...,*}_\text{max(0,kl-m+1)},A_{m,max(1,m-kl)},...,A_{m,min(ku+m,n)},*,...*}_\text{lda} - }_\text{lda x m}] - - The following program segment transfers a band matrix from - conventional full matrix storage (variable ``matrix``, with - leading dimension ``ldm``) to band storage (variable ``a``, with - leading dimension ``lda``): - - - - Using matrices stored with column major layout, - - :: - - for (j = 0; j < n; j++) { - k = ku – j; - for (i = max(0, j – ku); i < min(m, j + kl + 1); i++) { - a[(k + i) + j * lda] = matrix[i + j * ldm]; - } - } - - - Using matrices stored with row major layout, - - :: - - for (i = 0; i < m; i++) { - k = kl – i; - for (j = max(0, i – kl); j < min(n, i + ku + 1); j++) { - a[(k + j) + i * lda] = matrix[j + i * ldm]; - } - } - - - .. container:: section - - .. rubric:: Triangular Band Matrix - :name: triangular-band-matrix - :class: sectiontitle - - A triangular band matrix ``A`` of ``n`` rows and ``n`` columns - with ``k`` sub/super-diagonals and leading dimension ``lda`` is - represented as a one dimensional array ``a`` of size at least - ``lda`` \* ``n``. - - Before entry in any BLAS function using a triangular band matrix, - - - - If ``upper_lower = uplo::upper``, the leading (``k`` + 1) by ``n`` - part of the array ``a`` must contain the upper - triangular band part of the matrix ``A``. When using column - major layout, this matrix must be supplied column-by-column - (respectively row-by-row) with the main diagonal of the - matrix in row (``k``) (respectively column 0) of the array, - the first super-diagonal starting at position 1 - (respectively 0) in row (``k`` - 1) (respectively column 1), - and so on. Elements in the array ``a`` that do not correspond - to elements in the triangular band matrix (such as the top - left ``k`` by ``k`` triangle) are not referenced. - - Visually, the matrix - - .. math:: - - A = \left[\begin{smallmatrix} - A_{11} & A_{12} & A_{13} & \ldots & A_{1,k+1} & * & \ldots & \ldots & \ldots & \ldots & \ldots & * \\ - * & A_{22} & A_{23} & A_{24} & \ldots & A_{2,k+2} & * & \ldots & \ldots & \ldots & \ldots & * \\ - \vdots & * & A_{33} & A_{34} & A_{35} & \ldots & A_{3,k+3} & * & \ldots & \ldots & \ldots & * \\ - \vdots & \vdots & * & \ddots & \ddots & \ddots & \ddots & \ddots & * & \ldots & \ldots & \vdots \\ - \vdots & \vdots & \vdots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & * & \ldots & \vdots \\ - \vdots & \vdots & \vdots & \vdots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \vdots \\ - \vdots & \vdots & \vdots & \vdots & \vdots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & * \\ - \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \ddots & \ddots & \ddots & \ddots & \ddots & A_{n-k,n}\\ - \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \ddots & \ddots & \ddots & \ddots & \vdots \\ - \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \ddots & \ddots & \ddots & A_{n-2,n} \\ - \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \vdots & \ddots & \ddots & A_{n-1,n} \\ - * & * & * & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & * & A_{n,n} - \end{smallmatrix}\right] - - is stored as an array - - .. container:: fignone - - - For column major layout, - - .. math:: - - \scriptstyle a = - [\underbrace{ - \underbrace{\underbrace{*,...,*}_\text{ku},A_{11},*,...,*}_\text{lda}, - \underbrace{\underbrace{*,...,*}_\text{ku-1},A_{max(1,2-k),2},...,A_{2,2},*,...*}_\text{lda}, - ..., - \underbrace{\underbrace{*,...,*}_\text{max(0,k-n+1)},A_{max(1,n-k),n},...,A_{n,n},*,...*}_\text{lda} - }_\text{lda x n}] - - - - For row major layout, - - .. math:: - - \scriptstyle a = - [\underbrace{ - \underbrace{A_{11},A_{21},...,A_{min(k+1,n),1},*,...,*}_\text{lda}, - \underbrace{A_{2,2},...,A_{min(k+2,n),2},*,...,*}_\text{lda}, - ..., - \underbrace{A_{n,n},*,...*}_\text{lda} - }_\text{lda x n}] - - The following program segment transfers a band matrix from - conventional full matrix storage (variable ``matrix``, with - leading dimension ``ldm``) to band storage (variable ``a``, - with leading dimension ``lda``): - - - Using matrices stored with column major layout, - - :: - - for (j = 0; j < n; j++) { - m = k – j; - for (i = max(0, j – k); i <= j; i++) { - a[(m + i) + j * lda] = matrix[i + j * ldm]; - } - } - - - Using matrices stored with column major layout, - - :: - - for (i = 0; i < n; i++) { - m = –i; - for (j = i; j < min(n, i + k + 1); j++) { - a[(m + j) + i * lda] = matrix[j + i * ldm]; - } - } - - - If ``upper_lower = uplo::lower``, the leading (``k`` + 1) by ``n`` - part of the array ``a`` must contain the upper triangular - band part of the matrix ``A``. This matrix must be supplied - column-by-column with the main diagonal of the matrix in row 0 - of the array, the first sub-diagonal starting at position 0 in - row 1, and so on. Elements in the array ``a`` that do not - correspond to elements in the triangular band matrix (such as - the bottom right ``k`` by ``k`` triangle) are not referenced. - - That is, the matrix - - .. math:: - - A = \left[\begin{smallmatrix} - A_{11} & * & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & * \\ - A_{21} & A_{22} & * & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & * \\ - A_{31} & A_{32} & A_{33} & * & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & * \\ - \vdots & A_{42} & A_{43} & \ddots & \ddots & \ldots & \ldots & \ldots & \ldots & \ldots & \ldots & \vdots \\ - A_{k+1,1} & \vdots & A_{53} & \ddots & \ddots & \ddots & \ldots & \ldots & \ldots & \ldots & \ldots & \vdots \\ - * & A_{k+2,2} & \vdots & \ddots & \ddots & \ddots & \ddots & \ldots & \ldots & \ldots & \ldots & \vdots \\ - \vdots & * & A_{k+3,3} & \ddots & \ddots & \ddots & \ddots & \ddots & \ldots & \ldots & \ldots & \vdots \\ - \vdots & \vdots & * & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \ldots & \ldots & \vdots \\ - \vdots & \vdots & \vdots & * & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \ldots & \vdots \\ - \vdots & \vdots & \vdots & \vdots & * & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & \vdots \\ - \vdots & \vdots & \vdots & \vdots & \vdots & \ddots & \ddots & \ddots & \ddots & \ddots & \ddots & * \\ - * & * & * & \ldots & \ldots & \ldots & * & A_{n,n-k} & \ldots & A_{n,n-2} & A_{n,n-1} & A_{n,n} - \end{smallmatrix}\right] - - - is stored as the array - - - .. container:: fignone - - - For column major layout, - - .. math:: - - \scriptstyle a = - [\underbrace{ - \underbrace{A_{11},A_{21},...,A_{min(k+1,n),1},*,...,*}_\text{lda}, - \underbrace{A_{2,2},...,A_{min(k+2,n),2},*,...,*}_\text{lda}, - ..., - \underbrace{A_{n,n},*,...*}_\text{lda} - }_\text{lda x n}] - - - For row major layout, - - .. math:: - - \scriptstyle a = - [\underbrace{ - \underbrace{\underbrace{*,...,*}_\text{k},A_{11},*,...,*}_\text{lda}, - \underbrace{\underbrace{*,...,*}_\text{k-1},A_{max(1,2-k),2},...,A_{2,2},*,...*}_\text{lda}, - ..., - \underbrace{\underbrace{*,...,*}_\text{max(0,k-n+1)},A_{max(1,n-k),n},...,A_{n,n},*,...*}_\text{lda} - }_\text{lda x n}] - - - The following program segment transfers a band matrix from - conventional full matrix storage (variable ``matrix``, with - leading dimension ``ldm``) to band storage (variable ``a``, - with leading dimension ``lda``): - - - Using matrices stored with column major layout, - - :: - - for (j = 0; j < n; j++) { - m = –j; - for (i = j; i < min(n, j + k + 1); i++) { - a[(m + i) + j * lda] = matrix[i + j * ldm]; - } - } - - - Using matrices stored with row major layout, - - :: - - for (i = 0; i < n; i++) { - m = k – i; - for (j = max(0, i – k); j <= i; j++) { - a[(m + j) + i * lda] = matrix[j + i * ldm]; - } - } - - - .. container:: section - - .. rubric:: Packed Triangular Matrix - :name: packed-triangular-matrix - :class: sectiontitle - - A triangular matrix ``A`` of ``n`` rows and ``n`` columns is - represented in packed format as a one dimensional array ``a`` of - size at least (``n``\ \*(``n`` + 1))/2. All elements in the upper - or lower part of the matrix ``A`` are stored contiguously in the - array ``a``. - - Before entry in any BLAS function using a triangular packed - matrix, - - - If ``upper_lower = uplo::upper``, if column (respectively row) - major layout is used, the first (``n``\ \*(``n`` + 1))/2 - elements in the array ``a`` must contain the upper triangular - part of the matrix ``A`` packed sequentially, column by column - (respectively row by row) so that ``a``\ [0] contains ``A``\ - :sub:`11`, ``a``\ [1] and ``a``\ [2] contain ``A``\ :sub:`12` - and ``A``\ :sub:`22` (respectively ``A``\ :sub:`13`) - respectively, and so on. Hence, the matrix - - .. math:: - - A = \begin{bmatrix} - A_{11} & A_{12} & A_{13} & \ldots & A_{1n}\\ - * & A_{22} & A_{23} & \ldots & A_{2n}\\ - * & * & A_{33} & \ldots & A_{3n}\\ - \vdots & \vdots & \vdots & \ddots & \vdots\\ - * & * & * & \ldots & A_{nn} - \end{bmatrix} - - is stored as the array - - - For column major layout, - - .. math:: - - \scriptstyle a = [A_{11},A_{12},A_{22},A_{13},A_{23},A_{33},...,A_{(n-1),n},A_{nn}] - - - For row major layout, - - .. math:: - - \scriptstyle a = [A_{11},A_{12},A_{13},...,A_{1n}, - A_{22},A_{23},...,A_{2n},..., - A_{(n-1),(n-1)},A_{(n-1),n},A_{nn}] - - - If ``upper_lower = uplo::lower``, if column (respectively row) - major layout is used, the first (``n``\ \*(``n`` + 1))/2 - elements in the array ``a`` must contain the lower triangular - part of the matrix ``A`` packed sequentially, column by column - (row by row) so that ``a``\ [0] contains ``A``\ :sub:`11`, - ``a``\ [1] and ``a``\ [2] contain ``A``\ :sub:`21` and ``A``\ - :sub:`31` (respectively ``A``\ :sub:`22`) respectively, and so - on. The matrix - - .. math:: - - A = \begin{bmatrix} - A_{11} & * & * & \ldots & * \\ - A_{21} & A_{22} & * & \ldots & * \\ - A_{31} & A_{32} & A_{33} & \ldots & * \\ - \vdots & \vdots & \vdots & \ddots & \vdots\\ - A_{n1} & A_{n2} & A_{n3} & \ldots & A_{nn} - \end{bmatrix} - - is stored as the array - - - For column major layout, - - .. math:: - - \scriptstyle a = [A_{11},A_{21},A_{31},...,A_{n1}, - A_{22},A_{32},...,A_{n2},..., - A_{(n-1),(n-1)},A_{n,(n-1)},A_{nn}] - - - For row major layout, - - .. math:: - - \scriptstyle a = [A_{11},A_{21},A_{22},A_{31},A_{32},A_{33},...,A_{n,(n-1)},A_{nn}] - - .. container:: section - - .. rubric:: Vector - :name: vector - :class: sectiontitle - - A vector ``X`` of ``n`` elements with increment ``incx`` is - represented as a one dimensional array ``x`` of size at least (1 + - (``n`` - 1) \* abs(``incx``)). - - Visually, the vector - - .. math:: - - X = (X_{1},X_{2}, X_{3},...,X_{n}) - - is stored in memory as an array - - - .. math:: - - \scriptstyle x = [\underbrace{ - \underbrace{X_{1},*,...,*}_\text{incx}, - \underbrace{X_{2},*,...,*}_\text{incx}, - ..., - \underbrace{X_{n-1},*,...,*}_\text{incx},X_{n} - }_\text{1 + (n-1) x incx}] \quad if \:incx \:> \:0 - - .. math:: - - \scriptstyle x = [\underbrace{ - \underbrace{X_{n},*,...,*}_\text{|incx|}, - \underbrace{X_{n-1},*,...,*}_\text{|incx|}, - ..., - \underbrace{X_{2},*,...,*}_\text{|incx|},X_{1} - }_\text{1 + (1-n) x incx}] \quad if \:incx \:< \:0 - - - - diff --git a/docs/index.rst b/docs/index.rst index e17eeff6c..2a371255d 100644 --- a/docs/index.rst +++ b/docs/index.rst @@ -31,7 +31,5 @@ Contents :maxdepth: 2 :includehidden: - onemkl-datatypes.rst - domains/dense_linear_algebra.rst domains/sparse_linear_algebra.rst create_new_backend.rst diff --git a/docs/onemkl-datatypes.rst b/docs/onemkl-datatypes.rst deleted file mode 100644 index 33a08e834..000000000 --- a/docs/onemkl-datatypes.rst +++ /dev/null @@ -1,140 +0,0 @@ -.. _onemkl_datatypes: - -oneMKL Defined Datatypes -======================== - - -oneMKL BLAS and LAPACK for Data Parallel C++ (DPC++) introduces -several new enumeration data types, which are type-safe versions of -the traditional Fortran characters in BLAS and LAPACK. They are -declared in ``types.hpp``, which is included automatically when -you include ``mkl.hpp``. Like all oneMKL DPC++ functionality, they belong to the namespace ``oneapi::mkl``. - - -Each enumeration value comes with two names: A single-character name -(the traditional BLAS/LAPACK character) and a longer, descriptive -name. The two names are exactly equivalent and may be used -interchangeably. - - -transpose ---------- - -The ``transpose`` type specifies whether an input matrix should be -transposed and/or conjugated. It can take the following values: - - -.. list-table:: - :header-rows: 1 - - * - Short Name - - Long Name - - Description - * - ``transpose::N`` - - ``transpose::nontrans`` - - Do not transpose or conjugate the matrix. - * - ``transpose::T`` - - ``transpose::trans`` - - Transpose the matrix. - * - ``transpose::C`` - - ``transpose::conjtrans`` - - Perform Hermitian transpose (transpose and conjugate). Only applicable to complex matrices. - - - - -uplo ----- - -The ``uplo`` type specifies whether the lower or upper triangle of a riangular, symmetric, or Hermitian matrix should be accessed. - -It can take the following values: - - -.. list-table:: - :header-rows: 1 - - * - Short Name - - Long Name - - Description - * - ``uplo::U`` - - ``uplo::upper`` - - Access the upper triangle of the matrix. - * - ``uplo::L`` - - ``uplo::lower`` - - Access the lower triangle of the matrix. - - - - -In both cases, elements that are not in the selected triangle are not accessed or updated. - - -diag ----- - - -The ``diag`` type specifies the values on the diagonal of a triangular matrix. It can take the following values: - - -.. list-table:: - :header-rows: 1 - - * - Short Name - - Long Name - - Description - * - ``diag::N`` - - ``diag::nonunit`` - - The matrix is not unit triangular. The diagonal entries are stored with the matrix data. - * - ``diag::U`` - - ``diag::unit`` - - The matrix is unit triangular (the diagonal entries are all 1s). The diagonal entries in the matrix data are not accessed. - - - - -side ----- - - -The ``side`` type specifies the order of matrix multiplication when one matrix has a special form (triangular, symmetric, or Hermitian): - - -.. list-table:: - :header-rows: 1 - - * - Short Name - - Long Name - - Description - * - ``side::L`` - - ``side::left`` - - The special form matrix is on the left in the multiplication. - * - ``side::R`` - - ``side::right`` - - The special form matrix is on the right in the multiplication. - - -offset ------- - - -The ``offset`` type specifies whether the offset to apply to an output matrix is a fix offset, column offset or row offset. It can take the following values - - -.. list-table:: - :header-rows: 1 - - * - Short Name - - Long Name - - Description - * - ``offset::F`` - - ``offset::fix`` - - The offset to apply to the output matrix is fix, all the inputs in the ``C_offset`` matrix has the same value given by the first element in the ``co`` array. - * - ``offset::C`` - - ``offset::column`` - - The offset to apply to the output matrix is a column offset, that is to say all the columns in the ``C_offset`` matrix are the same and given by the elements in the ``co`` array. - * - ``offset::R`` - - ``offset::row`` - - The offset to apply to the output matrix is a row offset, that is to say all the rows in the ``C_offset`` matrix are the same and given by the elements in the ``co`` array. - -**Parent topic:** :ref:`onemkl` From 68d3966da511857c4b4a2eb8526510a822db4d53 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Mon, 7 Oct 2024 16:52:26 +0200 Subject: [PATCH 02/41] Move and rename files --- .github/scripts/domain-check.js | 6 +-- CONTRIBUTING.md | 2 +- README.md | 6 +-- cmake/CMakeLists.txt | 2 +- docs/create_new_backend.rst | 44 +++++++++---------- docs/index.rst | 2 +- ...cmake.rst => using_onemath_with_cmake.rst} | 0 .../level3/gemm_usm_mklcpu_cublas.cpp | 2 +- .../run_time_dispatching/level3/gemm_usm.cpp | 2 +- .../complex_fwd_usm_mklcpu_cufft.cpp | 2 +- .../dft/run_time_dispatching/real_fwd_usm.cpp | 2 +- .../getrs_usm_mklcpu_cusolver.cpp | 2 +- .../lapack/run_time_dispatching/getrs_usm.cpp | 2 +- .../uniform_usm_mklcpu_curand.cpp | 2 +- examples/rng/device/uniform.cpp | 2 +- .../rng/run_time_dispatching/uniform_usm.cpp | 2 +- .../sparse_blas_spmv_usm_mklcpu.cpp | 2 +- .../sparse_blas_spmv_usm.cpp | 2 +- include/oneapi/{mkl.hpp => math.hpp} | 18 ++++---- include/oneapi/{mkl => math}/bfloat16.hpp | 0 include/oneapi/{mkl => math}/blas.hpp | 20 ++++----- include/oneapi/{mkl => math}/blas.hxx | 0 .../blas/detail/blas_ct_backends.hpp | 4 +- .../blas/detail/blas_ct_backends.hxx | 0 .../{mkl => math}/blas/detail/blas_loader.hpp | 6 +-- .../{mkl => math}/blas/detail/blas_loader.hxx | 0 .../blas/detail/cublas/blas_ct.hpp | 8 ++-- .../blas/detail/cublas/blas_ct.hxx | 0 .../detail/cublas/onemath_blas_cublas.hpp} | 6 +-- .../detail/cublas/onemath_blas_cublas.hxx} | 0 .../blas/detail/mklcpu/blas_ct.hpp | 8 ++-- .../blas/detail/mklcpu/blas_ct.hxx | 0 .../detail/mklcpu/onemath_blas_mklcpu.hpp} | 8 ++-- .../blas/detail/mklgpu/blas_ct.hpp | 8 ++-- .../blas/detail/mklgpu/blas_ct.hxx | 0 .../detail/mklgpu/onemath_blas_mklgpu.hpp} | 8 ++-- .../blas/detail/netlib/blas_ct.hpp | 8 ++-- .../blas/detail/netlib/blas_ct.hxx | 0 .../detail/netlib/onemath_blas_netlib.hpp} | 8 ++-- .../blas/detail/onemath_blas_backends.hxx} | 0 .../blas/detail/portblas/blas_ct.hpp | 8 ++-- .../blas/detail/portblas/blas_ct.hxx | 0 .../portblas/onemath_blas_portblas.hpp} | 8 ++-- .../blas/detail/rocblas/blas_ct.hpp | 10 ++--- .../blas/detail/rocblas/blas_ct.hxx | 0 .../detail/rocblas/onemath_blas_rocblas.hpp} | 8 ++-- .../detail/rocblas/onemath_blas_rocblas.hxx} | 0 .../{mkl => math}/detail/backend_selector.hpp | 6 +-- .../detail/backend_selector_predicates.hpp | 6 +-- .../oneapi/{mkl => math}/detail/backends.hpp | 0 .../{mkl => math}/detail/backends_table.hpp | 2 +- .../{mkl => math}/detail/exceptions.hpp | 2 +- .../oneapi/{mkl => math}/detail/export.hpp | 2 +- .../{mkl => math}/detail/get_device_id.hpp | 4 +- include/oneapi/{mkl => math}/dft.hpp | 12 ++--- include/oneapi/{mkl => math}/dft/backward.hpp | 0 .../oneapi/{mkl => math}/dft/descriptor.hpp | 0 .../{mkl => math}/dft/detail/commit_impl.hpp | 0 .../dft/detail/cufft/onemath_dft_cufft.hpp} | 6 +-- .../dft/detail/descriptor_impl.hpp | 8 ++-- .../{mkl => math}/dft/detail/dft_ct.hxx | 0 .../{mkl => math}/dft/detail/dft_loader.hpp | 4 +- .../dft/detail/external_workspace_helper.hpp | 6 +-- .../dft/detail/mklcpu/onemath_dft_mklcpu.hpp} | 6 +-- .../dft/detail/mklgpu/onemath_dft_mklgpu.hpp} | 6 +-- .../detail/portfft/onemath_dft_portfft.hpp} | 6 +-- .../dft/detail/rocfft/onemath_dft_rocfft.hpp} | 6 +-- .../{mkl => math}/dft/detail/types_impl.hpp | 0 include/oneapi/{mkl => math}/dft/forward.hpp | 0 include/oneapi/{mkl => math}/dft/types.hpp | 0 include/oneapi/{mkl => math}/exceptions.hpp | 2 +- include/oneapi/{mkl => math}/lapack.hpp | 12 ++--- .../lapack/detail/cusolver/lapack_ct.hpp | 8 ++-- .../lapack/detail/cusolver/lapack_ct.hxx | 0 .../cusolver/onemath_lapack_cusolver.hpp} | 6 +-- .../cusolver/onemath_lapack_cusolver.hxx} | 0 .../lapack/detail/lapack_loader.hpp | 8 ++-- .../{mkl => math}/lapack/detail/lapack_rt.hpp | 10 ++--- .../lapack/detail/mkl_common/lapack_ct.hxx | 0 .../mkl_common/onemath_lapack_backends.hxx} | 0 .../lapack/detail/mklcpu/lapack_ct.hpp | 10 ++--- .../detail/mklcpu/onemath_lapack_mklcpu.hpp} | 8 ++-- .../lapack/detail/mklgpu/lapack_ct.hpp | 10 ++--- .../detail/mklgpu/onemath_lapack_mklgpu.hpp} | 8 ++-- .../lapack/detail/rocsolver/lapack_ct.hpp | 8 ++-- .../lapack/detail/rocsolver/lapack_ct.hxx | 0 .../rocsolver/onemath_lapack_rocsolver.hpp} | 6 +-- .../rocsolver/onemath_lapack_rocsolver.hxx} | 0 .../{mkl => math}/lapack/exceptions.hpp | 0 include/oneapi/{mkl => math}/lapack/types.hpp | 0 include/oneapi/{mkl => math}/rng.hpp | 14 +++--- .../rng/detail/curand/onemath_rng_curand.hpp} | 4 +- .../{mkl => math}/rng/detail/engine_impl.hpp | 8 ++-- .../rng/detail/mklcpu/onemath_rng_mklcpu.hpp} | 4 +- .../rng/detail/mklgpu/onemath_rng_mklgpu.hpp} | 4 +- .../{mkl => math}/rng/detail/rng_loader.hpp | 6 +-- .../detail/rocrand/onemath_rng_rocrand.hpp} | 4 +- include/oneapi/{mkl => math}/rng/device.hpp | 8 ++-- .../rng/device/detail/bernoulli_impl.hpp | 0 .../rng/device/detail/beta_impl.hpp | 0 .../rng/device/detail/bits_impl.hpp | 0 .../rng/device/detail/distribution_base.hpp | 24 +++++----- .../rng/device/detail/engine_base.hpp | 8 ++-- .../rng/device/detail/exponential_impl.hpp | 0 .../rng/device/detail/gamma_impl.hpp | 0 .../rng/device/detail/gaussian_impl.hpp | 0 .../rng/device/detail/lognormal_impl.hpp | 0 .../rng/device/detail/mcg31m1_impl.hpp | 0 .../rng/device/detail/mcg59_impl.hpp | 0 .../rng/device/detail/mrg32k3a_impl.hpp | 2 +- .../detail/mrg32k3a_skip_ahead_matrix.hpp | 0 .../rng/device/detail/philox4x32x10_impl.hpp | 0 .../rng/device/detail/poisson_impl.hpp | 0 .../rng/device/detail/uniform_bits_impl.hpp | 0 .../rng/device/detail/uniform_impl.hpp | 0 .../rng/device/detail/vm_wrappers.hpp | 0 .../rng/device/distributions.hpp | 4 +- .../{mkl => math}/rng/device/engines.hpp | 6 +-- .../{mkl => math}/rng/device/functions.hpp | 2 +- .../oneapi/{mkl => math}/rng/device/types.hpp | 0 .../{mkl => math}/rng/distributions.hpp | 2 +- include/oneapi/{mkl => math}/rng/engines.hpp | 16 +++---- .../oneapi/{mkl => math}/rng/functions.hpp | 4 +- .../oneapi/{mkl => math}/rng/predicates.hpp | 4 +- include/oneapi/{mkl => math}/sparse_blas.hpp | 2 +- .../sparse_blas/detail/handles.hpp | 0 .../sparse_blas/detail/helper_types.hpp | 0 .../mklcpu/onemath_sparse_blas_mklcpu.hpp} | 8 ++-- .../detail/mklcpu/sparse_blas_ct.hpp | 8 ++-- .../mklgpu/onemath_sparse_blas_mklgpu.hpp} | 8 ++-- .../detail/mklgpu/sparse_blas_ct.hpp | 8 ++-- .../detail/onemath_sparse_blas_backends.hxx} | 0 .../sparse_blas/detail/operation_types.hpp | 0 .../sparse_blas/detail/sparse_blas_ct.hxx | 2 +- .../sparse_blas/detail/sparse_blas_rt.hpp | 4 +- .../{mkl => math}/sparse_blas/matrix_view.hpp | 2 +- .../{mkl => math}/sparse_blas/types.hpp | 2 +- include/oneapi/{mkl => math}/types.hpp | 2 +- scripts/generate_backend_api.py | 4 +- scripts/generate_cmake.py | 2 +- scripts/generate_ct_instant.py | 6 +-- scripts/generate_ct_templates.py | 6 +-- scripts/generate_wrappers.py | 4 +- src/CMakeLists.txt | 10 ++--- src/blas/backends/cublas/cublas_batch.cpp | 4 +- .../backends/cublas/cublas_extensions.cpp | 4 +- src/blas/backends/cublas/cublas_helper.hpp | 2 +- src/blas/backends/cublas/cublas_level1.cpp | 4 +- src/blas/backends/cublas/cublas_level2.cpp | 4 +- src/blas/backends/cublas/cublas_level3.cpp | 4 +- src/blas/backends/cublas/cublas_task.hpp | 2 +- src/blas/backends/cublas/cublas_wrappers.cpp | 2 +- .../backends/mkl_common/mkl_blas_backend.hpp | 2 +- src/blas/backends/mklcpu/mklcpu_batch.cpp | 4 +- .../backends/mklcpu/mklcpu_extensions.cpp | 4 +- src/blas/backends/mklcpu/mklcpu_level1.cpp | 2 +- src/blas/backends/mklcpu/mklcpu_level2.cpp | 2 +- src/blas/backends/mklcpu/mklcpu_level3.cpp | 2 +- src/blas/backends/mklcpu/mklcpu_wrappers.cpp | 2 +- src/blas/backends/mklgpu/mklgpu_batch.cpp | 4 +- .../backends/mklgpu/mklgpu_extensions.cpp | 4 +- src/blas/backends/mklgpu/mklgpu_level1.cpp | 2 +- src/blas/backends/mklgpu/mklgpu_level2.cpp | 2 +- src/blas/backends/mklgpu/mklgpu_level3.cpp | 2 +- src/blas/backends/mklgpu/mklgpu_wrappers.cpp | 2 +- src/blas/backends/netlib/netlib_batch.cpp | 4 +- src/blas/backends/netlib/netlib_common.hpp | 4 +- .../backends/netlib/netlib_extensions.cpp | 4 +- src/blas/backends/netlib/netlib_level1.cpp | 4 +- src/blas/backends/netlib/netlib_level2.cpp | 2 +- src/blas/backends/netlib/netlib_level3.cpp | 4 +- src/blas/backends/netlib/netlib_wrappers.cpp | 2 +- src/blas/backends/portblas/portblas_batch.cpp | 4 +- .../backends/portblas/portblas_common.hpp | 4 +- .../portblas/portblas_level1_double.cpp | 4 +- .../portblas/portblas_level1_float.cpp | 4 +- .../portblas/portblas_level2_double.cpp | 4 +- .../portblas/portblas_level2_float.cpp | 4 +- .../portblas/portblas_level3_bfloat16.cpp | 4 +- .../portblas/portblas_level3_double.cpp | 4 +- .../portblas/portblas_level3_float.cpp | 4 +- .../portblas/portblas_level3_half.cpp | 4 +- .../backends/portblas/portblas_wrappers.cpp | 2 +- src/blas/backends/rocblas/rocblas_batch.cpp | 4 +- .../backends/rocblas/rocblas_extensions.cpp | 4 +- src/blas/backends/rocblas/rocblas_helper.hpp | 2 +- src/blas/backends/rocblas/rocblas_level1.cpp | 4 +- src/blas/backends/rocblas/rocblas_level2.cpp | 4 +- src/blas/backends/rocblas/rocblas_level3.cpp | 4 +- src/blas/backends/rocblas/rocblas_task.hpp | 2 +- .../backends/rocblas/rocblas_wrappers.cpp | 2 +- src/blas/blas_loader.cpp | 2 +- src/blas/function_table.hpp | 2 +- src/dft/backends/cufft/CMakeLists.txt | 2 +- src/dft/backends/cufft/backward.cpp | 8 ++-- src/dft/backends/cufft/commit.cpp | 10 ++--- ..._cufft_wrappers.cpp => cufft_wrappers.cpp} | 2 +- src/dft/backends/cufft/descriptor.cpp | 4 +- src/dft/backends/cufft/execute_helper.hpp | 8 ++-- src/dft/backends/cufft/forward.cpp | 8 ++-- src/dft/backends/descriptor.cpp | 4 +- src/dft/backends/mklcpu/backward.cpp | 8 ++-- src/dft/backends/mklcpu/commit.cpp | 12 ++--- .../backends/mklcpu/commit_derived_impl.hpp | 4 +- src/dft/backends/mklcpu/descriptor.cpp | 4 +- src/dft/backends/mklcpu/forward.cpp | 8 ++-- .../backends/mklcpu/mkl_dft_cpu_wrappers.cpp | 2 +- src/dft/backends/mklcpu/mklcpu_helpers.hpp | 4 +- src/dft/backends/mklgpu/backward.cpp | 6 +-- src/dft/backends/mklgpu/commit.cpp | 14 +++--- src/dft/backends/mklgpu/descriptor.cpp | 4 +- src/dft/backends/mklgpu/forward.cpp | 6 +-- .../backends/mklgpu/mkl_dft_gpu_wrappers.cpp | 2 +- src/dft/backends/mklgpu/mklgpu_helpers.hpp | 4 +- src/dft/backends/portfft/CMakeLists.txt | 2 +- src/dft/backends/portfft/commit.cpp | 10 ++--- src/dft/backends/portfft/descriptor.cpp | 4 +- src/dft/backends/portfft/portfft_helper.hpp | 4 +- ...tfft_wrappers.cpp => portfft_wrappers.cpp} | 2 +- src/dft/backends/rocfft/CMakeLists.txt | 2 +- src/dft/backends/rocfft/backward.cpp | 6 +-- src/dft/backends/rocfft/commit.cpp | 10 ++--- src/dft/backends/rocfft/descriptor.cpp | 4 +- src/dft/backends/rocfft/execute_helper.hpp | 8 ++-- src/dft/backends/rocfft/forward.cpp | 6 +-- ...ocfft_wrappers.cpp => rocfft_wrappers.cpp} | 2 +- src/dft/descriptor.cxx | 4 +- src/dft/descriptor_config_helper.hpp | 2 +- src/dft/dft_loader.cpp | 8 ++-- src/dft/function_table.hpp | 6 +-- src/include/function_table_initializer.hpp | 4 +- .../backends/cusolver/cusolver_batch.cpp | 4 +- .../backends/cusolver/cusolver_helper.hpp | 8 ++-- .../backends/cusolver/cusolver_lapack.cpp | 4 +- .../backends/cusolver/cusolver_task.hpp | 2 +- .../backends/cusolver/cusolver_wrappers.cpp | 2 +- .../mkl_common/mkl_lapack_backend.hpp | 4 +- .../backends/mklcpu/lapack_cpu_wrappers.cpp | 2 +- src/lapack/backends/mklcpu/mkl_lapack.cpp | 6 +-- .../backends/mklgpu/lapack_gpu_wrappers.cpp | 2 +- src/lapack/backends/mklgpu/mkl_lapack.cpp | 6 +-- .../backends/rocsolver/rocsolver_batch.cpp | 4 +- .../backends/rocsolver/rocsolver_helper.hpp | 8 ++-- .../backends/rocsolver/rocsolver_lapack.cpp | 4 +- .../backends/rocsolver/rocsolver_task.hpp | 2 +- .../backends/rocsolver/rocsolver_wrappers.cpp | 2 +- src/lapack/function_table.hpp | 2 +- src/lapack/lapack_loader.cpp | 2 +- src/rng/backends/curand/CMakeLists.txt | 2 +- src/rng/backends/curand/curand_helper.hpp | 2 +- ...urand_wrappers.cpp => curand_wrappers.cpp} | 2 +- src/rng/backends/curand/mrg32k3a.cpp | 8 ++-- src/rng/backends/curand/philox4x32x10.cpp | 8 ++-- .../backends/mklcpu/mkl_rng_cpu_wrappers.cpp | 2 +- src/rng/backends/mklcpu/mrg32k3a.cpp | 6 +-- src/rng/backends/mklcpu/philox4x32x10.cpp | 6 +-- .../backends/mklgpu/mkl_rng_gpu_wrappers.cpp | 2 +- src/rng/backends/mklgpu/mrg32k3a.cpp | 8 ++-- src/rng/backends/mklgpu/philox4x32x10.cpp | 8 ++-- src/rng/backends/rocrand/CMakeLists.txt | 2 +- src/rng/backends/rocrand/mrg32k3a.cpp | 8 ++-- src/rng/backends/rocrand/philox4x32x10.cpp | 8 ++-- src/rng/backends/rocrand/rocrand_helper.hpp | 2 +- ...rand_wrappers.cpp => rocrand_wrappers.cpp} | 2 +- src/rng/function_table.hpp | 2 +- src/rng/rng_loader.cpp | 2 +- .../backends/mkl_common/mkl_handles.hpp | 2 +- .../backends/mkl_common/mkl_helper.hpp | 4 +- .../backends/mklcpu/mklcpu_handles.cpp | 2 +- .../backends/mklcpu/mklcpu_operations.cpp | 2 +- .../backends/mklcpu/mklcpu_wrappers.cpp | 4 +- .../backends/mklgpu/mklgpu_handles.cpp | 2 +- .../backends/mklgpu/mklgpu_operations.cpp | 2 +- .../backends/mklgpu/mklgpu_wrappers.cpp | 4 +- src/sparse_blas/function_table.hpp | 2 +- src/sparse_blas/generic_container.hpp | 2 +- src/sparse_blas/matrix_view_comparison.hpp | 2 +- src/sparse_blas/sparse_blas_loader.cpp | 4 +- .../blas/batch/axpy_batch_stride.cpp | 6 +-- .../blas/batch/axpy_batch_stride_usm.cpp | 6 +-- .../unit_tests/blas/batch/axpy_batch_usm.cpp | 6 +-- .../blas/batch/copy_batch_stride.cpp | 6 +-- .../blas/batch/copy_batch_stride_usm.cpp | 6 +-- .../unit_tests/blas/batch/copy_batch_usm.cpp | 6 +-- .../blas/batch/dgmm_batch_stride.cpp | 6 +-- .../blas/batch/dgmm_batch_stride_usm.cpp | 6 +-- .../unit_tests/blas/batch/dgmm_batch_usm.cpp | 6 +-- .../blas/batch/gemm_batch_stride.cpp | 6 +-- .../blas/batch/gemm_batch_stride_usm.cpp | 6 +-- .../unit_tests/blas/batch/gemm_batch_usm.cpp | 6 +-- .../blas/batch/gemv_batch_stride.cpp | 6 +-- .../blas/batch/gemv_batch_stride_usm.cpp | 6 +-- .../unit_tests/blas/batch/gemv_batch_usm.cpp | 6 +-- .../blas/batch/imatcopy_batch_stride.cpp | 6 +-- .../blas/batch/imatcopy_batch_stride_usm.cpp | 6 +-- .../blas/batch/imatcopy_batch_usm.cpp | 6 +-- .../blas/batch/omatadd_batch_stride.cpp | 6 +-- .../blas/batch/omatadd_batch_stride_usm.cpp | 6 +-- .../blas/batch/omatcopy_batch_stride.cpp | 6 +-- .../blas/batch/omatcopy_batch_stride_usm.cpp | 6 +-- .../blas/batch/omatcopy_batch_usm.cpp | 6 +-- .../blas/batch/syrk_batch_stride.cpp | 6 +-- .../blas/batch/syrk_batch_stride_usm.cpp | 6 +-- .../unit_tests/blas/batch/syrk_batch_usm.cpp | 6 +-- .../blas/batch/trsm_batch_stride.cpp | 6 +-- .../blas/batch/trsm_batch_stride_usm.cpp | 6 +-- .../unit_tests/blas/batch/trsm_batch_usm.cpp | 6 +-- .../unit_tests/blas/extensions/gemm_bias.cpp | 6 +-- .../blas/extensions/gemm_bias_usm.cpp | 6 +-- tests/unit_tests/blas/extensions/gemmt.cpp | 6 +-- .../unit_tests/blas/extensions/gemmt_usm.cpp | 6 +-- tests/unit_tests/blas/extensions/imatcopy.cpp | 6 +-- .../blas/extensions/imatcopy_usm.cpp | 6 +-- tests/unit_tests/blas/extensions/omatadd.cpp | 6 +-- .../blas/extensions/omatadd_usm.cpp | 6 +-- tests/unit_tests/blas/extensions/omatcopy.cpp | 6 +-- .../unit_tests/blas/extensions/omatcopy2.cpp | 6 +-- .../blas/extensions/omatcopy2_usm.cpp | 6 +-- .../blas/extensions/omatcopy_usm.cpp | 6 +-- ...las_helper.hpp => onemath_blas_helper.hpp} | 8 ++-- .../blas/include/reference_blas_templates.hpp | 2 +- .../blas/include/reference_blas_wrappers.hpp | 4 +- tests/unit_tests/blas/level1/asum.cpp | 6 +-- tests/unit_tests/blas/level1/asum_usm.cpp | 6 +-- tests/unit_tests/blas/level1/axpby.cpp | 6 +-- tests/unit_tests/blas/level1/axpby_usm.cpp | 6 +-- tests/unit_tests/blas/level1/axpy.cpp | 6 +-- tests/unit_tests/blas/level1/axpy_usm.cpp | 6 +-- tests/unit_tests/blas/level1/copy.cpp | 6 +-- tests/unit_tests/blas/level1/copy_usm.cpp | 6 +-- tests/unit_tests/blas/level1/dot.cpp | 6 +-- tests/unit_tests/blas/level1/dot_usm.cpp | 6 +-- tests/unit_tests/blas/level1/dotc.cpp | 6 +-- tests/unit_tests/blas/level1/dotc_usm.cpp | 6 +-- tests/unit_tests/blas/level1/dotu.cpp | 6 +-- tests/unit_tests/blas/level1/dotu_usm.cpp | 6 +-- tests/unit_tests/blas/level1/iamax.cpp | 6 +-- tests/unit_tests/blas/level1/iamax_usm.cpp | 6 +-- tests/unit_tests/blas/level1/iamin.cpp | 6 +-- tests/unit_tests/blas/level1/iamin_usm.cpp | 6 +-- tests/unit_tests/blas/level1/nrm2.cpp | 6 +-- tests/unit_tests/blas/level1/nrm2_usm.cpp | 6 +-- tests/unit_tests/blas/level1/rot.cpp | 6 +-- tests/unit_tests/blas/level1/rot_usm.cpp | 6 +-- tests/unit_tests/blas/level1/rotg.cpp | 6 +-- tests/unit_tests/blas/level1/rotg_usm.cpp | 6 +-- tests/unit_tests/blas/level1/rotm.cpp | 6 +-- tests/unit_tests/blas/level1/rotm_usm.cpp | 6 +-- tests/unit_tests/blas/level1/rotmg.cpp | 6 +-- tests/unit_tests/blas/level1/rotmg_usm.cpp | 6 +-- tests/unit_tests/blas/level1/scal.cpp | 6 +-- tests/unit_tests/blas/level1/scal_usm.cpp | 6 +-- tests/unit_tests/blas/level1/sdsdot.cpp | 6 +-- tests/unit_tests/blas/level1/sdsdot_usm.cpp | 6 +-- tests/unit_tests/blas/level1/swap.cpp | 6 +-- tests/unit_tests/blas/level1/swap_usm.cpp | 6 +-- tests/unit_tests/blas/level2/gbmv.cpp | 6 +-- tests/unit_tests/blas/level2/gbmv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/gemv.cpp | 6 +-- tests/unit_tests/blas/level2/gemv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/ger.cpp | 6 +-- tests/unit_tests/blas/level2/ger_usm.cpp | 6 +-- tests/unit_tests/blas/level2/gerc.cpp | 6 +-- tests/unit_tests/blas/level2/gerc_usm.cpp | 6 +-- tests/unit_tests/blas/level2/geru.cpp | 6 +-- tests/unit_tests/blas/level2/geru_usm.cpp | 6 +-- tests/unit_tests/blas/level2/hbmv.cpp | 6 +-- tests/unit_tests/blas/level2/hbmv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/hemv.cpp | 6 +-- tests/unit_tests/blas/level2/hemv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/her.cpp | 6 +-- tests/unit_tests/blas/level2/her2.cpp | 6 +-- tests/unit_tests/blas/level2/her2_usm.cpp | 6 +-- tests/unit_tests/blas/level2/her_usm.cpp | 6 +-- tests/unit_tests/blas/level2/hpmv.cpp | 6 +-- tests/unit_tests/blas/level2/hpmv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/hpr.cpp | 6 +-- tests/unit_tests/blas/level2/hpr2.cpp | 6 +-- tests/unit_tests/blas/level2/hpr2_usm.cpp | 6 +-- tests/unit_tests/blas/level2/hpr_usm.cpp | 6 +-- tests/unit_tests/blas/level2/sbmv.cpp | 6 +-- tests/unit_tests/blas/level2/sbmv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/spmv.cpp | 6 +-- tests/unit_tests/blas/level2/spmv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/spr.cpp | 6 +-- tests/unit_tests/blas/level2/spr2.cpp | 6 +-- tests/unit_tests/blas/level2/spr2_usm.cpp | 6 +-- tests/unit_tests/blas/level2/spr_usm.cpp | 6 +-- tests/unit_tests/blas/level2/symv.cpp | 6 +-- tests/unit_tests/blas/level2/symv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/syr.cpp | 6 +-- tests/unit_tests/blas/level2/syr2.cpp | 6 +-- tests/unit_tests/blas/level2/syr2_usm.cpp | 6 +-- tests/unit_tests/blas/level2/syr_usm.cpp | 6 +-- tests/unit_tests/blas/level2/tbmv.cpp | 6 +-- tests/unit_tests/blas/level2/tbmv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/tbsv.cpp | 6 +-- tests/unit_tests/blas/level2/tbsv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/tpmv.cpp | 6 +-- tests/unit_tests/blas/level2/tpmv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/tpsv.cpp | 6 +-- tests/unit_tests/blas/level2/tpsv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/trmv.cpp | 6 +-- tests/unit_tests/blas/level2/trmv_usm.cpp | 6 +-- tests/unit_tests/blas/level2/trsv.cpp | 6 +-- tests/unit_tests/blas/level2/trsv_usm.cpp | 6 +-- tests/unit_tests/blas/level3/gemm.cpp | 6 +-- tests/unit_tests/blas/level3/gemm_usm.cpp | 6 +-- tests/unit_tests/blas/level3/hemm.cpp | 6 +-- tests/unit_tests/blas/level3/hemm_usm.cpp | 6 +-- tests/unit_tests/blas/level3/her2k.cpp | 6 +-- tests/unit_tests/blas/level3/her2k_usm.cpp | 6 +-- tests/unit_tests/blas/level3/herk.cpp | 6 +-- tests/unit_tests/blas/level3/herk_usm.cpp | 6 +-- tests/unit_tests/blas/level3/symm.cpp | 6 +-- tests/unit_tests/blas/level3/symm_usm.cpp | 6 +-- tests/unit_tests/blas/level3/syr2k.cpp | 6 +-- tests/unit_tests/blas/level3/syr2k_usm.cpp | 6 +-- tests/unit_tests/blas/level3/syrk.cpp | 6 +-- tests/unit_tests/blas/level3/syrk_usm.cpp | 6 +-- tests/unit_tests/blas/level3/trmm.cpp | 6 +-- tests/unit_tests/blas/level3/trmm_usm.cpp | 6 +-- tests/unit_tests/blas/level3/trsm.cpp | 6 +-- tests/unit_tests/blas/level3/trsm_usm.cpp | 6 +-- .../dft/include/compute_inplace.hpp | 2 +- .../unit_tests/dft/include/compute_tester.hpp | 2 +- .../unit_tests/dft/include/reference_dft.hpp | 2 +- tests/unit_tests/include/test_helper.hpp | 6 +-- .../lapack/include/lapack_common.hpp | 2 +- .../lapack/include/lapack_test_controller.hpp | 2 +- tests/unit_tests/lapack/source/gebrd.cpp | 2 +- tests/unit_tests/lapack/source/geqrf.cpp | 2 +- .../lapack/source/geqrf_batch_group.cpp | 2 +- .../lapack/source/geqrf_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/gerqf.cpp | 2 +- tests/unit_tests/lapack/source/gesvd.cpp | 2 +- tests/unit_tests/lapack/source/getrf.cpp | 2 +- .../lapack/source/getrf_batch_group.cpp | 2 +- .../lapack/source/getrf_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/getri.cpp | 2 +- .../lapack/source/getri_batch_group.cpp | 2 +- .../lapack/source/getri_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/getrs.cpp | 2 +- .../lapack/source/getrs_batch_group.cpp | 2 +- .../lapack/source/getrs_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/heevd.cpp | 2 +- tests/unit_tests/lapack/source/hegvd.cpp | 2 +- tests/unit_tests/lapack/source/hetrd.cpp | 2 +- tests/unit_tests/lapack/source/hetrf.cpp | 2 +- tests/unit_tests/lapack/source/orgbr.cpp | 2 +- tests/unit_tests/lapack/source/orgqr.cpp | 2 +- .../lapack/source/orgqr_batch_group.cpp | 2 +- .../lapack/source/orgqr_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/orgtr.cpp | 2 +- tests/unit_tests/lapack/source/ormqr.cpp | 2 +- tests/unit_tests/lapack/source/ormrq.cpp | 2 +- tests/unit_tests/lapack/source/ormtr.cpp | 2 +- tests/unit_tests/lapack/source/potrf.cpp | 2 +- .../lapack/source/potrf_batch_group.cpp | 2 +- .../lapack/source/potrf_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/potri.cpp | 2 +- tests/unit_tests/lapack/source/potrs.cpp | 2 +- .../lapack/source/potrs_batch_group.cpp | 2 +- .../lapack/source/potrs_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/syevd.cpp | 2 +- tests/unit_tests/lapack/source/sygvd.cpp | 2 +- tests/unit_tests/lapack/source/sytrd.cpp | 2 +- tests/unit_tests/lapack/source/sytrf.cpp | 2 +- tests/unit_tests/lapack/source/trtrs.cpp | 2 +- tests/unit_tests/lapack/source/ungbr.cpp | 2 +- tests/unit_tests/lapack/source/ungqr.cpp | 2 +- .../lapack/source/ungqr_batch_group.cpp | 2 +- .../lapack/source/ungqr_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/ungtr.cpp | 2 +- tests/unit_tests/lapack/source/unmqr.cpp | 2 +- tests/unit_tests/lapack/source/unmrq.cpp | 2 +- tests/unit_tests/lapack/source/unmtr.cpp | 2 +- tests/unit_tests/main_test.cpp | 4 +- .../unit_tests/rng/device/include/moments.hpp | 2 +- .../rng/device/include/skip_ahead_test.hpp | 2 +- .../rng/include/engines_api_tests.hpp | 2 +- .../rng/include/skip_ahead_test.hpp | 2 +- .../rng/include/statistics_check.hpp | 2 +- .../rng/include/statistics_check_test.hpp | 2 +- .../include/common_sparse_reference.hpp | 2 +- .../sparse_blas/include/test_spmm.hpp | 4 +- .../sparse_blas/include/test_spmv.hpp | 4 +- .../sparse_blas/include/test_spsv.hpp | 4 +- 488 files changed, 1080 insertions(+), 1080 deletions(-) rename docs/{using_onemkl_with_cmake.rst => using_onemath_with_cmake.rst} (100%) rename include/oneapi/{mkl.hpp => math.hpp} (74%) rename include/oneapi/{mkl => math}/bfloat16.hpp (100%) rename include/oneapi/{mkl => math}/blas.hpp (75%) rename include/oneapi/{mkl => math}/blas.hxx (100%) rename include/oneapi/{mkl => math}/blas/detail/blas_ct_backends.hpp (96%) rename include/oneapi/{mkl => math}/blas/detail/blas_ct_backends.hxx (100%) rename include/oneapi/{mkl => math}/blas/detail/blas_loader.hpp (91%) rename include/oneapi/{mkl => math}/blas/detail/blas_loader.hxx (100%) rename include/oneapi/{mkl => math}/blas/detail/cublas/blas_ct.hpp (87%) rename include/oneapi/{mkl => math}/blas/detail/cublas/blas_ct.hxx (100%) rename include/oneapi/{mkl/blas/detail/cublas/onemkl_blas_cublas.hpp => math/blas/detail/cublas/onemath_blas_cublas.hpp} (93%) rename include/oneapi/{mkl/blas/detail/cublas/onemkl_blas_cublas.hxx => math/blas/detail/cublas/onemath_blas_cublas.hxx} (100%) rename include/oneapi/{mkl => math}/blas/detail/mklcpu/blas_ct.hpp (86%) rename include/oneapi/{mkl => math}/blas/detail/mklcpu/blas_ct.hxx (100%) rename include/oneapi/{mkl/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp => math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp} (87%) rename include/oneapi/{mkl => math}/blas/detail/mklgpu/blas_ct.hpp (87%) rename include/oneapi/{mkl => math}/blas/detail/mklgpu/blas_ct.hxx (100%) rename include/oneapi/{mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp => math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp} (86%) rename include/oneapi/{mkl => math}/blas/detail/netlib/blas_ct.hpp (87%) rename include/oneapi/{mkl => math}/blas/detail/netlib/blas_ct.hxx (100%) rename include/oneapi/{mkl/blas/detail/netlib/onemkl_blas_netlib.hpp => math/blas/detail/netlib/onemath_blas_netlib.hpp} (87%) rename include/oneapi/{mkl/blas/detail/onemkl_blas_backends.hxx => math/blas/detail/onemath_blas_backends.hxx} (100%) rename include/oneapi/{mkl => math}/blas/detail/portblas/blas_ct.hpp (87%) rename include/oneapi/{mkl => math}/blas/detail/portblas/blas_ct.hxx (100%) rename include/oneapi/{mkl/blas/detail/portblas/onemkl_blas_portblas.hpp => math/blas/detail/portblas/onemath_blas_portblas.hpp} (87%) rename include/oneapi/{mkl => math}/blas/detail/rocblas/blas_ct.hpp (86%) rename include/oneapi/{mkl => math}/blas/detail/rocblas/blas_ct.hxx (100%) rename include/oneapi/{mkl/blas/detail/rocblas/onemkl_blas_rocblas.hpp => math/blas/detail/rocblas/onemath_blas_rocblas.hpp} (91%) rename include/oneapi/{mkl/blas/detail/rocblas/onemkl_blas_rocblas.hxx => math/blas/detail/rocblas/onemath_blas_rocblas.hxx} (100%) rename include/oneapi/{mkl => math}/detail/backend_selector.hpp (89%) rename include/oneapi/{mkl => math}/detail/backend_selector_predicates.hpp (97%) rename include/oneapi/{mkl => math}/detail/backends.hpp (100%) rename include/oneapi/{mkl => math}/detail/backends_table.hpp (99%) rename include/oneapi/{mkl => math}/detail/exceptions.hpp (98%) rename include/oneapi/{mkl => math}/detail/export.hpp (96%) rename include/oneapi/{mkl => math}/detail/get_device_id.hpp (95%) rename include/oneapi/{mkl => math}/dft.hpp (79%) rename include/oneapi/{mkl => math}/dft/backward.hpp (100%) rename include/oneapi/{mkl => math}/dft/descriptor.hpp (100%) rename include/oneapi/{mkl => math}/dft/detail/commit_impl.hpp (100%) rename include/oneapi/{mkl/dft/detail/cufft/onemkl_dft_cufft.hpp => math/dft/detail/cufft/onemath_dft_cufft.hpp} (88%) rename include/oneapi/{mkl => math}/dft/detail/descriptor_impl.hpp (94%) rename include/oneapi/{mkl => math}/dft/detail/dft_ct.hxx (100%) rename include/oneapi/{mkl => math}/dft/detail/dft_loader.hpp (94%) rename include/oneapi/{mkl => math}/dft/detail/external_workspace_helper.hpp (98%) rename include/oneapi/{mkl/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp => math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp} (88%) rename include/oneapi/{mkl/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp => math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp} (88%) rename include/oneapi/{mkl/dft/detail/portfft/onemkl_dft_portfft.hpp => math/dft/detail/portfft/onemath_dft_portfft.hpp} (90%) rename include/oneapi/{mkl/dft/detail/rocfft/onemkl_dft_rocfft.hpp => math/dft/detail/rocfft/onemath_dft_rocfft.hpp} (88%) rename include/oneapi/{mkl => math}/dft/detail/types_impl.hpp (100%) rename include/oneapi/{mkl => math}/dft/forward.hpp (100%) rename include/oneapi/{mkl => math}/dft/types.hpp (100%) rename include/oneapi/{mkl => math}/exceptions.hpp (99%) rename include/oneapi/{mkl => math}/lapack.hpp (74%) rename include/oneapi/{mkl => math}/lapack/detail/cusolver/lapack_ct.hpp (86%) rename include/oneapi/{mkl => math}/lapack/detail/cusolver/lapack_ct.hxx (100%) rename include/oneapi/{mkl/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp => math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp} (91%) rename include/oneapi/{mkl/lapack/detail/cusolver/onemkl_lapack_cusolver.hxx => math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx} (100%) rename include/oneapi/{mkl => math}/lapack/detail/lapack_loader.hpp (99%) rename include/oneapi/{mkl => math}/lapack/detail/lapack_rt.hpp (99%) rename include/oneapi/{mkl => math}/lapack/detail/mkl_common/lapack_ct.hxx (100%) rename include/oneapi/{mkl/lapack/detail/mkl_common/onemkl_lapack_backends.hxx => math/lapack/detail/mkl_common/onemath_lapack_backends.hxx} (100%) rename include/oneapi/{mkl => math}/lapack/detail/mklcpu/lapack_ct.hpp (81%) rename include/oneapi/{mkl/lapack/detail/mklcpu/onemkl_lapack_mklcpu.hpp => math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp} (85%) rename include/oneapi/{mkl => math}/lapack/detail/mklgpu/lapack_ct.hpp (81%) rename include/oneapi/{mkl/lapack/detail/mklgpu/onemkl_lapack_mklgpu.hpp => math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp} (85%) rename include/oneapi/{mkl => math}/lapack/detail/rocsolver/lapack_ct.hpp (87%) rename include/oneapi/{mkl => math}/lapack/detail/rocsolver/lapack_ct.hxx (100%) rename include/oneapi/{mkl/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp => math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp} (92%) rename include/oneapi/{mkl/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hxx => math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx} (100%) rename include/oneapi/{mkl => math}/lapack/exceptions.hpp (100%) rename include/oneapi/{mkl => math}/lapack/types.hpp (100%) rename include/oneapi/{mkl => math}/rng.hpp (76%) rename include/oneapi/{mkl/rng/detail/curand/onemkl_rng_curand.hpp => math/rng/detail/curand/onemath_rng_curand.hpp} (97%) rename include/oneapi/{mkl => math}/rng/detail/engine_impl.hpp (98%) rename include/oneapi/{mkl/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp => math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp} (95%) rename include/oneapi/{mkl/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp => math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp} (95%) rename include/oneapi/{mkl => math}/rng/detail/rng_loader.hpp (92%) rename include/oneapi/{mkl/rng/detail/rocrand/onemkl_rng_rocrand.hpp => math/rng/detail/rocrand/onemath_rng_rocrand.hpp} (98%) rename include/oneapi/{mkl => math}/rng/device.hpp (82%) rename include/oneapi/{mkl => math}/rng/device/detail/bernoulli_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/beta_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/bits_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/distribution_base.hpp (75%) rename include/oneapi/{mkl => math}/rng/device/detail/engine_base.hpp (83%) rename include/oneapi/{mkl => math}/rng/device/detail/exponential_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/gamma_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/gaussian_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/lognormal_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/mcg31m1_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/mcg59_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/mrg32k3a_impl.hpp (99%) rename include/oneapi/{mkl => math}/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/philox4x32x10_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/poisson_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/uniform_bits_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/uniform_impl.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/detail/vm_wrappers.hpp (100%) rename include/oneapi/{mkl => math}/rng/device/distributions.hpp (99%) rename include/oneapi/{mkl => math}/rng/device/engines.hpp (97%) rename include/oneapi/{mkl => math}/rng/device/functions.hpp (96%) rename include/oneapi/{mkl => math}/rng/device/types.hpp (100%) rename include/oneapi/{mkl => math}/rng/distributions.hpp (99%) rename include/oneapi/{mkl => math}/rng/engines.hpp (95%) rename include/oneapi/{mkl => math}/rng/functions.hpp (98%) rename include/oneapi/{mkl => math}/rng/predicates.hpp (96%) rename include/oneapi/{mkl => math}/sparse_blas.hpp (96%) rename include/oneapi/{mkl => math}/sparse_blas/detail/handles.hpp (100%) rename include/oneapi/{mkl => math}/sparse_blas/detail/helper_types.hpp (100%) rename include/oneapi/{mkl/sparse_blas/detail/mklcpu/onemkl_sparse_blas_mklcpu.hpp => math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp} (84%) rename include/oneapi/{mkl => math}/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp (85%) rename include/oneapi/{mkl/sparse_blas/detail/mklgpu/onemkl_sparse_blas_mklgpu.hpp => math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp} (84%) rename include/oneapi/{mkl => math}/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp (85%) rename include/oneapi/{mkl/sparse_blas/detail/onemkl_sparse_blas_backends.hxx => math/sparse_blas/detail/onemath_sparse_blas_backends.hxx} (100%) rename include/oneapi/{mkl => math}/sparse_blas/detail/operation_types.hpp (100%) rename include/oneapi/{mkl => math}/sparse_blas/detail/sparse_blas_ct.hxx (99%) rename include/oneapi/{mkl => math}/sparse_blas/detail/sparse_blas_rt.hpp (99%) rename include/oneapi/{mkl => math}/sparse_blas/matrix_view.hpp (97%) rename include/oneapi/{mkl => math}/sparse_blas/types.hpp (98%) rename include/oneapi/{mkl => math}/types.hpp (98%) rename src/dft/backends/cufft/{mkl_dft_cufft_wrappers.cpp => cufft_wrappers.cpp} (94%) rename src/dft/backends/portfft/{mkl_dft_portfft_wrappers.cpp => portfft_wrappers.cpp} (94%) rename src/dft/backends/rocfft/{mkl_dft_rocfft_wrappers.cpp => rocfft_wrappers.cpp} (94%) rename src/rng/backends/curand/{mkl_rng_curand_wrappers.cpp => curand_wrappers.cpp} (98%) rename src/rng/backends/rocrand/{mkl_rng_rocrand_wrappers.cpp => rocrand_wrappers.cpp} (98%) rename tests/unit_tests/blas/include/{onemkl_blas_helper.hpp => onemath_blas_helper.hpp} (95%) diff --git a/.github/scripts/domain-check.js b/.github/scripts/domain-check.js index ed68c2e42..7a6368ec2 100644 --- a/.github/scripts/domain-check.js +++ b/.github/scripts/domain-check.js @@ -14,7 +14,7 @@ function matchesPattern(domain, filePaths) { !filePath.startsWith("third-party-programs/"), ); // These directories contain domain specific code - const dirs = "(tests/unit_tests|examples|src|include/oneapi/mkl)"; + const dirs = "(tests/unit_tests|examples|src|include/oneapi/math)"; const domains = "(blas|lapack|rng|dft)"; // matches changes to the domain of interest or non domain-specific code const re = new RegExp(`^(${dirs}/${domain}|(?!${dirs}/${domains}))`); @@ -92,7 +92,7 @@ test_patterns = [ }, { domain: "lapack", - files: ["include/oneapi/mkl/lapack/lapack.hpp"], + files: ["include/oneapi/math/lapack/lapack.hpp"], expected: true, }, { @@ -117,7 +117,7 @@ test_patterns = [ }, { domain: "lapack", - files: ["include/oneapi/mkl/rng/lapack.hpp"], + files: ["include/oneapi/math/rng/lapack.hpp"], expected: false, }, { diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 2840fcbd6..f29d70bd6 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -195,7 +195,7 @@ class table { }; * **NS4:** Put each namespace on its own line when declaring nested namespaces. ```c -#include "oneapi/mkl/blas/path_to_some_header.hpp" +#include "oneapi/math/blas/path_to_some_header.hpp" namespace oneapi { namespace mkl { diff --git a/README.md b/README.md index 5dc8c9c3b..94ef2b85a 100644 --- a/README.md +++ b/README.md @@ -94,7 +94,7 @@ There are two oneMKL selector layer implementations: Example of app.cpp with run-time dispatching: ```cpp - #include "oneapi/mkl.hpp" + #include "oneapi/math.hpp" ... cpu_dev = sycl::device(sycl::cpu_selector()); @@ -120,7 +120,7 @@ There are two oneMKL selector layer implementations: Example of app.cpp with compile-time dispatching: ```cpp - #include "oneapi/mkl.hpp" + #include "oneapi/math.hpp" ... cpu_dev = sycl::device(sycl::cpu_selector()); @@ -147,7 +147,7 @@ There are two oneMKL selector layer implementations: Header-based and backend-independent Device API can be called within ```sycl kernel``` or work from Host code ([device-rng-usage-model-example](https://spec.oneapi.io/versions/latest/elements/oneMKL/source/domains/rng/device_api/device-rng-usage-model.html#id2)). Currently, the following domains support the Device API: -- **RNG**. To use RNG Device API functionality it's required to include ```oneapi/mkl/rng/device.hpp``` header file. +- **RNG**. To use RNG Device API functionality it's required to include ```oneapi/math/rng/device.hpp``` header file. ### Supported Configurations: diff --git a/cmake/CMakeLists.txt b/cmake/CMakeLists.txt index df7d2fc4c..b564414df 100644 --- a/cmake/CMakeLists.txt +++ b/cmake/CMakeLists.txt @@ -21,7 +21,7 @@ install(FILES FindCompiler.cmake DESTINATION "lib/cmake/${PROJECT_NAME}" ) if(ENABLE_MKLGPU_BACKEND OR ENABLE_MKLCPU_BACKEND) - install(FILES mkl/MKLConfig.cmake + install(FILES math/MKLConfig.cmake DESTINATION "lib/cmake/${PROJECT_NAME}" ) endif() diff --git a/docs/create_new_backend.rst b/docs/create_new_backend.rst index b22df014d..faccd6085 100644 --- a/docs/create_new_backend.rst +++ b/docs/create_new_backend.rst @@ -44,11 +44,11 @@ For each new backend library, you should create the following two header files: .. code-block:: bash - python scripts/generate_backend_api.py include/oneapi/mkl/blas.hpp \ # Base header file - include/oneapi/mkl/blas/detail/newlib/onemkl_blas_newlib.hpp \ # Output header file + python scripts/generate_backend_api.py include/oneapi/math/blas.hpp \ # Base header file + include/oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp \ # Output header file oneapi::mkl::newlib # Wrappers namespace -Code snippet of the generated header file ``include/oneapi/mkl/blas/detail/newlib/onemkl_blas_newlib.hpp`` +Code snippet of the generated header file ``include/oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp`` .. code-block:: cpp @@ -65,14 +65,14 @@ Code snippet of the generated header file ``include/oneapi/mkl/blas/detail/newli .. code-block:: bash - python scripts/generate_ct_instant.py include/oneapi/mkl/blas/detail/blas_ct_templates.hpp \ # Base header file - include/oneapi/mkl/blas/detail/newlib/blas_ct.hpp \ # Output header file - include/oneapi/mkl/blas/detail/newlib/onemkl_blas_newlib.hpp \ # Header file with declaration of entry points to wrappers + python scripts/generate_ct_instant.py include/oneapi/math/blas/detail/blas_ct_templates.hpp \ # Base header file + include/oneapi/math/blas/detail/newlib/blas_ct.hpp \ # Output header file + include/oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp \ # Header file with declaration of entry points to wrappers newlib \ # Library name newdevice \ # Backend name oneapi::mkl::newlib # Wrappers namespace -Code snippet of the generated header file ``include/oneapi/mkl/blas/detail/newlib/blas_ct.hpp`` +Code snippet of the generated header file ``include/oneapi/math/blas/detail/newlib/blas_ct.hpp`` .. code-block:: cpp @@ -101,8 +101,8 @@ Below you can see structure of oneMKL top-level include directory: include/ oneapi/ - mkl/ - mkl.hpp -> oneMKL spec APIs + math/ + math.hpp -> oneMKL spec APIs types.hpp -> oneMKL spec types blas.hpp -> oneMKL BLAS APIs w/ pre-check/dispatching/post-check detail/ -> implementation specific header files @@ -127,7 +127,7 @@ Below you can see structure of oneMKL top-level include directory: To integrate the new third-party library to a oneMKL header-based part, following files from this structure should be updated: -* ``include/oneapi/mkl/detail/backends.hpp``: add the new backend +* ``include/oneapi/math/detail/backends.hpp``: add the new backend **Example**: add the ``newbackend`` backend @@ -142,7 +142,7 @@ To integrate the new third-party library to a oneMKL header-based part, followin static backendmap backend_map = { { backend::mklcpu, "mklcpu" }, + { backend::newbackend, "newbackend" }, -* ``include/oneapi/mkl/detail/backends_table.hpp``: add new backend library for supported domain(s) and device(s) +* ``include/oneapi/math/detail/backends_table.hpp``: add new backend library for supported domain(s) and device(s) **Example**: enable ``newlib`` for ``blas`` domain and ``newdevice`` device @@ -168,7 +168,7 @@ To integrate the new third-party library to a oneMKL header-based part, followin + #endif + } }, -* ``include/oneapi/mkl/detail/get_device_id.hpp``: add new device detection mechanism for Run-time dispatching +* ``include/oneapi/math/detail/get_device_id.hpp``: add new device detection mechanism for Run-time dispatching **Example**: enable ``newdevice`` if the queue is targeted for the Host @@ -179,15 +179,15 @@ To integrate the new third-party library to a oneMKL header-based part, followin + if (queue.is_host()) + device_id=device::newdevice; -* ``include/oneapi/mkl/blas.hpp``: include the generated header file for the compile-time dispatching interface (see `oneMKL Usage Models <../README.md#supported-usage-models>`_) +* ``include/oneapi/math/blas.hpp``: include the generated header file for the compile-time dispatching interface (see `oneMKL Usage Models <../README.md#supported-usage-models>`_) - **Example**: add ``include/oneapi/mkl/blas/detail/newlib/blas_ct.hpp`` generated at the `1. Create Header Files`_ step + **Example**: add ``include/oneapi/math/blas/detail/newlib/blas_ct.hpp`` generated at the `1. Create Header Files`_ step .. code-block:: diff - #include "oneapi/mkl/blas/detail/mklcpu/blas_ct.hpp" - #include "oneapi/mkl/blas/detail/mklgpu/blas_ct.hpp" - + #include "oneapi/mkl/blas/detail/newlib/blas_ct.hpp" + #include "oneapi/math/blas/detail/mklcpu/blas_ct.hpp" + #include "oneapi/math/blas/detail/mklgpu/blas_ct.hpp" + + #include "oneapi/math/blas/detail/newlib/blas_ct.hpp" The new files generated at the `1. Create Header Files`_ step result in the following updated structure of the BLAS domain header files. @@ -196,7 +196,7 @@ The new files generated at the `1. Create Header Files`_ step result in the foll include/ oneapi/ - mkl/ + math/ blas.hpp -> oneMKL BLAS APIs w/ pre-check/dispatching/post-check blas/ predicates.hpp -> oneMKL BLAS pre-check post-check @@ -248,12 +248,12 @@ You can modify wrappers generated with this script to enable third-party library The command below generates two new files: -* ``src/blas/backends/newlib/newlib_wrappers.cpp`` - DPC++ wrappers for all functions from ``include/oneapi/mkl/blas/detail/newlib/onemkl_blas_newlib.hpp`` +* ``src/blas/backends/newlib/newlib_wrappers.cpp`` - DPC++ wrappers for all functions from ``include/oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp`` * ``src/blas/backends/newlib/newlib_wrappers_table_dyn.cpp`` - structure of symbols for run-time dispatcher (in the same location as wrappers), suffix ``_dyn`` indicates that this file is required for dynamic library only. .. code-block:: bash - python scripts/generate_wrappers.py include/oneapi/mkl/blas/detail/newlib/onemkl_blas_newlib.hpp \ # Base header file + python scripts/generate_wrappers.py include/oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp \ # Base header file src/blas/function_table.hpp \ # Declaration for structure of symbols src/blas/backends/newlib/newlib_wrappers.cpp \ # Output wrappers newlib # Library name @@ -276,9 +276,9 @@ The following code snippet is updated for ``src/blas/backends/newlib/newlib_wrap #include #endif - #include "oneapi/mkl/types.hpp" + #include "oneapi/math/types.hpp" - #include "oneapi/mkl/blas/detail/newlib/onemkl_blas_newlib.hpp" + #include "oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp" + + #include "newlib.h" diff --git a/docs/index.rst b/docs/index.rst index 2a371255d..96f5907af 100644 --- a/docs/index.rst +++ b/docs/index.rst @@ -24,7 +24,7 @@ Contents building_the_project_with_dpcpp.rst building_the_project_with_adaptivecpp.rst building_and_running_tests.rst - using_onemkl_with_cmake.rst + using_onemath_with_cmake.rst .. toctree:: :caption: Developer Reference diff --git a/docs/using_onemkl_with_cmake.rst b/docs/using_onemath_with_cmake.rst similarity index 100% rename from docs/using_onemkl_with_cmake.rst rename to docs/using_onemath_with_cmake.rst diff --git a/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp b/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp index 358c0b768..a8e5058e6 100644 --- a/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp +++ b/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp @@ -47,7 +47,7 @@ #else #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" // local includes #include "example_helper.hpp" diff --git a/examples/blas/run_time_dispatching/level3/gemm_usm.cpp b/examples/blas/run_time_dispatching/level3/gemm_usm.cpp index cd59e7b7f..e213e384b 100644 --- a/examples/blas/run_time_dispatching/level3/gemm_usm.cpp +++ b/examples/blas/run_time_dispatching/level3/gemm_usm.cpp @@ -47,7 +47,7 @@ #else #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "example_helper.hpp" diff --git a/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp b/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp index 59c810f3f..c7cc624b0 100644 --- a/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp +++ b/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp @@ -26,7 +26,7 @@ #else #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include void run_example(const sycl::device& cpu_device, const sycl::device& gpu_device) { diff --git a/examples/dft/run_time_dispatching/real_fwd_usm.cpp b/examples/dft/run_time_dispatching/real_fwd_usm.cpp index c220b0ee7..c5a8d50f2 100644 --- a/examples/dft/run_time_dispatching/real_fwd_usm.cpp +++ b/examples/dft/run_time_dispatching/real_fwd_usm.cpp @@ -28,7 +28,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" void run_example(const sycl::device& dev) { constexpr std::size_t N = 16; diff --git a/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp b/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp index 2d6017d08..9bb8cc646 100644 --- a/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp +++ b/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp @@ -41,7 +41,7 @@ #else #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" // local includes #include "example_helper.hpp" diff --git a/examples/lapack/run_time_dispatching/getrs_usm.cpp b/examples/lapack/run_time_dispatching/getrs_usm.cpp index 4cf851a7e..dde358f16 100644 --- a/examples/lapack/run_time_dispatching/getrs_usm.cpp +++ b/examples/lapack/run_time_dispatching/getrs_usm.cpp @@ -42,7 +42,7 @@ #else #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" // local includes #include "example_helper.hpp" diff --git a/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp b/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp index cdfd6c765..5698931d8 100644 --- a/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp +++ b/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp @@ -42,7 +42,7 @@ #else #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" // local includes #include "example_helper.hpp" diff --git a/examples/rng/device/uniform.cpp b/examples/rng/device/uniform.cpp index a1c097bba..9ccd37534 100644 --- a/examples/rng/device/uniform.cpp +++ b/examples/rng/device/uniform.cpp @@ -37,7 +37,7 @@ #include #endif -#include "oneapi/mkl/rng/device.hpp" +#include "oneapi/math/rng/device.hpp" #include "rng_example_helper.hpp" diff --git a/examples/rng/run_time_dispatching/uniform_usm.cpp b/examples/rng/run_time_dispatching/uniform_usm.cpp index 8ac7363c8..942ad5f11 100644 --- a/examples/rng/run_time_dispatching/uniform_usm.cpp +++ b/examples/rng/run_time_dispatching/uniform_usm.cpp @@ -42,7 +42,7 @@ #else #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" // local includes #include "example_helper.hpp" diff --git a/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp b/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp index 4ab078601..0f8956ad5 100644 --- a/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp +++ b/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp @@ -46,7 +46,7 @@ #else #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "example_helper.hpp" diff --git a/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp b/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp index d87297600..cc6908aae 100644 --- a/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp +++ b/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp @@ -47,7 +47,7 @@ #else #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "example_helper.hpp" diff --git a/include/oneapi/mkl.hpp b/include/oneapi/math.hpp similarity index 74% rename from include/oneapi/mkl.hpp rename to include/oneapi/math.hpp index f3e9b8618..229ae6c93 100644 --- a/include/oneapi/mkl.hpp +++ b/include/oneapi/math.hpp @@ -17,15 +17,15 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_HPP_ -#define _ONEMKL_HPP_ +#ifndef _ONEMATH_HPP_ +#define _ONEMATH_HPP_ -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" -#include "oneapi/mkl/blas.hpp" -#include "oneapi/mkl/dft.hpp" -#include "oneapi/mkl/lapack.hpp" -#include "oneapi/mkl/rng.hpp" -#include "oneapi/mkl/sparse_blas.hpp" +#include "oneapi/math/blas.hpp" +#include "oneapi/math/dft.hpp" +#include "oneapi/math/lapack.hpp" +#include "oneapi/math/rng.hpp" +#include "oneapi/math/sparse_blas.hpp" -#endif //_ONEMKL_HPP_ +#endif //_ONEMATH_HPP_ diff --git a/include/oneapi/mkl/bfloat16.hpp b/include/oneapi/math/bfloat16.hpp similarity index 100% rename from include/oneapi/mkl/bfloat16.hpp rename to include/oneapi/math/bfloat16.hpp diff --git a/include/oneapi/mkl/blas.hpp b/include/oneapi/math/blas.hpp similarity index 75% rename from include/oneapi/mkl/blas.hpp rename to include/oneapi/math/blas.hpp index 6b8066e69..0c0bd6bfa 100644 --- a/include/oneapi/mkl/blas.hpp +++ b/include/oneapi/math/blas.hpp @@ -28,29 +28,29 @@ #include #include -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math/types.hpp" -#include "oneapi/mkl/detail/get_device_id.hpp" +#include "oneapi/math/detail/get_device_id.hpp" -#include "oneapi/mkl/blas/detail/blas_loader.hpp" +#include "oneapi/math/blas/detail/blas_loader.hpp" #ifdef ONEMKL_ENABLE_CUBLAS_BACKEND -#include "oneapi/mkl/blas/detail/cublas/blas_ct.hpp" +#include "oneapi/math/blas/detail/cublas/blas_ct.hpp" #endif #ifdef ONEMKL_ENABLE_ROCBLAS_BACKEND -#include "oneapi/mkl/blas/detail/rocblas/blas_ct.hpp" +#include "oneapi/math/blas/detail/rocblas/blas_ct.hpp" #endif #ifdef ONEMKL_ENABLE_MKLCPU_BACKEND -#include "oneapi/mkl/blas/detail/mklcpu/blas_ct.hpp" +#include "oneapi/math/blas/detail/mklcpu/blas_ct.hpp" #endif #ifdef ONEMKL_ENABLE_MKLGPU_BACKEND -#include "oneapi/mkl/blas/detail/mklgpu/blas_ct.hpp" +#include "oneapi/math/blas/detail/mklgpu/blas_ct.hpp" #endif #ifdef ONEMKL_ENABLE_NETLIB_BACKEND -#include "oneapi/mkl/blas/detail/netlib/blas_ct.hpp" +#include "oneapi/math/blas/detail/netlib/blas_ct.hpp" #endif #ifdef ONEMKL_ENABLE_PORTBLAS_BACKEND -#include "oneapi/mkl/blas/detail/portblas/blas_ct.hpp" +#include "oneapi/math/blas/detail/portblas/blas_ct.hpp" #endif namespace oneapi { diff --git a/include/oneapi/mkl/blas.hxx b/include/oneapi/math/blas.hxx similarity index 100% rename from include/oneapi/mkl/blas.hxx rename to include/oneapi/math/blas.hxx diff --git a/include/oneapi/mkl/blas/detail/blas_ct_backends.hpp b/include/oneapi/math/blas/detail/blas_ct_backends.hpp similarity index 96% rename from include/oneapi/mkl/blas/detail/blas_ct_backends.hpp rename to include/oneapi/math/blas/detail/blas_ct_backends.hpp index eb894b5b9..a12e6778a 100644 --- a/include/oneapi/mkl/blas/detail/blas_ct_backends.hpp +++ b/include/oneapi/math/blas/detail/blas_ct_backends.hpp @@ -28,8 +28,8 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/backend_selector.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/blas/detail/blas_ct_backends.hxx b/include/oneapi/math/blas/detail/blas_ct_backends.hxx similarity index 100% rename from include/oneapi/mkl/blas/detail/blas_ct_backends.hxx rename to include/oneapi/math/blas/detail/blas_ct_backends.hxx diff --git a/include/oneapi/mkl/blas/detail/blas_loader.hpp b/include/oneapi/math/blas/detail/blas_loader.hpp similarity index 91% rename from include/oneapi/mkl/blas/detail/blas_loader.hpp rename to include/oneapi/math/blas/detail/blas_loader.hpp index 665f5dc80..0c1db264a 100644 --- a/include/oneapi/mkl/blas/detail/blas_loader.hpp +++ b/include/oneapi/math/blas/detail/blas_loader.hpp @@ -28,10 +28,10 @@ #include #endif -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/detail/get_device_id.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/detail/get_device_id.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/blas/detail/blas_loader.hxx b/include/oneapi/math/blas/detail/blas_loader.hxx similarity index 100% rename from include/oneapi/mkl/blas/detail/blas_loader.hxx rename to include/oneapi/math/blas/detail/blas_loader.hxx diff --git a/include/oneapi/mkl/blas/detail/cublas/blas_ct.hpp b/include/oneapi/math/blas/detail/cublas/blas_ct.hpp similarity index 87% rename from include/oneapi/mkl/blas/detail/cublas/blas_ct.hpp rename to include/oneapi/math/blas/detail/cublas/blas_ct.hpp index 2443c64d2..8b53854f4 100644 --- a/include/oneapi/mkl/blas/detail/cublas/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/cublas/blas_ct.hpp @@ -28,10 +28,10 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" -#include "oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hpp" -#include "oneapi/mkl/blas/detail/blas_ct_backends.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/backend_selector.hpp" +#include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" +#include "oneapi/math/blas/detail/blas_ct_backends.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/blas/detail/cublas/blas_ct.hxx b/include/oneapi/math/blas/detail/cublas/blas_ct.hxx similarity index 100% rename from include/oneapi/mkl/blas/detail/cublas/blas_ct.hxx rename to include/oneapi/math/blas/detail/cublas/blas_ct.hxx diff --git a/include/oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hpp b/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp similarity index 93% rename from include/oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hpp rename to include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp index 8a6d5448f..fa7360550 100644 --- a/include/oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hpp +++ b/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp @@ -26,7 +26,7 @@ #include #include #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" namespace oneapi { namespace mkl { @@ -39,12 +39,12 @@ namespace blas { namespace cublas { namespace column_major { -#include "onemkl_blas_cublas.hxx" +#include "onemath_blas_cublas.hxx" } //namespace column_major namespace row_major { -#include "onemkl_blas_cublas.hxx" +#include "onemath_blas_cublas.hxx" } //namespace row_major } //namespace cublas diff --git a/include/oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hxx b/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hxx similarity index 100% rename from include/oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hxx rename to include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hxx diff --git a/include/oneapi/mkl/blas/detail/mklcpu/blas_ct.hpp b/include/oneapi/math/blas/detail/mklcpu/blas_ct.hpp similarity index 86% rename from include/oneapi/mkl/blas/detail/mklcpu/blas_ct.hpp rename to include/oneapi/math/blas/detail/mklcpu/blas_ct.hpp index 24b0a9c7e..6f614fcda 100644 --- a/include/oneapi/mkl/blas/detail/mklcpu/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/mklcpu/blas_ct.hpp @@ -28,11 +28,11 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/backend_selector.hpp" -#include "oneapi/mkl/blas/detail/blas_ct_backends.hpp" -#include "oneapi/mkl/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/blas_ct_backends.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/blas/detail/mklcpu/blas_ct.hxx b/include/oneapi/math/blas/detail/mklcpu/blas_ct.hxx similarity index 100% rename from include/oneapi/mkl/blas/detail/mklcpu/blas_ct.hxx rename to include/oneapi/math/blas/detail/mklcpu/blas_ct.hxx diff --git a/include/oneapi/mkl/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp b/include/oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp similarity index 87% rename from include/oneapi/mkl/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp rename to include/oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp index db9e3f4d0..e0fcf3068 100644 --- a/include/oneapi/mkl/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp +++ b/include/oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp @@ -26,8 +26,8 @@ #endif #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/export.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/export.hpp" namespace oneapi { namespace mkl { @@ -40,12 +40,12 @@ namespace blas { namespace mklcpu { namespace column_major { -#include "oneapi/mkl/blas/detail/onemkl_blas_backends.hxx" +#include "oneapi/math/blas/detail/onemath_blas_backends.hxx" } // namespace column_major namespace row_major { -#include "oneapi/mkl/blas/detail/onemkl_blas_backends.hxx" +#include "oneapi/math/blas/detail/onemath_blas_backends.hxx" } // namespace row_major } // namespace mklcpu diff --git a/include/oneapi/mkl/blas/detail/mklgpu/blas_ct.hpp b/include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp similarity index 87% rename from include/oneapi/mkl/blas/detail/mklgpu/blas_ct.hpp rename to include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp index 10ceb3b73..f767a1678 100644 --- a/include/oneapi/mkl/blas/detail/mklgpu/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp @@ -28,11 +28,11 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/backends.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/backends.hpp" -#include "oneapi/mkl/blas/detail/blas_ct_backends.hpp" -#include "oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/blas_ct_backends.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/blas/detail/mklgpu/blas_ct.hxx b/include/oneapi/math/blas/detail/mklgpu/blas_ct.hxx similarity index 100% rename from include/oneapi/mkl/blas/detail/mklgpu/blas_ct.hxx rename to include/oneapi/math/blas/detail/mklgpu/blas_ct.hxx diff --git a/include/oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp b/include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp similarity index 86% rename from include/oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp rename to include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp index cf5a2a398..2263f2840 100644 --- a/include/oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp +++ b/include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp @@ -28,9 +28,9 @@ #include #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" -#include "oneapi/mkl/detail/export.hpp" +#include "oneapi/math/detail/export.hpp" namespace oneapi { namespace mkl { @@ -38,12 +38,12 @@ namespace blas { namespace mklgpu { namespace column_major { -#include "oneapi/mkl/blas/detail/onemkl_blas_backends.hxx" +#include "oneapi/math/blas/detail/onemath_blas_backends.hxx" } //namespace column_major namespace row_major { -#include "oneapi/mkl/blas/detail/onemkl_blas_backends.hxx" +#include "oneapi/math/blas/detail/onemath_blas_backends.hxx" } //namespace row_major } //namespace mklgpu diff --git a/include/oneapi/mkl/blas/detail/netlib/blas_ct.hpp b/include/oneapi/math/blas/detail/netlib/blas_ct.hpp similarity index 87% rename from include/oneapi/mkl/blas/detail/netlib/blas_ct.hpp rename to include/oneapi/math/blas/detail/netlib/blas_ct.hpp index cdfc79e7f..7c8b62405 100644 --- a/include/oneapi/mkl/blas/detail/netlib/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/netlib/blas_ct.hpp @@ -28,10 +28,10 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/backends.hpp" -#include "oneapi/mkl/blas/detail/netlib/onemkl_blas_netlib.hpp" -#include "oneapi/mkl/blas/detail/blas_ct_backends.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/backends.hpp" +#include "oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp" +#include "oneapi/math/blas/detail/blas_ct_backends.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/blas/detail/netlib/blas_ct.hxx b/include/oneapi/math/blas/detail/netlib/blas_ct.hxx similarity index 100% rename from include/oneapi/mkl/blas/detail/netlib/blas_ct.hxx rename to include/oneapi/math/blas/detail/netlib/blas_ct.hxx diff --git a/include/oneapi/mkl/blas/detail/netlib/onemkl_blas_netlib.hpp b/include/oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp similarity index 87% rename from include/oneapi/mkl/blas/detail/netlib/onemkl_blas_netlib.hpp rename to include/oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp index cb96c9a27..7e6e8e6a4 100644 --- a/include/oneapi/mkl/blas/detail/netlib/onemkl_blas_netlib.hpp +++ b/include/oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp @@ -29,9 +29,9 @@ #include #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" -#include "oneapi/mkl/detail/export.hpp" +#include "oneapi/math/detail/export.hpp" namespace oneapi { namespace mkl { @@ -46,12 +46,12 @@ namespace blas { namespace netlib { namespace column_major { -#include "oneapi/mkl/blas/detail/onemkl_blas_backends.hxx" +#include "oneapi/math/blas/detail/onemath_blas_backends.hxx" } //namespace column_major namespace row_major { -#include "oneapi/mkl/blas/detail/onemkl_blas_backends.hxx" +#include "oneapi/math/blas/detail/onemath_blas_backends.hxx" } //namespace row_major } //namespace netlib diff --git a/include/oneapi/mkl/blas/detail/onemkl_blas_backends.hxx b/include/oneapi/math/blas/detail/onemath_blas_backends.hxx similarity index 100% rename from include/oneapi/mkl/blas/detail/onemkl_blas_backends.hxx rename to include/oneapi/math/blas/detail/onemath_blas_backends.hxx diff --git a/include/oneapi/mkl/blas/detail/portblas/blas_ct.hpp b/include/oneapi/math/blas/detail/portblas/blas_ct.hpp similarity index 87% rename from include/oneapi/mkl/blas/detail/portblas/blas_ct.hpp rename to include/oneapi/math/blas/detail/portblas/blas_ct.hpp index 6d3b0b2c2..a7ffbe877 100644 --- a/include/oneapi/mkl/blas/detail/portblas/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/portblas/blas_ct.hpp @@ -28,10 +28,10 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" -#include "oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp" -#include "oneapi/mkl/blas/detail/blas_ct_backends.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/backend_selector.hpp" +#include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" +#include "oneapi/math/blas/detail/blas_ct_backends.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/blas/detail/portblas/blas_ct.hxx b/include/oneapi/math/blas/detail/portblas/blas_ct.hxx similarity index 100% rename from include/oneapi/mkl/blas/detail/portblas/blas_ct.hxx rename to include/oneapi/math/blas/detail/portblas/blas_ct.hxx diff --git a/include/oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp b/include/oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp similarity index 87% rename from include/oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp rename to include/oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp index c8d47d742..a8c29d02f 100644 --- a/include/oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp +++ b/include/oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp @@ -28,9 +28,9 @@ #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" -#include "oneapi/mkl/detail/export.hpp" +#include "oneapi/math/detail/export.hpp" namespace oneapi { namespace mkl { @@ -45,12 +45,12 @@ namespace blas { namespace portblas { namespace column_major { -#include "oneapi/mkl/blas/detail/onemkl_blas_backends.hxx" +#include "oneapi/math/blas/detail/onemath_blas_backends.hxx" } //namespace column_major namespace row_major { -#include "oneapi/mkl/blas/detail/onemkl_blas_backends.hxx" +#include "oneapi/math/blas/detail/onemath_blas_backends.hxx" } //namespace row_major } // namespace portblas diff --git a/include/oneapi/mkl/blas/detail/rocblas/blas_ct.hpp b/include/oneapi/math/blas/detail/rocblas/blas_ct.hpp similarity index 86% rename from include/oneapi/mkl/blas/detail/rocblas/blas_ct.hpp rename to include/oneapi/math/blas/detail/rocblas/blas_ct.hpp index 1a019b19e..e06bfb335 100644 --- a/include/oneapi/mkl/blas/detail/rocblas/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/rocblas/blas_ct.hpp @@ -30,11 +30,11 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hpp" -#include "oneapi/mkl/blas/detail/blas_ct_backends.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/backend_selector.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp" +#include "oneapi/math/blas/detail/blas_ct_backends.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/blas/detail/rocblas/blas_ct.hxx b/include/oneapi/math/blas/detail/rocblas/blas_ct.hxx similarity index 100% rename from include/oneapi/mkl/blas/detail/rocblas/blas_ct.hxx rename to include/oneapi/math/blas/detail/rocblas/blas_ct.hxx diff --git a/include/oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hpp b/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp similarity index 91% rename from include/oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hpp rename to include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp index a642e5609..0416d1562 100644 --- a/include/oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hpp +++ b/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp @@ -28,8 +28,8 @@ #include #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/config.hpp" namespace oneapi { namespace mkl { @@ -42,12 +42,12 @@ namespace blas { namespace rocblas { namespace column_major { -#include "onemkl_blas_rocblas.hxx" +#include "onemath_blas_rocblas.hxx" } //namespace column_major namespace row_major { -#include "onemkl_blas_rocblas.hxx" +#include "onemath_blas_rocblas.hxx" } //namespace row_major } //namespace rocblas diff --git a/include/oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hxx b/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hxx similarity index 100% rename from include/oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hxx rename to include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hxx diff --git a/include/oneapi/mkl/detail/backend_selector.hpp b/include/oneapi/math/detail/backend_selector.hpp similarity index 89% rename from include/oneapi/mkl/detail/backend_selector.hpp rename to include/oneapi/math/detail/backend_selector.hpp index b0c763ae0..78797551e 100644 --- a/include/oneapi/mkl/detail/backend_selector.hpp +++ b/include/oneapi/math/detail/backend_selector.hpp @@ -20,9 +20,9 @@ #ifndef _ONEMKL_BACKEND_SELECTOR_HPP_ #define _ONEMKL_BACKEND_SELECTOR_HPP_ -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/detail/backends.hpp" -#include "oneapi/mkl/detail/backend_selector_predicates.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/detail/backends.hpp" +#include "oneapi/math/detail/backend_selector_predicates.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/detail/backend_selector_predicates.hpp b/include/oneapi/math/detail/backend_selector_predicates.hpp similarity index 97% rename from include/oneapi/mkl/detail/backend_selector_predicates.hpp rename to include/oneapi/math/detail/backend_selector_predicates.hpp index 4ee3f3bb1..e3966cf29 100644 --- a/include/oneapi/mkl/detail/backend_selector_predicates.hpp +++ b/include/oneapi/math/detail/backend_selector_predicates.hpp @@ -27,9 +27,9 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/detail/backends.hpp" -#include "oneapi/mkl/detail/get_device_id.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/detail/backends.hpp" +#include "oneapi/math/detail/get_device_id.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/detail/backends.hpp b/include/oneapi/math/detail/backends.hpp similarity index 100% rename from include/oneapi/mkl/detail/backends.hpp rename to include/oneapi/math/detail/backends.hpp diff --git a/include/oneapi/mkl/detail/backends_table.hpp b/include/oneapi/math/detail/backends_table.hpp similarity index 99% rename from include/oneapi/mkl/detail/backends_table.hpp rename to include/oneapi/math/detail/backends_table.hpp index 731781375..bc9a1e78a 100644 --- a/include/oneapi/mkl/detail/backends_table.hpp +++ b/include/oneapi/math/detail/backends_table.hpp @@ -29,7 +29,7 @@ #include #endif -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math/detail/config.hpp" #ifdef __linux__ #define LIB_NAME(a) "libonemkl_" a ".so" diff --git a/include/oneapi/mkl/detail/exceptions.hpp b/include/oneapi/math/detail/exceptions.hpp similarity index 98% rename from include/oneapi/mkl/detail/exceptions.hpp rename to include/oneapi/math/detail/exceptions.hpp index 7767c2ac3..0b76d9d7e 100644 --- a/include/oneapi/mkl/detail/exceptions.hpp +++ b/include/oneapi/math/detail/exceptions.hpp @@ -24,7 +24,7 @@ #include #include -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/detail/export.hpp b/include/oneapi/math/detail/export.hpp similarity index 96% rename from include/oneapi/mkl/detail/export.hpp rename to include/oneapi/math/detail/export.hpp index 55f6a7b15..ca73e949b 100644 --- a/include/oneapi/mkl/detail/export.hpp +++ b/include/oneapi/math/detail/export.hpp @@ -20,7 +20,7 @@ #ifndef ONEMKL_EXPORT_H #define ONEMKL_EXPORT_H -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math/detail/config.hpp" #if !defined(ONEMKL_BUILD_SHARED_LIBS) || !defined(_WIN64) #define ONEMKL_EXPORT diff --git a/include/oneapi/mkl/detail/get_device_id.hpp b/include/oneapi/math/detail/get_device_id.hpp similarity index 95% rename from include/oneapi/mkl/detail/get_device_id.hpp rename to include/oneapi/math/detail/get_device_id.hpp index 32c30f1cb..dac0bf08f 100644 --- a/include/oneapi/mkl/detail/get_device_id.hpp +++ b/include/oneapi/math/detail/get_device_id.hpp @@ -26,8 +26,8 @@ #include #endif -#include "oneapi/mkl/detail/backends_table.hpp" -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/detail/backends_table.hpp" +#include "oneapi/math/exceptions.hpp" #define INTEL_ID 32902 #define NVIDIA_ID 4318 diff --git a/include/oneapi/mkl/dft.hpp b/include/oneapi/math/dft.hpp similarity index 79% rename from include/oneapi/mkl/dft.hpp rename to include/oneapi/math/dft.hpp index 17ee4e042..2041b296c 100644 --- a/include/oneapi/mkl/dft.hpp +++ b/include/oneapi/math/dft.hpp @@ -28,12 +28,12 @@ #include #include -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl/detail/get_device_id.hpp" -#include "oneapi/mkl/dft/detail/dft_loader.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math/detail/get_device_id.hpp" +#include "oneapi/math/dft/detail/dft_loader.hpp" -#include "oneapi/mkl/dft/descriptor.hpp" -#include "oneapi/mkl/dft/forward.hpp" -#include "oneapi/mkl/dft/backward.hpp" +#include "oneapi/math/dft/descriptor.hpp" +#include "oneapi/math/dft/forward.hpp" +#include "oneapi/math/dft/backward.hpp" #endif // _ONEMKL_DFT_HPP_ diff --git a/include/oneapi/mkl/dft/backward.hpp b/include/oneapi/math/dft/backward.hpp similarity index 100% rename from include/oneapi/mkl/dft/backward.hpp rename to include/oneapi/math/dft/backward.hpp diff --git a/include/oneapi/mkl/dft/descriptor.hpp b/include/oneapi/math/dft/descriptor.hpp similarity index 100% rename from include/oneapi/mkl/dft/descriptor.hpp rename to include/oneapi/math/dft/descriptor.hpp diff --git a/include/oneapi/mkl/dft/detail/commit_impl.hpp b/include/oneapi/math/dft/detail/commit_impl.hpp similarity index 100% rename from include/oneapi/mkl/dft/detail/commit_impl.hpp rename to include/oneapi/math/dft/detail/commit_impl.hpp diff --git a/include/oneapi/mkl/dft/detail/cufft/onemkl_dft_cufft.hpp b/include/oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp similarity index 88% rename from include/oneapi/mkl/dft/detail/cufft/onemkl_dft_cufft.hpp rename to include/oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp index 4e4ad2030..3d4470927 100644 --- a/include/oneapi/mkl/dft/detail/cufft/onemkl_dft_cufft.hpp +++ b/include/oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp @@ -26,12 +26,12 @@ #include #endif -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" namespace oneapi::mkl::dft::cufft { -#include "oneapi/mkl/dft/detail/dft_ct.hxx" +#include "oneapi/math/dft/detail/dft_ct.hxx" } // namespace oneapi::mkl::dft::cufft diff --git a/include/oneapi/mkl/dft/detail/descriptor_impl.hpp b/include/oneapi/math/dft/detail/descriptor_impl.hpp similarity index 94% rename from include/oneapi/mkl/dft/detail/descriptor_impl.hpp rename to include/oneapi/math/dft/detail/descriptor_impl.hpp index b2095c49b..36900fbc6 100644 --- a/include/oneapi/mkl/dft/detail/descriptor_impl.hpp +++ b/include/oneapi/math/dft/detail/descriptor_impl.hpp @@ -26,11 +26,11 @@ #include #endif -#include "oneapi/mkl/detail/backend_selector.hpp" -#include "oneapi/mkl/detail/export.hpp" +#include "oneapi/math/detail/backend_selector.hpp" +#include "oneapi/math/detail/export.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" -#include "oneapi/mkl/dft/detail/commit_impl.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/dft/detail/dft_ct.hxx b/include/oneapi/math/dft/detail/dft_ct.hxx similarity index 100% rename from include/oneapi/mkl/dft/detail/dft_ct.hxx rename to include/oneapi/math/dft/detail/dft_ct.hxx diff --git a/include/oneapi/mkl/dft/detail/dft_loader.hpp b/include/oneapi/math/dft/detail/dft_loader.hpp similarity index 94% rename from include/oneapi/mkl/dft/detail/dft_loader.hpp rename to include/oneapi/math/dft/detail/dft_loader.hpp index f84a4e01c..173ced31a 100644 --- a/include/oneapi/mkl/dft/detail/dft_loader.hpp +++ b/include/oneapi/math/dft/detail/dft_loader.hpp @@ -26,8 +26,8 @@ #include #endif -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/dft/detail/external_workspace_helper.hpp b/include/oneapi/math/dft/detail/external_workspace_helper.hpp similarity index 98% rename from include/oneapi/mkl/dft/detail/external_workspace_helper.hpp rename to include/oneapi/math/dft/detail/external_workspace_helper.hpp index b41dffc4c..8938b70e1 100644 --- a/include/oneapi/mkl/dft/detail/external_workspace_helper.hpp +++ b/include/oneapi/math/dft/detail/external_workspace_helper.hpp @@ -26,9 +26,9 @@ #include #endif -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" -#include "oneapi/mkl/dft/detail/commit_impl.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp b/include/oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp similarity index 88% rename from include/oneapi/mkl/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp rename to include/oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp index 00d4dd47b..063d7d976 100644 --- a/include/oneapi/mkl/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp +++ b/include/oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp @@ -26,12 +26,12 @@ #include #endif -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" namespace oneapi::mkl::dft::mklcpu { -#include "oneapi/mkl/dft/detail/dft_ct.hxx" +#include "oneapi/math/dft/detail/dft_ct.hxx" } // namespace oneapi::mkl::dft::mklcpu diff --git a/include/oneapi/mkl/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp b/include/oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp similarity index 88% rename from include/oneapi/mkl/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp rename to include/oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp index 56a55a9f7..7ea9487cd 100644 --- a/include/oneapi/mkl/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp +++ b/include/oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp @@ -26,12 +26,12 @@ #include #endif -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" namespace oneapi::mkl::dft::mklgpu { -#include "oneapi/mkl/dft/detail/dft_ct.hxx" +#include "oneapi/math/dft/detail/dft_ct.hxx" } // namespace oneapi::mkl::dft::mklgpu diff --git a/include/oneapi/mkl/dft/detail/portfft/onemkl_dft_portfft.hpp b/include/oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp similarity index 90% rename from include/oneapi/mkl/dft/detail/portfft/onemkl_dft_portfft.hpp rename to include/oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp index 4617e8a5c..4ef12e4bf 100644 --- a/include/oneapi/mkl/dft/detail/portfft/onemkl_dft_portfft.hpp +++ b/include/oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp @@ -26,13 +26,13 @@ #include #endif -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" namespace oneapi::mkl::dft::portfft { // We don't need the forward declarations of compute_xxxward templates (just need the create_commit template), but it doesn't hurt and keeps things simple. -#include "oneapi/mkl/dft/detail/dft_ct.hxx" +#include "oneapi/math/dft/detail/dft_ct.hxx" } // namespace oneapi::mkl::dft::portfft diff --git a/include/oneapi/mkl/dft/detail/rocfft/onemkl_dft_rocfft.hpp b/include/oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp similarity index 88% rename from include/oneapi/mkl/dft/detail/rocfft/onemkl_dft_rocfft.hpp rename to include/oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp index fe3305680..a7712858a 100644 --- a/include/oneapi/mkl/dft/detail/rocfft/onemkl_dft_rocfft.hpp +++ b/include/oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp @@ -26,12 +26,12 @@ #include #endif -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" namespace oneapi::mkl::dft::rocfft { -#include "oneapi/mkl/dft/detail/dft_ct.hxx" +#include "oneapi/math/dft/detail/dft_ct.hxx" } // namespace oneapi::mkl::dft::rocfft diff --git a/include/oneapi/mkl/dft/detail/types_impl.hpp b/include/oneapi/math/dft/detail/types_impl.hpp similarity index 100% rename from include/oneapi/mkl/dft/detail/types_impl.hpp rename to include/oneapi/math/dft/detail/types_impl.hpp diff --git a/include/oneapi/mkl/dft/forward.hpp b/include/oneapi/math/dft/forward.hpp similarity index 100% rename from include/oneapi/mkl/dft/forward.hpp rename to include/oneapi/math/dft/forward.hpp diff --git a/include/oneapi/mkl/dft/types.hpp b/include/oneapi/math/dft/types.hpp similarity index 100% rename from include/oneapi/mkl/dft/types.hpp rename to include/oneapi/math/dft/types.hpp diff --git a/include/oneapi/mkl/exceptions.hpp b/include/oneapi/math/exceptions.hpp similarity index 99% rename from include/oneapi/mkl/exceptions.hpp rename to include/oneapi/math/exceptions.hpp index 244c8c61d..7d3064237 100644 --- a/include/oneapi/mkl/exceptions.hpp +++ b/include/oneapi/math/exceptions.hpp @@ -28,7 +28,7 @@ #include #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" // These are oneAPI oneMKL Specification exceptions diff --git a/include/oneapi/mkl/lapack.hpp b/include/oneapi/math/lapack.hpp similarity index 74% rename from include/oneapi/mkl/lapack.hpp rename to include/oneapi/math/lapack.hpp index f9d331e42..dda570563 100644 --- a/include/oneapi/mkl/lapack.hpp +++ b/include/oneapi/math/lapack.hpp @@ -19,19 +19,19 @@ #pragma once -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math/detail/config.hpp" #ifdef ONEMKL_ENABLE_MKLCPU_BACKEND -#include "oneapi/mkl/lapack/detail/mklcpu/lapack_ct.hpp" +#include "oneapi/math/lapack/detail/mklcpu/lapack_ct.hpp" #endif #ifdef ONEMKL_ENABLE_MKLGPU_BACKEND -#include "oneapi/mkl/lapack/detail/mklgpu/lapack_ct.hpp" +#include "oneapi/math/lapack/detail/mklgpu/lapack_ct.hpp" #endif #ifdef ONEMKL_ENABLE_CUSOLVER_BACKEND -#include "oneapi/mkl/lapack/detail/cusolver/lapack_ct.hpp" +#include "oneapi/math/lapack/detail/cusolver/lapack_ct.hpp" #endif #ifdef ONEMKL_ENABLE_ROCSOLVER_BACKEND -#include "oneapi/mkl/lapack/detail/rocsolver/lapack_ct.hpp" +#include "oneapi/math/lapack/detail/rocsolver/lapack_ct.hpp" #endif -#include "oneapi/mkl/lapack/detail/lapack_rt.hpp" +#include "oneapi/math/lapack/detail/lapack_rt.hpp" diff --git a/include/oneapi/mkl/lapack/detail/cusolver/lapack_ct.hpp b/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hpp similarity index 86% rename from include/oneapi/mkl/lapack/detail/cusolver/lapack_ct.hpp rename to include/oneapi/math/lapack/detail/cusolver/lapack_ct.hpp index 1be0e5895..b25ec1c9e 100644 --- a/include/oneapi/mkl/lapack/detail/cusolver/lapack_ct.hpp +++ b/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hpp @@ -28,10 +28,10 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" -#include "oneapi/mkl/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/lapack/types.hpp" +#include "oneapi/math/detail/backend_selector.hpp" +#include "oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/lapack/detail/cusolver/lapack_ct.hxx b/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hxx similarity index 100% rename from include/oneapi/mkl/lapack/detail/cusolver/lapack_ct.hxx rename to include/oneapi/math/lapack/detail/cusolver/lapack_ct.hxx diff --git a/include/oneapi/mkl/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp similarity index 91% rename from include/oneapi/mkl/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp rename to include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp index 6d31a05af..3d363cb66 100644 --- a/include/oneapi/mkl/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp +++ b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp @@ -27,15 +27,15 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/export.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/export.hpp" namespace oneapi { namespace mkl { namespace lapack { namespace cusolver { -#include "onemkl_lapack_cusolver.hxx" +#include "onemath_lapack_cusolver.hxx" } // namespace cusolver } // namespace lapack diff --git a/include/oneapi/mkl/lapack/detail/cusolver/onemkl_lapack_cusolver.hxx b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx similarity index 100% rename from include/oneapi/mkl/lapack/detail/cusolver/onemkl_lapack_cusolver.hxx rename to include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx diff --git a/include/oneapi/mkl/lapack/detail/lapack_loader.hpp b/include/oneapi/math/lapack/detail/lapack_loader.hpp similarity index 99% rename from include/oneapi/mkl/lapack/detail/lapack_loader.hpp rename to include/oneapi/math/lapack/detail/lapack_loader.hpp index 2bb49364e..11a9331d4 100644 --- a/include/oneapi/mkl/lapack/detail/lapack_loader.hpp +++ b/include/oneapi/math/lapack/detail/lapack_loader.hpp @@ -28,10 +28,10 @@ #include #endif -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/detail/get_device_id.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/lapack/types.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/detail/get_device_id.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/lapack/detail/lapack_rt.hpp b/include/oneapi/math/lapack/detail/lapack_rt.hpp similarity index 99% rename from include/oneapi/mkl/lapack/detail/lapack_rt.hpp rename to include/oneapi/math/lapack/detail/lapack_rt.hpp index a96efe8d1..ef03694fd 100644 --- a/include/oneapi/mkl/lapack/detail/lapack_rt.hpp +++ b/include/oneapi/math/lapack/detail/lapack_rt.hpp @@ -28,11 +28,11 @@ #include #endif -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" -#include "oneapi/mkl/lapack/exceptions.hpp" -#include "oneapi/mkl/detail/get_device_id.hpp" -#include "oneapi/mkl/lapack/detail/lapack_loader.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/lapack/types.hpp" +#include "oneapi/math/lapack/exceptions.hpp" +#include "oneapi/math/detail/get_device_id.hpp" +#include "oneapi/math/lapack/detail/lapack_loader.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/lapack/detail/mkl_common/lapack_ct.hxx b/include/oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx similarity index 100% rename from include/oneapi/mkl/lapack/detail/mkl_common/lapack_ct.hxx rename to include/oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx diff --git a/include/oneapi/mkl/lapack/detail/mkl_common/onemkl_lapack_backends.hxx b/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx similarity index 100% rename from include/oneapi/mkl/lapack/detail/mkl_common/onemkl_lapack_backends.hxx rename to include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx diff --git a/include/oneapi/mkl/lapack/detail/mklcpu/lapack_ct.hpp b/include/oneapi/math/lapack/detail/mklcpu/lapack_ct.hpp similarity index 81% rename from include/oneapi/mkl/lapack/detail/mklcpu/lapack_ct.hpp rename to include/oneapi/math/lapack/detail/mklcpu/lapack_ct.hpp index 1a6c088d6..580ea72d1 100644 --- a/include/oneapi/mkl/lapack/detail/mklcpu/lapack_ct.hpp +++ b/include/oneapi/math/lapack/detail/mklcpu/lapack_ct.hpp @@ -28,17 +28,17 @@ #include #endif -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" -#include "oneapi/mkl/lapack/detail/mklcpu/onemkl_lapack_mklcpu.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/lapack/types.hpp" +#include "oneapi/math/detail/backend_selector.hpp" +#include "oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp" namespace oneapi { namespace mkl { namespace lapack { #define LAPACK_BACKEND mklcpu -#include "oneapi/mkl/lapack/detail/mkl_common/lapack_ct.hxx" +#include "oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx" #undef LAPACK_BACKEND } //namespace lapack diff --git a/include/oneapi/mkl/lapack/detail/mklcpu/onemkl_lapack_mklcpu.hpp b/include/oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp similarity index 85% rename from include/oneapi/mkl/lapack/detail/mklcpu/onemkl_lapack_mklcpu.hpp rename to include/oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp index fc52ce1db..a90c8eefe 100644 --- a/include/oneapi/mkl/lapack/detail/mklcpu/onemkl_lapack_mklcpu.hpp +++ b/include/oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp @@ -28,16 +28,16 @@ #include #endif -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" -#include "oneapi/mkl/detail/export.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/lapack/types.hpp" +#include "oneapi/math/detail/export.hpp" namespace oneapi { namespace mkl { namespace lapack { namespace mklcpu { -#include "oneapi/mkl/lapack/detail/mkl_common/onemkl_lapack_backends.hxx" +#include "oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx" } //namespace mklcpu } //namespace lapack diff --git a/include/oneapi/mkl/lapack/detail/mklgpu/lapack_ct.hpp b/include/oneapi/math/lapack/detail/mklgpu/lapack_ct.hpp similarity index 81% rename from include/oneapi/mkl/lapack/detail/mklgpu/lapack_ct.hpp rename to include/oneapi/math/lapack/detail/mklgpu/lapack_ct.hpp index e344966a0..23dc557e1 100644 --- a/include/oneapi/mkl/lapack/detail/mklgpu/lapack_ct.hpp +++ b/include/oneapi/math/lapack/detail/mklgpu/lapack_ct.hpp @@ -27,17 +27,17 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" -#include "oneapi/mkl/lapack/detail/mklgpu/onemkl_lapack_mklgpu.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/lapack/types.hpp" +#include "oneapi/math/detail/backend_selector.hpp" +#include "oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp" namespace oneapi { namespace mkl { namespace lapack { #define LAPACK_BACKEND mklgpu -#include "oneapi/mkl/lapack/detail/mkl_common/lapack_ct.hxx" +#include "oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx" #undef LAPACK_BACKEND } //namespace lapack diff --git a/include/oneapi/mkl/lapack/detail/mklgpu/onemkl_lapack_mklgpu.hpp b/include/oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp similarity index 85% rename from include/oneapi/mkl/lapack/detail/mklgpu/onemkl_lapack_mklgpu.hpp rename to include/oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp index 132431b7c..e0fda1d97 100644 --- a/include/oneapi/mkl/lapack/detail/mklgpu/onemkl_lapack_mklgpu.hpp +++ b/include/oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp @@ -28,16 +28,16 @@ #include #endif -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" -#include "oneapi/mkl/detail/export.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/lapack/types.hpp" +#include "oneapi/math/detail/export.hpp" namespace oneapi { namespace mkl { namespace lapack { namespace mklgpu { -#include "oneapi/mkl/lapack/detail/mkl_common/onemkl_lapack_backends.hxx" +#include "oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx" } //namespace mklgpu } //namespace lapack diff --git a/include/oneapi/mkl/lapack/detail/rocsolver/lapack_ct.hpp b/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hpp similarity index 87% rename from include/oneapi/mkl/lapack/detail/rocsolver/lapack_ct.hpp rename to include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hpp index 5e98b7c47..888863167 100644 --- a/include/oneapi/mkl/lapack/detail/rocsolver/lapack_ct.hpp +++ b/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hpp @@ -30,10 +30,10 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" -#include "oneapi/mkl/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/lapack/types.hpp" +#include "oneapi/math/detail/backend_selector.hpp" +#include "oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/lapack/detail/rocsolver/lapack_ct.hxx b/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hxx similarity index 100% rename from include/oneapi/mkl/lapack/detail/rocsolver/lapack_ct.hxx rename to include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hxx diff --git a/include/oneapi/mkl/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp similarity index 92% rename from include/oneapi/mkl/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp rename to include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp index f7e83f9a9..552235e3e 100644 --- a/include/oneapi/mkl/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp +++ b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp @@ -31,15 +31,15 @@ #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/export.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/export.hpp" namespace oneapi { namespace mkl { namespace lapack { namespace rocsolver { -#include "onemkl_lapack_rocsolver.hxx" +#include "onemath_lapack_rocsolver.hxx" } // namespace rocsolver } // namespace lapack diff --git a/include/oneapi/mkl/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hxx b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx similarity index 100% rename from include/oneapi/mkl/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hxx rename to include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx diff --git a/include/oneapi/mkl/lapack/exceptions.hpp b/include/oneapi/math/lapack/exceptions.hpp similarity index 100% rename from include/oneapi/mkl/lapack/exceptions.hpp rename to include/oneapi/math/lapack/exceptions.hpp diff --git a/include/oneapi/mkl/lapack/types.hpp b/include/oneapi/math/lapack/types.hpp similarity index 100% rename from include/oneapi/mkl/lapack/types.hpp rename to include/oneapi/math/lapack/types.hpp diff --git a/include/oneapi/mkl/rng.hpp b/include/oneapi/math/rng.hpp similarity index 76% rename from include/oneapi/mkl/rng.hpp rename to include/oneapi/math/rng.hpp index 5dfbdac81..2c741660c 100644 --- a/include/oneapi/mkl/rng.hpp +++ b/include/oneapi/math/rng.hpp @@ -28,14 +28,14 @@ #include #include -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl/detail/get_device_id.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math/detail/get_device_id.hpp" -#include "oneapi/mkl/rng/predicates.hpp" -#include "oneapi/mkl/rng/detail/rng_loader.hpp" +#include "oneapi/math/rng/predicates.hpp" +#include "oneapi/math/rng/detail/rng_loader.hpp" -#include "oneapi/mkl/rng/functions.hpp" -#include "oneapi/mkl/rng/distributions.hpp" -#include "oneapi/mkl/rng/engines.hpp" +#include "oneapi/math/rng/functions.hpp" +#include "oneapi/math/rng/distributions.hpp" +#include "oneapi/math/rng/engines.hpp" #endif // _ONEMKL_RNG_HPP_ diff --git a/include/oneapi/mkl/rng/detail/curand/onemkl_rng_curand.hpp b/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp similarity index 97% rename from include/oneapi/mkl/rng/detail/curand/onemkl_rng_curand.hpp rename to include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp index 062d21b61..156da53ad 100644 --- a/include/oneapi/mkl/rng/detail/curand/onemkl_rng_curand.hpp +++ b/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp @@ -66,8 +66,8 @@ #endif #include -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/rng/detail/engine_impl.hpp b/include/oneapi/math/rng/detail/engine_impl.hpp similarity index 98% rename from include/oneapi/mkl/rng/detail/engine_impl.hpp rename to include/oneapi/math/rng/detail/engine_impl.hpp index e76181e4c..0cdb05952 100644 --- a/include/oneapi/mkl/rng/detail/engine_impl.hpp +++ b/include/oneapi/math/rng/detail/engine_impl.hpp @@ -27,11 +27,11 @@ #include #endif -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/detail/get_device_id.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/detail/get_device_id.hpp" -#include "oneapi/mkl/rng/distributions.hpp" -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/rng/distributions.hpp" +#include "oneapi/math/types.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp b/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp similarity index 95% rename from include/oneapi/mkl/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp rename to include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp index e13b70148..4a46cc84a 100644 --- a/include/oneapi/mkl/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp +++ b/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp @@ -27,8 +27,8 @@ #include #endif -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp b/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp similarity index 95% rename from include/oneapi/mkl/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp rename to include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp index 4dd55f19b..83bfc7488 100644 --- a/include/oneapi/mkl/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp +++ b/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp @@ -27,8 +27,8 @@ #include #endif -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/rng/detail/rng_loader.hpp b/include/oneapi/math/rng/detail/rng_loader.hpp similarity index 92% rename from include/oneapi/mkl/rng/detail/rng_loader.hpp rename to include/oneapi/math/rng/detail/rng_loader.hpp index dc85df5d6..d03ea8ea9 100644 --- a/include/oneapi/mkl/rng/detail/rng_loader.hpp +++ b/include/oneapi/math/rng/detail/rng_loader.hpp @@ -27,10 +27,10 @@ #include #endif -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/detail/get_device_id.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/detail/get_device_id.hpp" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/rng/detail/rocrand/onemkl_rng_rocrand.hpp b/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp similarity index 98% rename from include/oneapi/mkl/rng/detail/rocrand/onemkl_rng_rocrand.hpp rename to include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp index 791bcc13b..05169e226 100644 --- a/include/oneapi/mkl/rng/detail/rocrand/onemkl_rng_rocrand.hpp +++ b/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp @@ -68,8 +68,8 @@ #endif #include -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/rng/device.hpp b/include/oneapi/math/rng/device.hpp similarity index 82% rename from include/oneapi/mkl/rng/device.hpp rename to include/oneapi/math/rng/device.hpp index a628395d2..88caa9fb9 100644 --- a/include/oneapi/mkl/rng/device.hpp +++ b/include/oneapi/math/rng/device.hpp @@ -20,9 +20,9 @@ #ifndef _MKL_RNG_SYCL_DEVICE_HPP__ #define _MKL_RNG_SYCL_DEVICE_HPP__ -#include "oneapi/mkl/rng/device/types.hpp" -#include "oneapi/mkl/rng/device/functions.hpp" -#include "oneapi/mkl/rng/device/distributions.hpp" -#include "oneapi/mkl/rng/device/engines.hpp" +#include "oneapi/math/rng/device/types.hpp" +#include "oneapi/math/rng/device/functions.hpp" +#include "oneapi/math/rng/device/distributions.hpp" +#include "oneapi/math/rng/device/engines.hpp" #endif // _MKL_RNG_SYCL_DEVICE_HPP__ diff --git a/include/oneapi/mkl/rng/device/detail/bernoulli_impl.hpp b/include/oneapi/math/rng/device/detail/bernoulli_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/bernoulli_impl.hpp rename to include/oneapi/math/rng/device/detail/bernoulli_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/beta_impl.hpp b/include/oneapi/math/rng/device/detail/beta_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/beta_impl.hpp rename to include/oneapi/math/rng/device/detail/beta_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/bits_impl.hpp b/include/oneapi/math/rng/device/detail/bits_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/bits_impl.hpp rename to include/oneapi/math/rng/device/detail/bits_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/distribution_base.hpp b/include/oneapi/math/rng/device/detail/distribution_base.hpp similarity index 75% rename from include/oneapi/mkl/rng/device/detail/distribution_base.hpp rename to include/oneapi/math/rng/device/detail/distribution_base.hpp index 575ea27f7..643d2fa6b 100644 --- a/include/oneapi/mkl/rng/device/detail/distribution_base.hpp +++ b/include/oneapi/math/rng/device/detail/distribution_base.hpp @@ -22,8 +22,8 @@ #include -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/rng/device/types.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/rng/device/types.hpp" namespace oneapi::mkl::rng::device { @@ -67,15 +67,15 @@ class bernoulli; } // namespace oneapi::mkl::rng::device -#include "oneapi/mkl/rng/device/detail/uniform_impl.hpp" -#include "oneapi/mkl/rng/device/detail/gaussian_impl.hpp" -#include "oneapi/mkl/rng/device/detail/lognormal_impl.hpp" -#include "oneapi/mkl/rng/device/detail/bits_impl.hpp" -#include "oneapi/mkl/rng/device/detail/uniform_bits_impl.hpp" -#include "oneapi/mkl/rng/device/detail/exponential_impl.hpp" -#include "oneapi/mkl/rng/device/detail/poisson_impl.hpp" -#include "oneapi/mkl/rng/device/detail/bernoulli_impl.hpp" -#include "oneapi/mkl/rng/device/detail/beta_impl.hpp" -#include "oneapi/mkl/rng/device/detail/gamma_impl.hpp" +#include "oneapi/math/rng/device/detail/uniform_impl.hpp" +#include "oneapi/math/rng/device/detail/gaussian_impl.hpp" +#include "oneapi/math/rng/device/detail/lognormal_impl.hpp" +#include "oneapi/math/rng/device/detail/bits_impl.hpp" +#include "oneapi/math/rng/device/detail/uniform_bits_impl.hpp" +#include "oneapi/math/rng/device/detail/exponential_impl.hpp" +#include "oneapi/math/rng/device/detail/poisson_impl.hpp" +#include "oneapi/math/rng/device/detail/bernoulli_impl.hpp" +#include "oneapi/math/rng/device/detail/beta_impl.hpp" +#include "oneapi/math/rng/device/detail/gamma_impl.hpp" #endif // _MKL_RNG_DISTRIBUTION_BASE_HPP_ diff --git a/include/oneapi/mkl/rng/device/detail/engine_base.hpp b/include/oneapi/math/rng/device/detail/engine_base.hpp similarity index 83% rename from include/oneapi/mkl/rng/device/detail/engine_base.hpp rename to include/oneapi/math/rng/device/detail/engine_base.hpp index fc1aee16a..6c754dcd4 100644 --- a/include/oneapi/mkl/rng/device/detail/engine_base.hpp +++ b/include/oneapi/math/rng/device/detail/engine_base.hpp @@ -35,9 +35,9 @@ class engine_base {}; } // namespace oneapi::mkl::rng::device::detail -#include "oneapi/mkl/rng/device/detail/philox4x32x10_impl.hpp" -#include "oneapi/mkl/rng/device/detail/mrg32k3a_impl.hpp" -#include "oneapi/mkl/rng/device/detail/mcg31m1_impl.hpp" -#include "oneapi/mkl/rng/device/detail/mcg59_impl.hpp" +#include "oneapi/math/rng/device/detail/philox4x32x10_impl.hpp" +#include "oneapi/math/rng/device/detail/mrg32k3a_impl.hpp" +#include "oneapi/math/rng/device/detail/mcg31m1_impl.hpp" +#include "oneapi/math/rng/device/detail/mcg59_impl.hpp" #endif // _MKL_RNG_DEVICE_ENGINE_BASE_HPP_ diff --git a/include/oneapi/mkl/rng/device/detail/exponential_impl.hpp b/include/oneapi/math/rng/device/detail/exponential_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/exponential_impl.hpp rename to include/oneapi/math/rng/device/detail/exponential_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/gamma_impl.hpp b/include/oneapi/math/rng/device/detail/gamma_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/gamma_impl.hpp rename to include/oneapi/math/rng/device/detail/gamma_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/gaussian_impl.hpp b/include/oneapi/math/rng/device/detail/gaussian_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/gaussian_impl.hpp rename to include/oneapi/math/rng/device/detail/gaussian_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/lognormal_impl.hpp b/include/oneapi/math/rng/device/detail/lognormal_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/lognormal_impl.hpp rename to include/oneapi/math/rng/device/detail/lognormal_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/mcg31m1_impl.hpp b/include/oneapi/math/rng/device/detail/mcg31m1_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/mcg31m1_impl.hpp rename to include/oneapi/math/rng/device/detail/mcg31m1_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/mcg59_impl.hpp b/include/oneapi/math/rng/device/detail/mcg59_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/mcg59_impl.hpp rename to include/oneapi/math/rng/device/detail/mcg59_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/mrg32k3a_impl.hpp b/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp similarity index 99% rename from include/oneapi/mkl/rng/device/detail/mrg32k3a_impl.hpp rename to include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp index 596e625ad..f285c7683 100644 --- a/include/oneapi/mkl/rng/device/detail/mrg32k3a_impl.hpp +++ b/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp @@ -25,7 +25,7 @@ #ifndef _MKL_RNG_DEVICE_MRG32K3A_IMPL_HPP_ #define _MKL_RNG_DEVICE_MRG32K3A_IMPL_HPP_ -#include "oneapi/mkl/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp" +#include "oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp" namespace oneapi::mkl::rng::device { diff --git a/include/oneapi/mkl/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp b/include/oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp rename to include/oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp diff --git a/include/oneapi/mkl/rng/device/detail/philox4x32x10_impl.hpp b/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/philox4x32x10_impl.hpp rename to include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/poisson_impl.hpp b/include/oneapi/math/rng/device/detail/poisson_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/poisson_impl.hpp rename to include/oneapi/math/rng/device/detail/poisson_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/uniform_bits_impl.hpp b/include/oneapi/math/rng/device/detail/uniform_bits_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/uniform_bits_impl.hpp rename to include/oneapi/math/rng/device/detail/uniform_bits_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/uniform_impl.hpp b/include/oneapi/math/rng/device/detail/uniform_impl.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/uniform_impl.hpp rename to include/oneapi/math/rng/device/detail/uniform_impl.hpp diff --git a/include/oneapi/mkl/rng/device/detail/vm_wrappers.hpp b/include/oneapi/math/rng/device/detail/vm_wrappers.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/detail/vm_wrappers.hpp rename to include/oneapi/math/rng/device/detail/vm_wrappers.hpp diff --git a/include/oneapi/mkl/rng/device/distributions.hpp b/include/oneapi/math/rng/device/distributions.hpp similarity index 99% rename from include/oneapi/mkl/rng/device/distributions.hpp rename to include/oneapi/math/rng/device/distributions.hpp index 121e81aa3..28bb439f1 100644 --- a/include/oneapi/mkl/rng/device/distributions.hpp +++ b/include/oneapi/math/rng/device/distributions.hpp @@ -22,8 +22,8 @@ #include -#include "oneapi/mkl/rng/device/detail/distribution_base.hpp" -#include "oneapi/mkl/rng/device/functions.hpp" +#include "oneapi/math/rng/device/detail/distribution_base.hpp" +#include "oneapi/math/rng/device/functions.hpp" namespace oneapi::mkl::rng::device { diff --git a/include/oneapi/mkl/rng/device/engines.hpp b/include/oneapi/math/rng/device/engines.hpp similarity index 97% rename from include/oneapi/mkl/rng/device/engines.hpp rename to include/oneapi/math/rng/device/engines.hpp index f1bcfd1b0..3eda19793 100644 --- a/include/oneapi/mkl/rng/device/engines.hpp +++ b/include/oneapi/math/rng/device/engines.hpp @@ -22,9 +22,9 @@ #include -#include "oneapi/mkl/rng/device/types.hpp" -#include "oneapi/mkl/rng/device/functions.hpp" -#include "oneapi/mkl/rng/device/detail/engine_base.hpp" +#include "oneapi/math/rng/device/types.hpp" +#include "oneapi/math/rng/device/functions.hpp" +#include "oneapi/math/rng/device/detail/engine_base.hpp" namespace oneapi::mkl::rng::device { diff --git a/include/oneapi/mkl/rng/device/functions.hpp b/include/oneapi/math/rng/device/functions.hpp similarity index 96% rename from include/oneapi/mkl/rng/device/functions.hpp rename to include/oneapi/math/rng/device/functions.hpp index d8542b836..89f322740 100644 --- a/include/oneapi/mkl/rng/device/functions.hpp +++ b/include/oneapi/math/rng/device/functions.hpp @@ -22,7 +22,7 @@ #include -#include "oneapi/mkl/rng/device/detail/distribution_base.hpp" +#include "oneapi/math/rng/device/detail/distribution_base.hpp" namespace oneapi::mkl::rng::device { diff --git a/include/oneapi/mkl/rng/device/types.hpp b/include/oneapi/math/rng/device/types.hpp similarity index 100% rename from include/oneapi/mkl/rng/device/types.hpp rename to include/oneapi/math/rng/device/types.hpp diff --git a/include/oneapi/mkl/rng/distributions.hpp b/include/oneapi/math/rng/distributions.hpp similarity index 99% rename from include/oneapi/mkl/rng/distributions.hpp rename to include/oneapi/math/rng/distributions.hpp index 88d1e46e7..58870c55a 100644 --- a/include/oneapi/mkl/rng/distributions.hpp +++ b/include/oneapi/math/rng/distributions.hpp @@ -28,7 +28,7 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/rng/engines.hpp b/include/oneapi/math/rng/engines.hpp similarity index 95% rename from include/oneapi/mkl/rng/engines.hpp rename to include/oneapi/math/rng/engines.hpp index 610fda70b..2edccdd7c 100644 --- a/include/oneapi/mkl/rng/engines.hpp +++ b/include/oneapi/math/rng/engines.hpp @@ -29,23 +29,23 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/detail/backend_selector.hpp" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" -#include "oneapi/mkl/rng/detail/rng_loader.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" +#include "oneapi/math/rng/detail/rng_loader.hpp" #ifdef ONEMKL_ENABLE_MKLCPU_BACKEND -#include "oneapi/mkl/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp" +#include "oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp" #endif #ifdef ONEMKL_ENABLE_MKLGPU_BACKEND -#include "oneapi/mkl/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp" +#include "oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp" #endif #ifdef ONEMKL_ENABLE_CURAND_BACKEND -#include "oneapi/mkl/rng/detail/curand/onemkl_rng_curand.hpp" +#include "oneapi/math/rng/detail/curand/onemath_rng_curand.hpp" #endif #ifdef ONEMKL_ENABLE_ROCRAND_BACKEND -#include "oneapi/mkl/rng/detail/rocrand/onemkl_rng_rocrand.hpp" +#include "oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp" #endif namespace oneapi { diff --git a/include/oneapi/mkl/rng/functions.hpp b/include/oneapi/math/rng/functions.hpp similarity index 98% rename from include/oneapi/mkl/rng/functions.hpp rename to include/oneapi/math/rng/functions.hpp index 028e13557..2bd55b180 100644 --- a/include/oneapi/mkl/rng/functions.hpp +++ b/include/oneapi/math/rng/functions.hpp @@ -27,8 +27,8 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/rng/predicates.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/rng/predicates.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/rng/predicates.hpp b/include/oneapi/math/rng/predicates.hpp similarity index 96% rename from include/oneapi/mkl/rng/predicates.hpp rename to include/oneapi/math/rng/predicates.hpp index 10422e543..bdddb5636 100644 --- a/include/oneapi/mkl/rng/predicates.hpp +++ b/include/oneapi/math/rng/predicates.hpp @@ -27,8 +27,8 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/types.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/sparse_blas.hpp b/include/oneapi/math/sparse_blas.hpp similarity index 96% rename from include/oneapi/mkl/sparse_blas.hpp rename to include/oneapi/math/sparse_blas.hpp index 004b79727..0d7308ceb 100644 --- a/include/oneapi/mkl/sparse_blas.hpp +++ b/include/oneapi/math/sparse_blas.hpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math/detail/config.hpp" #ifdef ONEMKL_ENABLE_MKLCPU_BACKEND #include "sparse_blas/detail/mklcpu/sparse_blas_ct.hpp" diff --git a/include/oneapi/mkl/sparse_blas/detail/handles.hpp b/include/oneapi/math/sparse_blas/detail/handles.hpp similarity index 100% rename from include/oneapi/mkl/sparse_blas/detail/handles.hpp rename to include/oneapi/math/sparse_blas/detail/handles.hpp diff --git a/include/oneapi/mkl/sparse_blas/detail/helper_types.hpp b/include/oneapi/math/sparse_blas/detail/helper_types.hpp similarity index 100% rename from include/oneapi/mkl/sparse_blas/detail/helper_types.hpp rename to include/oneapi/math/sparse_blas/detail/helper_types.hpp diff --git a/include/oneapi/mkl/sparse_blas/detail/mklcpu/onemkl_sparse_blas_mklcpu.hpp b/include/oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp similarity index 84% rename from include/oneapi/mkl/sparse_blas/detail/mklcpu/onemkl_sparse_blas_mklcpu.hpp rename to include/oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp index 8686d35bc..e3b035844 100644 --- a/include/oneapi/mkl/sparse_blas/detail/mklcpu/onemkl_sparse_blas_mklcpu.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp @@ -20,15 +20,15 @@ #ifndef _ONEMKL_SPARSE_BLAS_DETAIL_MKLCPU_ONEMKL_SPARSE_BLAS_MKLCPU_HPP_ #define _ONEMKL_SPARSE_BLAS_DETAIL_MKLCPU_ONEMKL_SPARSE_BLAS_MKLCPU_HPP_ -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/sparse_blas/detail/helper_types.hpp" -#include "oneapi/mkl/sparse_blas/types.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/sparse_blas/detail/helper_types.hpp" +#include "oneapi/math/sparse_blas/types.hpp" namespace oneapi::mkl::sparse::mklcpu { namespace detail = oneapi::mkl::sparse::detail; -#include "oneapi/mkl/sparse_blas/detail/onemkl_sparse_blas_backends.hxx" +#include "oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx" } // namespace oneapi::mkl::sparse::mklcpu diff --git a/include/oneapi/mkl/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp b/include/oneapi/math/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp similarity index 85% rename from include/oneapi/mkl/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp rename to include/oneapi/math/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp index ee127c3f8..89734a68a 100644 --- a/include/oneapi/mkl/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp @@ -20,17 +20,17 @@ #ifndef _ONEMKL_SPARSE_BLAS_DETAIL_MKLCPU_SPARSE_BLAS_CT_HPP_ #define _ONEMKL_SPARSE_BLAS_DETAIL_MKLCPU_SPARSE_BLAS_CT_HPP_ -#include "oneapi/mkl/detail/backends.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" +#include "oneapi/math/detail/backends.hpp" +#include "oneapi/math/detail/backend_selector.hpp" -#include "onemkl_sparse_blas_mklcpu.hpp" +#include "onemath_sparse_blas_mklcpu.hpp" namespace oneapi { namespace mkl { namespace sparse { #define BACKEND mklcpu -#include "oneapi/mkl/sparse_blas/detail/sparse_blas_ct.hxx" +#include "oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx" #undef BACKEND } //namespace sparse diff --git a/include/oneapi/mkl/sparse_blas/detail/mklgpu/onemkl_sparse_blas_mklgpu.hpp b/include/oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp similarity index 84% rename from include/oneapi/mkl/sparse_blas/detail/mklgpu/onemkl_sparse_blas_mklgpu.hpp rename to include/oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp index eb3aaa5ff..5e5ab093a 100644 --- a/include/oneapi/mkl/sparse_blas/detail/mklgpu/onemkl_sparse_blas_mklgpu.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp @@ -20,15 +20,15 @@ #ifndef _ONEMKL_SPARSE_BLAS_DETAIL_MKLGPU_ONEMKL_SPARSE_BLAS_MKLGPU_HPP_ #define _ONEMKL_SPARSE_BLAS_DETAIL_MKLGPU_ONEMKL_SPARSE_BLAS_MKLGPU_HPP_ -#include "oneapi/mkl/detail/export.hpp" -#include "oneapi/mkl/sparse_blas/detail/helper_types.hpp" -#include "oneapi/mkl/sparse_blas/types.hpp" +#include "oneapi/math/detail/export.hpp" +#include "oneapi/math/sparse_blas/detail/helper_types.hpp" +#include "oneapi/math/sparse_blas/types.hpp" namespace oneapi::mkl::sparse::mklgpu { namespace detail = oneapi::mkl::sparse::detail; -#include "oneapi/mkl/sparse_blas/detail/onemkl_sparse_blas_backends.hxx" +#include "oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx" } // namespace oneapi::mkl::sparse::mklgpu diff --git a/include/oneapi/mkl/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp b/include/oneapi/math/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp similarity index 85% rename from include/oneapi/mkl/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp rename to include/oneapi/math/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp index d3b0d365f..7b94241cf 100644 --- a/include/oneapi/mkl/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp @@ -20,17 +20,17 @@ #ifndef _ONEMKL_SPARSE_BLAS_DETAIL_MKLGPU_SPARSE_BLAS_CT_HPP_ #define _ONEMKL_SPARSE_BLAS_DETAIL_MKLGPU_SPARSE_BLAS_CT_HPP_ -#include "oneapi/mkl/detail/backends.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" +#include "oneapi/math/detail/backends.hpp" +#include "oneapi/math/detail/backend_selector.hpp" -#include "onemkl_sparse_blas_mklgpu.hpp" +#include "onemath_sparse_blas_mklgpu.hpp" namespace oneapi { namespace mkl { namespace sparse { #define BACKEND mklgpu -#include "oneapi/mkl/sparse_blas/detail/sparse_blas_ct.hxx" +#include "oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx" #undef BACKEND } //namespace sparse diff --git a/include/oneapi/mkl/sparse_blas/detail/onemkl_sparse_blas_backends.hxx b/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx similarity index 100% rename from include/oneapi/mkl/sparse_blas/detail/onemkl_sparse_blas_backends.hxx rename to include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx diff --git a/include/oneapi/mkl/sparse_blas/detail/operation_types.hpp b/include/oneapi/math/sparse_blas/detail/operation_types.hpp similarity index 100% rename from include/oneapi/mkl/sparse_blas/detail/operation_types.hpp rename to include/oneapi/math/sparse_blas/detail/operation_types.hpp diff --git a/include/oneapi/mkl/sparse_blas/detail/sparse_blas_ct.hxx b/include/oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx similarity index 99% rename from include/oneapi/mkl/sparse_blas/detail/sparse_blas_ct.hxx rename to include/oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx index ca09d09d4..8200964d2 100644 --- a/include/oneapi/mkl/sparse_blas/detail/sparse_blas_ct.hxx +++ b/include/oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx @@ -18,7 +18,7 @@ **************************************************************************/ // This file is meant to be included in each backend sparse_blas_ct.hpp files -// Each function calls the implementation from onemkl_sparse_blas_backends.hxx +// Each function calls the implementation from onemath_sparse_blas_backends.hxx #ifndef BACKEND #error "BACKEND is not defined" diff --git a/include/oneapi/mkl/sparse_blas/detail/sparse_blas_rt.hpp b/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp similarity index 99% rename from include/oneapi/mkl/sparse_blas/detail/sparse_blas_rt.hpp rename to include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp index 86a00f507..92e715dc6 100644 --- a/include/oneapi/mkl/sparse_blas/detail/sparse_blas_rt.hpp +++ b/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp @@ -20,8 +20,8 @@ #ifndef _ONEMKL_SPARSE_BLAS_DETAIL_SPARSE_BLAS_RT_HPP_ #define _ONEMKL_SPARSE_BLAS_DETAIL_SPARSE_BLAS_RT_HPP_ -#include "oneapi/mkl/sparse_blas/detail/helper_types.hpp" -#include "oneapi/mkl/sparse_blas/types.hpp" +#include "oneapi/math/sparse_blas/detail/helper_types.hpp" +#include "oneapi/math/sparse_blas/types.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/sparse_blas/matrix_view.hpp b/include/oneapi/math/sparse_blas/matrix_view.hpp similarity index 97% rename from include/oneapi/mkl/sparse_blas/matrix_view.hpp rename to include/oneapi/math/sparse_blas/matrix_view.hpp index 08762c5d7..196bd0271 100644 --- a/include/oneapi/mkl/sparse_blas/matrix_view.hpp +++ b/include/oneapi/math/sparse_blas/matrix_view.hpp @@ -20,7 +20,7 @@ #ifndef _ONEMKL_SPARSE_BLAS_MATRIX_VIEW_HPP_ #define _ONEMKL_SPARSE_BLAS_MATRIX_VIEW_HPP_ -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" namespace oneapi { namespace mkl { diff --git a/include/oneapi/mkl/sparse_blas/types.hpp b/include/oneapi/math/sparse_blas/types.hpp similarity index 98% rename from include/oneapi/mkl/sparse_blas/types.hpp rename to include/oneapi/math/sparse_blas/types.hpp index d619be4b3..da4e2340a 100644 --- a/include/oneapi/mkl/sparse_blas/types.hpp +++ b/include/oneapi/math/sparse_blas/types.hpp @@ -20,7 +20,7 @@ #ifndef _ONEMKL_SPARSE_BLAS_TYPES_HPP_ #define _ONEMKL_SPARSE_BLAS_TYPES_HPP_ -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" #include "matrix_view.hpp" #include "detail/handles.hpp" #include "detail/operation_types.hpp" diff --git a/include/oneapi/mkl/types.hpp b/include/oneapi/math/types.hpp similarity index 98% rename from include/oneapi/mkl/types.hpp rename to include/oneapi/math/types.hpp index 32d336e11..109bd6afe 100644 --- a/include/oneapi/mkl/types.hpp +++ b/include/oneapi/math/types.hpp @@ -21,7 +21,7 @@ #define _ONEMKL_TYPES_HPP_ #ifdef __HIPSYCL__ -#include "oneapi/mkl/bfloat16.hpp" +#include "oneapi/math/bfloat16.hpp" #endif #if __has_include() diff --git a/scripts/generate_backend_api.py b/scripts/generate_backend_api.py index c3866270e..b35165759 100755 --- a/scripts/generate_backend_api.py +++ b/scripts/generate_backend_api.py @@ -43,7 +43,7 @@ def usage(err = None): "onemkl_blas_mklgpu.hpp" header with declaration of all backend library APIs. API from backend library will be called from "oneapi::mkl::mklgpu::blas" namespace. -{script} include/oneapi/mkl/blas.hpp include/oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp oneapi::mkl::mklgpu::blas +{script} include/oneapi/math/blas.hpp include/oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp oneapi::mkl::mklgpu::blas '''.format(script = argv[0])) if len(argv) < 3: @@ -95,7 +95,7 @@ def print_declaration(func_list): #include #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" """.format(in_filename=in_filename)) for nmsp in namespace_list: diff --git a/scripts/generate_cmake.py b/scripts/generate_cmake.py index 36cac3669..5538e972d 100644 --- a/scripts/generate_cmake.py +++ b/scripts/generate_cmake.py @@ -38,7 +38,7 @@ def usage(err = None): Example: - {script} include/oneapi/mkl/blas/detail/mklgpu mklgpu + {script} include/oneapi/math/blas/detail/mklgpu mklgpu '''.format(script = argv[0])) if len(argv) <= 2: diff --git a/scripts/generate_ct_instant.py b/scripts/generate_ct_instant.py index acc02e3ea..afcd02e3e 100755 --- a/scripts/generate_ct_instant.py +++ b/scripts/generate_ct_instant.py @@ -43,7 +43,7 @@ def usage(err = None): "blas_ct.hpp" header with compile-time BLAS API based on "blas_ct_templates.hpp" for "mklgpu" backend. API from the backend library will be called from "oneapi::mkl::mklgpu::blas" namespace. -{script} include/oneapi/mkl/blas/detail/blas_ct_templates.hpp include/oneapi/mkl/blas/detail/mklgpu/blas_ct.hpp include/oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp mklgpu oneapi::mkl::mklgpu::blas +{script} include/oneapi/math/blas/detail/blas_ct_templates.hpp include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp include/oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp mklgpu oneapi::mkl::mklgpu::blas '''.format(script = argv[0])) if len(argv) < 6: @@ -101,8 +101,8 @@ def print_funcs(func_list): #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/backends.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/backends.hpp" #include "{internal_api}" #include "{ct_teplates}" diff --git a/scripts/generate_ct_templates.py b/scripts/generate_ct_templates.py index 7f54d4f08..9d7768727 100755 --- a/scripts/generate_ct_templates.py +++ b/scripts/generate_ct_templates.py @@ -42,7 +42,7 @@ def usage(err = None): The command below will generate: "blas_ct_templates.hpp" header with general templates for compile-time BLAS API based on "blas.hpp". - {script} include/oneapi/mkl/blas.hpp include/oneapi/mkl/blas/detail/blas_ct_templates.hpp + {script} include/oneapi/math/blas.hpp include/oneapi/math/blas/detail/blas_ct_templates.hpp '''.format(script = argv[0])) if len(argv) < 2: @@ -90,8 +90,8 @@ def print_funcs(func_list): #include #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/backends.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/backends.hpp" """.format(in_header=in_filename)) diff --git a/scripts/generate_wrappers.py b/scripts/generate_wrappers.py index 29947273f..06e021aea 100755 --- a/scripts/generate_wrappers.py +++ b/scripts/generate_wrappers.py @@ -40,7 +40,7 @@ def usage(err = None): Example: - {script} include/oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp src/blas/function_table.hpp src/blas/backend/mklgpu/wrappers.cpp mklgpu + {script} include/oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp src/blas/function_table.hpp src/blas/backend/mklgpu/wrappers.cpp mklgpu '''.format(script = argv[0])) if len(argv) <= 4: @@ -94,7 +94,7 @@ def print_funcs(func_list): #include #endif -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" #include "{header}" diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index 6ff8d5d11..7f089471a 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -60,13 +60,13 @@ function(generate_header_file) set(ONEMKL_ENABLE_ROCFFT_BACKEND ${ENABLE_ROCFFT_BACKEND}) set(ONEMKL_ENABLE_PORTFFT_BACKEND ${ENABLE_PORTFFT_BACKEND}) - configure_file(config.hpp.in "${CMAKE_CURRENT_BINARY_DIR}/oneapi/mkl/config.hpp.configured") + configure_file(config.hpp.in "${CMAKE_CURRENT_BINARY_DIR}/oneapi/math/config.hpp.configured") file(GENERATE - OUTPUT "${CMAKE_CURRENT_BINARY_DIR}/oneapi/mkl/detail/config.hpp" - INPUT "${CMAKE_CURRENT_BINARY_DIR}/oneapi/mkl/config.hpp.configured" + OUTPUT "${CMAKE_CURRENT_BINARY_DIR}/oneapi/math/detail/config.hpp" + INPUT "${CMAKE_CURRENT_BINARY_DIR}/oneapi/math/config.hpp.configured" ) - install(FILES "${CMAKE_CURRENT_BINARY_DIR}/oneapi/mkl/detail/config.hpp" - DESTINATION include/oneapi/mkl/detail + install(FILES "${CMAKE_CURRENT_BINARY_DIR}/oneapi/math/detail/config.hpp" + DESTINATION include/oneapi/math/detail COMPONENT Devel ) endfunction() diff --git a/src/blas/backends/cublas/cublas_batch.cpp b/src/blas/backends/cublas/cublas_batch.cpp index 2975e6c58..f86149f9f 100644 --- a/src/blas/backends/cublas/cublas_batch.cpp +++ b/src/blas/backends/cublas/cublas_batch.cpp @@ -18,8 +18,8 @@ **************************************************************************/ #include "cublas_helper.hpp" #include "cublas_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/cublas/cublas_extensions.cpp b/src/blas/backends/cublas/cublas_extensions.cpp index cc80b483d..edac33387 100644 --- a/src/blas/backends/cublas/cublas_extensions.cpp +++ b/src/blas/backends/cublas/cublas_extensions.cpp @@ -18,8 +18,8 @@ **************************************************************************/ #include "cublas_helper.hpp" #include "cublas_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/cublas/cublas_helper.hpp b/src/blas/backends/cublas/cublas_helper.hpp index 0bd4e6274..d8779d65f 100644 --- a/src/blas/backends/cublas/cublas_helper.hpp +++ b/src/blas/backends/cublas/cublas_helper.hpp @@ -33,7 +33,7 @@ #include #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" #include "runtime_support_helper.hpp" #include "dtype_string.hpp" diff --git a/src/blas/backends/cublas/cublas_level1.cpp b/src/blas/backends/cublas/cublas_level1.cpp index 3b0699c87..3670d97a5 100644 --- a/src/blas/backends/cublas/cublas_level1.cpp +++ b/src/blas/backends/cublas/cublas_level1.cpp @@ -19,8 +19,8 @@ #include "cublas_helper.hpp" #include "cublas_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/cublas/cublas_level2.cpp b/src/blas/backends/cublas/cublas_level2.cpp index 5ce6e5eaf..9eb223a4d 100644 --- a/src/blas/backends/cublas/cublas_level2.cpp +++ b/src/blas/backends/cublas/cublas_level2.cpp @@ -19,8 +19,8 @@ #include "cublas_helper.hpp" #include "cublas_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/cublas/cublas_level3.cpp b/src/blas/backends/cublas/cublas_level3.cpp index be634a15c..ca2324198 100644 --- a/src/blas/backends/cublas/cublas_level3.cpp +++ b/src/blas/backends/cublas/cublas_level3.cpp @@ -19,8 +19,8 @@ #include "cublas_helper.hpp" #include "cublas_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/cublas/cublas_task.hpp b/src/blas/backends/cublas/cublas_task.hpp index b3887fcd4..ff0ca0d8e 100644 --- a/src/blas/backends/cublas/cublas_task.hpp +++ b/src/blas/backends/cublas/cublas_task.hpp @@ -29,7 +29,7 @@ #else #include #endif -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" #ifndef __HIPSYCL__ #include "cublas_scope_handle.hpp" #else diff --git a/src/blas/backends/cublas/cublas_wrappers.cpp b/src/blas/backends/cublas/cublas_wrappers.cpp index ee5c7239f..2c4670a65 100644 --- a/src/blas/backends/cublas/cublas_wrappers.cpp +++ b/src/blas/backends/cublas/cublas_wrappers.cpp @@ -17,7 +17,7 @@ * **************************************************************************/ #include "blas/function_table.hpp" -#include "oneapi/mkl/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" #define WRAPPER_VERSION 1 diff --git a/src/blas/backends/mkl_common/mkl_blas_backend.hpp b/src/blas/backends/mkl_common/mkl_blas_backend.hpp index d45208a6d..e542e221a 100644 --- a/src/blas/backends/mkl_common/mkl_blas_backend.hpp +++ b/src/blas/backends/mkl_common/mkl_blas_backend.hpp @@ -22,7 +22,7 @@ #include #include "mkl_version.h" -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/mklcpu/mklcpu_batch.cpp b/src/blas/backends/mklcpu/mklcpu_batch.cpp index 5ecf4cc69..3be24e4d2 100644 --- a/src/blas/backends/mklcpu/mklcpu_batch.cpp +++ b/src/blas/backends/mklcpu/mklcpu_batch.cpp @@ -23,9 +23,9 @@ #include #endif -#include "oneapi/mkl/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { diff --git a/src/blas/backends/mklcpu/mklcpu_extensions.cpp b/src/blas/backends/mklcpu/mklcpu_extensions.cpp index 215addd5e..f41564eb7 100644 --- a/src/blas/backends/mklcpu/mklcpu_extensions.cpp +++ b/src/blas/backends/mklcpu/mklcpu_extensions.cpp @@ -23,9 +23,9 @@ #include #endif -#include "oneapi/mkl/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { diff --git a/src/blas/backends/mklcpu/mklcpu_level1.cpp b/src/blas/backends/mklcpu/mklcpu_level1.cpp index a4d786673..6beed0bf8 100644 --- a/src/blas/backends/mklcpu/mklcpu_level1.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level1.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/mkl/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklcpu/mklcpu_level2.cpp b/src/blas/backends/mklcpu/mklcpu_level2.cpp index 7bd46078c..545a14b4c 100644 --- a/src/blas/backends/mklcpu/mklcpu_level2.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level2.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/mkl/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklcpu/mklcpu_level3.cpp b/src/blas/backends/mklcpu/mklcpu_level3.cpp index 6433fc98b..435868908 100644 --- a/src/blas/backends/mklcpu/mklcpu_level3.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level3.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/mkl/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklcpu/mklcpu_wrappers.cpp b/src/blas/backends/mklcpu/mklcpu_wrappers.cpp index 527f38a18..bdbcfbc4c 100644 --- a/src/blas/backends/mklcpu/mklcpu_wrappers.cpp +++ b/src/blas/backends/mklcpu/mklcpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "blas/function_table.hpp" -#include "oneapi/mkl/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/blas/backends/mklgpu/mklgpu_batch.cpp b/src/blas/backends/mklgpu/mklgpu_batch.cpp index bad2db82c..b64e75afb 100644 --- a/src/blas/backends/mklgpu/mklgpu_batch.cpp +++ b/src/blas/backends/mklgpu/mklgpu_batch.cpp @@ -23,9 +23,9 @@ #include #endif -#include "oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { diff --git a/src/blas/backends/mklgpu/mklgpu_extensions.cpp b/src/blas/backends/mklgpu/mklgpu_extensions.cpp index c4b1635c8..8c27afa8a 100644 --- a/src/blas/backends/mklgpu/mklgpu_extensions.cpp +++ b/src/blas/backends/mklgpu/mklgpu_extensions.cpp @@ -23,9 +23,9 @@ #include #endif -#include "oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { diff --git a/src/blas/backends/mklgpu/mklgpu_level1.cpp b/src/blas/backends/mklgpu/mklgpu_level1.cpp index 9d853e23d..0031593ec 100644 --- a/src/blas/backends/mklgpu/mklgpu_level1.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level1.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklgpu/mklgpu_level2.cpp b/src/blas/backends/mklgpu/mklgpu_level2.cpp index 2d3fc6b39..8e28da243 100644 --- a/src/blas/backends/mklgpu/mklgpu_level2.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level2.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklgpu/mklgpu_level3.cpp b/src/blas/backends/mklgpu/mklgpu_level3.cpp index 6362c4eaa..d4937096e 100644 --- a/src/blas/backends/mklgpu/mklgpu_level3.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level3.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklgpu/mklgpu_wrappers.cpp b/src/blas/backends/mklgpu/mklgpu_wrappers.cpp index 39b49a93a..867729a5a 100644 --- a/src/blas/backends/mklgpu/mklgpu_wrappers.cpp +++ b/src/blas/backends/mklgpu/mklgpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "blas/function_table.hpp" -#include "oneapi/mkl/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/blas/backends/netlib/netlib_batch.cpp b/src/blas/backends/netlib/netlib_batch.cpp index 69197b09d..ad6aec748 100644 --- a/src/blas/backends/netlib/netlib_batch.cpp +++ b/src/blas/backends/netlib/netlib_batch.cpp @@ -24,8 +24,8 @@ #endif #include "netlib_common.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/netlib/netlib_common.hpp b/src/blas/backends/netlib/netlib_common.hpp index 3a69c70f8..a9208273e 100644 --- a/src/blas/backends/netlib/netlib_common.hpp +++ b/src/blas/backends/netlib/netlib_common.hpp @@ -29,8 +29,8 @@ #include "cblas.h" -#include "oneapi/mkl/blas/detail/netlib/onemkl_blas_netlib.hpp" -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/types.hpp" #define GET_MULTI_PTR template get_multi_ptr().get_raw() diff --git a/src/blas/backends/netlib/netlib_extensions.cpp b/src/blas/backends/netlib/netlib_extensions.cpp index 4815ba598..a6f265c63 100644 --- a/src/blas/backends/netlib/netlib_extensions.cpp +++ b/src/blas/backends/netlib/netlib_extensions.cpp @@ -24,8 +24,8 @@ #endif #include "netlib_common.hpp" -#include "oneapi/mkl/blas/detail/netlib/onemkl_blas_netlib.hpp" -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/exceptions.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/netlib/netlib_level1.cpp b/src/blas/backends/netlib/netlib_level1.cpp index 59830db81..b73fc1a4e 100644 --- a/src/blas/backends/netlib/netlib_level1.cpp +++ b/src/blas/backends/netlib/netlib_level1.cpp @@ -24,8 +24,8 @@ #endif #include "netlib_common.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" inline float abs_val(float val) { return std::abs(val); diff --git a/src/blas/backends/netlib/netlib_level2.cpp b/src/blas/backends/netlib/netlib_level2.cpp index fb63bf3a9..2a893e41c 100644 --- a/src/blas/backends/netlib/netlib_level2.cpp +++ b/src/blas/backends/netlib/netlib_level2.cpp @@ -24,7 +24,7 @@ #endif #include "netlib_common.hpp" -#include "oneapi/mkl/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/netlib/netlib_level3.cpp b/src/blas/backends/netlib/netlib_level3.cpp index c41f78205..04d559e69 100644 --- a/src/blas/backends/netlib/netlib_level3.cpp +++ b/src/blas/backends/netlib/netlib_level3.cpp @@ -24,8 +24,8 @@ #endif #include "netlib_common.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/netlib/netlib_wrappers.cpp b/src/blas/backends/netlib/netlib_wrappers.cpp index 1a377f647..c99875b0f 100644 --- a/src/blas/backends/netlib/netlib_wrappers.cpp +++ b/src/blas/backends/netlib/netlib_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "blas/function_table.hpp" -#include "oneapi/mkl/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" #define WRAPPER_VERSION 1 diff --git a/src/blas/backends/portblas/portblas_batch.cpp b/src/blas/backends/portblas/portblas_batch.cpp index 65f0cd59e..ee5a47d9d 100644 --- a/src/blas/backends/portblas/portblas_batch.cpp +++ b/src/blas/backends/portblas/portblas_batch.cpp @@ -24,8 +24,8 @@ #endif #include "portblas_common.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_common.hpp b/src/blas/backends/portblas/portblas_common.hpp index 1624749e8..709bfdd0c 100644 --- a/src/blas/backends/portblas/portblas_common.hpp +++ b/src/blas/backends/portblas/portblas_common.hpp @@ -21,8 +21,8 @@ #define _PORTBLAS_COMMON_HPP_ #include "portblas.hpp" -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/exceptions.hpp" #include #include diff --git a/src/blas/backends/portblas/portblas_level1_double.cpp b/src/blas/backends/portblas/portblas_level1_double.cpp index 4c99f98c6..070633a9e 100644 --- a/src/blas/backends/portblas/portblas_level1_double.cpp +++ b/src/blas/backends/portblas/portblas_level1_double.cpp @@ -24,8 +24,8 @@ #endif #include "portblas_common.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level1_float.cpp b/src/blas/backends/portblas/portblas_level1_float.cpp index 744729f1a..244d9b4d3 100644 --- a/src/blas/backends/portblas/portblas_level1_float.cpp +++ b/src/blas/backends/portblas/portblas_level1_float.cpp @@ -24,8 +24,8 @@ #endif #include "portblas_common.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level2_double.cpp b/src/blas/backends/portblas/portblas_level2_double.cpp index 092aa0c59..9d67b9bb5 100644 --- a/src/blas/backends/portblas/portblas_level2_double.cpp +++ b/src/blas/backends/portblas/portblas_level2_double.cpp @@ -24,8 +24,8 @@ #endif #include "portblas_common.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level2_float.cpp b/src/blas/backends/portblas/portblas_level2_float.cpp index 7308c05da..8a5579595 100644 --- a/src/blas/backends/portblas/portblas_level2_float.cpp +++ b/src/blas/backends/portblas/portblas_level2_float.cpp @@ -24,8 +24,8 @@ #endif #include "portblas_common.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level3_bfloat16.cpp b/src/blas/backends/portblas/portblas_level3_bfloat16.cpp index 1684b1b3e..afb39b623 100644 --- a/src/blas/backends/portblas/portblas_level3_bfloat16.cpp +++ b/src/blas/backends/portblas/portblas_level3_bfloat16.cpp @@ -23,8 +23,8 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level3_double.cpp b/src/blas/backends/portblas/portblas_level3_double.cpp index 9f9d82d37..1b3d4f31b 100644 --- a/src/blas/backends/portblas/portblas_level3_double.cpp +++ b/src/blas/backends/portblas/portblas_level3_double.cpp @@ -24,8 +24,8 @@ #endif #include "portblas_common.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level3_float.cpp b/src/blas/backends/portblas/portblas_level3_float.cpp index 53a5a1697..d0d461e16 100644 --- a/src/blas/backends/portblas/portblas_level3_float.cpp +++ b/src/blas/backends/portblas/portblas_level3_float.cpp @@ -24,8 +24,8 @@ #endif #include "portblas_common.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level3_half.cpp b/src/blas/backends/portblas/portblas_level3_half.cpp index 0e42528fa..73eb5cd35 100644 --- a/src/blas/backends/portblas/portblas_level3_half.cpp +++ b/src/blas/backends/portblas/portblas_level3_half.cpp @@ -23,8 +23,8 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_wrappers.cpp b/src/blas/backends/portblas/portblas_wrappers.cpp index 3f6170bb7..1afb250fd 100644 --- a/src/blas/backends/portblas/portblas_wrappers.cpp +++ b/src/blas/backends/portblas/portblas_wrappers.cpp @@ -4,7 +4,7 @@ #include "blas/function_table.hpp" -#include "oneapi/mkl/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" #define WRAPPER_VERSION 1 diff --git a/src/blas/backends/rocblas/rocblas_batch.cpp b/src/blas/backends/rocblas/rocblas_batch.cpp index 225616427..f267d1f1a 100644 --- a/src/blas/backends/rocblas/rocblas_batch.cpp +++ b/src/blas/backends/rocblas/rocblas_batch.cpp @@ -22,8 +22,8 @@ #include "rocblas_helper.hpp" #include "rocblas_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" // Helper Functions diff --git a/src/blas/backends/rocblas/rocblas_extensions.cpp b/src/blas/backends/rocblas/rocblas_extensions.cpp index a1fd1df1c..334a3708a 100644 --- a/src/blas/backends/rocblas/rocblas_extensions.cpp +++ b/src/blas/backends/rocblas/rocblas_extensions.cpp @@ -22,8 +22,8 @@ #include "rocblas_helper.hpp" #include "rocblas_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/rocblas/rocblas_helper.hpp b/src/blas/backends/rocblas/rocblas_helper.hpp index 7fd45bdca..2f61d92d5 100644 --- a/src/blas/backends/rocblas/rocblas_helper.hpp +++ b/src/blas/backends/rocblas/rocblas_helper.hpp @@ -29,7 +29,7 @@ #include #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" #include #include "dtype_string.hpp" diff --git a/src/blas/backends/rocblas/rocblas_level1.cpp b/src/blas/backends/rocblas/rocblas_level1.cpp index 9a4f509dc..9d3adf46c 100644 --- a/src/blas/backends/rocblas/rocblas_level1.cpp +++ b/src/blas/backends/rocblas/rocblas_level1.cpp @@ -22,8 +22,8 @@ #include "rocblas_helper.hpp" #include "rocblas_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/rocblas/rocblas_level2.cpp b/src/blas/backends/rocblas/rocblas_level2.cpp index aaddd0d37..009100dc6 100644 --- a/src/blas/backends/rocblas/rocblas_level2.cpp +++ b/src/blas/backends/rocblas/rocblas_level2.cpp @@ -22,8 +22,8 @@ #include "rocblas_helper.hpp" #include "rocblas_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" // Helper Functions diff --git a/src/blas/backends/rocblas/rocblas_level3.cpp b/src/blas/backends/rocblas/rocblas_level3.cpp index 04f67ef5a..5b9f57411 100644 --- a/src/blas/backends/rocblas/rocblas_level3.cpp +++ b/src/blas/backends/rocblas/rocblas_level3.cpp @@ -22,8 +22,8 @@ #include "rocblas_helper.hpp" #include "rocblas_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/rocblas/rocblas_task.hpp b/src/blas/backends/rocblas/rocblas_task.hpp index 3b872e536..e9d8386c2 100644 --- a/src/blas/backends/rocblas/rocblas_task.hpp +++ b/src/blas/backends/rocblas/rocblas_task.hpp @@ -27,7 +27,7 @@ #else #include #endif -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" #ifndef __HIPSYCL__ #include "rocblas_scope_handle.hpp" #else diff --git a/src/blas/backends/rocblas/rocblas_wrappers.cpp b/src/blas/backends/rocblas/rocblas_wrappers.cpp index ce4c92da5..f7122cae5 100644 --- a/src/blas/backends/rocblas/rocblas_wrappers.cpp +++ b/src/blas/backends/rocblas/rocblas_wrappers.cpp @@ -19,7 +19,7 @@ * **************************************************************************/ #include "blas/function_table.hpp" -#include "oneapi/mkl/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" #define WRAPPER_VERSION 1 diff --git a/src/blas/blas_loader.cpp b/src/blas/blas_loader.cpp index 1c3cfcb71..c638770ec 100644 --- a/src/blas/blas_loader.cpp +++ b/src/blas/blas_loader.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/blas/detail/blas_loader.hpp" +#include "oneapi/math/blas/detail/blas_loader.hpp" #include "function_table_initializer.hpp" #include "blas/function_table.hpp" diff --git a/src/blas/function_table.hpp b/src/blas/function_table.hpp index a242fd0c0..68ca9c282 100644 --- a/src/blas/function_table.hpp +++ b/src/blas/function_table.hpp @@ -27,7 +27,7 @@ #else #include #endif -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" typedef struct { int version; diff --git a/src/dft/backends/cufft/CMakeLists.txt b/src/dft/backends/cufft/CMakeLists.txt index 010905546..365642766 100644 --- a/src/dft/backends/cufft/CMakeLists.txt +++ b/src/dft/backends/cufft/CMakeLists.txt @@ -27,7 +27,7 @@ add_library(${LIB_OBJ} OBJECT commit.cpp forward.cpp backward.cpp - $<$: mkl_dft_cufft_wrappers.cpp> + $<$: cufft_wrappers.cpp> ) add_dependencies(onemkl_backend_libs_dft ${LIB_NAME}) diff --git a/src/dft/backends/cufft/backward.cpp b/src/dft/backends/cufft/backward.cpp index 80e475991..86d4a8b81 100644 --- a/src/dft/backends/cufft/backward.cpp +++ b/src/dft/backends/cufft/backward.cpp @@ -23,11 +23,11 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" -#include "oneapi/mkl/dft/detail/commit_impl.hpp" -#include "oneapi/mkl/dft/detail/cufft/onemkl_dft_cufft.hpp" -#include "oneapi/mkl/dft/types.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" +#include "oneapi/math/dft/detail/cufft/onemkl_dft_cufft.hpp" +#include "oneapi/math/dft/types.hpp" #include "execute_helper.hpp" #include "../../execute_helper_generic.hpp" diff --git a/src/dft/backends/cufft/commit.cpp b/src/dft/backends/cufft/commit.cpp index 8dd9b225b..2de2c882b 100644 --- a/src/dft/backends/cufft/commit.cpp +++ b/src/dft/backends/cufft/commit.cpp @@ -27,12 +27,12 @@ #include #include -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" -#include "oneapi/mkl/dft/detail/commit_impl.hpp" -#include "oneapi/mkl/dft/detail/descriptor_impl.hpp" -#include "oneapi/mkl/dft/detail/cufft/onemkl_dft_cufft.hpp" -#include "oneapi/mkl/dft/types.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" +#include "oneapi/math/dft/detail/descriptor_impl.hpp" +#include "oneapi/math/dft/detail/cufft/onemkl_dft_cufft.hpp" +#include "oneapi/math/dft/types.hpp" #include "../stride_helper.hpp" diff --git a/src/dft/backends/cufft/mkl_dft_cufft_wrappers.cpp b/src/dft/backends/cufft/cufft_wrappers.cpp similarity index 94% rename from src/dft/backends/cufft/mkl_dft_cufft_wrappers.cpp rename to src/dft/backends/cufft/cufft_wrappers.cpp index 93d3aae11..3401859db 100644 --- a/src/dft/backends/cufft/mkl_dft_cufft_wrappers.cpp +++ b/src/dft/backends/cufft/cufft_wrappers.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/detail/cufft/onemkl_dft_cufft.hpp" +#include "oneapi/math/dft/detail/cufft/onemkl_dft_cufft.hpp" #include "dft/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/dft/backends/cufft/descriptor.cpp b/src/dft/backends/cufft/descriptor.cpp index bf26b600f..f61dd1b91 100644 --- a/src/dft/backends/cufft/descriptor.cpp +++ b/src/dft/backends/cufft/descriptor.cpp @@ -17,10 +17,10 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/descriptor.hpp" +#include "oneapi/math/dft/descriptor.hpp" #include "../../descriptor.cxx" -#include "oneapi/mkl/dft/detail/cufft/onemkl_dft_cufft.hpp" +#include "oneapi/math/dft/detail/cufft/onemkl_dft_cufft.hpp" namespace oneapi::mkl::dft::detail { diff --git a/src/dft/backends/cufft/execute_helper.hpp b/src/dft/backends/cufft/execute_helper.hpp index 7b7d946db..e6014478c 100644 --- a/src/dft/backends/cufft/execute_helper.hpp +++ b/src/dft/backends/cufft/execute_helper.hpp @@ -26,10 +26,10 @@ #include #endif -#include "oneapi/mkl/dft/detail/commit_impl.hpp" -#include "oneapi/mkl/dft/detail/descriptor_impl.hpp" -#include "oneapi/mkl/dft/types.hpp" -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" +#include "oneapi/math/dft/detail/descriptor_impl.hpp" +#include "oneapi/math/dft/types.hpp" +#include "oneapi/math/exceptions.hpp" #include #include diff --git a/src/dft/backends/cufft/forward.cpp b/src/dft/backends/cufft/forward.cpp index 7cf73976d..e104e1602 100644 --- a/src/dft/backends/cufft/forward.cpp +++ b/src/dft/backends/cufft/forward.cpp @@ -24,11 +24,11 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" -#include "oneapi/mkl/dft/detail/commit_impl.hpp" -#include "oneapi/mkl/dft/detail/cufft/onemkl_dft_cufft.hpp" -#include "oneapi/mkl/dft/types.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" +#include "oneapi/math/dft/detail/cufft/onemkl_dft_cufft.hpp" +#include "oneapi/math/dft/types.hpp" #include "execute_helper.hpp" #include "../../execute_helper_generic.hpp" diff --git a/src/dft/backends/descriptor.cpp b/src/dft/backends/descriptor.cpp index c6f6884f8..dbc4a7f71 100644 --- a/src/dft/backends/descriptor.cpp +++ b/src/dft/backends/descriptor.cpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/descriptor.hpp" -#include "oneapi/mkl/dft/detail/dft_loader.hpp" +#include "oneapi/math/dft/descriptor.hpp" +#include "oneapi/math/dft/detail/dft_loader.hpp" #include "../descriptor.cxx" diff --git a/src/dft/backends/mklcpu/backward.cpp b/src/dft/backends/mklcpu/backward.cpp index fe7186630..77365e1d0 100644 --- a/src/dft/backends/mklcpu/backward.cpp +++ b/src/dft/backends/mklcpu/backward.cpp @@ -23,11 +23,11 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" -#include "oneapi/mkl/dft/descriptor.hpp" -#include "oneapi/mkl/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" -#include "oneapi/mkl/dft/types.hpp" +#include "oneapi/math/dft/descriptor.hpp" +#include "oneapi/math/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" +#include "oneapi/math/dft/types.hpp" #include "dft/backends/mklcpu/commit_derived_impl.hpp" diff --git a/src/dft/backends/mklcpu/commit.cpp b/src/dft/backends/mklcpu/commit.cpp index 1ec8aef9c..be9cf375c 100644 --- a/src/dft/backends/mklcpu/commit.cpp +++ b/src/dft/backends/mklcpu/commit.cpp @@ -23,14 +23,14 @@ #include #endif -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/detail/backends.hpp" -#include "oneapi/mkl/dft/types.hpp" -#include "oneapi/mkl/dft/descriptor.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/detail/backends.hpp" +#include "oneapi/math/dft/types.hpp" +#include "oneapi/math/dft/descriptor.hpp" -#include "oneapi/mkl/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" +#include "oneapi/math/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" -#include "oneapi/mkl/dft/detail/commit_impl.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" #include "dft/backends/mklcpu/commit_derived_impl.hpp" #include "../stride_helper.hpp" diff --git a/src/dft/backends/mklcpu/commit_derived_impl.hpp b/src/dft/backends/mklcpu/commit_derived_impl.hpp index 3551758a0..61b55942b 100644 --- a/src/dft/backends/mklcpu/commit_derived_impl.hpp +++ b/src/dft/backends/mklcpu/commit_derived_impl.hpp @@ -20,8 +20,8 @@ #ifndef _ONEMKL_DFT_COMMIT_DERIVED_IMPL_HPP_ #define _ONEMKL_DFT_COMMIT_DERIVED_IMPL_HPP_ -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" #include "dft/backends/mklcpu/mklcpu_helpers.hpp" // MKLCPU header diff --git a/src/dft/backends/mklcpu/descriptor.cpp b/src/dft/backends/mklcpu/descriptor.cpp index a72fdcfc3..95a49229b 100644 --- a/src/dft/backends/mklcpu/descriptor.cpp +++ b/src/dft/backends/mklcpu/descriptor.cpp @@ -17,10 +17,10 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/descriptor.hpp" +#include "oneapi/math/dft/descriptor.hpp" #include "../../descriptor.cxx" -#include "oneapi/mkl/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" +#include "oneapi/math/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" namespace oneapi::mkl::dft::detail { diff --git a/src/dft/backends/mklcpu/forward.cpp b/src/dft/backends/mklcpu/forward.cpp index 2e5e2fa88..7a1ac1524 100644 --- a/src/dft/backends/mklcpu/forward.cpp +++ b/src/dft/backends/mklcpu/forward.cpp @@ -23,11 +23,11 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" -#include "oneapi/mkl/dft/descriptor.hpp" -#include "oneapi/mkl/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" -#include "oneapi/mkl/dft/types.hpp" +#include "oneapi/math/dft/descriptor.hpp" +#include "oneapi/math/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" +#include "oneapi/math/dft/types.hpp" #include "dft/backends/mklcpu/commit_derived_impl.hpp" diff --git a/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp b/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp index 463ab80f4..ad0687b5b 100644 --- a/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp +++ b/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" +#include "oneapi/math/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" #include "dft/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/dft/backends/mklcpu/mklcpu_helpers.hpp b/src/dft/backends/mklcpu/mklcpu_helpers.hpp index 55a8345c2..8c0f1cf1a 100644 --- a/src/dft/backends/mklcpu/mklcpu_helpers.hpp +++ b/src/dft/backends/mklcpu/mklcpu_helpers.hpp @@ -20,8 +20,8 @@ #ifndef _ONEMKL_DFT_SRC_MKLCPU_HELPERS_HPP_ #define _ONEMKL_DFT_SRC_MKLCPU_HELPERS_HPP_ -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" // MKLCPU header #include "mkl_dfti.h" diff --git a/src/dft/backends/mklgpu/backward.cpp b/src/dft/backends/mklgpu/backward.cpp index 4899ed3e7..1811cb65d 100644 --- a/src/dft/backends/mklgpu/backward.cpp +++ b/src/dft/backends/mklgpu/backward.cpp @@ -23,10 +23,10 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" -#include "oneapi/mkl/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" -#include "oneapi/mkl/dft/detail/descriptor_impl.hpp" +#include "oneapi/math/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" +#include "oneapi/math/dft/detail/descriptor_impl.hpp" #include "mklgpu_helpers.hpp" diff --git a/src/dft/backends/mklgpu/commit.cpp b/src/dft/backends/mklgpu/commit.cpp index bae1eb69f..a70c5223d 100644 --- a/src/dft/backends/mklgpu/commit.cpp +++ b/src/dft/backends/mklgpu/commit.cpp @@ -23,14 +23,14 @@ #include #endif -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/detail/backends.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/detail/backends.hpp" -#include "oneapi/mkl/dft/detail/commit_impl.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" -#include "oneapi/mkl/dft/detail/descriptor_impl.hpp" -#include "oneapi/mkl/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" +#include "oneapi/math/dft/detail/descriptor_impl.hpp" +#include "oneapi/math/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" #include "dft/backends/mklgpu/mklgpu_helpers.hpp" #include "../stride_helper.hpp" diff --git a/src/dft/backends/mklgpu/descriptor.cpp b/src/dft/backends/mklgpu/descriptor.cpp index 7f7f0bf70..bd129b4ff 100644 --- a/src/dft/backends/mklgpu/descriptor.cpp +++ b/src/dft/backends/mklgpu/descriptor.cpp @@ -17,10 +17,10 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/descriptor.hpp" +#include "oneapi/math/dft/descriptor.hpp" #include "../../descriptor.cxx" -#include "oneapi/mkl/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" +#include "oneapi/math/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" namespace oneapi::mkl::dft::detail { diff --git a/src/dft/backends/mklgpu/forward.cpp b/src/dft/backends/mklgpu/forward.cpp index 9dbffa081..b8c09dc77 100644 --- a/src/dft/backends/mklgpu/forward.cpp +++ b/src/dft/backends/mklgpu/forward.cpp @@ -24,10 +24,10 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" -#include "oneapi/mkl/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" -#include "oneapi/mkl/dft/detail/descriptor_impl.hpp" +#include "oneapi/math/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" +#include "oneapi/math/dft/detail/descriptor_impl.hpp" #include "mklgpu_helpers.hpp" diff --git a/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp b/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp index 8d2fa111d..fc04491a0 100644 --- a/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp +++ b/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" +#include "oneapi/math/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" #include "dft/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/dft/backends/mklgpu/mklgpu_helpers.hpp b/src/dft/backends/mklgpu/mklgpu_helpers.hpp index 4b93d1423..08119253c 100644 --- a/src/dft/backends/mklgpu/mklgpu_helpers.hpp +++ b/src/dft/backends/mklgpu/mklgpu_helpers.hpp @@ -20,8 +20,8 @@ #ifndef _ONEMKL_DFT_SRC_MKLGPU_HELPERS_HPP_ #define _ONEMKL_DFT_SRC_MKLGPU_HELPERS_HPP_ -#include "oneapi/mkl/detail/exceptions.hpp" -#include "oneapi/mkl/dft/detail/types_impl.hpp" +#include "oneapi/math/detail/exceptions.hpp" +#include "oneapi/math/dft/detail/types_impl.hpp" #include "mkl_version.h" // MKLGPU header diff --git a/src/dft/backends/portfft/CMakeLists.txt b/src/dft/backends/portfft/CMakeLists.txt index 752fae8d5..2491a6587 100644 --- a/src/dft/backends/portfft/CMakeLists.txt +++ b/src/dft/backends/portfft/CMakeLists.txt @@ -65,7 +65,7 @@ add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT descriptor.cpp commit.cpp - $<$: mkl_dft_portfft_wrappers.cpp> + $<$: portfft_wrappers.cpp> ) add_dependencies(onemkl_backend_libs_dft ${LIB_NAME}) diff --git a/src/dft/backends/portfft/commit.cpp b/src/dft/backends/portfft/commit.cpp index a2c80e91a..aa92256ab 100644 --- a/src/dft/backends/portfft/commit.cpp +++ b/src/dft/backends/portfft/commit.cpp @@ -28,12 +28,12 @@ #include -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" -#include "oneapi/mkl/dft/detail/commit_impl.hpp" -#include "oneapi/mkl/dft/detail/descriptor_impl.hpp" -#include "oneapi/mkl/dft/detail/portfft/onemkl_dft_portfft.hpp" -#include "oneapi/mkl/dft/types.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" +#include "oneapi/math/dft/detail/descriptor_impl.hpp" +#include "oneapi/math/dft/detail/portfft/onemkl_dft_portfft.hpp" +#include "oneapi/math/dft/types.hpp" #include "../stride_helper.hpp" diff --git a/src/dft/backends/portfft/descriptor.cpp b/src/dft/backends/portfft/descriptor.cpp index c45b9f2c5..7bdcf21cc 100644 --- a/src/dft/backends/portfft/descriptor.cpp +++ b/src/dft/backends/portfft/descriptor.cpp @@ -17,10 +17,10 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/descriptor.hpp" +#include "oneapi/math/dft/descriptor.hpp" #include "../../descriptor.cxx" -#include "oneapi/mkl/dft/detail/portfft/onemkl_dft_portfft.hpp" +#include "oneapi/math/dft/detail/portfft/onemkl_dft_portfft.hpp" namespace oneapi::mkl::dft::detail { diff --git a/src/dft/backends/portfft/portfft_helper.hpp b/src/dft/backends/portfft/portfft_helper.hpp index 373865f49..4322987a8 100644 --- a/src/dft/backends/portfft/portfft_helper.hpp +++ b/src/dft/backends/portfft/portfft_helper.hpp @@ -24,8 +24,8 @@ #include -#include "oneapi/mkl/dft/detail/commit_impl.hpp" -#include "oneapi/mkl/dft/detail/descriptor_impl.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" +#include "oneapi/math/dft/detail/descriptor_impl.hpp" namespace pfft = portfft; diff --git a/src/dft/backends/portfft/mkl_dft_portfft_wrappers.cpp b/src/dft/backends/portfft/portfft_wrappers.cpp similarity index 94% rename from src/dft/backends/portfft/mkl_dft_portfft_wrappers.cpp rename to src/dft/backends/portfft/portfft_wrappers.cpp index 28996b0a1..22abfcb81 100644 --- a/src/dft/backends/portfft/mkl_dft_portfft_wrappers.cpp +++ b/src/dft/backends/portfft/portfft_wrappers.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/detail/portfft/onemkl_dft_portfft.hpp" +#include "oneapi/math/dft/detail/portfft/onemkl_dft_portfft.hpp" #include "dft/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/dft/backends/rocfft/CMakeLists.txt b/src/dft/backends/rocfft/CMakeLists.txt index 1380c8f0a..cae182254 100644 --- a/src/dft/backends/rocfft/CMakeLists.txt +++ b/src/dft/backends/rocfft/CMakeLists.txt @@ -27,7 +27,7 @@ add_library(${LIB_OBJ} OBJECT commit.cpp forward.cpp backward.cpp - $<$: mkl_dft_rocfft_wrappers.cpp> + $<$: rocfft_wrappers.cpp> ) add_dependencies(onemkl_backend_libs_dft ${LIB_NAME}) diff --git a/src/dft/backends/rocfft/backward.cpp b/src/dft/backends/rocfft/backward.cpp index e76437ee2..c835cdf34 100644 --- a/src/dft/backends/rocfft/backward.cpp +++ b/src/dft/backends/rocfft/backward.cpp @@ -23,10 +23,10 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" -#include "oneapi/mkl/dft/detail/rocfft/onemkl_dft_rocfft.hpp" -#include "oneapi/mkl/dft/descriptor.hpp" +#include "oneapi/math/dft/detail/rocfft/onemkl_dft_rocfft.hpp" +#include "oneapi/math/dft/descriptor.hpp" #include "execute_helper.hpp" #include "../../execute_helper_generic.hpp" diff --git a/src/dft/backends/rocfft/commit.cpp b/src/dft/backends/rocfft/commit.cpp index 991b0c471..54b7b5e9c 100644 --- a/src/dft/backends/rocfft/commit.cpp +++ b/src/dft/backends/rocfft/commit.cpp @@ -27,12 +27,12 @@ #include #include -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" -#include "oneapi/mkl/dft/detail/commit_impl.hpp" -#include "oneapi/mkl/dft/detail/descriptor_impl.hpp" -#include "oneapi/mkl/dft/detail/rocfft/onemkl_dft_rocfft.hpp" -#include "oneapi/mkl/dft/types.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" +#include "oneapi/math/dft/detail/descriptor_impl.hpp" +#include "oneapi/math/dft/detail/rocfft/onemkl_dft_rocfft.hpp" +#include "oneapi/math/dft/types.hpp" #include "../stride_helper.hpp" diff --git a/src/dft/backends/rocfft/descriptor.cpp b/src/dft/backends/rocfft/descriptor.cpp index 22f21590a..9a235c46f 100644 --- a/src/dft/backends/rocfft/descriptor.cpp +++ b/src/dft/backends/rocfft/descriptor.cpp @@ -17,10 +17,10 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/descriptor.hpp" +#include "oneapi/math/dft/descriptor.hpp" #include "../../descriptor.cxx" -#include "oneapi/mkl/dft/detail/rocfft/onemkl_dft_rocfft.hpp" +#include "oneapi/math/dft/detail/rocfft/onemkl_dft_rocfft.hpp" namespace oneapi::mkl::dft::detail { diff --git a/src/dft/backends/rocfft/execute_helper.hpp b/src/dft/backends/rocfft/execute_helper.hpp index 78663a090..fa2eecb24 100644 --- a/src/dft/backends/rocfft/execute_helper.hpp +++ b/src/dft/backends/rocfft/execute_helper.hpp @@ -26,10 +26,10 @@ #include #endif -#include "oneapi/mkl/dft/detail/commit_impl.hpp" -#include "oneapi/mkl/dft/detail/descriptor_impl.hpp" -#include "oneapi/mkl/dft/types.hpp" -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/dft/detail/commit_impl.hpp" +#include "oneapi/math/dft/detail/descriptor_impl.hpp" +#include "oneapi/math/dft/types.hpp" +#include "oneapi/math/exceptions.hpp" #include #include diff --git a/src/dft/backends/rocfft/forward.cpp b/src/dft/backends/rocfft/forward.cpp index d9a576720..5412f4678 100644 --- a/src/dft/backends/rocfft/forward.cpp +++ b/src/dft/backends/rocfft/forward.cpp @@ -24,10 +24,10 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" -#include "oneapi/mkl/dft/detail/rocfft/onemkl_dft_rocfft.hpp" -#include "oneapi/mkl/dft/descriptor.hpp" +#include "oneapi/math/dft/detail/rocfft/onemkl_dft_rocfft.hpp" +#include "oneapi/math/dft/descriptor.hpp" #include "execute_helper.hpp" #include "../../execute_helper_generic.hpp" diff --git a/src/dft/backends/rocfft/mkl_dft_rocfft_wrappers.cpp b/src/dft/backends/rocfft/rocfft_wrappers.cpp similarity index 94% rename from src/dft/backends/rocfft/mkl_dft_rocfft_wrappers.cpp rename to src/dft/backends/rocfft/rocfft_wrappers.cpp index c8f0e35c7..c2afb0fb2 100644 --- a/src/dft/backends/rocfft/mkl_dft_rocfft_wrappers.cpp +++ b/src/dft/backends/rocfft/rocfft_wrappers.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/detail/rocfft/onemkl_dft_rocfft.hpp" +#include "oneapi/math/dft/detail/rocfft/onemkl_dft_rocfft.hpp" #include "dft/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/dft/descriptor.cxx b/src/dft/descriptor.cxx index aa0d9d70e..a823817c0 100644 --- a/src/dft/descriptor.cxx +++ b/src/dft/descriptor.cxx @@ -18,8 +18,8 @@ *******************************************************************************/ #include -#include "oneapi/mkl/detail/exceptions.hpp" -#include "oneapi/mkl/dft/descriptor.hpp" +#include "oneapi/math/detail/exceptions.hpp" +#include "oneapi/math/dft/descriptor.hpp" #include "dft/descriptor_config_helper.hpp" diff --git a/src/dft/descriptor_config_helper.hpp b/src/dft/descriptor_config_helper.hpp index dc8c97ac2..109e39e80 100644 --- a/src/dft/descriptor_config_helper.hpp +++ b/src/dft/descriptor_config_helper.hpp @@ -23,7 +23,7 @@ #include #include -#include "oneapi/mkl/dft/descriptor.hpp" +#include "oneapi/math/dft/descriptor.hpp" namespace oneapi { namespace mkl { diff --git a/src/dft/dft_loader.cpp b/src/dft/dft_loader.cpp index 55a280388..f5adec10b 100644 --- a/src/dft/dft_loader.cpp +++ b/src/dft/dft_loader.cpp @@ -17,13 +17,13 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/dft/detail/dft_loader.hpp" -#include "oneapi/mkl/dft/forward.hpp" -#include "oneapi/mkl/dft/backward.hpp" +#include "oneapi/math/dft/detail/dft_loader.hpp" +#include "oneapi/math/dft/forward.hpp" +#include "oneapi/math/dft/backward.hpp" #include "function_table_initializer.hpp" #include "dft/function_table.hpp" -#include "oneapi/mkl/detail/get_device_id.hpp" +#include "oneapi/math/detail/get_device_id.hpp" namespace oneapi::mkl::dft::detail { diff --git a/src/dft/function_table.hpp b/src/dft/function_table.hpp index 9146f239e..9a491361a 100644 --- a/src/dft/function_table.hpp +++ b/src/dft/function_table.hpp @@ -29,9 +29,9 @@ #include #endif -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/dft/types.hpp" -#include "oneapi/mkl/dft/descriptor.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/dft/types.hpp" +#include "oneapi/math/dft/descriptor.hpp" typedef struct { int version; diff --git a/src/include/function_table_initializer.hpp b/src/include/function_table_initializer.hpp index 8a870e218..7655b5f4c 100644 --- a/src/include/function_table_initializer.hpp +++ b/src/include/function_table_initializer.hpp @@ -23,8 +23,8 @@ #include #include -#include "oneapi/mkl/detail/backends_table.hpp" -#include "oneapi/mkl/detail/exceptions.hpp" +#include "oneapi/math/detail/backends_table.hpp" +#include "oneapi/math/detail/exceptions.hpp" #define SPEC_VERSION 1 diff --git a/src/lapack/backends/cusolver/cusolver_batch.cpp b/src/lapack/backends/cusolver/cusolver_batch.cpp index f4017f873..da4ec8907 100644 --- a/src/lapack/backends/cusolver/cusolver_batch.cpp +++ b/src/lapack/backends/cusolver/cusolver_batch.cpp @@ -20,8 +20,8 @@ #include "cusolver_helper.hpp" #include "cusolver_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp" namespace oneapi { namespace mkl { diff --git a/src/lapack/backends/cusolver/cusolver_helper.hpp b/src/lapack/backends/cusolver/cusolver_helper.hpp index 954d41246..861e45ef8 100644 --- a/src/lapack/backends/cusolver/cusolver_helper.hpp +++ b/src/lapack/backends/cusolver/cusolver_helper.hpp @@ -33,10 +33,10 @@ #include #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" #include "runtime_support_helper.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/lapack/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/lapack/exceptions.hpp" namespace oneapi { namespace mkl { @@ -44,7 +44,7 @@ namespace lapack { namespace cusolver { // The static assert to make sure that all index types used in -// oneMKL/include/oneapi/mkl/lapack.hpp interface are int64_t +// oneMKL/include/oneapi/math/lapack.hpp interface are int64_t template struct is_int64 : std::false_type {}; diff --git a/src/lapack/backends/cusolver/cusolver_lapack.cpp b/src/lapack/backends/cusolver/cusolver_lapack.cpp index c8190f50d..9c57fdeb6 100644 --- a/src/lapack/backends/cusolver/cusolver_lapack.cpp +++ b/src/lapack/backends/cusolver/cusolver_lapack.cpp @@ -19,8 +19,8 @@ #include "cusolver_helper.hpp" #include "cusolver_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp" namespace oneapi { namespace mkl { diff --git a/src/lapack/backends/cusolver/cusolver_task.hpp b/src/lapack/backends/cusolver/cusolver_task.hpp index 96107f959..2985e935e 100644 --- a/src/lapack/backends/cusolver/cusolver_task.hpp +++ b/src/lapack/backends/cusolver/cusolver_task.hpp @@ -28,7 +28,7 @@ #else #include #endif -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" #include "cusolver_scope_handle.hpp" // After Plugin Interface removal in DPC++ ur.hpp is the new include diff --git a/src/lapack/backends/cusolver/cusolver_wrappers.cpp b/src/lapack/backends/cusolver/cusolver_wrappers.cpp index 4b5ab8e2c..60e4a57a9 100644 --- a/src/lapack/backends/cusolver/cusolver_wrappers.cpp +++ b/src/lapack/backends/cusolver/cusolver_wrappers.cpp @@ -17,7 +17,7 @@ * **************************************************************************/ #include "lapack/function_table.hpp" -#include "oneapi/mkl/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp" +#include "oneapi/math/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp" #define WRAPPER_VERSION 1 diff --git a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp b/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp index 1932bb959..706a76dcc 100644 --- a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp +++ b/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp @@ -21,8 +21,8 @@ #include -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/lapack/types.hpp" namespace oneapi { namespace mkl { diff --git a/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp b/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp index 4bd0713fa..c4ee847f1 100644 --- a/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp +++ b/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "lapack/function_table.hpp" -#include "oneapi/mkl/lapack/detail/mklcpu/onemkl_lapack_mklcpu.hpp" +#include "oneapi/math/lapack/detail/mklcpu/onemkl_lapack_mklcpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/lapack/backends/mklcpu/mkl_lapack.cpp b/src/lapack/backends/mklcpu/mkl_lapack.cpp index cbd3aaa84..1f94a05e3 100644 --- a/src/lapack/backends/mklcpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklcpu/mkl_lapack.cpp @@ -23,9 +23,9 @@ #include #endif -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" -#include "oneapi/mkl/lapack/detail/mklcpu/onemkl_lapack_mklcpu.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/lapack/types.hpp" +#include "oneapi/math/lapack/detail/mklcpu/onemkl_lapack_mklcpu.hpp" #include "../mkl_common/mkl_lapack_backend.hpp" namespace oneapi { diff --git a/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp b/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp index cd3933274..4a75f1c66 100644 --- a/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp +++ b/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "lapack/function_table.hpp" -#include "oneapi/mkl/lapack/detail/mklgpu/onemkl_lapack_mklgpu.hpp" +#include "oneapi/math/lapack/detail/mklgpu/onemkl_lapack_mklgpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/lapack/backends/mklgpu/mkl_lapack.cpp b/src/lapack/backends/mklgpu/mkl_lapack.cpp index f7bc2a7e6..ea5b5749d 100644 --- a/src/lapack/backends/mklgpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklgpu/mkl_lapack.cpp @@ -23,9 +23,9 @@ #include #endif -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" -#include "oneapi/mkl/lapack/detail/mklgpu/onemkl_lapack_mklgpu.hpp" +#include "oneapi/math/types.hpp" +#include "oneapi/math/lapack/types.hpp" +#include "oneapi/math/lapack/detail/mklgpu/onemkl_lapack_mklgpu.hpp" #include "../mkl_common/mkl_lapack_backend.hpp" namespace oneapi { diff --git a/src/lapack/backends/rocsolver/rocsolver_batch.cpp b/src/lapack/backends/rocsolver/rocsolver_batch.cpp index 4363be306..a3d2381da 100644 --- a/src/lapack/backends/rocsolver/rocsolver_batch.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_batch.cpp @@ -21,8 +21,8 @@ #include "rocsolver_helper.hpp" #include "rocsolver_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp" namespace oneapi { namespace mkl { diff --git a/src/lapack/backends/rocsolver/rocsolver_helper.hpp b/src/lapack/backends/rocsolver/rocsolver_helper.hpp index 34064fb09..d51c240a9 100644 --- a/src/lapack/backends/rocsolver/rocsolver_helper.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_helper.hpp @@ -32,10 +32,10 @@ #include #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" #include "runtime_support_helper.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/lapack/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/lapack/exceptions.hpp" namespace oneapi { namespace mkl { @@ -43,7 +43,7 @@ namespace lapack { namespace rocsolver { // The static assert to make sure that all index types used in -// oneMKL/include/oneapi/mkl/lapack.hpp interface are int64_t +// oneMKL/include/oneapi/math/lapack.hpp interface are int64_t template struct is_int64 : std::false_type {}; diff --git a/src/lapack/backends/rocsolver/rocsolver_lapack.cpp b/src/lapack/backends/rocsolver/rocsolver_lapack.cpp index d3e1b9e26..d3cf85d03 100644 --- a/src/lapack/backends/rocsolver/rocsolver_lapack.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_lapack.cpp @@ -21,8 +21,8 @@ #include "rocsolver_helper.hpp" #include "rocsolver_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp" namespace oneapi { namespace mkl { diff --git a/src/lapack/backends/rocsolver/rocsolver_task.hpp b/src/lapack/backends/rocsolver/rocsolver_task.hpp index 4842a18e9..a5a30b3d1 100644 --- a/src/lapack/backends/rocsolver/rocsolver_task.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_task.hpp @@ -30,7 +30,7 @@ #else #include #endif -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" #include "rocsolver_scope_handle.hpp" // After Plugin Interface removal in DPC++ ur.hpp is the new include diff --git a/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp b/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp index 8613cc05e..b0c2fa489 100644 --- a/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp @@ -19,7 +19,7 @@ * **************************************************************************/ #include "lapack/function_table.hpp" -#include "oneapi/mkl/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp" +#include "oneapi/math/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp" #define WRAPPER_VERSION 1 diff --git a/src/lapack/function_table.hpp b/src/lapack/function_table.hpp index e034fe357..4a0b9293e 100644 --- a/src/lapack/function_table.hpp +++ b/src/lapack/function_table.hpp @@ -28,7 +28,7 @@ #include #endif -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" typedef struct { int version; diff --git a/src/lapack/lapack_loader.cpp b/src/lapack/lapack_loader.cpp index f26e5f5ad..a27c40626 100644 --- a/src/lapack/lapack_loader.cpp +++ b/src/lapack/lapack_loader.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/lapack/detail/lapack_loader.hpp" +#include "oneapi/math/lapack/detail/lapack_loader.hpp" #include "function_table_initializer.hpp" #include "lapack/function_table.hpp" diff --git a/src/rng/backends/curand/CMakeLists.txt b/src/rng/backends/curand/CMakeLists.txt index f37a34f1d..f9fbf8168 100644 --- a/src/rng/backends/curand/CMakeLists.txt +++ b/src/rng/backends/curand/CMakeLists.txt @@ -62,7 +62,7 @@ find_package(cuRAND REQUIRED) set(SOURCES philox4x32x10.cpp mrg32k3a.cpp - $<$: mkl_rng_curand_wrappers.cpp>) + $<$: curand_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) diff --git a/src/rng/backends/curand/curand_helper.hpp b/src/rng/backends/curand/curand_helper.hpp index 3926e6283..b38946260 100644 --- a/src/rng/backends/curand/curand_helper.hpp +++ b/src/rng/backends/curand/curand_helper.hpp @@ -67,7 +67,7 @@ #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" namespace oneapi { namespace mkl { diff --git a/src/rng/backends/curand/mkl_rng_curand_wrappers.cpp b/src/rng/backends/curand/curand_wrappers.cpp similarity index 98% rename from src/rng/backends/curand/mkl_rng_curand_wrappers.cpp rename to src/rng/backends/curand/curand_wrappers.cpp index 393433c81..f46227344 100644 --- a/src/rng/backends/curand/mkl_rng_curand_wrappers.cpp +++ b/src/rng/backends/curand/curand_wrappers.cpp @@ -56,7 +56,7 @@ * so. ******************************************************************************/ -#include "oneapi/mkl/rng/detail/curand/onemkl_rng_curand.hpp" +#include "oneapi/math/rng/detail/curand/onemkl_rng_curand.hpp" #include "rng/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/rng/backends/curand/mrg32k3a.cpp b/src/rng/backends/curand/mrg32k3a.cpp index dd44f4def..2f761f95e 100644 --- a/src/rng/backends/curand/mrg32k3a.cpp +++ b/src/rng/backends/curand/mrg32k3a.cpp @@ -74,10 +74,10 @@ #include "curand_helper.hpp" #include "curand_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/rng/detail/curand/onemkl_rng_curand.hpp" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" -#include "oneapi/mkl/rng/engines.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/rng/detail/curand/onemkl_rng_curand.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" +#include "oneapi/math/rng/engines.hpp" namespace oneapi { namespace mkl { diff --git a/src/rng/backends/curand/philox4x32x10.cpp b/src/rng/backends/curand/philox4x32x10.cpp index c3d4393d2..4a0fcd476 100644 --- a/src/rng/backends/curand/philox4x32x10.cpp +++ b/src/rng/backends/curand/philox4x32x10.cpp @@ -72,12 +72,12 @@ #endif #include -#include "oneapi/mkl/rng/detail/engine_impl.hpp" -// #include "oneapi/mkl/rng/engines.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" +// #include "oneapi/math/rng/engines.hpp" #include "curand_helper.hpp" #include "curand_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/rng/detail/curand/onemkl_rng_curand.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/rng/detail/curand/onemkl_rng_curand.hpp" namespace oneapi { namespace mkl { diff --git a/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp b/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp index 840205db7..bfc701780 100644 --- a/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp +++ b/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "rng/function_table.hpp" -#include "oneapi/mkl/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp" +#include "oneapi/math/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/rng/backends/mklcpu/mrg32k3a.cpp b/src/rng/backends/mklcpu/mrg32k3a.cpp index cc234de45..ca31f7409 100644 --- a/src/rng/backends/mklcpu/mrg32k3a.cpp +++ b/src/rng/backends/mklcpu/mrg32k3a.cpp @@ -26,9 +26,9 @@ #include "mkl_vsl.h" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" -#include "oneapi/mkl/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" +#include "oneapi/math/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp" #include "cpu_common.hpp" diff --git a/src/rng/backends/mklcpu/philox4x32x10.cpp b/src/rng/backends/mklcpu/philox4x32x10.cpp index 3f8e5e89b..f9af3597a 100644 --- a/src/rng/backends/mklcpu/philox4x32x10.cpp +++ b/src/rng/backends/mklcpu/philox4x32x10.cpp @@ -26,9 +26,9 @@ #include "mkl_vsl.h" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" -#include "oneapi/mkl/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" +#include "oneapi/math/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp" #include "cpu_common.hpp" diff --git a/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp b/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp index 6754b2749..0bd5dcf53 100644 --- a/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp +++ b/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "rng/function_table.hpp" -#include "oneapi/mkl/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp" +#include "oneapi/math/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/rng/backends/mklgpu/mrg32k3a.cpp b/src/rng/backends/mklgpu/mrg32k3a.cpp index 05d24e1a4..f603b69b0 100644 --- a/src/rng/backends/mklgpu/mrg32k3a.cpp +++ b/src/rng/backends/mklgpu/mrg32k3a.cpp @@ -26,10 +26,10 @@ #include "mkl_version.h" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" -#include "oneapi/mkl/rng/engines.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" +#include "oneapi/math/rng/engines.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp" #include "mkl_internal_rng_gpu.hpp" diff --git a/src/rng/backends/mklgpu/philox4x32x10.cpp b/src/rng/backends/mklgpu/philox4x32x10.cpp index bcf869c61..25e318f90 100644 --- a/src/rng/backends/mklgpu/philox4x32x10.cpp +++ b/src/rng/backends/mklgpu/philox4x32x10.cpp @@ -26,10 +26,10 @@ #include "mkl_version.h" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" -#include "oneapi/mkl/rng/engines.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" +#include "oneapi/math/rng/engines.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp" #include "mkl_internal_rng_gpu.hpp" diff --git a/src/rng/backends/rocrand/CMakeLists.txt b/src/rng/backends/rocrand/CMakeLists.txt index 47929703b..509996611 100644 --- a/src/rng/backends/rocrand/CMakeLists.txt +++ b/src/rng/backends/rocrand/CMakeLists.txt @@ -59,7 +59,7 @@ find_package(rocrand REQUIRED) find_package(Threads REQUIRED) set(SOURCES philox4x32x10.cpp mrg32k3a.cpp $<$: - mkl_rng_rocrand_wrappers.cpp>) + rocrand_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) diff --git a/src/rng/backends/rocrand/mrg32k3a.cpp b/src/rng/backends/rocrand/mrg32k3a.cpp index 424f14caf..d0c2fef59 100644 --- a/src/rng/backends/rocrand/mrg32k3a.cpp +++ b/src/rng/backends/rocrand/mrg32k3a.cpp @@ -74,10 +74,10 @@ #include "rocrand_helper.hpp" #include "rocrand_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/rng/detail/rocrand/onemkl_rng_rocrand.hpp" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" -#include "oneapi/mkl/rng/engines.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/rng/detail/rocrand/onemkl_rng_rocrand.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" +#include "oneapi/math/rng/engines.hpp" namespace oneapi { namespace mkl { diff --git a/src/rng/backends/rocrand/philox4x32x10.cpp b/src/rng/backends/rocrand/philox4x32x10.cpp index 5bc241360..2e21534a1 100644 --- a/src/rng/backends/rocrand/philox4x32x10.cpp +++ b/src/rng/backends/rocrand/philox4x32x10.cpp @@ -74,10 +74,10 @@ #include "rocrand_helper.hpp" #include "rocrand_task.hpp" -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/rng/detail/rocrand/onemkl_rng_rocrand.hpp" -#include "oneapi/mkl/rng/detail/engine_impl.hpp" -// #include "oneapi/mkl/rng/engines.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/rng/detail/rocrand/onemkl_rng_rocrand.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" +// #include "oneapi/math/rng/engines.hpp" namespace oneapi { namespace mkl { diff --git a/src/rng/backends/rocrand/rocrand_helper.hpp b/src/rng/backends/rocrand/rocrand_helper.hpp index 205429ee8..e1d51dee3 100644 --- a/src/rng/backends/rocrand/rocrand_helper.hpp +++ b/src/rng/backends/rocrand/rocrand_helper.hpp @@ -67,7 +67,7 @@ #include #include -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" namespace oneapi { namespace mkl { diff --git a/src/rng/backends/rocrand/mkl_rng_rocrand_wrappers.cpp b/src/rng/backends/rocrand/rocrand_wrappers.cpp similarity index 98% rename from src/rng/backends/rocrand/mkl_rng_rocrand_wrappers.cpp rename to src/rng/backends/rocrand/rocrand_wrappers.cpp index 5450f47b4..bdbfa6fb1 100644 --- a/src/rng/backends/rocrand/mkl_rng_rocrand_wrappers.cpp +++ b/src/rng/backends/rocrand/rocrand_wrappers.cpp @@ -58,7 +58,7 @@ * so. ******************************************************************************/ -#include "oneapi/mkl/rng/detail/rocrand/onemkl_rng_rocrand.hpp" +#include "oneapi/math/rng/detail/rocrand/onemkl_rng_rocrand.hpp" #include "rng/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/rng/function_table.hpp b/src/rng/function_table.hpp index c94757250..0708d5fbb 100644 --- a/src/rng/function_table.hpp +++ b/src/rng/function_table.hpp @@ -27,7 +27,7 @@ #include #endif -#include "oneapi/mkl/rng/detail/engine_impl.hpp" +#include "oneapi/math/rng/detail/engine_impl.hpp" typedef struct { int version; diff --git a/src/rng/rng_loader.cpp b/src/rng/rng_loader.cpp index 68e3a5ba5..fa969bb3c 100644 --- a/src/rng/rng_loader.cpp +++ b/src/rng/rng_loader.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/mkl/rng/detail/rng_loader.hpp" +#include "oneapi/math/rng/detail/rng_loader.hpp" #include "function_table_initializer.hpp" #include "rng/function_table.hpp" diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp index 44b12e8df..feaf0fdfc 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp @@ -23,7 +23,7 @@ // MKLCPU and MKLGPU backends include // This include defines its own oneapi::mkl::sparse namespace with some of the // types that are used here: matrix_handle_t, index_base, transpose, uplo, diag. -#include +#include #include "sparse_blas/generic_container.hpp" diff --git a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp index ca15c5b4f..42d8abbb0 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp @@ -26,8 +26,8 @@ #include #endif -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/sparse_blas/detail/helper_types.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/sparse_blas/detail/helper_types.hpp" #include "sparse_blas/enum_data_types.hpp" #include "sparse_blas/macros.hpp" diff --git a/src/sparse_blas/backends/mklcpu/mklcpu_handles.cpp b/src/sparse_blas/backends/mklcpu/mklcpu_handles.cpp index a6ea51629..526fd45a9 100644 --- a/src/sparse_blas/backends/mklcpu/mklcpu_handles.cpp +++ b/src/sparse_blas/backends/mklcpu/mklcpu_handles.cpp @@ -17,7 +17,7 @@ * **************************************************************************/ -#include "oneapi/mkl/sparse_blas/detail/mklcpu/onemkl_sparse_blas_mklcpu.hpp" +#include "oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp" #include "sparse_blas/backends/mkl_common/mkl_helper.hpp" #include "sparse_blas/backends/mkl_common/mkl_handles.hpp" diff --git a/src/sparse_blas/backends/mklcpu/mklcpu_operations.cpp b/src/sparse_blas/backends/mklcpu/mklcpu_operations.cpp index 0929a7ef4..a5bb1043c 100644 --- a/src/sparse_blas/backends/mklcpu/mklcpu_operations.cpp +++ b/src/sparse_blas/backends/mklcpu/mklcpu_operations.cpp @@ -22,7 +22,7 @@ #include "sparse_blas/macros.hpp" #include "sparse_blas/matrix_view_comparison.hpp" -#include "oneapi/mkl/sparse_blas/detail/mklcpu/onemkl_sparse_blas_mklcpu.hpp" +#include "oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp" #define BACKEND mklcpu diff --git a/src/sparse_blas/backends/mklcpu/mklcpu_wrappers.cpp b/src/sparse_blas/backends/mklcpu/mklcpu_wrappers.cpp index 1a6217684..d67d1d239 100644 --- a/src/sparse_blas/backends/mklcpu/mklcpu_wrappers.cpp +++ b/src/sparse_blas/backends/mklcpu/mklcpu_wrappers.cpp @@ -17,9 +17,9 @@ * **************************************************************************/ -#include "oneapi/mkl/sparse_blas/types.hpp" +#include "oneapi/math/sparse_blas/types.hpp" -#include "oneapi/mkl/sparse_blas/detail/mklcpu/onemkl_sparse_blas_mklcpu.hpp" +#include "oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp" #include "sparse_blas/function_table.hpp" diff --git a/src/sparse_blas/backends/mklgpu/mklgpu_handles.cpp b/src/sparse_blas/backends/mklgpu/mklgpu_handles.cpp index 7cb9853a7..c5b33a7da 100644 --- a/src/sparse_blas/backends/mklgpu/mklgpu_handles.cpp +++ b/src/sparse_blas/backends/mklgpu/mklgpu_handles.cpp @@ -17,7 +17,7 @@ * **************************************************************************/ -#include "oneapi/mkl/sparse_blas/detail/mklgpu/onemkl_sparse_blas_mklgpu.hpp" +#include "oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp" #include "sparse_blas/backends/mkl_common/mkl_handles.hpp" #include "sparse_blas/backends/mkl_common/mkl_helper.hpp" diff --git a/src/sparse_blas/backends/mklgpu/mklgpu_operations.cpp b/src/sparse_blas/backends/mklgpu/mklgpu_operations.cpp index be5e0c0aa..bfdb68a1f 100644 --- a/src/sparse_blas/backends/mklgpu/mklgpu_operations.cpp +++ b/src/sparse_blas/backends/mklgpu/mklgpu_operations.cpp @@ -22,7 +22,7 @@ #include "sparse_blas/macros.hpp" #include "sparse_blas/matrix_view_comparison.hpp" -#include "oneapi/mkl/sparse_blas/detail/mklgpu/onemkl_sparse_blas_mklgpu.hpp" +#include "oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp" #define BACKEND mklgpu diff --git a/src/sparse_blas/backends/mklgpu/mklgpu_wrappers.cpp b/src/sparse_blas/backends/mklgpu/mklgpu_wrappers.cpp index 4a261f64e..17e433663 100644 --- a/src/sparse_blas/backends/mklgpu/mklgpu_wrappers.cpp +++ b/src/sparse_blas/backends/mklgpu/mklgpu_wrappers.cpp @@ -17,9 +17,9 @@ * **************************************************************************/ -#include "oneapi/mkl/sparse_blas/types.hpp" +#include "oneapi/math/sparse_blas/types.hpp" -#include "oneapi/mkl/sparse_blas/detail/mklgpu/onemkl_sparse_blas_mklgpu.hpp" +#include "oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp" #include "sparse_blas/function_table.hpp" diff --git a/src/sparse_blas/function_table.hpp b/src/sparse_blas/function_table.hpp index d1e3d8189..e48f50ccc 100644 --- a/src/sparse_blas/function_table.hpp +++ b/src/sparse_blas/function_table.hpp @@ -20,7 +20,7 @@ #ifndef _ONEMKL_SPARSE_BLAS_FUNCTION_TABLE_HPP_ #define _ONEMKL_SPARSE_BLAS_FUNCTION_TABLE_HPP_ -#include "oneapi/mkl/sparse_blas/types.hpp" +#include "oneapi/math/sparse_blas/types.hpp" #include "sparse_blas/macros.hpp" // Dense vector diff --git a/src/sparse_blas/generic_container.hpp b/src/sparse_blas/generic_container.hpp index 53bd50837..7e3d60359 100644 --- a/src/sparse_blas/generic_container.hpp +++ b/src/sparse_blas/generic_container.hpp @@ -29,7 +29,7 @@ #include #endif -#include "oneapi/mkl/sparse_blas/types.hpp" +#include "oneapi/math/sparse_blas/types.hpp" #include "enum_data_types.hpp" namespace oneapi::mkl::sparse::detail { diff --git a/src/sparse_blas/matrix_view_comparison.hpp b/src/sparse_blas/matrix_view_comparison.hpp index e01be7311..31d2193b3 100644 --- a/src/sparse_blas/matrix_view_comparison.hpp +++ b/src/sparse_blas/matrix_view_comparison.hpp @@ -20,7 +20,7 @@ #ifndef _ONEMKL_SRC_SPARSE_BLAS_MATRIX_VIEW_COMPARISON_HPP_ #define _ONEMKL_SRC_SPARSE_BLAS_MATRIX_VIEW_COMPARISON_HPP_ -#include "oneapi/mkl/sparse_blas/matrix_view.hpp" +#include "oneapi/math/sparse_blas/matrix_view.hpp" inline bool operator==(const oneapi::mkl::sparse::matrix_view& lhs, const oneapi::mkl::sparse::matrix_view& rhs) { diff --git a/src/sparse_blas/sparse_blas_loader.cpp b/src/sparse_blas/sparse_blas_loader.cpp index 4304479d1..2cf6fba74 100644 --- a/src/sparse_blas/sparse_blas_loader.cpp +++ b/src/sparse_blas/sparse_blas_loader.cpp @@ -17,12 +17,12 @@ * **************************************************************************/ -#include "oneapi/mkl/sparse_blas/detail/sparse_blas_rt.hpp" +#include "oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp" #include "function_table_initializer.hpp" #include "sparse_blas/function_table.hpp" #include "sparse_blas/macros.hpp" -#include "oneapi/mkl/detail/get_device_id.hpp" +#include "oneapi/math/detail/get_device_id.hpp" namespace oneapi::mkl::sparse { diff --git a/tests/unit_tests/blas/batch/axpy_batch_stride.cpp b/tests/unit_tests/blas/batch/axpy_batch_stride.cpp index 9bb1406ef..b49305662 100644 --- a/tests/unit_tests/blas/batch/axpy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/axpy_batch_stride.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp index 9ebc82abe..7fbde684d 100644 --- a/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/axpy_batch_usm.cpp b/tests/unit_tests/blas/batch/axpy_batch_usm.cpp index 4dacf8ddb..c511789fa 100644 --- a/tests/unit_tests/blas/batch/axpy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/axpy_batch_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/copy_batch_stride.cpp b/tests/unit_tests/blas/batch/copy_batch_stride.cpp index a1da595f6..a78663d8a 100644 --- a/tests/unit_tests/blas/batch/copy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/copy_batch_stride.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp index 569293be1..dd1a7a9f3 100644 --- a/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/copy_batch_usm.cpp b/tests/unit_tests/blas/batch/copy_batch_usm.cpp index 8cac23704..ef1a30d27 100644 --- a/tests/unit_tests/blas/batch/copy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/copy_batch_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp b/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp index bb642c3ee..16d335fcb 100644 --- a/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp index bb9cf0df3..64843b7d9 100644 --- a/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp b/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp index 1f568580f..18234d879 100644 --- a/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp @@ -30,10 +30,10 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" #include "allocator_helper.hpp" -#include "onemkl_blas_helper.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/gemm_batch_stride.cpp b/tests/unit_tests/blas/batch/gemm_batch_stride.cpp index 5241cb822..537d39ee1 100644 --- a/tests/unit_tests/blas/batch/gemm_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/gemm_batch_stride.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp index 97f2dd086..b4f10dc7d 100644 --- a/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/gemm_batch_usm.cpp b/tests/unit_tests/blas/batch/gemm_batch_usm.cpp index a651f9ae3..fa7ee938d 100644 --- a/tests/unit_tests/blas/batch/gemm_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/gemm_batch_usm.cpp @@ -30,10 +30,10 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" #include "allocator_helper.hpp" -#include "onemkl_blas_helper.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/gemv_batch_stride.cpp b/tests/unit_tests/blas/batch/gemv_batch_stride.cpp index bd92f70ca..98677a0bd 100644 --- a/tests/unit_tests/blas/batch/gemv_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/gemv_batch_stride.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp index d6eb47887..0539f7eac 100644 --- a/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/gemv_batch_usm.cpp b/tests/unit_tests/blas/batch/gemv_batch_usm.cpp index 4ad661f5b..f239c17cf 100644 --- a/tests/unit_tests/blas/batch/gemv_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/gemv_batch_usm.cpp @@ -30,10 +30,10 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" #include "allocator_helper.hpp" -#include "onemkl_blas_helper.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp b/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp index ac8bbb2b4..fa287d187 100644 --- a/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp index b3099d309..ab9fb95f4 100644 --- a/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp b/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp index 74c9881af..a4b8093c9 100644 --- a/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp @@ -32,9 +32,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp b/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp index cc20d0e3b..4bd5fcce3 100644 --- a/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp index 7388084cb..de6267978 100644 --- a/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp b/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp index d08329fc6..bd024bbc1 100644 --- a/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp index 7479b57db..2fc0a32a0 100644 --- a/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp @@ -32,9 +32,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp b/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp index 7f1e4a103..e1c0b8a3c 100644 --- a/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp @@ -32,9 +32,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/syrk_batch_stride.cpp b/tests/unit_tests/blas/batch/syrk_batch_stride.cpp index 58dc4d7dc..4b7b8d56d 100644 --- a/tests/unit_tests/blas/batch/syrk_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/syrk_batch_stride.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp index 31aa09b79..bd11b1765 100644 --- a/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/syrk_batch_usm.cpp b/tests/unit_tests/blas/batch/syrk_batch_usm.cpp index 36d0d6dd5..d308f98ca 100644 --- a/tests/unit_tests/blas/batch/syrk_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/syrk_batch_usm.cpp @@ -30,10 +30,10 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" #include "allocator_helper.hpp" -#include "onemkl_blas_helper.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/trsm_batch_stride.cpp b/tests/unit_tests/blas/batch/trsm_batch_stride.cpp index cde6aa367..fcafbf4d6 100644 --- a/tests/unit_tests/blas/batch/trsm_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/trsm_batch_stride.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp index d99836f87..4dde42ef7 100644 --- a/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/batch/trsm_batch_usm.cpp b/tests/unit_tests/blas/batch/trsm_batch_usm.cpp index 747f59433..1e3d2980d 100644 --- a/tests/unit_tests/blas/batch/trsm_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/trsm_batch_usm.cpp @@ -30,10 +30,10 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" #include "allocator_helper.hpp" -#include "onemkl_blas_helper.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/gemm_bias.cpp b/tests/unit_tests/blas/extensions/gemm_bias.cpp index c6e99e829..f08015827 100644 --- a/tests/unit_tests/blas/extensions/gemm_bias.cpp +++ b/tests/unit_tests/blas/extensions/gemm_bias.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp b/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp index 908eed909..b223c5700 100644 --- a/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp +++ b/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/gemmt.cpp b/tests/unit_tests/blas/extensions/gemmt.cpp index 228a85d33..f83e2d1b0 100644 --- a/tests/unit_tests/blas/extensions/gemmt.cpp +++ b/tests/unit_tests/blas/extensions/gemmt.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/gemmt_usm.cpp b/tests/unit_tests/blas/extensions/gemmt_usm.cpp index dac300ae2..1a3b3e450 100644 --- a/tests/unit_tests/blas/extensions/gemmt_usm.cpp +++ b/tests/unit_tests/blas/extensions/gemmt_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/imatcopy.cpp b/tests/unit_tests/blas/extensions/imatcopy.cpp index e21702775..739612791 100644 --- a/tests/unit_tests/blas/extensions/imatcopy.cpp +++ b/tests/unit_tests/blas/extensions/imatcopy.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/imatcopy_usm.cpp b/tests/unit_tests/blas/extensions/imatcopy_usm.cpp index dc3d43d2e..a8f3745bf 100644 --- a/tests/unit_tests/blas/extensions/imatcopy_usm.cpp +++ b/tests/unit_tests/blas/extensions/imatcopy_usm.cpp @@ -32,9 +32,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/omatadd.cpp b/tests/unit_tests/blas/extensions/omatadd.cpp index b2af98935..9cebed350 100644 --- a/tests/unit_tests/blas/extensions/omatadd.cpp +++ b/tests/unit_tests/blas/extensions/omatadd.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/omatadd_usm.cpp b/tests/unit_tests/blas/extensions/omatadd_usm.cpp index 783f985b2..1cd838c17 100644 --- a/tests/unit_tests/blas/extensions/omatadd_usm.cpp +++ b/tests/unit_tests/blas/extensions/omatadd_usm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/omatcopy.cpp b/tests/unit_tests/blas/extensions/omatcopy.cpp index 122ba2c79..022cf2e63 100644 --- a/tests/unit_tests/blas/extensions/omatcopy.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/omatcopy2.cpp b/tests/unit_tests/blas/extensions/omatcopy2.cpp index d0407c324..93038c63b 100644 --- a/tests/unit_tests/blas/extensions/omatcopy2.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy2.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp b/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp index d2103d243..1e880cf5b 100644 --- a/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp @@ -32,9 +32,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/extensions/omatcopy_usm.cpp b/tests/unit_tests/blas/extensions/omatcopy_usm.cpp index ac9ba2d5c..beddf9c5d 100644 --- a/tests/unit_tests/blas/extensions/omatcopy_usm.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy_usm.cpp @@ -32,9 +32,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/include/onemkl_blas_helper.hpp b/tests/unit_tests/blas/include/onemath_blas_helper.hpp similarity index 95% rename from tests/unit_tests/blas/include/onemkl_blas_helper.hpp rename to tests/unit_tests/blas/include/onemath_blas_helper.hpp index 5489aaa61..30645840f 100644 --- a/tests/unit_tests/blas/include/onemkl_blas_helper.hpp +++ b/tests/unit_tests/blas/include/onemath_blas_helper.hpp @@ -17,12 +17,12 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef ONEMKL_BLAS_HELPER_HPP -#define ONEMKL_BLAS_HELPER_HPP +#ifndef ONEMATH_BLAS_HELPER_HPP +#define ONEMATH_BLAS_HELPER_HPP #include "cblas.h" -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" typedef enum { CblasRowOffset = 101, CblasColOffset = 102, CblasFixOffset = 103 } CBLAS_OFFSET; @@ -82,4 +82,4 @@ static const CBLAS_TRANSPOSE fcblastrans_c[] = { CblasConjTrans, CblasNoTrans, C static const CBLAS_OFFSET fcblasoffset[] = { CblasColOffset, CblasRowOffset, CblasFixOffset }; -#endif // ONEMKL_BLAS_HELPER_HPP +#endif // ONEMATH_BLAS_HELPER_HPP diff --git a/tests/unit_tests/blas/include/reference_blas_templates.hpp b/tests/unit_tests/blas/include/reference_blas_templates.hpp index 6d184ba75..182fcfad6 100644 --- a/tests/unit_tests/blas/include/reference_blas_templates.hpp +++ b/tests/unit_tests/blas/include/reference_blas_templates.hpp @@ -24,7 +24,7 @@ #include #include #include "cblas.h" -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" #include "test_helper.hpp" #include "reference_blas_wrappers.hpp" diff --git a/tests/unit_tests/blas/include/reference_blas_wrappers.hpp b/tests/unit_tests/blas/include/reference_blas_wrappers.hpp index 977d14ca4..e5b4c3a7d 100644 --- a/tests/unit_tests/blas/include/reference_blas_wrappers.hpp +++ b/tests/unit_tests/blas/include/reference_blas_wrappers.hpp @@ -20,8 +20,8 @@ #ifndef _REFERENCE_BLAS_WRAPPERS_HPP__ #define _REFERENCE_BLAS_WRAPPERS_HPP__ -#include "oneapi/mkl/exceptions.hpp" -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/detail/config.hpp" #include #include "cblas.h" diff --git a/tests/unit_tests/blas/level1/asum.cpp b/tests/unit_tests/blas/level1/asum.cpp index 6969789e3..d0a9869d6 100644 --- a/tests/unit_tests/blas/level1/asum.cpp +++ b/tests/unit_tests/blas/level1/asum.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/asum_usm.cpp b/tests/unit_tests/blas/level1/asum_usm.cpp index b42799abd..076de8f0b 100644 --- a/tests/unit_tests/blas/level1/asum_usm.cpp +++ b/tests/unit_tests/blas/level1/asum_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/axpby.cpp b/tests/unit_tests/blas/level1/axpby.cpp index d43f9beda..df3550efe 100644 --- a/tests/unit_tests/blas/level1/axpby.cpp +++ b/tests/unit_tests/blas/level1/axpby.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/axpby_usm.cpp b/tests/unit_tests/blas/level1/axpby_usm.cpp index ae85ca8f1..348731324 100644 --- a/tests/unit_tests/blas/level1/axpby_usm.cpp +++ b/tests/unit_tests/blas/level1/axpby_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/axpy.cpp b/tests/unit_tests/blas/level1/axpy.cpp index c81f2902d..ebcb6476c 100644 --- a/tests/unit_tests/blas/level1/axpy.cpp +++ b/tests/unit_tests/blas/level1/axpy.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/axpy_usm.cpp b/tests/unit_tests/blas/level1/axpy_usm.cpp index da68f173c..5bcf9874d 100644 --- a/tests/unit_tests/blas/level1/axpy_usm.cpp +++ b/tests/unit_tests/blas/level1/axpy_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/copy.cpp b/tests/unit_tests/blas/level1/copy.cpp index 87a1c2f1b..f5e2c81a6 100644 --- a/tests/unit_tests/blas/level1/copy.cpp +++ b/tests/unit_tests/blas/level1/copy.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/copy_usm.cpp b/tests/unit_tests/blas/level1/copy_usm.cpp index 0f491015b..ac457acf9 100644 --- a/tests/unit_tests/blas/level1/copy_usm.cpp +++ b/tests/unit_tests/blas/level1/copy_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/dot.cpp b/tests/unit_tests/blas/level1/dot.cpp index 11cb09bcc..0ac335641 100644 --- a/tests/unit_tests/blas/level1/dot.cpp +++ b/tests/unit_tests/blas/level1/dot.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/dot_usm.cpp b/tests/unit_tests/blas/level1/dot_usm.cpp index b8780c75d..94ad1c008 100644 --- a/tests/unit_tests/blas/level1/dot_usm.cpp +++ b/tests/unit_tests/blas/level1/dot_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/dotc.cpp b/tests/unit_tests/blas/level1/dotc.cpp index cb8d0fc37..544e007ea 100644 --- a/tests/unit_tests/blas/level1/dotc.cpp +++ b/tests/unit_tests/blas/level1/dotc.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/dotc_usm.cpp b/tests/unit_tests/blas/level1/dotc_usm.cpp index ad05c9d3b..f24b5e222 100644 --- a/tests/unit_tests/blas/level1/dotc_usm.cpp +++ b/tests/unit_tests/blas/level1/dotc_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/dotu.cpp b/tests/unit_tests/blas/level1/dotu.cpp index bbef3ad8c..6f3163da5 100644 --- a/tests/unit_tests/blas/level1/dotu.cpp +++ b/tests/unit_tests/blas/level1/dotu.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/dotu_usm.cpp b/tests/unit_tests/blas/level1/dotu_usm.cpp index 3f30bf5ff..c67871c2c 100644 --- a/tests/unit_tests/blas/level1/dotu_usm.cpp +++ b/tests/unit_tests/blas/level1/dotu_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/iamax.cpp b/tests/unit_tests/blas/level1/iamax.cpp index 977f12b5d..41d79748c 100644 --- a/tests/unit_tests/blas/level1/iamax.cpp +++ b/tests/unit_tests/blas/level1/iamax.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/iamax_usm.cpp b/tests/unit_tests/blas/level1/iamax_usm.cpp index 405a79532..86e8ab9fa 100644 --- a/tests/unit_tests/blas/level1/iamax_usm.cpp +++ b/tests/unit_tests/blas/level1/iamax_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/iamin.cpp b/tests/unit_tests/blas/level1/iamin.cpp index a52862cb6..147ab4f51 100644 --- a/tests/unit_tests/blas/level1/iamin.cpp +++ b/tests/unit_tests/blas/level1/iamin.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/iamin_usm.cpp b/tests/unit_tests/blas/level1/iamin_usm.cpp index a3523c8e7..437d60aed 100644 --- a/tests/unit_tests/blas/level1/iamin_usm.cpp +++ b/tests/unit_tests/blas/level1/iamin_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/nrm2.cpp b/tests/unit_tests/blas/level1/nrm2.cpp index 423cecb59..f9b905421 100644 --- a/tests/unit_tests/blas/level1/nrm2.cpp +++ b/tests/unit_tests/blas/level1/nrm2.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/nrm2_usm.cpp b/tests/unit_tests/blas/level1/nrm2_usm.cpp index 8628738f4..065b784e1 100644 --- a/tests/unit_tests/blas/level1/nrm2_usm.cpp +++ b/tests/unit_tests/blas/level1/nrm2_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/rot.cpp b/tests/unit_tests/blas/level1/rot.cpp index f65540182..e8cdbe993 100644 --- a/tests/unit_tests/blas/level1/rot.cpp +++ b/tests/unit_tests/blas/level1/rot.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/rot_usm.cpp b/tests/unit_tests/blas/level1/rot_usm.cpp index 287ac285b..e54bb483b 100644 --- a/tests/unit_tests/blas/level1/rot_usm.cpp +++ b/tests/unit_tests/blas/level1/rot_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/rotg.cpp b/tests/unit_tests/blas/level1/rotg.cpp index 1a0d569d8..289287d2d 100644 --- a/tests/unit_tests/blas/level1/rotg.cpp +++ b/tests/unit_tests/blas/level1/rotg.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/rotg_usm.cpp b/tests/unit_tests/blas/level1/rotg_usm.cpp index de71a793d..e4c38ffe5 100644 --- a/tests/unit_tests/blas/level1/rotg_usm.cpp +++ b/tests/unit_tests/blas/level1/rotg_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/rotm.cpp b/tests/unit_tests/blas/level1/rotm.cpp index ab2c599bf..31cbd54d2 100644 --- a/tests/unit_tests/blas/level1/rotm.cpp +++ b/tests/unit_tests/blas/level1/rotm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/rotm_usm.cpp b/tests/unit_tests/blas/level1/rotm_usm.cpp index 7723e096c..02472bce1 100644 --- a/tests/unit_tests/blas/level1/rotm_usm.cpp +++ b/tests/unit_tests/blas/level1/rotm_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/rotmg.cpp b/tests/unit_tests/blas/level1/rotmg.cpp index f62bd1cf9..bb30a7a32 100644 --- a/tests/unit_tests/blas/level1/rotmg.cpp +++ b/tests/unit_tests/blas/level1/rotmg.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/rotmg_usm.cpp b/tests/unit_tests/blas/level1/rotmg_usm.cpp index 92eeee491..3da6cdbf8 100644 --- a/tests/unit_tests/blas/level1/rotmg_usm.cpp +++ b/tests/unit_tests/blas/level1/rotmg_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/scal.cpp b/tests/unit_tests/blas/level1/scal.cpp index 8901bb424..5c4f8ee35 100644 --- a/tests/unit_tests/blas/level1/scal.cpp +++ b/tests/unit_tests/blas/level1/scal.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/scal_usm.cpp b/tests/unit_tests/blas/level1/scal_usm.cpp index e669deb2d..83a936717 100644 --- a/tests/unit_tests/blas/level1/scal_usm.cpp +++ b/tests/unit_tests/blas/level1/scal_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/sdsdot.cpp b/tests/unit_tests/blas/level1/sdsdot.cpp index 7293a3699..90bc2d767 100644 --- a/tests/unit_tests/blas/level1/sdsdot.cpp +++ b/tests/unit_tests/blas/level1/sdsdot.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/sdsdot_usm.cpp b/tests/unit_tests/blas/level1/sdsdot_usm.cpp index a5740516c..fc4f9189d 100644 --- a/tests/unit_tests/blas/level1/sdsdot_usm.cpp +++ b/tests/unit_tests/blas/level1/sdsdot_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/swap.cpp b/tests/unit_tests/blas/level1/swap.cpp index 6c6721537..a83400233 100644 --- a/tests/unit_tests/blas/level1/swap.cpp +++ b/tests/unit_tests/blas/level1/swap.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level1/swap_usm.cpp b/tests/unit_tests/blas/level1/swap_usm.cpp index de20f3eb7..4a5c4f375 100644 --- a/tests/unit_tests/blas/level1/swap_usm.cpp +++ b/tests/unit_tests/blas/level1/swap_usm.cpp @@ -29,9 +29,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/gbmv.cpp b/tests/unit_tests/blas/level2/gbmv.cpp index 94fcbc906..8a532ef39 100644 --- a/tests/unit_tests/blas/level2/gbmv.cpp +++ b/tests/unit_tests/blas/level2/gbmv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/gbmv_usm.cpp b/tests/unit_tests/blas/level2/gbmv_usm.cpp index 9d92fcf7e..36f412340 100644 --- a/tests/unit_tests/blas/level2/gbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/gbmv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/gemv.cpp b/tests/unit_tests/blas/level2/gemv.cpp index 3bfff4324..33681d68f 100644 --- a/tests/unit_tests/blas/level2/gemv.cpp +++ b/tests/unit_tests/blas/level2/gemv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/gemv_usm.cpp b/tests/unit_tests/blas/level2/gemv_usm.cpp index d1e726e38..f11815bb8 100644 --- a/tests/unit_tests/blas/level2/gemv_usm.cpp +++ b/tests/unit_tests/blas/level2/gemv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/ger.cpp b/tests/unit_tests/blas/level2/ger.cpp index 3b32d2827..62341bf85 100644 --- a/tests/unit_tests/blas/level2/ger.cpp +++ b/tests/unit_tests/blas/level2/ger.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/ger_usm.cpp b/tests/unit_tests/blas/level2/ger_usm.cpp index 87087f026..26f3b7c53 100644 --- a/tests/unit_tests/blas/level2/ger_usm.cpp +++ b/tests/unit_tests/blas/level2/ger_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/gerc.cpp b/tests/unit_tests/blas/level2/gerc.cpp index c19c9f029..45b42adcf 100644 --- a/tests/unit_tests/blas/level2/gerc.cpp +++ b/tests/unit_tests/blas/level2/gerc.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/gerc_usm.cpp b/tests/unit_tests/blas/level2/gerc_usm.cpp index b6473484d..5691e74cb 100644 --- a/tests/unit_tests/blas/level2/gerc_usm.cpp +++ b/tests/unit_tests/blas/level2/gerc_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/geru.cpp b/tests/unit_tests/blas/level2/geru.cpp index e0cb7c45d..333826b5a 100644 --- a/tests/unit_tests/blas/level2/geru.cpp +++ b/tests/unit_tests/blas/level2/geru.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/geru_usm.cpp b/tests/unit_tests/blas/level2/geru_usm.cpp index 1e882bd97..4748cc995 100644 --- a/tests/unit_tests/blas/level2/geru_usm.cpp +++ b/tests/unit_tests/blas/level2/geru_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/hbmv.cpp b/tests/unit_tests/blas/level2/hbmv.cpp index 119aef32a..efc9b8c17 100644 --- a/tests/unit_tests/blas/level2/hbmv.cpp +++ b/tests/unit_tests/blas/level2/hbmv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/hbmv_usm.cpp b/tests/unit_tests/blas/level2/hbmv_usm.cpp index 60305cb93..ef60731c4 100644 --- a/tests/unit_tests/blas/level2/hbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/hbmv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/hemv.cpp b/tests/unit_tests/blas/level2/hemv.cpp index 3636e3774..b0dab2e95 100644 --- a/tests/unit_tests/blas/level2/hemv.cpp +++ b/tests/unit_tests/blas/level2/hemv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/hemv_usm.cpp b/tests/unit_tests/blas/level2/hemv_usm.cpp index a1b8093fc..d78935e71 100644 --- a/tests/unit_tests/blas/level2/hemv_usm.cpp +++ b/tests/unit_tests/blas/level2/hemv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/her.cpp b/tests/unit_tests/blas/level2/her.cpp index 46ae9a879..951f37a6d 100644 --- a/tests/unit_tests/blas/level2/her.cpp +++ b/tests/unit_tests/blas/level2/her.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/her2.cpp b/tests/unit_tests/blas/level2/her2.cpp index e98c5cc8b..2a42bef17 100644 --- a/tests/unit_tests/blas/level2/her2.cpp +++ b/tests/unit_tests/blas/level2/her2.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/her2_usm.cpp b/tests/unit_tests/blas/level2/her2_usm.cpp index c732331ee..2940e3004 100644 --- a/tests/unit_tests/blas/level2/her2_usm.cpp +++ b/tests/unit_tests/blas/level2/her2_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/her_usm.cpp b/tests/unit_tests/blas/level2/her_usm.cpp index 9e1f5099e..589fed553 100644 --- a/tests/unit_tests/blas/level2/her_usm.cpp +++ b/tests/unit_tests/blas/level2/her_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/hpmv.cpp b/tests/unit_tests/blas/level2/hpmv.cpp index 69e6ea9b2..7f43d82a0 100644 --- a/tests/unit_tests/blas/level2/hpmv.cpp +++ b/tests/unit_tests/blas/level2/hpmv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/hpmv_usm.cpp b/tests/unit_tests/blas/level2/hpmv_usm.cpp index 743194b18..6be30eadf 100644 --- a/tests/unit_tests/blas/level2/hpmv_usm.cpp +++ b/tests/unit_tests/blas/level2/hpmv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/hpr.cpp b/tests/unit_tests/blas/level2/hpr.cpp index b2e5548bd..714da1780 100644 --- a/tests/unit_tests/blas/level2/hpr.cpp +++ b/tests/unit_tests/blas/level2/hpr.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/hpr2.cpp b/tests/unit_tests/blas/level2/hpr2.cpp index e2b19e2fd..fe03b6d48 100644 --- a/tests/unit_tests/blas/level2/hpr2.cpp +++ b/tests/unit_tests/blas/level2/hpr2.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/hpr2_usm.cpp b/tests/unit_tests/blas/level2/hpr2_usm.cpp index 6dc60dbf6..d58d4163d 100644 --- a/tests/unit_tests/blas/level2/hpr2_usm.cpp +++ b/tests/unit_tests/blas/level2/hpr2_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/hpr_usm.cpp b/tests/unit_tests/blas/level2/hpr_usm.cpp index b90b0ee63..15dcbd4ac 100644 --- a/tests/unit_tests/blas/level2/hpr_usm.cpp +++ b/tests/unit_tests/blas/level2/hpr_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/sbmv.cpp b/tests/unit_tests/blas/level2/sbmv.cpp index c0347dfda..b0291945b 100644 --- a/tests/unit_tests/blas/level2/sbmv.cpp +++ b/tests/unit_tests/blas/level2/sbmv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/sbmv_usm.cpp b/tests/unit_tests/blas/level2/sbmv_usm.cpp index 4fb7d46ad..b14464788 100644 --- a/tests/unit_tests/blas/level2/sbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/sbmv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/spmv.cpp b/tests/unit_tests/blas/level2/spmv.cpp index 799e7d775..78d413853 100644 --- a/tests/unit_tests/blas/level2/spmv.cpp +++ b/tests/unit_tests/blas/level2/spmv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/spmv_usm.cpp b/tests/unit_tests/blas/level2/spmv_usm.cpp index ae38ada4a..12e1eae39 100644 --- a/tests/unit_tests/blas/level2/spmv_usm.cpp +++ b/tests/unit_tests/blas/level2/spmv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/spr.cpp b/tests/unit_tests/blas/level2/spr.cpp index 4e4b5d8a9..1d954af7a 100644 --- a/tests/unit_tests/blas/level2/spr.cpp +++ b/tests/unit_tests/blas/level2/spr.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/spr2.cpp b/tests/unit_tests/blas/level2/spr2.cpp index d9d00a4e8..e1f1735ce 100644 --- a/tests/unit_tests/blas/level2/spr2.cpp +++ b/tests/unit_tests/blas/level2/spr2.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/spr2_usm.cpp b/tests/unit_tests/blas/level2/spr2_usm.cpp index 683288775..dd5e0fd84 100644 --- a/tests/unit_tests/blas/level2/spr2_usm.cpp +++ b/tests/unit_tests/blas/level2/spr2_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/spr_usm.cpp b/tests/unit_tests/blas/level2/spr_usm.cpp index 3a23a33b4..38f695b21 100644 --- a/tests/unit_tests/blas/level2/spr_usm.cpp +++ b/tests/unit_tests/blas/level2/spr_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/symv.cpp b/tests/unit_tests/blas/level2/symv.cpp index a22e48ff7..226fe7981 100644 --- a/tests/unit_tests/blas/level2/symv.cpp +++ b/tests/unit_tests/blas/level2/symv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/symv_usm.cpp b/tests/unit_tests/blas/level2/symv_usm.cpp index f33c0d25f..aee51965c 100644 --- a/tests/unit_tests/blas/level2/symv_usm.cpp +++ b/tests/unit_tests/blas/level2/symv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/syr.cpp b/tests/unit_tests/blas/level2/syr.cpp index 6b305582b..b7cf0aad7 100644 --- a/tests/unit_tests/blas/level2/syr.cpp +++ b/tests/unit_tests/blas/level2/syr.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/syr2.cpp b/tests/unit_tests/blas/level2/syr2.cpp index 5da1e0106..b58af622d 100644 --- a/tests/unit_tests/blas/level2/syr2.cpp +++ b/tests/unit_tests/blas/level2/syr2.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/syr2_usm.cpp b/tests/unit_tests/blas/level2/syr2_usm.cpp index a1e2cba7d..357200f79 100644 --- a/tests/unit_tests/blas/level2/syr2_usm.cpp +++ b/tests/unit_tests/blas/level2/syr2_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/syr_usm.cpp b/tests/unit_tests/blas/level2/syr_usm.cpp index 5a9f5034d..1c10bd493 100644 --- a/tests/unit_tests/blas/level2/syr_usm.cpp +++ b/tests/unit_tests/blas/level2/syr_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/tbmv.cpp b/tests/unit_tests/blas/level2/tbmv.cpp index 554082a01..c22f03c65 100644 --- a/tests/unit_tests/blas/level2/tbmv.cpp +++ b/tests/unit_tests/blas/level2/tbmv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/tbmv_usm.cpp b/tests/unit_tests/blas/level2/tbmv_usm.cpp index 808c5d1c3..d3de28f21 100644 --- a/tests/unit_tests/blas/level2/tbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/tbmv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/tbsv.cpp b/tests/unit_tests/blas/level2/tbsv.cpp index e653105e8..305254936 100644 --- a/tests/unit_tests/blas/level2/tbsv.cpp +++ b/tests/unit_tests/blas/level2/tbsv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/tbsv_usm.cpp b/tests/unit_tests/blas/level2/tbsv_usm.cpp index 1b77997eb..dbbadad7e 100644 --- a/tests/unit_tests/blas/level2/tbsv_usm.cpp +++ b/tests/unit_tests/blas/level2/tbsv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/tpmv.cpp b/tests/unit_tests/blas/level2/tpmv.cpp index ce45279bb..04fbf1986 100644 --- a/tests/unit_tests/blas/level2/tpmv.cpp +++ b/tests/unit_tests/blas/level2/tpmv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/tpmv_usm.cpp b/tests/unit_tests/blas/level2/tpmv_usm.cpp index 74ebc2502..9760da308 100644 --- a/tests/unit_tests/blas/level2/tpmv_usm.cpp +++ b/tests/unit_tests/blas/level2/tpmv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/tpsv.cpp b/tests/unit_tests/blas/level2/tpsv.cpp index 2a12ab1da..348ce132b 100644 --- a/tests/unit_tests/blas/level2/tpsv.cpp +++ b/tests/unit_tests/blas/level2/tpsv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/tpsv_usm.cpp b/tests/unit_tests/blas/level2/tpsv_usm.cpp index bcb676843..9cdeb57aa 100644 --- a/tests/unit_tests/blas/level2/tpsv_usm.cpp +++ b/tests/unit_tests/blas/level2/tpsv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/trmv.cpp b/tests/unit_tests/blas/level2/trmv.cpp index 8dfc517eb..ef6fc4845 100644 --- a/tests/unit_tests/blas/level2/trmv.cpp +++ b/tests/unit_tests/blas/level2/trmv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/trmv_usm.cpp b/tests/unit_tests/blas/level2/trmv_usm.cpp index af3e4b898..c444fe786 100644 --- a/tests/unit_tests/blas/level2/trmv_usm.cpp +++ b/tests/unit_tests/blas/level2/trmv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/trsv.cpp b/tests/unit_tests/blas/level2/trsv.cpp index fb1e39e06..137f4379e 100644 --- a/tests/unit_tests/blas/level2/trsv.cpp +++ b/tests/unit_tests/blas/level2/trsv.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level2/trsv_usm.cpp b/tests/unit_tests/blas/level2/trsv_usm.cpp index 2e6242d58..908b36dbd 100644 --- a/tests/unit_tests/blas/level2/trsv_usm.cpp +++ b/tests/unit_tests/blas/level2/trsv_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/gemm.cpp b/tests/unit_tests/blas/level3/gemm.cpp index 564700b16..f892ce19d 100644 --- a/tests/unit_tests/blas/level3/gemm.cpp +++ b/tests/unit_tests/blas/level3/gemm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/gemm_usm.cpp b/tests/unit_tests/blas/level3/gemm_usm.cpp index 9d5d8d048..a07092bf3 100644 --- a/tests/unit_tests/blas/level3/gemm_usm.cpp +++ b/tests/unit_tests/blas/level3/gemm_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/hemm.cpp b/tests/unit_tests/blas/level3/hemm.cpp index ce050e97d..5e661c81f 100644 --- a/tests/unit_tests/blas/level3/hemm.cpp +++ b/tests/unit_tests/blas/level3/hemm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/hemm_usm.cpp b/tests/unit_tests/blas/level3/hemm_usm.cpp index eafb06ea5..af0cf5217 100644 --- a/tests/unit_tests/blas/level3/hemm_usm.cpp +++ b/tests/unit_tests/blas/level3/hemm_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/her2k.cpp b/tests/unit_tests/blas/level3/her2k.cpp index ce57041d9..a2639426d 100644 --- a/tests/unit_tests/blas/level3/her2k.cpp +++ b/tests/unit_tests/blas/level3/her2k.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/her2k_usm.cpp b/tests/unit_tests/blas/level3/her2k_usm.cpp index a4ada6cb2..0bed61698 100644 --- a/tests/unit_tests/blas/level3/her2k_usm.cpp +++ b/tests/unit_tests/blas/level3/her2k_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/herk.cpp b/tests/unit_tests/blas/level3/herk.cpp index f908a77b7..8ba8dd041 100644 --- a/tests/unit_tests/blas/level3/herk.cpp +++ b/tests/unit_tests/blas/level3/herk.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/herk_usm.cpp b/tests/unit_tests/blas/level3/herk_usm.cpp index 470159c63..fec764421 100644 --- a/tests/unit_tests/blas/level3/herk_usm.cpp +++ b/tests/unit_tests/blas/level3/herk_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/symm.cpp b/tests/unit_tests/blas/level3/symm.cpp index 3f6920370..08a7dcd2e 100644 --- a/tests/unit_tests/blas/level3/symm.cpp +++ b/tests/unit_tests/blas/level3/symm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/symm_usm.cpp b/tests/unit_tests/blas/level3/symm_usm.cpp index f774e82e3..7785c922b 100644 --- a/tests/unit_tests/blas/level3/symm_usm.cpp +++ b/tests/unit_tests/blas/level3/symm_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/syr2k.cpp b/tests/unit_tests/blas/level3/syr2k.cpp index 0153e9ec0..2709c9e87 100644 --- a/tests/unit_tests/blas/level3/syr2k.cpp +++ b/tests/unit_tests/blas/level3/syr2k.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/syr2k_usm.cpp b/tests/unit_tests/blas/level3/syr2k_usm.cpp index efa3f07d3..403f5e89a 100644 --- a/tests/unit_tests/blas/level3/syr2k_usm.cpp +++ b/tests/unit_tests/blas/level3/syr2k_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/syrk.cpp b/tests/unit_tests/blas/level3/syrk.cpp index a6b28735d..74ec2be38 100644 --- a/tests/unit_tests/blas/level3/syrk.cpp +++ b/tests/unit_tests/blas/level3/syrk.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/syrk_usm.cpp b/tests/unit_tests/blas/level3/syrk_usm.cpp index e5569eb78..00ed1cab4 100644 --- a/tests/unit_tests/blas/level3/syrk_usm.cpp +++ b/tests/unit_tests/blas/level3/syrk_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/trmm.cpp b/tests/unit_tests/blas/level3/trmm.cpp index 2a02aa0d1..8629a58df 100644 --- a/tests/unit_tests/blas/level3/trmm.cpp +++ b/tests/unit_tests/blas/level3/trmm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/trmm_usm.cpp b/tests/unit_tests/blas/level3/trmm_usm.cpp index 1fa9bbdb0..3dec90f5b 100644 --- a/tests/unit_tests/blas/level3/trmm_usm.cpp +++ b/tests/unit_tests/blas/level3/trmm_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/trsm.cpp b/tests/unit_tests/blas/level3/trsm.cpp index 90b8d5c93..f8110ce2b 100644 --- a/tests/unit_tests/blas/level3/trsm.cpp +++ b/tests/unit_tests/blas/level3/trsm.cpp @@ -31,9 +31,9 @@ #endif #include "allocator_helper.hpp" #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/blas/level3/trsm_usm.cpp b/tests/unit_tests/blas/level3/trsm_usm.cpp index f84b0ed61..dd041f765 100644 --- a/tests/unit_tests/blas/level3/trsm_usm.cpp +++ b/tests/unit_tests/blas/level3/trsm_usm.cpp @@ -30,9 +30,9 @@ #include #endif #include "cblas.h" -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "onemkl_blas_helper.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "onemath_blas_helper.hpp" #include "reference_blas_templates.hpp" #include "test_common.hpp" #include "test_helper.hpp" diff --git a/tests/unit_tests/dft/include/compute_inplace.hpp b/tests/unit_tests/dft/include/compute_inplace.hpp index 9cc161c34..718bb687c 100644 --- a/tests/unit_tests/dft/include/compute_inplace.hpp +++ b/tests/unit_tests/dft/include/compute_inplace.hpp @@ -21,7 +21,7 @@ #define ONEMKL_COMPUTE_INPLACE_HPP #include "compute_tester.hpp" -#include +#include template int DFT_Test::test_in_place_buffer() { diff --git a/tests/unit_tests/dft/include/compute_tester.hpp b/tests/unit_tests/dft/include/compute_tester.hpp index 17ffac0cb..ea6a1b212 100644 --- a/tests/unit_tests/dft/include/compute_tester.hpp +++ b/tests/unit_tests/dft/include/compute_tester.hpp @@ -27,7 +27,7 @@ #else #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "test_helper.hpp" #include "test_common.hpp" #include "reference_dft.hpp" diff --git a/tests/unit_tests/dft/include/reference_dft.hpp b/tests/unit_tests/dft/include/reference_dft.hpp index 236edc7b0..a80dee83e 100644 --- a/tests/unit_tests/dft/include/reference_dft.hpp +++ b/tests/unit_tests/dft/include/reference_dft.hpp @@ -26,7 +26,7 @@ #include #include -#include +#include #include "test_common.hpp" namespace detail { diff --git a/tests/unit_tests/include/test_helper.hpp b/tests/unit_tests/include/test_helper.hpp index d4f79c5ce..63d94a493 100644 --- a/tests/unit_tests/include/test_helper.hpp +++ b/tests/unit_tests/include/test_helper.hpp @@ -30,9 +30,9 @@ #include #endif -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl/detail/backend_selector.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math/detail/backend_selector.hpp" #ifdef _WIN64 #include diff --git a/tests/unit_tests/lapack/include/lapack_common.hpp b/tests/unit_tests/lapack/include/lapack_common.hpp index 1cebb7553..1db0f9121 100644 --- a/tests/unit_tests/lapack/include/lapack_common.hpp +++ b/tests/unit_tests/lapack/include/lapack_common.hpp @@ -31,7 +31,7 @@ #include #endif -#include "oneapi/mkl/types.hpp" +#include "oneapi/math/types.hpp" namespace test_log { diff --git a/tests/unit_tests/lapack/include/lapack_test_controller.hpp b/tests/unit_tests/lapack/include/lapack_test_controller.hpp index 918060959..807698cf8 100644 --- a/tests/unit_tests/lapack/include/lapack_test_controller.hpp +++ b/tests/unit_tests/lapack/include/lapack_test_controller.hpp @@ -31,7 +31,7 @@ #endif #include "lapack_common.hpp" -#include "oneapi/mkl/exceptions.hpp" +#include "oneapi/math/exceptions.hpp" template std::istream& operator>>(std::istream& is, T& t) { diff --git a/tests/unit_tests/lapack/source/gebrd.cpp b/tests/unit_tests/lapack/source/gebrd.cpp index 66eb0b231..b54b25da3 100644 --- a/tests/unit_tests/lapack/source/gebrd.cpp +++ b/tests/unit_tests/lapack/source/gebrd.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/geqrf.cpp b/tests/unit_tests/lapack/source/geqrf.cpp index 27577e972..66c1bc8c8 100644 --- a/tests/unit_tests/lapack/source/geqrf.cpp +++ b/tests/unit_tests/lapack/source/geqrf.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/geqrf_batch_group.cpp b/tests/unit_tests/lapack/source/geqrf_batch_group.cpp index 416466028..78f0388a5 100644 --- a/tests/unit_tests/lapack/source/geqrf_batch_group.cpp +++ b/tests/unit_tests/lapack/source/geqrf_batch_group.cpp @@ -28,7 +28,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp b/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp index 16ceef63a..78dbcec3f 100644 --- a/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/gerqf.cpp b/tests/unit_tests/lapack/source/gerqf.cpp index dac6d79aa..ed46758df 100644 --- a/tests/unit_tests/lapack/source/gerqf.cpp +++ b/tests/unit_tests/lapack/source/gerqf.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/gesvd.cpp b/tests/unit_tests/lapack/source/gesvd.cpp index 1e143315b..56f590605 100644 --- a/tests/unit_tests/lapack/source/gesvd.cpp +++ b/tests/unit_tests/lapack/source/gesvd.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/getrf.cpp b/tests/unit_tests/lapack/source/getrf.cpp index 4537ef665..159d01e2d 100644 --- a/tests/unit_tests/lapack/source/getrf.cpp +++ b/tests/unit_tests/lapack/source/getrf.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/getrf_batch_group.cpp b/tests/unit_tests/lapack/source/getrf_batch_group.cpp index 12e651746..d16840ae5 100644 --- a/tests/unit_tests/lapack/source/getrf_batch_group.cpp +++ b/tests/unit_tests/lapack/source/getrf_batch_group.cpp @@ -28,7 +28,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/getrf_batch_stride.cpp b/tests/unit_tests/lapack/source/getrf_batch_stride.cpp index 3e4ef6589..67f21654b 100644 --- a/tests/unit_tests/lapack/source/getrf_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/getrf_batch_stride.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/getri.cpp b/tests/unit_tests/lapack/source/getri.cpp index a1aa2deda..9037d33c8 100644 --- a/tests/unit_tests/lapack/source/getri.cpp +++ b/tests/unit_tests/lapack/source/getri.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/getri_batch_group.cpp b/tests/unit_tests/lapack/source/getri_batch_group.cpp index 244acfcc8..6de30e512 100644 --- a/tests/unit_tests/lapack/source/getri_batch_group.cpp +++ b/tests/unit_tests/lapack/source/getri_batch_group.cpp @@ -28,7 +28,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/getri_batch_stride.cpp b/tests/unit_tests/lapack/source/getri_batch_stride.cpp index 5a71d2d7e..22f709785 100644 --- a/tests/unit_tests/lapack/source/getri_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/getri_batch_stride.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/getrs.cpp b/tests/unit_tests/lapack/source/getrs.cpp index bfc271758..767e9248d 100644 --- a/tests/unit_tests/lapack/source/getrs.cpp +++ b/tests/unit_tests/lapack/source/getrs.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/getrs_batch_group.cpp b/tests/unit_tests/lapack/source/getrs_batch_group.cpp index 2027663e4..5149738d4 100644 --- a/tests/unit_tests/lapack/source/getrs_batch_group.cpp +++ b/tests/unit_tests/lapack/source/getrs_batch_group.cpp @@ -28,7 +28,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/getrs_batch_stride.cpp b/tests/unit_tests/lapack/source/getrs_batch_stride.cpp index 1faf3d3e6..bd7c8f281 100644 --- a/tests/unit_tests/lapack/source/getrs_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/getrs_batch_stride.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/heevd.cpp b/tests/unit_tests/lapack/source/heevd.cpp index 62c23c3ad..34f2c15d1 100644 --- a/tests/unit_tests/lapack/source/heevd.cpp +++ b/tests/unit_tests/lapack/source/heevd.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/hegvd.cpp b/tests/unit_tests/lapack/source/hegvd.cpp index 9a109e6b8..96585cc19 100644 --- a/tests/unit_tests/lapack/source/hegvd.cpp +++ b/tests/unit_tests/lapack/source/hegvd.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/hetrd.cpp b/tests/unit_tests/lapack/source/hetrd.cpp index 13172d64f..d6ccca707 100644 --- a/tests/unit_tests/lapack/source/hetrd.cpp +++ b/tests/unit_tests/lapack/source/hetrd.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/hetrf.cpp b/tests/unit_tests/lapack/source/hetrf.cpp index 73535a77f..dc6d7676c 100644 --- a/tests/unit_tests/lapack/source/hetrf.cpp +++ b/tests/unit_tests/lapack/source/hetrf.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/orgbr.cpp b/tests/unit_tests/lapack/source/orgbr.cpp index 274cafce0..d7fec952b 100644 --- a/tests/unit_tests/lapack/source/orgbr.cpp +++ b/tests/unit_tests/lapack/source/orgbr.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/orgqr.cpp b/tests/unit_tests/lapack/source/orgqr.cpp index 9d62daf5f..95d218992 100644 --- a/tests/unit_tests/lapack/source/orgqr.cpp +++ b/tests/unit_tests/lapack/source/orgqr.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/orgqr_batch_group.cpp b/tests/unit_tests/lapack/source/orgqr_batch_group.cpp index 3af796e7d..b08c0c9aa 100644 --- a/tests/unit_tests/lapack/source/orgqr_batch_group.cpp +++ b/tests/unit_tests/lapack/source/orgqr_batch_group.cpp @@ -28,7 +28,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp b/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp index 1cf3471c5..c78e7e82c 100644 --- a/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/orgtr.cpp b/tests/unit_tests/lapack/source/orgtr.cpp index 5a01745d5..4021ccc0e 100644 --- a/tests/unit_tests/lapack/source/orgtr.cpp +++ b/tests/unit_tests/lapack/source/orgtr.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/ormqr.cpp b/tests/unit_tests/lapack/source/ormqr.cpp index e2ed49b96..ea64e26be 100644 --- a/tests/unit_tests/lapack/source/ormqr.cpp +++ b/tests/unit_tests/lapack/source/ormqr.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/ormrq.cpp b/tests/unit_tests/lapack/source/ormrq.cpp index 4882e5bc7..d9736da19 100644 --- a/tests/unit_tests/lapack/source/ormrq.cpp +++ b/tests/unit_tests/lapack/source/ormrq.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/ormtr.cpp b/tests/unit_tests/lapack/source/ormtr.cpp index 4e8dd95b9..5af36c851 100644 --- a/tests/unit_tests/lapack/source/ormtr.cpp +++ b/tests/unit_tests/lapack/source/ormtr.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/potrf.cpp b/tests/unit_tests/lapack/source/potrf.cpp index 7d2df8ea9..d3b98d8c5 100644 --- a/tests/unit_tests/lapack/source/potrf.cpp +++ b/tests/unit_tests/lapack/source/potrf.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/potrf_batch_group.cpp b/tests/unit_tests/lapack/source/potrf_batch_group.cpp index 4a5b8dd58..97833e08b 100644 --- a/tests/unit_tests/lapack/source/potrf_batch_group.cpp +++ b/tests/unit_tests/lapack/source/potrf_batch_group.cpp @@ -28,7 +28,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/potrf_batch_stride.cpp b/tests/unit_tests/lapack/source/potrf_batch_stride.cpp index fae4f0bcc..f2a606e06 100644 --- a/tests/unit_tests/lapack/source/potrf_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/potrf_batch_stride.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/potri.cpp b/tests/unit_tests/lapack/source/potri.cpp index cd2f86449..93387b4a2 100644 --- a/tests/unit_tests/lapack/source/potri.cpp +++ b/tests/unit_tests/lapack/source/potri.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/potrs.cpp b/tests/unit_tests/lapack/source/potrs.cpp index c534ec8ba..209387345 100644 --- a/tests/unit_tests/lapack/source/potrs.cpp +++ b/tests/unit_tests/lapack/source/potrs.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/potrs_batch_group.cpp b/tests/unit_tests/lapack/source/potrs_batch_group.cpp index 35c5ead0c..d4d69edee 100644 --- a/tests/unit_tests/lapack/source/potrs_batch_group.cpp +++ b/tests/unit_tests/lapack/source/potrs_batch_group.cpp @@ -28,7 +28,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/potrs_batch_stride.cpp b/tests/unit_tests/lapack/source/potrs_batch_stride.cpp index de2568e86..2eedc3190 100644 --- a/tests/unit_tests/lapack/source/potrs_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/potrs_batch_stride.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/syevd.cpp b/tests/unit_tests/lapack/source/syevd.cpp index 291713354..bf5f2d61b 100644 --- a/tests/unit_tests/lapack/source/syevd.cpp +++ b/tests/unit_tests/lapack/source/syevd.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/sygvd.cpp b/tests/unit_tests/lapack/source/sygvd.cpp index f800b03dd..8ea658299 100644 --- a/tests/unit_tests/lapack/source/sygvd.cpp +++ b/tests/unit_tests/lapack/source/sygvd.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/sytrd.cpp b/tests/unit_tests/lapack/source/sytrd.cpp index 01ffe0dff..e6864e09b 100644 --- a/tests/unit_tests/lapack/source/sytrd.cpp +++ b/tests/unit_tests/lapack/source/sytrd.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/sytrf.cpp b/tests/unit_tests/lapack/source/sytrf.cpp index 81d7fdb2d..3d04c528a 100644 --- a/tests/unit_tests/lapack/source/sytrf.cpp +++ b/tests/unit_tests/lapack/source/sytrf.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/trtrs.cpp b/tests/unit_tests/lapack/source/trtrs.cpp index 4018a2c51..70946ca11 100644 --- a/tests/unit_tests/lapack/source/trtrs.cpp +++ b/tests/unit_tests/lapack/source/trtrs.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/ungbr.cpp b/tests/unit_tests/lapack/source/ungbr.cpp index 7cdf8e52a..3fd840406 100644 --- a/tests/unit_tests/lapack/source/ungbr.cpp +++ b/tests/unit_tests/lapack/source/ungbr.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/ungqr.cpp b/tests/unit_tests/lapack/source/ungqr.cpp index 08b8b1192..dce9a9308 100644 --- a/tests/unit_tests/lapack/source/ungqr.cpp +++ b/tests/unit_tests/lapack/source/ungqr.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/ungqr_batch_group.cpp b/tests/unit_tests/lapack/source/ungqr_batch_group.cpp index ddb350828..99c7303b2 100644 --- a/tests/unit_tests/lapack/source/ungqr_batch_group.cpp +++ b/tests/unit_tests/lapack/source/ungqr_batch_group.cpp @@ -28,7 +28,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp b/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp index e656b9fb7..1da5a60df 100644 --- a/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/ungtr.cpp b/tests/unit_tests/lapack/source/ungtr.cpp index b0ad8e8f2..b6204e606 100644 --- a/tests/unit_tests/lapack/source/ungtr.cpp +++ b/tests/unit_tests/lapack/source/ungtr.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/unmqr.cpp b/tests/unit_tests/lapack/source/unmqr.cpp index 2f555c1ca..95f2ef2f1 100644 --- a/tests/unit_tests/lapack/source/unmqr.cpp +++ b/tests/unit_tests/lapack/source/unmqr.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/unmrq.cpp b/tests/unit_tests/lapack/source/unmrq.cpp index 628063837..8a26c7657 100644 --- a/tests/unit_tests/lapack/source/unmrq.cpp +++ b/tests/unit_tests/lapack/source/unmrq.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/lapack/source/unmtr.cpp b/tests/unit_tests/lapack/source/unmtr.cpp index 8148c644d..31a2935d5 100644 --- a/tests/unit_tests/lapack/source/unmtr.cpp +++ b/tests/unit_tests/lapack/source/unmtr.cpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "lapack_common.hpp" #include "lapack_test_controller.hpp" #include "lapack_accuracy_checks.hpp" diff --git a/tests/unit_tests/main_test.cpp b/tests/unit_tests/main_test.cpp index 25277e4e0..da8a38216 100644 --- a/tests/unit_tests/main_test.cpp +++ b/tests/unit_tests/main_test.cpp @@ -25,8 +25,8 @@ #endif #include #include "test_helper.hpp" -#include "oneapi/mkl/detail/config.hpp" -#include "oneapi/mkl.hpp" +#include "oneapi/math/detail/config.hpp" +#include "oneapi/math.hpp" #define MAX_STR 128 diff --git a/tests/unit_tests/rng/device/include/moments.hpp b/tests/unit_tests/rng/device/include/moments.hpp index 8acf20bf9..df1441ff6 100644 --- a/tests/unit_tests/rng/device/include/moments.hpp +++ b/tests/unit_tests/rng/device/include/moments.hpp @@ -35,7 +35,7 @@ #include #endif -#include "oneapi/mkl/rng/device.hpp" +#include "oneapi/math/rng/device.hpp" #include "rng_device_test_common.hpp" diff --git a/tests/unit_tests/rng/device/include/skip_ahead_test.hpp b/tests/unit_tests/rng/device/include/skip_ahead_test.hpp index 0b3bcf8a7..d1ff3dc8e 100644 --- a/tests/unit_tests/rng/device/include/skip_ahead_test.hpp +++ b/tests/unit_tests/rng/device/include/skip_ahead_test.hpp @@ -38,7 +38,7 @@ #include #endif -#include "oneapi/mkl/rng/device.hpp" +#include "oneapi/math/rng/device.hpp" #include "rng_device_test_common.hpp" diff --git a/tests/unit_tests/rng/include/engines_api_tests.hpp b/tests/unit_tests/rng/include/engines_api_tests.hpp index 2469c3023..567707afc 100644 --- a/tests/unit_tests/rng/include/engines_api_tests.hpp +++ b/tests/unit_tests/rng/include/engines_api_tests.hpp @@ -30,7 +30,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "rng_test_common.hpp" diff --git a/tests/unit_tests/rng/include/skip_ahead_test.hpp b/tests/unit_tests/rng/include/skip_ahead_test.hpp index efec71dde..2cc657549 100644 --- a/tests/unit_tests/rng/include/skip_ahead_test.hpp +++ b/tests/unit_tests/rng/include/skip_ahead_test.hpp @@ -30,7 +30,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "rng_test_common.hpp" diff --git a/tests/unit_tests/rng/include/statistics_check.hpp b/tests/unit_tests/rng/include/statistics_check.hpp index 8a1d045f0..2f8c68816 100644 --- a/tests/unit_tests/rng/include/statistics_check.hpp +++ b/tests/unit_tests/rng/include/statistics_check.hpp @@ -26,7 +26,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "rng_test_common.hpp" diff --git a/tests/unit_tests/rng/include/statistics_check_test.hpp b/tests/unit_tests/rng/include/statistics_check_test.hpp index 14a637d7a..53ad61460 100644 --- a/tests/unit_tests/rng/include/statistics_check_test.hpp +++ b/tests/unit_tests/rng/include/statistics_check_test.hpp @@ -30,7 +30,7 @@ #include #endif -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "statistics_check.hpp" diff --git a/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp b/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp index 62b213100..777eb5e45 100644 --- a/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp +++ b/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp @@ -24,7 +24,7 @@ #include #include -#include "oneapi/mkl.hpp" +#include "oneapi/math.hpp" #include "test_common.hpp" diff --git a/tests/unit_tests/sparse_blas/include/test_spmm.hpp b/tests/unit_tests/sparse_blas/include/test_spmm.hpp index 6188d4268..09eb24369 100644 --- a/tests/unit_tests/sparse_blas/include/test_spmm.hpp +++ b/tests/unit_tests/sparse_blas/include/test_spmm.hpp @@ -26,8 +26,8 @@ #include #endif -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" #include "common_sparse_reference.hpp" #include "test_common.hpp" diff --git a/tests/unit_tests/sparse_blas/include/test_spmv.hpp b/tests/unit_tests/sparse_blas/include/test_spmv.hpp index f141db893..e6e4f60bc 100644 --- a/tests/unit_tests/sparse_blas/include/test_spmv.hpp +++ b/tests/unit_tests/sparse_blas/include/test_spmv.hpp @@ -26,8 +26,8 @@ #include #endif -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" #include "common_sparse_reference.hpp" #include "test_common.hpp" diff --git a/tests/unit_tests/sparse_blas/include/test_spsv.hpp b/tests/unit_tests/sparse_blas/include/test_spsv.hpp index bdf9210f8..d301b0d48 100644 --- a/tests/unit_tests/sparse_blas/include/test_spsv.hpp +++ b/tests/unit_tests/sparse_blas/include/test_spsv.hpp @@ -26,8 +26,8 @@ #include #endif -#include "oneapi/mkl.hpp" -#include "oneapi/mkl/detail/config.hpp" +#include "oneapi/math.hpp" +#include "oneapi/math/detail/config.hpp" #include "common_sparse_reference.hpp" #include "test_common.hpp" From 600cd44c87e8c9d523450b612496b51bb8e7502a Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 8 Oct 2024 11:59:52 +0200 Subject: [PATCH 03/41] onemkl -> onemath --- .github/workflows/slack-pr.yaml | 2 +- README.md | 2 +- cmake/WarningsUtils.cmake | 8 +- docs/create_new_backend.rst | 40 ++--- docs/domains/sparse_linear_algebra.rst | 4 +- docs/index.rst | 8 +- docs/using_onemath_with_cmake.rst | 28 ++-- .../level3/CMakeLists.txt | 4 +- .../level3/CMakeLists.txt | 4 +- .../compile_time_dispatching/CMakeLists.txt | 6 +- .../dft/run_time_dispatching/CMakeLists.txt | 6 +- .../compile_time_dispatching/CMakeLists.txt | 4 +- .../run_time_dispatching/CMakeLists.txt | 4 +- .../compile_time_dispatching/CMakeLists.txt | 8 +- .../rng/run_time_dispatching/CMakeLists.txt | 4 +- .../compile_time_dispatching/CMakeLists.txt | 4 +- .../run_time_dispatching/CMakeLists.txt | 6 +- include/oneapi/math/detail/backends_table.hpp | 4 +- include/oneapi/math/detail/export.hpp | 2 +- .../detail/onemath_sparse_blas_backends.hxx | 4 +- scripts/generate_backend_api.py | 4 +- scripts/generate_cmake.py | 2 +- scripts/generate_ct_instant.py | 2 +- scripts/generate_wrappers.py | 2 +- src/CMakeLists.txt | 24 +-- src/blas/CMakeLists.txt | 16 +- src/blas/backends/CMakeLists.txt | 4 +- src/blas/backends/cublas/CMakeLists.txt | 4 +- src/blas/backends/cublas/cublas_batch.cpp | 12 +- .../backends/cublas/cublas_extensions.cpp | 18 +-- src/blas/backends/cublas/cublas_level1.cpp | 58 +++---- src/blas/backends/cublas/cublas_level2.cpp | 94 +++++------ src/blas/backends/cublas/cublas_level3.cpp | 42 ++--- src/blas/backends/cublas/cublas_task.hpp | 2 +- src/blas/backends/cublas/cublas_wrappers.cpp | 2 +- src/blas/backends/mklcpu/CMakeLists.txt | 4 +- src/blas/backends/mklcpu/mklcpu_batch.cpp | 2 +- .../backends/mklcpu/mklcpu_extensions.cpp | 2 +- src/blas/backends/mklcpu/mklcpu_level1.cpp | 2 +- src/blas/backends/mklcpu/mklcpu_level2.cpp | 2 +- src/blas/backends/mklcpu/mklcpu_level3.cpp | 2 +- src/blas/backends/mklcpu/mklcpu_wrappers.cpp | 2 +- src/blas/backends/mklgpu/CMakeLists.txt | 4 +- src/blas/backends/mklgpu/mklgpu_batch.cpp | 2 +- .../backends/mklgpu/mklgpu_extensions.cpp | 2 +- src/blas/backends/mklgpu/mklgpu_level1.cpp | 2 +- src/blas/backends/mklgpu/mklgpu_level2.cpp | 2 +- src/blas/backends/mklgpu/mklgpu_level3.cpp | 2 +- src/blas/backends/mklgpu/mklgpu_wrappers.cpp | 2 +- src/blas/backends/netlib/CMakeLists.txt | 4 +- src/blas/backends/netlib/netlib_batch.cpp | 2 +- src/blas/backends/netlib/netlib_common.hpp | 4 +- .../backends/netlib/netlib_extensions.cpp | 2 +- src/blas/backends/netlib/netlib_level1.cpp | 2 +- src/blas/backends/netlib/netlib_level2.cpp | 2 +- src/blas/backends/netlib/netlib_level3.cpp | 2 +- src/blas/backends/netlib/netlib_wrappers.cpp | 2 +- src/blas/backends/portblas/CMakeLists.txt | 4 +- src/blas/backends/portblas/portblas_batch.cpp | 2 +- .../backends/portblas/portblas_common.hpp | 4 +- .../portblas/portblas_level1_double.cpp | 2 +- .../portblas/portblas_level1_float.cpp | 2 +- .../portblas/portblas_level2_double.cpp | 2 +- .../portblas/portblas_level2_float.cpp | 2 +- .../portblas/portblas_level3_bfloat16.cpp | 2 +- .../portblas/portblas_level3_double.cpp | 2 +- .../portblas/portblas_level3_float.cpp | 2 +- .../portblas/portblas_level3_half.cpp | 2 +- .../backends/portblas/portblas_wrappers.cpp | 2 +- src/blas/backends/rocblas/CMakeLists.txt | 4 +- src/blas/backends/rocblas/rocblas_batch.cpp | 54 +++---- .../backends/rocblas/rocblas_extensions.cpp | 10 +- src/blas/backends/rocblas/rocblas_level1.cpp | 58 +++---- src/blas/backends/rocblas/rocblas_level2.cpp | 94 +++++------ src/blas/backends/rocblas/rocblas_level3.cpp | 42 ++--- src/blas/backends/rocblas/rocblas_task.hpp | 2 +- .../backends/rocblas/rocblas_wrappers.cpp | 2 +- src/dft/CMakeLists.txt | 18 +-- src/dft/backends/CMakeLists.txt | 4 +- src/dft/backends/cufft/CMakeLists.txt | 4 +- src/dft/backends/cufft/backward.cpp | 2 +- src/dft/backends/cufft/commit.cpp | 2 +- src/dft/backends/cufft/cufft_wrappers.cpp | 2 +- src/dft/backends/cufft/descriptor.cpp | 2 +- src/dft/backends/cufft/forward.cpp | 2 +- src/dft/backends/mklcpu/CMakeLists.txt | 8 +- src/dft/backends/mklcpu/backward.cpp | 2 +- src/dft/backends/mklcpu/commit.cpp | 2 +- src/dft/backends/mklcpu/descriptor.cpp | 2 +- src/dft/backends/mklcpu/forward.cpp | 2 +- .../backends/mklcpu/mkl_dft_cpu_wrappers.cpp | 2 +- src/dft/backends/mklgpu/CMakeLists.txt | 8 +- src/dft/backends/mklgpu/backward.cpp | 2 +- src/dft/backends/mklgpu/commit.cpp | 12 +- src/dft/backends/mklgpu/descriptor.cpp | 2 +- src/dft/backends/mklgpu/forward.cpp | 2 +- .../backends/mklgpu/mkl_dft_gpu_wrappers.cpp | 2 +- src/dft/backends/portfft/CMakeLists.txt | 6 +- src/dft/backends/portfft/commit.cpp | 2 +- src/dft/backends/portfft/descriptor.cpp | 2 +- src/dft/backends/portfft/portfft_wrappers.cpp | 2 +- src/dft/backends/rocfft/CMakeLists.txt | 4 +- src/dft/backends/rocfft/backward.cpp | 2 +- src/dft/backends/rocfft/commit.cpp | 2 +- src/dft/backends/rocfft/descriptor.cpp | 2 +- src/dft/backends/rocfft/forward.cpp | 2 +- src/dft/backends/rocfft/rocfft_wrappers.cpp | 2 +- src/lapack/CMakeLists.txt | 14 +- src/lapack/backends/CMakeLists.txt | 4 +- src/lapack/backends/cusolver/CMakeLists.txt | 4 +- .../backends/cusolver/cusolver_batch.cpp | 64 ++++---- .../backends/cusolver/cusolver_lapack.cpp | 148 +++++++++--------- .../backends/cusolver/cusolver_task.hpp | 2 +- .../backends/cusolver/cusolver_wrappers.cpp | 2 +- src/lapack/backends/mklcpu/CMakeLists.txt | 4 +- .../backends/mklcpu/lapack_cpu_wrappers.cpp | 2 +- src/lapack/backends/mklcpu/mkl_lapack.cpp | 2 +- src/lapack/backends/mklgpu/CMakeLists.txt | 4 +- .../backends/mklgpu/lapack_gpu_wrappers.cpp | 2 +- src/lapack/backends/mklgpu/mkl_lapack.cpp | 2 +- src/lapack/backends/rocsolver/CMakeLists.txt | 4 +- .../backends/rocsolver/rocsolver_batch.cpp | 6 +- .../backends/rocsolver/rocsolver_lapack.cpp | 102 ++++++------ .../backends/rocsolver/rocsolver_task.hpp | 2 +- .../backends/rocsolver/rocsolver_wrappers.cpp | 2 +- src/rng/CMakeLists.txt | 16 +- src/rng/backends/CMakeLists.txt | 4 +- src/rng/backends/curand/CMakeLists.txt | 4 +- src/rng/backends/curand/curand_task.hpp | 4 +- src/rng/backends/curand/curand_wrappers.cpp | 2 +- src/rng/backends/curand/mrg32k3a.cpp | 42 ++--- src/rng/backends/curand/philox4x32x10.cpp | 42 ++--- src/rng/backends/mklcpu/CMakeLists.txt | 4 +- .../backends/mklcpu/mkl_rng_cpu_wrappers.cpp | 2 +- src/rng/backends/mklcpu/mrg32k3a.cpp | 2 +- src/rng/backends/mklcpu/philox4x32x10.cpp | 2 +- src/rng/backends/mklgpu/CMakeLists.txt | 4 +- .../backends/mklgpu/mkl_rng_gpu_wrappers.cpp | 2 +- src/rng/backends/mklgpu/mrg32k3a.cpp | 2 +- src/rng/backends/mklgpu/philox4x32x10.cpp | 2 +- src/rng/backends/rocrand/CMakeLists.txt | 4 +- src/rng/backends/rocrand/mrg32k3a.cpp | 66 ++++---- src/rng/backends/rocrand/philox4x32x10.cpp | 66 ++++---- src/rng/backends/rocrand/rocrand_task.hpp | 4 +- src/rng/backends/rocrand/rocrand_wrappers.cpp | 2 +- src/sparse_blas/CMakeLists.txt | 18 +-- src/sparse_blas/backends/CMakeLists.txt | 4 +- .../backends/mklcpu/CMakeLists.txt | 8 +- .../backends/mklgpu/CMakeLists.txt | 8 +- tests/README.md | 2 +- tests/unit_tests/CMakeLists.txt | 56 +++---- .../blas/include/onemath_blas_helper.hpp | 2 +- tests/unit_tests/dft/source/CMakeLists.txt | 4 +- .../sparse_blas/source/CMakeLists.txt | 4 +- 154 files changed, 822 insertions(+), 822 deletions(-) diff --git a/.github/workflows/slack-pr.yaml b/.github/workflows/slack-pr.yaml index 4c5f3df7d..5c8b466a6 100644 --- a/.github/workflows/slack-pr.yaml +++ b/.github/workflows/slack-pr.yaml @@ -22,7 +22,7 @@ on: env: SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }} - channel: "onemkl" + channel: "onemath" permissions: pull-requests: read diff --git a/README.md b/README.md index 94ef2b85a..d8f228b96 100644 --- a/README.md +++ b/README.md @@ -138,7 +138,7 @@ There are two oneMKL selector layer implementations: ```cmd $> clang++ -fsycl –I$ONEMKL/include app.cpp - $> clang++ -fsycl app.o –L$ONEMKL/lib –lonemkl_blas_mklcpu –lonemkl_blas_cublas + $> clang++ -fsycl app.o –L$ONEMKL/lib –lonemath_blas_mklcpu –lonemath_blas_cublas ``` *Refer to [Selecting a Compiler](https://oneapi-src.github.io/oneMKL/selecting_a_compiler.html) for the choice between `icpx/icx` and `clang++` compilers.* diff --git a/cmake/WarningsUtils.cmake b/cmake/WarningsUtils.cmake index 3b5f76afb..8de2858a6 100644 --- a/cmake/WarningsUtils.cmake +++ b/cmake/WarningsUtils.cmake @@ -19,7 +19,7 @@ include_guard(GLOBAL) -add_library(onemkl_warnings INTERFACE) +add_library(onemath_warnings INTERFACE) set(ONEMKL_WARNINGS "") @@ -41,8 +41,8 @@ add_warning("-Wpedantic") message(VERBOSE "Domains with warnings enabled use: ${ONEMKL_WARNINGS}") -# The onemkl_warnings target can be linked to any other target to enable warnings. -target_compile_options(onemkl_warnings INTERFACE ${ONEMKL_WARNINGS}) +# The onemath_warnings target can be linked to any other target to enable warnings. +target_compile_options(onemath_warnings INTERFACE ${ONEMKL_WARNINGS}) # Add the library to install package -install(TARGETS onemkl_warnings EXPORT oneMKLTargets) +install(TARGETS onemath_warnings EXPORT oneMKLTargets) diff --git a/docs/create_new_backend.rst b/docs/create_new_backend.rst index faccd6085..15b9ba3ae 100644 --- a/docs/create_new_backend.rst +++ b/docs/create_new_backend.rst @@ -14,11 +14,11 @@ For the new backend library and header naming please use the following template: .. code-block:: - onemkl__<3rd-party library short name>[] + onemath__<3rd-party library short name>[] -Where ```` is required only if multiple wrappers are provided from the same 3rd-party library, e.g., wrappers with Intel oneMKL C API for CPU target ``onemkl_blas_mklcpu.so`` and wrappers with Intel oneMKL DPC++ API for GPU target ``onemkl_blas_mklgpu.so``. +Where ```` is required only if multiple wrappers are provided from the same 3rd-party library, e.g., wrappers with Intel oneMKL C API for CPU target ``onemath_blas_mklcpu.so`` and wrappers with Intel oneMKL DPC++ API for GPU target ``onemath_blas_mklgpu.so``. -If there is no need for multiple wrappers only ```` and ``<3rd-party library short name>`` are required, e.g. ``onemkl_rng_curand.so`` +If there is no need for multiple wrappers only ```` and ``<3rd-party library short name>`` are required, e.g. ``onemath_rng_curand.so`` `1. Create Header Files`_ @@ -45,10 +45,10 @@ For each new backend library, you should create the following two header files: .. code-block:: bash python scripts/generate_backend_api.py include/oneapi/math/blas.hpp \ # Base header file - include/oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp \ # Output header file + include/oneapi/math/blas/detail/newlib/onemath_blas_newlib.hpp \ # Output header file oneapi::mkl::newlib # Wrappers namespace -Code snippet of the generated header file ``include/oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp`` +Code snippet of the generated header file ``include/oneapi/math/blas/detail/newlib/onemath_blas_newlib.hpp`` .. code-block:: cpp @@ -67,7 +67,7 @@ Code snippet of the generated header file ``include/oneapi/math/blas/detail/newl python scripts/generate_ct_instant.py include/oneapi/math/blas/detail/blas_ct_templates.hpp \ # Base header file include/oneapi/math/blas/detail/newlib/blas_ct.hpp \ # Output header file - include/oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp \ # Header file with declaration of entry points to wrappers + include/oneapi/math/blas/detail/newlib/onemath_blas_newlib.hpp \ # Header file with declaration of entry points to wrappers newlib \ # Library name newdevice \ # Backend name oneapi::mkl::newlib # Wrappers namespace @@ -117,10 +117,10 @@ Below you can see structure of oneMKL top-level include directory: blas_ct_templates.hpp -> oneMKL Compile-time BLAS API general templates cublas/ blas_ct.hpp -> oneMKL Compile-time BLAS API template instantiations for - onemkl_blas_cublas.hpp -> backend wrappers library API + onemath_blas_cublas.hpp -> backend wrappers library API mklcpu/ blas_ct.hpp -> oneMKL Compile-time BLAS API template instantiations for - onemkl_blas_mklcpu.hpp -> backend wrappers library API + onemath_blas_mklcpu.hpp -> backend wrappers library API / / @@ -205,13 +205,13 @@ The new files generated at the `1. Create Header Files`_ step result in the foll blas_ct_templates.hpp -> oneMKL Compile-time BLAS API general templates cublas/ blas_ct.hpp -> oneMKL Compile-time BLAS API template instantiations for - onemkl_blas_cublas.hpp -> backend wrappers library API + onemath_blas_cublas.hpp -> backend wrappers library API mklcpu/ blas_ct.hpp -> oneMKL Compile-time BLAS API template instantiations for - onemkl_blas_mklcpu.hpp -> backend wrappers library API + onemath_blas_mklcpu.hpp -> backend wrappers library API + newlib/ + blas_ct.hpp -> oneMKL Compile-time BLAS API template instantiations for - + onemkl_blas_newlib.hpp -> backend wrappers library API + + onemath_blas_newlib.hpp -> backend wrappers library API / / @@ -248,12 +248,12 @@ You can modify wrappers generated with this script to enable third-party library The command below generates two new files: -* ``src/blas/backends/newlib/newlib_wrappers.cpp`` - DPC++ wrappers for all functions from ``include/oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp`` +* ``src/blas/backends/newlib/newlib_wrappers.cpp`` - DPC++ wrappers for all functions from ``include/oneapi/math/blas/detail/newlib/onemath_blas_newlib.hpp`` * ``src/blas/backends/newlib/newlib_wrappers_table_dyn.cpp`` - structure of symbols for run-time dispatcher (in the same location as wrappers), suffix ``_dyn`` indicates that this file is required for dynamic library only. .. code-block:: bash - python scripts/generate_wrappers.py include/oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp \ # Base header file + python scripts/generate_wrappers.py include/oneapi/math/blas/detail/newlib/onemath_blas_newlib.hpp \ # Base header file src/blas/function_table.hpp \ # Declaration for structure of symbols src/blas/backends/newlib/newlib_wrappers.cpp \ # Output wrappers newlib # Library name @@ -278,7 +278,7 @@ The following code snippet is updated for ``src/blas/backends/newlib/newlib_wrap #include "oneapi/math/types.hpp" - #include "oneapi/math/blas/detail/newlib/onemkl_blas_newlib.hpp" + #include "oneapi/math/blas/detail/newlib/onemath_blas_newlib.hpp" + + #include "newlib.h" @@ -437,22 +437,22 @@ Update the following files to enable the new third-party library for unit tests: .. code-block:: diff if(ENABLE_MKLCPU_BACKEND) - add_dependencies(test_main_ct onemkl_blas_mklcpu) + add_dependencies(test_main_ct onemath_blas_mklcpu) if(BUILD_SHARED_LIBS) - list(APPEND ONEMKL_LIBRARIES onemkl_blas_mklcpu) + list(APPEND ONEMKL_LIBRARIES onemath_blas_mklcpu) else() - list(APPEND ONEMKL_LIBRARIES -foffload-static-lib=${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libonemkl_blas_mklcpu.a) + list(APPEND ONEMKL_LIBRARIES -foffload-static-lib=${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libonemath_blas_mklcpu.a) find_package(MKL REQUIRED) list(APPEND ONEMKL_LIBRARIES ${MKL_LINK_C}) endif() endif() + + if(ENABLE_NEWLIB_BACKEND) - + add_dependencies(test_main_ct onemkl_blas_newlib) + + add_dependencies(test_main_ct onemath_blas_newlib) + if(BUILD_SHARED_LIBS) - + list(APPEND ONEMKL_LIBRARIES onemkl_blas_newlib) + + list(APPEND ONEMKL_LIBRARIES onemath_blas_newlib) + else() - + list(APPEND ONEMKL_LIBRARIES -foffload-static-lib=${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libonemkl_blas_newlib.a) + + list(APPEND ONEMKL_LIBRARIES -foffload-static-lib=${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libonemath_blas_newlib.a) + find_package(NEWLIB REQUIRED) + list(APPEND ONEMKL_LIBRARIES ONEMKL::NEWLIB::NEWLIB) + endif() diff --git a/docs/domains/sparse_linear_algebra.rst b/docs/domains/sparse_linear_algebra.rst index eab5afd56..ba2699c76 100644 --- a/docs/domains/sparse_linear_algebra.rst +++ b/docs/domains/sparse_linear_algebra.rst @@ -1,10 +1,10 @@ -.. _onemkl_sparse_linear_algebra: +.. _onemath_sparse_linear_algebra: Sparse Linear Algebra ===================== See the latest specification for the sparse domain `here -`_. +`_. This page documents implementation specific or backend specific details of the sparse domain. diff --git a/docs/index.rst b/docs/index.rst index 96f5907af..7a906f1c9 100644 --- a/docs/index.rst +++ b/docs/index.rst @@ -1,11 +1,11 @@ .. Copyright 2020-2024 Intel Corporation -.. _onemkl: +.. _onemath: -***************** -oneMKL Interfaces -***************** +******* +oneMath +******* Contents ======== diff --git a/docs/using_onemath_with_cmake.rst b/docs/using_onemath_with_cmake.rst index 03f828bf4..b1c80b7a8 100644 --- a/docs/using_onemath_with_cmake.rst +++ b/docs/using_onemath_with_cmake.rst @@ -1,4 +1,4 @@ -.. _using_onemkl_interface_library_with_cmake: +.. _using_onemath_interface_library_with_cmake: Using the oneMKL Interfaces in your project with CMake ============================================================= @@ -31,12 +31,12 @@ For example: .. code-block:: cmake find_package(oneMKL REQUIRED) - target_link_libraries(myTarget PRIVATE MKL::onemkl) + target_link_libraries(myTarget PRIVATE MKL::onemath) Different targets can be used depending on the requirements of oneMKL. -To link against the entire library, the ``MKL::onemkl`` target should be used. -For specific domains, ``MKL::onemkl_`` should be used. -And for specific backends, ``MKL::onemkl__`` should be used. +To link against the entire library, the ``MKL::onemath`` target should be used. +For specific domains, ``MKL::onemath_`` should be used. +And for specific backends, ``MKL::onemath__`` should be used. When using a binary, it may be useful to know the backends that were enabled during the build. To check for the existence of backends, CMake's ``if(TARGET @@ -44,8 +44,8 @@ during the build. To check for the existence of backends, CMake's ``if(TARGET .. code-block:: cmake - if(TARGET MKL::onemkl_dft_cufft) - target_link_libraries(myTarget PRIVATE MKL::onemkl_dft_cufft) + if(TARGET MKL::onemath_dft_cufft) + target_link_libraries(myTarget PRIVATE MKL::onemath_dft_cufft) else() message(FATAL_ERROR "oneMKL Interfaces was not built with CuFFT backend") endif() @@ -68,23 +68,23 @@ For example: set(BUILD_EXAMPLES False) set(ENABLE__BACKEND True) FetchContent_Declare( - onemkl_interface_library + onemath_interface_library GIT_REPOSITORY https://github.com/oneapi-src/oneMKL.git GIT_TAG develop ) - FetchContent_MakeAvailable(onemkl_interface_library) + FetchContent_MakeAvailable(onemath_interface_library) - target_link_libraries(myTarget PRIVATE onemkl) + target_link_libraries(myTarget PRIVATE onemath) The build parameters should be appropriately set before ``FetchContent_Declare``. See :ref:`building_the_project_with_dpcpp` or :ref:`building_the_project_with_adaptivecpp`. To link against the main library with run-time dispatching, use the target -``onemkl``. To link against particular domains, use the target -``onemkl_``. For example, ``onemkl_blas`` or ``onemkl_dft``. To link +``onemath``. To link against particular domains, use the target +``onemath_``. For example, ``onemath_blas`` or ``onemath_dft``. To link against particular backends (as required for static dispatch of oneAPI calls to -a particular backend), use the target ``onemkl__``. For -example, ``onemkl_dft_cufft``. +a particular backend), use the target ``onemath__``. For +example, ``onemath_dft_cufft``. .. _FetchContent: https://cmake.org/cmake/help/latest/module/FetchContent.html diff --git a/examples/blas/compile_time_dispatching/level3/CMakeLists.txt b/examples/blas/compile_time_dispatching/level3/CMakeLists.txt index 294e000de..7aaebdf1a 100644 --- a/examples/blas/compile_time_dispatching/level3/CMakeLists.txt +++ b/examples/blas/compile_time_dispatching/level3/CMakeLists.txt @@ -32,8 +32,8 @@ foreach(blas_ct_source ${BLAS_CT_SOURCES}) ) if(domain STREQUAL "blas" AND ENABLE_MKLCPU_BACKEND AND ENABLE_CUBLAS_BACKEND) - add_dependencies(example_${domain}_${blas_ct_source} onemkl_${domain}_mklcpu onemkl_${domain}_cublas) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_mklcpu onemkl_${domain}_cublas) + add_dependencies(example_${domain}_${blas_ct_source} onemath_${domain}_mklcpu onemath_${domain}_cublas) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_mklcpu onemath_${domain}_cublas) endif() target_link_libraries(example_${domain}_${blas_ct_source} PUBLIC diff --git a/examples/blas/run_time_dispatching/level3/CMakeLists.txt b/examples/blas/run_time_dispatching/level3/CMakeLists.txt index d0d35fc0d..193e5db95 100644 --- a/examples/blas/run_time_dispatching/level3/CMakeLists.txt +++ b/examples/blas/run_time_dispatching/level3/CMakeLists.txt @@ -64,14 +64,14 @@ foreach(blas_rt_source ${BLAS_RT_SOURCES}) PUBLIC ${CMAKE_BINARY_DIR}/bin ) - add_dependencies(example_${domain}_${blas_rt_source} onemkl) + add_dependencies(example_${domain}_${blas_rt_source} onemath) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET example_${domain}_${blas_rt_source} SOURCES ${BLAS_RT_SOURCES}) endif() target_link_libraries(example_${domain}_${blas_rt_source} PUBLIC - onemkl + onemath ONEMKL::SYCL::SYCL ${CMAKE_DL_LIBS} ) diff --git a/examples/dft/compile_time_dispatching/CMakeLists.txt b/examples/dft/compile_time_dispatching/CMakeLists.txt index ed0ca2922..5684b7cdb 100644 --- a/examples/dft/compile_time_dispatching/CMakeLists.txt +++ b/examples/dft/compile_time_dispatching/CMakeLists.txt @@ -34,13 +34,13 @@ foreach(dft_ct_source ${DFT_CT_SOURCES}) ) if(domain STREQUAL "dft" AND ENABLE_MKLCPU_BACKEND AND ENABLE_CUFFT_BACKEND) - add_dependencies(${EXAMPLE_NAME} onemkl_${domain}_mklcpu onemkl_${domain}_cufft) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_mklcpu onemkl_${domain}_cufft) + add_dependencies(${EXAMPLE_NAME} onemath_${domain}_mklcpu onemath_${domain}_cufft) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_mklcpu onemath_${domain}_cufft) endif() target_link_libraries(${EXAMPLE_NAME} PUBLIC ${ONEMKL_LIBRARIES_${domain}} - onemkl_warnings + onemath_warnings ) # Register example as ctest diff --git a/examples/dft/run_time_dispatching/CMakeLists.txt b/examples/dft/run_time_dispatching/CMakeLists.txt index e221c7950..c1f89338d 100644 --- a/examples/dft/run_time_dispatching/CMakeLists.txt +++ b/examples/dft/run_time_dispatching/CMakeLists.txt @@ -57,17 +57,17 @@ foreach(dft_rt_sources ${DFT_RT_SOURCES}) PUBLIC ${CMAKE_BINARY_DIR}/bin ) - add_dependencies(example_${domain}_${dft_rt_sources} onemkl) + add_dependencies(example_${domain}_${dft_rt_sources} onemath) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET example_${domain}_${dft_rt_sources} SOURCES ${DFT_RT_SOURCES}) endif() target_link_libraries(example_${domain}_${dft_rt_sources} - PUBLIC onemkl + PUBLIC onemath PUBLIC ONEMKL::SYCL::SYCL PUBLIC ${CMAKE_DL_LIBS} - PRIVATE onemkl_warnings + PRIVATE onemath_warnings ) # Register example as ctest diff --git a/examples/lapack/compile_time_dispatching/CMakeLists.txt b/examples/lapack/compile_time_dispatching/CMakeLists.txt index cc126674f..69acae8e0 100644 --- a/examples/lapack/compile_time_dispatching/CMakeLists.txt +++ b/examples/lapack/compile_time_dispatching/CMakeLists.txt @@ -36,8 +36,8 @@ foreach(lapack_ct_source ${LAPACK_CT_SOURCES}) PUBLIC ${CMAKE_BINARY_DIR}/bin ) if(domain STREQUAL "lapack" AND ENABLE_MKLCPU_BACKEND AND ENABLE_CUSOLVER_BACKEND) - add_dependencies(example_${domain}_${lapack_ct_source} onemkl_${domain}_mklcpu onemkl_${domain}_cusolver) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_mklcpu onemkl_${domain}_cusolver) + add_dependencies(example_${domain}_${lapack_ct_source} onemath_${domain}_mklcpu onemath_${domain}_cusolver) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_mklcpu onemath_${domain}_cusolver) target_link_libraries(example_${domain}_${lapack_ct_source} PUBLIC ${OPENCL_LIBRARY}) endif() target_link_libraries(example_${domain}_${lapack_ct_source} PUBLIC diff --git a/examples/lapack/run_time_dispatching/CMakeLists.txt b/examples/lapack/run_time_dispatching/CMakeLists.txt index 5fcf6a311..5a2affce2 100644 --- a/examples/lapack/run_time_dispatching/CMakeLists.txt +++ b/examples/lapack/run_time_dispatching/CMakeLists.txt @@ -49,14 +49,14 @@ foreach(lapack_rt_source ${LAPACK_RT_SOURCES}) PUBLIC ${CMAKE_BINARY_DIR}/bin ) - add_dependencies(example_${domain}_${lapack_rt_source} onemkl) + add_dependencies(example_${domain}_${lapack_rt_source} onemath) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET example_${domain}_${lapack_rt_source} SOURCES ${LAPACK_RT_SOURCES}) endif() target_link_libraries(example_${domain}_${lapack_rt_source} PUBLIC - onemkl + onemath ONEMKL::SYCL::SYCL ${CMAKE_DL_LIBS} ) diff --git a/examples/rng/compile_time_dispatching/CMakeLists.txt b/examples/rng/compile_time_dispatching/CMakeLists.txt index 4f57db38c..bcbc9d4ea 100644 --- a/examples/rng/compile_time_dispatching/CMakeLists.txt +++ b/examples/rng/compile_time_dispatching/CMakeLists.txt @@ -33,11 +33,11 @@ foreach(rng_ct_source ${RNG_CT_SOURCES}) if(domain STREQUAL "rng" AND ENABLE_MKLCPU_BACKEND AND ENABLE_CURAND_BACKEND) add_dependencies(example_${domain}_${rng_ct_source} - onemkl_${domain}_mklcpu - onemkl_${domain}_curand) + onemath_${domain}_mklcpu + onemath_${domain}_curand) list(APPEND ONEMKL_LIBRARIES_${domain} - onemkl_${domain}_mklcpu - onemkl_${domain}_curand) + onemath_${domain}_mklcpu + onemath_${domain}_curand) endif() target_link_libraries(example_${domain}_${rng_ct_source} PUBLIC diff --git a/examples/rng/run_time_dispatching/CMakeLists.txt b/examples/rng/run_time_dispatching/CMakeLists.txt index d3bcc0f19..5a373efaa 100644 --- a/examples/rng/run_time_dispatching/CMakeLists.txt +++ b/examples/rng/run_time_dispatching/CMakeLists.txt @@ -50,14 +50,14 @@ foreach(rng_rt_source ${RNG_RT_SOURCES}) PUBLIC ${CMAKE_BINARY_DIR}/bin ) - add_dependencies(example_${domain}_${rng_rt_source} onemkl) + add_dependencies(example_${domain}_${rng_rt_source} onemath) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET example_${domain}_${rng_rt_source} SOURCES ${RNG_RT_SOURCES}) endif() target_link_libraries(example_${domain}_${rng_rt_source} PUBLIC - onemkl + onemath ONEMKL::SYCL::SYCL ${CMAKE_DL_LIBS} ) diff --git a/examples/sparse_blas/compile_time_dispatching/CMakeLists.txt b/examples/sparse_blas/compile_time_dispatching/CMakeLists.txt index 5dbbba8a4..42516da05 100644 --- a/examples/sparse_blas/compile_time_dispatching/CMakeLists.txt +++ b/examples/sparse_blas/compile_time_dispatching/CMakeLists.txt @@ -35,8 +35,8 @@ foreach(backend ${SPARSE_BLAS_BACKENDS}) PUBLIC ${CMAKE_BINARY_DIR}/bin ) - add_dependencies(${EXAMPLE_NAME} onemkl_sparse_blas_${backend}) - target_link_libraries(${EXAMPLE_NAME} PRIVATE ONEMKL::SYCL::SYCL onemkl_sparse_blas_${backend}) + add_dependencies(${EXAMPLE_NAME} onemath_sparse_blas_${backend}) + target_link_libraries(${EXAMPLE_NAME} PRIVATE ONEMKL::SYCL::SYCL onemath_sparse_blas_${backend}) # Register example as ctest add_test(NAME sparse_blas/EXAMPLE/CT/sparse_blas_spmv_usm_${backend} COMMAND ${EXAMPLE_NAME}) diff --git a/examples/sparse_blas/run_time_dispatching/CMakeLists.txt b/examples/sparse_blas/run_time_dispatching/CMakeLists.txt index 398f3e0f2..1d64928e5 100644 --- a/examples/sparse_blas/run_time_dispatching/CMakeLists.txt +++ b/examples/sparse_blas/run_time_dispatching/CMakeLists.txt @@ -44,17 +44,17 @@ foreach(sparse_blas_rt_sources ${SPARSE_BLAS_RT_SOURCES}) PUBLIC ${CMAKE_BINARY_DIR}/bin ) - add_dependencies(example_${sparse_blas_rt_sources} onemkl) + add_dependencies(example_${sparse_blas_rt_sources} onemath) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET example_${sparse_blas_rt_sources} SOURCES ${SPARSE_BLAS_RT_SOURCES}) endif() target_link_libraries(example_${sparse_blas_rt_sources} - PUBLIC onemkl + PUBLIC onemath PUBLIC ONEMKL::SYCL::SYCL PUBLIC ${CMAKE_DL_LIBS} - PRIVATE onemkl_warnings + PRIVATE onemath_warnings ) # Register example as ctest diff --git a/include/oneapi/math/detail/backends_table.hpp b/include/oneapi/math/detail/backends_table.hpp index bc9a1e78a..b59ce2b10 100644 --- a/include/oneapi/math/detail/backends_table.hpp +++ b/include/oneapi/math/detail/backends_table.hpp @@ -32,9 +32,9 @@ #include "oneapi/math/detail/config.hpp" #ifdef __linux__ -#define LIB_NAME(a) "libonemkl_" a ".so" +#define LIB_NAME(a) "libonemath_" a ".so" #elif defined(_WIN64) -#define LIB_NAME(a) "onemkl_" a ".dll" +#define LIB_NAME(a) "onemath_" a ".dll" #endif namespace oneapi { diff --git a/include/oneapi/math/detail/export.hpp b/include/oneapi/math/detail/export.hpp index ca73e949b..2b2596e09 100644 --- a/include/oneapi/math/detail/export.hpp +++ b/include/oneapi/math/detail/export.hpp @@ -27,7 +27,7 @@ #define ONEMKL_NO_EXPORT #else #ifndef ONEMKL_EXPORT -#ifdef onemkl_EXPORTS +#ifdef onemath_EXPORTS /* We are building this library */ #define ONEMKL_EXPORT __declspec(dllexport) #else diff --git a/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx b/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx index 4b701eb6f..a0987b022 100644 --- a/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx +++ b/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx @@ -17,8 +17,8 @@ * **************************************************************************/ -// This file is meant to be included in each backend onemkl_sparse_blas_BACKEND.hpp files. -// It is used to exports each symbol to the onemkl_sparse_blas_BACKEND library. +// This file is meant to be included in each backend onemath_sparse_blas_BACKEND.hpp files. +// It is used to exports each symbol to the onemath_sparse_blas_BACKEND library. // Dense vector template diff --git a/scripts/generate_backend_api.py b/scripts/generate_backend_api.py index b35165759..ab392230c 100755 --- a/scripts/generate_backend_api.py +++ b/scripts/generate_backend_api.py @@ -40,10 +40,10 @@ def usage(err = None): Example: The command below will generate: -"onemkl_blas_mklgpu.hpp" header with declaration of all backend library APIs. +"onemath_blas_mklgpu.hpp" header with declaration of all backend library APIs. API from backend library will be called from "oneapi::mkl::mklgpu::blas" namespace. -{script} include/oneapi/math/blas.hpp include/oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp oneapi::mkl::mklgpu::blas +{script} include/oneapi/math/blas.hpp include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp oneapi::mkl::mklgpu::blas '''.format(script = argv[0])) if len(argv) < 3: diff --git a/scripts/generate_cmake.py b/scripts/generate_cmake.py index 5538e972d..c245bc0a8 100644 --- a/scripts/generate_cmake.py +++ b/scripts/generate_cmake.py @@ -72,7 +72,7 @@ def usage(err = None): # generated file # -set(LIB_NAME onemkl_blas_{libname}) +set(LIB_NAME onemath_blas_{libname}) set(LIB_OBJ ${{LIB_NAME}}_obj) # Add third-party library diff --git a/scripts/generate_ct_instant.py b/scripts/generate_ct_instant.py index afcd02e3e..875d66533 100755 --- a/scripts/generate_ct_instant.py +++ b/scripts/generate_ct_instant.py @@ -43,7 +43,7 @@ def usage(err = None): "blas_ct.hpp" header with compile-time BLAS API based on "blas_ct_templates.hpp" for "mklgpu" backend. API from the backend library will be called from "oneapi::mkl::mklgpu::blas" namespace. -{script} include/oneapi/math/blas/detail/blas_ct_templates.hpp include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp include/oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp mklgpu oneapi::mkl::mklgpu::blas +{script} include/oneapi/math/blas/detail/blas_ct_templates.hpp include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp mklgpu oneapi::mkl::mklgpu::blas '''.format(script = argv[0])) if len(argv) < 6: diff --git a/scripts/generate_wrappers.py b/scripts/generate_wrappers.py index 06e021aea..61602128e 100755 --- a/scripts/generate_wrappers.py +++ b/scripts/generate_wrappers.py @@ -40,7 +40,7 @@ def usage(err = None): Example: - {script} include/oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp src/blas/function_table.hpp src/blas/backend/mklgpu/wrappers.cpp mklgpu + {script} include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp src/blas/function_table.hpp src/blas/backend/mklgpu/wrappers.cpp mklgpu '''.format(script = argv[0])) if len(argv) <= 4: diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index 7f089471a..d5cce0a2e 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -20,7 +20,7 @@ # Define common build flags for oneMKL libraries set(ONEMKL_BUILD_COPT "") if(WIN32 AND BUILD_SHARED_LIBS) - list(APPEND ONEMKL_BUILD_COPT "-Donemkl_EXPORTS") + list(APPEND ONEMKL_BUILD_COPT "-Donemath_EXPORTS") endif() # store path to CMAKE_CURRENT_BINARY_DIR to use it later (makes FetchContent_Declare workable) @@ -34,7 +34,7 @@ set(ONEMKL_INTERFACE_INCLUDE_DIRS ) # Build loader and backends for each domain -add_custom_target(onemkl_backend_libs) +add_custom_target(onemath_backend_libs) foreach(domain ${TARGET_DOMAINS}) add_subdirectory(${domain}) endforeach() @@ -73,32 +73,32 @@ endfunction() generate_header_file() -# Add recipe for onemkl loader library +# Add recipe for onemath loader library if(BUILD_SHARED_LIBS) - add_library(onemkl SHARED) + add_library(onemath SHARED) # The loader library depends on all the backend libraries as it uses # dlopen to load them at runtime. # Use add_dependencies to ensure that all the backend libraries are # (re-)built when compiling the loader or runtime binaries. - add_dependencies(onemkl onemkl_backend_libs) + add_dependencies(onemath onemath_backend_libs) - target_include_directories(onemkl + target_include_directories(onemath PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} ) - set_target_properties(onemkl PROPERTIES + set_target_properties(onemath PROPERTIES SOVERSION ${PROJECT_VERSION_MAJOR} ) # w/a for setting oneMKL Interfaces installed headers as -I instead of -isystem for cmake >= 3.25 for workable find_package(MKL) combination if(CMAKE_VERSION VERSION_GREATER_EQUAL "3.25.0") - set_target_properties(onemkl PROPERTIES EXPORT_NO_SYSTEM true) + set_target_properties(onemath PROPERTIES EXPORT_NO_SYSTEM true) endif() # Build dispatcher library set (ONEMKL_LIBS ${TARGET_DOMAINS}) - list(TRANSFORM ONEMKL_LIBS PREPEND onemkl_) - target_link_libraries(onemkl PUBLIC ${ONEMKL_LIBS}) - set_target_properties(onemkl PROPERTIES + list(TRANSFORM ONEMKL_LIBS PREPEND onemath_) + target_link_libraries(onemath PUBLIC ${ONEMKL_LIBS}) + set_target_properties(onemath PROPERTIES INSTALL_RPATH "\$ORIGIN" BUILD_WITH_INSTALL_RPATH TRUE ) @@ -107,7 +107,7 @@ if(BUILD_SHARED_LIBS) foreach(domain_lib ${ONEMKL_LIBS}) install(TARGETS ${domain_lib} EXPORT oneMKLTargets) endforeach() - install(TARGETS onemkl EXPORT oneMKLTargets + install(TARGETS onemath EXPORT oneMKLTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/blas/CMakeLists.txt b/src/blas/CMakeLists.txt index 1edf2e445..e40de2ac1 100644 --- a/src/blas/CMakeLists.txt +++ b/src/blas/CMakeLists.txt @@ -22,26 +22,26 @@ add_subdirectory(backends) # Recipe for BLAS loader object if(BUILD_SHARED_LIBS) -add_library(onemkl_blas OBJECT) -target_sources(onemkl_blas PRIVATE blas_loader.cpp) -target_include_directories(onemkl_blas +add_library(onemath_blas OBJECT) +target_sources(onemath_blas PRIVATE blas_loader.cpp) +target_include_directories(onemath_blas PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMKL_GENERATED_INCLUDE_PATH} - $ + $ ) -target_compile_options(onemkl_blas PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(onemath_blas PRIVATE ${ONEMKL_BUILD_COPT}) -set_target_properties(onemkl_blas PROPERTIES +set_target_properties(onemath_blas PROPERTIES POSITION_INDEPENDENT_CODE ON ) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) - add_sycl_to_target(TARGET onemkl_blas SOURCES blas_loader.cpp) + add_sycl_to_target(TARGET onemath_blas SOURCES blas_loader.cpp) else() - target_link_libraries(onemkl_blas PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(onemath_blas PUBLIC ONEMKL::SYCL::SYCL) endif() endif() diff --git a/src/blas/backends/CMakeLists.txt b/src/blas/backends/CMakeLists.txt index 351f4b0e5..ac7595fc6 100644 --- a/src/blas/backends/CMakeLists.txt +++ b/src/blas/backends/CMakeLists.txt @@ -17,8 +17,8 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -add_custom_target(onemkl_backend_libs_blas) -add_dependencies(onemkl_backend_libs onemkl_backend_libs_blas) +add_custom_target(onemath_backend_libs_blas) +add_dependencies(onemath_backend_libs onemath_backend_libs_blas) if(ENABLE_MKLCPU_BACKEND) add_subdirectory(mklcpu) diff --git a/src/blas/backends/cublas/CMakeLists.txt b/src/blas/backends/cublas/CMakeLists.txt index b64e7c37d..63b061fe3 100644 --- a/src/blas/backends/cublas/CMakeLists.txt +++ b/src/blas/backends/cublas/CMakeLists.txt @@ -17,7 +17,7 @@ # #========================================================================= -set(LIB_NAME onemkl_blas_cublas) +set(LIB_NAME onemath_blas_cublas) set(LIB_OBJ ${LIB_NAME}_obj) find_package(cuBLAS REQUIRED) set(SOURCES cublas_level1.cpp @@ -30,7 +30,7 @@ set(SOURCES cublas_level1.cpp $<$: cublas_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) -add_dependencies(onemkl_backend_libs_blas ${LIB_NAME}) +add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/blas/backends/cublas/cublas_batch.cpp b/src/blas/backends/cublas/cublas_batch.cpp index f86149f9f..b8d768a1d 100644 --- a/src/blas/backends/cublas/cublas_batch.cpp +++ b/src/blas/backends/cublas/cublas_batch.cpp @@ -19,7 +19,7 @@ #include "cublas_helper.hpp" #include "cublas_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" namespace oneapi { namespace mkl { @@ -161,7 +161,7 @@ inline void gemm_batch_impl(sycl::queue &queue, transpose transa, transpose tran auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -513,7 +513,7 @@ inline sycl::event gemv_batch(const char *func_name, Func func, sycl::queue &que } auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; cublasStatus_t err; @@ -632,7 +632,7 @@ inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose tra for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cublasStatus_t err; #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND @@ -718,7 +718,7 @@ inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, tr for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; cublasStatus_t err; @@ -832,7 +832,7 @@ inline sycl::event trsm_batch(const char *func_name, Func func, sycl::queue &que for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; cublasStatus_t err; diff --git a/src/blas/backends/cublas/cublas_extensions.cpp b/src/blas/backends/cublas/cublas_extensions.cpp index edac33387..d6e358cbc 100644 --- a/src/blas/backends/cublas/cublas_extensions.cpp +++ b/src/blas/backends/cublas/cublas_extensions.cpp @@ -19,7 +19,7 @@ #include "cublas_helper.hpp" #include "cublas_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" namespace oneapi { namespace mkl { @@ -96,7 +96,7 @@ void omatcopy(const char *func_name, Func func, sycl::queue &queue, transpose tr auto b_acc = b.template get_access(cgh); const int64_t logical_m = (trans == oneapi::mkl::transpose::nontrans ? m : n); const int64_t logical_n = (trans == oneapi::mkl::transpose::nontrans ? n : m); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -172,7 +172,7 @@ void omatadd(const char *func_name, Func func, sycl::queue &queue, transpose tra auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -274,7 +274,7 @@ sycl::event omatcopy(const char *func_name, Func func, sycl::queue &queue, trans cgh.depends_on(dependencies); const int64_t logical_m = (trans == oneapi::mkl::transpose::nontrans ? m : n); const int64_t logical_n = (trans == oneapi::mkl::transpose::nontrans ? n : m); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -356,7 +356,7 @@ inline sycl::event omatadd(const char *func_name, Func func, sycl::queue &queue, overflow_check(m, n, lda, ldb, ldc); auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -459,7 +459,7 @@ void omatcopy(const char *func_name, Func func, sycl::queue &queue, transpose tr auto b_acc = b.template get_access(cgh); const int64_t logical_m = (trans == oneapi::mkl::transpose::nontrans ? n : m); const int64_t logical_n = (trans == oneapi::mkl::transpose::nontrans ? m : n); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -535,7 +535,7 @@ void omatadd(const char *func_name, Func func, sycl::queue &queue, transpose tra auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -637,7 +637,7 @@ sycl::event omatcopy(const char *func_name, Func func, sycl::queue &queue, trans cgh.depends_on(dependencies); const int64_t logical_m = (trans == oneapi::mkl::transpose::nontrans ? n : m); const int64_t logical_n = (trans == oneapi::mkl::transpose::nontrans ? m : n); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -719,7 +719,7 @@ inline sycl::event omatadd(const char *func_name, Func func, sycl::queue &queue, overflow_check(m, n, lda, ldb, ldc); auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); diff --git a/src/blas/backends/cublas/cublas_level1.cpp b/src/blas/backends/cublas/cublas_level1.cpp index 3670d97a5..77dd6476b 100644 --- a/src/blas/backends/cublas/cublas_level1.cpp +++ b/src/blas/backends/cublas/cublas_level1.cpp @@ -20,7 +20,7 @@ #include "cublas_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" namespace oneapi { namespace mkl { @@ -41,7 +41,7 @@ inline void asum(const char *func_name, Func func, sycl::queue &queue, int64_t n queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -81,7 +81,7 @@ inline void scal(const char *func_name, Func func, sycl::queue &queue, int64_t n overflow_check(n, incx); queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = sc.get_mem(x_acc); cublasStatus_t err; @@ -112,7 +112,7 @@ inline void axpy(const char *func_name, Func func, sycl::queue &queue, int64_t n queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = sc.get_mem(x_acc); auto y_ = sc.get_mem(y_acc); @@ -167,7 +167,7 @@ inline void rotg(const char *func_name, Func func, sycl::queue &queue, sycl::buf auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto s_acc = s.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -211,7 +211,7 @@ inline void rotm(const char *func_name, Func func, sycl::queue &queue, int64_t n auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); auto param_acc = param.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -250,7 +250,7 @@ inline void copy(const char *func_name, Func func, sycl::queue &queue, int64_t n queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = sc.get_mem(x_acc); auto y_ = sc.get_mem(y_acc); @@ -282,7 +282,7 @@ inline void dot(const char *func_name, Func func, sycl::queue &queue, int64_t n, auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -327,7 +327,7 @@ inline void rot(const char *func_name, Func func, sycl::queue &queue, int64_t n, queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -364,7 +364,7 @@ void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, auto x_acc = x.get_access(cgh); auto y_acc = y.get_access(cgh); auto res_acc = result.get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -404,7 +404,7 @@ inline void rotmg(const char *func_name, Func func, sycl::queue &queue, sycl::bu auto x1_acc = x1.template get_access(cgh); auto y1_acc = y1_buff.template get_access(cgh); auto param_acc = param.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -453,7 +453,7 @@ inline void iamax(const char *func_name, Func func, sycl::queue &queue, int64_t queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto int_res_acc = int_res_buff.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -501,7 +501,7 @@ inline void swap(const char *func_name, Func func, sycl::queue &queue, int64_t n queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = sc.get_mem(x_acc); auto y_ = sc.get_mem(y_acc); @@ -539,7 +539,7 @@ inline void iamin(const char *func_name, Func func, sycl::queue &queue, int64_t queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto int_res_acc = int_res_buff.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -589,7 +589,7 @@ inline void nrm2(const char *func_name, Func func, sycl::queue &queue, int64_t n queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -639,7 +639,7 @@ inline sycl::event asum(const char *func_name, Func func, sycl::queue &queue, in for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); auto res_ = reinterpret_cast(result); @@ -679,7 +679,7 @@ inline sycl::event scal(const char *func_name, Func func, sycl::queue &queue, in for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); cublasStatus_t err; @@ -715,7 +715,7 @@ inline sycl::event axpy(const char *func_name, Func func, sycl::queue &queue, in for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); auto y_ = reinterpret_cast(y); @@ -788,7 +788,7 @@ inline sycl::event rotg(const char *func_name, Func func, sycl::queue &queue, T1 for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -830,7 +830,7 @@ inline sycl::event rotm(const char *func_name, Func func, sycl::queue &queue, in for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); auto y_ = reinterpret_cast(y); @@ -864,7 +864,7 @@ inline sycl::event copy(const char *func_name, Func func, sycl::queue &queue, in for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); auto y_ = reinterpret_cast(y); @@ -900,7 +900,7 @@ inline sycl::event dot(const char *func_name, Func func, sycl::queue &queue, int for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); auto y_ = reinterpret_cast(y); @@ -946,7 +946,7 @@ inline sycl::event rot(const char *func_name, Func func, sycl::queue &queue, int for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); auto y_ = reinterpret_cast(y); @@ -984,7 +984,7 @@ sycl::event sdsdot(sycl::queue &queue, int64_t n, float sb, const float *x, int6 for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); auto y_ = reinterpret_cast(y); @@ -1049,7 +1049,7 @@ inline sycl::event rotmg(const char *func_name, Func func, sycl::queue &queue, T for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto d1_ = reinterpret_cast(d1); auto d2_ = reinterpret_cast(d2); @@ -1111,7 +1111,7 @@ inline sycl::event iamax(const char *func_name, Func func, sycl::queue &queue, i for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); if (result_on_device) { @@ -1163,7 +1163,7 @@ inline sycl::event swap(const char *func_name, Func func, sycl::queue &queue, in for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); auto y_ = reinterpret_cast(y); @@ -1212,7 +1212,7 @@ inline sycl::event iamin(const char *func_name, Func func, sycl::queue &queue, i for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); if (result_on_device) { @@ -1268,7 +1268,7 @@ inline sycl::event nrm2(const char *func_name, Func func, sycl::queue &queue, in for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); auto res_ = reinterpret_cast(result); diff --git a/src/blas/backends/cublas/cublas_level2.cpp b/src/blas/backends/cublas/cublas_level2.cpp index 9eb223a4d..0f9992df2 100644 --- a/src/blas/backends/cublas/cublas_level2.cpp +++ b/src/blas/backends/cublas/cublas_level2.cpp @@ -20,7 +20,7 @@ #include "cublas_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" namespace oneapi { namespace mkl { @@ -40,7 +40,7 @@ inline void gemv(const char *func_name, Func func, sycl::queue &queue, transpose auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -77,7 +77,7 @@ inline void gbmv(const char *func_name, Func func, sycl::queue &queue, transpose auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -114,7 +114,7 @@ inline void ger(const char *func_name, Func func, sycl::queue &queue, int64_t m, auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -151,7 +151,7 @@ inline void hbmv(const char *func_name, Func func, sycl::queue &queue, uplo uppe auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -186,7 +186,7 @@ inline void hemv(const char *func_name, Func func, sycl::queue &queue, uplo uppe auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -222,7 +222,7 @@ inline void her(const char *func_name, Func func, sycl::queue &queue, uplo upper queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -256,7 +256,7 @@ inline void her2(const char *func_name, Func func, sycl::queue &queue, uplo uppe auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -292,7 +292,7 @@ inline void hpmv(const char *func_name, Func func, sycl::queue &queue, uplo uppe auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -328,7 +328,7 @@ inline void hpr(const char *func_name, Func func, sycl::queue &queue, uplo upper queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -361,7 +361,7 @@ inline void hpr2(const char *func_name, Func func, sycl::queue &queue, uplo uppe auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -396,7 +396,7 @@ inline void sbmv(const char *func_name, Func func, sycl::queue &queue, uplo uppe auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -432,7 +432,7 @@ inline void symv(const char *func_name, Func func, sycl::queue &queue, uplo uppe auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -466,7 +466,7 @@ inline void syr(const char *func_name, Func func, sycl::queue &queue, uplo upper queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -501,7 +501,7 @@ inline void syr2(const char *func_name, Func func, sycl::queue &queue, uplo uppe auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -540,7 +540,7 @@ inline void spmv(const char *func_name, Func func, sycl::queue &queue, uplo uppe auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -574,7 +574,7 @@ inline void spr(const char *func_name, Func func, sycl::queue &queue, uplo upper queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -607,7 +607,7 @@ inline void spr2(const char *func_name, Func func, sycl::queue &queue, uplo uppe auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -641,7 +641,7 @@ inline void tbmv(const char *func_name, Func func, sycl::queue &queue, uplo uppe queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -677,7 +677,7 @@ inline void tbsv(const char *func_name, Func func, sycl::queue &queue, uplo uppe queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -713,7 +713,7 @@ inline void tpmv(const char *func_name, Func func, sycl::queue &queue, uplo uppe queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -748,7 +748,7 @@ inline void tpsv(const char *func_name, Func func, sycl::queue &queue, uplo uppe queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -783,7 +783,7 @@ inline void trmv(const char *func_name, Func func, sycl::queue &queue, uplo uppe queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -818,7 +818,7 @@ inline void trsv(const char *func_name, Func func, sycl::queue &queue, uplo uppe queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto x_ = sc.get_mem(x_acc); @@ -858,7 +858,7 @@ inline sycl::event gemv(const char *func_name, Func func, sycl::queue &queue, tr for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -898,7 +898,7 @@ inline sycl::event gbmv(const char *func_name, Func func, sycl::queue &queue, tr for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -938,7 +938,7 @@ inline sycl::event ger(const char *func_name, Func func, sycl::queue &queue, int for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -979,7 +979,7 @@ inline sycl::event hbmv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1016,7 +1016,7 @@ inline sycl::event hemv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1055,7 +1055,7 @@ inline sycl::event her(const char *func_name, Func func, sycl::queue &queue, upl for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1092,7 +1092,7 @@ inline sycl::event her2(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1130,7 +1130,7 @@ inline sycl::event hpmv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1169,7 +1169,7 @@ inline sycl::event hpr(const char *func_name, Func func, sycl::queue &queue, upl for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1206,7 +1206,7 @@ inline sycl::event hpr2(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1245,7 +1245,7 @@ inline sycl::event sbmv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1283,7 +1283,7 @@ inline sycl::event symv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1321,7 +1321,7 @@ inline sycl::event syr(const char *func_name, Func func, sycl::queue &queue, upl for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1360,7 +1360,7 @@ inline sycl::event syr2(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1401,7 +1401,7 @@ inline sycl::event spmv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1439,7 +1439,7 @@ inline sycl::event spr(const char *func_name, Func func, sycl::queue &queue, upl for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1475,7 +1475,7 @@ inline sycl::event spr2(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1514,7 +1514,7 @@ inline sycl::event tbmv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1554,7 +1554,7 @@ inline sycl::event tbsv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1593,7 +1593,7 @@ inline sycl::event tpmv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1632,7 +1632,7 @@ inline sycl::event tpsv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1671,7 +1671,7 @@ inline sycl::event trmv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); @@ -1710,7 +1710,7 @@ inline sycl::event trsv(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto x_ = reinterpret_cast(x); diff --git a/src/blas/backends/cublas/cublas_level3.cpp b/src/blas/backends/cublas/cublas_level3.cpp index ca2324198..5ef3ea9bb 100644 --- a/src/blas/backends/cublas/cublas_level3.cpp +++ b/src/blas/backends/cublas/cublas_level3.cpp @@ -20,7 +20,7 @@ #include "cublas_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" namespace oneapi { namespace mkl { @@ -41,7 +41,7 @@ inline void gemm(const char *func_name, Func func, sycl::queue &queue, transpose auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -88,7 +88,7 @@ inline void gemm_ex(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, sycl::que auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -140,7 +140,7 @@ inline void symm(const char *func_name, Func func, sycl::queue &queue, side left auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -179,7 +179,7 @@ inline void hemm(const char *func_name, Func func, sycl::queue &queue, side left auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -213,7 +213,7 @@ inline void syrk(const char *func_name, Func func, sycl::queue &queue, uplo uppe queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto c_ = sc.get_mem(c_acc); @@ -252,7 +252,7 @@ inline void herk(const char *func_name, Func func, sycl::queue &queue, uplo uppe queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto c_ = sc.get_mem(c_acc); @@ -289,7 +289,7 @@ inline void syr2k(const char *func_name, Func func, sycl::queue &queue, uplo upp auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -329,7 +329,7 @@ inline void her2k(const char *func_name, Func func, sycl::queue &queue, uplo upp auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -370,7 +370,7 @@ inline void trmm(const char *func_name, Func func, sycl::queue &queue, side left queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -406,7 +406,7 @@ inline void trsm(const char *func_name, Func func, sycl::queue &queue, side left queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -447,7 +447,7 @@ inline sycl::event gemm(const char *func_name, Func func, sycl::queue &queue, tr for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -493,7 +493,7 @@ inline sycl::event gemm_ex_usm(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -549,7 +549,7 @@ inline sycl::event symm(const char *func_name, Func func, sycl::queue &queue, si for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -591,7 +591,7 @@ inline sycl::event hemm(const char *func_name, Func func, sycl::queue &queue, si for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -629,7 +629,7 @@ inline sycl::event syrk(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto c_ = reinterpret_cast(c); @@ -671,7 +671,7 @@ inline sycl::event herk(const char *func_name, Func func, sycl::queue &queue, up for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto c_ = reinterpret_cast(c); @@ -711,7 +711,7 @@ inline sycl::event syr2k(const char *func_name, Func func, sycl::queue &queue, u for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -755,7 +755,7 @@ inline sycl::event her2k(const char *func_name, Func func, sycl::queue &queue, u for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -800,7 +800,7 @@ inline sycl::event trmm(const char *func_name, Func func, sycl::queue &queue, si for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -840,7 +840,7 @@ inline sycl::event trsm(const char *func_name, Func func, sycl::queue &queue, si for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); diff --git a/src/blas/backends/cublas/cublas_task.hpp b/src/blas/backends/cublas/cublas_task.hpp index ff0ca0d8e..ee9b84e8d 100644 --- a/src/blas/backends/cublas/cublas_task.hpp +++ b/src/blas/backends/cublas/cublas_task.hpp @@ -78,7 +78,7 @@ static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { } #endif template -static inline void onemkl_cublas_host_task(H &cgh, sycl::queue queue, F f) { +static inline void onemath_cublas_host_task(H &cgh, sycl::queue queue, F f) { (void)host_task_internal(cgh, queue, f); } diff --git a/src/blas/backends/cublas/cublas_wrappers.cpp b/src/blas/backends/cublas/cublas_wrappers.cpp index 2c4670a65..0752fb247 100644 --- a/src/blas/backends/cublas/cublas_wrappers.cpp +++ b/src/blas/backends/cublas/cublas_wrappers.cpp @@ -17,7 +17,7 @@ * **************************************************************************/ #include "blas/function_table.hpp" -#include "oneapi/math/blas/detail/cublas/onemkl_blas_cublas.hpp" +#include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" #define WRAPPER_VERSION 1 diff --git a/src/blas/backends/mklcpu/CMakeLists.txt b/src/blas/backends/mklcpu/CMakeLists.txt index 322741d26..cfb1481b5 100644 --- a/src/blas/backends/mklcpu/CMakeLists.txt +++ b/src/blas/backends/mklcpu/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_blas_mklcpu) +set(LIB_NAME onemath_blas_mklcpu) set(LIB_OBJ ${LIB_NAME}_obj) set(SOURCES @@ -25,7 +25,7 @@ set(SOURCES $<$: mklcpu_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) -add_dependencies(onemkl_backend_libs_blas ${LIB_NAME}) +add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET ${LIB_OBJ} SOURCES ${SOURCES}) endif() diff --git a/src/blas/backends/mklcpu/mklcpu_batch.cpp b/src/blas/backends/mklcpu/mklcpu_batch.cpp index 3be24e4d2..64ed392ef 100644 --- a/src/blas/backends/mklcpu/mklcpu_batch.cpp +++ b/src/blas/backends/mklcpu/mklcpu_batch.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp" #include "oneapi/math/exceptions.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklcpu/mklcpu_extensions.cpp b/src/blas/backends/mklcpu/mklcpu_extensions.cpp index f41564eb7..0daa063e8 100644 --- a/src/blas/backends/mklcpu/mklcpu_extensions.cpp +++ b/src/blas/backends/mklcpu/mklcpu_extensions.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp" #include "oneapi/math/exceptions.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklcpu/mklcpu_level1.cpp b/src/blas/backends/mklcpu/mklcpu_level1.cpp index 6beed0bf8..edddb1afc 100644 --- a/src/blas/backends/mklcpu/mklcpu_level1.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level1.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklcpu/mklcpu_level2.cpp b/src/blas/backends/mklcpu/mklcpu_level2.cpp index 545a14b4c..8fa321ff9 100644 --- a/src/blas/backends/mklcpu/mklcpu_level2.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level2.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklcpu/mklcpu_level3.cpp b/src/blas/backends/mklcpu/mklcpu_level3.cpp index 435868908..debdc802f 100644 --- a/src/blas/backends/mklcpu/mklcpu_level3.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level3.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklcpu/mklcpu_wrappers.cpp b/src/blas/backends/mklcpu/mklcpu_wrappers.cpp index bdbcfbc4c..21dd03534 100644 --- a/src/blas/backends/mklcpu/mklcpu_wrappers.cpp +++ b/src/blas/backends/mklcpu/mklcpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "blas/function_table.hpp" -#include "oneapi/math/blas/detail/mklcpu/onemkl_blas_mklcpu.hpp" +#include "oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/blas/backends/mklgpu/CMakeLists.txt b/src/blas/backends/mklgpu/CMakeLists.txt index c971d1afd..d197f19d8 100644 --- a/src/blas/backends/mklgpu/CMakeLists.txt +++ b/src/blas/backends/mklgpu/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_blas_mklgpu) +set(LIB_NAME onemath_blas_mklgpu) set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) @@ -25,7 +25,7 @@ add_library(${LIB_OBJ} OBJECT mklgpu_level1.cpp mklgpu_level2.cpp mklgpu_level3.cpp mklgpu_batch.cpp mklgpu_extensions.cpp $<$: mklgpu_wrappers.cpp> ) -add_dependencies(onemkl_backend_libs_blas ${LIB_NAME}) +add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/blas/backends/mklgpu/mklgpu_batch.cpp b/src/blas/backends/mklgpu/mklgpu_batch.cpp index b64e75afb..5767126a8 100644 --- a/src/blas/backends/mklgpu/mklgpu_batch.cpp +++ b/src/blas/backends/mklgpu/mklgpu_batch.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp" #include "oneapi/math/exceptions.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklgpu/mklgpu_extensions.cpp b/src/blas/backends/mklgpu/mklgpu_extensions.cpp index 8c27afa8a..da9e615c4 100644 --- a/src/blas/backends/mklgpu/mklgpu_extensions.cpp +++ b/src/blas/backends/mklgpu/mklgpu_extensions.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp" #include "oneapi/math/exceptions.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklgpu/mklgpu_level1.cpp b/src/blas/backends/mklgpu/mklgpu_level1.cpp index 0031593ec..928860eb6 100644 --- a/src/blas/backends/mklgpu/mklgpu_level1.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level1.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklgpu/mklgpu_level2.cpp b/src/blas/backends/mklgpu/mklgpu_level2.cpp index 8e28da243..51cd0c119 100644 --- a/src/blas/backends/mklgpu/mklgpu_level2.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level2.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklgpu/mklgpu_level3.cpp b/src/blas/backends/mklgpu/mklgpu_level3.cpp index d4937096e..9f35affe5 100644 --- a/src/blas/backends/mklgpu/mklgpu_level3.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level3.cpp @@ -23,7 +23,7 @@ #include #endif -#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp" #include "../mkl_common/mkl_blas_backend.hpp" diff --git a/src/blas/backends/mklgpu/mklgpu_wrappers.cpp b/src/blas/backends/mklgpu/mklgpu_wrappers.cpp index 867729a5a..8ce5be8df 100644 --- a/src/blas/backends/mklgpu/mklgpu_wrappers.cpp +++ b/src/blas/backends/mklgpu/mklgpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "blas/function_table.hpp" -#include "oneapi/math/blas/detail/mklgpu/onemkl_blas_mklgpu.hpp" +#include "oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/blas/backends/netlib/CMakeLists.txt b/src/blas/backends/netlib/CMakeLists.txt index fd5275fc0..5a4bb76f2 100644 --- a/src/blas/backends/netlib/CMakeLists.txt +++ b/src/blas/backends/netlib/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_blas_netlib) +set(LIB_NAME onemath_blas_netlib) set(LIB_OBJ ${LIB_NAME}_obj) # Add third-party library @@ -29,7 +29,7 @@ set(SOURCES netlib_common.hpp ) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) -add_dependencies(onemkl_backend_libs_blas ${LIB_NAME}) +add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET ${LIB_OBJ} SOURCES ${SOURCES}) diff --git a/src/blas/backends/netlib/netlib_batch.cpp b/src/blas/backends/netlib/netlib_batch.cpp index ad6aec748..ae492ee6d 100644 --- a/src/blas/backends/netlib/netlib_batch.cpp +++ b/src/blas/backends/netlib/netlib_batch.cpp @@ -25,7 +25,7 @@ #include "netlib_common.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/netlib/netlib_common.hpp b/src/blas/backends/netlib/netlib_common.hpp index a9208273e..d202d0faf 100644 --- a/src/blas/backends/netlib/netlib_common.hpp +++ b/src/blas/backends/netlib/netlib_common.hpp @@ -29,7 +29,7 @@ #include "cblas.h" -#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp" #include "oneapi/math/types.hpp" #define GET_MULTI_PTR template get_multi_ptr().get_raw() @@ -42,7 +42,7 @@ namespace netlib { typedef enum { CblasFixOffset = 101, CblasColOffset = 102, CblasRowOffset = 103 } CBLAS_OFFSET; /** - * Helper methods for converting between onemkl types and their BLAS + * Helper methods for converting between onemath types and their BLAS * equivalents. */ diff --git a/src/blas/backends/netlib/netlib_extensions.cpp b/src/blas/backends/netlib/netlib_extensions.cpp index a6f265c63..888129e44 100644 --- a/src/blas/backends/netlib/netlib_extensions.cpp +++ b/src/blas/backends/netlib/netlib_extensions.cpp @@ -24,7 +24,7 @@ #endif #include "netlib_common.hpp" -#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp" #include "oneapi/math/exceptions.hpp" namespace oneapi { diff --git a/src/blas/backends/netlib/netlib_level1.cpp b/src/blas/backends/netlib/netlib_level1.cpp index b73fc1a4e..d067c5c15 100644 --- a/src/blas/backends/netlib/netlib_level1.cpp +++ b/src/blas/backends/netlib/netlib_level1.cpp @@ -25,7 +25,7 @@ #include "netlib_common.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp" inline float abs_val(float val) { return std::abs(val); diff --git a/src/blas/backends/netlib/netlib_level2.cpp b/src/blas/backends/netlib/netlib_level2.cpp index 2a893e41c..f455d609f 100644 --- a/src/blas/backends/netlib/netlib_level2.cpp +++ b/src/blas/backends/netlib/netlib_level2.cpp @@ -24,7 +24,7 @@ #endif #include "netlib_common.hpp" -#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/netlib/netlib_level3.cpp b/src/blas/backends/netlib/netlib_level3.cpp index 04d559e69..952d16690 100644 --- a/src/blas/backends/netlib/netlib_level3.cpp +++ b/src/blas/backends/netlib/netlib_level3.cpp @@ -25,7 +25,7 @@ #include "netlib_common.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/netlib/netlib_wrappers.cpp b/src/blas/backends/netlib/netlib_wrappers.cpp index c99875b0f..08de7eb39 100644 --- a/src/blas/backends/netlib/netlib_wrappers.cpp +++ b/src/blas/backends/netlib/netlib_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "blas/function_table.hpp" -#include "oneapi/math/blas/detail/netlib/onemkl_blas_netlib.hpp" +#include "oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp" #define WRAPPER_VERSION 1 diff --git a/src/blas/backends/portblas/CMakeLists.txt b/src/blas/backends/portblas/CMakeLists.txt index 3256554a3..f1f0cef8b 100644 --- a/src/blas/backends/portblas/CMakeLists.txt +++ b/src/blas/backends/portblas/CMakeLists.txt @@ -17,7 +17,7 @@ # #========================================================================= -set(LIB_NAME onemkl_blas_portblas) +set(LIB_NAME onemath_blas_portblas) set(LIB_OBJ ${LIB_NAME}_obj) if(NOT DEFINED PORTBLAS_TUNING_TARGET) @@ -177,7 +177,7 @@ set(SOURCES $<$: portblas_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) -add_dependencies(onemkl_backend_libs_blas ${LIB_NAME}) +add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET ${LIB_OBJ} SOURCES ${SOURCES}) diff --git a/src/blas/backends/portblas/portblas_batch.cpp b/src/blas/backends/portblas/portblas_batch.cpp index ee5a47d9d..c4d2b427b 100644 --- a/src/blas/backends/portblas/portblas_batch.cpp +++ b/src/blas/backends/portblas/portblas_batch.cpp @@ -25,7 +25,7 @@ #include "portblas_common.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_common.hpp b/src/blas/backends/portblas/portblas_common.hpp index 709bfdd0c..58ef28d3c 100644 --- a/src/blas/backends/portblas/portblas_common.hpp +++ b/src/blas/backends/portblas/portblas_common.hpp @@ -53,9 +53,9 @@ using sycl_complex_t = sycl::ext::oneapi::experimental::complex; template struct portblas_type; -#define DEF_PORTBLAS_TYPE(onemkl_t, portblas_t) \ +#define DEF_PORTBLAS_TYPE(onemath_t, portblas_t) \ template <> \ - struct portblas_type { \ + struct portblas_type { \ using type = portblas_t; \ }; diff --git a/src/blas/backends/portblas/portblas_level1_double.cpp b/src/blas/backends/portblas/portblas_level1_double.cpp index 070633a9e..9f7bd9428 100644 --- a/src/blas/backends/portblas/portblas_level1_double.cpp +++ b/src/blas/backends/portblas/portblas_level1_double.cpp @@ -25,7 +25,7 @@ #include "portblas_common.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level1_float.cpp b/src/blas/backends/portblas/portblas_level1_float.cpp index 244d9b4d3..b3f9cc5cf 100644 --- a/src/blas/backends/portblas/portblas_level1_float.cpp +++ b/src/blas/backends/portblas/portblas_level1_float.cpp @@ -25,7 +25,7 @@ #include "portblas_common.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level2_double.cpp b/src/blas/backends/portblas/portblas_level2_double.cpp index 9d67b9bb5..67a606cb1 100644 --- a/src/blas/backends/portblas/portblas_level2_double.cpp +++ b/src/blas/backends/portblas/portblas_level2_double.cpp @@ -25,7 +25,7 @@ #include "portblas_common.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level2_float.cpp b/src/blas/backends/portblas/portblas_level2_float.cpp index 8a5579595..ccb422f39 100644 --- a/src/blas/backends/portblas/portblas_level2_float.cpp +++ b/src/blas/backends/portblas/portblas_level2_float.cpp @@ -25,7 +25,7 @@ #include "portblas_common.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level3_bfloat16.cpp b/src/blas/backends/portblas/portblas_level3_bfloat16.cpp index afb39b623..f0ff06452 100644 --- a/src/blas/backends/portblas/portblas_level3_bfloat16.cpp +++ b/src/blas/backends/portblas/portblas_level3_bfloat16.cpp @@ -24,7 +24,7 @@ #endif #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level3_double.cpp b/src/blas/backends/portblas/portblas_level3_double.cpp index 1b3d4f31b..1ca4beb4b 100644 --- a/src/blas/backends/portblas/portblas_level3_double.cpp +++ b/src/blas/backends/portblas/portblas_level3_double.cpp @@ -25,7 +25,7 @@ #include "portblas_common.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level3_float.cpp b/src/blas/backends/portblas/portblas_level3_float.cpp index d0d461e16..d0168f7d6 100644 --- a/src/blas/backends/portblas/portblas_level3_float.cpp +++ b/src/blas/backends/portblas/portblas_level3_float.cpp @@ -25,7 +25,7 @@ #include "portblas_common.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_level3_half.cpp b/src/blas/backends/portblas/portblas_level3_half.cpp index 73eb5cd35..c262f9e5b 100644 --- a/src/blas/backends/portblas/portblas_level3_half.cpp +++ b/src/blas/backends/portblas/portblas_level3_half.cpp @@ -24,7 +24,7 @@ #endif #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { namespace mkl { diff --git a/src/blas/backends/portblas/portblas_wrappers.cpp b/src/blas/backends/portblas/portblas_wrappers.cpp index 1afb250fd..69ec16016 100644 --- a/src/blas/backends/portblas/portblas_wrappers.cpp +++ b/src/blas/backends/portblas/portblas_wrappers.cpp @@ -4,7 +4,7 @@ #include "blas/function_table.hpp" -#include "oneapi/math/blas/detail/portblas/onemkl_blas_portblas.hpp" +#include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" #define WRAPPER_VERSION 1 diff --git a/src/blas/backends/rocblas/CMakeLists.txt b/src/blas/backends/rocblas/CMakeLists.txt index 76dc126ad..f8d42b312 100644 --- a/src/blas/backends/rocblas/CMakeLists.txt +++ b/src/blas/backends/rocblas/CMakeLists.txt @@ -19,7 +19,7 @@ # #========================================================================= -set(LIB_NAME onemkl_blas_rocblas) +set(LIB_NAME onemath_blas_rocblas) set(LIB_OBJ ${LIB_NAME}_obj) find_package(hip REQUIRED) find_package(rocblas REQUIRED) @@ -35,7 +35,7 @@ set(SOURCES rocblas_level1.cpp $<$: rocblas_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) -add_dependencies(onemkl_backend_libs_blas ${LIB_NAME}) +add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/blas/backends/rocblas/rocblas_batch.cpp b/src/blas/backends/rocblas/rocblas_batch.cpp index f267d1f1a..5e8be4f29 100644 --- a/src/blas/backends/rocblas/rocblas_batch.cpp +++ b/src/blas/backends/rocblas/rocblas_batch.cpp @@ -23,7 +23,7 @@ #include "rocblas_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp" // Helper Functions @@ -81,7 +81,7 @@ inline void copy_batch(Func func, sycl::queue &queue, int64_t n, sycl::buffer(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = sc.get_mem(x_acc); @@ -117,7 +117,7 @@ inline void axpy_batch(Func func, sycl::queue &queue, int64_t n, T alpha, sycl:: queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = sc.get_mem(x_acc); @@ -156,7 +156,7 @@ inline void gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -198,7 +198,7 @@ inline void dgmm_batch(Func func, sycl::queue &queue, side left_right, int64_t m auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -245,7 +245,7 @@ inline void gemm_batch_impl(sycl::queue &queue, transpose transa, transpose tran auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -314,7 +314,7 @@ inline void trsm_batch(Func func, sycl::queue &queue, side left_right, uplo uppe queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -356,7 +356,7 @@ inline void syrk_batch(Func func, sycl::queue &queue, uplo upper_lower, transpos queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -400,7 +400,7 @@ inline void omatcopy_batch(Func func, sycl::queue &queue, transpose trans, int64 queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -467,7 +467,7 @@ inline void omatadd_batch(Func func, sycl::queue &queue, transpose transa, trans auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -512,7 +512,7 @@ inline sycl::event copy_batch(Func func, sycl::queue &queue, int64_t *n, const T auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; @@ -554,7 +554,7 @@ inline sycl::event copy_batch(Func func, sycl::queue &queue, int64_t n, const T auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); @@ -594,7 +594,7 @@ inline sycl::event axpy_batch(Func func, sycl::queue &queue, int64_t *n, T *alph auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; @@ -637,7 +637,7 @@ inline sycl::event axpy_batch(Func func, sycl::queue &queue, int64_t n, T alpha, auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); @@ -677,7 +677,7 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose trans, in auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -722,7 +722,7 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, i auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; @@ -769,7 +769,7 @@ inline sycl::event dgmm_batch(Func func, sycl::queue &queue, side left_right, in auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -812,7 +812,7 @@ inline sycl::event dgmm_batch(Func func, sycl::queue &queue, side *left_right, i auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; @@ -866,7 +866,7 @@ inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose tra rocblas_gemm_flags flags = rocblas_gemm_flags_none; auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -944,7 +944,7 @@ inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, tr rocblas_gemm_flags flags = rocblas_gemm_flags_none; auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; @@ -1019,7 +1019,7 @@ inline sycl::event trsm_batch(Func func, sycl::queue &queue, side left_right, up auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1064,7 +1064,7 @@ inline sycl::event trsm_batch(Func func, sycl::queue &queue, side *left_right, u auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; @@ -1115,7 +1115,7 @@ inline sycl::event syrk_batch(Func func, sycl::queue &queue, uplo *upper_lower, auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; @@ -1162,7 +1162,7 @@ inline sycl::event syrk_batch(Func func, sycl::queue &queue, uplo upper_lower, t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1210,7 +1210,7 @@ inline sycl::event omatcopy_batch(Func func, sycl::queue &queue, transpose trans auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1279,7 +1279,7 @@ inline sycl::event omatadd_batch(Func func, sycl::queue &queue, transpose transa auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1325,7 +1325,7 @@ inline sycl::event omatcopy_batch(Func func, sycl::queue &queue, transpose *tran auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; diff --git a/src/blas/backends/rocblas/rocblas_extensions.cpp b/src/blas/backends/rocblas/rocblas_extensions.cpp index 334a3708a..7d4343f4d 100644 --- a/src/blas/backends/rocblas/rocblas_extensions.cpp +++ b/src/blas/backends/rocblas/rocblas_extensions.cpp @@ -23,7 +23,7 @@ #include "rocblas_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp" namespace oneapi { namespace mkl { @@ -103,7 +103,7 @@ inline void omatcopy(Func func, sycl::queue &queue, transpose trans, int64_t m, queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -183,7 +183,7 @@ inline void omatadd(Func func, sycl::queue &queue, transpose transa, transpose t auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -289,7 +289,7 @@ inline sycl::event omatcopy(Func func, sycl::queue &queue, transpose trans, int6 auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -375,7 +375,7 @@ inline sycl::event omatadd(Func func, sycl::queue &queue, transpose transa, tran auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); diff --git a/src/blas/backends/rocblas/rocblas_level1.cpp b/src/blas/backends/rocblas/rocblas_level1.cpp index 9d3adf46c..69f0e2883 100644 --- a/src/blas/backends/rocblas/rocblas_level1.cpp +++ b/src/blas/backends/rocblas/rocblas_level1.cpp @@ -23,7 +23,7 @@ #include "rocblas_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp" namespace oneapi { namespace mkl { @@ -43,7 +43,7 @@ inline void asum(Func func, sycl::queue &queue, int64_t n, sycl::buffer & queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host // when the data is on buffer, it must be set to @@ -86,7 +86,7 @@ inline void scal(Func func, sycl::queue &queue, int64_t n, T1 a, sycl::buffer(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = sc.get_mem(x_acc); rocblas_status err; @@ -119,7 +119,7 @@ inline void axpy(Func func, sycl::queue &queue, int64_t n, T alpha, sycl::buffer queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = sc.get_mem(x_acc); @@ -177,7 +177,7 @@ inline void rotg(Func func, sycl::queue &queue, sycl::buffer &a, sycl::bu auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto s_acc = s.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host // when the data is on buffer, it must be set to @@ -222,7 +222,7 @@ inline void rotm(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); auto param_acc = param.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -264,7 +264,7 @@ inline void copy(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = sc.get_mem(x_acc); @@ -298,7 +298,7 @@ inline void dot(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -351,7 +351,7 @@ inline void rot(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host // when the data is on buffer, it must be set to @@ -390,7 +390,7 @@ void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, auto x_acc = x.get_access(cgh); auto y_acc = y.get_access(cgh); auto res_acc = result.get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -428,7 +428,7 @@ inline void rotmg(Func func, sycl::queue &queue, sycl::buffer &d1, sycl::b auto x1_acc = x1.template get_access(cgh); auto y1_acc = y1_buff.template get_access(cgh); auto param_acc = param.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -480,7 +480,7 @@ inline void iamax(Func func, sycl::queue &queue, int64_t n, sycl::buffer & queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto int_res_acc = int_res_buff.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -532,7 +532,7 @@ inline void swap(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = sc.get_mem(x_acc); @@ -573,7 +573,7 @@ inline void iamin(Func func, sycl::queue &queue, int64_t n, sycl::buffer & queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto int_res_acc = int_res_buff.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -626,7 +626,7 @@ inline void nrm2(Func func, sycl::queue &queue, int64_t n, sycl::buffer & queue.submit([&](sycl::handler &cgh) { auto x_acc = x.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -672,7 +672,7 @@ inline sycl::event asum(Func func, sycl::queue &queue, int64_t n, const T1 *x, c auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); @@ -710,7 +710,7 @@ inline sycl::event scal(Func func, sycl::queue &queue, int64_t n, T1 a, T2 *x, i auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); @@ -746,7 +746,7 @@ inline sycl::event axpy(Func func, sycl::queue &queue, int64_t n, T alpha, const auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); @@ -804,7 +804,7 @@ inline sycl::event rotg(Func func, sycl::queue &queue, T1 *a, T1 *b, T2 *c, T1 * auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -840,7 +840,7 @@ inline sycl::event rotm(Func func, sycl::queue &queue, int64_t n, T *x, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); @@ -873,7 +873,7 @@ inline sycl::event copy(Func func, sycl::queue &queue, int64_t n, const T *x, in auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); @@ -908,7 +908,7 @@ inline sycl::event dot(Func func, sycl::queue &queue, int64_t n, const T *x, con auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); @@ -953,7 +953,7 @@ inline sycl::event rot(Func func, sycl::queue &queue, int64_t n, T1 *x, const in auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); @@ -989,7 +989,7 @@ sycl::event sdsdot(sycl::queue &queue, int64_t n, float sb, const float *x, int6 // rocBLAS does not support sdot so we need to mimic sdot. auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); @@ -1012,7 +1012,7 @@ inline sycl::event rotmg(Func func, sycl::queue &queue, T *d1, T *d2, T *x1, T y auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto d1_ = reinterpret_cast(d1); @@ -1055,7 +1055,7 @@ inline sycl::event iamax(Func func, sycl::queue &queue, int64_t n, const T *x, c auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); auto x_ = reinterpret_cast(x); @@ -1094,7 +1094,7 @@ inline sycl::event swap(Func func, sycl::queue &queue, int64_t n, T *x, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto x_ = reinterpret_cast(x); @@ -1136,7 +1136,7 @@ inline sycl::event iamin(Func func, sycl::queue &queue, int64_t n, const T *x, c auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); @@ -1177,7 +1177,7 @@ inline sycl::event nrm2(Func func, sycl::queue &queue, int64_t n, const T1 *x, c auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); diff --git a/src/blas/backends/rocblas/rocblas_level2.cpp b/src/blas/backends/rocblas/rocblas_level2.cpp index 009100dc6..c887fbd61 100644 --- a/src/blas/backends/rocblas/rocblas_level2.cpp +++ b/src/blas/backends/rocblas/rocblas_level2.cpp @@ -23,7 +23,7 @@ #include "rocblas_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp" // Helper Functions @@ -92,7 +92,7 @@ inline void gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int6 auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -131,7 +131,7 @@ inline void gbmv(Func func, sycl::queue &queue, transpose trans, int64_t m, int6 auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -170,7 +170,7 @@ inline void ger(Func func, sycl::queue &queue, int64_t m, int64_t n, T alpha, sy auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -210,7 +210,7 @@ inline void hbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -247,7 +247,7 @@ inline void hemv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T a auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -284,7 +284,7 @@ inline void her(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, Scal queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -319,7 +319,7 @@ inline void her2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T a auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -355,7 +355,7 @@ inline void hpmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T a auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -391,7 +391,7 @@ inline void hpr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, Scal queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -425,7 +425,7 @@ inline void hpr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T a auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -461,7 +461,7 @@ inline void sbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -498,7 +498,7 @@ inline void symv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T a auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -533,7 +533,7 @@ inline void syr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T al queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -570,7 +570,7 @@ inline void syr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T a auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -609,7 +609,7 @@ inline void spmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T a auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -644,7 +644,7 @@ inline void spr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T al queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -678,7 +678,7 @@ inline void spr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T a auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -713,7 +713,7 @@ inline void tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -750,7 +750,7 @@ inline void tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -786,7 +786,7 @@ inline void tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -821,7 +821,7 @@ inline void tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -857,7 +857,7 @@ inline void trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -893,7 +893,7 @@ inline void trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -930,7 +930,7 @@ inline sycl::event gemv(Func func, sycl::queue &queue, transpose trans, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -971,7 +971,7 @@ inline sycl::event gbmv(Func func, sycl::queue &queue, transpose trans, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1012,7 +1012,7 @@ inline sycl::event ger(Func func, sycl::queue &queue, int64_t m, int64_t n, T al auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1052,7 +1052,7 @@ inline sycl::event hbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1090,7 +1090,7 @@ inline sycl::event hemv(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1129,7 +1129,7 @@ inline sycl::event her(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1164,7 +1164,7 @@ inline sycl::event her2(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1201,7 +1201,7 @@ inline sycl::event hpmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1240,7 +1240,7 @@ inline sycl::event hpr(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1275,7 +1275,7 @@ inline sycl::event hpr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1312,7 +1312,7 @@ inline sycl::event sbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1350,7 +1350,7 @@ inline sycl::event symv(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1388,7 +1388,7 @@ inline sycl::event syr(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1426,7 +1426,7 @@ inline sycl::event syr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1466,7 +1466,7 @@ inline sycl::event spmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1503,7 +1503,7 @@ inline sycl::event spr(Func func, sycl::queue &queue, uplo upper_lower, int64_t overflow_check(n, incx); auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1537,7 +1537,7 @@ inline sycl::event spr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1574,7 +1574,7 @@ inline sycl::event tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpo auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1613,7 +1613,7 @@ inline sycl::event tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpo auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1652,7 +1652,7 @@ inline sycl::event tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpo auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1691,7 +1691,7 @@ inline sycl::event tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpo auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1730,7 +1730,7 @@ inline sycl::event trmv(Func func, sycl::queue &queue, uplo upper_lower, transpo auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -1769,7 +1769,7 @@ inline sycl::event trsv(Func func, sycl::queue &queue, uplo upper_lower, transpo auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); diff --git a/src/blas/backends/rocblas/rocblas_level3.cpp b/src/blas/backends/rocblas/rocblas_level3.cpp index 5b9f57411..085df77ed 100644 --- a/src/blas/backends/rocblas/rocblas_level3.cpp +++ b/src/blas/backends/rocblas/rocblas_level3.cpp @@ -23,7 +23,7 @@ #include "rocblas_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp" namespace oneapi { namespace mkl { @@ -44,7 +44,7 @@ inline void gemm(Func func, sycl::queue &queue, transpose transa, transpose tran auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -91,7 +91,7 @@ inline void gemm_ex(Func func, DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -140,7 +140,7 @@ inline void symm(Func func, sycl::queue &queue, side left_right, uplo upper_lowe auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -180,7 +180,7 @@ inline void hemm(Func func, sycl::queue &queue, side left_right, uplo upper_lowe auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -217,7 +217,7 @@ inline void syrk(Func func, sycl::queue &queue, uplo upper_lower, transpose tran queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -255,7 +255,7 @@ inline void herk(Func func, sycl::queue &queue, uplo upper_lower, transpose tran queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -291,7 +291,7 @@ inline void syr2k(Func func, sycl::queue &queue, uplo upper_lower, transpose tra auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -333,7 +333,7 @@ inline void her2k(Func func, sycl::queue &queue, uplo upper_lower, transpose tra auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -375,7 +375,7 @@ inline void trmm(Func func, sycl::queue &queue, side left_right, uplo upper_lowe queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -421,7 +421,7 @@ inline void trsm(Func func, sycl::queue &queue, side left_right, uplo upper_lowe queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -462,7 +462,7 @@ inline sycl::event gemm(Func func, sycl::queue &queue, transpose transa, transpo auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -509,7 +509,7 @@ inline sycl::event gemm_ex(Func func, DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -559,7 +559,7 @@ inline sycl::event symm(Func func, sycl::queue &queue, side left_right, uplo upp auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -600,7 +600,7 @@ inline sycl::event hemm(Func func, sycl::queue &queue, side left_right, uplo upp auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -639,7 +639,7 @@ inline sycl::event syrk(Func func, sycl::queue &queue, uplo upper_lower, transpo auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -680,7 +680,7 @@ inline sycl::event herk(Func func, sycl::queue &queue, uplo upper_lower, transpo auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -719,7 +719,7 @@ inline sycl::event syr2k(Func func, sycl::queue &queue, uplo upper_lower, transp auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -762,7 +762,7 @@ inline sycl::event her2k(Func func, sycl::queue &queue, uplo upper_lower, transp auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -806,7 +806,7 @@ inline sycl::event trmm(Func func, sycl::queue &queue, side left_right, uplo upp auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -854,7 +854,7 @@ inline sycl::event trsm(Func func, sycl::queue &queue, side left_right, uplo upp auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); diff --git a/src/blas/backends/rocblas/rocblas_task.hpp b/src/blas/backends/rocblas/rocblas_task.hpp index e9d8386c2..588c9029b 100644 --- a/src/blas/backends/rocblas/rocblas_task.hpp +++ b/src/blas/backends/rocblas/rocblas_task.hpp @@ -73,7 +73,7 @@ static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { } #endif template -static inline void onemkl_rocblas_host_task(H &cgh, sycl::queue queue, F f) { +static inline void onemath_rocblas_host_task(H &cgh, sycl::queue queue, F f) { (void)host_task_internal(cgh, queue, f); } diff --git a/src/blas/backends/rocblas/rocblas_wrappers.cpp b/src/blas/backends/rocblas/rocblas_wrappers.cpp index f7122cae5..8d4915ea0 100644 --- a/src/blas/backends/rocblas/rocblas_wrappers.cpp +++ b/src/blas/backends/rocblas/rocblas_wrappers.cpp @@ -19,7 +19,7 @@ * **************************************************************************/ #include "blas/function_table.hpp" -#include "oneapi/math/blas/detail/rocblas/onemkl_blas_rocblas.hpp" +#include "oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp" #define WRAPPER_VERSION 1 diff --git a/src/dft/CMakeLists.txt b/src/dft/CMakeLists.txt index e3b373645..d4c44747a 100644 --- a/src/dft/CMakeLists.txt +++ b/src/dft/CMakeLists.txt @@ -22,29 +22,29 @@ add_subdirectory(backends) # Recipe for DFT loader object if(BUILD_SHARED_LIBS) -add_library(onemkl_dft OBJECT) -target_sources(onemkl_dft PRIVATE backends/descriptor.cpp dft_loader.cpp) -target_include_directories(onemkl_dft +add_library(onemath_dft OBJECT) +target_sources(onemath_dft PRIVATE backends/descriptor.cpp dft_loader.cpp) +target_include_directories(onemath_dft PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMKL_GENERATED_INCLUDE_PATH} - $ + $ ) -target_compile_options(onemkl_dft PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(onemath_dft PRIVATE ${ONEMKL_BUILD_COPT}) -set_target_properties(onemkl_dft PROPERTIES +set_target_properties(onemath_dft PROPERTIES POSITION_INDEPENDENT_CODE ON ) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) - add_sycl_to_target(TARGET onemkl_dft SOURCES backends/descriptor.cxx dft_loader.cpp) + add_sycl_to_target(TARGET onemath_dft SOURCES backends/descriptor.cxx dft_loader.cpp) else() - target_link_libraries(onemkl_dft PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(onemath_dft PUBLIC ONEMKL::SYCL::SYCL) endif() include(WarningsUtils) -target_link_libraries(onemkl_dft PRIVATE onemkl_warnings) +target_link_libraries(onemath_dft PRIVATE onemath_warnings) endif() diff --git a/src/dft/backends/CMakeLists.txt b/src/dft/backends/CMakeLists.txt index b03a63e8a..262bfe3cc 100644 --- a/src/dft/backends/CMakeLists.txt +++ b/src/dft/backends/CMakeLists.txt @@ -17,8 +17,8 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -add_custom_target(onemkl_backend_libs_dft) -add_dependencies(onemkl_backend_libs onemkl_backend_libs_dft) +add_custom_target(onemath_backend_libs_dft) +add_dependencies(onemath_backend_libs onemath_backend_libs_dft) if(ENABLE_MKLGPU_BACKEND) add_subdirectory(mklgpu) diff --git a/src/dft/backends/cufft/CMakeLists.txt b/src/dft/backends/cufft/CMakeLists.txt index 365642766..b315c5f4f 100644 --- a/src/dft/backends/cufft/CMakeLists.txt +++ b/src/dft/backends/cufft/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_dft_cufft) +set(LIB_NAME onemath_dft_cufft) set(LIB_OBJ ${LIB_NAME}_obj) @@ -29,7 +29,7 @@ add_library(${LIB_OBJ} OBJECT backward.cpp $<$: cufft_wrappers.cpp> ) -add_dependencies(onemkl_backend_libs_dft ${LIB_NAME}) +add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_OBJ} PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} diff --git a/src/dft/backends/cufft/backward.cpp b/src/dft/backends/cufft/backward.cpp index 86d4a8b81..6453b3301 100644 --- a/src/dft/backends/cufft/backward.cpp +++ b/src/dft/backends/cufft/backward.cpp @@ -26,7 +26,7 @@ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/dft/detail/commit_impl.hpp" -#include "oneapi/math/dft/detail/cufft/onemkl_dft_cufft.hpp" +#include "oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp" #include "oneapi/math/dft/types.hpp" #include "execute_helper.hpp" diff --git a/src/dft/backends/cufft/commit.cpp b/src/dft/backends/cufft/commit.cpp index 2de2c882b..ad6f4abe4 100644 --- a/src/dft/backends/cufft/commit.cpp +++ b/src/dft/backends/cufft/commit.cpp @@ -31,7 +31,7 @@ #include "oneapi/math/dft/detail/commit_impl.hpp" #include "oneapi/math/dft/detail/descriptor_impl.hpp" -#include "oneapi/math/dft/detail/cufft/onemkl_dft_cufft.hpp" +#include "oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp" #include "oneapi/math/dft/types.hpp" #include "../stride_helper.hpp" diff --git a/src/dft/backends/cufft/cufft_wrappers.cpp b/src/dft/backends/cufft/cufft_wrappers.cpp index 3401859db..fed14d6fc 100644 --- a/src/dft/backends/cufft/cufft_wrappers.cpp +++ b/src/dft/backends/cufft/cufft_wrappers.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/math/dft/detail/cufft/onemkl_dft_cufft.hpp" +#include "oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp" #include "dft/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/dft/backends/cufft/descriptor.cpp b/src/dft/backends/cufft/descriptor.cpp index f61dd1b91..48e22b7e7 100644 --- a/src/dft/backends/cufft/descriptor.cpp +++ b/src/dft/backends/cufft/descriptor.cpp @@ -20,7 +20,7 @@ #include "oneapi/math/dft/descriptor.hpp" #include "../../descriptor.cxx" -#include "oneapi/math/dft/detail/cufft/onemkl_dft_cufft.hpp" +#include "oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp" namespace oneapi::mkl::dft::detail { diff --git a/src/dft/backends/cufft/forward.cpp b/src/dft/backends/cufft/forward.cpp index e104e1602..5f75752c2 100644 --- a/src/dft/backends/cufft/forward.cpp +++ b/src/dft/backends/cufft/forward.cpp @@ -27,7 +27,7 @@ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/dft/detail/commit_impl.hpp" -#include "oneapi/math/dft/detail/cufft/onemkl_dft_cufft.hpp" +#include "oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp" #include "oneapi/math/dft/types.hpp" #include "execute_helper.hpp" diff --git a/src/dft/backends/mklcpu/CMakeLists.txt b/src/dft/backends/mklcpu/CMakeLists.txt index 6d0f1276d..2d1617854 100644 --- a/src/dft/backends/mklcpu/CMakeLists.txt +++ b/src/dft/backends/mklcpu/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_dft_mklcpu) +set(LIB_NAME onemath_dft_mklcpu) set(LIB_OBJ ${LIB_NAME}_obj) include(WarningsUtils) @@ -30,7 +30,7 @@ add_library(${LIB_OBJ} OBJECT backward.cpp $<$: mkl_dft_cpu_wrappers.cpp> ) -add_dependencies(onemkl_backend_libs_dft ${LIB_NAME}) +add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_OBJ} PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} @@ -54,13 +54,13 @@ if(TARGET MKL::MKL_SYCL::DFT) target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL PUBLIC MKL::MKL_SYCL::DFT - PRIVATE onemkl_warnings + PRIVATE onemath_warnings ) else() target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL PUBLIC MKL::MKL_DPCPP - PRIVATE onemkl_warnings + PRIVATE onemath_warnings ) endif() diff --git a/src/dft/backends/mklcpu/backward.cpp b/src/dft/backends/mklcpu/backward.cpp index 77365e1d0..b7a517849 100644 --- a/src/dft/backends/mklcpu/backward.cpp +++ b/src/dft/backends/mklcpu/backward.cpp @@ -26,7 +26,7 @@ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/dft/descriptor.hpp" -#include "oneapi/math/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" +#include "oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp" #include "oneapi/math/dft/types.hpp" #include "dft/backends/mklcpu/commit_derived_impl.hpp" diff --git a/src/dft/backends/mklcpu/commit.cpp b/src/dft/backends/mklcpu/commit.cpp index be9cf375c..381bfbbeb 100644 --- a/src/dft/backends/mklcpu/commit.cpp +++ b/src/dft/backends/mklcpu/commit.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/dft/types.hpp" #include "oneapi/math/dft/descriptor.hpp" -#include "oneapi/math/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" +#include "oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp" #include "oneapi/math/dft/detail/commit_impl.hpp" diff --git a/src/dft/backends/mklcpu/descriptor.cpp b/src/dft/backends/mklcpu/descriptor.cpp index 95a49229b..9a6acd7cf 100644 --- a/src/dft/backends/mklcpu/descriptor.cpp +++ b/src/dft/backends/mklcpu/descriptor.cpp @@ -20,7 +20,7 @@ #include "oneapi/math/dft/descriptor.hpp" #include "../../descriptor.cxx" -#include "oneapi/math/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" +#include "oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp" namespace oneapi::mkl::dft::detail { diff --git a/src/dft/backends/mklcpu/forward.cpp b/src/dft/backends/mklcpu/forward.cpp index 7a1ac1524..62b5755d1 100644 --- a/src/dft/backends/mklcpu/forward.cpp +++ b/src/dft/backends/mklcpu/forward.cpp @@ -26,7 +26,7 @@ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/dft/descriptor.hpp" -#include "oneapi/math/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" +#include "oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp" #include "oneapi/math/dft/types.hpp" #include "dft/backends/mklcpu/commit_derived_impl.hpp" diff --git a/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp b/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp index ad0687b5b..10f86017c 100644 --- a/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp +++ b/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/math/dft/detail/mklcpu/onemkl_dft_mklcpu.hpp" +#include "oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp" #include "dft/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/dft/backends/mklgpu/CMakeLists.txt b/src/dft/backends/mklgpu/CMakeLists.txt index 8ec322de8..c1a0c2c60 100644 --- a/src/dft/backends/mklgpu/CMakeLists.txt +++ b/src/dft/backends/mklgpu/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_dft_mklgpu) +set(LIB_NAME onemath_dft_mklgpu) set(LIB_OBJ ${LIB_NAME}_obj) include(WarningsUtils) @@ -30,7 +30,7 @@ add_library(${LIB_OBJ} OBJECT backward.cpp $<$: mkl_dft_gpu_wrappers.cpp> ) -add_dependencies(onemkl_backend_libs_dft ${LIB_NAME}) +add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_NAME} PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} @@ -56,13 +56,13 @@ if(TARGET MKL::MKL_SYCL::DFT) target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL PUBLIC MKL::MKL_SYCL::DFT - PRIVATE onemkl_warnings + PRIVATE onemath_warnings ) else() target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL PUBLIC MKL::MKL_DPCPP - PRIVATE onemkl_warnings + PRIVATE onemath_warnings ) endif() diff --git a/src/dft/backends/mklgpu/backward.cpp b/src/dft/backends/mklgpu/backward.cpp index 1811cb65d..5ca2d68ab 100644 --- a/src/dft/backends/mklgpu/backward.cpp +++ b/src/dft/backends/mklgpu/backward.cpp @@ -25,7 +25,7 @@ #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" +#include "oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp" #include "oneapi/math/dft/detail/descriptor_impl.hpp" #include "mklgpu_helpers.hpp" diff --git a/src/dft/backends/mklgpu/commit.cpp b/src/dft/backends/mklgpu/commit.cpp index a70c5223d..7bbdc7bec 100644 --- a/src/dft/backends/mklgpu/commit.cpp +++ b/src/dft/backends/mklgpu/commit.cpp @@ -30,7 +30,7 @@ #include "oneapi/math/dft/detail/commit_impl.hpp" #include "oneapi/math/dft/detail/types_impl.hpp" #include "oneapi/math/dft/detail/descriptor_impl.hpp" -#include "oneapi/math/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" +#include "oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp" #include "dft/backends/mklgpu/mklgpu_helpers.hpp" #include "../stride_helper.hpp" @@ -170,7 +170,7 @@ class mklgpu_commit final : public dft::detail::commit_impl { void set_value(mklgpu_descriptor_t& desc, const dft::detail::dft_values& config, bool assume_fwd_dft, dft::detail::stride_api stride_choice) { - using onemkl_param = dft::detail::config_param; + using onemath_param = dft::detail::config_param; using backend_param = dft::config_param; // The following are read-only: @@ -180,15 +180,15 @@ class mklgpu_commit final : public dft::detail::commit_impl { desc.set_value(backend_param::BACKWARD_SCALE, config.bwd_scale); desc.set_value(backend_param::NUMBER_OF_TRANSFORMS, config.number_of_transforms); desc.set_value(backend_param::COMPLEX_STORAGE, - to_mklgpu(config.complex_storage)); + to_mklgpu(config.complex_storage)); if (config.real_storage != dft::detail::config_value::REAL_REAL) { throw mkl::invalid_argument("dft/backends/mklgpu", "commit", "MKLGPU only supports real-real real storage."); } desc.set_value(backend_param::CONJUGATE_EVEN_STORAGE, - to_mklgpu(config.conj_even_storage)); + to_mklgpu(config.conj_even_storage)); desc.set_value(backend_param::PLACEMENT, - to_mklgpu(config.placement)); + to_mklgpu(config.placement)); if (stride_choice == dft::detail::stride_api::FB_STRIDES) { if (config.fwd_strides[0] != 0 || config.fwd_strides[0] != 0) { @@ -217,7 +217,7 @@ class mklgpu_commit final : public dft::detail::commit_impl { if (config.workspace_placement == dft::detail::config_value::WORKSPACE_EXTERNAL) { // Setting WORKSPACE_INTERNAL (default) causes FFT_INVALID_DESCRIPTOR. desc.set_value(backend_param::WORKSPACE, - to_mklgpu_config_value( + to_mklgpu_config_value( config.workspace_placement)); } // Setting the ordering causes an FFT_INVALID_DESCRIPTOR. Check that default is used: diff --git a/src/dft/backends/mklgpu/descriptor.cpp b/src/dft/backends/mklgpu/descriptor.cpp index bd129b4ff..ca25a13ec 100644 --- a/src/dft/backends/mklgpu/descriptor.cpp +++ b/src/dft/backends/mklgpu/descriptor.cpp @@ -20,7 +20,7 @@ #include "oneapi/math/dft/descriptor.hpp" #include "../../descriptor.cxx" -#include "oneapi/math/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" +#include "oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp" namespace oneapi::mkl::dft::detail { diff --git a/src/dft/backends/mklgpu/forward.cpp b/src/dft/backends/mklgpu/forward.cpp index b8c09dc77..45b527f52 100644 --- a/src/dft/backends/mklgpu/forward.cpp +++ b/src/dft/backends/mklgpu/forward.cpp @@ -26,7 +26,7 @@ #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" +#include "oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp" #include "oneapi/math/dft/detail/descriptor_impl.hpp" #include "mklgpu_helpers.hpp" diff --git a/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp b/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp index fc04491a0..bcadc5464 100644 --- a/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp +++ b/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/math/dft/detail/mklgpu/onemkl_dft_mklgpu.hpp" +#include "oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp" #include "dft/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/dft/backends/portfft/CMakeLists.txt b/src/dft/backends/portfft/CMakeLists.txt index 2491a6587..4aec1d11a 100644 --- a/src/dft/backends/portfft/CMakeLists.txt +++ b/src/dft/backends/portfft/CMakeLists.txt @@ -58,7 +58,7 @@ if (IS_DPCPP AND UNIX AND NOT FOUND_TARGETS) target_link_options(ONEMKL::SYCL::SYCL INTERFACE ${TARGETS_LINK_OPTIONS}) endif() -set(LIB_NAME onemkl_dft_portfft) +set(LIB_NAME onemath_dft_portfft) set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) @@ -67,7 +67,7 @@ add_library(${LIB_OBJ} OBJECT commit.cpp $<$: portfft_wrappers.cpp> ) -add_dependencies(onemkl_backend_libs_dft ${LIB_NAME}) +add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) find_package(portfft QUIET) if (NOT portfft_FOUND) @@ -88,7 +88,7 @@ else() target_link_libraries(${LIB_OBJ} PRIVATE portfft::portfft) endif() -target_link_libraries(${LIB_OBJ} PRIVATE onemkl_warnings) +target_link_libraries(${LIB_OBJ} PRIVATE onemath_warnings) target_include_directories(${LIB_OBJ} PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} diff --git a/src/dft/backends/portfft/commit.cpp b/src/dft/backends/portfft/commit.cpp index aa92256ab..4d20831c9 100644 --- a/src/dft/backends/portfft/commit.cpp +++ b/src/dft/backends/portfft/commit.cpp @@ -32,7 +32,7 @@ #include "oneapi/math/dft/detail/commit_impl.hpp" #include "oneapi/math/dft/detail/descriptor_impl.hpp" -#include "oneapi/math/dft/detail/portfft/onemkl_dft_portfft.hpp" +#include "oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp" #include "oneapi/math/dft/types.hpp" #include "../stride_helper.hpp" diff --git a/src/dft/backends/portfft/descriptor.cpp b/src/dft/backends/portfft/descriptor.cpp index 7bdcf21cc..21c59e8d3 100644 --- a/src/dft/backends/portfft/descriptor.cpp +++ b/src/dft/backends/portfft/descriptor.cpp @@ -20,7 +20,7 @@ #include "oneapi/math/dft/descriptor.hpp" #include "../../descriptor.cxx" -#include "oneapi/math/dft/detail/portfft/onemkl_dft_portfft.hpp" +#include "oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp" namespace oneapi::mkl::dft::detail { diff --git a/src/dft/backends/portfft/portfft_wrappers.cpp b/src/dft/backends/portfft/portfft_wrappers.cpp index 22abfcb81..c702d3696 100644 --- a/src/dft/backends/portfft/portfft_wrappers.cpp +++ b/src/dft/backends/portfft/portfft_wrappers.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/math/dft/detail/portfft/onemkl_dft_portfft.hpp" +#include "oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp" #include "dft/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/dft/backends/rocfft/CMakeLists.txt b/src/dft/backends/rocfft/CMakeLists.txt index cae182254..0c3f2f5af 100644 --- a/src/dft/backends/rocfft/CMakeLists.txt +++ b/src/dft/backends/rocfft/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_dft_rocfft) +set(LIB_NAME onemath_dft_rocfft) set(LIB_OBJ ${LIB_NAME}_obj) @@ -29,7 +29,7 @@ add_library(${LIB_OBJ} OBJECT backward.cpp $<$: rocfft_wrappers.cpp> ) -add_dependencies(onemkl_backend_libs_dft ${LIB_NAME}) +add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_OBJ} PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} diff --git a/src/dft/backends/rocfft/backward.cpp b/src/dft/backends/rocfft/backward.cpp index c835cdf34..938108e21 100644 --- a/src/dft/backends/rocfft/backward.cpp +++ b/src/dft/backends/rocfft/backward.cpp @@ -25,7 +25,7 @@ #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/dft/detail/rocfft/onemkl_dft_rocfft.hpp" +#include "oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp" #include "oneapi/math/dft/descriptor.hpp" #include "execute_helper.hpp" diff --git a/src/dft/backends/rocfft/commit.cpp b/src/dft/backends/rocfft/commit.cpp index 54b7b5e9c..2ce871447 100644 --- a/src/dft/backends/rocfft/commit.cpp +++ b/src/dft/backends/rocfft/commit.cpp @@ -31,7 +31,7 @@ #include "oneapi/math/dft/detail/commit_impl.hpp" #include "oneapi/math/dft/detail/descriptor_impl.hpp" -#include "oneapi/math/dft/detail/rocfft/onemkl_dft_rocfft.hpp" +#include "oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp" #include "oneapi/math/dft/types.hpp" #include "../stride_helper.hpp" diff --git a/src/dft/backends/rocfft/descriptor.cpp b/src/dft/backends/rocfft/descriptor.cpp index 9a235c46f..ba5fbb139 100644 --- a/src/dft/backends/rocfft/descriptor.cpp +++ b/src/dft/backends/rocfft/descriptor.cpp @@ -20,7 +20,7 @@ #include "oneapi/math/dft/descriptor.hpp" #include "../../descriptor.cxx" -#include "oneapi/math/dft/detail/rocfft/onemkl_dft_rocfft.hpp" +#include "oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp" namespace oneapi::mkl::dft::detail { diff --git a/src/dft/backends/rocfft/forward.cpp b/src/dft/backends/rocfft/forward.cpp index 5412f4678..ca9f8ad43 100644 --- a/src/dft/backends/rocfft/forward.cpp +++ b/src/dft/backends/rocfft/forward.cpp @@ -26,7 +26,7 @@ #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/dft/detail/rocfft/onemkl_dft_rocfft.hpp" +#include "oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp" #include "oneapi/math/dft/descriptor.hpp" #include "execute_helper.hpp" diff --git a/src/dft/backends/rocfft/rocfft_wrappers.cpp b/src/dft/backends/rocfft/rocfft_wrappers.cpp index c2afb0fb2..7554f0829 100644 --- a/src/dft/backends/rocfft/rocfft_wrappers.cpp +++ b/src/dft/backends/rocfft/rocfft_wrappers.cpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#include "oneapi/math/dft/detail/rocfft/onemkl_dft_rocfft.hpp" +#include "oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp" #include "dft/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/lapack/CMakeLists.txt b/src/lapack/CMakeLists.txt index 524edde03..cd5c28332 100644 --- a/src/lapack/CMakeLists.txt +++ b/src/lapack/CMakeLists.txt @@ -22,22 +22,22 @@ add_subdirectory(backends) # Recipe for LAPACK loader object if(BUILD_SHARED_LIBS) -add_library(onemkl_lapack OBJECT) -target_sources(onemkl_lapack PRIVATE lapack_loader.cpp) -target_include_directories(onemkl_lapack +add_library(onemath_lapack OBJECT) +target_sources(onemath_lapack PRIVATE lapack_loader.cpp) +target_include_directories(onemath_lapack PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMKL_GENERATED_INCLUDE_PATH} - $ + $ ) -target_compile_options(onemkl_lapack PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(onemath_lapack PRIVATE ${ONEMKL_BUILD_COPT}) -set_target_properties(onemkl_lapack PROPERTIES +set_target_properties(onemath_lapack PROPERTIES POSITION_INDEPENDENT_CODE ON ) -target_link_libraries(onemkl_lapack PUBLIC ONEMKL::SYCL::SYCL) +target_link_libraries(onemath_lapack PUBLIC ONEMKL::SYCL::SYCL) endif() diff --git a/src/lapack/backends/CMakeLists.txt b/src/lapack/backends/CMakeLists.txt index 636f6728f..44ac43b38 100644 --- a/src/lapack/backends/CMakeLists.txt +++ b/src/lapack/backends/CMakeLists.txt @@ -17,8 +17,8 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -add_custom_target(onemkl_backend_libs_lapack) -add_dependencies(onemkl_backend_libs onemkl_backend_libs_lapack) +add_custom_target(onemath_backend_libs_lapack) +add_dependencies(onemath_backend_libs onemath_backend_libs_lapack) if(ENABLE_MKLCPU_BACKEND) add_subdirectory(mklcpu) diff --git a/src/lapack/backends/cusolver/CMakeLists.txt b/src/lapack/backends/cusolver/CMakeLists.txt index dfd1267d7..bfcdc3721 100644 --- a/src/lapack/backends/cusolver/CMakeLists.txt +++ b/src/lapack/backends/cusolver/CMakeLists.txt @@ -17,7 +17,7 @@ # #========================================================================= -set(LIB_NAME onemkl_lapack_cusolver) +set(LIB_NAME onemath_lapack_cusolver) set(LIB_OBJ ${LIB_NAME}_obj) find_package(cuSOLVER REQUIRED) find_package(cuBLAS REQUIRED) @@ -27,7 +27,7 @@ set(SOURCES cusolver_lapack.cpp $<$: cusolver_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) -add_dependencies(onemkl_backend_libs_lapack ${LIB_NAME}) +add_dependencies(onemath_backend_libs_lapack ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/lapack/backends/cusolver/cusolver_batch.cpp b/src/lapack/backends/cusolver/cusolver_batch.cpp index da4ec8907..b193113e6 100644 --- a/src/lapack/backends/cusolver/cusolver_batch.cpp +++ b/src/lapack/backends/cusolver/cusolver_batch.cpp @@ -21,7 +21,7 @@ #include "cusolver_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp" +#include "oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp" namespace oneapi { namespace mkl { @@ -44,7 +44,7 @@ inline void geqrf_batch(const char *func_name, Func func, sycl::queue &queue, st auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -110,7 +110,7 @@ inline void getri_batch(const char *func_name, Func func, sycl::queue &queue, st sycl::accessor ipiv32_acc{ ipiv32, cgh }; sycl::accessor devInfo_acc{ devInfo, cgh, sycl::write_only }; - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { cublasStatus_t err; CUresult cuda_result; cublasHandle_t cublas_handle; @@ -214,7 +214,7 @@ inline void getrs_batch(const char *func_name, Func func, sycl::queue &queue, auto ipiv_acc = ipiv32.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto ipiv_ = sc.get_mem(ipiv_acc); @@ -275,7 +275,7 @@ inline void getrf_batch(const char *func_name, Func func, sycl::queue &queue, st auto ipiv32_acc = ipiv32.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto ipiv_ = sc.get_mem(ipiv32_acc); @@ -333,7 +333,7 @@ inline void orgqr_batch(const char *func_name, Func func, sycl::queue &queue, st auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -375,7 +375,7 @@ inline void potrf_batch(const char *func_name, Func func, sycl::queue &queue, queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); CUdeviceptr a_dev; CUresult cuda_result; @@ -434,7 +434,7 @@ inline void potrs_batch(const char *func_name, Func func, sycl::queue &queue, queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); CUdeviceptr a_dev, b_dev; cusolverStatus_t err; @@ -499,7 +499,7 @@ inline void ungqr_batch(const char *func_name, Func func, sycl::queue &queue, st auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -544,7 +544,7 @@ inline sycl::event geqrf_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -595,7 +595,7 @@ inline sycl::event geqrf_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -653,7 +653,7 @@ inline sycl::event getrf_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto devInfo_ = reinterpret_cast(devInfo); @@ -735,7 +735,7 @@ inline sycl::event getrf_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto scratch_ = reinterpret_cast(scratchpad); @@ -835,7 +835,7 @@ sycl::event getri_batch(const char *func_name, Func func, sycl::queue &queue, st cgh.depends_on(done_casting); cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { cublasStatus_t err; CUresult cuda_result; cublasHandle_t cublas_handle; @@ -961,7 +961,7 @@ inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); cgh.depends_on(done_casting); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto ipiv_ = reinterpret_cast(ipiv32); @@ -1049,7 +1049,7 @@ inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &qu cgh.depends_on(dependencies); cgh.depends_on(casting_dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -1109,7 +1109,7 @@ inline sycl::event orgqr_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1159,7 +1159,7 @@ inline sycl::event orgqr_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1210,7 +1210,7 @@ inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); CUdeviceptr a_dev; cusolverStatus_t err; @@ -1268,7 +1268,7 @@ inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; CUdeviceptr a_dev; @@ -1331,7 +1331,7 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); CUresult cuda_result; CUdeviceptr a_dev, b_dev; @@ -1414,7 +1414,7 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu cgh.depends_on(dependencies); cgh.depends_on(done_cpy_a); cgh.depends_on(done_cpy_b); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; cusolverStatus_t err; @@ -1468,7 +1468,7 @@ inline sycl::event ungqr_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1518,7 +1518,7 @@ inline sycl::event ungqr_batch(const char *func_name, Func func, sycl::queue &qu auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1565,7 +1565,7 @@ inline void getrf_batch_scratchpad_size(const char *func_name, Func func, sycl:: std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size, int *scratch_size) { auto e = queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; @@ -1633,7 +1633,7 @@ inline void geqrf_batch_scratchpad_size(const char *func_name, Func func, sycl:: std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size, int *scratch_size) { auto e = queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; @@ -1701,7 +1701,7 @@ inline void orgqr_batch_scratchpad_size(const char *func_name, Func func, sycl:: std::int64_t stride_tau, std::int64_t batch_size, int *scratch_size) { auto e = queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; @@ -1735,7 +1735,7 @@ inline void ungqr_batch_scratchpad_size(const char *func_name, Func func, sycl:: std::int64_t stride_tau, std::int64_t batch_size, int *scratch_size) { auto e = queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; @@ -1768,7 +1768,7 @@ inline void getrf_batch_scratchpad_size(const char *func_name, Func func, sycl:: std::int64_t group_count, std::int64_t *group_sizes, int *scratch_size) { auto e = queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int group_scratch_size = 0; *scratch_size = 0; @@ -1847,7 +1847,7 @@ inline void geqrf_batch_scratchpad_size(const char *func_name, Func func, sycl:: std::int64_t group_count, std::int64_t *group_sizes, int *scratch_size) { auto e = queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int group_scratch_size = 0; *scratch_size = 0; @@ -1889,7 +1889,7 @@ inline void orgqr_batch_scratchpad_size(const char *func_name, Func func, sycl:: std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, int *scratch_size) { auto e = queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int group_scratch_size = 0; *scratch_size = 0; @@ -1963,7 +1963,7 @@ inline void ungqr_batch_scratchpad_size(const char *func_name, Func func, sycl:: std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, int *scratch_size) { auto e = queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int group_scratch_size = 0; *scratch_size = 0; diff --git a/src/lapack/backends/cusolver/cusolver_lapack.cpp b/src/lapack/backends/cusolver/cusolver_lapack.cpp index 9c57fdeb6..d9624cdaf 100644 --- a/src/lapack/backends/cusolver/cusolver_lapack.cpp +++ b/src/lapack/backends/cusolver/cusolver_lapack.cpp @@ -20,7 +20,7 @@ #include "cusolver_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp" +#include "oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp" namespace oneapi { namespace mkl { @@ -48,7 +48,7 @@ inline void gebrd(const char *func_name, Func func, sycl::queue &queue, std::int auto tauq_acc = tauq.template get_access(cgh); auto taup_acc = taup.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto d_ = sc.get_mem(d_acc); @@ -111,7 +111,7 @@ inline void geqrf(const char *func_name, Func func, sycl::queue &queue, std::int auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -157,7 +157,7 @@ void getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, auto ipiv32_acc = ipiv32.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto ipiv32_ = sc.get_mem(ipiv32_acc); @@ -237,7 +237,7 @@ inline void getrs(const char *func_name, Func func, sycl::queue &queue, auto a_acc = a.template get_access(cgh); auto ipiv_acc = ipiv32.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto ipiv_ = sc.get_mem(ipiv_acc); @@ -282,7 +282,7 @@ inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto vt_acc = vt.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto s_ = sc.get_mem(s_acc); @@ -331,7 +331,7 @@ inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto w_ = sc.get_mem(w_acc); @@ -375,7 +375,7 @@ inline void hegvd(const char *func_name, Func func, sycl::queue &queue, std::int auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -421,7 +421,7 @@ inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto tau_acc = tau.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto d_ = sc.get_mem(d_acc); @@ -474,7 +474,7 @@ inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -509,7 +509,7 @@ inline void orgqr(const char *func_name, Func func, sycl::queue &queue, std::int auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -544,7 +544,7 @@ inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -582,7 +582,7 @@ inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -637,7 +637,7 @@ inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -676,7 +676,7 @@ inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto a_acc = a.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto devInfo_ = sc.get_mem(devInfo_acc); @@ -714,7 +714,7 @@ inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto a_acc = a.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto devInfo_ = sc.get_mem(devInfo_acc); @@ -752,7 +752,7 @@ inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi:: queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -790,7 +790,7 @@ inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto w_ = sc.get_mem(w_acc); @@ -832,7 +832,7 @@ inline void sygvd(const char *func_name, Func func, sycl::queue &queue, std::int auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -877,7 +877,7 @@ inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto tau_acc = tau.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto d_ = sc.get_mem(d_acc); @@ -927,7 +927,7 @@ inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto ipiv32_acc = ipiv32.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto ipiv32_ = sc.get_mem(ipiv32_acc); @@ -1003,7 +1003,7 @@ inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -1038,7 +1038,7 @@ inline void ungqr(const char *func_name, Func func, sycl::queue &queue, std::int auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -1073,7 +1073,7 @@ inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -1125,7 +1125,7 @@ inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -1166,7 +1166,7 @@ inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -1215,7 +1215,7 @@ inline sycl::event gebrd(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto d_ = reinterpret_cast(d); @@ -1280,7 +1280,7 @@ inline sycl::event geqrf(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1328,7 +1328,7 @@ inline sycl::event getrf(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto devInfo_ = reinterpret_cast(devInfo); @@ -1416,7 +1416,7 @@ inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, cgh.depends_on(dependencies[i]); } cgh.depends_on(done_casting); - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto ipiv_ = reinterpret_cast(ipiv32); @@ -1465,7 +1465,7 @@ inline sycl::event gesvd(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto s_ = reinterpret_cast(s); @@ -1516,7 +1516,7 @@ inline sycl::event heevd(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto w_ = reinterpret_cast(w); @@ -1562,7 +1562,7 @@ inline sycl::event hegvd(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -1609,7 +1609,7 @@ inline sycl::event hetrd(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto d_ = reinterpret_cast(d); @@ -1667,7 +1667,7 @@ inline sycl::event orgbr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1706,7 +1706,7 @@ inline sycl::event orgqr(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1744,7 +1744,7 @@ inline sycl::event orgtr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1784,7 +1784,7 @@ inline sycl::event ormtr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1841,7 +1841,7 @@ inline sycl::event ormqr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1884,7 +1884,7 @@ inline sycl::event potrf(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto devInfo_ = reinterpret_cast(devInfo); @@ -1927,7 +1927,7 @@ inline sycl::event potri(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto scratch_ = reinterpret_cast(scratchpad); @@ -1971,7 +1971,7 @@ inline sycl::event potrs(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -2012,7 +2012,7 @@ inline sycl::event syevd(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto w_ = reinterpret_cast(w); @@ -2057,7 +2057,7 @@ inline sycl::event sygvd(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -2102,7 +2102,7 @@ inline sycl::event sytrd(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto d_ = reinterpret_cast(d); @@ -2154,7 +2154,7 @@ inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto scratch_ = reinterpret_cast(scratchpad); @@ -2239,7 +2239,7 @@ inline sycl::event ungbr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -2278,7 +2278,7 @@ inline sycl::event ungqr(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -2316,7 +2316,7 @@ inline sycl::event ungtr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -2370,7 +2370,7 @@ inline sycl::event unmqr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -2414,7 +2414,7 @@ inline sycl::event unmtr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -2452,7 +2452,7 @@ inline void gebrd_scratchpad_size(const char *func_name, Func func, sycl::queue std::int64_t m, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, scratch_size); @@ -2503,7 +2503,7 @@ inline void geqrf_scratchpad_size(const char *func_name, Func func, sycl::queue std::int64_t m, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, nullptr, lda, scratch_size); @@ -2534,7 +2534,7 @@ inline void gesvd_scratchpad_size(const char *func_name, Func func, sycl::queue std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, scratch_size); @@ -2565,7 +2565,7 @@ inline void getrf_scratchpad_size(const char *func_name, Func func, sycl::queue std::int64_t m, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, nullptr, lda, scratch_size); @@ -2625,7 +2625,7 @@ inline void heevd_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_job(jobz), @@ -2657,7 +2657,7 @@ inline void hegvd_scratchpad_size(const char *func_name, Func func, sycl::queue std::int64_t n, std::int64_t lda, std::int64_t ldb, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_itype(itype), @@ -2688,7 +2688,7 @@ inline void hetrd_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, @@ -2728,7 +2728,7 @@ inline void orgbr_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_generate(vec), m, n, k, @@ -2758,7 +2758,7 @@ inline void orgtr_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, @@ -2787,7 +2787,7 @@ inline void orgqr_scratchpad_size(const char *func_name, Func func, sycl::queue std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, k, nullptr, lda, nullptr, @@ -2832,7 +2832,7 @@ inline void ormqr_scratchpad_size(const char *func_name, Func func, sycl::queue std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), @@ -2864,7 +2864,7 @@ inline void ormtr_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), @@ -2896,7 +2896,7 @@ inline void potrf_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, @@ -2943,7 +2943,7 @@ inline void potri_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, @@ -2974,7 +2974,7 @@ inline void sytrf_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, n, nullptr, lda, scratch_size); @@ -3004,7 +3004,7 @@ inline void syevd_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_job(jobz), @@ -3036,7 +3036,7 @@ inline void sygvd_scratchpad_size(const char *func_name, Func func, sycl::queue std::int64_t n, std::int64_t lda, std::int64_t ldb, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_itype(itype), @@ -3067,7 +3067,7 @@ inline void sytrd_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, @@ -3127,7 +3127,7 @@ inline void ungbr_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_generate(vec), m, n, k, @@ -3157,7 +3157,7 @@ inline void ungqr_scratchpad_size(const char *func_name, Func func, sycl::queue std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, k, nullptr, lda, nullptr, @@ -3186,7 +3186,7 @@ inline void ungtr_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, @@ -3233,7 +3233,7 @@ inline void unmqr_scratchpad_size(const char *func_name, Func func, sycl::queue std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), @@ -3265,7 +3265,7 @@ inline void unmtr_scratchpad_size(const char *func_name, Func func, sycl::queue oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { - onemkl_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), diff --git a/src/lapack/backends/cusolver/cusolver_task.hpp b/src/lapack/backends/cusolver/cusolver_task.hpp index 2985e935e..972b41ae9 100644 --- a/src/lapack/backends/cusolver/cusolver_task.hpp +++ b/src/lapack/backends/cusolver/cusolver_task.hpp @@ -61,7 +61,7 @@ static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { } template -static inline void onemkl_cusolver_host_task(H &cgh, sycl::queue queue, F f) { +static inline void onemath_cusolver_host_task(H &cgh, sycl::queue queue, F f) { (void)host_task_internal(cgh, queue, f); } diff --git a/src/lapack/backends/cusolver/cusolver_wrappers.cpp b/src/lapack/backends/cusolver/cusolver_wrappers.cpp index 60e4a57a9..acfff6c65 100644 --- a/src/lapack/backends/cusolver/cusolver_wrappers.cpp +++ b/src/lapack/backends/cusolver/cusolver_wrappers.cpp @@ -17,7 +17,7 @@ * **************************************************************************/ #include "lapack/function_table.hpp" -#include "oneapi/math/lapack/detail/cusolver/onemkl_lapack_cusolver.hpp" +#include "oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp" #define WRAPPER_VERSION 1 diff --git a/src/lapack/backends/mklcpu/CMakeLists.txt b/src/lapack/backends/mklcpu/CMakeLists.txt index fcc60a8e7..d53c0711f 100644 --- a/src/lapack/backends/mklcpu/CMakeLists.txt +++ b/src/lapack/backends/mklcpu/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_lapack_mklcpu) +set(LIB_NAME onemath_lapack_mklcpu) set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) @@ -25,7 +25,7 @@ add_library(${LIB_OBJ} OBJECT mkl_lapack.cpp $<$: lapack_cpu_wrappers.cpp> ) -add_dependencies(onemkl_backend_libs_lapack ${LIB_NAME}) +add_dependencies(onemath_backend_libs_lapack ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp b/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp index c4ee847f1..a3c1cf6fb 100644 --- a/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp +++ b/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "lapack/function_table.hpp" -#include "oneapi/math/lapack/detail/mklcpu/onemkl_lapack_mklcpu.hpp" +#include "oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/lapack/backends/mklcpu/mkl_lapack.cpp b/src/lapack/backends/mklcpu/mkl_lapack.cpp index 1f94a05e3..f64b538a8 100644 --- a/src/lapack/backends/mklcpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklcpu/mkl_lapack.cpp @@ -25,7 +25,7 @@ #include "oneapi/math/types.hpp" #include "oneapi/math/lapack/types.hpp" -#include "oneapi/math/lapack/detail/mklcpu/onemkl_lapack_mklcpu.hpp" +#include "oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp" #include "../mkl_common/mkl_lapack_backend.hpp" namespace oneapi { diff --git a/src/lapack/backends/mklgpu/CMakeLists.txt b/src/lapack/backends/mklgpu/CMakeLists.txt index e11592f82..213cbb73d 100644 --- a/src/lapack/backends/mklgpu/CMakeLists.txt +++ b/src/lapack/backends/mklgpu/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_lapack_mklgpu) +set(LIB_NAME onemath_lapack_mklgpu) set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) @@ -25,7 +25,7 @@ add_library(${LIB_OBJ} OBJECT mkl_lapack.cpp $<$: lapack_gpu_wrappers.cpp> ) -add_dependencies(onemkl_backend_libs_lapack ${LIB_NAME}) +add_dependencies(onemath_backend_libs_lapack ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp b/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp index 4a75f1c66..675c5653c 100644 --- a/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp +++ b/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "lapack/function_table.hpp" -#include "oneapi/math/lapack/detail/mklgpu/onemkl_lapack_mklgpu.hpp" +#include "oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/lapack/backends/mklgpu/mkl_lapack.cpp b/src/lapack/backends/mklgpu/mkl_lapack.cpp index ea5b5749d..8d24c0eb9 100644 --- a/src/lapack/backends/mklgpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklgpu/mkl_lapack.cpp @@ -25,7 +25,7 @@ #include "oneapi/math/types.hpp" #include "oneapi/math/lapack/types.hpp" -#include "oneapi/math/lapack/detail/mklgpu/onemkl_lapack_mklgpu.hpp" +#include "oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp" #include "../mkl_common/mkl_lapack_backend.hpp" namespace oneapi { diff --git a/src/lapack/backends/rocsolver/CMakeLists.txt b/src/lapack/backends/rocsolver/CMakeLists.txt index c91089118..f62f56ec6 100644 --- a/src/lapack/backends/rocsolver/CMakeLists.txt +++ b/src/lapack/backends/rocsolver/CMakeLists.txt @@ -19,7 +19,7 @@ # #========================================================================= -set(LIB_NAME onemkl_lapack_rocsolver) +set(LIB_NAME onemath_lapack_rocsolver) set(LIB_OBJ ${LIB_NAME}_obj) find_package(hip REQUIRED) find_package(rocsolver REQUIRED) @@ -31,7 +31,7 @@ set(SOURCES rocsolver_lapack.cpp $<$: rocsolver_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) -add_dependencies(onemkl_backend_libs_lapack ${LIB_NAME}) +add_dependencies(onemath_backend_libs_lapack ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/lapack/backends/rocsolver/rocsolver_batch.cpp b/src/lapack/backends/rocsolver/rocsolver_batch.cpp index a3d2381da..03217b85b 100644 --- a/src/lapack/backends/rocsolver/rocsolver_batch.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_batch.cpp @@ -22,7 +22,7 @@ #include "rocsolver_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp" +#include "oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp" namespace oneapi { namespace mkl { @@ -520,7 +520,7 @@ inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &qu cgh.depends_on(dependencies[i]); } cgh.depends_on(done_cpy); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; @@ -620,7 +620,7 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu } cgh.depends_on(done_cpy_a); cgh.depends_on(done_cpy_b); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; diff --git a/src/lapack/backends/rocsolver/rocsolver_lapack.cpp b/src/lapack/backends/rocsolver/rocsolver_lapack.cpp index d3cf85d03..311c9985a 100644 --- a/src/lapack/backends/rocsolver/rocsolver_lapack.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_lapack.cpp @@ -22,7 +22,7 @@ #include "rocsolver_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp" +#include "oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp" namespace oneapi { namespace mkl { @@ -46,7 +46,7 @@ inline void gebrd(const char *func_name, Func func, sycl::queue &queue, std::int auto e_acc = e.template get_access(cgh); auto tauq_acc = tauq.template get_access(cgh); auto taup_acc = taup.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto d_ = sc.get_mem(d_acc); @@ -107,7 +107,7 @@ inline void geqrf(const char *func_name, Func func, sycl::queue &queue, std::int queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -150,7 +150,7 @@ void getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, auto a_acc = a.template get_access(cgh); auto ipiv32_acc = ipiv32.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto ipiv32_ = sc.get_mem(ipiv32_acc); @@ -236,7 +236,7 @@ inline void getrs(const char *func_name, Func func, sycl::queue &queue, auto a_acc = a.template get_access(cgh); auto ipiv_acc = ipiv32.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto ipiv_ = sc.get_mem(ipiv_acc); @@ -281,7 +281,7 @@ inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto vt_acc = vt.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto s_ = sc.get_mem(s_acc); @@ -330,7 +330,7 @@ inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto w_ = sc.get_mem(w_acc); @@ -374,7 +374,7 @@ inline void hegvd(const char *func_name, Func func, sycl::queue &queue, std::int auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -417,7 +417,7 @@ inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto d_acc = d.template get_access(cgh); auto e_acc = e.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto d_ = sc.get_mem(d_acc); @@ -466,7 +466,7 @@ inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -499,7 +499,7 @@ inline void orgqr(const char *func_name, Func func, sycl::queue &queue, std::int queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -531,7 +531,7 @@ inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -567,7 +567,7 @@ inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -619,7 +619,7 @@ inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -656,7 +656,7 @@ inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto devInfo_ = sc.get_mem(devInfo_acc); @@ -692,7 +692,7 @@ inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi:: queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto devInfo_ = sc.get_mem(devInfo_acc); @@ -728,7 +728,7 @@ inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi:: queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -766,7 +766,7 @@ inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto w_ = sc.get_mem(w_acc); @@ -808,7 +808,7 @@ inline void sygvd(const char *func_name, Func func, sycl::queue &queue, std::int auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto b_ = sc.get_mem(b_acc); @@ -850,7 +850,7 @@ inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto d_acc = d.template get_access(cgh); auto e_acc = e.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto d_ = sc.get_mem(d_acc); @@ -896,7 +896,7 @@ inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto a_acc = a.template get_access(cgh); auto ipiv32_acc = ipiv32.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto ipiv32_ = sc.get_mem(ipiv32_acc); @@ -971,7 +971,7 @@ inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -1004,7 +1004,7 @@ inline void ungqr(const char *func_name, Func func, sycl::queue &queue, std::int queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -1036,7 +1036,7 @@ inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -1086,7 +1086,7 @@ inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -1125,7 +1125,7 @@ inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); auto tau_ = sc.get_mem(tau_acc); @@ -1169,7 +1169,7 @@ inline sycl::event gebrd(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto d_ = reinterpret_cast(d); @@ -1233,7 +1233,7 @@ inline sycl::event geqrf(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1279,7 +1279,7 @@ inline sycl::event getrf(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto devInfo_ = reinterpret_cast(devInfo); @@ -1368,7 +1368,7 @@ inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, cgh.depends_on(dependencies[i]); } cgh.depends_on(done_casting); - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto ipiv_ = reinterpret_cast(ipiv32); @@ -1417,7 +1417,7 @@ inline sycl::event gesvd(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto s_ = reinterpret_cast(s); @@ -1468,7 +1468,7 @@ inline sycl::event heevd(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto w_ = reinterpret_cast(w); @@ -1514,7 +1514,7 @@ inline sycl::event hegvd(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -1560,7 +1560,7 @@ inline sycl::event hetrd(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto d_ = reinterpret_cast(d); @@ -1614,7 +1614,7 @@ inline sycl::event orgbr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1652,7 +1652,7 @@ inline sycl::event orgqr(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1688,7 +1688,7 @@ inline sycl::event orgtr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1727,7 +1727,7 @@ inline sycl::event ormtr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1782,7 +1782,7 @@ inline sycl::event ormqr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -1824,7 +1824,7 @@ inline sycl::event potrf(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto devInfo_ = reinterpret_cast(devInfo); @@ -1866,7 +1866,7 @@ inline sycl::event potri(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto scratch_ = reinterpret_cast(scratchpad); @@ -1909,7 +1909,7 @@ inline sycl::event potrs(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -1950,7 +1950,7 @@ inline sycl::event syevd(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto w_ = reinterpret_cast(w); @@ -1995,7 +1995,7 @@ inline sycl::event sygvd(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto b_ = reinterpret_cast(b); @@ -2039,7 +2039,7 @@ inline sycl::event sytrd(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto d_ = reinterpret_cast(d); @@ -2087,7 +2087,7 @@ inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto ipiv_ = reinterpret_cast(ipiv32); @@ -2168,7 +2168,7 @@ inline sycl::event ungbr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -2206,7 +2206,7 @@ inline sycl::event ungqr(const char *func_name, Func func, sycl::queue &queue, s for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -2242,7 +2242,7 @@ inline sycl::event ungtr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -2295,7 +2295,7 @@ inline sycl::event unmqr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); @@ -2338,7 +2338,7 @@ inline sycl::event unmtr(const char *func_name, Func func, sycl::queue &queue, for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemkl_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); auto tau_ = reinterpret_cast(tau); diff --git a/src/lapack/backends/rocsolver/rocsolver_task.hpp b/src/lapack/backends/rocsolver/rocsolver_task.hpp index a5a30b3d1..f6b110259 100644 --- a/src/lapack/backends/rocsolver/rocsolver_task.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_task.hpp @@ -63,7 +63,7 @@ static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { } template -static inline void onemkl_rocsolver_host_task(H &cgh, sycl::queue queue, F f) { +static inline void onemath_rocsolver_host_task(H &cgh, sycl::queue queue, F f) { (void)host_task_internal(cgh, queue, f); } diff --git a/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp b/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp index b0c2fa489..f361687fe 100644 --- a/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp @@ -19,7 +19,7 @@ * **************************************************************************/ #include "lapack/function_table.hpp" -#include "oneapi/math/lapack/detail/rocsolver/onemkl_lapack_rocsolver.hpp" +#include "oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp" #define WRAPPER_VERSION 1 diff --git a/src/rng/CMakeLists.txt b/src/rng/CMakeLists.txt index 30df39403..9f6775e72 100644 --- a/src/rng/CMakeLists.txt +++ b/src/rng/CMakeLists.txt @@ -22,26 +22,26 @@ add_subdirectory(backends) # Recipe for RNG loader object if(BUILD_SHARED_LIBS) -add_library(onemkl_rng OBJECT) -target_sources(onemkl_rng PRIVATE rng_loader.cpp) -target_include_directories(onemkl_rng +add_library(onemath_rng OBJECT) +target_sources(onemath_rng PRIVATE rng_loader.cpp) +target_include_directories(onemath_rng PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMKL_GENERATED_INCLUDE_PATH} - $ + $ ) -target_compile_options(onemkl_rng PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(onemath_rng PRIVATE ${ONEMKL_BUILD_COPT}) -set_target_properties(onemkl_rng PROPERTIES +set_target_properties(onemath_rng PROPERTIES POSITION_INDEPENDENT_CODE ON ) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) - add_sycl_to_target(TARGET onemkl_rng SOURCES rng_loader.cpp) + add_sycl_to_target(TARGET onemath_rng SOURCES rng_loader.cpp) else() - target_link_libraries(onemkl_rng PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(onemath_rng PUBLIC ONEMKL::SYCL::SYCL) endif() endif() diff --git a/src/rng/backends/CMakeLists.txt b/src/rng/backends/CMakeLists.txt index 9045f7e75..52ddcdd3c 100644 --- a/src/rng/backends/CMakeLists.txt +++ b/src/rng/backends/CMakeLists.txt @@ -17,8 +17,8 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -add_custom_target(onemkl_backend_libs_rng) -add_dependencies(onemkl_backend_libs onemkl_backend_libs_rng) +add_custom_target(onemath_backend_libs_rng) +add_dependencies(onemath_backend_libs onemath_backend_libs_rng) if(ENABLE_MKLCPU_BACKEND) add_subdirectory(mklcpu) diff --git a/src/rng/backends/curand/CMakeLists.txt b/src/rng/backends/curand/CMakeLists.txt index f9fbf8168..3990f8086 100644 --- a/src/rng/backends/curand/CMakeLists.txt +++ b/src/rng/backends/curand/CMakeLists.txt @@ -56,7 +56,7 @@ # so. #================================================================================= -set(LIB_NAME onemkl_rng_curand) +set(LIB_NAME onemath_rng_curand) set(LIB_OBJ ${LIB_NAME}_obj) find_package(cuRAND REQUIRED) @@ -66,7 +66,7 @@ set(SOURCES philox4x32x10.cpp add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) -add_dependencies(onemkl_backend_libs_rng ${LIB_NAME}) +add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/rng/backends/curand/curand_task.hpp b/src/rng/backends/curand/curand_task.hpp index 240ced805..49455cfbc 100644 --- a/src/rng/backends/curand/curand_task.hpp +++ b/src/rng/backends/curand/curand_task.hpp @@ -73,12 +73,12 @@ static inline void host_task_internal(H &cgh, E e, F f) { } #endif template -static inline void onemkl_curand_host_task(H &cgh, A acc, E e, F f) { +static inline void onemath_curand_host_task(H &cgh, A acc, E e, F f) { host_task_internal(cgh, acc, e, f); } template -static inline void onemkl_curand_host_task(H &cgh, Engine e, F f) { +static inline void onemath_curand_host_task(H &cgh, Engine e, F f) { host_task_internal(cgh, e, f); } diff --git a/src/rng/backends/curand/curand_wrappers.cpp b/src/rng/backends/curand/curand_wrappers.cpp index f46227344..92ef359eb 100644 --- a/src/rng/backends/curand/curand_wrappers.cpp +++ b/src/rng/backends/curand/curand_wrappers.cpp @@ -56,7 +56,7 @@ * so. ******************************************************************************/ -#include "oneapi/math/rng/detail/curand/onemkl_rng_curand.hpp" +#include "oneapi/math/rng/detail/curand/onemath_rng_curand.hpp" #include "rng/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/rng/backends/curand/mrg32k3a.cpp b/src/rng/backends/curand/mrg32k3a.cpp index 2f761f95e..d3e77a810 100644 --- a/src/rng/backends/curand/mrg32k3a.cpp +++ b/src/rng/backends/curand/mrg32k3a.cpp @@ -75,7 +75,7 @@ #include "curand_helper.hpp" #include "curand_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/rng/detail/curand/onemkl_rng_curand.hpp" +#include "oneapi/math/rng/detail/curand/onemath_rng_curand.hpp" #include "oneapi/math/rng/detail/engine_impl.hpp" #include "oneapi/math/rng/engines.hpp" @@ -112,7 +112,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateUniform, status, engine_, r_ptr, n); }); @@ -125,7 +125,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateUniformDouble, status, engine_, r_ptr, n); }); @@ -139,7 +139,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::buffer ib(n); queue_.submit([&](sycl::handler& cgh) { auto acc = ib.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerate, status, engine_, r_ptr, n); }); @@ -152,7 +152,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateUniform, status, engine_, r_ptr, n); }); @@ -165,7 +165,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateUniformDouble, status, engine_, r_ptr, n); }); @@ -178,7 +178,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateNormal, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -191,7 +191,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateNormalDouble, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -220,7 +220,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateLogNormal, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -233,7 +233,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateLogNormalDouble, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -289,7 +289,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.template get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerate, status, engine_, r_ptr, n); }); @@ -303,7 +303,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateUniform, status, engine_, r, n); }); @@ -316,7 +316,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateUniformDouble, status, engine_, r, n); }); @@ -337,7 +337,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerate, status, engine_, ib, n); }); @@ -352,7 +352,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateUniform, status, engine_, r, n); }); @@ -366,7 +366,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateUniformDouble, status, engine_, r, n); }); @@ -380,7 +380,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateNormal, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -394,7 +394,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateNormalDouble, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -426,7 +426,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateLogNormal, status, engine_, r, n, distr.m(), distr.s()); }); @@ -439,7 +439,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateLogNormalDouble, status, engine_, r, n, distr.m(), distr.s()); @@ -505,7 +505,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerate, status, engine_, r, n); }); diff --git a/src/rng/backends/curand/philox4x32x10.cpp b/src/rng/backends/curand/philox4x32x10.cpp index 4a0fcd476..53fef8f60 100644 --- a/src/rng/backends/curand/philox4x32x10.cpp +++ b/src/rng/backends/curand/philox4x32x10.cpp @@ -77,7 +77,7 @@ #include "curand_helper.hpp" #include "curand_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/rng/detail/curand/onemkl_rng_curand.hpp" +#include "oneapi/math/rng/detail/curand/onemath_rng_curand.hpp" namespace oneapi { namespace mkl { @@ -135,7 +135,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateUniform, status, engine_, r_ptr, n); }); @@ -148,7 +148,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateUniformDouble, status, engine_, r_ptr, n); }); @@ -162,7 +162,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::buffer ib(n); queue_.submit([&](sycl::handler& cgh) { auto acc = ib.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerate, status, engine_, r_ptr, n); }); @@ -175,7 +175,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateUniform, status, engine_, r_ptr, n); }); @@ -188,7 +188,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateUniformDouble, status, engine_, r_ptr, n); }); @@ -201,7 +201,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateNormal, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -214,7 +214,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateNormalDouble, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -243,7 +243,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](float* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateLogNormal, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -256,7 +256,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerateLogNormalDouble, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -312,7 +312,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.template get_access(cgh); - onemkl_curand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { + onemath_curand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { curandStatus_t status; CURAND_CALL(curandGenerate, status, engine_, r_ptr, n); }); @@ -326,7 +326,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateUniform, status, engine_, r, n); }); @@ -339,7 +339,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateUniformDouble, status, engine_, r, n); }); @@ -360,7 +360,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerate, status, engine_, ib, n); }); @@ -375,7 +375,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateUniform, status, engine_, r, n); }); @@ -388,7 +388,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateUniformDouble, status, engine_, r, n); }); @@ -402,7 +402,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateNormal, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -416,7 +416,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateNormalDouble, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -448,7 +448,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateLogNormal, status, engine_, r, n, distr.m(), distr.s()); }); @@ -461,7 +461,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerateLogNormalDouble, status, engine_, r, n, distr.m(), distr.s()); @@ -527,7 +527,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { - onemkl_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_curand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandGenerate, status, engine_, r, n); }); diff --git a/src/rng/backends/mklcpu/CMakeLists.txt b/src/rng/backends/mklcpu/CMakeLists.txt index e72ce048f..b977dc28d 100644 --- a/src/rng/backends/mklcpu/CMakeLists.txt +++ b/src/rng/backends/mklcpu/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_rng_mklcpu) +set(LIB_NAME onemath_rng_mklcpu) set(LIB_OBJ ${LIB_NAME}_obj) set(SOURCES cpu_common.hpp @@ -28,7 +28,7 @@ set(SOURCES cpu_common.hpp add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) -add_dependencies(onemkl_backend_libs_rng ${LIB_NAME}) +add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src diff --git a/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp b/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp index bfc701780..cf6d56cc3 100644 --- a/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp +++ b/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "rng/function_table.hpp" -#include "oneapi/math/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp" +#include "oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/rng/backends/mklcpu/mrg32k3a.cpp b/src/rng/backends/mklcpu/mrg32k3a.cpp index ca31f7409..27bb86203 100644 --- a/src/rng/backends/mklcpu/mrg32k3a.cpp +++ b/src/rng/backends/mklcpu/mrg32k3a.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/rng/detail/engine_impl.hpp" -#include "oneapi/math/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp" +#include "oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp" #include "cpu_common.hpp" diff --git a/src/rng/backends/mklcpu/philox4x32x10.cpp b/src/rng/backends/mklcpu/philox4x32x10.cpp index f9af3597a..2d3e660fd 100644 --- a/src/rng/backends/mklcpu/philox4x32x10.cpp +++ b/src/rng/backends/mklcpu/philox4x32x10.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/rng/detail/engine_impl.hpp" -#include "oneapi/math/rng/detail/mklcpu/onemkl_rng_mklcpu.hpp" +#include "oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp" #include "cpu_common.hpp" diff --git a/src/rng/backends/mklgpu/CMakeLists.txt b/src/rng/backends/mklgpu/CMakeLists.txt index 150f90136..01849fbb1 100644 --- a/src/rng/backends/mklgpu/CMakeLists.txt +++ b/src/rng/backends/mklgpu/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_rng_mklgpu) +set(LIB_NAME onemath_rng_mklgpu) set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) @@ -27,7 +27,7 @@ add_library(${LIB_OBJ} OBJECT mrg32k3a.cpp $<$: mkl_rng_gpu_wrappers.cpp> ) -add_dependencies(onemkl_backend_libs_rng ${LIB_NAME}) +add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp b/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp index 0bd5dcf53..2d0c76cac 100644 --- a/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp +++ b/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp @@ -18,7 +18,7 @@ *******************************************************************************/ #include "rng/function_table.hpp" -#include "oneapi/math/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp" +#include "oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp" #define WRAPPER_VERSION 1 diff --git a/src/rng/backends/mklgpu/mrg32k3a.cpp b/src/rng/backends/mklgpu/mrg32k3a.cpp index f603b69b0..1dd44775d 100644 --- a/src/rng/backends/mklgpu/mrg32k3a.cpp +++ b/src/rng/backends/mklgpu/mrg32k3a.cpp @@ -29,7 +29,7 @@ #include "oneapi/math/rng/detail/engine_impl.hpp" #include "oneapi/math/rng/engines.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp" +#include "oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp" #include "mkl_internal_rng_gpu.hpp" diff --git a/src/rng/backends/mklgpu/philox4x32x10.cpp b/src/rng/backends/mklgpu/philox4x32x10.cpp index 25e318f90..6410876fb 100644 --- a/src/rng/backends/mklgpu/philox4x32x10.cpp +++ b/src/rng/backends/mklgpu/philox4x32x10.cpp @@ -29,7 +29,7 @@ #include "oneapi/math/rng/detail/engine_impl.hpp" #include "oneapi/math/rng/engines.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/rng/detail/mklgpu/onemkl_rng_mklgpu.hpp" +#include "oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp" #include "mkl_internal_rng_gpu.hpp" diff --git a/src/rng/backends/rocrand/CMakeLists.txt b/src/rng/backends/rocrand/CMakeLists.txt index 509996611..f4417f3c1 100644 --- a/src/rng/backends/rocrand/CMakeLists.txt +++ b/src/rng/backends/rocrand/CMakeLists.txt @@ -52,7 +52,7 @@ # perform publicly and display publicly, and to permit others to do so. # ================================================================================= -set(LIB_NAME onemkl_rng_rocrand) +set(LIB_NAME onemath_rng_rocrand) set(LIB_OBJ ${LIB_NAME}_obj) find_package(hip REQUIRED) find_package(rocrand REQUIRED) @@ -63,7 +63,7 @@ set(SOURCES philox4x32x10.cpp mrg32k3a.cpp $<$: add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) -add_dependencies(onemkl_backend_libs_rng ${LIB_NAME}) +add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) target_include_directories( ${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src diff --git a/src/rng/backends/rocrand/mrg32k3a.cpp b/src/rng/backends/rocrand/mrg32k3a.cpp index d0c2fef59..419434159 100644 --- a/src/rng/backends/rocrand/mrg32k3a.cpp +++ b/src/rng/backends/rocrand/mrg32k3a.cpp @@ -75,7 +75,7 @@ #include "rocrand_helper.hpp" #include "rocrand_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/rng/detail/rocrand/onemkl_rng_rocrand.hpp" +#include "oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp" #include "oneapi/math/rng/detail/engine_impl.hpp" #include "oneapi/math/rng/engines.hpp" @@ -122,7 +122,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform, status, engine_, r_ptr, n); }); @@ -140,7 +140,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform_double, status, engine_, r_ptr, n); }); @@ -159,7 +159,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = ib.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate, status, engine_, r_ptr, n); }); @@ -177,7 +177,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform, status, engine_, r_ptr, n); }); @@ -195,7 +195,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform_double, status, engine_, r_ptr, n); }); @@ -213,7 +213,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -230,7 +230,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal_double, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -247,7 +247,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -264,7 +264,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal_double, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -281,7 +281,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -298,7 +298,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal_double, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -315,7 +315,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -332,7 +332,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal_double, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -362,7 +362,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](std::int32_t* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](std::int32_t* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_poisson, status, engine_, (std::uint32_t*)r_ptr, n, distr.lambda()); @@ -378,7 +378,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_poisson, status, engine_, r_ptr, n, distr.lambda()); @@ -394,7 +394,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.template get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](uint32_t* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](uint32_t* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate, status, engine_, r_ptr, n); }); @@ -412,7 +412,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::event::wait_and_throw(dependencies); queue_ .submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform, status, engine_, r, n); }); @@ -430,7 +430,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::event::wait_and_throw(dependencies); queue_ .submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform_double, status, engine_, r, n); }); @@ -450,7 +450,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { n * sizeof(std::uint32_t), queue_.get_device(), queue_.get_context()); queue_ .submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate, status, engine_, ib, n); }); @@ -468,7 +468,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::event::wait_and_throw(dependencies); queue_ .submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform, status, engine_, r, n); }); @@ -486,7 +486,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::event::wait_and_throw(dependencies); queue_ .submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform_double, status, engine_, r, n); }); @@ -504,7 +504,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -522,7 +522,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal_double, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -539,7 +539,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -556,7 +556,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal_double, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -574,7 +574,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal, status, engine_, r, n, distr.m(), distr.s()); @@ -592,7 +592,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal_double, status, engine_, r, n, distr.m(), distr.s()); @@ -609,7 +609,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal, status, engine_, r, n, distr.m(), distr.s()); @@ -626,7 +626,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal_double, status, engine_, r, n, distr.m(), distr.s()); @@ -661,7 +661,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int32_t* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_poisson, status, engine_, (std::uint32_t*)r, n, distr.lambda()); @@ -679,7 +679,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_poisson, status, engine_, r, n, distr.lambda()); }); @@ -694,7 +694,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate, status, engine_, r, n); }); diff --git a/src/rng/backends/rocrand/philox4x32x10.cpp b/src/rng/backends/rocrand/philox4x32x10.cpp index 2e21534a1..3bacaa46d 100644 --- a/src/rng/backends/rocrand/philox4x32x10.cpp +++ b/src/rng/backends/rocrand/philox4x32x10.cpp @@ -75,7 +75,7 @@ #include "rocrand_helper.hpp" #include "rocrand_task.hpp" #include "oneapi/math/exceptions.hpp" -#include "oneapi/math/rng/detail/rocrand/onemkl_rng_rocrand.hpp" +#include "oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp" #include "oneapi/math/rng/detail/engine_impl.hpp" // #include "oneapi/math/rng/engines.hpp" @@ -144,7 +144,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform, status, engine_, r_ptr, n); }); @@ -162,7 +162,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform_double, status, engine_, r_ptr, n); }); @@ -181,7 +181,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = ib.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate, status, engine_, r_ptr, n); }); @@ -199,7 +199,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform, status, engine_, r_ptr, n); }); @@ -217,7 +217,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform_double, status, engine_, r_ptr, n); }); @@ -235,7 +235,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -252,7 +252,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal_double, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -269,7 +269,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -286,7 +286,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal_double, status, engine_, r_ptr, n, distr.mean(), distr.stddev()); @@ -303,7 +303,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -320,7 +320,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal_double, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -337,7 +337,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](float* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -354,7 +354,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](double* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal_double, status, engine_, r_ptr, n, distr.m(), distr.s()); @@ -384,7 +384,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](std::int32_t* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](std::int32_t* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_poisson, status, engine_, (std::uint32_t*)r_ptr, n, distr.lambda()); @@ -400,7 +400,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate_poisson, status, engine_, r_ptr, n, distr.lambda()); @@ -416,7 +416,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.template get_access(cgh); - onemkl_rocrand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { + onemath_rocrand_host_task(cgh, acc, engine_, [=](std::uint32_t* r_ptr) { rocrand_status status; ROCRAND_CALL(rocrand_generate, status, engine_, r_ptr, n); }); @@ -434,7 +434,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::event::wait_and_throw(dependencies); queue_ .submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform, status, engine_, r, n); }); @@ -452,7 +452,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::event::wait_and_throw(dependencies); queue_ .submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform_double, status, engine_, r, n); }); @@ -472,7 +472,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { n * sizeof(std::uint32_t), queue_.get_device(), queue_.get_context()); queue_ .submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate, status, engine_, ib, n); }); @@ -490,7 +490,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::event::wait_and_throw(dependencies); queue_ .submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform, status, engine_, r, n); }); @@ -508,7 +508,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { sycl::event::wait_and_throw(dependencies); queue_ .submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_uniform_double, status, engine_, r, n); }); @@ -526,7 +526,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -544,7 +544,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal_double, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -561,7 +561,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -578,7 +578,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_normal_double, status, engine_, r, n, distr.mean(), distr.stddev()); @@ -596,7 +596,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal, status, engine_, r, n, distr.m(), distr.s()); @@ -614,7 +614,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal_double, status, engine_, r, n, distr.m(), distr.s()); @@ -631,7 +631,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal, status, engine_, r, n, distr.m(), distr.s()); @@ -648,7 +648,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_log_normal_double, status, engine_, r, n, distr.m(), distr.s()); @@ -683,7 +683,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int32_t* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_poisson, status, engine_, (std::uint32_t*)r, n, distr.lambda()); @@ -700,7 +700,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::uint32_t* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate_poisson, status, engine_, r, n, distr.lambda()); }); @@ -715,7 +715,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { - onemkl_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { + onemath_rocrand_host_task(cgh, engine_, [=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_generate, status, engine_, r, n); }); diff --git a/src/rng/backends/rocrand/rocrand_task.hpp b/src/rng/backends/rocrand/rocrand_task.hpp index bad40a9e5..3256a6e96 100644 --- a/src/rng/backends/rocrand/rocrand_task.hpp +++ b/src/rng/backends/rocrand/rocrand_task.hpp @@ -73,12 +73,12 @@ static inline void host_task_internal(H &cgh, E e, F f) { } #endif template -static inline void onemkl_rocrand_host_task(H &cgh, A acc, E e, F f) { +static inline void onemath_rocrand_host_task(H &cgh, A acc, E e, F f) { host_task_internal(cgh, acc, e, f); } template -static inline void onemkl_rocrand_host_task(H &cgh, Engine e, F f) { +static inline void onemath_rocrand_host_task(H &cgh, Engine e, F f) { host_task_internal(cgh, e, f); } diff --git a/src/rng/backends/rocrand/rocrand_wrappers.cpp b/src/rng/backends/rocrand/rocrand_wrappers.cpp index bdbfa6fb1..7528b7c0f 100644 --- a/src/rng/backends/rocrand/rocrand_wrappers.cpp +++ b/src/rng/backends/rocrand/rocrand_wrappers.cpp @@ -58,7 +58,7 @@ * so. ******************************************************************************/ -#include "oneapi/math/rng/detail/rocrand/onemkl_rng_rocrand.hpp" +#include "oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp" #include "rng/function_table.hpp" #define WRAPPER_VERSION 1 diff --git a/src/sparse_blas/CMakeLists.txt b/src/sparse_blas/CMakeLists.txt index b01cc63fd..cec5ad720 100644 --- a/src/sparse_blas/CMakeLists.txt +++ b/src/sparse_blas/CMakeLists.txt @@ -20,29 +20,29 @@ add_subdirectory(backends) if(BUILD_SHARED_LIBS) - add_library(onemkl_sparse_blas OBJECT) - target_sources(onemkl_sparse_blas PRIVATE sparse_blas_loader.cpp) - target_include_directories(onemkl_sparse_blas + add_library(onemath_sparse_blas OBJECT) + target_sources(onemath_sparse_blas PRIVATE sparse_blas_loader.cpp) + target_include_directories(onemath_sparse_blas PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMKL_GENERATED_INCLUDE_PATH} - $ + $ ) - target_compile_options(onemkl_sparse_blas PRIVATE ${ONEMKL_BUILD_COPT}) + target_compile_options(onemath_sparse_blas PRIVATE ${ONEMKL_BUILD_COPT}) - set_target_properties(onemkl_sparse_blas PROPERTIES + set_target_properties(onemath_sparse_blas PROPERTIES POSITION_INDEPENDENT_CODE ON ) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) - add_sycl_to_target(TARGET onemkl_sparse_blas SOURCES sparse_blas_loader.cpp) + add_sycl_to_target(TARGET onemath_sparse_blas SOURCES sparse_blas_loader.cpp) else() - target_link_libraries(onemkl_sparse_blas PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(onemath_sparse_blas PUBLIC ONEMKL::SYCL::SYCL) endif() include(WarningsUtils) - target_link_libraries(onemkl_sparse_blas PRIVATE onemkl_warnings) + target_link_libraries(onemath_sparse_blas PRIVATE onemath_warnings) endif() diff --git a/src/sparse_blas/backends/CMakeLists.txt b/src/sparse_blas/backends/CMakeLists.txt index 294040808..5d123cced 100644 --- a/src/sparse_blas/backends/CMakeLists.txt +++ b/src/sparse_blas/backends/CMakeLists.txt @@ -17,8 +17,8 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -add_custom_target(onemkl_backend_libs_sparse_blas) -add_dependencies(onemkl_backend_libs onemkl_backend_libs_sparse_blas) +add_custom_target(onemath_backend_libs_sparse_blas) +add_dependencies(onemath_backend_libs onemath_backend_libs_sparse_blas) if(ENABLE_MKLCPU_BACKEND) add_subdirectory(mklcpu) diff --git a/src/sparse_blas/backends/mklcpu/CMakeLists.txt b/src/sparse_blas/backends/mklcpu/CMakeLists.txt index e41cae268..0f78e2e04 100644 --- a/src/sparse_blas/backends/mklcpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklcpu/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_sparse_blas_mklcpu) +set(LIB_NAME onemath_sparse_blas_mklcpu) set(LIB_OBJ ${LIB_NAME}_obj) include(WarningsUtils) @@ -28,7 +28,7 @@ add_library(${LIB_OBJ} OBJECT mklcpu_operations.cpp $<$: mklcpu_wrappers.cpp> ) -add_dependencies(onemkl_backend_libs_sparse_blas ${LIB_NAME}) +add_dependencies(onemath_backend_libs_sparse_blas ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include @@ -43,13 +43,13 @@ if(TARGET MKL::MKL_SYCL::SPARSE) target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL PUBLIC MKL::MKL_SYCL::SPARSE - PRIVATE onemkl_warnings + PRIVATE onemath_warnings ) else() target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL PUBLIC MKL::MKL_DPCPP - PRIVATE onemkl_warnings + PRIVATE onemath_warnings ) endif() diff --git a/src/sparse_blas/backends/mklgpu/CMakeLists.txt b/src/sparse_blas/backends/mklgpu/CMakeLists.txt index cd25babc2..15fc23a4c 100644 --- a/src/sparse_blas/backends/mklgpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklgpu/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -set(LIB_NAME onemkl_sparse_blas_mklgpu) +set(LIB_NAME onemath_sparse_blas_mklgpu) set(LIB_OBJ ${LIB_NAME}_obj) include(WarningsUtils) @@ -28,7 +28,7 @@ add_library(${LIB_OBJ} OBJECT mklgpu_operations.cpp $<$: mklgpu_wrappers.cpp> ) -add_dependencies(onemkl_backend_libs_sparse_blas ${LIB_NAME}) +add_dependencies(onemath_backend_libs_sparse_blas ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include @@ -43,13 +43,13 @@ if(TARGET MKL::MKL_SYCL::SPARSE) target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL PUBLIC MKL::MKL_SYCL::SPARSE - PRIVATE onemkl_warnings + PRIVATE onemath_warnings ) else() target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL PUBLIC MKL::MKL_DPCPP - PRIVATE onemkl_warnings + PRIVATE onemath_warnings ) endif() diff --git a/tests/README.md b/tests/README.md index 3a8346057..71511711a 100644 --- a/tests/README.md +++ b/tests/README.md @@ -8,4 +8,4 @@ See [Building and Running Tests](https://oneapi-src.github.io/oneMKL/building_an [GoogleTest](https://github.com/google/googletest) is used as the unit-testing framework. -*Refer to `/deps/googletest/LICENSE` for GoogleTest license.* +*Refer to `/deps/googletest/LICENSE` for GoogleTest license.* diff --git a/tests/unit_tests/CMakeLists.txt b/tests/unit_tests/CMakeLists.txt index 235f8c8e5..9d3f4da8d 100644 --- a/tests/unit_tests/CMakeLists.txt +++ b/tests/unit_tests/CMakeLists.txt @@ -119,7 +119,7 @@ foreach(domain ${TEST_TARGET_DOMAINS}) ${CMAKE_DL_LIBS} ${${domain}_TEST_LINK} ONEMKL::SYCL::SYCL - onemkl + onemath ${${domain}_TEST_LIST_RT} ) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) @@ -128,68 +128,68 @@ foreach(domain ${TEST_TARGET_DOMAINS}) endif() if(ENABLE_MKLCPU_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_${domain}_mklcpu) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_mklcpu) + add_dependencies(test_main_${domain}_ct onemath_${domain}_mklcpu) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_mklcpu) endif() if(ENABLE_MKLGPU_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_${domain}_mklgpu) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_mklgpu) + add_dependencies(test_main_${domain}_ct onemath_${domain}_mklgpu) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_mklgpu) endif() if(domain STREQUAL "blas" AND ENABLE_CUBLAS_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_${domain}_cublas) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_cublas) + add_dependencies(test_main_${domain}_ct onemath_${domain}_cublas) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_cublas) endif() if(domain STREQUAL "blas" AND ENABLE_ROCBLAS_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_${domain}_rocblas) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_rocblas) + add_dependencies(test_main_${domain}_ct onemath_${domain}_rocblas) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_rocblas) endif() if(domain STREQUAL "blas" AND ENABLE_NETLIB_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_${domain}_netlib) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_netlib) + add_dependencies(test_main_${domain}_ct onemath_${domain}_netlib) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_netlib) endif() if(domain STREQUAL "blas" AND ENABLE_PORTBLAS_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_${domain}_portblas) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_portblas) + add_dependencies(test_main_${domain}_ct onemath_${domain}_portblas) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_portblas) endif() if(domain STREQUAL "lapack" AND ENABLE_CUSOLVER_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_${domain}_cusolver) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_cusolver) + add_dependencies(test_main_${domain}_ct onemath_${domain}_cusolver) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_cusolver) endif() if(domain STREQUAL "lapack" AND ENABLE_ROCSOLVER_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_${domain}_rocsolver) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_rocsolver) + add_dependencies(test_main_${domain}_ct onemath_${domain}_rocsolver) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_rocsolver) endif() if(domain STREQUAL "rng" AND ENABLE_CURAND_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_${domain}_curand) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_curand) + add_dependencies(test_main_${domain}_ct onemath_${domain}_curand) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_curand) endif() if(domain STREQUAL "rng" AND ENABLE_ROCRAND_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_${domain}_rocrand) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_rocrand) + add_dependencies(test_main_${domain}_ct onemath_${domain}_rocrand) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_rocrand) endif() if(domain STREQUAL "dft" AND ENABLE_CUFFT_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_${domain}_cufft) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_${domain}_cufft) + add_dependencies(test_main_${domain}_ct onemath_${domain}_cufft) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_cufft) endif() if(domain STREQUAL "dft" AND ENABLE_ROCFFT_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_dft_rocfft) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_dft_rocfft) + add_dependencies(test_main_${domain}_ct onemath_dft_rocfft) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_dft_rocfft) endif() if(domain STREQUAL "dft" AND ENABLE_PORTFFT_BACKEND) - add_dependencies(test_main_${domain}_ct onemkl_dft_portfft) - list(APPEND ONEMKL_LIBRARIES_${domain} onemkl_dft_portfft) + add_dependencies(test_main_${domain}_ct onemath_dft_portfft) + list(APPEND ONEMKL_LIBRARIES_${domain} onemath_dft_portfft) endif() target_link_libraries(test_main_${domain}_ct PUBLIC @@ -217,7 +217,7 @@ foreach(domain ${TEST_TARGET_DOMAINS}) if(BUILD_SHARED_LIBS) set_target_properties(test_main_${domain}_rt - PROPERTIES BUILD_RPATH $) + PROPERTIES BUILD_RPATH $) # Find individual tests within executable gtest_discover_tests(test_main_${domain}_rt PROPERTIES BUILD_RPATH ${CMAKE_BINARY_DIR}/lib diff --git a/tests/unit_tests/blas/include/onemath_blas_helper.hpp b/tests/unit_tests/blas/include/onemath_blas_helper.hpp index 30645840f..081fb2234 100644 --- a/tests/unit_tests/blas/include/onemath_blas_helper.hpp +++ b/tests/unit_tests/blas/include/onemath_blas_helper.hpp @@ -27,7 +27,7 @@ typedef enum { CblasRowOffset = 101, CblasColOffset = 102, CblasFixOffset = 103 } CBLAS_OFFSET; /** - * Helper methods for converting between onemkl types and their BLAS + * Helper methods for converting between onemath types and their BLAS * equivalents. */ diff --git a/tests/unit_tests/dft/source/CMakeLists.txt b/tests/unit_tests/dft/source/CMakeLists.txt index 364ad564f..e8fc29eca 100644 --- a/tests/unit_tests/dft/source/CMakeLists.txt +++ b/tests/unit_tests/dft/source/CMakeLists.txt @@ -36,7 +36,7 @@ if (BUILD_SHARED_LIBS) else () target_link_libraries(dft_source_rt PUBLIC ONEMKL::SYCL::SYCL) endif () - target_link_libraries(dft_source_rt PRIVATE onemkl_warnings) + target_link_libraries(dft_source_rt PRIVATE onemath_warnings) endif () add_library(dft_source_ct OBJECT ${DFT_SOURCES}) @@ -53,5 +53,5 @@ if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) else () target_link_libraries(dft_source_ct PUBLIC ONEMKL::SYCL::SYCL) endif () -target_link_libraries(dft_source_ct PRIVATE onemkl_warnings) +target_link_libraries(dft_source_ct PRIVATE onemath_warnings) diff --git a/tests/unit_tests/sparse_blas/source/CMakeLists.txt b/tests/unit_tests/sparse_blas/source/CMakeLists.txt index a9271bfa2..53b34e2ce 100644 --- a/tests/unit_tests/sparse_blas/source/CMakeLists.txt +++ b/tests/unit_tests/sparse_blas/source/CMakeLists.txt @@ -43,7 +43,7 @@ if (BUILD_SHARED_LIBS) else () target_link_libraries(spblas_source_rt PUBLIC ONEMKL::SYCL::SYCL) endif () - target_link_libraries(spblas_source_rt PRIVATE onemkl_warnings) + target_link_libraries(spblas_source_rt PRIVATE onemath_warnings) endif () add_library(spblas_source_ct OBJECT ${SPBLAS_SOURCES}) @@ -60,4 +60,4 @@ if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) else () target_link_libraries(spblas_source_ct PUBLIC ONEMKL::SYCL::SYCL) endif () -target_link_libraries(spblas_source_ct PRIVATE onemkl_warnings) +target_link_libraries(spblas_source_ct PRIVATE onemath_warnings) From fb37bae5cb28fa7138c52ff78e502c9d469c9160 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 8 Oct 2024 12:01:50 +0200 Subject: [PATCH 04/41] ONEMKL -> ONEMATH --- CMakeLists.txt | 34 +- cmake/FindCompiler.cmake | 8 +- cmake/FindNETLIB.cmake | 4 +- cmake/FindcuBLAS.cmake | 8 +- cmake/FindcuRAND.cmake | 16 +- cmake/FindcuSOLVER.cmake | 6 +- cmake/WarningsUtils.cmake | 8 +- .../building_the_project_with_adaptivecpp.rst | 2 +- docs/create_new_backend.rst | 34 +- .../level3/CMakeLists.txt | 6 +- .../level3/CMakeLists.txt | 2 +- .../compile_time_dispatching/CMakeLists.txt | 4 +- .../dft/run_time_dispatching/CMakeLists.txt | 2 +- .../compile_time_dispatching/CMakeLists.txt | 6 +- .../run_time_dispatching/CMakeLists.txt | 2 +- .../compile_time_dispatching/CMakeLists.txt | 6 +- examples/rng/device/CMakeLists.txt | 4 +- .../rng/run_time_dispatching/CMakeLists.txt | 2 +- .../compile_time_dispatching/CMakeLists.txt | 2 +- .../run_time_dispatching/CMakeLists.txt | 2 +- include/oneapi/math/blas.hpp | 18 +- .../oneapi/math/blas/detail/blas_loader.hpp | 6 +- .../oneapi/math/blas/detail/blas_loader.hxx | 980 +++++++++--------- .../detail/cublas/onemath_blas_cublas.hpp | 6 +- .../detail/mklcpu/onemath_blas_mklcpu.hpp | 6 +- .../detail/mklgpu/onemath_blas_mklgpu.hpp | 6 +- .../detail/netlib/onemath_blas_netlib.hpp | 6 +- .../blas/detail/onemath_blas_backends.hxx | 980 +++++++++--------- .../detail/portblas/onemath_blas_portblas.hpp | 6 +- .../detail/rocblas/onemath_blas_rocblas.hpp | 6 +- .../oneapi/math/detail/backend_selector.hpp | 6 +- .../detail/backend_selector_predicates.hpp | 22 +- include/oneapi/math/detail/backends.hpp | 6 +- include/oneapi/math/detail/backends_table.hpp | 60 +- include/oneapi/math/detail/exceptions.hpp | 6 +- include/oneapi/math/detail/export.hpp | 18 +- include/oneapi/math/detail/get_device_id.hpp | 6 +- include/oneapi/math/dft.hpp | 6 +- include/oneapi/math/dft/backward.hpp | 6 +- include/oneapi/math/dft/descriptor.hpp | 6 +- .../oneapi/math/dft/detail/commit_impl.hpp | 6 +- .../dft/detail/cufft/onemath_dft_cufft.hpp | 6 +- .../math/dft/detail/descriptor_impl.hpp | 16 +- include/oneapi/math/dft/detail/dft_ct.hxx | 34 +- include/oneapi/math/dft/detail/dft_loader.hpp | 8 +- .../dft/detail/external_workspace_helper.hpp | 6 +- .../dft/detail/mklcpu/onemath_dft_mklcpu.hpp | 6 +- .../dft/detail/mklgpu/onemath_dft_mklgpu.hpp | 6 +- .../detail/portfft/onemath_dft_portfft.hpp | 6 +- .../dft/detail/rocfft/onemath_dft_rocfft.hpp | 6 +- include/oneapi/math/dft/detail/types_impl.hpp | 6 +- include/oneapi/math/dft/forward.hpp | 6 +- include/oneapi/math/dft/types.hpp | 6 +- include/oneapi/math/exceptions.hpp | 6 +- include/oneapi/math/lapack.hpp | 8 +- .../cusolver/onemath_lapack_cusolver.hpp | 6 +- .../cusolver/onemath_lapack_cusolver.hxx | 606 +++++------ .../math/lapack/detail/lapack_loader.hpp | 796 +++++++------- .../mkl_common/onemath_lapack_backends.hxx | 796 +++++++------- .../rocsolver/onemath_lapack_rocsolver.hpp | 6 +- .../rocsolver/onemath_lapack_rocsolver.hxx | 606 +++++------ include/oneapi/math/rng.hpp | 6 +- .../rng/detail/curand/onemath_rng_curand.hpp | 14 +- .../oneapi/math/rng/detail/engine_impl.hpp | 6 +- .../rng/detail/mklcpu/onemath_rng_mklcpu.hpp | 14 +- .../rng/detail/mklgpu/onemath_rng_mklgpu.hpp | 14 +- include/oneapi/math/rng/detail/rng_loader.hpp | 14 +- .../detail/rocrand/onemath_rng_rocrand.hpp | 14 +- include/oneapi/math/rng/distributions.hpp | 6 +- include/oneapi/math/rng/engines.hpp | 30 +- include/oneapi/math/rng/functions.hpp | 6 +- include/oneapi/math/rng/predicates.hpp | 10 +- include/oneapi/math/sparse_blas.hpp | 10 +- .../math/sparse_blas/detail/handles.hpp | 6 +- .../math/sparse_blas/detail/helper_types.hpp | 6 +- .../mklcpu/onemath_sparse_blas_mklcpu.hpp | 6 +- .../detail/mklcpu/sparse_blas_ct.hpp | 6 +- .../mklgpu/onemath_sparse_blas_mklgpu.hpp | 6 +- .../detail/mklgpu/sparse_blas_ct.hpp | 6 +- .../detail/onemath_sparse_blas_backends.hxx | 74 +- .../sparse_blas/detail/operation_types.hpp | 6 +- .../sparse_blas/detail/sparse_blas_rt.hpp | 6 +- .../oneapi/math/sparse_blas/matrix_view.hpp | 6 +- include/oneapi/math/sparse_blas/types.hpp | 6 +- include/oneapi/math/types.hpp | 6 +- scripts/generate_cmake.py | 4 +- src/CMakeLists.txt | 46 +- src/blas/CMakeLists.txt | 6 +- src/blas/backends/cublas/CMakeLists.txt | 16 +- .../backends/cublas/cublas_scope_handle.cpp | 4 +- .../backends/cublas/cublas_scope_handle.hpp | 6 +- src/blas/backends/cublas/cublas_task.hpp | 4 +- .../backends/mkl_common/mkl_blas_backend.hxx | 16 +- src/blas/backends/mklcpu/CMakeLists.txt | 10 +- src/blas/backends/mklcpu/mklcpu_wrappers.cpp | 2 +- src/blas/backends/mklgpu/CMakeLists.txt | 10 +- src/blas/backends/mklgpu/mklgpu_wrappers.cpp | 2 +- src/blas/backends/netlib/CMakeLists.txt | 8 +- src/blas/backends/netlib/netlib_wrappers.cpp | 2 +- src/blas/backends/portblas/CMakeLists.txt | 58 +- .../backends/portblas/portblas_wrappers.cpp | 2 +- src/blas/backends/rocblas/CMakeLists.txt | 22 +- .../backends/rocblas/rocblas_scope_handle.cpp | 4 +- .../backends/rocblas/rocblas_scope_handle.hpp | 6 +- src/blas/backends/rocblas/rocblas_task.hpp | 4 +- src/config.hpp.in | 40 +- src/dft/CMakeLists.txt | 6 +- .../backend_backward_instantiations.cxx | 30 +- .../backend_forward_instantiations.cxx | 30 +- src/dft/backends/cufft/CMakeLists.txt | 12 +- src/dft/backends/cufft/backward.cpp | 16 +- src/dft/backends/cufft/execute_helper.hpp | 6 +- src/dft/backends/cufft/forward.cpp | 16 +- src/dft/backends/mklcpu/CMakeLists.txt | 14 +- src/dft/backends/mklcpu/backward.cpp | 16 +- .../backends/mklcpu/commit_derived_impl.hpp | 6 +- src/dft/backends/mklcpu/forward.cpp | 16 +- src/dft/backends/mklcpu/mklcpu_helpers.hpp | 6 +- src/dft/backends/mklgpu/CMakeLists.txt | 12 +- src/dft/backends/mklgpu/backward.cpp | 16 +- src/dft/backends/mklgpu/forward.cpp | 16 +- src/dft/backends/mklgpu/mklgpu_helpers.hpp | 6 +- src/dft/backends/portfft/CMakeLists.txt | 14 +- src/dft/backends/portfft/portfft_helper.hpp | 4 +- src/dft/backends/rocfft/CMakeLists.txt | 12 +- src/dft/backends/rocfft/backward.cpp | 16 +- src/dft/backends/rocfft/execute_helper.hpp | 6 +- src/dft/backends/rocfft/forward.cpp | 16 +- src/dft/backends/rocfft/rocfft_handle.hpp | 4 +- src/dft/descriptor_config_helper.hpp | 6 +- src/dft/execute_helper_generic.hpp | 6 +- src/include/dtype_string.hpp | 6 +- src/include/runtime_support_helper.hpp | 6 +- src/lapack/CMakeLists.txt | 6 +- src/lapack/backends/cusolver/CMakeLists.txt | 12 +- .../cusolver/cusolver_scope_handle.cpp | 4 +- .../cusolver/cusolver_scope_handle.hpp | 6 +- .../backends/cusolver/cusolver_task.hpp | 4 +- .../backends/cusolver/cusolver_wrappers.cpp | 2 +- src/lapack/backends/mklcpu/CMakeLists.txt | 10 +- .../backends/mklcpu/lapack_cpu_wrappers.cpp | 2 +- src/lapack/backends/mklgpu/CMakeLists.txt | 10 +- .../backends/mklgpu/lapack_gpu_wrappers.cpp | 2 +- src/lapack/backends/rocsolver/CMakeLists.txt | 8 +- .../rocsolver/rocsolver_scope_handle.cpp | 4 +- .../rocsolver/rocsolver_scope_handle.hpp | 6 +- .../backends/rocsolver/rocsolver_task.hpp | 4 +- .../backends/rocsolver/rocsolver_wrappers.cpp | 2 +- src/rng/CMakeLists.txt | 6 +- src/rng/backends/curand/CMakeLists.txt | 4 +- src/rng/backends/curand/curand_wrappers.cpp | 2 +- src/rng/backends/mklcpu/CMakeLists.txt | 8 +- .../backends/mklcpu/mkl_rng_cpu_wrappers.cpp | 2 +- src/rng/backends/mklgpu/CMakeLists.txt | 10 +- .../backends/mklgpu/mkl_rng_gpu_wrappers.cpp | 2 +- src/rng/backends/rocrand/CMakeLists.txt | 4 +- src/rng/backends/rocrand/rocrand_wrappers.cpp | 2 +- src/sparse_blas/CMakeLists.txt | 6 +- .../backends/mkl_common/mkl_handles.hpp | 6 +- .../backends/mkl_common/mkl_helper.hpp | 6 +- .../backends/mklcpu/CMakeLists.txt | 10 +- .../backends/mklgpu/CMakeLists.txt | 10 +- src/sparse_blas/enum_data_types.hpp | 6 +- src/sparse_blas/function_table.hpp | 6 +- src/sparse_blas/generic_container.hpp | 6 +- src/sparse_blas/macros.hpp | 6 +- src/sparse_blas/matrix_view_comparison.hpp | 6 +- tests/unit_tests/CMakeLists.txt | 36 +- tests/unit_tests/blas/CMakeLists.txt | 4 +- tests/unit_tests/blas/batch/CMakeLists.txt | 4 +- .../unit_tests/blas/extensions/CMakeLists.txt | 4 +- .../blas/include/reference_blas_wrappers.hpp | 8 +- tests/unit_tests/blas/level1/CMakeLists.txt | 4 +- tests/unit_tests/blas/level2/CMakeLists.txt | 4 +- tests/unit_tests/blas/level3/CMakeLists.txt | 4 +- .../dft/include/compute_inplace.hpp | 6 +- .../dft/include/compute_inplace_real_real.hpp | 6 +- .../dft/include/compute_out_of_place.hpp | 6 +- .../compute_out_of_place_real_real.hpp | 6 +- .../unit_tests/dft/include/compute_tester.hpp | 6 +- .../unit_tests/dft/include/parseval_check.hpp | 6 +- .../unit_tests/dft/include/reference_dft.hpp | 6 +- tests/unit_tests/dft/include/test_common.hpp | 6 +- tests/unit_tests/dft/source/CMakeLists.txt | 4 +- tests/unit_tests/include/test_helper.hpp | 34 +- tests/unit_tests/lapack/source/CMakeLists.txt | 4 +- tests/unit_tests/main_test.cpp | 36 +- .../rng/device/moments/CMakeLists.txt | 4 +- .../rng/device/service/CMakeLists.txt | 4 +- .../rng/include/rng_test_common.hpp | 2 +- tests/unit_tests/rng/service/CMakeLists.txt | 4 +- .../rng/statistics_check/CMakeLists.txt | 4 +- .../sparse_blas/source/CMakeLists.txt | 4 +- 193 files changed, 3275 insertions(+), 3275 deletions(-) diff --git a/CMakeLists.txt b/CMakeLists.txt index c4cad8e17..e78525b68 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -60,7 +60,7 @@ option(ENABLE_CUFFT_BACKEND "Enable the cuFFT backend for the DFT interface" OFF option(ENABLE_ROCFFT_BACKEND "Enable the rocFFT backend for the DFT interface" OFF) option(ENABLE_PORTFFT_BACKEND "Enable the portFFT DFT backend for the DFT interface. Cannot be used with other DFT backends." OFF) -set(ONEMKL_SYCL_IMPLEMENTATION "dpc++" CACHE STRING "Name of the SYCL compiler") +set(ONEMATH_SYCL_IMPLEMENTATION "dpc++" CACHE STRING "Name of the SYCL compiler") set(HIP_TARGETS "" CACHE STRING "Target HIP architectures") ## Testing @@ -124,7 +124,7 @@ if (ENABLE_PORTFFT_BACKEND endif() # Define required CXX compilers before project -if(CMAKE_CXX_COMPILER OR NOT ONEMKL_SYCL_IMPLEMENTATION STREQUAL "dpc++") +if(CMAKE_CXX_COMPILER OR NOT ONEMATH_SYCL_IMPLEMENTATION STREQUAL "dpc++") if(WIN32) string(REPLACE "\\" "/" CMAKE_CXX_COMPILER ${CMAKE_CXX_COMPILER}) endif() @@ -165,7 +165,7 @@ else() endif() # Define required C compilers before project -if(CMAKE_C_COMPILER OR NOT ONEMKL_SYCL_IMPLEMENTATION STREQUAL "dpc++") +if(CMAKE_C_COMPILER OR NOT ONEMATH_SYCL_IMPLEMENTATION STREQUAL "dpc++") if(WIN32) string(REPLACE "\\" "/" CMAKE_C_COMPILER ${CMAKE_C_COMPILER}) endif() @@ -188,7 +188,7 @@ endif() project(oneMKL VERSION 0.5.0 LANGUAGES CXX) # Override default CXX compile/link lines for Windows after project -if(WIN32 AND ONEMKL_SYCL_IMPLEMENTATION STREQUAL "dpc++") +if(WIN32 AND ONEMATH_SYCL_IMPLEMENTATION STREQUAL "dpc++") set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wno-unused-function -w") foreach (flag_var CMAKE_CXX_FLAGS CMAKE_CXX_FLAGS_DEBUG CMAKE_CXX_FLAGS_RELEASE @@ -204,7 +204,7 @@ if(WIN32 AND ONEMKL_SYCL_IMPLEMENTATION STREQUAL "dpc++") endif() # Temporary disable sycl 2020 deprecations warnings for cuSOLVER and rocSOLVER -if(ONEMKL_SYCL_IMPLEMENTATION STREQUAL "dpc++" AND (ENABLE_ROCSOLVER_BACKEND)) +if(ONEMATH_SYCL_IMPLEMENTATION STREQUAL "dpc++" AND (ENABLE_ROCSOLVER_BACKEND)) set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -DSYCL2020_DISABLE_DEPRECATION_WARNINGS") endif() @@ -268,23 +268,23 @@ list(APPEND CMAKE_MODULE_PATH "${PROJECT_SOURCE_DIR}/cmake") # Add DPC++ options for Linux if(WIN32) - add_library(ONEMKL::SYCL::SYCL INTERFACE IMPORTED) + add_library(ONEMATH::SYCL::SYCL INTERFACE IMPORTED) else() # Find necessary packages - if(ONEMKL_SYCL_IMPLEMENTATION) - string( TOLOWER "${ONEMKL_SYCL_IMPLEMENTATION}" ONEMKL_SYCL_IMPLEMENTATION) - if (ONEMKL_SYCL_IMPLEMENTATION STREQUAL "hipsycl") + if(ONEMATH_SYCL_IMPLEMENTATION) + string( TOLOWER "${ONEMATH_SYCL_IMPLEMENTATION}" ONEMATH_SYCL_IMPLEMENTATION) + if (ONEMATH_SYCL_IMPLEMENTATION STREQUAL "hipsycl") message(STATUS "Looking for hipSYCL") find_package(hipSYCL CONFIG REQUIRED) set(USE_ADD_SYCL_TO_TARGET_INTEGRATION true) set (CMAKE_CXX_STANDARD 17) - add_library(ONEMKL::SYCL::SYCL INTERFACE IMPORTED) - elseif(ONEMKL_SYCL_IMPLEMENTATION STREQUAL "dpc++") + add_library(ONEMATH::SYCL::SYCL INTERFACE IMPORTED) + elseif(ONEMATH_SYCL_IMPLEMENTATION STREQUAL "dpc++") message(STATUS "Looking for dpc++") set(USE_ADD_SYCL_TO_TARGET_INTEGRATION false) find_package(Compiler REQUIRED) else() - message(FATAL_ERROR "SYCL implementation ${ONEMKL_SYCL_IMPLEMENTATION} is not known") + message(FATAL_ERROR "SYCL implementation ${ONEMATH_SYCL_IMPLEMENTATION} is not known") endif() else() message(STATUS "Looking for dpc++") @@ -294,8 +294,8 @@ else() endif() if(DEFINED REF_BLAS_ROOT) - find_file(ONEMKL_REF_BLAS_LIBNAME NAMES blas.dll libblas.so HINTS ${REF_BLAS_ROOT} PATH_SUFFIXES lib lib64) - find_file(ONEMKL_REF_CBLAS_LIBNAME NAMES cblas.dll libcblas.so HINTS ${REF_BLAS_ROOT} PATH_SUFFIXES lib lib64) + find_file(ONEMATH_REF_BLAS_LIBNAME NAMES blas.dll libblas.so HINTS ${REF_BLAS_ROOT} PATH_SUFFIXES lib lib64) + find_file(ONEMATH_REF_CBLAS_LIBNAME NAMES cblas.dll libcblas.so HINTS ${REF_BLAS_ROOT} PATH_SUFFIXES lib lib64) endif() # Add source directory and output to bin/ @@ -331,9 +331,9 @@ write_basic_package_version_file( COMPATIBILITY AnyNewerVersion ) -export(EXPORT oneMKLTargets - FILE "${CMAKE_CURRENT_BINARY_DIR}/oneMKLTargets.cmake" - NAMESPACE ONEMKL:: +export(EXPORT oneMathTargets + FILE "${CMAKE_CURRENT_BINARY_DIR}/oneMathTargets.cmake" + NAMESPACE ONEMATH:: ) configure_file("${PROJECT_SOURCE_DIR}/cmake/oneMKLConfig.cmake" "${CMAKE_CURRENT_BINARY_DIR}/oneMKLConfig.cmake" diff --git a/cmake/FindCompiler.cmake b/cmake/FindCompiler.cmake index 556211999..49cce8996 100644 --- a/cmake/FindCompiler.cmake +++ b/cmake/FindCompiler.cmake @@ -30,7 +30,7 @@ if(is_dpcpp) message(FATAL_ERROR "SYCL library is not found in ${SYCL_BINARY_DIR}/../lib, PATH, and LIBRARY_PATH") endif() - add_library(ONEMKL::SYCL::SYCL INTERFACE IMPORTED) + add_library(ONEMATH::SYCL::SYCL INTERFACE IMPORTED) if(UNIX) set(UNIX_INTERFACE_COMPILE_OPTIONS -fsycl) set(UNIX_INTERFACE_LINK_OPTIONS -fsycl) @@ -53,18 +53,18 @@ if(is_dpcpp) endif() if(ENABLE_CURAND_BACKEND OR ENABLE_CUSOLVER_BACKEND OR ENABLE_ROCBLAS_BACKEND OR ENABLE_ROCRAND_BACKEND OR ENABLE_ROCSOLVER_BACKEND) - set_target_properties(ONEMKL::SYCL::SYCL PROPERTIES + set_target_properties(ONEMATH::SYCL::SYCL PROPERTIES INTERFACE_COMPILE_OPTIONS "${UNIX_INTERFACE_COMPILE_OPTIONS}" INTERFACE_LINK_OPTIONS "${UNIX_INTERFACE_LINK_OPTIONS}" INTERFACE_LINK_LIBRARIES ${SYCL_LIBRARY}) else() - set_target_properties(ONEMKL::SYCL::SYCL PROPERTIES + set_target_properties(ONEMATH::SYCL::SYCL PROPERTIES INTERFACE_COMPILE_OPTIONS "-fsycl" INTERFACE_LINK_OPTIONS "-fsycl" INTERFACE_LINK_LIBRARIES ${SYCL_LIBRARY}) endif() else() - set_target_properties(ONEMKL::SYCL::SYCL PROPERTIES + set_target_properties(ONEMATH::SYCL::SYCL PROPERTIES INTERFACE_COMPILE_OPTIONS "-fsycl" INTERFACE_LINK_LIBRARIES ${SYCL_LIBRARY}) endif() diff --git a/cmake/FindNETLIB.cmake b/cmake/FindNETLIB.cmake index a066f4043..788e92ec4 100644 --- a/cmake/FindNETLIB.cmake +++ b/cmake/FindNETLIB.cmake @@ -36,6 +36,6 @@ list(APPEND NETLIB_LINK ${NETLIB_BLAS_LIBRARY}) find_package_handle_standard_args(NETLIB REQUIRED_VARS NETLIB_INCLUDE NETLIB_LINK) -add_library(ONEMKL::NETLIB::NETLIB UNKNOWN IMPORTED) -set_target_properties(ONEMKL::NETLIB::NETLIB PROPERTIES IMPORTED_LOCATION ${NETLIB_CBLAS_LIBRARY}) +add_library(ONEMATH::NETLIB::NETLIB UNKNOWN IMPORTED) +set_target_properties(ONEMATH::NETLIB::NETLIB PROPERTIES IMPORTED_LOCATION ${NETLIB_CBLAS_LIBRARY}) diff --git a/cmake/FindcuBLAS.cmake b/cmake/FindcuBLAS.cmake index c26a62f6b..1357c331e 100644 --- a/cmake/FindcuBLAS.cmake +++ b/cmake/FindcuBLAS.cmake @@ -35,8 +35,8 @@ find_package(Threads REQUIRED) include(FindPackageHandleStandardArgs) -if(NOT TARGET ONEMKL::cuBLAS::cuBLAS) - add_library(ONEMKL::cuBLAS::cuBLAS SHARED IMPORTED) +if(NOT TARGET ONEMATH::cuBLAS::cuBLAS) + add_library(ONEMATH::cuBLAS::cuBLAS SHARED IMPORTED) if(USE_ADD_SYCL_TO_TARGET_INTEGRATION) find_package_handle_standard_args(cuBLAS REQUIRED_VARS @@ -46,7 +46,7 @@ if(NOT TARGET ONEMKL::cuBLAS::cuBLAS) CUDA_CUDART_LIBRARY CUDA_CUDA_LIBRARY ) - set_target_properties(ONEMKL::cuBLAS::cuBLAS PROPERTIES + set_target_properties(ONEMATH::cuBLAS::cuBLAS PROPERTIES IMPORTED_LOCATION ${CUDA_cublas_LIBRARY} INTERFACE_INCLUDE_DIRECTORIES "${CUDA_TOOLKIT_INCLUDE}" INTERFACE_LINK_LIBRARIES "Threads::Threads;${CUDA_LIBRARIES};${CUDA_CUDART_LIBRARY};${CUDA_CUDA_LIBRARY}" @@ -60,7 +60,7 @@ if(NOT TARGET ONEMKL::cuBLAS::cuBLAS) CUDA_CUDA_LIBRARY OPENCL_INCLUDE_DIR ) - set_target_properties(ONEMKL::cuBLAS::cuBLAS PROPERTIES + set_target_properties(ONEMATH::cuBLAS::cuBLAS PROPERTIES IMPORTED_LOCATION ${CUDA_cublas_LIBRARY} INTERFACE_INCLUDE_DIRECTORIES "${OPENCL_INCLUDE_DIR};${CUDA_TOOLKIT_INCLUDE}" INTERFACE_LINK_LIBRARIES "Threads::Threads;${CUDA_CUDA_LIBRARY};${CUDA_LIBRARIES}" diff --git a/cmake/FindcuRAND.cmake b/cmake/FindcuRAND.cmake index df353077c..cd8a673da 100644 --- a/cmake/FindcuRAND.cmake +++ b/cmake/FindcuRAND.cmake @@ -59,7 +59,7 @@ find_package(CUDA 10.0 REQUIRED) get_filename_component(SYCL_BINARY_DIR ${CMAKE_CXX_COMPILER} DIRECTORY) -if (NOT (ONEMKL_SYCL_IMPLEMENTATION STREQUAL "hipsycl")) +if (NOT (ONEMATH_SYCL_IMPLEMENTATION STREQUAL "hipsycl")) # the OpenCL include file from cuda is opencl 1.1 and it is not compatible with DPC++ # the OpenCL include headers 1.2 onward is required. This is used to bypass NVIDIA OpenCL headers find_path(OPENCL_INCLUDE_DIR CL/cl.h OpenCL/cl.h @@ -76,7 +76,7 @@ find_package(Threads REQUIRED) include(FindPackageHandleStandardArgs) -if (ONEMKL_SYCL_IMPLEMENTATION STREQUAL "hipsycl") +if (ONEMATH_SYCL_IMPLEMENTATION STREQUAL "hipsycl") find_package_handle_standard_args(cuRAND REQUIRED_VARS CUDA_TOOLKIT_INCLUDE @@ -85,9 +85,9 @@ find_package_handle_standard_args(cuRAND CUDA_CUDA_LIBRARY ) - if(NOT TARGET ONEMKL::cuRAND::cuRAND) - add_library(ONEMKL::cuRAND::cuRAND SHARED IMPORTED) - set_target_properties(ONEMKL::cuRAND::cuRAND PROPERTIES + if(NOT TARGET ONEMATH::cuRAND::cuRAND) + add_library(ONEMATH::cuRAND::cuRAND SHARED IMPORTED) + set_target_properties(ONEMATH::cuRAND::cuRAND PROPERTIES IMPORTED_LOCATION ${CUDA_curand_LIBRARY} INTERFACE_INCLUDE_DIRECTORIES "${CUDA_TOOLKIT_INCLUDE}" INTERFACE_LINK_LIBRARIES "Threads::Threads;${CUDA_CUDA_LIBRARY};${CUDA_LIBRARIES}" @@ -103,9 +103,9 @@ find_package_handle_standard_args(cuRAND OPENCL_INCLUDE_DIR ) - if(NOT TARGET ONEMKL::cuRAND::cuRAND) - add_library(ONEMKL::cuRAND::cuRAND SHARED IMPORTED) - set_target_properties(ONEMKL::cuRAND::cuRAND PROPERTIES + if(NOT TARGET ONEMATH::cuRAND::cuRAND) + add_library(ONEMATH::cuRAND::cuRAND SHARED IMPORTED) + set_target_properties(ONEMATH::cuRAND::cuRAND PROPERTIES IMPORTED_LOCATION ${CUDA_curand_LIBRARY} INTERFACE_INCLUDE_DIRECTORIES "${OPENCL_INCLUDE_DIR};${CUDA_TOOLKIT_INCLUDE}" INTERFACE_LINK_LIBRARIES "Threads::Threads;${CUDA_CUDA_LIBRARY};${CUDA_LIBRARIES}" diff --git a/cmake/FindcuSOLVER.cmake b/cmake/FindcuSOLVER.cmake index 0f7f59564..8662679fd 100644 --- a/cmake/FindcuSOLVER.cmake +++ b/cmake/FindcuSOLVER.cmake @@ -40,9 +40,9 @@ find_package_handle_standard_args(cuSOLVER CUDA_CUDA_LIBRARY OPENCL_INCLUDE_DIR ) -if(NOT TARGET ONEMKL::cuSOLVER::cuSOLVER) - add_library(ONEMKL::cuSOLVER::cuSOLVER SHARED IMPORTED) - set_target_properties(ONEMKL::cuSOLVER::cuSOLVER PROPERTIES +if(NOT TARGET ONEMATH::cuSOLVER::cuSOLVER) + add_library(ONEMATH::cuSOLVER::cuSOLVER SHARED IMPORTED) + set_target_properties(ONEMATH::cuSOLVER::cuSOLVER PROPERTIES IMPORTED_LOCATION ${CUDA_cusolver_LIBRARY} INTERFACE_INCLUDE_DIRECTORIES "${OPENCL_INCLUDE_DIR};${CUDA_TOOLKIT_INCLUDE}" INTERFACE_LINK_LIBRARIES "Threads::Threads;${CUDA_CUDA_LIBRARY};${CUDA_LIBRARIES}" diff --git a/cmake/WarningsUtils.cmake b/cmake/WarningsUtils.cmake index 8de2858a6..43e19d8ef 100644 --- a/cmake/WarningsUtils.cmake +++ b/cmake/WarningsUtils.cmake @@ -21,13 +21,13 @@ include_guard(GLOBAL) add_library(onemath_warnings INTERFACE) -set(ONEMKL_WARNINGS "") +set(ONEMATH_WARNINGS "") include(CheckCXXCompilerFlag) macro(add_warning flag) check_cxx_compiler_flag(${flag} IS_SUPPORTED) if(${IS_SUPPORTED}) - list(APPEND ONEMKL_WARNINGS ${flag}) + list(APPEND ONEMATH_WARNINGS ${flag}) else() message(WARNING "Compiler does not support ${flag}") endif() @@ -39,10 +39,10 @@ add_warning("-Wshadow") add_warning("-Wconversion") add_warning("-Wpedantic") -message(VERBOSE "Domains with warnings enabled use: ${ONEMKL_WARNINGS}") +message(VERBOSE "Domains with warnings enabled use: ${ONEMATH_WARNINGS}") # The onemath_warnings target can be linked to any other target to enable warnings. -target_compile_options(onemath_warnings INTERFACE ${ONEMKL_WARNINGS}) +target_compile_options(onemath_warnings INTERFACE ${ONEMATH_WARNINGS}) # Add the library to install package install(TARGETS onemath_warnings EXPORT oneMKLTargets) diff --git a/docs/building_the_project_with_adaptivecpp.rst b/docs/building_the_project_with_adaptivecpp.rst index 98c763b90..5be18fb5b 100644 --- a/docs/building_the_project_with_adaptivecpp.rst +++ b/docs/building_the_project_with_adaptivecpp.rst @@ -35,7 +35,7 @@ On Linux (other OSes are not supported with the AdaptiveCpp compiler): # Inside mkdir build && cd build - cmake .. -DONEMKL_SYCL_IMPLEMENTATION=hipsycl \ # Indicate that AdaptiveCpp is being used. + cmake .. -DONEMATH_SYCL_IMPLEMENTATION=hipsycl \ # Indicate that AdaptiveCpp is being used. -DENABLE_MKLGPU_BACKEND=False \ # MKLGPU backend is not supported by AdaptiveCpp -DENABLE__BACKEND=True \ # Enable backend(s) (optional) -DENABLE__BACKEND=True \ # Multiple backends can be enabled at once. diff --git a/docs/create_new_backend.rst b/docs/create_new_backend.rst index 15b9ba3ae..98c9d8877 100644 --- a/docs/create_new_backend.rst +++ b/docs/create_new_backend.rst @@ -157,13 +157,13 @@ To integrate the new third-party library to a oneMKL header-based part, followin { domain::blas, { { device::x86cpu, { - #ifdef ONEMKL_ENABLE_MKLCPU_BACKEND + #ifdef ONEMATH_ENABLE_MKLCPU_BACKEND LIB_NAME("blas_mklcpu") #endif } }, + { device::newdevice, + { - + #ifdef ONEMKL_ENABLE_NEWLIB_BACKEND + + #ifdef ONEMATH_ENABLE_NEWLIB_BACKEND + LIB_NAME("blas_newlib") + #endif + } }, @@ -368,8 +368,8 @@ Here is the list of files that should be created/updated to integrate the new wr include(FindPackageHandleStandardArgs) find_package_handle_standard_args(NEWLIB REQUIRED_VARS NEWLIB_LIBRARY) # Set cmake target for the library - add_library(ONEMKL::NEWLIB::NEWLIB UNKNOWN IMPORTED) - set_target_properties(ONEMKL::NEWLIB::NEWLIB PROPERTIES + add_library(ONEMATH::NEWLIB::NEWLIB UNKNOWN IMPORTED) + set_target_properties(ONEMATH::NEWLIB::NEWLIB PROPERTIES IMPORTED_LOCATION ${NEWLIB_LIBRARY}) * Create the ``src//backends//CMakeList.txt`` cmake config file to specify how to build the backend layer for the new third-party library. @@ -397,9 +397,9 @@ Here is the list of files that should be created/updated to integrate the new wr .. code-block:: diff target_link_libraries(${LIB_OBJ} - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL - # Add third-party library to link with here - + PUBLIC ONEMKL::NEWLIB::NEWLIB + + PUBLIC ONEMATH::NEWLIB::NEWLIB ) Now you can build the backend library for ``newlib`` to make sure the third-party library integration was completed successfully (for more information, see `Build with cmake <../README.md#building-with-cmake>`_) @@ -427,8 +427,8 @@ Update the following files to enable the new third-party library for unit tests: .. code-block:: diff - #cmakedefine ONEMKL_ENABLE_MKLCPU_BACKEND - + #cmakedefine ONEMKL_ENABLE_NEWLIB_BACKEND + #cmakedefine ONEMATH_ENABLE_MKLCPU_BACKEND + + #cmakedefine ONEMATH_ENABLE_NEWLIB_BACKEND * ``tests/unit_tests/CMakeLists.txt``: add instructions about how to link tests with the new backend library @@ -439,22 +439,22 @@ Update the following files to enable the new third-party library for unit tests: if(ENABLE_MKLCPU_BACKEND) add_dependencies(test_main_ct onemath_blas_mklcpu) if(BUILD_SHARED_LIBS) - list(APPEND ONEMKL_LIBRARIES onemath_blas_mklcpu) + list(APPEND ONEMATH_LIBRARIES onemath_blas_mklcpu) else() - list(APPEND ONEMKL_LIBRARIES -foffload-static-lib=${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libonemath_blas_mklcpu.a) + list(APPEND ONEMATH_LIBRARIES -foffload-static-lib=${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libonemath_blas_mklcpu.a) find_package(MKL REQUIRED) - list(APPEND ONEMKL_LIBRARIES ${MKL_LINK_C}) + list(APPEND ONEMATH_LIBRARIES ${MKL_LINK_C}) endif() endif() + + if(ENABLE_NEWLIB_BACKEND) + add_dependencies(test_main_ct onemath_blas_newlib) + if(BUILD_SHARED_LIBS) - + list(APPEND ONEMKL_LIBRARIES onemath_blas_newlib) + + list(APPEND ONEMATH_LIBRARIES onemath_blas_newlib) + else() - + list(APPEND ONEMKL_LIBRARIES -foffload-static-lib=${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libonemath_blas_newlib.a) + + list(APPEND ONEMATH_LIBRARIES -foffload-static-lib=${CMAKE_LIBRARY_OUTPUT_DIRECTORY}/libonemath_blas_newlib.a) + find_package(NEWLIB REQUIRED) - + list(APPEND ONEMKL_LIBRARIES ONEMKL::NEWLIB::NEWLIB) + + list(APPEND ONEMATH_LIBRARIES ONEMATH::NEWLIB::NEWLIB) + endif() + endif() @@ -464,14 +464,14 @@ Update the following files to enable the new third-party library for unit tests: .. code-block:: diff - #ifdef ONEMKL_ENABLE_MKLGPU_BACKEND + #ifdef ONEMATH_ENABLE_MKLGPU_BACKEND #define TEST_RUN_INTELGPU(q, func, args) \ func args #else #define TEST_RUN_INTELGPU(q, func, args) #endif + - + #ifdef ONEMKL_ENABLE_NEWLIB_BACKEND + + #ifdef ONEMATH_ENABLE_NEWLIB_BACKEND + #define TEST_RUN_NEWDEVICE(q, func, args) \ + func args + #else @@ -495,7 +495,7 @@ Update the following files to enable the new third-party library for unit tests: } } + - + #ifdef ONEMKL_ENABLE_NEWLIB_BACKEND + + #ifdef ONEMATH_ENABLE_NEWLIB_BACKEND + devices.push_back(sycl::device(sycl::host_selector())); + #endif diff --git a/examples/blas/compile_time_dispatching/level3/CMakeLists.txt b/examples/blas/compile_time_dispatching/level3/CMakeLists.txt index 7aaebdf1a..eb33a7497 100644 --- a/examples/blas/compile_time_dispatching/level3/CMakeLists.txt +++ b/examples/blas/compile_time_dispatching/level3/CMakeLists.txt @@ -33,12 +33,12 @@ foreach(blas_ct_source ${BLAS_CT_SOURCES}) if(domain STREQUAL "blas" AND ENABLE_MKLCPU_BACKEND AND ENABLE_CUBLAS_BACKEND) add_dependencies(example_${domain}_${blas_ct_source} onemath_${domain}_mklcpu onemath_${domain}_cublas) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_mklcpu onemath_${domain}_cublas) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_mklcpu onemath_${domain}_cublas) endif() target_link_libraries(example_${domain}_${blas_ct_source} PUBLIC - ${ONEMKL_LIBRARIES_${domain}} - ONEMKL::SYCL::SYCL + ${ONEMATH_LIBRARIES_${domain}} + ONEMATH::SYCL::SYCL ) # Register example as ctest diff --git a/examples/blas/run_time_dispatching/level3/CMakeLists.txt b/examples/blas/run_time_dispatching/level3/CMakeLists.txt index 193e5db95..1b3f992c4 100644 --- a/examples/blas/run_time_dispatching/level3/CMakeLists.txt +++ b/examples/blas/run_time_dispatching/level3/CMakeLists.txt @@ -72,7 +72,7 @@ foreach(blas_rt_source ${BLAS_RT_SOURCES}) target_link_libraries(example_${domain}_${blas_rt_source} PUBLIC onemath - ONEMKL::SYCL::SYCL + ONEMATH::SYCL::SYCL ${CMAKE_DL_LIBS} ) diff --git a/examples/dft/compile_time_dispatching/CMakeLists.txt b/examples/dft/compile_time_dispatching/CMakeLists.txt index 5684b7cdb..8c7665cb2 100644 --- a/examples/dft/compile_time_dispatching/CMakeLists.txt +++ b/examples/dft/compile_time_dispatching/CMakeLists.txt @@ -35,11 +35,11 @@ foreach(dft_ct_source ${DFT_CT_SOURCES}) if(domain STREQUAL "dft" AND ENABLE_MKLCPU_BACKEND AND ENABLE_CUFFT_BACKEND) add_dependencies(${EXAMPLE_NAME} onemath_${domain}_mklcpu onemath_${domain}_cufft) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_mklcpu onemath_${domain}_cufft) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_mklcpu onemath_${domain}_cufft) endif() target_link_libraries(${EXAMPLE_NAME} PUBLIC - ${ONEMKL_LIBRARIES_${domain}} + ${ONEMATH_LIBRARIES_${domain}} onemath_warnings ) diff --git a/examples/dft/run_time_dispatching/CMakeLists.txt b/examples/dft/run_time_dispatching/CMakeLists.txt index c1f89338d..5d9f201cc 100644 --- a/examples/dft/run_time_dispatching/CMakeLists.txt +++ b/examples/dft/run_time_dispatching/CMakeLists.txt @@ -65,7 +65,7 @@ foreach(dft_rt_sources ${DFT_RT_SOURCES}) target_link_libraries(example_${domain}_${dft_rt_sources} PUBLIC onemath - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL PUBLIC ${CMAKE_DL_LIBS} PRIVATE onemath_warnings ) diff --git a/examples/lapack/compile_time_dispatching/CMakeLists.txt b/examples/lapack/compile_time_dispatching/CMakeLists.txt index 69acae8e0..f72ea3c7b 100644 --- a/examples/lapack/compile_time_dispatching/CMakeLists.txt +++ b/examples/lapack/compile_time_dispatching/CMakeLists.txt @@ -37,12 +37,12 @@ foreach(lapack_ct_source ${LAPACK_CT_SOURCES}) ) if(domain STREQUAL "lapack" AND ENABLE_MKLCPU_BACKEND AND ENABLE_CUSOLVER_BACKEND) add_dependencies(example_${domain}_${lapack_ct_source} onemath_${domain}_mklcpu onemath_${domain}_cusolver) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_mklcpu onemath_${domain}_cusolver) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_mklcpu onemath_${domain}_cusolver) target_link_libraries(example_${domain}_${lapack_ct_source} PUBLIC ${OPENCL_LIBRARY}) endif() target_link_libraries(example_${domain}_${lapack_ct_source} PUBLIC - ${ONEMKL_LIBRARIES_${domain}} - ONEMKL::SYCL::SYCL + ${ONEMATH_LIBRARIES_${domain}} + ONEMATH::SYCL::SYCL ) # Register example as ctest add_test(NAME ${domain}/EXAMPLE/CT/${lapack_ct_source} COMMAND example_${domain}_${lapack_ct_source}) diff --git a/examples/lapack/run_time_dispatching/CMakeLists.txt b/examples/lapack/run_time_dispatching/CMakeLists.txt index 5a2affce2..5026a6b49 100644 --- a/examples/lapack/run_time_dispatching/CMakeLists.txt +++ b/examples/lapack/run_time_dispatching/CMakeLists.txt @@ -57,7 +57,7 @@ foreach(lapack_rt_source ${LAPACK_RT_SOURCES}) target_link_libraries(example_${domain}_${lapack_rt_source} PUBLIC onemath - ONEMKL::SYCL::SYCL + ONEMATH::SYCL::SYCL ${CMAKE_DL_LIBS} ) diff --git a/examples/rng/compile_time_dispatching/CMakeLists.txt b/examples/rng/compile_time_dispatching/CMakeLists.txt index bcbc9d4ea..bfd37eead 100644 --- a/examples/rng/compile_time_dispatching/CMakeLists.txt +++ b/examples/rng/compile_time_dispatching/CMakeLists.txt @@ -35,14 +35,14 @@ foreach(rng_ct_source ${RNG_CT_SOURCES}) add_dependencies(example_${domain}_${rng_ct_source} onemath_${domain}_mklcpu onemath_${domain}_curand) - list(APPEND ONEMKL_LIBRARIES_${domain} + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_mklcpu onemath_${domain}_curand) endif() target_link_libraries(example_${domain}_${rng_ct_source} PUBLIC - ${ONEMKL_LIBRARIES_${domain}} - ONEMKL::SYCL::SYCL + ${ONEMATH_LIBRARIES_${domain}} + ONEMATH::SYCL::SYCL ) # Register example as ctest diff --git a/examples/rng/device/CMakeLists.txt b/examples/rng/device/CMakeLists.txt index 1b6ecf2dd..b7257d627 100644 --- a/examples/rng/device/CMakeLists.txt +++ b/examples/rng/device/CMakeLists.txt @@ -56,10 +56,10 @@ foreach(rng_device_source ${RNG_DEVICE_SOURCES}) endif() target_link_libraries(example_${domain}_${rng_device_source} PUBLIC - ONEMKL::SYCL::SYCL + ONEMATH::SYCL::SYCL ) - if(NOT ${ONEMKL_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") + if(NOT ${ONEMATH_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") target_link_options(example_${domain}_${rng_device_source} PUBLIC -fsycl -fsycl-device-code-split=per_kernel) endif() diff --git a/examples/rng/run_time_dispatching/CMakeLists.txt b/examples/rng/run_time_dispatching/CMakeLists.txt index 5a373efaa..8de795c02 100644 --- a/examples/rng/run_time_dispatching/CMakeLists.txt +++ b/examples/rng/run_time_dispatching/CMakeLists.txt @@ -58,7 +58,7 @@ foreach(rng_rt_source ${RNG_RT_SOURCES}) target_link_libraries(example_${domain}_${rng_rt_source} PUBLIC onemath - ONEMKL::SYCL::SYCL + ONEMATH::SYCL::SYCL ${CMAKE_DL_LIBS} ) diff --git a/examples/sparse_blas/compile_time_dispatching/CMakeLists.txt b/examples/sparse_blas/compile_time_dispatching/CMakeLists.txt index 42516da05..40e682870 100644 --- a/examples/sparse_blas/compile_time_dispatching/CMakeLists.txt +++ b/examples/sparse_blas/compile_time_dispatching/CMakeLists.txt @@ -36,7 +36,7 @@ foreach(backend ${SPARSE_BLAS_BACKENDS}) ) add_dependencies(${EXAMPLE_NAME} onemath_sparse_blas_${backend}) - target_link_libraries(${EXAMPLE_NAME} PRIVATE ONEMKL::SYCL::SYCL onemath_sparse_blas_${backend}) + target_link_libraries(${EXAMPLE_NAME} PRIVATE ONEMATH::SYCL::SYCL onemath_sparse_blas_${backend}) # Register example as ctest add_test(NAME sparse_blas/EXAMPLE/CT/sparse_blas_spmv_usm_${backend} COMMAND ${EXAMPLE_NAME}) diff --git a/examples/sparse_blas/run_time_dispatching/CMakeLists.txt b/examples/sparse_blas/run_time_dispatching/CMakeLists.txt index 1d64928e5..fea7f2eb6 100644 --- a/examples/sparse_blas/run_time_dispatching/CMakeLists.txt +++ b/examples/sparse_blas/run_time_dispatching/CMakeLists.txt @@ -52,7 +52,7 @@ foreach(sparse_blas_rt_sources ${SPARSE_BLAS_RT_SOURCES}) target_link_libraries(example_${sparse_blas_rt_sources} PUBLIC onemath - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL PUBLIC ${CMAKE_DL_LIBS} PRIVATE onemath_warnings ) diff --git a/include/oneapi/math/blas.hpp b/include/oneapi/math/blas.hpp index 0c0bd6bfa..66f31c4d4 100644 --- a/include/oneapi/math/blas.hpp +++ b/include/oneapi/math/blas.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_BLAS_HPP_ -#define _ONEMKL_BLAS_HPP_ +#ifndef _ONEMATH_BLAS_HPP_ +#define _ONEMATH_BLAS_HPP_ #if __has_include() #include @@ -34,22 +34,22 @@ #include "oneapi/math/detail/get_device_id.hpp" #include "oneapi/math/blas/detail/blas_loader.hpp" -#ifdef ONEMKL_ENABLE_CUBLAS_BACKEND +#ifdef ONEMATH_ENABLE_CUBLAS_BACKEND #include "oneapi/math/blas/detail/cublas/blas_ct.hpp" #endif -#ifdef ONEMKL_ENABLE_ROCBLAS_BACKEND +#ifdef ONEMATH_ENABLE_ROCBLAS_BACKEND #include "oneapi/math/blas/detail/rocblas/blas_ct.hpp" #endif -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND #include "oneapi/math/blas/detail/mklcpu/blas_ct.hpp" #endif -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND #include "oneapi/math/blas/detail/mklgpu/blas_ct.hpp" #endif -#ifdef ONEMKL_ENABLE_NETLIB_BACKEND +#ifdef ONEMATH_ENABLE_NETLIB_BACKEND #include "oneapi/math/blas/detail/netlib/blas_ct.hpp" #endif -#ifdef ONEMKL_ENABLE_PORTBLAS_BACKEND +#ifdef ONEMATH_ENABLE_PORTBLAS_BACKEND #include "oneapi/math/blas/detail/portblas/blas_ct.hpp" #endif @@ -70,4 +70,4 @@ namespace row_major { } //namespace mkl } //namespace oneapi -#endif //_ONEMKL_BLAS_LOADER_HPP_ +#endif //_ONEMATH_BLAS_LOADER_HPP_ diff --git a/include/oneapi/math/blas/detail/blas_loader.hpp b/include/oneapi/math/blas/detail/blas_loader.hpp index 0c1db264a..078334154 100644 --- a/include/oneapi/math/blas/detail/blas_loader.hpp +++ b/include/oneapi/math/blas/detail/blas_loader.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_BLAS_LOADER_HPP_ -#define _ONEMKL_BLAS_LOADER_HPP_ +#ifndef _ONEMATH_BLAS_LOADER_HPP_ +#define _ONEMATH_BLAS_LOADER_HPP_ #include #include @@ -54,4 +54,4 @@ namespace detail { } //namespace mkl } //namespace oneapi -#endif //_ONEMKL_BLAS_LOADER_HPP_ +#endif //_ONEMATH_BLAS_LOADER_HPP_ diff --git a/include/oneapi/math/blas/detail/blas_loader.hxx b/include/oneapi/math/blas/detail/blas_loader.hxx index 98d93b2ad..bebe32c02 100644 --- a/include/oneapi/math/blas/detail/blas_loader.hxx +++ b/include/oneapi/math/blas/detail/blas_loader.hxx @@ -19,88 +19,88 @@ // Buffer APIs -ONEMKL_EXPORT void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); -ONEMKL_EXPORT void spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); -ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -108,7 +108,7 @@ ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, tr std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -116,7 +116,7 @@ ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, tr std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, @@ -124,21 +124,21 @@ ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, tr std::int64_t stride_b, sycl::half beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, @@ -146,43 +146,43 @@ ONEMKL_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, tr sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -190,158 +190,158 @@ ONEMKL_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, up sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, float s); -ONEMKL_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, double s); -ONEMKL_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s); -ONEMKL_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s); -ONEMKL_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, @@ -349,7 +349,7 @@ ONEMKL_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, tr sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -357,7 +357,7 @@ ONEMKL_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, tr sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -365,21 +365,21 @@ ONEMKL_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, tr sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMKL_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMKL_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -387,7 +387,7 @@ ONEMKL_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, si sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMKL_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -395,44 +395,44 @@ ONEMKL_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, si sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMKL_EXPORT void her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); -ONEMKL_EXPORT void hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); -ONEMKL_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMKL_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMKL_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMKL_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, @@ -440,606 +440,606 @@ ONEMKL_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, tra sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMKL_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void rotmg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotmg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m); -ONEMKL_EXPORT void rotmg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotmg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m); -ONEMKL_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMKL_EXPORT void dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMKL_EXPORT void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); -ONEMKL_EXPORT void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); -ONEMKL_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMKL_EXPORT void dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMKL_EXPORT void syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); -ONEMKL_EXPORT void spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); -ONEMKL_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m); -ONEMKL_EXPORT void rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m); -ONEMKL_EXPORT void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMKL_EXPORT void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMKL_EXPORT void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMKL_EXPORT void sdsdot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void sdsdot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float sb, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMKL_EXPORT void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s); -ONEMKL_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s); -ONEMKL_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s); -ONEMKL_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s); -ONEMKL_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMKL_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMKL_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMKL_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMKL_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -1048,73 +1048,73 @@ ONEMKL_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); -ONEMKL_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); -ONEMKL_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); -ONEMKL_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); -ONEMKL_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); -ONEMKL_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); -ONEMKL_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); -ONEMKL_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); -ONEMKL_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, @@ -1123,92 +1123,92 @@ ONEMKL_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, trans // USM APIs -ONEMKL_EXPORT sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, const float **b, @@ -1216,7 +1216,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &qu std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, const double **b, @@ -1224,7 +1224,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &qu std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, @@ -1233,14 +1233,14 @@ ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &qu std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch( +ONEMATH_EXPORT sycl::event gemm_batch( oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, const sycl::half **a, std::int64_t *lda, @@ -1248,14 +1248,14 @@ ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &qu sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, @@ -1263,7 +1263,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &qu float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, @@ -1271,7 +1271,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &qu std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, @@ -1279,7 +1279,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &qu float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, @@ -1287,27 +1287,27 @@ ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &qu double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch( +ONEMATH_EXPORT sycl::event gemm_batch( oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch( +ONEMATH_EXPORT sycl::event gemm_batch( oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch( +ONEMATH_EXPORT sycl::event gemm_batch( oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, @@ -1315,7 +1315,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &qu float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, @@ -1323,7 +1323,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &qu float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, @@ -1332,24 +1332,24 @@ ONEMKL_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &qu std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1357,21 +1357,21 @@ ONEMKL_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, @@ -1379,7 +1379,7 @@ ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &qu std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, @@ -1388,21 +1388,21 @@ ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &qu std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1410,7 +1410,7 @@ ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &qu std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1419,27 +1419,27 @@ ONEMKL_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &qu std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, @@ -1447,84 +1447,84 @@ ONEMKL_EXPORT sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, double c, double s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float c, float s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double c, double s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, float *alpha, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, double *alpha, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, std::complex *alpha, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, std::complex *alpha, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, @@ -1532,54 +1532,54 @@ ONEMKL_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &qu std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1587,7 +1587,7 @@ ONEMKL_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1596,24 +1596,24 @@ ONEMKL_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, @@ -1621,7 +1621,7 @@ ONEMKL_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, @@ -1629,7 +1629,7 @@ ONEMKL_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &qu std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, @@ -1637,33 +1637,33 @@ ONEMKL_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &qu std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch( +ONEMATH_EXPORT sycl::event gemv_batch( oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch( +ONEMATH_EXPORT sycl::event gemv_batch( oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, @@ -1671,28 +1671,28 @@ ONEMKL_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &qu std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch( +ONEMATH_EXPORT sycl::event gemv_batch( oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, @@ -1700,7 +1700,7 @@ ONEMKL_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &qu std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, @@ -1708,26 +1708,26 @@ ONEMKL_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &qu std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, @@ -1735,135 +1735,135 @@ ONEMKL_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &qu std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotmg(oneapi::mkl::device libkey, sycl::queue &queue, float *d1, +ONEMATH_EXPORT sycl::event rotmg(oneapi::mkl::device libkey, sycl::queue &queue, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotmg(oneapi::mkl::device libkey, sycl::queue &queue, double *d1, +ONEMATH_EXPORT sycl::event rotmg(oneapi::mkl::device libkey, sycl::queue &queue, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1871,7 +1871,7 @@ ONEMKL_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1880,19 +1880,19 @@ ONEMKL_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1900,7 +1900,7 @@ ONEMKL_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1908,27 +1908,27 @@ ONEMKL_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, const std::int8_t *a, int64_t lda, std::int8_t ao, @@ -1936,12 +1936,12 @@ ONEMKL_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &que float beta, std::int32_t *c, int64_t ldc, const std::int32_t *co, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_bias( +ONEMATH_EXPORT sycl::event gemm_bias( oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, const std::int8_t *a, int64_t lda, std::int8_t ao, const std::int8_t *b, int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, int64_t ldc, const std::int32_t *co, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, const std::uint8_t *a, int64_t lda, std::uint8_t ao, @@ -1949,7 +1949,7 @@ ONEMKL_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &que float beta, std::int32_t *c, int64_t ldc, const std::int32_t *co, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, const std::uint8_t *a, int64_t lda, std::uint8_t ao, @@ -1958,97 +1958,97 @@ ONEMKL_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &que const std::int32_t *co, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, float *alpha, const float **a, int64_t *lda, float **b, int64_t *ldb, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, double *alpha, const double **a, int64_t *lda, double **b, int64_t *ldb, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, std::complex *alpha, const std::complex **a, int64_t *lda, std::complex **b, int64_t *ldb, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, std::complex *alpha, @@ -2057,18 +2057,18 @@ ONEMKL_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &qu int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2076,7 +2076,7 @@ ONEMKL_EXPORT sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2085,32 +2085,32 @@ ONEMKL_EXPORT sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2118,7 +2118,7 @@ ONEMKL_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2127,40 +2127,40 @@ ONEMKL_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2168,7 +2168,7 @@ ONEMKL_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2177,171 +2177,171 @@ ONEMKL_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -2349,115 +2349,115 @@ ONEMKL_EXPORT sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float *param, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double *param, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sdsdot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sdsdot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float sb, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2465,41 +2465,41 @@ ONEMKL_EXPORT sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, float *a, +ONEMATH_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, float *a, float *b, float *c, float *s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, double *a, +ONEMATH_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, double *a, double *b, double *c, double *s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *a, std::complex *b, float *c, std::complex *s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *a, std::complex *b, double *c, std::complex *s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, @@ -2507,31 +2507,31 @@ ONEMKL_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, @@ -2539,7 +2539,7 @@ ONEMKL_EXPORT sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, @@ -2547,14 +2547,14 @@ ONEMKL_EXPORT sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd_batch( +ONEMATH_EXPORT sycl::event omatadd_batch( oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd_batch( +ONEMATH_EXPORT sycl::event omatadd_batch( oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -2562,83 +2562,83 @@ ONEMKL_EXPORT sycl::event omatadd_batch( std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, @@ -2646,26 +2646,26 @@ ONEMKL_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, @@ -2673,25 +2673,25 @@ ONEMKL_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, float* alpha, float** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, double* alpha, double** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, std::complex** ab, std::int64_t* lda, std::int64_t* ldb, diff --git a/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp b/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp index fa7360550..234782ca4 100644 --- a/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp +++ b/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp @@ -16,8 +16,8 @@ * limitations under the License. * **************************************************************************/ -#ifndef _ONEMKL_BLAS_CUBLAS_HPP_ -#define _ONEMKL_BLAS_CUBLAS_HPP_ +#ifndef _ONEMATH_BLAS_CUBLAS_HPP_ +#define _ONEMATH_BLAS_CUBLAS_HPP_ #if __has_include() #include #else @@ -52,4 +52,4 @@ namespace row_major { } //namespace mkl } //namespace oneapi -#endif //_ONEMKL_BLAS_CUBLAS_HPP_ +#endif //_ONEMATH_BLAS_CUBLAS_HPP_ diff --git a/include/oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp b/include/oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp index e0fcf3068..1c6eefe36 100644 --- a/include/oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp +++ b/include/oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_BLAS_MKLCPU_HPP_ -#define _ONEMKL_BLAS_MKLCPU_HPP_ +#ifndef _ONEMATH_BLAS_MKLCPU_HPP_ +#define _ONEMATH_BLAS_MKLCPU_HPP_ #if __has_include() #include #else @@ -53,4 +53,4 @@ namespace row_major { } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_BLAS_MKLCPU_HPP_ +#endif //_ONEMATH_BLAS_MKLCPU_HPP_ diff --git a/include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp b/include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp index 2263f2840..6624a79a3 100644 --- a/include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp +++ b/include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_BLAS_MKLGPU_HPP_ -#define _ONEMKL_BLAS_MKLGPU_HPP_ +#ifndef _ONEMATH_BLAS_MKLGPU_HPP_ +#define _ONEMATH_BLAS_MKLGPU_HPP_ #if __has_include() #include @@ -51,4 +51,4 @@ namespace row_major { } //namespace mkl } //namespace oneapi -#endif //_ONEMKL_BLAS_MKLGPU_HPP_ +#endif //_ONEMATH_BLAS_MKLGPU_HPP_ diff --git a/include/oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp b/include/oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp index 7e6e8e6a4..404355d33 100644 --- a/include/oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp +++ b/include/oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_BLAS_NETLIB_HPP_ -#define _ONEMKL_BLAS_NETLIB_HPP_ +#ifndef _ONEMATH_BLAS_NETLIB_HPP_ +#define _ONEMATH_BLAS_NETLIB_HPP_ #if __has_include() #include @@ -59,4 +59,4 @@ namespace row_major { } //namespace mkl } //namespace oneapi -#endif //_ONEMKL_BLAS_NETLIB_HPP_ +#endif //_ONEMATH_BLAS_NETLIB_HPP_ diff --git a/include/oneapi/math/blas/detail/onemath_blas_backends.hxx b/include/oneapi/math/blas/detail/onemath_blas_backends.hxx index fbb64a6a0..28ac984c6 100644 --- a/include/oneapi/math/blas/detail/onemath_blas_backends.hxx +++ b/include/oneapi/math/blas/detail/onemath_blas_backends.hxx @@ -19,19 +19,19 @@ // Buffer APIs -ONEMKL_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -39,7 +39,7 @@ ONEMKL_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -47,99 +47,99 @@ ONEMKL_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void hemm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void hemm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void hemm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void hemm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -147,7 +147,7 @@ ONEMKL_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -155,147 +155,147 @@ ONEMKL_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -303,7 +303,7 @@ ONEMKL_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -311,626 +311,626 @@ ONEMKL_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMKL_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMKL_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMKL_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMKL_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, +ONEMATH_EXPORT void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, +ONEMATH_EXPORT void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMKL_EXPORT void hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); -ONEMKL_EXPORT void hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); -ONEMKL_EXPORT void hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); -ONEMKL_EXPORT void hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); -ONEMKL_EXPORT void sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMKL_EXPORT void spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); -ONEMKL_EXPORT void spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); -ONEMKL_EXPORT void spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); -ONEMKL_EXPORT void spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); -ONEMKL_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void dotc(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotc(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMKL_EXPORT void dotc(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotc(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMKL_EXPORT void dotu(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotu(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMKL_EXPORT void dotu(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotu(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMKL_EXPORT void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamax(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamax(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamin(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void iamin(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void asum(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void asum(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void asum(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void asum(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void axpy(sycl::queue &queue, std::int64_t n, float alpha, +ONEMATH_EXPORT void axpy(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void axpy(sycl::queue &queue, std::int64_t n, double alpha, +ONEMATH_EXPORT void axpy(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, +ONEMATH_EXPORT void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, +ONEMATH_EXPORT void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, +ONEMATH_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, +ONEMATH_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, +ONEMATH_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, +ONEMATH_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void axpby(sycl::queue &queue, std::int64_t n, float alpha, +ONEMATH_EXPORT void axpby(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void axpby(sycl::queue &queue, std::int64_t n, double alpha, +ONEMATH_EXPORT void axpby(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, +ONEMATH_EXPORT void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, +ONEMATH_EXPORT void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void copy(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void copy(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMKL_EXPORT void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMKL_EXPORT void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMKL_EXPORT void sdsdot(sycl::queue &queue, std::int64_t n, float sb, +ONEMATH_EXPORT void sdsdot(sycl::queue &queue, std::int64_t n, float sb, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMKL_EXPORT void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMKL_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMKL_EXPORT void rot(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rot(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, float s); -ONEMKL_EXPORT void rot(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rot(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, double s); -ONEMKL_EXPORT void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s); -ONEMKL_EXPORT void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s); -ONEMKL_EXPORT void rotg(sycl::queue &queue, sycl::buffer &a, +ONEMATH_EXPORT void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s); -ONEMKL_EXPORT void rotg(sycl::queue &queue, sycl::buffer &a, +ONEMATH_EXPORT void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s); -ONEMKL_EXPORT void rotg(sycl::queue &queue, sycl::buffer, 1> &a, +ONEMATH_EXPORT void rotg(sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s); -ONEMKL_EXPORT void rotg(sycl::queue &queue, sycl::buffer, 1> &a, +ONEMATH_EXPORT void rotg(sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s); -ONEMKL_EXPORT void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m); -ONEMKL_EXPORT void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m); -ONEMKL_EXPORT void rotmg(sycl::queue &queue, sycl::buffer &d1, +ONEMATH_EXPORT void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m); -ONEMKL_EXPORT void rotmg(sycl::queue &queue, sycl::buffer &d1, +ONEMATH_EXPORT void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m); -ONEMKL_EXPORT void scal(sycl::queue &queue, std::int64_t n, float alpha, +ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void scal(sycl::queue &queue, std::int64_t n, double alpha, +ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx); -ONEMKL_EXPORT void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, +ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, +ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void scal(sycl::queue &queue, std::int64_t n, float alpha, +ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void scal(sycl::queue &queue, std::int64_t n, double alpha, +ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMKL_EXPORT void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, +ONEMATH_EXPORT void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMKL_EXPORT void swap(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void swap(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, @@ -938,7 +938,7 @@ ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, @@ -946,7 +946,7 @@ ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, std::int64_t stride_b, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -955,7 +955,7 @@ ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -964,7 +964,7 @@ ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, @@ -973,7 +973,7 @@ ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, @@ -981,7 +981,7 @@ ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, @@ -989,7 +989,7 @@ ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, @@ -997,21 +997,21 @@ ONEMKL_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, @@ -1019,7 +1019,7 @@ ONEMKL_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, @@ -1027,20 +1027,20 @@ ONEMKL_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -1048,7 +1048,7 @@ ONEMKL_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -1056,7 +1056,7 @@ ONEMKL_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, @@ -1064,7 +1064,7 @@ ONEMKL_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMKL_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, @@ -1072,7 +1072,7 @@ ONEMKL_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMKL_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, @@ -1080,7 +1080,7 @@ ONEMKL_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMKL_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, @@ -1088,58 +1088,58 @@ ONEMKL_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMKL_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMKL_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMKL_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMKL_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMKL_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMKL_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, @@ -1147,7 +1147,7 @@ ONEMKL_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose tran sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -1155,7 +1155,7 @@ ONEMKL_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose tran std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -1164,84 +1164,84 @@ ONEMKL_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose tran std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMKL_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMKL_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMKL_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); -ONEMKL_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); -ONEMKL_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); -ONEMKL_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); -ONEMKL_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); -ONEMKL_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); -ONEMKL_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); -ONEMKL_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); -ONEMKL_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); -ONEMKL_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMKL_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, @@ -1250,21 +1250,21 @@ ONEMKL_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, // USM APIs -ONEMKL_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1273,7 +1273,7 @@ ONEMKL_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1282,28 +1282,28 @@ ONEMKL_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, @@ -1312,7 +1312,7 @@ ONEMKL_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose t const std::int32_t *co, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, @@ -1321,7 +1321,7 @@ ONEMKL_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose t const std::int32_t *co, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, @@ -1330,7 +1330,7 @@ ONEMKL_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose t const std::int32_t *co, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, @@ -1339,19 +1339,19 @@ ONEMKL_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose t const std::int32_t *co, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, @@ -1359,7 +1359,7 @@ ONEMKL_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, @@ -1367,7 +1367,7 @@ ONEMKL_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, @@ -1375,7 +1375,7 @@ ONEMKL_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, @@ -1383,33 +1383,33 @@ ONEMKL_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::mkl::side left_right, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, float *beta, float **c, @@ -1417,7 +1417,7 @@ ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *uppe std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, double *beta, double **c, @@ -1425,7 +1425,7 @@ ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *uppe std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, @@ -1434,7 +1434,7 @@ ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *uppe std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, @@ -1443,7 +1443,7 @@ ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *uppe std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, @@ -1451,7 +1451,7 @@ ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, @@ -1459,7 +1459,7 @@ ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1468,7 +1468,7 @@ ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1477,32 +1477,32 @@ ONEMKL_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, @@ -1510,7 +1510,7 @@ ONEMKL_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lowe std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, @@ -1518,7 +1518,7 @@ ONEMKL_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lowe std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, @@ -1526,7 +1526,7 @@ ONEMKL_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lowe std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, @@ -1534,63 +1534,63 @@ ONEMKL_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lowe std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, @@ -1599,7 +1599,7 @@ ONEMKL_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_ std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, @@ -1608,21 +1608,21 @@ ONEMKL_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_ std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch( +ONEMATH_EXPORT sycl::event trsm_batch( sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch( +ONEMATH_EXPORT sycl::event trsm_batch( sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, @@ -1631,7 +1631,7 @@ ONEMKL_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, @@ -1640,33 +1640,33 @@ ONEMKL_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch( +ONEMATH_EXPORT sycl::event trsm_batch( sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsm_batch( +ONEMATH_EXPORT sycl::event trsm_batch( sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1674,7 +1674,7 @@ ONEMKL_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1682,7 +1682,7 @@ ONEMKL_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, @@ -1690,7 +1690,7 @@ ONEMKL_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, @@ -1698,21 +1698,21 @@ ONEMKL_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch( +ONEMATH_EXPORT sycl::event gemv_batch( sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch( +ONEMATH_EXPORT sycl::event gemv_batch( sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, float **y, @@ -1720,7 +1720,7 @@ ONEMKL_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, double **y, @@ -1728,7 +1728,7 @@ ONEMKL_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, @@ -1737,14 +1737,14 @@ ONEMKL_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemv_batch( +ONEMATH_EXPORT sycl::event gemv_batch( sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float *c, @@ -1752,7 +1752,7 @@ ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_ std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double *c, @@ -1760,7 +1760,7 @@ ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_ std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, @@ -1769,7 +1769,7 @@ ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_ std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, @@ -1778,21 +1778,21 @@ ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_ std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, @@ -1800,7 +1800,7 @@ ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, @@ -1808,19 +1808,19 @@ ONEMKL_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, @@ -1828,7 +1828,7 @@ ONEMKL_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, @@ -1836,41 +1836,41 @@ ONEMKL_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1878,7 +1878,7 @@ ONEMKL_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1886,7 +1886,7 @@ ONEMKL_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1894,7 +1894,7 @@ ONEMKL_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1902,642 +1902,642 @@ ONEMKL_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dotc(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event dotc(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dotc(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event dotc(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dotu(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event dotu(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dotu(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event dotu(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, const float *x, +ONEMATH_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, const double *x, +ONEMATH_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, const float *x, +ONEMATH_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, const double *x, +ONEMATH_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, const float *x, +ONEMATH_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, const double *x, +ONEMATH_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, float alpha, +ONEMATH_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, double alpha, +ONEMATH_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, float *alpha, +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, float *alpha, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, double *alpha, +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, double *alpha, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, std::complex *alpha, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, std::complex *alpha, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, float alpha, +ONEMATH_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, double alpha, +ONEMATH_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, const float *x, +ONEMATH_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, const double *x, +ONEMATH_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const float **x, +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const double **x, +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const float *x, +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const double *x, +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, +ONEMATH_EXPORT sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dot(sycl::queue &queue, std::int64_t n, const double *x, +ONEMATH_EXPORT sycl::event dot(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sdsdot(sycl::queue &queue, std::int64_t n, float sb, +ONEMATH_EXPORT sycl::event sdsdot(sycl::queue &queue, std::int64_t n, float sb, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, +ONEMATH_EXPORT sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, const float *x, +ONEMATH_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, const double *x, +ONEMATH_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, +ONEMATH_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, +ONEMATH_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, double c, double s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, float *x, +ONEMATH_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float c, float s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, double *x, +ONEMATH_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double c, double s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotg(sycl::queue &queue, float *a, float *b, float *c, float *s, +ONEMATH_EXPORT sycl::event rotg(sycl::queue &queue, float *a, float *b, float *c, float *s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotg(sycl::queue &queue, double *a, double *b, double *c, +ONEMATH_EXPORT sycl::event rotg(sycl::queue &queue, double *a, double *b, double *c, double *s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotg(sycl::queue &queue, std::complex *a, +ONEMATH_EXPORT sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, float *c, std::complex *s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotg(sycl::queue &queue, std::complex *a, +ONEMATH_EXPORT sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, double *c, std::complex *s, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotm(sycl::queue &queue, std::int64_t n, float *x, +ONEMATH_EXPORT sycl::event rotm(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float *param, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotm(sycl::queue &queue, std::int64_t n, double *x, +ONEMATH_EXPORT sycl::event rotm(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double *param, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, +ONEMATH_EXPORT sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, +ONEMATH_EXPORT sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, float *x, +ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, double *x, +ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, +ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, +ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, float *x, +ONEMATH_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, double *x, +ONEMATH_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, +ONEMATH_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, +ONEMATH_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, const float **b, @@ -2546,7 +2546,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, const double **b, @@ -2555,7 +2555,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, @@ -2565,7 +2565,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch( +ONEMATH_EXPORT sycl::event gemm_batch( sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, @@ -2573,7 +2573,7 @@ ONEMKL_EXPORT sycl::event gemm_batch( std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, const sycl::half **a, std::int64_t *lda, @@ -2582,7 +2582,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, @@ -2590,7 +2590,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, @@ -2599,7 +2599,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, @@ -2608,7 +2608,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, @@ -2617,7 +2617,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, @@ -2626,7 +2626,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch( +ONEMATH_EXPORT sycl::event gemm_batch( sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, @@ -2634,7 +2634,7 @@ ONEMKL_EXPORT sycl::event gemm_batch( std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch( +ONEMATH_EXPORT sycl::event gemm_batch( sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, @@ -2642,14 +2642,14 @@ ONEMKL_EXPORT sycl::event gemm_batch( std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch( +ONEMATH_EXPORT sycl::event gemm_batch( sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, @@ -2658,7 +2658,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, @@ -2667,7 +2667,7 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, @@ -2676,21 +2676,21 @@ ONEMKL_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2699,7 +2699,7 @@ ONEMKL_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lowe std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2708,21 +2708,21 @@ ONEMKL_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lowe std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, @@ -2730,7 +2730,7 @@ ONEMKL_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transp std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, @@ -2738,33 +2738,33 @@ ONEMKL_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transp std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, @@ -2773,7 +2773,7 @@ ONEMKL_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpo std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, @@ -2782,7 +2782,7 @@ ONEMKL_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpo std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd_batch( +ONEMATH_EXPORT sycl::event omatadd_batch( sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, @@ -2790,7 +2790,7 @@ ONEMKL_EXPORT sycl::event omatadd_batch( std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd_batch( +ONEMATH_EXPORT sycl::event omatadd_batch( sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -2798,86 +2798,86 @@ ONEMKL_EXPORT sycl::event omatadd_batch( std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, @@ -2885,7 +2885,7 @@ ONEMKL_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose tra std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, +ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, @@ -2893,52 +2893,52 @@ ONEMKL_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose tra std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, float* alpha, float** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, double* alpha, double** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMKL_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, diff --git a/include/oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp b/include/oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp index a8c29d02f..2c7cf0905 100644 --- a/include/oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp +++ b/include/oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_BLAS_PORTBLAS_HPP_ -#define _ONEMKL_BLAS_PORTBLAS_HPP_ +#ifndef _ONEMATH_BLAS_PORTBLAS_HPP_ +#define _ONEMATH_BLAS_PORTBLAS_HPP_ #if __has_include() #include @@ -58,4 +58,4 @@ namespace row_major { } // namespace mkl } // namespace oneapi -#endif // _ONEMKL_BLAS_PORTBLAS_HPP_ +#endif // _ONEMATH_BLAS_PORTBLAS_HPP_ diff --git a/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp b/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp index 0416d1562..e2e8c8f97 100644 --- a/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp +++ b/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp @@ -18,8 +18,8 @@ * limitations under the License. * **************************************************************************/ -#ifndef _ONEMKL_BLAS_ROCBLAS_HPP_ -#define _ONEMKL_BLAS_ROCBLAS_HPP_ +#ifndef _ONEMATH_BLAS_ROCBLAS_HPP_ +#define _ONEMATH_BLAS_ROCBLAS_HPP_ #if __has_include() #include #else @@ -55,4 +55,4 @@ namespace row_major { } //namespace mkl } //namespace oneapi -#endif //_ONEMKL_BLAS_ROCBLAS_HPP_ +#endif //_ONEMATH_BLAS_ROCBLAS_HPP_ diff --git a/include/oneapi/math/detail/backend_selector.hpp b/include/oneapi/math/detail/backend_selector.hpp index 78797551e..688e8934c 100644 --- a/include/oneapi/math/detail/backend_selector.hpp +++ b/include/oneapi/math/detail/backend_selector.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_BACKEND_SELECTOR_HPP_ -#define _ONEMKL_BACKEND_SELECTOR_HPP_ +#ifndef _ONEMATH_BACKEND_SELECTOR_HPP_ +#define _ONEMATH_BACKEND_SELECTOR_HPP_ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/detail/backends.hpp" @@ -44,4 +44,4 @@ class backend_selector { } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_BACKEND_SELECTOR_HPP_ +#endif //_ONEMATH_BACKEND_SELECTOR_HPP_ diff --git a/include/oneapi/math/detail/backend_selector_predicates.hpp b/include/oneapi/math/detail/backend_selector_predicates.hpp index e3966cf29..ec995121e 100644 --- a/include/oneapi/math/detail/backend_selector_predicates.hpp +++ b/include/oneapi/math/detail/backend_selector_predicates.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_BACKEND_SELECTOR_PREDICATES_HPP_ -#define _ONEMKL_BACKEND_SELECTOR_PREDICATES_HPP_ +#ifndef _ONEMATH_BACKEND_SELECTOR_PREDICATES_HPP_ +#define _ONEMATH_BACKEND_SELECTOR_PREDICATES_HPP_ #include #if __has_include() @@ -39,7 +39,7 @@ inline void backend_selector_precondition(sycl::queue&) {} template <> inline void backend_selector_precondition(sycl::queue& queue) { -#ifndef ONEMKL_DISABLE_PREDICATES +#ifndef ONEMATH_DISABLE_PREDICATES #ifdef __HIPSYCL__ if (!(queue.is_host() || queue.get_device().is_cpu())) { #else @@ -54,7 +54,7 @@ inline void backend_selector_precondition(sycl::queue& queue) { template <> inline void backend_selector_precondition(sycl::queue& queue) { -#ifndef ONEMKL_DISABLE_PREDICATES +#ifndef ONEMATH_DISABLE_PREDICATES #ifdef __HIPSYCL__ if (!(queue.is_host() || queue.get_device().is_cpu())) { #else @@ -69,7 +69,7 @@ inline void backend_selector_precondition(sycl::queue& queue) { template <> inline void backend_selector_precondition(sycl::queue& queue) { -#ifndef ONEMKL_DISABLE_PREDICATES +#ifndef ONEMATH_DISABLE_PREDICATES unsigned int vendor_id = static_cast(queue.get_device().get_info()); if (!(queue.get_device().is_gpu() && vendor_id == INTEL_ID)) { @@ -82,7 +82,7 @@ inline void backend_selector_precondition(sycl::queue& queue) { template <> inline void backend_selector_precondition(sycl::queue& queue) { -#ifndef ONEMKL_DISABLE_PREDICATES +#ifndef ONEMATH_DISABLE_PREDICATES unsigned int vendor_id = static_cast(queue.get_device().get_info()); if (!(queue.get_device().is_gpu() && vendor_id == NVIDIA_ID)) { @@ -95,7 +95,7 @@ inline void backend_selector_precondition(sycl::queue& queue) { template <> inline void backend_selector_precondition(sycl::queue& queue) { -#ifndef ONEMKL_DISABLE_PREDICATES +#ifndef ONEMATH_DISABLE_PREDICATES unsigned int vendor_id = static_cast(queue.get_device().get_info()); if (!(queue.get_device().is_gpu() && vendor_id == NVIDIA_ID)) { @@ -108,7 +108,7 @@ inline void backend_selector_precondition(sycl::queue& queue) template <> inline void backend_selector_precondition(sycl::queue& queue) { -#ifndef ONEMKL_DISABLE_PREDICATES +#ifndef ONEMATH_DISABLE_PREDICATES unsigned int vendor_id = static_cast(queue.get_device().get_info()); if (!(queue.get_device().is_gpu() && vendor_id == AMD_ID)) { @@ -121,7 +121,7 @@ inline void backend_selector_precondition(sycl::queue& queue) template <> inline void backend_selector_precondition(sycl::queue& queue) { -#ifndef ONEMKL_DISABLE_PREDICATES +#ifndef ONEMATH_DISABLE_PREDICATES unsigned int vendor_id = static_cast(queue.get_device().get_info()); if (!(queue.get_device().is_gpu() && vendor_id == AMD_ID)) { @@ -134,7 +134,7 @@ inline void backend_selector_precondition(sycl::queue& queue) template <> inline void backend_selector_precondition(sycl::queue& queue) { -#ifndef ONEMKL_DISABLE_PREDICATES +#ifndef ONEMATH_DISABLE_PREDICATES unsigned int vendor_id = static_cast(queue.get_device().get_info()); if (!(queue.get_device().is_gpu() && vendor_id == AMD_ID)) { @@ -147,4 +147,4 @@ inline void backend_selector_precondition(sycl::queue& queue } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_BACKEND_SELECTOR_PREDICATES_HPP_ +#endif //_ONEMATH_BACKEND_SELECTOR_PREDICATES_HPP_ diff --git a/include/oneapi/math/detail/backends.hpp b/include/oneapi/math/detail/backends.hpp index 32b7c2614..0925149d1 100644 --- a/include/oneapi/math/detail/backends.hpp +++ b/include/oneapi/math/detail/backends.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_BACKENDS_HPP_ -#define _ONEMKL_BACKENDS_HPP_ +#ifndef _ONEMATH_BACKENDS_HPP_ +#define _ONEMATH_BACKENDS_HPP_ #include #include @@ -58,4 +58,4 @@ static backendmap backend_map = { } //namespace mkl } //namespace oneapi -#endif //_ONEMKL_BACKENDS_HPP_ +#endif //_ONEMATH_BACKENDS_HPP_ diff --git a/include/oneapi/math/detail/backends_table.hpp b/include/oneapi/math/detail/backends_table.hpp index b59ce2b10..4ed59b679 100644 --- a/include/oneapi/math/detail/backends_table.hpp +++ b/include/oneapi/math/detail/backends_table.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_BACKENDS_TABLE_HPP_ -#define _ONEMKL_BACKENDS_TABLE_HPP_ +#ifndef _ONEMATH_BACKENDS_TABLE_HPP_ +#define _ONEMATH_BACKENDS_TABLE_HPP_ #include #include @@ -47,40 +47,40 @@ static std::map>> libraries = { domain::blas, { { device::x86cpu, { -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND LIB_NAME("blas_mklcpu"), #endif -#ifdef ONEMKL_ENABLE_NETLIB_BACKEND +#ifdef ONEMATH_ENABLE_NETLIB_BACKEND LIB_NAME("blas_netlib"), #endif -#ifdef ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_CPU +#ifdef ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_CPU LIB_NAME("blas_portblas"), #endif } }, { device::intelgpu, { -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND LIB_NAME("blas_mklgpu"), #endif -#ifdef ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_GPU +#ifdef ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_GPU LIB_NAME("blas_portblas"), #endif } }, { device::amdgpu, { -#ifdef ONEMKL_ENABLE_ROCBLAS_BACKEND +#ifdef ONEMATH_ENABLE_ROCBLAS_BACKEND LIB_NAME("blas_rocblas"), #endif -#ifdef ONEMKL_ENABLE_PORTBLAS_BACKEND_AMD_GPU +#ifdef ONEMATH_ENABLE_PORTBLAS_BACKEND_AMD_GPU LIB_NAME("blas_portblas"), #endif } }, { device::nvidiagpu, { -#ifdef ONEMKL_ENABLE_CUBLAS_BACKEND +#ifdef ONEMATH_ENABLE_CUBLAS_BACKEND LIB_NAME("blas_cublas"), #endif -#ifdef ONEMKL_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU +#ifdef ONEMATH_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU LIB_NAME("blas_portblas"), #endif } }, @@ -94,37 +94,37 @@ static std::map>> libraries = { domain::dft, { { device::x86cpu, { -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND LIB_NAME("dft_mklcpu") #endif -#ifdef ONEMKL_ENABLE_PORTFFT_BACKEND +#ifdef ONEMATH_ENABLE_PORTFFT_BACKEND LIB_NAME("dft_portfft") #endif } }, { device::intelgpu, { -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND LIB_NAME("dft_mklgpu") #endif -#ifdef ONEMKL_ENABLE_PORTFFT_BACKEND +#ifdef ONEMATH_ENABLE_PORTFFT_BACKEND LIB_NAME("dft_portfft") #endif } }, { device::amdgpu, { -#ifdef ONEMKL_ENABLE_ROCFFT_BACKEND +#ifdef ONEMATH_ENABLE_ROCFFT_BACKEND LIB_NAME("dft_rocfft") #endif -#ifdef ONEMKL_ENABLE_PORTFFT_BACKEND +#ifdef ONEMATH_ENABLE_PORTFFT_BACKEND LIB_NAME("dft_portfft") #endif } }, { device::nvidiagpu, { -#ifdef ONEMKL_ENABLE_CUFFT_BACKEND +#ifdef ONEMATH_ENABLE_CUFFT_BACKEND LIB_NAME("dft_cufft") #endif -#ifdef ONEMKL_ENABLE_PORTFFT_BACKEND +#ifdef ONEMATH_ENABLE_PORTFFT_BACKEND LIB_NAME("dft_portfft") #endif } }, @@ -138,25 +138,25 @@ static std::map>> libraries = { domain::lapack, { { device::x86cpu, { -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND LIB_NAME("lapack_mklcpu") #endif } }, { device::intelgpu, { -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND LIB_NAME("lapack_mklgpu") #endif } }, { device::amdgpu, { -#ifdef ONEMKL_ENABLE_ROCSOLVER_BACKEND +#ifdef ONEMATH_ENABLE_ROCSOLVER_BACKEND LIB_NAME("lapack_rocsolver") #endif } }, { device::nvidiagpu, { -#ifdef ONEMKL_ENABLE_CUSOLVER_BACKEND +#ifdef ONEMATH_ENABLE_CUSOLVER_BACKEND LIB_NAME("lapack_cusolver") #endif } } } }, @@ -164,25 +164,25 @@ static std::map>> libraries = { domain::rng, { { device::x86cpu, { -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND LIB_NAME("rng_mklcpu") #endif } }, { device::intelgpu, { -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND LIB_NAME("rng_mklgpu") #endif } }, { device::amdgpu, { -#ifdef ONEMKL_ENABLE_ROCRAND_BACKEND +#ifdef ONEMATH_ENABLE_ROCRAND_BACKEND LIB_NAME("rng_rocrand") #endif } }, { device::nvidiagpu, { -#ifdef ONEMKL_ENABLE_CURAND_BACKEND +#ifdef ONEMATH_ENABLE_CURAND_BACKEND LIB_NAME("rng_curand") #endif } } } }, @@ -190,13 +190,13 @@ static std::map>> libraries = { domain::sparse_blas, { { device::x86cpu, { -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND LIB_NAME("sparse_blas_mklcpu") #endif } }, { device::intelgpu, { -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND LIB_NAME("sparse_blas_mklgpu") #endif } } } }, @@ -212,4 +212,4 @@ static std::map table_names = { { domain::blas, "mkl_blas_t } //namespace mkl } //namespace oneapi -#endif //_ONEMKL_BACKENDS_TABLE_HPP_ +#endif //_ONEMATH_BACKENDS_TABLE_HPP_ diff --git a/include/oneapi/math/detail/exceptions.hpp b/include/oneapi/math/detail/exceptions.hpp index 0b76d9d7e..7d77d1f43 100644 --- a/include/oneapi/math/detail/exceptions.hpp +++ b/include/oneapi/math/detail/exceptions.hpp @@ -19,8 +19,8 @@ // These are oneAPI Math Kernel Library (oneMKL) Interfaces specific exceptions -#ifndef _ONEMKL_DETAIL_EXCEPTIONS_HPP_ -#define _ONEMKL_DETAIL_EXCEPTIONS_HPP_ +#ifndef _ONEMATH_DETAIL_EXCEPTIONS_HPP_ +#define _ONEMATH_DETAIL_EXCEPTIONS_HPP_ #include #include @@ -56,4 +56,4 @@ class specification_mismatch : public oneapi::mkl::exception { } // namespace mkl } // namespace oneapi -#endif // _ONEMKL_DETAIL_EXCEPTIONS_HPP_ +#endif // _ONEMATH_DETAIL_EXCEPTIONS_HPP_ diff --git a/include/oneapi/math/detail/export.hpp b/include/oneapi/math/detail/export.hpp index 2b2596e09..851fe9d2f 100644 --- a/include/oneapi/math/detail/export.hpp +++ b/include/oneapi/math/detail/export.hpp @@ -17,24 +17,24 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef ONEMKL_EXPORT_H -#define ONEMKL_EXPORT_H +#ifndef ONEMATH_EXPORT_H +#define ONEMATH_EXPORT_H #include "oneapi/math/detail/config.hpp" -#if !defined(ONEMKL_BUILD_SHARED_LIBS) || !defined(_WIN64) -#define ONEMKL_EXPORT -#define ONEMKL_NO_EXPORT +#if !defined(ONEMATH_BUILD_SHARED_LIBS) || !defined(_WIN64) +#define ONEMATH_EXPORT +#define ONEMATH_NO_EXPORT #else -#ifndef ONEMKL_EXPORT +#ifndef ONEMATH_EXPORT #ifdef onemath_EXPORTS /* We are building this library */ -#define ONEMKL_EXPORT __declspec(dllexport) +#define ONEMATH_EXPORT __declspec(dllexport) #else /* We are using this library */ -#define ONEMKL_EXPORT __declspec(dllimport) +#define ONEMATH_EXPORT __declspec(dllimport) #endif #endif #endif -#endif /* ONEMKL_EXPORT_H */ +#endif /* ONEMATH_EXPORT_H */ diff --git a/include/oneapi/math/detail/get_device_id.hpp b/include/oneapi/math/detail/get_device_id.hpp index dac0bf08f..8521fb5b8 100644 --- a/include/oneapi/math/detail/get_device_id.hpp +++ b/include/oneapi/math/detail/get_device_id.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_GET_DEVICE_ID_HPP_ -#define _ONEMKL_GET_DEVICE_ID_HPP_ +#ifndef _ONEMATH_GET_DEVICE_ID_HPP_ +#define _ONEMATH_GET_DEVICE_ID_HPP_ #if __has_include() #include @@ -70,4 +70,4 @@ inline oneapi::mkl::device get_device_id(sycl::queue &queue) { } //namespace mkl } //namespace oneapi -#endif //_ONEMKL_GET_DEVICE_ID_HPP_ +#endif //_ONEMATH_GET_DEVICE_ID_HPP_ diff --git a/include/oneapi/math/dft.hpp b/include/oneapi/math/dft.hpp index 2041b296c..99d830bee 100644 --- a/include/oneapi/math/dft.hpp +++ b/include/oneapi/math/dft.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_HPP_ -#define _ONEMKL_DFT_HPP_ +#ifndef _ONEMATH_DFT_HPP_ +#define _ONEMATH_DFT_HPP_ #if __has_include() #include @@ -36,4 +36,4 @@ #include "oneapi/math/dft/forward.hpp" #include "oneapi/math/dft/backward.hpp" -#endif // _ONEMKL_DFT_HPP_ +#endif // _ONEMATH_DFT_HPP_ diff --git a/include/oneapi/math/dft/backward.hpp b/include/oneapi/math/dft/backward.hpp index 3cd03e13b..2cff87055 100644 --- a/include/oneapi/math/dft/backward.hpp +++ b/include/oneapi/math/dft/backward.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_BACKWARD_HPP_ -#define _ONEMKL_DFT_BACKWARD_HPP_ +#ifndef _ONEMATH_DFT_BACKWARD_HPP_ +#define _ONEMATH_DFT_BACKWARD_HPP_ #if __has_include() #include @@ -161,4 +161,4 @@ sycl::event compute_backward(descriptor_type &desc, input_type *in_re, input_typ } } // namespace oneapi::mkl::dft -#endif // _ONEMKL_DFT_BACKWARD_HPP_ +#endif // _ONEMATH_DFT_BACKWARD_HPP_ diff --git a/include/oneapi/math/dft/descriptor.hpp b/include/oneapi/math/dft/descriptor.hpp index fb618fd42..3d1fa1996 100644 --- a/include/oneapi/math/dft/descriptor.hpp +++ b/include/oneapi/math/dft/descriptor.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_DESCRIPTOR_HPP_ -#define _ONEMKL_DFT_DESCRIPTOR_HPP_ +#ifndef _ONEMATH_DFT_DESCRIPTOR_HPP_ +#define _ONEMATH_DFT_DESCRIPTOR_HPP_ #include "detail/descriptor_impl.hpp" #include "types.hpp" @@ -35,4 +35,4 @@ using descriptor = detail::descriptor; } // namespace mkl } // namespace oneapi -#endif // _ONEMKL_DFT_DESCRIPTOR_HPP_ +#endif // _ONEMATH_DFT_DESCRIPTOR_HPP_ diff --git a/include/oneapi/math/dft/detail/commit_impl.hpp b/include/oneapi/math/dft/detail/commit_impl.hpp index 9e827f357..1a2c013bc 100644 --- a/include/oneapi/math/dft/detail/commit_impl.hpp +++ b/include/oneapi/math/dft/detail/commit_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_COMMIT_IMPL_HPP_ -#define _ONEMKL_DFT_COMMIT_IMPL_HPP_ +#ifndef _ONEMATH_DFT_COMMIT_IMPL_HPP_ +#define _ONEMATH_DFT_COMMIT_IMPL_HPP_ #if __has_include() #include @@ -181,4 +181,4 @@ class commit_impl { } // namespace oneapi::mkl::dft::detail -#endif //_ONEMKL_DFT_COMMIT_IMPL_HPP_ +#endif //_ONEMATH_DFT_COMMIT_IMPL_HPP_ diff --git a/include/oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp b/include/oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp index 3d4470927..d053f7f55 100644 --- a/include/oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp +++ b/include/oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_CUFFT_HPP_ -#define _ONEMKL_DFT_CUFFT_HPP_ +#ifndef _ONEMATH_DFT_CUFFT_HPP_ +#define _ONEMATH_DFT_CUFFT_HPP_ #if __has_include() #include @@ -35,4 +35,4 @@ namespace oneapi::mkl::dft::cufft { } // namespace oneapi::mkl::dft::cufft -#endif // _ONEMKL_DFT_CUFFT_HPP_ +#endif // _ONEMATH_DFT_CUFFT_HPP_ diff --git a/include/oneapi/math/dft/detail/descriptor_impl.hpp b/include/oneapi/math/dft/detail/descriptor_impl.hpp index 36900fbc6..174c1a22d 100644 --- a/include/oneapi/math/dft/detail/descriptor_impl.hpp +++ b/include/oneapi/math/dft/detail/descriptor_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_DETAIL_DESCRIPTOR_IMPL_HPP_ -#define _ONEMKL_DFT_DETAIL_DESCRIPTOR_IMPL_HPP_ +#ifndef _ONEMATH_DFT_DETAIL_DESCRIPTOR_IMPL_HPP_ +#define _ONEMATH_DFT_DETAIL_DESCRIPTOR_IMPL_HPP_ #if __has_include() #include @@ -71,23 +71,23 @@ class descriptor { void commit(sycl::queue& queue); -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND void commit(backend_selector selector); #endif -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND void commit(backend_selector selector); #endif -#ifdef ONEMKL_ENABLE_CUFFT_BACKEND +#ifdef ONEMATH_ENABLE_CUFFT_BACKEND void commit(backend_selector selector); #endif -#ifdef ONEMKL_ENABLE_ROCFFT_BACKEND +#ifdef ONEMATH_ENABLE_ROCFFT_BACKEND void commit(backend_selector selector); #endif -#ifdef ONEMKL_ENABLE_PORTFFT_BACKEND +#ifdef ONEMATH_ENABLE_PORTFFT_BACKEND void commit(backend_selector selector); #endif @@ -121,4 +121,4 @@ inline commit_impl* get_commit(descriptor& desc) { } // namespace mkl } // namespace oneapi -#endif // _ONEMKL_DFT_DETAIL_DESCRIPTOR_IMPL_HPP_ +#endif // _ONEMATH_DFT_DETAIL_DESCRIPTOR_IMPL_HPP_ diff --git a/include/oneapi/math/dft/detail/dft_ct.hxx b/include/oneapi/math/dft/detail/dft_ct.hxx index 20cd537d8..bf479bde1 100644 --- a/include/oneapi/math/dft/detail/dft_ct.hxx +++ b/include/oneapi/math/dft/detail/dft_ct.hxx @@ -20,7 +20,7 @@ // Commit template -ONEMKL_EXPORT dft::detail::commit_impl *create_commit( +ONEMATH_EXPORT dft::detail::commit_impl *create_commit( const dft::detail::descriptor &desc, sycl::queue &sycl_queue); // BUFFER version @@ -34,23 +34,23 @@ using bwd = typename detail::descriptor_info::backward_type; //In-place transform template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &inout); //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &inout_re, sycl::buffer, 1> &inout_im); //Out-of-place transform template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, sycl::buffer, 1> &out); //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in_re, sycl::buffer, 1> &in_im, sycl::buffer, 1> &out_re, @@ -60,24 +60,24 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, //In-place transform template -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, const std::vector &dependencies); //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *inout_re, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *inout_re, scalar *inout_im, const std::vector &dependencies); //Out-of-place transform template -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, bwd *out, const std::vector &dependencies); //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *in_re, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *in_re, scalar *in_im, scalar *out_re, scalar *out_im, @@ -87,24 +87,24 @@ ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &inout); //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &inout_re, sycl::buffer, 1> &inout_im); //Out-of-place transform template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &in, sycl::buffer, 1> &out); //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &in_re, sycl::buffer, 1> &in_im, sycl::buffer, 1> &out_re, @@ -114,24 +114,24 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &desc, //In-place transform template -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, const std::vector &dependencies); //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *inout_re, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *inout_re, scalar *inout_im, const std::vector &dependencies); //Out-of-place transform template -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, fwd *out, const std::vector &dependencies); //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *in_re, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *in_re, scalar *in_im, scalar *out_re, scalar *out_im, diff --git a/include/oneapi/math/dft/detail/dft_loader.hpp b/include/oneapi/math/dft/detail/dft_loader.hpp index 173ced31a..a04f3e3cf 100644 --- a/include/oneapi/math/dft/detail/dft_loader.hpp +++ b/include/oneapi/math/dft/detail/dft_loader.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_LOADER_HPP_ -#define _ONEMKL_DFT_LOADER_HPP_ +#ifndef _ONEMATH_DFT_LOADER_HPP_ +#define _ONEMATH_DFT_LOADER_HPP_ #if __has_include() #include @@ -41,7 +41,7 @@ template class descriptor; template -ONEMKL_EXPORT commit_impl* create_commit(const descriptor& desc, +ONEMATH_EXPORT commit_impl* create_commit(const descriptor& desc, sycl::queue& queue); } // namespace detail @@ -49,4 +49,4 @@ ONEMKL_EXPORT commit_impl* create_commit(const descriptor& } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_DFT_LOADER_HPP_ +#endif //_ONEMATH_DFT_LOADER_HPP_ diff --git a/include/oneapi/math/dft/detail/external_workspace_helper.hpp b/include/oneapi/math/dft/detail/external_workspace_helper.hpp index 8938b70e1..b0ae40966 100644 --- a/include/oneapi/math/dft/detail/external_workspace_helper.hpp +++ b/include/oneapi/math/dft/detail/external_workspace_helper.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_EXTERNAL_WORKSPACE_HELPER_HPP_ -#define _ONEMKL_DFT_EXTERNAL_WORKSPACE_HELPER_HPP_ +#ifndef _ONEMATH_DFT_EXTERNAL_WORKSPACE_HELPER_HPP_ +#define _ONEMATH_DFT_EXTERNAL_WORKSPACE_HELPER_HPP_ #if __has_include() #include @@ -191,4 +191,4 @@ class external_workspace_helper { } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_DFT_EXTERNAL_WORKSPACE_HELPER_HPP_ +#endif //_ONEMATH_DFT_EXTERNAL_WORKSPACE_HELPER_HPP_ diff --git a/include/oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp b/include/oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp index 063d7d976..466bb59db 100644 --- a/include/oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp +++ b/include/oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_MKLCPU_HPP_ -#define _ONEMKL_DFT_MKLCPU_HPP_ +#ifndef _ONEMATH_DFT_MKLCPU_HPP_ +#define _ONEMATH_DFT_MKLCPU_HPP_ #if __has_include() #include @@ -35,4 +35,4 @@ namespace oneapi::mkl::dft::mklcpu { } // namespace oneapi::mkl::dft::mklcpu -#endif // _ONEMKL_DFT_MKLCPU_HPP_ +#endif // _ONEMATH_DFT_MKLCPU_HPP_ diff --git a/include/oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp b/include/oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp index 7ea9487cd..97dfd188a 100644 --- a/include/oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp +++ b/include/oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_MKLGPU_HPP_ -#define _ONEMKL_DFT_MKLGPU_HPP_ +#ifndef _ONEMATH_DFT_MKLGPU_HPP_ +#define _ONEMATH_DFT_MKLGPU_HPP_ #if __has_include() #include @@ -35,4 +35,4 @@ namespace oneapi::mkl::dft::mklgpu { } // namespace oneapi::mkl::dft::mklgpu -#endif // _ONEMKL_DFT_MKLGPU_HPP_ +#endif // _ONEMATH_DFT_MKLGPU_HPP_ diff --git a/include/oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp b/include/oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp index 4ef12e4bf..9ad6d762f 100644 --- a/include/oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp +++ b/include/oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_PORTFFT_HPP_ -#define _ONEMKL_DFT_PORTFFT_HPP_ +#ifndef _ONEMATH_DFT_PORTFFT_HPP_ +#define _ONEMATH_DFT_PORTFFT_HPP_ #if __has_include() #include @@ -36,4 +36,4 @@ namespace oneapi::mkl::dft::portfft { } // namespace oneapi::mkl::dft::portfft -#endif // _ONEMKL_DFT_PORTFFT_HPP_ +#endif // _ONEMATH_DFT_PORTFFT_HPP_ diff --git a/include/oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp b/include/oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp index a7712858a..0c3f79182 100644 --- a/include/oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp +++ b/include/oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_ROCFFT_HPP_ -#define _ONEMKL_DFT_ROCFFT_HPP_ +#ifndef _ONEMATH_DFT_ROCFFT_HPP_ +#define _ONEMATH_DFT_ROCFFT_HPP_ #if __has_include() #include @@ -35,4 +35,4 @@ namespace oneapi::mkl::dft::rocfft { } // namespace oneapi::mkl::dft::rocfft -#endif // _ONEMKL_DFT_ROCFFT_HPP_ +#endif // _ONEMATH_DFT_ROCFFT_HPP_ diff --git a/include/oneapi/math/dft/detail/types_impl.hpp b/include/oneapi/math/dft/detail/types_impl.hpp index 60eb922ab..aa9096301 100644 --- a/include/oneapi/math/dft/detail/types_impl.hpp +++ b/include/oneapi/math/dft/detail/types_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DETAIL_TYPES_IMPL_HPP_ -#define _ONEMKL_DETAIL_TYPES_IMPL_HPP_ +#ifndef _ONEMATH_DETAIL_TYPES_IMPL_HPP_ +#define _ONEMATH_DETAIL_TYPES_IMPL_HPP_ #if __has_include() #include @@ -231,4 +231,4 @@ class dft_values { } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_DETAIL_TYPES_IMPL_HPP_ +#endif //_ONEMATH_DETAIL_TYPES_IMPL_HPP_ diff --git a/include/oneapi/math/dft/forward.hpp b/include/oneapi/math/dft/forward.hpp index e43c39ce0..457e7c723 100644 --- a/include/oneapi/math/dft/forward.hpp +++ b/include/oneapi/math/dft/forward.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_FORWARD_HPP_ -#define _ONEMKL_DFT_FORWARD_HPP_ +#ifndef _ONEMATH_DFT_FORWARD_HPP_ +#define _ONEMATH_DFT_FORWARD_HPP_ #if __has_include() #include @@ -158,4 +158,4 @@ sycl::event compute_forward(descriptor_type &desc, input_type *in_re, input_type } } // namespace oneapi::mkl::dft -#endif // _ONEMKL_DFT_FORWARD_HPP_ +#endif // _ONEMATH_DFT_FORWARD_HPP_ diff --git a/include/oneapi/math/dft/types.hpp b/include/oneapi/math/dft/types.hpp index dfbcd3c28..106faeb33 100644 --- a/include/oneapi/math/dft/types.hpp +++ b/include/oneapi/math/dft/types.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_TYPES_HPP_ -#define _ONEMKL_DFT_TYPES_HPP_ +#ifndef _ONEMATH_DFT_TYPES_HPP_ +#define _ONEMATH_DFT_TYPES_HPP_ #include "detail/types_impl.hpp" @@ -39,4 +39,4 @@ using DFT_ERROR = detail::DFT_ERROR; } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_TYPES_HPP_ +#endif //_ONEMATH_TYPES_HPP_ diff --git a/include/oneapi/math/exceptions.hpp b/include/oneapi/math/exceptions.hpp index 7d3064237..1a5a7addf 100644 --- a/include/oneapi/math/exceptions.hpp +++ b/include/oneapi/math/exceptions.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_EXCEPTIONS_HPP_ -#define _ONEMKL_EXCEPTIONS_HPP_ +#ifndef _ONEMATH_EXCEPTIONS_HPP_ +#define _ONEMATH_EXCEPTIONS_HPP_ #if __has_include() #include @@ -127,4 +127,4 @@ class library_not_found : public oneapi::mkl::exception { } // namespace mkl } // namespace oneapi -#endif // _ONEMKL_EXCEPTIONS_HPP_ +#endif // _ONEMATH_EXCEPTIONS_HPP_ diff --git a/include/oneapi/math/lapack.hpp b/include/oneapi/math/lapack.hpp index dda570563..c862513c4 100644 --- a/include/oneapi/math/lapack.hpp +++ b/include/oneapi/math/lapack.hpp @@ -21,16 +21,16 @@ #include "oneapi/math/detail/config.hpp" -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND #include "oneapi/math/lapack/detail/mklcpu/lapack_ct.hpp" #endif -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND #include "oneapi/math/lapack/detail/mklgpu/lapack_ct.hpp" #endif -#ifdef ONEMKL_ENABLE_CUSOLVER_BACKEND +#ifdef ONEMATH_ENABLE_CUSOLVER_BACKEND #include "oneapi/math/lapack/detail/cusolver/lapack_ct.hpp" #endif -#ifdef ONEMKL_ENABLE_ROCSOLVER_BACKEND +#ifdef ONEMATH_ENABLE_ROCSOLVER_BACKEND #include "oneapi/math/lapack/detail/rocsolver/lapack_ct.hpp" #endif diff --git a/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp index 3d363cb66..7c27483e8 100644 --- a/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp +++ b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp @@ -16,8 +16,8 @@ * limitations under the License. * **************************************************************************/ -#ifndef _ONEMKL_LAPACK_CUSOLVER_HPP_ -#define _ONEMKL_LAPACK_CUSOLVER_HPP_ +#ifndef _ONEMATH_LAPACK_CUSOLVER_HPP_ +#define _ONEMATH_LAPACK_CUSOLVER_HPP_ #if __has_include() #include #else @@ -42,4 +42,4 @@ namespace cusolver { } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_LAPACK_CUSOLVER_HPP_ +#endif //_ONEMATH_LAPACK_CUSOLVER_HPP_ diff --git a/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx index ffa9c3007..26c3323b1 100644 --- a/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx +++ b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx @@ -19,7 +19,7 @@ // Buffer APIs -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, @@ -27,18 +27,18 @@ ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, @@ -46,121 +46,121 @@ ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, @@ -168,7 +168,7 @@ ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::m sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, @@ -176,292 +176,292 @@ ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::m sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -469,7 +469,7 @@ ONEMKL_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -477,7 +477,7 @@ ONEMKL_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -485,7 +485,7 @@ ONEMKL_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -493,7 +493,7 @@ ONEMKL_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -501,7 +501,7 @@ ONEMKL_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -509,73 +509,73 @@ ONEMKL_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -584,7 +584,7 @@ ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -593,79 +593,79 @@ ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -673,7 +673,7 @@ ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std:: sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -681,14 +681,14 @@ ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std:: sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, @@ -697,147 +697,147 @@ ONEMKL_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t // USM APIs -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tauq, double *taup, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tauq, float *taup, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, @@ -845,7 +845,7 @@ ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, @@ -853,303 +853,303 @@ ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, @@ -1157,14 +1157,14 @@ ONEMKL_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, @@ -1172,14 +1172,14 @@ ONEMKL_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, @@ -1187,26 +1187,26 @@ ONEMKL_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, onea std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, @@ -1214,52 +1214,52 @@ ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::i std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, @@ -1267,83 +1267,83 @@ ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::i std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, @@ -1351,7 +1351,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, @@ -1359,7 +1359,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1368,7 +1368,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1377,7 +1377,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, @@ -1385,7 +1385,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, @@ -1393,7 +1393,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1401,7 +1401,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1410,97 +1410,97 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -1508,7 +1508,7 @@ ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -1516,28 +1516,28 @@ ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, @@ -1545,14 +1545,14 @@ ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *upl std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, @@ -1560,14 +1560,14 @@ ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::i std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, @@ -1577,254 +1577,254 @@ ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std:: // SCRATCHPAD APIs template -ONEMKL_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template -ONEMKL_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMKL_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template -ONEMKL_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMKL_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template -ONEMKL_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMKL_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, +ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); diff --git a/include/oneapi/math/lapack/detail/lapack_loader.hpp b/include/oneapi/math/lapack/detail/lapack_loader.hpp index 11a9331d4..88d92688c 100644 --- a/include/oneapi/math/lapack/detail/lapack_loader.hpp +++ b/include/oneapi/math/lapack/detail/lapack_loader.hpp @@ -38,496 +38,496 @@ namespace mkl { namespace lapack { namespace detail { -ONEMKL_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, +ONEMATH_EXPORT void heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, +ONEMATH_EXPORT void heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT void hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT void hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +ONEMATH_EXPORT void orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +ONEMATH_EXPORT void orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, +ONEMATH_EXPORT void syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, +ONEMATH_EXPORT void syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT void sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT void sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +ONEMATH_EXPORT void ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +ONEMATH_EXPORT void ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT void unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, @@ -535,7 +535,7 @@ ONEMKL_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, @@ -543,516 +543,516 @@ ONEMKL_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tauq, double *taup, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tauq, float *taup, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event heevd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event heevd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT sycl::event hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT sycl::event hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrf(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrf(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgbr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgbr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgtr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgtr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormtr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormtr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormrq(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormrq(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormqr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormqr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syevd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syevd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT sycl::event sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT sycl::event sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrd(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungbr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungbr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungtr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungtr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmrq(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmrq(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, @@ -1060,14 +1060,14 @@ ONEMKL_EXPORT sycl::event unmrq(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmqr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmqr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, @@ -1075,14 +1075,14 @@ ONEMKL_EXPORT sycl::event unmqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmtr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmtr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, @@ -1090,157 +1090,157 @@ ONEMKL_EXPORT sycl::event unmtr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, @@ -1248,7 +1248,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, @@ -1256,33 +1256,33 @@ ONEMKL_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch( +ONEMATH_EXPORT sycl::event getrs_batch( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch( +ONEMATH_EXPORT sycl::event getrs_batch( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, @@ -1290,101 +1290,101 @@ ONEMKL_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch( +ONEMATH_EXPORT sycl::event getrs_batch( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, @@ -1392,7 +1392,7 @@ ONEMKL_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, @@ -1400,26 +1400,26 @@ ONEMKL_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, @@ -1427,26 +1427,26 @@ ONEMKL_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &q std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch( +ONEMATH_EXPORT sycl::event ungqr_batch( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, @@ -1681,699 +1681,699 @@ std::int64_t ungqr_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t gebrd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gebrd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gebrd_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gebrd_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gerqf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gerqf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gerqf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gerqf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t geqrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t geqrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t geqrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t geqrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gesvd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> -ONEMKL_EXPORT std::int64_t gesvd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> -ONEMKL_EXPORT std::int64_t gesvd_scratchpad_size>( +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> -ONEMKL_EXPORT std::int64_t gesvd_scratchpad_size>( +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> -ONEMKL_EXPORT std::int64_t getrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getri_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getri_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getri_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getri_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getrs_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t getrs_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t getrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t getrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t heevd_scratchpad_size>( +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t heevd_scratchpad_size>( +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t hegvd_scratchpad_size>( +ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t hegvd_scratchpad_size>( +ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t hetrd_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t hetrd_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgbr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgbr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgtr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgtr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgqr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgqr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ormrq_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t ormrq_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t ormqr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t ormqr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t ormtr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t ormtr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t potrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potrs_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t potrs_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t potrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t potrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t potri_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potri_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potri_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potri_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sytrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sytrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sytrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sytrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t syevd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t syevd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t sytrd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sytrd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t trtrs_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t trtrs_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t trtrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t trtrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t ungbr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ungbr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ungqr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ungqr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ungtr_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ungtr_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t unmrq_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t unmrq_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t unmqr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t unmqr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t unmtr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t unmtr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); } //namespace detail diff --git a/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx b/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx index 372e2646b..ef6908cc5 100644 --- a/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx +++ b/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx @@ -17,467 +17,467 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -485,7 +485,7 @@ ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -493,643 +493,643 @@ ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tauq, double *taup, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tauq, float *taup, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1137,7 +1137,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1145,28 +1145,28 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1174,141 +1174,141 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, @@ -1506,634 +1506,634 @@ std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, st std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gebrd_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gebrd_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t geqrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t geqrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> -ONEMKL_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> -ONEMKL_EXPORT std::int64_t gesvd_scratchpad_size>( +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> -ONEMKL_EXPORT std::int64_t gesvd_scratchpad_size>( +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> -ONEMKL_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getri_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getri_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t getrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t getrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t heevd_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t heevd_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t hegvd_scratchpad_size>( +ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t hegvd_scratchpad_size>( +ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t potrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t potrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potri_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t potri_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t trtrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t trtrs_scratchpad_size>( +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMKL_EXPORT std::int64_t ungbr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ungbr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ungqr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ungqr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMKL_EXPORT std::int64_t unmrq_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t unmrq_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t unmqr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t unmqr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t unmtr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t unmtr_scratchpad_size>( +ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size>( +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); diff --git a/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp index 552235e3e..28d02a542 100644 --- a/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp +++ b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp @@ -19,8 +19,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_LAPACK_ROCSOLVER_HPP_ -#define _ONEMKL_LAPACK_ROCSOLVER_HPP_ +#ifndef _ONEMATH_LAPACK_ROCSOLVER_HPP_ +#define _ONEMATH_LAPACK_ROCSOLVER_HPP_ #if __has_include() #include @@ -46,4 +46,4 @@ namespace rocsolver { } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_LAPACK_ROCSOLVER_HPP_ +#endif //_ONEMATH_LAPACK_ROCSOLVER_HPP_ diff --git a/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx index c68009e54..819bb1cd6 100644 --- a/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx +++ b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx @@ -21,7 +21,7 @@ // Buffer APIs -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, @@ -29,18 +29,18 @@ ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, @@ -48,121 +48,121 @@ ONEMKL_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, +ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, @@ -170,7 +170,7 @@ ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::m sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, @@ -178,292 +178,292 @@ ONEMKL_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::m sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -471,7 +471,7 @@ ONEMKL_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -479,7 +479,7 @@ ONEMKL_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -487,7 +487,7 @@ ONEMKL_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -495,7 +495,7 @@ ONEMKL_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -503,7 +503,7 @@ ONEMKL_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -511,73 +511,73 @@ ONEMKL_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -586,7 +586,7 @@ ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -595,79 +595,79 @@ ONEMKL_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -675,7 +675,7 @@ ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std:: sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -683,14 +683,14 @@ ONEMKL_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std:: sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMKL_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, @@ -699,147 +699,147 @@ ONEMKL_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t // USM APIs -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tauq, double *taup, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tauq, float *taup, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, @@ -847,7 +847,7 @@ ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, @@ -855,303 +855,303 @@ ONEMKL_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, @@ -1159,14 +1159,14 @@ ONEMKL_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, @@ -1174,14 +1174,14 @@ ONEMKL_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, @@ -1189,26 +1189,26 @@ ONEMKL_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, onea std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, @@ -1216,52 +1216,52 @@ ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::i std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, @@ -1269,83 +1269,83 @@ ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::i std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, @@ -1353,7 +1353,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, @@ -1361,7 +1361,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1370,7 +1370,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1379,7 +1379,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, @@ -1387,7 +1387,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, @@ -1395,7 +1395,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1403,7 +1403,7 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1412,97 +1412,97 @@ ONEMKL_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -1510,7 +1510,7 @@ ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -1518,28 +1518,28 @@ ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, @@ -1547,14 +1547,14 @@ ONEMKL_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *upl std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, @@ -1562,14 +1562,14 @@ ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::i std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, @@ -1579,173 +1579,173 @@ ONEMKL_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std:: // SCRATCHPAD APIs template -ONEMKL_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template -ONEMKL_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMKL_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template -ONEMKL_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMKL_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template -ONEMKL_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMKL_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, +ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMKL_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, +ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, @@ -1754,82 +1754,82 @@ ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template -ONEMKL_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t getrs_batch_scratchpad_size( +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMKL_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); diff --git a/include/oneapi/math/rng.hpp b/include/oneapi/math/rng.hpp index 2c741660c..71c9e57b9 100644 --- a/include/oneapi/math/rng.hpp +++ b/include/oneapi/math/rng.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_RNG_HPP_ -#define _ONEMKL_RNG_HPP_ +#ifndef _ONEMATH_RNG_HPP_ +#define _ONEMATH_RNG_HPP_ #if __has_include() #include @@ -38,4 +38,4 @@ #include "oneapi/math/rng/distributions.hpp" #include "oneapi/math/rng/engines.hpp" -#endif // _ONEMKL_RNG_HPP_ +#endif // _ONEMATH_RNG_HPP_ diff --git a/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp b/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp index 156da53ad..7148f8a24 100644 --- a/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp +++ b/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp @@ -56,8 +56,8 @@ * so. ******************************************************************************/ -#ifndef _ONEMKL_RNG_CURAND_HPP_ -#define _ONEMKL_RNG_CURAND_HPP_ +#ifndef _ONEMATH_RNG_CURAND_HPP_ +#define _ONEMATH_RNG_CURAND_HPP_ #if __has_include() #include @@ -74,16 +74,16 @@ namespace mkl { namespace rng { namespace curand { -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); } // namespace curand @@ -91,4 +91,4 @@ ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_RNG_CURAND_HPP_ +#endif //_ONEMATH_RNG_CURAND_HPP_ diff --git a/include/oneapi/math/rng/detail/engine_impl.hpp b/include/oneapi/math/rng/detail/engine_impl.hpp index 0cdb05952..36429b066 100644 --- a/include/oneapi/math/rng/detail/engine_impl.hpp +++ b/include/oneapi/math/rng/detail/engine_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_RNG_ENGINE_IMPL_HPP_ -#define _ONEMKL_RNG_ENGINE_IMPL_HPP_ +#ifndef _ONEMATH_RNG_ENGINE_IMPL_HPP_ +#define _ONEMATH_RNG_ENGINE_IMPL_HPP_ #include #if __has_include() @@ -194,4 +194,4 @@ class engine_impl { } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_RNG_ENGINE_IMPL_HPP_ +#endif //_ONEMATH_RNG_ENGINE_IMPL_HPP_ diff --git a/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp b/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp index 4a46cc84a..0e97d6003 100644 --- a/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp +++ b/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_RNG_MKLCPU_HPP_ -#define _ONEMKL_RNG_MKLCPU_HPP_ +#ifndef _ONEMATH_RNG_MKLCPU_HPP_ +#define _ONEMATH_RNG_MKLCPU_HPP_ #include #if __has_include() @@ -35,16 +35,16 @@ namespace mkl { namespace rng { namespace mklcpu { -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); } // namespace mklcpu @@ -52,4 +52,4 @@ ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_RNG_MKLCPU_HPP_ +#endif //_ONEMATH_RNG_MKLCPU_HPP_ diff --git a/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp b/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp index 83bfc7488..0de6f9b45 100644 --- a/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp +++ b/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_RNG_MKLGPU_HPP_ -#define _ONEMKL_RNG_MKLGPU_HPP_ +#ifndef _ONEMATH_RNG_MKLGPU_HPP_ +#define _ONEMATH_RNG_MKLGPU_HPP_ #include #if __has_include() @@ -35,16 +35,16 @@ namespace mkl { namespace rng { namespace mklgpu { -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); } // namespace mklgpu @@ -52,4 +52,4 @@ ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_RNG_MKLGPU_HPP_ +#endif //_ONEMATH_RNG_MKLGPU_HPP_ diff --git a/include/oneapi/math/rng/detail/rng_loader.hpp b/include/oneapi/math/rng/detail/rng_loader.hpp index d03ea8ea9..dcb92ea84 100644 --- a/include/oneapi/math/rng/detail/rng_loader.hpp +++ b/include/oneapi/math/rng/detail/rng_loader.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_RNG_LOADER_HPP_ -#define _ONEMKL_RNG_LOADER_HPP_ +#ifndef _ONEMATH_RNG_LOADER_HPP_ +#define _ONEMATH_RNG_LOADER_HPP_ #include #if __has_include() @@ -37,16 +37,16 @@ namespace mkl { namespace rng { namespace detail { -ONEMKL_EXPORT engine_impl* create_philox4x32x10(oneapi::mkl::device libkey, sycl::queue queue, +ONEMATH_EXPORT engine_impl* create_philox4x32x10(oneapi::mkl::device libkey, sycl::queue queue, std::uint64_t seed); -ONEMKL_EXPORT engine_impl* create_philox4x32x10(oneapi::mkl::device libkey, sycl::queue queue, +ONEMATH_EXPORT engine_impl* create_philox4x32x10(oneapi::mkl::device libkey, sycl::queue queue, std::initializer_list seed); -ONEMKL_EXPORT engine_impl* create_mrg32k3a(oneapi::mkl::device libkey, sycl::queue queue, +ONEMATH_EXPORT engine_impl* create_mrg32k3a(oneapi::mkl::device libkey, sycl::queue queue, std::uint32_t seed); -ONEMKL_EXPORT engine_impl* create_mrg32k3a(oneapi::mkl::device libkey, sycl::queue queue, +ONEMATH_EXPORT engine_impl* create_mrg32k3a(oneapi::mkl::device libkey, sycl::queue queue, std::initializer_list seed); } // namespace detail @@ -54,4 +54,4 @@ ONEMKL_EXPORT engine_impl* create_mrg32k3a(oneapi::mkl::device libkey, sycl::que } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_RNG_LOADER_HPP_ +#endif //_ONEMATH_RNG_LOADER_HPP_ diff --git a/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp b/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp index 05169e226..00b0865af 100644 --- a/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp +++ b/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp @@ -58,8 +58,8 @@ * so. ******************************************************************************/ -#ifndef _ONEMKL_RNG_ROCRAND_HPP_ -#define _ONEMKL_RNG_ROCRAND_HPP_ +#ifndef _ONEMATH_RNG_ROCRAND_HPP_ +#define _ONEMATH_RNG_ROCRAND_HPP_ #if __has_include() #include @@ -76,16 +76,16 @@ namespace mkl { namespace rng { namespace rocrand { -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed); -ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( +ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); } // namespace rocrand @@ -93,4 +93,4 @@ ONEMKL_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_RNG_ROCRAND_HPP_ +#endif //_ONEMATH_RNG_ROCRAND_HPP_ diff --git a/include/oneapi/math/rng/distributions.hpp b/include/oneapi/math/rng/distributions.hpp index 58870c55a..1167b62a8 100644 --- a/include/oneapi/math/rng/distributions.hpp +++ b/include/oneapi/math/rng/distributions.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_RNG_DISTRIBUTIONS_HPP_ -#define _ONEMKL_RNG_DISTRIBUTIONS_HPP_ +#ifndef _ONEMATH_RNG_DISTRIBUTIONS_HPP_ +#define _ONEMATH_RNG_DISTRIBUTIONS_HPP_ #include #include @@ -370,4 +370,4 @@ class bits { } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_RNG_DISTRIBUTIONS_HPP_ +#endif //_ONEMATH_RNG_DISTRIBUTIONS_HPP_ diff --git a/include/oneapi/math/rng/engines.hpp b/include/oneapi/math/rng/engines.hpp index 2edccdd7c..e72948ae1 100644 --- a/include/oneapi/math/rng/engines.hpp +++ b/include/oneapi/math/rng/engines.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_RNG_ENGINES_HPP_ -#define _ONEMKL_RNG_ENGINES_HPP_ +#ifndef _ONEMATH_RNG_ENGINES_HPP_ +#define _ONEMATH_RNG_ENGINES_HPP_ #include #include @@ -35,16 +35,16 @@ #include "oneapi/math/rng/detail/engine_impl.hpp" #include "oneapi/math/rng/detail/rng_loader.hpp" -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND #include "oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp" #endif -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND #include "oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp" #endif -#ifdef ONEMKL_ENABLE_CURAND_BACKEND +#ifdef ONEMATH_ENABLE_CURAND_BACKEND #include "oneapi/math/rng/detail/curand/onemath_rng_curand.hpp" #endif -#ifdef ONEMKL_ENABLE_ROCRAND_BACKEND +#ifdef ONEMATH_ENABLE_ROCRAND_BACKEND #include "oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp" #endif @@ -68,7 +68,7 @@ class philox4x32x10 { philox4x32x10(sycl::queue queue, std::initializer_list seed) : pimpl_(detail::create_philox4x32x10(get_device_id(queue), queue, seed)) {} -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND philox4x32x10(backend_selector selector, std::uint64_t seed = default_seed) : pimpl_(mklcpu::create_philox4x32x10(selector.get_queue(), seed)) {} @@ -77,7 +77,7 @@ class philox4x32x10 { : pimpl_(mklcpu::create_philox4x32x10(selector.get_queue(), seed)) {} #endif -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND philox4x32x10(backend_selector selector, std::uint64_t seed = default_seed) : pimpl_(mklgpu::create_philox4x32x10(selector.get_queue(), seed)) {} @@ -86,7 +86,7 @@ class philox4x32x10 { : pimpl_(mklgpu::create_philox4x32x10(selector.get_queue(), seed)) {} #endif -#ifdef ONEMKL_ENABLE_CURAND_BACKEND +#ifdef ONEMATH_ENABLE_CURAND_BACKEND philox4x32x10(backend_selector selector, std::uint64_t seed = default_seed) : pimpl_(curand::create_philox4x32x10(selector.get_queue(), seed)) {} @@ -94,7 +94,7 @@ class philox4x32x10 { std::initializer_list seed) : pimpl_(curand::create_philox4x32x10(selector.get_queue(), seed)) {} #endif -#ifdef ONEMKL_ENABLE_ROCRAND_BACKEND +#ifdef ONEMATH_ENABLE_ROCRAND_BACKEND philox4x32x10(backend_selector selector, std::uint64_t seed = default_seed) : pimpl_(rocrand::create_philox4x32x10(selector.get_queue(), seed)) {} @@ -160,7 +160,7 @@ class mrg32k3a { mrg32k3a(sycl::queue queue, std::initializer_list seed) : pimpl_(detail::create_mrg32k3a(get_device_id(queue), queue, seed)) {} -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND mrg32k3a(backend_selector selector, std::uint32_t seed = default_seed) : pimpl_(mklcpu::create_mrg32k3a(selector.get_queue(), seed)) {} @@ -168,7 +168,7 @@ class mrg32k3a { : pimpl_(mklcpu::create_mrg32k3a(selector.get_queue(), seed)) {} #endif -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND mrg32k3a(backend_selector selector, std::uint32_t seed = default_seed) : pimpl_(mklgpu::create_mrg32k3a(selector.get_queue(), seed)) {} @@ -176,7 +176,7 @@ class mrg32k3a { : pimpl_(mklgpu::create_mrg32k3a(selector.get_queue(), seed)) {} #endif -#ifdef ONEMKL_ENABLE_CURAND_BACKEND +#ifdef ONEMATH_ENABLE_CURAND_BACKEND mrg32k3a(backend_selector selector, std::uint32_t seed = default_seed) : pimpl_(curand::create_mrg32k3a(selector.get_queue(), seed)) {} @@ -184,7 +184,7 @@ class mrg32k3a { : pimpl_(curand::create_mrg32k3a(selector.get_queue(), seed)) {} #endif -#ifdef ONEMKL_ENABLE_ROCRAND_BACKEND +#ifdef ONEMATH_ENABLE_ROCRAND_BACKEND mrg32k3a(backend_selector selector, std::uint32_t seed = default_seed) : pimpl_(rocrand::create_mrg32k3a(selector.get_queue(), seed)) {} @@ -240,4 +240,4 @@ using default_engine = philox4x32x10; } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_RNG_ENGINES_HPP_ +#endif //_ONEMATH_RNG_ENGINES_HPP_ diff --git a/include/oneapi/math/rng/functions.hpp b/include/oneapi/math/rng/functions.hpp index 2bd55b180..e84762b4e 100644 --- a/include/oneapi/math/rng/functions.hpp +++ b/include/oneapi/math/rng/functions.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_RNG_FUNCTIONS_HPP_ -#define _ONEMKL_RNG_FUNCTIONS_HPP_ +#ifndef _ONEMATH_RNG_FUNCTIONS_HPP_ +#define _ONEMATH_RNG_FUNCTIONS_HPP_ #include #if __has_include() @@ -117,4 +117,4 @@ static inline void leapfrog(Engine& engine, std::uint64_t idx, std::uint64_t str } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_RNG_FUNCTIONS_HPP_ +#endif //_ONEMATH_RNG_FUNCTIONS_HPP_ diff --git a/include/oneapi/math/rng/predicates.hpp b/include/oneapi/math/rng/predicates.hpp index bdddb5636..798825bbd 100644 --- a/include/oneapi/math/rng/predicates.hpp +++ b/include/oneapi/math/rng/predicates.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_RNG_PREDICATES_HPP_ -#define _ONEMKL_RNG_PREDICATES_HPP_ +#ifndef _ONEMATH_RNG_PREDICATES_HPP_ +#define _ONEMATH_RNG_PREDICATES_HPP_ #include #if __has_include() @@ -39,7 +39,7 @@ namespace rng { template inline void generate_precondition(const Distr& /*distr*/, Engine& /*engine*/, std::int64_t n, sycl::buffer& r) { -#ifndef ONEMKL_DISABLE_PREDICATES +#ifndef ONEMATH_DISABLE_PREDICATES if (n < 0 || n > r.size()) { throw oneapi::mkl::invalid_argument("rng", "generate", "n"); } @@ -52,7 +52,7 @@ template inline void generate_precondition(const Distr& /*distr*/, Engine& /*engine*/, std::int64_t n, typename Distr::result_type* r, const std::vector& /*dependencies*/) { -#ifndef ONEMKL_DISABLE_PREDICATES +#ifndef ONEMATH_DISABLE_PREDICATES if (n < 0) { throw oneapi::mkl::invalid_argument("rng", "generate", "n"); } @@ -66,4 +66,4 @@ inline void generate_precondition(const Distr& /*distr*/, Engine& /*engine*/, st } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_RNG_PREDICATES_HPP_ +#endif //_ONEMATH_RNG_PREDICATES_HPP_ diff --git a/include/oneapi/math/sparse_blas.hpp b/include/oneapi/math/sparse_blas.hpp index 0d7308ceb..0d4d2eff3 100644 --- a/include/oneapi/math/sparse_blas.hpp +++ b/include/oneapi/math/sparse_blas.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_HPP_ -#define _ONEMKL_SPARSE_BLAS_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_HPP_ +#define _ONEMATH_SPARSE_BLAS_HPP_ #if __has_include() #include @@ -28,13 +28,13 @@ #include "oneapi/math/detail/config.hpp" -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND #include "sparse_blas/detail/mklcpu/sparse_blas_ct.hpp" #endif -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND #include "sparse_blas/detail/mklgpu/sparse_blas_ct.hpp" #endif #include "sparse_blas/detail/sparse_blas_rt.hpp" -#endif // _ONEMKL_SPARSE_BLAS_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/handles.hpp b/include/oneapi/math/sparse_blas/detail/handles.hpp index 0566f93b4..792cfe3a1 100644 --- a/include/oneapi/math/sparse_blas/detail/handles.hpp +++ b/include/oneapi/math/sparse_blas/detail/handles.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_DETAIL_HANDLES_HPP_ -#define _ONEMKL_SPARSE_BLAS_DETAIL_HANDLES_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_DETAIL_HANDLES_HPP_ +#define _ONEMATH_SPARSE_BLAS_DETAIL_HANDLES_HPP_ namespace oneapi::mkl::sparse { @@ -35,4 +35,4 @@ using matrix_handle_t = matrix_handle*; } // namespace oneapi::mkl::sparse -#endif // _ONEMKL_SPARSE_BLAS_DETAIL_HANDLES_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_DETAIL_HANDLES_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/helper_types.hpp b/include/oneapi/math/sparse_blas/detail/helper_types.hpp index ace216f00..95b5a6d48 100644 --- a/include/oneapi/math/sparse_blas/detail/helper_types.hpp +++ b/include/oneapi/math/sparse_blas/detail/helper_types.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_DETAIL_HELPER_TYPES_HPP_ -#define _ONEMKL_SPARSE_BLAS_DETAIL_HELPER_TYPES_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_DETAIL_HELPER_TYPES_HPP_ +#define _ONEMATH_SPARSE_BLAS_DETAIL_HELPER_TYPES_HPP_ #include #include @@ -47,4 +47,4 @@ inline constexpr bool are_fp_int_supported_v = } // namespace mkl } // namespace oneapi -#endif // _ONEMKL_SPARSE_BLAS_DETAIL_HELPER_TYPES_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_DETAIL_HELPER_TYPES_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp b/include/oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp index e3b035844..710c93d2a 100644 --- a/include/oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_DETAIL_MKLCPU_ONEMKL_SPARSE_BLAS_MKLCPU_HPP_ -#define _ONEMKL_SPARSE_BLAS_DETAIL_MKLCPU_ONEMKL_SPARSE_BLAS_MKLCPU_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_DETAIL_MKLCPU_ONEMATH_SPARSE_BLAS_MKLCPU_HPP_ +#define _ONEMATH_SPARSE_BLAS_DETAIL_MKLCPU_ONEMATH_SPARSE_BLAS_MKLCPU_HPP_ #include "oneapi/math/detail/export.hpp" #include "oneapi/math/sparse_blas/detail/helper_types.hpp" @@ -32,4 +32,4 @@ namespace detail = oneapi::mkl::sparse::detail; } // namespace oneapi::mkl::sparse::mklcpu -#endif // _ONEMKL_SPARSE_BLAS_DETAIL_MKLCPU_ONEMKL_SPARSE_BLAS_MKLCPU_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_DETAIL_MKLCPU_ONEMATH_SPARSE_BLAS_MKLCPU_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp b/include/oneapi/math/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp index 89734a68a..d4678ce1e 100644 --- a/include/oneapi/math/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_DETAIL_MKLCPU_SPARSE_BLAS_CT_HPP_ -#define _ONEMKL_SPARSE_BLAS_DETAIL_MKLCPU_SPARSE_BLAS_CT_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_DETAIL_MKLCPU_SPARSE_BLAS_CT_HPP_ +#define _ONEMATH_SPARSE_BLAS_DETAIL_MKLCPU_SPARSE_BLAS_CT_HPP_ #include "oneapi/math/detail/backends.hpp" #include "oneapi/math/detail/backend_selector.hpp" @@ -37,4 +37,4 @@ namespace sparse { } //namespace mkl } //namespace oneapi -#endif // _ONEMKL_SPARSE_BLAS_DETAIL_MKLCPU_SPARSE_BLAS_CT_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_DETAIL_MKLCPU_SPARSE_BLAS_CT_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp b/include/oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp index 5e5ab093a..03c3592b2 100644 --- a/include/oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_DETAIL_MKLGPU_ONEMKL_SPARSE_BLAS_MKLGPU_HPP_ -#define _ONEMKL_SPARSE_BLAS_DETAIL_MKLGPU_ONEMKL_SPARSE_BLAS_MKLGPU_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_DETAIL_MKLGPU_ONEMATH_SPARSE_BLAS_MKLGPU_HPP_ +#define _ONEMATH_SPARSE_BLAS_DETAIL_MKLGPU_ONEMATH_SPARSE_BLAS_MKLGPU_HPP_ #include "oneapi/math/detail/export.hpp" #include "oneapi/math/sparse_blas/detail/helper_types.hpp" @@ -32,4 +32,4 @@ namespace detail = oneapi::mkl::sparse::detail; } // namespace oneapi::mkl::sparse::mklgpu -#endif // _ONEMKL_SPARSE_BLAS_DETAIL_MKLGPU_ONEMKL_SPARSE_BLAS_MKLGPU_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_DETAIL_MKLGPU_ONEMATH_SPARSE_BLAS_MKLGPU_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp b/include/oneapi/math/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp index 7b94241cf..19b4805c9 100644 --- a/include/oneapi/math/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_DETAIL_MKLGPU_SPARSE_BLAS_CT_HPP_ -#define _ONEMKL_SPARSE_BLAS_DETAIL_MKLGPU_SPARSE_BLAS_CT_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_DETAIL_MKLGPU_SPARSE_BLAS_CT_HPP_ +#define _ONEMATH_SPARSE_BLAS_DETAIL_MKLGPU_SPARSE_BLAS_CT_HPP_ #include "oneapi/math/detail/backends.hpp" #include "oneapi/math/detail/backend_selector.hpp" @@ -37,4 +37,4 @@ namespace sparse { } //namespace mkl } //namespace oneapi -#endif // _ONEMKL_SPARSE_BLAS_DETAIL_MKLGPU_SPARSE_BLAS_CT_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_DETAIL_MKLGPU_SPARSE_BLAS_CT_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx b/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx index a0987b022..65c30750c 100644 --- a/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx +++ b/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx @@ -22,123 +22,123 @@ // Dense vector template -ONEMKL_EXPORT void init_dense_vector(sycl::queue &queue, dense_vector_handle_t *p_dvhandle, +ONEMATH_EXPORT void init_dense_vector(sycl::queue &queue, dense_vector_handle_t *p_dvhandle, std::int64_t size, sycl::buffer val); template -ONEMKL_EXPORT void init_dense_vector(sycl::queue &queue, dense_vector_handle_t *p_dvhandle, +ONEMATH_EXPORT void init_dense_vector(sycl::queue &queue, dense_vector_handle_t *p_dvhandle, std::int64_t size, dataType *val); template -ONEMKL_EXPORT void set_dense_vector_data(sycl::queue &queue, dense_vector_handle_t dvhandle, +ONEMATH_EXPORT void set_dense_vector_data(sycl::queue &queue, dense_vector_handle_t dvhandle, std::int64_t size, sycl::buffer val); template -ONEMKL_EXPORT void set_dense_vector_data(sycl::queue &queue, dense_vector_handle_t dvhandle, +ONEMATH_EXPORT void set_dense_vector_data(sycl::queue &queue, dense_vector_handle_t dvhandle, std::int64_t size, dataType *val); -ONEMKL_EXPORT sycl::event release_dense_vector(sycl::queue &queue, dense_vector_handle_t dvhandle, +ONEMATH_EXPORT sycl::event release_dense_vector(sycl::queue &queue, dense_vector_handle_t dvhandle, const std::vector &dependencies = {}); // Dense matrix template -ONEMKL_EXPORT void init_dense_matrix(sycl::queue &queue, dense_matrix_handle_t *p_dmhandle, +ONEMATH_EXPORT void init_dense_matrix(sycl::queue &queue, dense_matrix_handle_t *p_dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, layout dense_layout, sycl::buffer val); template -ONEMKL_EXPORT void init_dense_matrix(sycl::queue &queue, dense_matrix_handle_t *p_dmhandle, +ONEMATH_EXPORT void init_dense_matrix(sycl::queue &queue, dense_matrix_handle_t *p_dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, layout dense_layout, dataType *val); template -ONEMKL_EXPORT void set_dense_matrix_data(sycl::queue &queue, dense_matrix_handle_t dmhandle, +ONEMATH_EXPORT void set_dense_matrix_data(sycl::queue &queue, dense_matrix_handle_t dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, layout dense_layout, sycl::buffer val); template -ONEMKL_EXPORT void set_dense_matrix_data(sycl::queue &queue, dense_matrix_handle_t dmhandle, +ONEMATH_EXPORT void set_dense_matrix_data(sycl::queue &queue, dense_matrix_handle_t dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, layout dense_layout, dataType *val); -ONEMKL_EXPORT sycl::event release_dense_matrix(sycl::queue &queue, dense_matrix_handle_t dmhandle, +ONEMATH_EXPORT sycl::event release_dense_matrix(sycl::queue &queue, dense_matrix_handle_t dmhandle, const std::vector &dependencies = {}); // COO matrix template -ONEMKL_EXPORT void init_coo_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, +ONEMATH_EXPORT void init_coo_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, sycl::buffer row_ind, sycl::buffer col_ind, sycl::buffer val); template -ONEMKL_EXPORT void init_coo_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, +ONEMATH_EXPORT void init_coo_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, indexType *row_ind, indexType *col_ind, dataType *val); template -ONEMKL_EXPORT void set_coo_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, +ONEMATH_EXPORT void set_coo_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, sycl::buffer row_ind, sycl::buffer col_ind, sycl::buffer val); template -ONEMKL_EXPORT void set_coo_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, +ONEMATH_EXPORT void set_coo_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, indexType *row_ind, indexType *col_ind, dataType *val); // CSR matrix template -ONEMKL_EXPORT void init_csr_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, +ONEMATH_EXPORT void init_csr_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, sycl::buffer row_ptr, sycl::buffer col_ind, sycl::buffer val); template -ONEMKL_EXPORT void init_csr_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, +ONEMATH_EXPORT void init_csr_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, indexType *row_ptr, indexType *col_ind, dataType *val); template -ONEMKL_EXPORT void set_csr_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, +ONEMATH_EXPORT void set_csr_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, sycl::buffer row_ptr, sycl::buffer col_ind, sycl::buffer val); template -ONEMKL_EXPORT void set_csr_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, +ONEMATH_EXPORT void set_csr_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, indexType *row_ptr, indexType *col_ind, dataType *val); // Common sparse matrix functions -ONEMKL_EXPORT sycl::event release_sparse_matrix(sycl::queue &queue, matrix_handle_t smhandle, +ONEMATH_EXPORT sycl::event release_sparse_matrix(sycl::queue &queue, matrix_handle_t smhandle, const std::vector &dependencies = {}); bool set_matrix_property(sycl::queue &queue, matrix_handle_t smhandle, matrix_property property); // SPMM -ONEMKL_EXPORT void init_spmm_descr(sycl::queue &queue, spmm_descr_t *p_spmm_descr); +ONEMATH_EXPORT void init_spmm_descr(sycl::queue &queue, spmm_descr_t *p_spmm_descr); -ONEMKL_EXPORT sycl::event release_spmm_descr(sycl::queue &queue, spmm_descr_t spmm_descr, +ONEMATH_EXPORT sycl::event release_spmm_descr(sycl::queue &queue, spmm_descr_t spmm_descr, const std::vector &dependencies = {}); -ONEMKL_EXPORT void spmm_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT void spmm_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, std::size_t &temp_buffer_size); -ONEMKL_EXPORT void spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT void spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, sycl::buffer workspace); -ONEMKL_EXPORT sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, @@ -146,7 +146,7 @@ ONEMKL_EXPORT sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpo spmm_descr_t spmm_descr, void *workspace, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spmm(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT sycl::event spmm(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, @@ -154,31 +154,31 @@ ONEMKL_EXPORT sycl::event spmm(sycl::queue &queue, oneapi::mkl::transpose opA, const std::vector &dependencies = {}); // SPMV -ONEMKL_EXPORT void init_spmv_descr(sycl::queue &queue, spmv_descr_t *p_spmv_descr); +ONEMATH_EXPORT void init_spmv_descr(sycl::queue &queue, spmv_descr_t *p_spmv_descr); -ONEMKL_EXPORT sycl::event release_spmv_descr(sycl::queue &queue, spmv_descr_t spmv_descr, +ONEMATH_EXPORT sycl::event release_spmv_descr(sycl::queue &queue, spmv_descr_t spmv_descr, const std::vector &dependencies = {}); -ONEMKL_EXPORT void spmv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT void spmv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, std::size_t &temp_buffer_size); -ONEMKL_EXPORT void spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +ONEMATH_EXPORT void spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, sycl::buffer workspace); -ONEMKL_EXPORT sycl::event spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT sycl::event spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, void *workspace, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, @@ -186,31 +186,31 @@ ONEMKL_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::mkl::transpose opA, c const std::vector &dependencies = {}); // SPSV -ONEMKL_EXPORT void init_spsv_descr(sycl::queue &queue, spsv_descr_t *p_spsv_descr); +ONEMATH_EXPORT void init_spsv_descr(sycl::queue &queue, spsv_descr_t *p_spsv_descr); -ONEMKL_EXPORT sycl::event release_spsv_descr(sycl::queue &queue, spsv_descr_t spsv_descr, +ONEMATH_EXPORT sycl::event release_spsv_descr(sycl::queue &queue, spsv_descr_t spsv_descr, const std::vector &dependencies = {}); -ONEMKL_EXPORT void spsv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT void spsv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, std::size_t &temp_buffer_size); -ONEMKL_EXPORT void spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +ONEMATH_EXPORT void spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, sycl::buffer workspace); -ONEMKL_EXPORT sycl::event spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT sycl::event spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, void *workspace, const std::vector &dependencies = {}); -ONEMKL_EXPORT sycl::event spsv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +ONEMATH_EXPORT sycl::event spsv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, diff --git a/include/oneapi/math/sparse_blas/detail/operation_types.hpp b/include/oneapi/math/sparse_blas/detail/operation_types.hpp index b79036830..51611070a 100644 --- a/include/oneapi/math/sparse_blas/detail/operation_types.hpp +++ b/include/oneapi/math/sparse_blas/detail/operation_types.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_DETAIL_OPERATION_TYPES_HPP_ -#define _ONEMKL_SPARSE_BLAS_DETAIL_OPERATION_TYPES_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_DETAIL_OPERATION_TYPES_HPP_ +#define _ONEMATH_SPARSE_BLAS_DETAIL_OPERATION_TYPES_HPP_ namespace oneapi::mkl::sparse { @@ -35,4 +35,4 @@ using spsv_descr_t = spsv_descr*; } // namespace oneapi::mkl::sparse -#endif // _ONEMKL_SPARSE_BLAS_DETAIL_OPERATION_TYPES_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_DETAIL_OPERATION_TYPES_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp b/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp index 92e715dc6..d220979ce 100644 --- a/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp +++ b/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_DETAIL_SPARSE_BLAS_RT_HPP_ -#define _ONEMKL_SPARSE_BLAS_DETAIL_SPARSE_BLAS_RT_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_DETAIL_SPARSE_BLAS_RT_HPP_ +#define _ONEMATH_SPARSE_BLAS_DETAIL_SPARSE_BLAS_RT_HPP_ #include "oneapi/math/sparse_blas/detail/helper_types.hpp" #include "oneapi/math/sparse_blas/types.hpp" @@ -205,4 +205,4 @@ sycl::event spsv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alp } // namespace mkl } // namespace oneapi -#endif // _ONEMKL_SPARSE_BLAS_DETAIL_SPARSE_BLAS_RT_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_DETAIL_SPARSE_BLAS_RT_HPP_ diff --git a/include/oneapi/math/sparse_blas/matrix_view.hpp b/include/oneapi/math/sparse_blas/matrix_view.hpp index 196bd0271..8625b2ab9 100644 --- a/include/oneapi/math/sparse_blas/matrix_view.hpp +++ b/include/oneapi/math/sparse_blas/matrix_view.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_MATRIX_VIEW_HPP_ -#define _ONEMKL_SPARSE_BLAS_MATRIX_VIEW_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_MATRIX_VIEW_HPP_ +#define _ONEMATH_SPARSE_BLAS_MATRIX_VIEW_HPP_ #include "oneapi/math/types.hpp" @@ -48,4 +48,4 @@ struct matrix_view { } // namespace mkl } // namespace oneapi -#endif // _ONEMKL_SPARSE_BLAS_MATRIX_VIEW_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_MATRIX_VIEW_HPP_ diff --git a/include/oneapi/math/sparse_blas/types.hpp b/include/oneapi/math/sparse_blas/types.hpp index da4e2340a..986d447c9 100644 --- a/include/oneapi/math/sparse_blas/types.hpp +++ b/include/oneapi/math/sparse_blas/types.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_TYPES_HPP_ -#define _ONEMKL_SPARSE_BLAS_TYPES_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_TYPES_HPP_ +#define _ONEMATH_SPARSE_BLAS_TYPES_HPP_ #include "oneapi/math/types.hpp" #include "matrix_view.hpp" @@ -69,4 +69,4 @@ enum class spsv_alg { } // namespace mkl } // namespace oneapi -#endif // _ONEMKL_SPARSE_BLAS_TYPES_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_TYPES_HPP_ diff --git a/include/oneapi/math/types.hpp b/include/oneapi/math/types.hpp index 109bd6afe..bb58ff647 100644 --- a/include/oneapi/math/types.hpp +++ b/include/oneapi/math/types.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_TYPES_HPP_ -#define _ONEMKL_TYPES_HPP_ +#ifndef _ONEMATH_TYPES_HPP_ +#define _ONEMATH_TYPES_HPP_ #ifdef __HIPSYCL__ #include "oneapi/math/bfloat16.hpp" @@ -119,4 +119,4 @@ enum class order : char { } //namespace mkl } //namespace oneapi -#endif //_ONEMKL_TYPES_HPP_ +#endif //_ONEMATH_TYPES_HPP_ diff --git a/scripts/generate_cmake.py b/scripts/generate_cmake.py index c245bc0a8..17bc3a279 100644 --- a/scripts/generate_cmake.py +++ b/scripts/generate_cmake.py @@ -84,7 +84,7 @@ def usage(err = None): for f in file_list: if re.search('_dyn.c', f): - out_file.write(""" $<$: {filename}> + out_file.write(""" $<$: {filename}> """.format(filename=f)) else: out_file.write(""" {filename} @@ -99,7 +99,7 @@ def usage(err = None): ) target_link_libraries(${{LIB_OBJ}} - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL # Add third party library to link with here ) diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index d5cce0a2e..0b59cd968 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -18,16 +18,16 @@ #=============================================================================== # Define common build flags for oneMKL libraries -set(ONEMKL_BUILD_COPT "") +set(ONEMATH_BUILD_COPT "") if(WIN32 AND BUILD_SHARED_LIBS) - list(APPEND ONEMKL_BUILD_COPT "-Donemath_EXPORTS") + list(APPEND ONEMATH_BUILD_COPT "-Donemath_EXPORTS") endif() # store path to CMAKE_CURRENT_BINARY_DIR to use it later (makes FetchContent_Declare workable) -set(ONEMKL_GENERATED_INCLUDE_PATH ${CMAKE_CURRENT_BINARY_DIR}) +set(ONEMATH_GENERATED_INCLUDE_PATH ${CMAKE_CURRENT_BINARY_DIR}) -set(ONEMKL_INTERFACE_INCLUDE_DIRS +set(ONEMATH_INTERFACE_INCLUDE_DIRS $ $ $ @@ -45,20 +45,20 @@ function(generate_header_file) # Following if-conditions allow to decouple cmake configuration variables with # the corresponding generated macro. This is done to be conformant with ES.33 # C++ Core Guidelines - set(ONEMKL_BUILD_SHARED_LIBS ${BUILD_SHARED_LIBS}) - set(ONEMKL_ENABLE_MKLCPU_BACKEND ${ENABLE_MKLCPU_BACKEND}) - set(ONEMKL_ENABLE_MKLGPU_BACKEND ${ENABLE_MKLGPU_BACKEND}) - set(ONEMKL_ENABLE_CUBLAS_BACKEND ${ENABLE_CUBLAS_BACKEND}) - set(ONEMKL_ENABLE_ROCBLAS_BACKEND ${ENABLE_ROCBLAS_BACKEND}) - set(ONEMKL_ENABLE_NETLIB_BACKEND ${ENABLE_NETLIB_BACKEND}) - set(ONEMKL_ENABLE_PORTBLAS_BACKEND ${ENABLE_PORTBLAS_BACKEND}) - set(ONEMKL_ENABLE_CURAND_BACKEND ${ENABLE_CURAND_BACKEND}) - set(ONEMKL_ENABLE_ROCRAND_BACKEND ${ENABLE_ROCRAND_BACKEND}) - set(ONEMKL_ENABLE_CUSOLVER_BACKEND ${ENABLE_CUSOLVER_BACKEND}) - set(ONEMKL_ENABLE_ROCSOLVER_BACKEND ${ENABLE_ROCSOLVER_BACKEND}) - set(ONEMKL_ENABLE_CUFFT_BACKEND ${ENABLE_CUFFT_BACKEND}) - set(ONEMKL_ENABLE_ROCFFT_BACKEND ${ENABLE_ROCFFT_BACKEND}) - set(ONEMKL_ENABLE_PORTFFT_BACKEND ${ENABLE_PORTFFT_BACKEND}) + set(ONEMATH_BUILD_SHARED_LIBS ${BUILD_SHARED_LIBS}) + set(ONEMATH_ENABLE_MKLCPU_BACKEND ${ENABLE_MKLCPU_BACKEND}) + set(ONEMATH_ENABLE_MKLGPU_BACKEND ${ENABLE_MKLGPU_BACKEND}) + set(ONEMATH_ENABLE_CUBLAS_BACKEND ${ENABLE_CUBLAS_BACKEND}) + set(ONEMATH_ENABLE_ROCBLAS_BACKEND ${ENABLE_ROCBLAS_BACKEND}) + set(ONEMATH_ENABLE_NETLIB_BACKEND ${ENABLE_NETLIB_BACKEND}) + set(ONEMATH_ENABLE_PORTBLAS_BACKEND ${ENABLE_PORTBLAS_BACKEND}) + set(ONEMATH_ENABLE_CURAND_BACKEND ${ENABLE_CURAND_BACKEND}) + set(ONEMATH_ENABLE_ROCRAND_BACKEND ${ENABLE_ROCRAND_BACKEND}) + set(ONEMATH_ENABLE_CUSOLVER_BACKEND ${ENABLE_CUSOLVER_BACKEND}) + set(ONEMATH_ENABLE_ROCSOLVER_BACKEND ${ENABLE_ROCSOLVER_BACKEND}) + set(ONEMATH_ENABLE_CUFFT_BACKEND ${ENABLE_CUFFT_BACKEND}) + set(ONEMATH_ENABLE_ROCFFT_BACKEND ${ENABLE_ROCFFT_BACKEND}) + set(ONEMATH_ENABLE_PORTFFT_BACKEND ${ENABLE_PORTFFT_BACKEND}) configure_file(config.hpp.in "${CMAKE_CURRENT_BINARY_DIR}/oneapi/math/config.hpp.configured") file(GENERATE @@ -84,7 +84,7 @@ if(BUILD_SHARED_LIBS) add_dependencies(onemath onemath_backend_libs) target_include_directories(onemath - PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} ) set_target_properties(onemath PROPERTIES SOVERSION ${PROJECT_VERSION_MAJOR} @@ -95,16 +95,16 @@ if(BUILD_SHARED_LIBS) endif() # Build dispatcher library - set (ONEMKL_LIBS ${TARGET_DOMAINS}) - list(TRANSFORM ONEMKL_LIBS PREPEND onemath_) - target_link_libraries(onemath PUBLIC ${ONEMKL_LIBS}) + set (ONEMATH_LIBS ${TARGET_DOMAINS}) + list(TRANSFORM ONEMATH_LIBS PREPEND onemath_) + target_link_libraries(onemath PUBLIC ${ONEMATH_LIBS}) set_target_properties(onemath PROPERTIES INSTALL_RPATH "\$ORIGIN" BUILD_WITH_INSTALL_RPATH TRUE ) # Add the library to install package - foreach(domain_lib ${ONEMKL_LIBS}) + foreach(domain_lib ${ONEMATH_LIBS}) install(TARGETS ${domain_lib} EXPORT oneMKLTargets) endforeach() install(TARGETS onemath EXPORT oneMKLTargets diff --git a/src/blas/CMakeLists.txt b/src/blas/CMakeLists.txt index e40de2ac1..fe0499f1e 100644 --- a/src/blas/CMakeLists.txt +++ b/src/blas/CMakeLists.txt @@ -29,11 +29,11 @@ target_include_directories(onemath_blas ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} $ ) -target_compile_options(onemath_blas PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(onemath_blas PRIVATE ${ONEMATH_BUILD_COPT}) set_target_properties(onemath_blas PROPERTIES POSITION_INDEPENDENT_CODE ON @@ -41,7 +41,7 @@ set_target_properties(onemath_blas PROPERTIES if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET onemath_blas SOURCES blas_loader.cpp) else() - target_link_libraries(onemath_blas PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(onemath_blas PUBLIC ONEMATH::SYCL::SYCL) endif() endif() diff --git a/src/blas/backends/cublas/CMakeLists.txt b/src/blas/backends/cublas/CMakeLists.txt index 63b061fe3..d9270c124 100644 --- a/src/blas/backends/cublas/CMakeLists.txt +++ b/src/blas/backends/cublas/CMakeLists.txt @@ -25,8 +25,8 @@ set(SOURCES cublas_level1.cpp cublas_level3.cpp cublas_batch.cpp cublas_extensions.cpp - $<$:cublas_scope_handle.cpp > - $<$:cublas_scope_handle_hipsycl.cpp > + $<$:cublas_scope_handle.cpp > + $<$:cublas_scope_handle_hipsycl.cpp > $<$: cublas_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) @@ -36,17 +36,17 @@ target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src/include ${PROJECT_SOURCE_DIR}/src - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) -if(NOT ${ONEMKL_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") - target_compile_options(ONEMKL::SYCL::SYCL INTERFACE +if(NOT ${ONEMATH_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") + target_compile_options(ONEMATH::SYCL::SYCL INTERFACE -fsycl-targets=nvptx64-nvidia-cuda -fsycl-unnamed-lambda) - target_link_options(ONEMKL::SYCL::SYCL INTERFACE + target_link_options(ONEMATH::SYCL::SYCL INTERFACE -fsycl-targets=nvptx64-nvidia-cuda) endif() -target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL ONEMKL::cuBLAS::cuBLAS) +target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL ONEMATH::cuBLAS::cuBLAS) target_compile_features(${LIB_OBJ} PUBLIC cxx_std_11) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON) diff --git a/src/blas/backends/cublas/cublas_scope_handle.cpp b/src/blas/backends/cublas/cublas_scope_handle.cpp index dd12552fb..947cd1da1 100644 --- a/src/blas/backends/cublas/cublas_scope_handle.cpp +++ b/src/blas/backends/cublas/cublas_scope_handle.cpp @@ -35,7 +35,7 @@ namespace cublas { * takes place if no other element in the container has a key equivalent to * the one being emplaced (keys in a map container are unique). */ -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED thread_local cublas_handle CublasScopedContextHandler::handle_helper = cublas_handle{}; #else @@ -97,7 +97,7 @@ cublasHandle_t CublasScopedContextHandler::get_handle(const sycl::queue &queue) CUresult cuErr; CUcontext desired; CUDA_ERROR_FUNC(cuDevicePrimaryCtxRetain, cuErr, &desired, cudaDevice); -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED auto piPlacedContext_ = reinterpret_cast(desired); #else auto piPlacedContext_ = reinterpret_cast(desired); diff --git a/src/blas/backends/cublas/cublas_scope_handle.hpp b/src/blas/backends/cublas/cublas_scope_handle.hpp index d88124ac1..27bcfd495 100644 --- a/src/blas/backends/cublas/cublas_scope_handle.hpp +++ b/src/blas/backends/cublas/cublas_scope_handle.hpp @@ -36,8 +36,8 @@ // After Plugin Interface removal in DPC++ ur.hpp is the new include #if __has_include() #include -#ifndef ONEMKL_PI_INTERFACE_REMOVED -#define ONEMKL_PI_INTERFACE_REMOVED +#ifndef ONEMATH_PI_INTERFACE_REMOVED +#define ONEMATH_PI_INTERFACE_REMOVED #endif #elif __has_include() #include @@ -88,7 +88,7 @@ class CublasScopedContextHandler { sycl::context *placedContext_; bool needToRecover_; sycl::interop_handle &ih; -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED static thread_local cublas_handle handle_helper; #else static thread_local cublas_handle handle_helper; diff --git a/src/blas/backends/cublas/cublas_task.hpp b/src/blas/backends/cublas/cublas_task.hpp index ee9b84e8d..63afea9c9 100644 --- a/src/blas/backends/cublas/cublas_task.hpp +++ b/src/blas/backends/cublas/cublas_task.hpp @@ -38,8 +38,8 @@ // After Plugin Interface removal in DPC++ ur.hpp is the new include #if __has_include() #include -#ifndef ONEMKL_PI_INTERFACE_REMOVED -#define ONEMKL_PI_INTERFACE_REMOVED +#ifndef ONEMATH_PI_INTERFACE_REMOVED +#define ONEMATH_PI_INTERFACE_REMOVED #endif #elif __has_include() #include diff --git a/src/blas/backends/mkl_common/mkl_blas_backend.hxx b/src/blas/backends/mkl_common/mkl_blas_backend.hxx index 10e441bd7..82588a66a 100644 --- a/src/blas/backends/mkl_common/mkl_blas_backend.hxx +++ b/src/blas/backends/mkl_common/mkl_blas_backend.hxx @@ -1361,22 +1361,22 @@ sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, value_or_ sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, value_or_pointer y1, double *param, const std::vector &dependencies = {}); -#define ONEMKL_DECLARE_SCAL(T, Ts) \ +#define ONEMATH_DECLARE_SCAL(T, Ts) \ sycl::event scal(sycl::queue &queue, std::int64_t n, value_or_pointer alpha, T *x, \ std::int64_t incx, const std::vector &dependencies = {}); -ONEMKL_DECLARE_SCAL(float, float) -ONEMKL_DECLARE_SCAL(double, double) -ONEMKL_DECLARE_SCAL(std::complex, std::complex) -ONEMKL_DECLARE_SCAL(std::complex, std::complex) -ONEMKL_DECLARE_SCAL(std::complex, float) -ONEMKL_DECLARE_SCAL(std::complex, double) +ONEMATH_DECLARE_SCAL(float, float) +ONEMATH_DECLARE_SCAL(double, double) +ONEMATH_DECLARE_SCAL(std::complex, std::complex) +ONEMATH_DECLARE_SCAL(std::complex, std::complex) +ONEMATH_DECLARE_SCAL(std::complex, float) +ONEMATH_DECLARE_SCAL(std::complex, double) sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -#undef ONEMKL_DECLARE_SCAL +#undef ONEMATH_DECLARE_SCAL sycl::event swap(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies = {}); diff --git a/src/blas/backends/mklcpu/CMakeLists.txt b/src/blas/backends/mklcpu/CMakeLists.txt index cfb1481b5..05e4e4a3a 100644 --- a/src/blas/backends/mklcpu/CMakeLists.txt +++ b/src/blas/backends/mklcpu/CMakeLists.txt @@ -35,15 +35,15 @@ target_include_directories(${LIB_OBJ} ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::BLAS) - target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL MKL::MKL_SYCL::BLAS) + target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_SYCL::BLAS) else() - target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL MKL::MKL_DPCPP) + target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_DPCPP) endif() set_target_properties(${LIB_OBJ} PROPERTIES @@ -54,7 +54,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) #Set libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/blas/backends/mklcpu/mklcpu_wrappers.cpp b/src/blas/backends/mklcpu/mklcpu_wrappers.cpp index 21dd03534..6b5caea4e 100644 --- a/src/blas/backends/mklcpu/mklcpu_wrappers.cpp +++ b/src/blas/backends/mklcpu/mklcpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT blas_function_table_t mkl_blas_table = { +extern "C" ONEMATH_EXPORT blas_function_table_t mkl_blas_table = { WRAPPER_VERSION, #define BACKEND mklcpu #define MAJOR column_major diff --git a/src/blas/backends/mklgpu/CMakeLists.txt b/src/blas/backends/mklgpu/CMakeLists.txt index d197f19d8..a5cc7abf3 100644 --- a/src/blas/backends/mklgpu/CMakeLists.txt +++ b/src/blas/backends/mklgpu/CMakeLists.txt @@ -32,15 +32,15 @@ target_include_directories(${LIB_OBJ} ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::BLAS) - target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL MKL::MKL_SYCL::BLAS) + target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_SYCL::BLAS) else() - target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL MKL::MKL_DPCPP) + target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_DPCPP) endif() set_target_properties(${LIB_OBJ} PROPERTIES @@ -51,7 +51,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) #Set libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/blas/backends/mklgpu/mklgpu_wrappers.cpp b/src/blas/backends/mklgpu/mklgpu_wrappers.cpp index 8ce5be8df..6328196b8 100644 --- a/src/blas/backends/mklgpu/mklgpu_wrappers.cpp +++ b/src/blas/backends/mklgpu/mklgpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT blas_function_table_t mkl_blas_table = { +extern "C" ONEMATH_EXPORT blas_function_table_t mkl_blas_table = { WRAPPER_VERSION, #define BACKEND mklgpu #define MAJOR column_major diff --git a/src/blas/backends/netlib/CMakeLists.txt b/src/blas/backends/netlib/CMakeLists.txt index 5a4bb76f2..6ecab7591 100644 --- a/src/blas/backends/netlib/CMakeLists.txt +++ b/src/blas/backends/netlib/CMakeLists.txt @@ -41,12 +41,12 @@ target_include_directories(${LIB_OBJ} ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${NETLIB_INCLUDE} - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) -target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL ${NETLIB_LINK}) +target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL ${NETLIB_LINK}) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON @@ -55,7 +55,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/blas/backends/netlib/netlib_wrappers.cpp b/src/blas/backends/netlib/netlib_wrappers.cpp index 08de7eb39..6d067e904 100644 --- a/src/blas/backends/netlib/netlib_wrappers.cpp +++ b/src/blas/backends/netlib/netlib_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT blas_function_table_t mkl_blas_table = { +extern "C" ONEMATH_EXPORT blas_function_table_t mkl_blas_table = { WRAPPER_VERSION, #define BACKEND netlib #define MAJOR column_major diff --git a/src/blas/backends/portblas/CMakeLists.txt b/src/blas/backends/portblas/CMakeLists.txt index f1f0cef8b..0809ceed3 100644 --- a/src/blas/backends/portblas/CMakeLists.txt +++ b/src/blas/backends/portblas/CMakeLists.txt @@ -36,10 +36,10 @@ endfunction(get_sycl_targets) # portBLAS supports tuning for some device types, but can only be compiled # for one at a time currently. Work out which device to tune for based on the # DPC++ target triple specified via -fsycl-targets -if(TARGET ONEMKL::SYCL::SYCL) - get_target_property(ONEMKL_COMPILE_OPTIONS ONEMKL::SYCL::SYCL INTERFACE_COMPILE_OPTIONS) +if(TARGET ONEMATH::SYCL::SYCL) + get_target_property(ONEMATH_COMPILE_OPTIONS ONEMATH::SYCL::SYCL INTERFACE_COMPILE_OPTIONS) endif() -get_sycl_targets("${ONEMKL_COMPILE_OPTIONS}") +get_sycl_targets("${ONEMATH_COMPILE_OPTIONS}") list(LENGTH SYCL_TARGETS NUM_TARGETS) if(NUM_TARGETS EQUAL 0) get_sycl_targets("${CMAKE_CXX_FLAGS}") @@ -50,21 +50,21 @@ if(PORTBLAS_TUNING_TARGET) # Allow the user to manually enable a specific device type # for tuned portBLAS configurations and sets sycl-target. if(PORTBLAS_TUNING_TARGET STREQUAL "INTEL_CPU") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_CPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_CPU "ON" CACHE INTERNAL "") set(PORTBLAS_TUNING_TARGET "") - target_compile_options(ONEMKL::SYCL::SYCL INTERFACE + target_compile_options(ONEMATH::SYCL::SYCL INTERFACE -fsycl-targets=spir64_x86_64 -fsycl-unnamed-lambda) - target_link_options(ONEMKL::SYCL::SYCL INTERFACE + target_link_options(ONEMATH::SYCL::SYCL INTERFACE -fsycl-targets=spir64_x86_64) elseif(PORTBLAS_TUNING_TARGET STREQUAL "INTEL_GPU") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_GPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_GPU "ON" CACHE INTERNAL "") elseif(PORTBLAS_TUNING_TARGET STREQUAL "AMD_GPU") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_AMD_GPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_AMD_GPU "ON" CACHE INTERNAL "") if (is_dpcpp) - target_compile_options(ONEMKL::SYCL::SYCL INTERFACE + target_compile_options(ONEMATH::SYCL::SYCL INTERFACE -fsycl-targets=amdgcn-amd-amdhsa -fsycl-unnamed-lambda -Xsycl-target-backend --offload-arch=${HIP_TARGETS}) - target_link_options(ONEMKL::SYCL::SYCL INTERFACE + target_link_options(ONEMATH::SYCL::SYCL INTERFACE -fsycl-targets=amdgcn-amd-amdhsa -Xsycl-target-backend --offload-arch=${HIP_TARGETS}) else() message(WARNING "Compiler is not supported." @@ -72,16 +72,16 @@ if(PORTBLAS_TUNING_TARGET) " Compilation may fail.") endif() elseif(PORTBLAS_TUNING_TARGET STREQUAL "NVIDIA_GPU") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU "ON" CACHE INTERNAL "") if (is_dpcpp) - target_compile_options(ONEMKL::SYCL::SYCL INTERFACE + target_compile_options(ONEMATH::SYCL::SYCL INTERFACE -fsycl-targets=nvptx64-nvidia-cuda -fsycl-unnamed-lambda) - target_link_options(ONEMKL::SYCL::SYCL INTERFACE + target_link_options(ONEMATH::SYCL::SYCL INTERFACE -fsycl-targets=nvptx64-nvidia-cuda) if(DEFINED CUDA_TARGET) - target_compile_options(ONEMKL::SYCL::SYCL INTERFACE + target_compile_options(ONEMATH::SYCL::SYCL INTERFACE -Xsycl-target-backend --cuda-gpu-arch=${CUDA_TARGET}) - target_link_options(ONEMKL::SYCL::SYCL INTERFACE + target_link_options(ONEMATH::SYCL::SYCL INTERFACE -Xsycl-target-backend --cuda-gpu-arch=${CUDA_TARGET}) endif() else() @@ -94,27 +94,27 @@ if(PORTBLAS_TUNING_TARGET) endif() elseif(NUM_TARGETS EQUAL 0) # Enable portBLAS backend for all devices types - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_CPU "ON" CACHE INTERNAL "") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_GPU "ON" CACHE INTERNAL "") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_AMD_GPU "ON" CACHE INTERNAL "") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_CPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_GPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_AMD_GPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU "ON" CACHE INTERNAL "") else() # Try to automatically detect the PORTBLAS_TUNING_TARGET foreach(SYCL_TARGET IN LISTS SYCL_TARGETS) if(SYCL_TARGETS MATCHES "^intel_gpu" OR SYCL_TARGETS MATCHES "^spir64_gen") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_GPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_GPU "ON" CACHE INTERNAL "") set(PORTBLAS_TUNING_TARGET "INTEL_GPU") elseif(SYCL_TARGETS MATCHES "^spir64_x86_64") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_CPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_CPU "ON" CACHE INTERNAL "") elseif(SYCL_TARGETS MATCHES "^spir64") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_CPU "ON" CACHE INTERNAL "") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_GPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_CPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_GPU "ON" CACHE INTERNAL "") set(PORTBLAS_TUNING_TARGET "INTEL_GPU") elseif(SYCL_TARGETS MATCHES "^amd_gpu" OR SYCL_TARGETS MATCHES "-amd-") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_AMD_GPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_AMD_GPU "ON" CACHE INTERNAL "") set(PORTBLAS_TUNING_TARGET "AMD_GPU") elseif(SYCL_TARGETS MATCHES "^nvidia_gpu" OR SYCL_TARGETS MATCHES "-nvidia-") - set(ONEMKL_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU "ON" CACHE INTERNAL "") + set(ONEMATH_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU "ON" CACHE INTERNAL "") set(PORTBLAS_TUNING_TARGET "NVIDIA_GPU") endif() endforeach() @@ -188,11 +188,11 @@ target_include_directories(${LIB_OBJ} ${PROJECT_SOURCE_DIR}/src/include ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) -target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL portblas) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) +target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL portblas) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON) @@ -201,7 +201,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/blas/backends/portblas/portblas_wrappers.cpp b/src/blas/backends/portblas/portblas_wrappers.cpp index 69ec16016..bf9d2591b 100644 --- a/src/blas/backends/portblas/portblas_wrappers.cpp +++ b/src/blas/backends/portblas/portblas_wrappers.cpp @@ -8,7 +8,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT blas_function_table_t mkl_blas_table = { +extern "C" ONEMATH_EXPORT blas_function_table_t mkl_blas_table = { WRAPPER_VERSION, #define BACKEND portblas #define MAJOR column_major diff --git a/src/blas/backends/rocblas/CMakeLists.txt b/src/blas/backends/rocblas/CMakeLists.txt index f8d42b312..6fc87fc5d 100644 --- a/src/blas/backends/rocblas/CMakeLists.txt +++ b/src/blas/backends/rocblas/CMakeLists.txt @@ -30,8 +30,8 @@ set(SOURCES rocblas_level1.cpp rocblas_level3.cpp rocblas_batch.cpp rocblas_extensions.cpp - $<$:rocblas_scope_handle.cpp > - $<$:rocblas_scope_handle_hipsycl.cpp > + $<$:rocblas_scope_handle.cpp > + $<$:rocblas_scope_handle_hipsycl.cpp > $<$: rocblas_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) @@ -42,25 +42,25 @@ target_include_directories(${LIB_OBJ} ${PROJECT_SOURCE_DIR}/src/include ${PROJECT_SOURCE_DIR}/src ${PROJECT_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -if(NOT ${ONEMKL_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") - target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) - target_compile_options(ONEMKL::SYCL::SYCL INTERFACE +if(NOT ${ONEMATH_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") + target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) + target_compile_options(ONEMATH::SYCL::SYCL INTERFACE -fsycl-targets=amdgcn-amd-amdhsa -fsycl-unnamed-lambda -Xsycl-target-backend --offload-arch=${HIP_TARGETS}) - target_link_options(ONEMKL::SYCL::SYCL INTERFACE + target_link_options(ONEMATH::SYCL::SYCL INTERFACE -fsycl-targets=amdgcn-amd-amdhsa -Xsycl-target-backend --offload-arch=${HIP_TARGETS}) else() - target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) - target_compile_options(ONEMKL::SYCL::SYCL INTERFACE) - target_link_options(ONEMKL::SYCL::SYCL INTERFACE) + target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) + target_compile_options(ONEMATH::SYCL::SYCL INTERFACE) + target_link_options(ONEMATH::SYCL::SYCL INTERFACE) endif() target_link_libraries(${LIB_OBJ} PRIVATE roc::rocblas hip::host Threads::Threads) -target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL) +target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL) target_compile_features(${LIB_OBJ} PUBLIC cxx_std_17) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON) diff --git a/src/blas/backends/rocblas/rocblas_scope_handle.cpp b/src/blas/backends/rocblas/rocblas_scope_handle.cpp index 8f48aed3d..ecec4dded 100644 --- a/src/blas/backends/rocblas/rocblas_scope_handle.cpp +++ b/src/blas/backends/rocblas/rocblas_scope_handle.cpp @@ -50,7 +50,7 @@ rocblas_handle_container::~rocblas_handle_container() noexcept(false) { * takes place if no other element in the container has a key equivalent to * the one being emplaced (keys in a map container are unique). */ -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED thread_local rocblas_handle_container RocblasScopedContextHandler::handle_helper = rocblas_handle_container{}; #else @@ -113,7 +113,7 @@ rocblas_handle RocblasScopedContextHandler::get_handle(const sycl::queue &queue) hipError_t hipErr; hipCtx_t desired; HIP_ERROR_FUNC(hipDevicePrimaryCtxRetain, hipErr, &desired, hipDevice); -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED auto piPlacedContext_ = reinterpret_cast(desired); #else auto piPlacedContext_ = reinterpret_cast(desired); diff --git a/src/blas/backends/rocblas/rocblas_scope_handle.hpp b/src/blas/backends/rocblas/rocblas_scope_handle.hpp index 76c849ef8..50f650299 100644 --- a/src/blas/backends/rocblas/rocblas_scope_handle.hpp +++ b/src/blas/backends/rocblas/rocblas_scope_handle.hpp @@ -29,8 +29,8 @@ // After Plugin Interface removal in DPC++ ur.hpp is the new include #if __has_include() #include -#ifndef ONEMKL_PI_INTERFACE_REMOVED -#define ONEMKL_PI_INTERFACE_REMOVED +#ifndef ONEMATH_PI_INTERFACE_REMOVED +#define ONEMATH_PI_INTERFACE_REMOVED #endif #elif __has_include() #include @@ -55,7 +55,7 @@ class RocblasScopedContextHandler { sycl::context *placedContext_; bool needToRecover_; sycl::interop_handle &interop_h; -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED static thread_local rocblas_handle_container handle_helper; #else static thread_local rocblas_handle_container handle_helper; diff --git a/src/blas/backends/rocblas/rocblas_task.hpp b/src/blas/backends/rocblas/rocblas_task.hpp index 588c9029b..a74063ad8 100644 --- a/src/blas/backends/rocblas/rocblas_task.hpp +++ b/src/blas/backends/rocblas/rocblas_task.hpp @@ -37,8 +37,8 @@ // After Plugin Interface removal in DPC++ ur.hpp is the new include #if __has_include() #include -#ifndef ONEMKL_PI_INTERFACE_REMOVED -#define ONEMKL_PI_INTERFACE_REMOVED +#ifndef ONEMATH_PI_INTERFACE_REMOVED +#define ONEMATH_PI_INTERFACE_REMOVED #endif #elif __has_include() #include diff --git a/src/config.hpp.in b/src/config.hpp.in index de44cb16b..758ca63a2 100644 --- a/src/config.hpp.in +++ b/src/config.hpp.in @@ -17,26 +17,26 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef ONEMKL_CONFIG_H -#define ONEMKL_CONFIG_H +#ifndef ONEMATH_CONFIG_H +#define ONEMATH_CONFIG_H -#cmakedefine ONEMKL_ENABLE_CUBLAS_BACKEND -#cmakedefine ONEMKL_ENABLE_CUFFT_BACKEND -#cmakedefine ONEMKL_ENABLE_CURAND_BACKEND -#cmakedefine ONEMKL_ENABLE_CUSOLVER_BACKEND -#cmakedefine ONEMKL_ENABLE_MKLCPU_BACKEND -#cmakedefine ONEMKL_ENABLE_MKLGPU_BACKEND -#cmakedefine ONEMKL_ENABLE_NETLIB_BACKEND -#cmakedefine ONEMKL_ENABLE_PORTBLAS_BACKEND -#cmakedefine ONEMKL_ENABLE_PORTBLAS_BACKEND_AMD_GPU -#cmakedefine ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_CPU -#cmakedefine ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_GPU -#cmakedefine ONEMKL_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU -#cmakedefine ONEMKL_ENABLE_PORTFFT_BACKEND -#cmakedefine ONEMKL_ENABLE_ROCBLAS_BACKEND -#cmakedefine ONEMKL_ENABLE_ROCFFT_BACKEND -#cmakedefine ONEMKL_ENABLE_ROCRAND_BACKEND -#cmakedefine ONEMKL_ENABLE_ROCSOLVER_BACKEND -#cmakedefine ONEMKL_BUILD_SHARED_LIBS +#cmakedefine ONEMATH_ENABLE_CUBLAS_BACKEND +#cmakedefine ONEMATH_ENABLE_CUFFT_BACKEND +#cmakedefine ONEMATH_ENABLE_CURAND_BACKEND +#cmakedefine ONEMATH_ENABLE_CUSOLVER_BACKEND +#cmakedefine ONEMATH_ENABLE_MKLCPU_BACKEND +#cmakedefine ONEMATH_ENABLE_MKLGPU_BACKEND +#cmakedefine ONEMATH_ENABLE_NETLIB_BACKEND +#cmakedefine ONEMATH_ENABLE_PORTBLAS_BACKEND +#cmakedefine ONEMATH_ENABLE_PORTBLAS_BACKEND_AMD_GPU +#cmakedefine ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_CPU +#cmakedefine ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_GPU +#cmakedefine ONEMATH_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU +#cmakedefine ONEMATH_ENABLE_PORTFFT_BACKEND +#cmakedefine ONEMATH_ENABLE_ROCBLAS_BACKEND +#cmakedefine ONEMATH_ENABLE_ROCFFT_BACKEND +#cmakedefine ONEMATH_ENABLE_ROCRAND_BACKEND +#cmakedefine ONEMATH_ENABLE_ROCSOLVER_BACKEND +#cmakedefine ONEMATH_BUILD_SHARED_LIBS #endif diff --git a/src/dft/CMakeLists.txt b/src/dft/CMakeLists.txt index d4c44747a..1b41cef72 100644 --- a/src/dft/CMakeLists.txt +++ b/src/dft/CMakeLists.txt @@ -29,11 +29,11 @@ target_include_directories(onemath_dft ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} $ ) -target_compile_options(onemath_dft PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(onemath_dft PRIVATE ${ONEMATH_BUILD_COPT}) set_target_properties(onemath_dft PROPERTIES POSITION_INDEPENDENT_CODE ON @@ -41,7 +41,7 @@ set_target_properties(onemath_dft PROPERTIES if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET onemath_dft SOURCES backends/descriptor.cxx dft_loader.cpp) else() - target_link_libraries(onemath_dft PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(onemath_dft PUBLIC ONEMATH::SYCL::SYCL) endif() include(WarningsUtils) diff --git a/src/dft/backends/backend_backward_instantiations.cxx b/src/dft/backends/backend_backward_instantiations.cxx index a6aeaf71b..eab281a96 100644 --- a/src/dft/backends/backend_backward_instantiations.cxx +++ b/src/dft/backends/backend_backward_instantiations.cxx @@ -27,32 +27,32 @@ using desc_cd_t = dft::detail::descriptor; using depends_vec_t = const std::vector &; -#define ONEMKL_DFT_BACKWARD_INSTANTIATIONS(DESCRIPTOR_T, SCALAR_T, FORWARD_T, BACKWARD_T) \ +#define ONEMATH_DFT_BACKWARD_INSTANTIATIONS(DESCRIPTOR_T, SCALAR_T, FORWARD_T, BACKWARD_T) \ /* Buffer API */ \ - template ONEMKL_EXPORT void compute_backward(DESCRIPTOR_T &, \ + template ONEMATH_EXPORT void compute_backward(DESCRIPTOR_T &, \ sycl::buffer &); \ - template ONEMKL_EXPORT void compute_backward( \ + template ONEMATH_EXPORT void compute_backward( \ DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &); \ - template ONEMKL_EXPORT void compute_backward( \ + template ONEMATH_EXPORT void compute_backward( \ DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &); \ - template ONEMKL_EXPORT void compute_backward( \ + template ONEMATH_EXPORT void compute_backward( \ DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &, \ sycl::buffer &, sycl::buffer &); \ \ /* USM API */ \ - template ONEMKL_EXPORT sycl::event compute_backward(DESCRIPTOR_T &, FORWARD_T *, \ + template ONEMATH_EXPORT sycl::event compute_backward(DESCRIPTOR_T &, FORWARD_T *, \ depends_vec_t); \ - template ONEMKL_EXPORT sycl::event compute_backward(DESCRIPTOR_T &, SCALAR_T *, \ + template ONEMATH_EXPORT sycl::event compute_backward(DESCRIPTOR_T &, SCALAR_T *, \ SCALAR_T *, depends_vec_t); \ - template ONEMKL_EXPORT sycl::event compute_backward( \ + template ONEMATH_EXPORT sycl::event compute_backward( \ DESCRIPTOR_T &, BACKWARD_T *, FORWARD_T *, depends_vec_t); \ - template ONEMKL_EXPORT sycl::event compute_backward( \ + template ONEMATH_EXPORT sycl::event compute_backward( \ DESCRIPTOR_T &, SCALAR_T *, SCALAR_T *, SCALAR_T *, SCALAR_T *, depends_vec_t); -ONEMKL_DFT_BACKWARD_INSTANTIATIONS(desc_rf_t, float, float, std::complex) -ONEMKL_DFT_BACKWARD_INSTANTIATIONS(desc_cf_t, float, std::complex, std::complex) -ONEMKL_DFT_BACKWARD_INSTANTIATIONS(desc_rd_t, double, double, std::complex) -ONEMKL_DFT_BACKWARD_INSTANTIATIONS(desc_cd_t, double, std::complex, std::complex) +ONEMATH_DFT_BACKWARD_INSTANTIATIONS(desc_rf_t, float, float, std::complex) +ONEMATH_DFT_BACKWARD_INSTANTIATIONS(desc_cf_t, float, std::complex, std::complex) +ONEMATH_DFT_BACKWARD_INSTANTIATIONS(desc_rd_t, double, double, std::complex) +ONEMATH_DFT_BACKWARD_INSTANTIATIONS(desc_cd_t, double, std::complex, std::complex) -#undef ONEMKL_DFT_BACKWARD_INSTANTIATIONS -#undef ONEMKL_DFT_BACKWARD_INSTANTIATIONS_REAL_ONLY +#undef ONEMATH_DFT_BACKWARD_INSTANTIATIONS +#undef ONEMATH_DFT_BACKWARD_INSTANTIATIONS_REAL_ONLY diff --git a/src/dft/backends/backend_forward_instantiations.cxx b/src/dft/backends/backend_forward_instantiations.cxx index a6ed371d5..fc4084633 100644 --- a/src/dft/backends/backend_forward_instantiations.cxx +++ b/src/dft/backends/backend_forward_instantiations.cxx @@ -27,32 +27,32 @@ using desc_cd_t = dft::detail::descriptor; using depends_vec_t = const std::vector &; -#define ONEMKL_DFT_FORWARD_INSTANTIATIONS(DESCRIPTOR_T, SCALAR_T, FORWARD_T, BACKWARD_T) \ +#define ONEMATH_DFT_FORWARD_INSTANTIATIONS(DESCRIPTOR_T, SCALAR_T, FORWARD_T, BACKWARD_T) \ /* Buffer API */ \ - template ONEMKL_EXPORT void compute_forward(DESCRIPTOR_T &, \ + template ONEMATH_EXPORT void compute_forward(DESCRIPTOR_T &, \ sycl::buffer &); \ - template ONEMKL_EXPORT void compute_forward( \ + template ONEMATH_EXPORT void compute_forward( \ DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &); \ - template ONEMKL_EXPORT void compute_forward( \ + template ONEMATH_EXPORT void compute_forward( \ DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &); \ - template ONEMKL_EXPORT void compute_forward( \ + template ONEMATH_EXPORT void compute_forward( \ DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &, \ sycl::buffer &, sycl::buffer &); \ \ /* USM API */ \ - template ONEMKL_EXPORT sycl::event compute_forward(DESCRIPTOR_T &, FORWARD_T *, \ + template ONEMATH_EXPORT sycl::event compute_forward(DESCRIPTOR_T &, FORWARD_T *, \ depends_vec_t); \ - template ONEMKL_EXPORT sycl::event compute_forward(DESCRIPTOR_T &, SCALAR_T *, \ + template ONEMATH_EXPORT sycl::event compute_forward(DESCRIPTOR_T &, SCALAR_T *, \ SCALAR_T *, depends_vec_t); \ - template ONEMKL_EXPORT sycl::event compute_forward(DESCRIPTOR_T &, FORWARD_T *, \ + template ONEMATH_EXPORT sycl::event compute_forward(DESCRIPTOR_T &, FORWARD_T *, \ BACKWARD_T *, depends_vec_t); \ - template ONEMKL_EXPORT sycl::event compute_forward( \ + template ONEMATH_EXPORT sycl::event compute_forward( \ DESCRIPTOR_T &, SCALAR_T *, SCALAR_T *, SCALAR_T *, SCALAR_T *, depends_vec_t); -ONEMKL_DFT_FORWARD_INSTANTIATIONS(desc_rf_t, float, float, std::complex) -ONEMKL_DFT_FORWARD_INSTANTIATIONS(desc_cf_t, float, std::complex, std::complex) -ONEMKL_DFT_FORWARD_INSTANTIATIONS(desc_rd_t, double, double, std::complex) -ONEMKL_DFT_FORWARD_INSTANTIATIONS(desc_cd_t, double, std::complex, std::complex) +ONEMATH_DFT_FORWARD_INSTANTIATIONS(desc_rf_t, float, float, std::complex) +ONEMATH_DFT_FORWARD_INSTANTIATIONS(desc_cf_t, float, std::complex, std::complex) +ONEMATH_DFT_FORWARD_INSTANTIATIONS(desc_rd_t, double, double, std::complex) +ONEMATH_DFT_FORWARD_INSTANTIATIONS(desc_cd_t, double, std::complex, std::complex) -#undef ONEMKL_DFT_FORWARD_INSTANTIATIONS -#undef ONEMKL_DFT_FORWARD_INSTANTIATIONS_REAL_ONLY +#undef ONEMATH_DFT_FORWARD_INSTANTIATIONS +#undef ONEMATH_DFT_FORWARD_INSTANTIATIONS_REAL_ONLY diff --git a/src/dft/backends/cufft/CMakeLists.txt b/src/dft/backends/cufft/CMakeLists.txt index b315c5f4f..2f9e908c7 100644 --- a/src/dft/backends/cufft/CMakeLists.txt +++ b/src/dft/backends/cufft/CMakeLists.txt @@ -32,18 +32,18 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} ) target_include_directories(${LIB_NAME} - PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} ) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if (${CMAKE_VERSION} VERSION_LESS "3.17.0") find_package(CUDA REQUIRED) @@ -54,7 +54,7 @@ else() target_link_libraries(${LIB_OBJ} PRIVATE CUDA::cufft CUDA::cuda_driver) endif() -target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL) +target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON @@ -64,7 +64,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) #Set oneMKL libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/dft/backends/cufft/backward.cpp b/src/dft/backends/cufft/backward.cpp index 6453b3301..d70c49611 100644 --- a/src/dft/backends/cufft/backward.cpp +++ b/src/dft/backends/cufft/backward.cpp @@ -49,7 +49,7 @@ cufftHandle get_bwd_plan(dft::detail::commit_impl *commit) { //In-place transform template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &inout) { const std::string func_name = "compute_backward(desc, inout)"; detail::expect_config( @@ -86,7 +86,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_backward(descriptor_type &, sycl::buffer, 1> &, +ONEMATH_EXPORT void compute_backward(descriptor_type &, sycl::buffer, 1> &, sycl::buffer, 1> &) { throw oneapi::mkl::unimplemented("DFT", "compute_backward(desc, inout_re, inout_im)", "cuFFT does not support real-real complex storage."); @@ -94,7 +94,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &, sycl::buffer -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &in, sycl::buffer, 1> &out) { const std::string func_name = "compute_backward(desc, in, out)"; @@ -137,7 +137,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &desc, //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_backward(descriptor_type &, sycl::buffer, 1> &, +ONEMATH_EXPORT void compute_backward(descriptor_type &, sycl::buffer, 1> &, sycl::buffer, 1> &, sycl::buffer, 1> &, sycl::buffer, 1> &) { @@ -149,7 +149,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &, sycl::buffer -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, const std::vector &dependencies) { const std::string func_name = "compute_backward(desc, inout, dependencies)"; detail::expect_config( @@ -185,7 +185,7 @@ ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &, scalar *, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &, scalar *, scalar *, const std::vector &) { throw oneapi::mkl::unimplemented("DFT", @@ -195,7 +195,7 @@ ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &, scalar -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, fwd *out, const std::vector &dependencies) { const std::string func_name = "compute_backward(desc, in, out, dependencies)"; @@ -231,7 +231,7 @@ ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &, scalar *, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &, scalar *, scalar *, scalar *, scalar *, const std::vector &) { diff --git a/src/dft/backends/cufft/execute_helper.hpp b/src/dft/backends/cufft/execute_helper.hpp index e6014478c..629cf1767 100644 --- a/src/dft/backends/cufft/execute_helper.hpp +++ b/src/dft/backends/cufft/execute_helper.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_SRC_EXECUTE_HELPER_CUFFT_HPP_ -#define _ONEMKL_DFT_SRC_EXECUTE_HELPER_CUFFT_HPP_ +#ifndef _ONEMATH_DFT_SRC_EXECUTE_HELPER_CUFFT_HPP_ +#define _ONEMATH_DFT_SRC_EXECUTE_HELPER_CUFFT_HPP_ #if __has_include() #include @@ -149,4 +149,4 @@ inline CUstream setup_stream(const std::string &func, sycl::interop_handle ih, c } // namespace oneapi::mkl::dft::cufft::detail -#endif // _ONEMKL_DFT_SRC_EXECUTE_HELPER_CUFFT_HPP_ +#endif // _ONEMATH_DFT_SRC_EXECUTE_HELPER_CUFFT_HPP_ diff --git a/src/dft/backends/cufft/forward.cpp b/src/dft/backends/cufft/forward.cpp index 5f75752c2..757c2b97d 100644 --- a/src/dft/backends/cufft/forward.cpp +++ b/src/dft/backends/cufft/forward.cpp @@ -52,7 +52,7 @@ cufftHandle get_fwd_plan(dft::detail::commit_impl *commit) { //In-place transform template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &inout) { const std::string func_name = "compute_forward(desc, inout)"; detail::expect_config( @@ -89,7 +89,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_forward(descriptor_type &, sycl::buffer, 1> &, +ONEMATH_EXPORT void compute_forward(descriptor_type &, sycl::buffer, 1> &, sycl::buffer, 1> &) { throw oneapi::mkl::unimplemented("DFT", "compute_forward(desc, inout_re, inout_im)", "cuFFT does not support real-real complex storage."); @@ -97,7 +97,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &, sycl::buffer -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, sycl::buffer, 1> &out) { const std::string func_name = "compute_forward(desc, in, out)"; detail::expect_config( @@ -139,7 +139,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer -ONEMKL_EXPORT void compute_forward(descriptor_type &, sycl::buffer, 1> &, +ONEMATH_EXPORT void compute_forward(descriptor_type &, sycl::buffer, 1> &, sycl::buffer, 1> &, sycl::buffer, 1> &, sycl::buffer, 1> &) { @@ -151,7 +151,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &, sycl::buffer -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, const std::vector &dependencies) { const std::string func_name = "compute_forward(desc, inout, dependencies)"; detail::expect_config( @@ -187,7 +187,7 @@ ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &, scalar *, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &, scalar *, scalar *, const std::vector &) { throw oneapi::mkl::unimplemented("DFT", @@ -197,7 +197,7 @@ ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &, scalar -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, bwd *out, const std::vector &dependencies) { const std::string func_name = "compute_forward(desc, in, out, dependencies)"; @@ -233,7 +233,7 @@ ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &, scalar *, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &, scalar *, scalar *, scalar *, scalar *, const std::vector &) { diff --git a/src/dft/backends/mklcpu/CMakeLists.txt b/src/dft/backends/mklcpu/CMakeLists.txt index 2d1617854..f1bcdd9d5 100644 --- a/src/dft/backends/mklcpu/CMakeLists.txt +++ b/src/dft/backends/mklcpu/CMakeLists.txt @@ -33,32 +33,32 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} ) target_include_directories(${LIB_NAME} - PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} ) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET ${LIB_OBJ} SOURCES ${SOURCES}) endif() if(TARGET MKL::MKL_SYCL::DFT) target_link_libraries(${LIB_OBJ} - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL PUBLIC MKL::MKL_SYCL::DFT PRIVATE onemath_warnings ) else() target_link_libraries(${LIB_OBJ} - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL PUBLIC MKL::MKL_DPCPP PRIVATE onemath_warnings ) @@ -72,7 +72,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) #Set oneMKL libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/dft/backends/mklcpu/backward.cpp b/src/dft/backends/mklcpu/backward.cpp index b7a517849..0cf42948e 100644 --- a/src/dft/backends/mklcpu/backward.cpp +++ b/src/dft/backends/mklcpu/backward.cpp @@ -77,7 +77,7 @@ auto get_buffer(commit_t *commit_handle) { //In-place transform template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &inout) { detail::expect_config( desc, "Unexpected value for placement"); @@ -104,7 +104,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &inout_re, sycl::buffer, 1> &inout_im) { detail::expect_config -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &in, sycl::buffer, 1> &out) { detail::expect_config -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &in_re, sycl::buffer, 1> &in_im, sycl::buffer, 1> &out_re, @@ -210,7 +210,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &desc, //In-place transform template -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, const std::vector &dependencies) { detail::expect_config( desc, "Unexpected value for placement"); @@ -237,7 +237,7 @@ ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *inout_re, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *inout_re, scalar *inout_im, const std::vector &dependencies) { detail::expect_config -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, fwd *out, const std::vector &dependencies) { // Check: inplace, complex storage @@ -295,7 +295,7 @@ ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *in_re, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *in_re, scalar *in_im, scalar *out_re, scalar *out_im, diff --git a/src/dft/backends/mklcpu/commit_derived_impl.hpp b/src/dft/backends/mklcpu/commit_derived_impl.hpp index 61b55942b..a2e3320c5 100644 --- a/src/dft/backends/mklcpu/commit_derived_impl.hpp +++ b/src/dft/backends/mklcpu/commit_derived_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_COMMIT_DERIVED_IMPL_HPP_ -#define _ONEMKL_DFT_COMMIT_DERIVED_IMPL_HPP_ +#ifndef _ONEMATH_DFT_COMMIT_DERIVED_IMPL_HPP_ +#define _ONEMATH_DFT_COMMIT_DERIVED_IMPL_HPP_ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/dft/detail/types_impl.hpp" @@ -85,4 +85,4 @@ using commit_derived_t = detail::commit_derived_impl; } // namespace mkl } // namespace oneapi -#endif // _ONEMKL_DFT_COMMIT_DERIVED_IMPL_HPP_ +#endif // _ONEMATH_DFT_COMMIT_DERIVED_IMPL_HPP_ diff --git a/src/dft/backends/mklcpu/forward.cpp b/src/dft/backends/mklcpu/forward.cpp index 62b5755d1..7152f3322 100644 --- a/src/dft/backends/mklcpu/forward.cpp +++ b/src/dft/backends/mklcpu/forward.cpp @@ -78,7 +78,7 @@ auto get_buffer(commit_t *commit_handle) { //In-place transform template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &inout) { detail::expect_config( desc, "Unexpected value for placement"); @@ -105,7 +105,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &inout_re, sycl::buffer, 1> &inout_im) { detail::expect_config -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, sycl::buffer, 1> &out) { detail::expect_config(desc, @@ -169,7 +169,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in_re, sycl::buffer, 1> &in_im, sycl::buffer, 1> &out_re, @@ -210,7 +210,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, //In-place transform template -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, const std::vector &dependencies) { detail::expect_config( desc, "Unexpected value for placement"); @@ -238,7 +238,7 @@ ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *inout_re, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *inout_re, scalar *inout_im, const std::vector &dependencies) { detail::expect_config -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, bwd *out, const std::vector &dependencies) { // Check: inplace @@ -299,7 +299,7 @@ ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *in_re, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *in_re, scalar *in_im, scalar *out_re, scalar *out_im, diff --git a/src/dft/backends/mklcpu/mklcpu_helpers.hpp b/src/dft/backends/mklcpu/mklcpu_helpers.hpp index 8c0f1cf1a..2986806ed 100644 --- a/src/dft/backends/mklcpu/mklcpu_helpers.hpp +++ b/src/dft/backends/mklcpu/mklcpu_helpers.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_SRC_MKLCPU_HELPERS_HPP_ -#define _ONEMKL_DFT_SRC_MKLCPU_HELPERS_HPP_ +#ifndef _ONEMATH_DFT_SRC_MKLCPU_HELPERS_HPP_ +#define _ONEMATH_DFT_SRC_MKLCPU_HELPERS_HPP_ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/dft/detail/types_impl.hpp" @@ -175,4 +175,4 @@ typename AccType::value_type* acc_to_ptr(AccType acc) { } // namespace oneapi::mkl::dft::mklcpu::detail -#endif // _ONEMKL_DFT_SRC_MKLCPU_HELPERS_HPP_ +#endif // _ONEMATH_DFT_SRC_MKLCPU_HELPERS_HPP_ diff --git a/src/dft/backends/mklgpu/CMakeLists.txt b/src/dft/backends/mklgpu/CMakeLists.txt index c1a0c2c60..d0f21c8a7 100644 --- a/src/dft/backends/mklgpu/CMakeLists.txt +++ b/src/dft/backends/mklgpu/CMakeLists.txt @@ -33,7 +33,7 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_NAME} - PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} ) # Due to using the same file name for different headers in this library and in @@ -47,20 +47,20 @@ target_compile_options(${LIB_OBJ} target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::DFT) target_link_libraries(${LIB_OBJ} - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL PUBLIC MKL::MKL_SYCL::DFT PRIVATE onemath_warnings ) else() target_link_libraries(${LIB_OBJ} - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL PUBLIC MKL::MKL_DPCPP PRIVATE onemath_warnings ) @@ -74,7 +74,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) #Set oneMKL libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/dft/backends/mklgpu/backward.cpp b/src/dft/backends/mklgpu/backward.cpp index 5ca2d68ab..4a8bb1671 100644 --- a/src/dft/backends/mklgpu/backward.cpp +++ b/src/dft/backends/mklgpu/backward.cpp @@ -83,7 +83,7 @@ inline auto expect_config(DescT &desc, const char *message) { //In-place transform template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &inout) { detail::expect_config( desc, "Unexpected value for placement"); @@ -92,7 +92,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_backward(descriptor_type & /*desc*/, +ONEMATH_EXPORT void compute_backward(descriptor_type & /*desc*/, sycl::buffer, 1> & /*inout_re*/, sycl::buffer, 1> & /*inout_im*/) { throw mkl::unimplemented("DFT", "compute_backward", @@ -101,7 +101,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type & /*desc*/, //Out-of-place transform template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &in, sycl::buffer, 1> &out) { detail::expect_config -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> & /*in_re*/, sycl::buffer, 1> & /*in_im*/, sycl::buffer, 1> & /*out_re*/, @@ -129,7 +129,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &desc, //In-place transform template -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, const std::vector &dependencies) { detail::expect_config( desc, "Unexpected value for placement"); @@ -138,7 +138,7 @@ ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type & /*desc*/, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type & /*desc*/, scalar * /*inout_re*/, scalar * /*inout_im*/, const std::vector & /*dependencies*/) { @@ -149,7 +149,7 @@ ONEMKL_EXPORT sycl::event compute_backward(descriptor_type & /*desc*/, //Out-of-place transform template -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, fwd *out, const std::vector &dependencies) { detail::expect_config -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar * /*in_re*/, scalar * /*in_im*/, scalar * /*out_re*/, diff --git a/src/dft/backends/mklgpu/forward.cpp b/src/dft/backends/mklgpu/forward.cpp index 45b527f52..5037241c9 100644 --- a/src/dft/backends/mklgpu/forward.cpp +++ b/src/dft/backends/mklgpu/forward.cpp @@ -90,7 +90,7 @@ inline auto expect_config(DescT &desc, const char *message) { //In-place transform template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &inout) { detail::expect_config( desc, "Unexpected value for placement"); @@ -99,7 +99,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_forward(descriptor_type & /*desc*/, +ONEMATH_EXPORT void compute_forward(descriptor_type & /*desc*/, sycl::buffer, 1> & /*inout_re*/, sycl::buffer, 1> & /*inout_im*/) { throw mkl::unimplemented("DFT", "compute_forward", @@ -108,7 +108,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type & /*desc*/, //Out-of-place transform template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, sycl::buffer, 1> &out) { detail::expect_config(desc, @@ -118,7 +118,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> & /*in_re*/, sycl::buffer, 1> & /*in_im*/, sycl::buffer, 1> & /*out_re*/, @@ -135,7 +135,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, //In-place transform template -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, const std::vector &dependencies) { detail::expect_config( desc, "Unexpected value for placement"); @@ -144,7 +144,7 @@ ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type & /*desc*/, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type & /*desc*/, scalar * /*inout_re*/, scalar * /*inout_im*/, const std::vector & /*dependencies*/) { @@ -155,7 +155,7 @@ ONEMKL_EXPORT sycl::event compute_forward(descriptor_type & /*desc*/, //Out-of-place transform template -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, bwd *out, const std::vector &dependencies) { detail::expect_config -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar * /*in_re*/, scalar * /*in_im*/, scalar * /*out_re*/, diff --git a/src/dft/backends/mklgpu/mklgpu_helpers.hpp b/src/dft/backends/mklgpu/mklgpu_helpers.hpp index 08119253c..afecb984d 100644 --- a/src/dft/backends/mklgpu/mklgpu_helpers.hpp +++ b/src/dft/backends/mklgpu/mklgpu_helpers.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_SRC_MKLGPU_HELPERS_HPP_ -#define _ONEMKL_DFT_SRC_MKLGPU_HELPERS_HPP_ +#ifndef _ONEMATH_DFT_SRC_MKLGPU_HELPERS_HPP_ +#define _ONEMATH_DFT_SRC_MKLGPU_HELPERS_HPP_ #include "oneapi/math/detail/exceptions.hpp" #include "oneapi/math/dft/detail/types_impl.hpp" @@ -176,4 +176,4 @@ to_mklgpu_config_value( } // namespace mkl } // namespace oneapi -#endif // _ONEMKL_DFT_SRC_MKLGPU_HELPERS_HPP_ +#endif // _ONEMATH_DFT_SRC_MKLGPU_HELPERS_HPP_ diff --git a/src/dft/backends/portfft/CMakeLists.txt b/src/dft/backends/portfft/CMakeLists.txt index 4aec1d11a..da351deb2 100644 --- a/src/dft/backends/portfft/CMakeLists.txt +++ b/src/dft/backends/portfft/CMakeLists.txt @@ -54,8 +54,8 @@ if (IS_DPCPP AND UNIX AND NOT FOUND_TARGETS) list(APPEND TARGETS_COMPILE_OPTIONS -fsycl-targets=${TARGETS_TRIPLES}) list(APPEND TARGETS_LINK_OPTIONS -fsycl-targets=${TARGETS_TRIPLES}) - target_compile_options(ONEMKL::SYCL::SYCL INTERFACE ${TARGETS_COMPILE_OPTIONS}) - target_link_options(ONEMKL::SYCL::SYCL INTERFACE ${TARGETS_LINK_OPTIONS}) + target_compile_options(ONEMATH::SYCL::SYCL INTERFACE ${TARGETS_COMPILE_OPTIONS}) + target_link_options(ONEMATH::SYCL::SYCL INTERFACE ${TARGETS_LINK_OPTIONS}) endif() set(LIB_NAME onemath_dft_portfft) @@ -91,19 +91,19 @@ endif() target_link_libraries(${LIB_OBJ} PRIVATE onemath_warnings) target_include_directories(${LIB_OBJ} - PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} ) target_include_directories(${LIB_NAME} - PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} ) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) -target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL) +target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON @@ -113,7 +113,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) #Set oneMKL libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/dft/backends/portfft/portfft_helper.hpp b/src/dft/backends/portfft/portfft_helper.hpp index 4322987a8..446dd3b3f 100644 --- a/src/dft/backends/portfft/portfft_helper.hpp +++ b/src/dft/backends/portfft/portfft_helper.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_SRC_PORTFFT_HELPERS_HPP_ -#define _ONEMKL_DFT_SRC_PORTFFT_HELPERS_HPP_ +#ifndef _ONEMATH_DFT_SRC_PORTFFT_HELPERS_HPP_ +#define _ONEMATH_DFT_SRC_PORTFFT_HELPERS_HPP_ #include diff --git a/src/dft/backends/rocfft/CMakeLists.txt b/src/dft/backends/rocfft/CMakeLists.txt index 0c3f2f5af..0e2548d95 100644 --- a/src/dft/backends/rocfft/CMakeLists.txt +++ b/src/dft/backends/rocfft/CMakeLists.txt @@ -32,18 +32,18 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} ) target_include_directories(${LIB_NAME} - PUBLIC ${ONEMKL_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} ) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) find_package(HIP REQUIRED) # Require the minimum rocFFT version matching with ROCm 5.4.3. @@ -64,7 +64,7 @@ find_path( ) target_include_directories(${LIB_OBJ} PRIVATE ${rocfft_EXTRA_INCLUDE_DIR}) -target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL) +target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON @@ -74,7 +74,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) # Set oneMKL libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/dft/backends/rocfft/backward.cpp b/src/dft/backends/rocfft/backward.cpp index 938108e21..d9583a1c6 100644 --- a/src/dft/backends/rocfft/backward.cpp +++ b/src/dft/backends/rocfft/backward.cpp @@ -55,7 +55,7 @@ rocfft_execution_info get_bwd_info(dft::detail::commit_impl *commit) //In-place transform template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &inout) { const std::string func_name = "compute_backward(desc, inout)"; detail::expect_config( @@ -92,7 +92,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &inout_re, sycl::buffer, 1> &inout_im) { const std::string func_name = "compute_backward(desc, inout_re, inout_im)"; @@ -131,7 +131,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &desc, //Out-of-place transform template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &in, sycl::buffer, 1> &out) { detail::expect_config( @@ -164,7 +164,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &desc, //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_backward(descriptor_type &desc, +ONEMATH_EXPORT void compute_backward(descriptor_type &desc, sycl::buffer, 1> &in_re, sycl::buffer, 1> &in_im, sycl::buffer, 1> &out_re, @@ -211,7 +211,7 @@ ONEMKL_EXPORT void compute_backward(descriptor_type &desc, //In-place transform template -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, const std::vector &deps) { const std::string func_name = "compute_backward(desc, inout, deps)"; detail::expect_config( @@ -249,7 +249,7 @@ ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *inout_re, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *inout_re, scalar *inout_im, const std::vector &deps) { const std::string func_name = "compute_backward(desc, inout_re, inout_im, deps)"; @@ -283,7 +283,7 @@ ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, fwd *out, const std::vector &deps) { detail::expect_config( @@ -316,7 +316,7 @@ ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd -ONEMKL_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *in_re, +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *in_re, scalar *in_im, scalar *out_re, scalar *out_im, diff --git a/src/dft/backends/rocfft/execute_helper.hpp b/src/dft/backends/rocfft/execute_helper.hpp index fa2eecb24..b5b4cc353 100644 --- a/src/dft/backends/rocfft/execute_helper.hpp +++ b/src/dft/backends/rocfft/execute_helper.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_SRC_EXECUTE_HELPER_ROCFFT_HPP_ -#define _ONEMKL_DFT_SRC_EXECUTE_HELPER_ROCFFT_HPP_ +#ifndef _ONEMATH_DFT_SRC_EXECUTE_HELPER_ROCFFT_HPP_ +#define _ONEMATH_DFT_SRC_EXECUTE_HELPER_ROCFFT_HPP_ #if __has_include() #include @@ -100,4 +100,4 @@ inline void execute_checked(const std::string &func, hipStream_t stream, const r } // namespace oneapi::mkl::dft::rocfft::detail -#endif // _ONEMKL_DFT_SRC_EXECUTE_HELPER_ROCFFT_HPP_ +#endif // _ONEMATH_DFT_SRC_EXECUTE_HELPER_ROCFFT_HPP_ diff --git a/src/dft/backends/rocfft/forward.cpp b/src/dft/backends/rocfft/forward.cpp index ca9f8ad43..ebafc9e2f 100644 --- a/src/dft/backends/rocfft/forward.cpp +++ b/src/dft/backends/rocfft/forward.cpp @@ -58,7 +58,7 @@ rocfft_execution_info get_fwd_info(dft::detail::commit_impl *commit) //In-place transform template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &inout) { const std::string func_name = "compute_forward(desc, inout)"; detail::expect_config( @@ -95,7 +95,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &inout_re, sycl::buffer, 1> &inout_im) { const std::string func_name = "compute_forward(desc, inout_re, inout_im)"; @@ -134,7 +134,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, //Out-of-place transform template -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, sycl::buffer, 1> &out) { detail::expect_config( desc, "Unexpected value for placement"); @@ -166,7 +166,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer -ONEMKL_EXPORT void compute_forward(descriptor_type &desc, +ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in_re, sycl::buffer, 1> &in_im, sycl::buffer, 1> &out_re, @@ -213,7 +213,7 @@ ONEMKL_EXPORT void compute_forward(descriptor_type &desc, //In-place transform template -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, const std::vector &deps) { const std::string func_name = "compute_forward(desc, inout, deps)"; detail::expect_config( @@ -251,7 +251,7 @@ ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *inout_re, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *inout_re, scalar *inout_im, const std::vector &deps) { const std::string func_name = "compute_forward(desc, inout_re, inout_im, deps)"; @@ -283,7 +283,7 @@ ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, bwd *out, const std::vector &deps) { detail::expect_config( @@ -316,7 +316,7 @@ ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMKL_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *in_re, +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *in_re, scalar *in_im, scalar *out_re, scalar *out_im, diff --git a/src/dft/backends/rocfft/rocfft_handle.hpp b/src/dft/backends/rocfft/rocfft_handle.hpp index ea4f44d68..04eca565d 100644 --- a/src/dft/backends/rocfft/rocfft_handle.hpp +++ b/src/dft/backends/rocfft/rocfft_handle.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_SRC_ROCFFT_ROCFFT_HANDLE_HPP_ -#define _ONEMKL_DFT_SRC_ROCFFT_ROCFFT_HANDLE_HPP_ +#ifndef _ONEMATH_DFT_SRC_ROCFFT_ROCFFT_HANDLE_HPP_ +#define _ONEMATH_DFT_SRC_ROCFFT_ROCFFT_HANDLE_HPP_ #include diff --git a/src/dft/descriptor_config_helper.hpp b/src/dft/descriptor_config_helper.hpp index 109e39e80..6a6c4cced 100644 --- a/src/dft/descriptor_config_helper.hpp +++ b/src/dft/descriptor_config_helper.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DETAIL_DESCRIPTOR_CONFIG_HELPER_HPP_ -#define _ONEMKL_DETAIL_DESCRIPTOR_CONFIG_HELPER_HPP_ +#ifndef _ONEMATH_DETAIL_DESCRIPTOR_CONFIG_HELPER_HPP_ +#define _ONEMATH_DETAIL_DESCRIPTOR_CONFIG_HELPER_HPP_ #include #include @@ -259,4 +259,4 @@ void set_value(dft_values& vals, } // namespace mkl } // namespace oneapi -#endif //_ONEMKL_DETAIL_DESCRIPTOR_CONFIG_HELPER_HPP_ +#endif //_ONEMATH_DETAIL_DESCRIPTOR_CONFIG_HELPER_HPP_ diff --git a/src/dft/execute_helper_generic.hpp b/src/dft/execute_helper_generic.hpp index 22fe0cb33..b857328df 100644 --- a/src/dft/execute_helper_generic.hpp +++ b/src/dft/execute_helper_generic.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_DFT_SRC_EXECUTE_HELPER_GENERIC_HPP_ -#define _ONEMKL_DFT_SRC_EXECUTE_HELPER_GENERIC_HPP_ +#ifndef _ONEMATH_DFT_SRC_EXECUTE_HELPER_GENERIC_HPP_ +#define _ONEMATH_DFT_SRC_EXECUTE_HELPER_GENERIC_HPP_ #if __has_include() #include @@ -50,4 +50,4 @@ static inline void fft_enqueue_task(HandlerT&& cgh, FnT&& f) { } // namespace oneapi::mkl::dft::detail -#endif // _ONEMKL_DFT_SRC_EXECUTE_HELPER_GENERIC_HPP_ +#endif // _ONEMATH_DFT_SRC_EXECUTE_HELPER_GENERIC_HPP_ diff --git a/src/include/dtype_string.hpp b/src/include/dtype_string.hpp index 6f2a87feb..528b7cee7 100644 --- a/src/include/dtype_string.hpp +++ b/src/include/dtype_string.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_ERROR_HELPER_HPP_ -#define _ONEMKL_ERROR_HELPER_HPP_ +#ifndef _ONEMATH_ERROR_HELPER_HPP_ +#define _ONEMATH_ERROR_HELPER_HPP_ #include @@ -53,4 +53,4 @@ inline const std::string dtype_string() { return "int8"; } -#endif //_ONEMKL_ERROR_HELPER_HPP_ +#endif //_ONEMATH_ERROR_HELPER_HPP_ diff --git a/src/include/runtime_support_helper.hpp b/src/include/runtime_support_helper.hpp index 7c3514673..cabc4ea8d 100644 --- a/src/include/runtime_support_helper.hpp +++ b/src/include/runtime_support_helper.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_RUNTIME_SUPPORT_HELPER_HPP_ -#define _ONEMKL_RUNTIME_SUPPORT_HELPER_HPP_ +#ifndef _ONEMATH_RUNTIME_SUPPORT_HELPER_HPP_ +#define _ONEMATH_RUNTIME_SUPPORT_HELPER_HPP_ #if __has_include() #include @@ -50,4 +50,4 @@ bool verify_support(sycl::queue q, sycl::aspect aspect) { } } -#endif //_ONEMKL_RUNTIME_SUPPORT_HELPER_HPP_ +#endif //_ONEMATH_RUNTIME_SUPPORT_HELPER_HPP_ diff --git a/src/lapack/CMakeLists.txt b/src/lapack/CMakeLists.txt index cd5c28332..0bb134db6 100644 --- a/src/lapack/CMakeLists.txt +++ b/src/lapack/CMakeLists.txt @@ -29,15 +29,15 @@ target_include_directories(onemath_lapack ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} $ ) -target_compile_options(onemath_lapack PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(onemath_lapack PRIVATE ${ONEMATH_BUILD_COPT}) set_target_properties(onemath_lapack PROPERTIES POSITION_INDEPENDENT_CODE ON ) -target_link_libraries(onemath_lapack PUBLIC ONEMKL::SYCL::SYCL) +target_link_libraries(onemath_lapack PUBLIC ONEMATH::SYCL::SYCL) endif() diff --git a/src/lapack/backends/cusolver/CMakeLists.txt b/src/lapack/backends/cusolver/CMakeLists.txt index bfcdc3721..1828c64df 100644 --- a/src/lapack/backends/cusolver/CMakeLists.txt +++ b/src/lapack/backends/cusolver/CMakeLists.txt @@ -23,7 +23,7 @@ find_package(cuSOLVER REQUIRED) find_package(cuBLAS REQUIRED) set(SOURCES cusolver_lapack.cpp cusolver_batch.cpp - $<$:cusolver_scope_handle.cpp > + $<$:cusolver_scope_handle.cpp > $<$: cusolver_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) @@ -35,13 +35,13 @@ target_include_directories(${LIB_OBJ} ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/blas/backends/cublas ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) target_link_libraries(${LIB_OBJ} - PUBLIC ONEMKL::SYCL::SYCL - ONEMKL::cuSOLVER::cuSOLVER - ONEMKL::cuBLAS::cuBLAS) + PUBLIC ONEMATH::SYCL::SYCL + ONEMATH::cuSOLVER::cuSOLVER + ONEMATH::cuBLAS::cuBLAS) target_compile_features(${LIB_OBJ} PUBLIC cxx_std_11) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON) diff --git a/src/lapack/backends/cusolver/cusolver_scope_handle.cpp b/src/lapack/backends/cusolver/cusolver_scope_handle.cpp index f381336d6..513254851 100644 --- a/src/lapack/backends/cusolver/cusolver_scope_handle.cpp +++ b/src/lapack/backends/cusolver/cusolver_scope_handle.cpp @@ -35,7 +35,7 @@ namespace cusolver { * takes place if no other element in the container has a key equivalent to * the one being emplaced (keys in a map container are unique). */ -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED thread_local cusolver_handle CusolverScopedContextHandler::handle_helper = cusolver_handle{}; #else @@ -98,7 +98,7 @@ cusolverDnHandle_t CusolverScopedContextHandler::get_handle(const sycl::queue &q CUresult cuErr; CUcontext desired; CUDA_ERROR_FUNC(cuDevicePrimaryCtxRetain, cuErr, &desired, cudaDevice); -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED auto piPlacedContext_ = reinterpret_cast(desired); #else auto piPlacedContext_ = reinterpret_cast(desired); diff --git a/src/lapack/backends/cusolver/cusolver_scope_handle.hpp b/src/lapack/backends/cusolver/cusolver_scope_handle.hpp index 1bd916f71..c25cc2c35 100644 --- a/src/lapack/backends/cusolver/cusolver_scope_handle.hpp +++ b/src/lapack/backends/cusolver/cusolver_scope_handle.hpp @@ -42,8 +42,8 @@ // After Plugin Interface removal in DPC++ ur.hpp is the new include #if __has_include() #include -#ifndef ONEMKL_PI_INTERFACE_REMOVED -#define ONEMKL_PI_INTERFACE_REMOVED +#ifndef ONEMATH_PI_INTERFACE_REMOVED +#define ONEMATH_PI_INTERFACE_REMOVED #endif #elif __has_include() #include @@ -92,7 +92,7 @@ class CusolverScopedContextHandler { sycl::context *placedContext_; bool needToRecover_; sycl::interop_handle &ih; -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED static thread_local cusolver_handle handle_helper; #else static thread_local cusolver_handle handle_helper; diff --git a/src/lapack/backends/cusolver/cusolver_task.hpp b/src/lapack/backends/cusolver/cusolver_task.hpp index 972b41ae9..7059b5063 100644 --- a/src/lapack/backends/cusolver/cusolver_task.hpp +++ b/src/lapack/backends/cusolver/cusolver_task.hpp @@ -34,8 +34,8 @@ // After Plugin Interface removal in DPC++ ur.hpp is the new include #if __has_include() #include -#ifndef ONEMKL_PI_INTERFACE_REMOVED -#define ONEMKL_PI_INTERFACE_REMOVED +#ifndef ONEMATH_PI_INTERFACE_REMOVED +#define ONEMATH_PI_INTERFACE_REMOVED #endif #elif __has_include() #include diff --git a/src/lapack/backends/cusolver/cusolver_wrappers.cpp b/src/lapack/backends/cusolver/cusolver_wrappers.cpp index acfff6c65..34dd68da2 100644 --- a/src/lapack/backends/cusolver/cusolver_wrappers.cpp +++ b/src/lapack/backends/cusolver/cusolver_wrappers.cpp @@ -21,7 +21,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT lapack_function_table_t mkl_lapack_table = { +extern "C" ONEMATH_EXPORT lapack_function_table_t mkl_lapack_table = { WRAPPER_VERSION, #define LAPACK_BACKEND cusolver oneapi::mkl::lapack::cusolver::gebrd, diff --git a/src/lapack/backends/mklcpu/CMakeLists.txt b/src/lapack/backends/mklcpu/CMakeLists.txt index d53c0711f..688dba71d 100644 --- a/src/lapack/backends/mklcpu/CMakeLists.txt +++ b/src/lapack/backends/mklcpu/CMakeLists.txt @@ -31,15 +31,15 @@ target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::LAPACK) - target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL MKL::MKL_SYCL::LAPACK) + target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_SYCL::LAPACK) else() - target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL MKL::MKL_DPCPP) + target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_DPCPP) endif() set_target_properties(${LIB_OBJ} PROPERTIES @@ -50,7 +50,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) #Set oneMKL libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp b/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp index a3c1cf6fb..30195d3a1 100644 --- a/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp +++ b/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT lapack_function_table_t mkl_lapack_table = { +extern "C" ONEMATH_EXPORT lapack_function_table_t mkl_lapack_table = { WRAPPER_VERSION, #define LAPACK_BACKEND mklcpu #include "../mkl_common/lapack_wrappers.cxx" diff --git a/src/lapack/backends/mklgpu/CMakeLists.txt b/src/lapack/backends/mklgpu/CMakeLists.txt index 213cbb73d..2be7539e9 100644 --- a/src/lapack/backends/mklgpu/CMakeLists.txt +++ b/src/lapack/backends/mklgpu/CMakeLists.txt @@ -31,15 +31,15 @@ target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::LAPACK) - target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL MKL::MKL_SYCL::LAPACK) + target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_SYCL::LAPACK) else() - target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL MKL::MKL_DPCPP) + target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_DPCPP) endif() set_target_properties(${LIB_OBJ} PROPERTIES @@ -50,7 +50,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) #Set oneMKL libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp b/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp index 675c5653c..e6435b898 100644 --- a/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp +++ b/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT lapack_function_table_t mkl_lapack_table = { +extern "C" ONEMATH_EXPORT lapack_function_table_t mkl_lapack_table = { WRAPPER_VERSION, #define LAPACK_BACKEND mklgpu #include "../mkl_common/lapack_wrappers.cxx" diff --git a/src/lapack/backends/rocsolver/CMakeLists.txt b/src/lapack/backends/rocsolver/CMakeLists.txt index f62f56ec6..a7dde150c 100644 --- a/src/lapack/backends/rocsolver/CMakeLists.txt +++ b/src/lapack/backends/rocsolver/CMakeLists.txt @@ -27,7 +27,7 @@ find_package(Threads REQUIRED) set(SOURCES rocsolver_lapack.cpp rocsolver_batch.cpp - $<$:rocsolver_scope_handle.cpp> + $<$:rocsolver_scope_handle.cpp> $<$: rocsolver_wrappers.cpp>) add_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) @@ -38,11 +38,11 @@ target_include_directories(${LIB_OBJ} ${PROJECT_SOURCE_DIR}/src/include ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) target_link_libraries(${LIB_OBJ} PRIVATE roc::rocsolver hip::host Threads::Threads) -target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL) +target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL) target_compile_features(${LIB_OBJ} PUBLIC cxx_std_17) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON) diff --git a/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp b/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp index 2a2e7d1ab..ee3de2367 100644 --- a/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp @@ -37,7 +37,7 @@ namespace rocsolver { * takes place if no other element in the container has a key equivalent to * the one being emplaced (keys in a map container are unique). */ -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED thread_local rocsolver_handle RocsolverScopedContextHandler::handle_helper = rocsolver_handle{}; #else @@ -100,7 +100,7 @@ rocblas_handle RocsolverScopedContextHandler::get_handle(const sycl::queue &queu hipError_t hipErr; hipCtx_t desired; HIP_ERROR_FUNC(hipDevicePrimaryCtxRetain, hipErr, &desired, hipDevice); -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED auto piPlacedContext_ = reinterpret_cast(desired); #else auto piPlacedContext_ = reinterpret_cast(desired); diff --git a/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp b/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp index 443a1ca2a..98368f0db 100644 --- a/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp @@ -36,8 +36,8 @@ // After Plugin Interface removal in DPC++ ur.hpp is the new include #if __has_include() #include -#ifndef ONEMKL_PI_INTERFACE_REMOVED -#define ONEMKL_PI_INTERFACE_REMOVED +#ifndef ONEMATH_PI_INTERFACE_REMOVED +#define ONEMATH_PI_INTERFACE_REMOVED #endif #elif __has_include() #include @@ -55,7 +55,7 @@ class RocsolverScopedContextHandler { sycl::context *placedContext_; bool needToRecover_; sycl::interop_handle &ih; -#ifdef ONEMKL_PI_INTERFACE_REMOVED +#ifdef ONEMATH_PI_INTERFACE_REMOVED static thread_local rocsolver_handle handle_helper; #else static thread_local rocsolver_handle handle_helper; diff --git a/src/lapack/backends/rocsolver/rocsolver_task.hpp b/src/lapack/backends/rocsolver/rocsolver_task.hpp index f6b110259..965484690 100644 --- a/src/lapack/backends/rocsolver/rocsolver_task.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_task.hpp @@ -36,8 +36,8 @@ // After Plugin Interface removal in DPC++ ur.hpp is the new include #if __has_include() #include -#ifndef ONEMKL_PI_INTERFACE_REMOVED -#define ONEMKL_PI_INTERFACE_REMOVED +#ifndef ONEMATH_PI_INTERFACE_REMOVED +#define ONEMATH_PI_INTERFACE_REMOVED #endif #elif __has_include() #include diff --git a/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp b/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp index f361687fe..539f46922 100644 --- a/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp @@ -23,7 +23,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT lapack_function_table_t mkl_lapack_table = { +extern "C" ONEMATH_EXPORT lapack_function_table_t mkl_lapack_table = { WRAPPER_VERSION, #define LAPACK_BACKEND rocsolver oneapi::mkl::lapack::rocsolver::gebrd, diff --git a/src/rng/CMakeLists.txt b/src/rng/CMakeLists.txt index 9f6775e72..bc5161b1f 100644 --- a/src/rng/CMakeLists.txt +++ b/src/rng/CMakeLists.txt @@ -29,11 +29,11 @@ target_include_directories(onemath_rng ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} $ ) -target_compile_options(onemath_rng PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(onemath_rng PRIVATE ${ONEMATH_BUILD_COPT}) set_target_properties(onemath_rng PROPERTIES POSITION_INDEPENDENT_CODE ON @@ -41,7 +41,7 @@ set_target_properties(onemath_rng PROPERTIES if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET onemath_rng SOURCES rng_loader.cpp) else() - target_link_libraries(onemath_rng PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(onemath_rng PUBLIC ONEMATH::SYCL::SYCL) endif() endif() diff --git a/src/rng/backends/curand/CMakeLists.txt b/src/rng/backends/curand/CMakeLists.txt index 3990f8086..1eee395e7 100644 --- a/src/rng/backends/curand/CMakeLists.txt +++ b/src/rng/backends/curand/CMakeLists.txt @@ -73,10 +73,10 @@ target_include_directories(${LIB_OBJ} ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin ${MKL_INCLUDE} - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL ONEMKL::cuRAND::cuRAND) +target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL ONEMATH::cuRAND::cuRAND) target_compile_features(${LIB_OBJ} PUBLIC cxx_std_11) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON diff --git a/src/rng/backends/curand/curand_wrappers.cpp b/src/rng/backends/curand/curand_wrappers.cpp index 92ef359eb..4b30b9a76 100644 --- a/src/rng/backends/curand/curand_wrappers.cpp +++ b/src/rng/backends/curand/curand_wrappers.cpp @@ -61,7 +61,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT rng_function_table_t mkl_rng_table = { +extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { WRAPPER_VERSION, oneapi::mkl::rng::curand::create_philox4x32x10, oneapi::mkl::rng::curand::create_philox4x32x10, oneapi::mkl::rng::curand::create_mrg32k3a, oneapi::mkl::rng::curand::create_mrg32k3a diff --git a/src/rng/backends/mklcpu/CMakeLists.txt b/src/rng/backends/mklcpu/CMakeLists.txt index b977dc28d..b96b8ae27 100644 --- a/src/rng/backends/mklcpu/CMakeLists.txt +++ b/src/rng/backends/mklcpu/CMakeLists.txt @@ -33,14 +33,14 @@ target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET ${LIB_OBJ} SOURCES ${SOURCES}) endif() -target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL MKL::MKL) +target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON @@ -50,7 +50,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) # Set oneMKL libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp b/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp index cf6d56cc3..04d4a7c7b 100644 --- a/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp +++ b/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT rng_function_table_t mkl_rng_table = { +extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { WRAPPER_VERSION, oneapi::mkl::rng::mklcpu::create_philox4x32x10, oneapi::mkl::rng::mklcpu::create_philox4x32x10, oneapi::mkl::rng::mklcpu::create_mrg32k3a, oneapi::mkl::rng::mklcpu::create_mrg32k3a diff --git a/src/rng/backends/mklgpu/CMakeLists.txt b/src/rng/backends/mklgpu/CMakeLists.txt index 01849fbb1..891aff8be 100644 --- a/src/rng/backends/mklgpu/CMakeLists.txt +++ b/src/rng/backends/mklgpu/CMakeLists.txt @@ -33,15 +33,15 @@ target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::RNG) - target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL MKL::MKL_SYCL::RNG) + target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_SYCL::RNG) else() - target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL MKL::MKL_DPCPP) + target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_DPCPP) endif() set_target_properties(${LIB_OBJ} PROPERTIES @@ -52,7 +52,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) # Set oneMKL libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp b/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp index 2d0c76cac..216aab951 100644 --- a/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp +++ b/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT rng_function_table_t mkl_rng_table = { +extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { WRAPPER_VERSION, oneapi::mkl::rng::mklgpu::create_philox4x32x10, oneapi::mkl::rng::mklgpu::create_philox4x32x10, oneapi::mkl::rng::mklgpu::create_mrg32k3a, oneapi::mkl::rng::mklgpu::create_mrg32k3a diff --git a/src/rng/backends/rocrand/CMakeLists.txt b/src/rng/backends/rocrand/CMakeLists.txt index f4417f3c1..50211fe02 100644 --- a/src/rng/backends/rocrand/CMakeLists.txt +++ b/src/rng/backends/rocrand/CMakeLists.txt @@ -67,10 +67,10 @@ add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) target_include_directories( ${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src - ${CMAKE_BINARY_DIR}/bin ${MKL_INCLUDE} ${ONEMKL_GENERATED_INCLUDE_PATH}) + ${CMAKE_BINARY_DIR}/bin ${MKL_INCLUDE} ${ONEMATH_GENERATED_INCLUDE_PATH}) target_link_libraries(${LIB_OBJ} PRIVATE roc::rocrand hip::host Threads::Threads) -target_link_libraries(${LIB_OBJ} PUBLIC ONEMKL::SYCL::SYCL) +target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL) target_compile_features(${LIB_OBJ} PUBLIC cxx_std_11) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON) diff --git a/src/rng/backends/rocrand/rocrand_wrappers.cpp b/src/rng/backends/rocrand/rocrand_wrappers.cpp index 7528b7c0f..5a399a6af 100644 --- a/src/rng/backends/rocrand/rocrand_wrappers.cpp +++ b/src/rng/backends/rocrand/rocrand_wrappers.cpp @@ -63,7 +63,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMKL_EXPORT rng_function_table_t mkl_rng_table = { +extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { WRAPPER_VERSION, oneapi::mkl::rng::rocrand::create_philox4x32x10, oneapi::mkl::rng::rocrand::create_philox4x32x10, oneapi::mkl::rng::rocrand::create_mrg32k3a, oneapi::mkl::rng::rocrand::create_mrg32k3a diff --git a/src/sparse_blas/CMakeLists.txt b/src/sparse_blas/CMakeLists.txt index cec5ad720..3b1d3d1af 100644 --- a/src/sparse_blas/CMakeLists.txt +++ b/src/sparse_blas/CMakeLists.txt @@ -27,11 +27,11 @@ if(BUILD_SHARED_LIBS) ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} $ ) - target_compile_options(onemath_sparse_blas PRIVATE ${ONEMKL_BUILD_COPT}) + target_compile_options(onemath_sparse_blas PRIVATE ${ONEMATH_BUILD_COPT}) set_target_properties(onemath_sparse_blas PROPERTIES POSITION_INDEPENDENT_CODE ON @@ -39,7 +39,7 @@ if(BUILD_SHARED_LIBS) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET onemath_sparse_blas SOURCES sparse_blas_loader.cpp) else() - target_link_libraries(onemath_sparse_blas PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(onemath_sparse_blas PUBLIC ONEMATH::SYCL::SYCL) endif() include(WarningsUtils) diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp index feaf0fdfc..117faa630 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HANDLES_HPP_ -#define _ONEMKL_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HANDLES_HPP_ +#ifndef _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HANDLES_HPP_ +#define _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HANDLES_HPP_ // MKLCPU and MKLGPU backends include // This include defines its own oneapi::mkl::sparse namespace with some of the @@ -77,4 +77,4 @@ inline auto get_internal_handle(matrix_handle_t handle) { } // namespace oneapi::mkl::sparse::detail -#endif // _ONEMKL_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HANDLES_HPP_ +#endif // _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HANDLES_HPP_ diff --git a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp index 42d8abbb0..f55c42a1d 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HELPER_HPP_ -#define _ONEMKL_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HELPER_HPP_ +#ifndef _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HELPER_HPP_ +#define _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HELPER_HPP_ #if __has_include() #include @@ -108,4 +108,4 @@ inline sycl::event collapse_dependencies(sycl::queue &queue, } // namespace oneapi::mkl::sparse::detail -#endif // _ONEMKL_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HELPER_HPP_ +#endif // _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HELPER_HPP_ diff --git a/src/sparse_blas/backends/mklcpu/CMakeLists.txt b/src/sparse_blas/backends/mklcpu/CMakeLists.txt index 0f78e2e04..de1cbf6a8 100644 --- a/src/sparse_blas/backends/mklcpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklcpu/CMakeLists.txt @@ -34,20 +34,20 @@ target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::SPARSE) target_link_libraries(${LIB_OBJ} - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL PUBLIC MKL::MKL_SYCL::SPARSE PRIVATE onemath_warnings ) else() target_link_libraries(${LIB_OBJ} - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL PUBLIC MKL::MKL_DPCPP PRIVATE onemath_warnings ) @@ -61,7 +61,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) #Set oneMKL libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/sparse_blas/backends/mklgpu/CMakeLists.txt b/src/sparse_blas/backends/mklgpu/CMakeLists.txt index 15fc23a4c..5d86d4f65 100644 --- a/src/sparse_blas/backends/mklgpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklgpu/CMakeLists.txt @@ -34,20 +34,20 @@ target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${ONEMKL_GENERATED_INCLUDE_PATH} + ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_compile_options(${LIB_OBJ} PRIVATE ${ONEMKL_BUILD_COPT}) +target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::SPARSE) target_link_libraries(${LIB_OBJ} - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL PUBLIC MKL::MKL_SYCL::SPARSE PRIVATE onemath_warnings ) else() target_link_libraries(${LIB_OBJ} - PUBLIC ONEMKL::SYCL::SYCL + PUBLIC ONEMATH::SYCL::SYCL PUBLIC MKL::MKL_DPCPP PRIVATE onemath_warnings ) @@ -61,7 +61,7 @@ target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) #Set oneMKL libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES - INTERFACE_LINK_LIBRARIES ONEMKL::SYCL::SYCL + INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL ) endif() diff --git a/src/sparse_blas/enum_data_types.hpp b/src/sparse_blas/enum_data_types.hpp index 26946facb..17f8b35b6 100644 --- a/src/sparse_blas/enum_data_types.hpp +++ b/src/sparse_blas/enum_data_types.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SRC_SPARSE_BLAS_ENUM_DATA_TYPES_HPP_ -#define _ONEMKL_SRC_SPARSE_BLAS_ENUM_DATA_TYPES_HPP_ +#ifndef _ONEMATH_SRC_SPARSE_BLAS_ENUM_DATA_TYPES_HPP_ +#define _ONEMATH_SRC_SPARSE_BLAS_ENUM_DATA_TYPES_HPP_ #include @@ -66,4 +66,4 @@ constexpr data_type get_data_type() { } // namespace oneapi::mkl::sparse::detail -#endif // _ONEMKL_SRC_SPARSE_BLAS_ENUM_DATA_TYPES_HPP_ +#endif // _ONEMATH_SRC_SPARSE_BLAS_ENUM_DATA_TYPES_HPP_ diff --git a/src/sparse_blas/function_table.hpp b/src/sparse_blas/function_table.hpp index e48f50ccc..591f89f9d 100644 --- a/src/sparse_blas/function_table.hpp +++ b/src/sparse_blas/function_table.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_FUNCTION_TABLE_HPP_ -#define _ONEMKL_SPARSE_BLAS_FUNCTION_TABLE_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_FUNCTION_TABLE_HPP_ +#define _ONEMATH_SPARSE_BLAS_FUNCTION_TABLE_HPP_ #include "oneapi/math/sparse_blas/types.hpp" #include "sparse_blas/macros.hpp" @@ -264,4 +264,4 @@ typedef struct { #undef DEFINE_COO_MATRIX_FUNCS #undef DEFINE_CSR_MATRIX_FUNCS -#endif // _ONEMKL_SPARSE_BLAS_FUNCTION_TABLE_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_FUNCTION_TABLE_HPP_ diff --git a/src/sparse_blas/generic_container.hpp b/src/sparse_blas/generic_container.hpp index 7e3d60359..0b0f87c22 100644 --- a/src/sparse_blas/generic_container.hpp +++ b/src/sparse_blas/generic_container.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SRC_SPARSE_BLAS_GENERIC_CONTAINER_HPP_ -#define _ONEMKL_SRC_SPARSE_BLAS_GENERIC_CONTAINER_HPP_ +#ifndef _ONEMATH_SRC_SPARSE_BLAS_GENERIC_CONTAINER_HPP_ +#define _ONEMATH_SRC_SPARSE_BLAS_GENERIC_CONTAINER_HPP_ #include #include @@ -331,4 +331,4 @@ sycl::event submit_release(sycl::queue& queue, T* ptr, const DependenciesT& depe } // namespace oneapi::mkl::sparse::detail -#endif // _ONEMKL_SRC_SPARSE_BLAS_GENERIC_CONTAINER_HPP_ +#endif // _ONEMATH_SRC_SPARSE_BLAS_GENERIC_CONTAINER_HPP_ diff --git a/src/sparse_blas/macros.hpp b/src/sparse_blas/macros.hpp index 7eba01390..9aa08503f 100644 --- a/src/sparse_blas/macros.hpp +++ b/src/sparse_blas/macros.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMKL_SPARSE_BLAS_MACROS_HPP_ -#define _ONEMKL_SPARSE_BLAS_MACROS_HPP_ +#ifndef _ONEMATH_SPARSE_BLAS_MACROS_HPP_ +#define _ONEMATH_SPARSE_BLAS_MACROS_HPP_ #define FOR_EACH_FP_TYPE(DEFINE_MACRO) \ DEFINE_MACRO(float, _rf); \ @@ -42,4 +42,4 @@ std::string(#PTR) + " must not be nullptr."); \ } -#endif // _ONEMKL_SPARSE_BLAS_MACROS_HPP_ +#endif // _ONEMATH_SPARSE_BLAS_MACROS_HPP_ diff --git a/src/sparse_blas/matrix_view_comparison.hpp b/src/sparse_blas/matrix_view_comparison.hpp index 31d2193b3..69ea94b7e 100644 --- a/src/sparse_blas/matrix_view_comparison.hpp +++ b/src/sparse_blas/matrix_view_comparison.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _ONEMKL_SRC_SPARSE_BLAS_MATRIX_VIEW_COMPARISON_HPP_ -#define _ONEMKL_SRC_SPARSE_BLAS_MATRIX_VIEW_COMPARISON_HPP_ +#ifndef _ONEMATH_SRC_SPARSE_BLAS_MATRIX_VIEW_COMPARISON_HPP_ +#define _ONEMATH_SRC_SPARSE_BLAS_MATRIX_VIEW_COMPARISON_HPP_ #include "oneapi/math/sparse_blas/matrix_view.hpp" @@ -33,4 +33,4 @@ inline bool operator!=(const oneapi::mkl::sparse::matrix_view& lhs, return !(lhs == rhs); } -#endif // _ONEMKL_SRC_SPARSE_BLAS_MATRIX_VIEW_COMPARISON_HPP_ \ No newline at end of file +#endif // _ONEMATH_SRC_SPARSE_BLAS_MATRIX_VIEW_COMPARISON_HPP_ \ No newline at end of file diff --git a/tests/unit_tests/CMakeLists.txt b/tests/unit_tests/CMakeLists.txt index 9d3f4da8d..baac3dc53 100644 --- a/tests/unit_tests/CMakeLists.txt +++ b/tests/unit_tests/CMakeLists.txt @@ -110,7 +110,7 @@ foreach(domain ${TEST_TARGET_DOMAINS}) if(BUILD_SHARED_LIBS) add_executable(test_main_${domain}_rt main_test.cpp) target_include_directories(test_main_${domain}_rt PUBLIC ${GTEST_INCLUDE_DIR}) - if(NOT ${ONEMKL_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") + if(NOT ${ONEMATH_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") target_compile_options(test_main_${domain}_rt PRIVATE -fsycl) endif() target_link_libraries(test_main_${domain}_rt PUBLIC @@ -118,7 +118,7 @@ foreach(domain ${TEST_TARGET_DOMAINS}) gtest_main ${CMAKE_DL_LIBS} ${${domain}_TEST_LINK} - ONEMKL::SYCL::SYCL + ONEMATH::SYCL::SYCL onemath ${${domain}_TEST_LIST_RT} ) @@ -129,67 +129,67 @@ foreach(domain ${TEST_TARGET_DOMAINS}) if(ENABLE_MKLCPU_BACKEND) add_dependencies(test_main_${domain}_ct onemath_${domain}_mklcpu) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_mklcpu) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_mklcpu) endif() if(ENABLE_MKLGPU_BACKEND) add_dependencies(test_main_${domain}_ct onemath_${domain}_mklgpu) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_mklgpu) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_mklgpu) endif() if(domain STREQUAL "blas" AND ENABLE_CUBLAS_BACKEND) add_dependencies(test_main_${domain}_ct onemath_${domain}_cublas) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_cublas) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_cublas) endif() if(domain STREQUAL "blas" AND ENABLE_ROCBLAS_BACKEND) add_dependencies(test_main_${domain}_ct onemath_${domain}_rocblas) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_rocblas) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_rocblas) endif() if(domain STREQUAL "blas" AND ENABLE_NETLIB_BACKEND) add_dependencies(test_main_${domain}_ct onemath_${domain}_netlib) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_netlib) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_netlib) endif() if(domain STREQUAL "blas" AND ENABLE_PORTBLAS_BACKEND) add_dependencies(test_main_${domain}_ct onemath_${domain}_portblas) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_portblas) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_portblas) endif() if(domain STREQUAL "lapack" AND ENABLE_CUSOLVER_BACKEND) add_dependencies(test_main_${domain}_ct onemath_${domain}_cusolver) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_cusolver) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_cusolver) endif() if(domain STREQUAL "lapack" AND ENABLE_ROCSOLVER_BACKEND) add_dependencies(test_main_${domain}_ct onemath_${domain}_rocsolver) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_rocsolver) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_rocsolver) endif() if(domain STREQUAL "rng" AND ENABLE_CURAND_BACKEND) add_dependencies(test_main_${domain}_ct onemath_${domain}_curand) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_curand) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_curand) endif() if(domain STREQUAL "rng" AND ENABLE_ROCRAND_BACKEND) add_dependencies(test_main_${domain}_ct onemath_${domain}_rocrand) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_rocrand) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_rocrand) endif() if(domain STREQUAL "dft" AND ENABLE_CUFFT_BACKEND) add_dependencies(test_main_${domain}_ct onemath_${domain}_cufft) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_${domain}_cufft) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_${domain}_cufft) endif() if(domain STREQUAL "dft" AND ENABLE_ROCFFT_BACKEND) add_dependencies(test_main_${domain}_ct onemath_dft_rocfft) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_dft_rocfft) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_dft_rocfft) endif() if(domain STREQUAL "dft" AND ENABLE_PORTFFT_BACKEND) add_dependencies(test_main_${domain}_ct onemath_dft_portfft) - list(APPEND ONEMKL_LIBRARIES_${domain} onemath_dft_portfft) + list(APPEND ONEMATH_LIBRARIES_${domain} onemath_dft_portfft) endif() target_link_libraries(test_main_${domain}_ct PUBLIC @@ -197,13 +197,13 @@ foreach(domain ${TEST_TARGET_DOMAINS}) gtest_main ${CMAKE_DL_LIBS} ${${domain}_TEST_LINK} - ${ONEMKL_LIBRARIES_${domain}} - ONEMKL::SYCL::SYCL + ${ONEMATH_LIBRARIES_${domain}} + ONEMATH::SYCL::SYCL ${${domain}_TEST_LIST_CT} ${${domain}_DEVICE_TEST_LIST_CT} ) - if(NOT ${ONEMKL_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") + if(NOT ${ONEMATH_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") target_link_options(test_main_${domain}_ct PUBLIC -fsycl-device-code-split=per_kernel) endif() diff --git a/tests/unit_tests/blas/CMakeLists.txt b/tests/unit_tests/blas/CMakeLists.txt index bd37d55f3..c80d0043a 100644 --- a/tests/unit_tests/blas/CMakeLists.txt +++ b/tests/unit_tests/blas/CMakeLists.txt @@ -19,8 +19,8 @@ # Add compile definitions for the reference libraries add_compile_definitions( - ONEMKL_REF_BLAS_LIBNAME="${ONEMKL_REF_BLAS_LIBNAME}" - ONEMKL_REF_CBLAS_LIBNAME="${ONEMKL_REF_CBLAS_LIBNAME}") + ONEMATH_REF_BLAS_LIBNAME="${ONEMATH_REF_BLAS_LIBNAME}" + ONEMATH_REF_CBLAS_LIBNAME="${ONEMATH_REF_CBLAS_LIBNAME}") add_subdirectory(level1) add_subdirectory(level2) diff --git a/tests/unit_tests/blas/batch/CMakeLists.txt b/tests/unit_tests/blas/batch/CMakeLists.txt index a3c6cb252..eb4e45935 100644 --- a/tests/unit_tests/blas/batch/CMakeLists.txt +++ b/tests/unit_tests/blas/batch/CMakeLists.txt @@ -41,7 +41,7 @@ if(BUILD_SHARED_LIBS) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET blas_batch_rt SOURCES ${BATCH_SOURCES}) else() - target_link_libraries(blas_batch_rt PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(blas_batch_rt PUBLIC ONEMATH::SYCL::SYCL) endif() endif() @@ -58,5 +58,5 @@ target_include_directories(blas_batch_ct if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET blas_batch_ct SOURCES ${BATCH_SOURCES}) else() - target_link_libraries(blas_batch_ct PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(blas_batch_ct PUBLIC ONEMATH::SYCL::SYCL) endif() diff --git a/tests/unit_tests/blas/extensions/CMakeLists.txt b/tests/unit_tests/blas/extensions/CMakeLists.txt index 47af93e96..e9e3b6ac5 100644 --- a/tests/unit_tests/blas/extensions/CMakeLists.txt +++ b/tests/unit_tests/blas/extensions/CMakeLists.txt @@ -42,7 +42,7 @@ if(BUILD_SHARED_LIBS) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET blas_extensions_rt SOURCES ${EXTENSIONS_SOURCES}) else() - target_link_libraries(blas_extensions_rt PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(blas_extensions_rt PUBLIC ONEMATH::SYCL::SYCL) endif() endif() @@ -59,5 +59,5 @@ target_include_directories(blas_extensions_ct if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET blas_extensions_ct SOURCES ${EXTENSIONS_SOURCES}) else() - target_link_libraries(blas_extensions_ct PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(blas_extensions_ct PUBLIC ONEMATH::SYCL::SYCL) endif() diff --git a/tests/unit_tests/blas/include/reference_blas_wrappers.hpp b/tests/unit_tests/blas/include/reference_blas_wrappers.hpp index e5b4c3a7d..90186d8d8 100644 --- a/tests/unit_tests/blas/include/reference_blas_wrappers.hpp +++ b/tests/unit_tests/blas/include/reference_blas_wrappers.hpp @@ -41,11 +41,11 @@ extern "C" { static LIB_TYPE h_libblas = NULL; static LIB_TYPE blas_library() { if (h_libblas == NULL) { - h_libblas = GET_LIB_HANDLE(ONEMKL_REF_BLAS_LIBNAME); + h_libblas = GET_LIB_HANDLE(ONEMATH_REF_BLAS_LIBNAME); if (h_libblas == NULL) { throw oneapi::mkl::library_not_found( "BLAS", "blas_library()", - std::string("failed to load BLAS library ") + ONEMKL_REF_BLAS_LIBNAME); + std::string("failed to load BLAS library ") + ONEMATH_REF_BLAS_LIBNAME); } } return h_libblas; @@ -54,11 +54,11 @@ static LIB_TYPE blas_library() { static LIB_TYPE h_libcblas = NULL; static LIB_TYPE cblas_library() { if (h_libcblas == NULL) { - h_libcblas = GET_LIB_HANDLE(ONEMKL_REF_CBLAS_LIBNAME); + h_libcblas = GET_LIB_HANDLE(ONEMATH_REF_CBLAS_LIBNAME); if (h_libcblas == NULL) { throw oneapi::mkl::library_not_found( "BLAS", "cblas_library()", - std::string("failed to load CBLAS library ") + ONEMKL_REF_CBLAS_LIBNAME); + std::string("failed to load CBLAS library ") + ONEMATH_REF_CBLAS_LIBNAME); } } return h_libcblas; diff --git a/tests/unit_tests/blas/level1/CMakeLists.txt b/tests/unit_tests/blas/level1/CMakeLists.txt index 6a67ef2e8..21d70657e 100644 --- a/tests/unit_tests/blas/level1/CMakeLists.txt +++ b/tests/unit_tests/blas/level1/CMakeLists.txt @@ -41,7 +41,7 @@ if(BUILD_SHARED_LIBS) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET blas_level1_rt SOURCES ${L1_SOURCES}) else() - target_link_libraries(blas_level1_rt PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(blas_level1_rt PUBLIC ONEMATH::SYCL::SYCL) endif() endif() @@ -58,5 +58,5 @@ target_include_directories(blas_level1_ct if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET blas_level1_ct SOURCES ${L1_SOURCES}) else() - target_link_libraries(blas_level1_ct PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(blas_level1_ct PUBLIC ONEMATH::SYCL::SYCL) endif() diff --git a/tests/unit_tests/blas/level2/CMakeLists.txt b/tests/unit_tests/blas/level2/CMakeLists.txt index 5ab82a8b5..8d7bc9d63 100644 --- a/tests/unit_tests/blas/level2/CMakeLists.txt +++ b/tests/unit_tests/blas/level2/CMakeLists.txt @@ -41,7 +41,7 @@ if(BUILD_SHARED_LIBS) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET blas_level2_rt SOURCES ${L2_SOURCES}) else() - target_link_libraries(blas_level2_rt PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(blas_level2_rt PUBLIC ONEMATH::SYCL::SYCL) endif() endif() @@ -58,6 +58,6 @@ target_include_directories(blas_level2_ct if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET blas_level2_ct SOURCES ${L2_SOURCES}) else() - target_link_libraries(blas_level2_ct PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(blas_level2_ct PUBLIC ONEMATH::SYCL::SYCL) endif() diff --git a/tests/unit_tests/blas/level3/CMakeLists.txt b/tests/unit_tests/blas/level3/CMakeLists.txt index ad6d2f273..ee2e8e3f8 100644 --- a/tests/unit_tests/blas/level3/CMakeLists.txt +++ b/tests/unit_tests/blas/level3/CMakeLists.txt @@ -41,7 +41,7 @@ if(BUILD_SHARED_LIBS) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET blas_level3_rt SOURCES ${L3_SOURCES}) else() - target_link_libraries(blas_level3_rt PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(blas_level3_rt PUBLIC ONEMATH::SYCL::SYCL) endif() endif() @@ -58,7 +58,7 @@ target_include_directories(blas_level3_ct if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET blas_level3_ct SOURCES ${L3_SOURCES}) else() - target_link_libraries(blas_level3_ct PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(blas_level3_ct PUBLIC ONEMATH::SYCL::SYCL) endif() diff --git a/tests/unit_tests/dft/include/compute_inplace.hpp b/tests/unit_tests/dft/include/compute_inplace.hpp index 718bb687c..d0f3bbb0f 100644 --- a/tests/unit_tests/dft/include/compute_inplace.hpp +++ b/tests/unit_tests/dft/include/compute_inplace.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef ONEMKL_COMPUTE_INPLACE_HPP -#define ONEMKL_COMPUTE_INPLACE_HPP +#ifndef ONEMATH_COMPUTE_INPLACE_HPP +#define ONEMATH_COMPUTE_INPLACE_HPP #include "compute_tester.hpp" #include @@ -211,4 +211,4 @@ int DFT_Test::test_in_place_USM() { return !::testing::Test::HasFailure(); } -#endif //ONEMKL_COMPUTE_INPLACE_HPP +#endif //ONEMATH_COMPUTE_INPLACE_HPP diff --git a/tests/unit_tests/dft/include/compute_inplace_real_real.hpp b/tests/unit_tests/dft/include/compute_inplace_real_real.hpp index d4af1a44a..410440a02 100644 --- a/tests/unit_tests/dft/include/compute_inplace_real_real.hpp +++ b/tests/unit_tests/dft/include/compute_inplace_real_real.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef ONEMKL_COMPUTE_INPLACE_REAL_REAL_HPP -#define ONEMKL_COMPUTE_INPLACE_REAL_REAL_HPP +#ifndef ONEMATH_COMPUTE_INPLACE_REAL_REAL_HPP +#define ONEMATH_COMPUTE_INPLACE_REAL_REAL_HPP #include "compute_tester.hpp" @@ -149,4 +149,4 @@ int DFT_Test::test_in_place_real_real_buffer() { } } -#endif //ONEMKL_COMPUTE_INPLACE_REAL_REAL_HPP +#endif //ONEMATH_COMPUTE_INPLACE_REAL_REAL_HPP diff --git a/tests/unit_tests/dft/include/compute_out_of_place.hpp b/tests/unit_tests/dft/include/compute_out_of_place.hpp index bcfd09dda..c99da10e1 100644 --- a/tests/unit_tests/dft/include/compute_out_of_place.hpp +++ b/tests/unit_tests/dft/include/compute_out_of_place.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef ONEMKL_COMPUTE_OUT_OF_PLACE_HPP -#define ONEMKL_COMPUTE_OUT_OF_PLACE_HPP +#ifndef ONEMATH_COMPUTE_OUT_OF_PLACE_HPP +#define ONEMATH_COMPUTE_OUT_OF_PLACE_HPP #include "compute_tester.hpp" #include @@ -187,4 +187,4 @@ int DFT_Test::test_out_of_place_USM() { return !::testing::Test::HasFailure(); } -#endif //ONEMKL_COMPUTE_OUT_OF_PLACE_HPP +#endif //ONEMATH_COMPUTE_OUT_OF_PLACE_HPP diff --git a/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp b/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp index fb3ecb4f2..6ca8da1aa 100644 --- a/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp +++ b/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef ONEMKL_COMPUTE_OUT_OF_PLACE_REAL_REAL_HPP -#define ONEMKL_COMPUTE_OUT_OF_PLACE_REAL_REAL_HPP +#ifndef ONEMATH_COMPUTE_OUT_OF_PLACE_REAL_REAL_HPP +#define ONEMATH_COMPUTE_OUT_OF_PLACE_REAL_REAL_HPP #include "compute_tester.hpp" @@ -157,4 +157,4 @@ int DFT_Test::test_out_of_place_real_real_buffer() { return !::testing::Test::HasFailure(); } -#endif //ONEMKL_COMPUTE_OUT_OF_PLACE_REAL_REAL_HPP +#endif //ONEMATH_COMPUTE_OUT_OF_PLACE_REAL_REAL_HPP diff --git a/tests/unit_tests/dft/include/compute_tester.hpp b/tests/unit_tests/dft/include/compute_tester.hpp index ea6a1b212..8b0737a9e 100644 --- a/tests/unit_tests/dft/include/compute_tester.hpp +++ b/tests/unit_tests/dft/include/compute_tester.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef ONEMKL_COMPUTE_TESTER_HPP -#define ONEMKL_COMPUTE_TESTER_HPP +#ifndef ONEMATH_COMPUTE_TESTER_HPP +#define ONEMATH_COMPUTE_TESTER_HPP #include @@ -151,4 +151,4 @@ struct DFT_Test { int test_out_of_place_real_real_USM(); }; -#endif //ONEMKL_COMPUTE_TESTER_HPP +#endif //ONEMATH_COMPUTE_TESTER_HPP diff --git a/tests/unit_tests/dft/include/parseval_check.hpp b/tests/unit_tests/dft/include/parseval_check.hpp index ece6f7d31..8b8f4dfc5 100644 --- a/tests/unit_tests/dft/include/parseval_check.hpp +++ b/tests/unit_tests/dft/include/parseval_check.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef ONEMKL_PARSEVAL_CHECK_HPP -#define ONEMKL_PARSEVAL_CHECK_HPP +#ifndef ONEMATH_PARSEVAL_CHECK_HPP +#define ONEMATH_PARSEVAL_CHECK_HPP #include #include @@ -78,4 +78,4 @@ bool parseval_check(std::size_t dft_len, const TypeFwd* in, TypeBwd* out, } return true; } -#endif // ONEMKL_PARSEVAL_CHECK_HPP +#endif // ONEMATH_PARSEVAL_CHECK_HPP diff --git a/tests/unit_tests/dft/include/reference_dft.hpp b/tests/unit_tests/dft/include/reference_dft.hpp index a80dee83e..448fc3363 100644 --- a/tests/unit_tests/dft/include/reference_dft.hpp +++ b/tests/unit_tests/dft/include/reference_dft.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef ONEMKL_REFERENCE_DFT_HPP -#define ONEMKL_REFERENCE_DFT_HPP +#ifndef ONEMATH_REFERENCE_DFT_HPP +#define ONEMATH_REFERENCE_DFT_HPP #include #include @@ -127,4 +127,4 @@ void reference_forward_dft(const std::vector &sizes, const TypeIn } } -#endif //ONEMKL_REFERENCE_DFT_HPP +#endif //ONEMATH_REFERENCE_DFT_HPP diff --git a/tests/unit_tests/dft/include/test_common.hpp b/tests/unit_tests/dft/include/test_common.hpp index b13723105..5b74d80c9 100644 --- a/tests/unit_tests/dft/include/test_common.hpp +++ b/tests/unit_tests/dft/include/test_common.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef ONEMKL_TEST_COMMON_HPP -#define ONEMKL_TEST_COMMON_HPP +#ifndef ONEMATH_TEST_COMMON_HPP +#define ONEMATH_TEST_COMMON_HPP #include #include @@ -388,4 +388,4 @@ class DFTParamsPrint { } }; -#endif //ONEMKL_TEST_COMMON_HPP +#endif //ONEMATH_TEST_COMMON_HPP diff --git a/tests/unit_tests/dft/source/CMakeLists.txt b/tests/unit_tests/dft/source/CMakeLists.txt index e8fc29eca..1ff996afb 100644 --- a/tests/unit_tests/dft/source/CMakeLists.txt +++ b/tests/unit_tests/dft/source/CMakeLists.txt @@ -34,7 +34,7 @@ if (BUILD_SHARED_LIBS) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET dft_source_rt SOURCES ${DFT_SOURCES}) else () - target_link_libraries(dft_source_rt PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(dft_source_rt PUBLIC ONEMATH::SYCL::SYCL) endif () target_link_libraries(dft_source_rt PRIVATE onemath_warnings) endif () @@ -51,7 +51,7 @@ target_include_directories(dft_source_ct if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET dft_source_ct SOURCES ${DFT_SOURCES}) else () - target_link_libraries(dft_source_ct PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(dft_source_ct PUBLIC ONEMATH::SYCL::SYCL) endif () target_link_libraries(dft_source_ct PRIVATE onemath_warnings) diff --git a/tests/unit_tests/include/test_helper.hpp b/tests/unit_tests/include/test_helper.hpp index 63d94a493..01b277cc9 100644 --- a/tests/unit_tests/include/test_helper.hpp +++ b/tests/unit_tests/include/test_helper.hpp @@ -73,8 +73,8 @@ if (d->get_info().size() == 0) \ GTEST_SKIP() << "Double precision is not supported on the device" -#if defined(ONEMKL_ENABLE_MKLCPU_BACKEND) || defined(ONEMKL_ENABLE_NETLIB_BACKEND) -#ifdef ONEMKL_ENABLE_MKLCPU_BACKEND +#if defined(ONEMATH_ENABLE_MKLCPU_BACKEND) || defined(ONEMATH_ENABLE_NETLIB_BACKEND) +#ifdef ONEMATH_ENABLE_MKLCPU_BACKEND #define TEST_RUN_INTELCPU_SELECT_NO_ARGS(q, func) \ func(oneapi::mkl::backend_selector{ q }) #define TEST_RUN_INTELCPU_SELECT(q, func, ...) \ @@ -88,7 +88,7 @@ #define TEST_RUN_INTELCPU_SELECT(q, func, ...) #endif -#ifdef ONEMKL_ENABLE_MKLGPU_BACKEND +#ifdef ONEMATH_ENABLE_MKLGPU_BACKEND #define TEST_RUN_INTELGPU_SELECT_NO_ARGS(q, func) \ func(oneapi::mkl::backend_selector{ q }) #define TEST_RUN_INTELGPU_SELECT(q, func, ...) \ @@ -98,55 +98,55 @@ #define TEST_RUN_INTELGPU_SELECT(q, func, ...) #endif -#ifdef ONEMKL_ENABLE_CUBLAS_BACKEND +#ifdef ONEMATH_ENABLE_CUBLAS_BACKEND #define TEST_RUN_NVIDIAGPU_CUBLAS_SELECT(q, func, ...) \ func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_NVIDIAGPU_CUBLAS_SELECT(q, func, ...) #endif -#ifdef ONEMKL_ENABLE_CUSOLVER_BACKEND +#ifdef ONEMATH_ENABLE_CUSOLVER_BACKEND #define TEST_RUN_NVIDIAGPU_CUSOLVER_SELECT(q, func, ...) \ func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_NVIDIAGPU_CUSOLVER_SELECT(q, func, ...) #endif -#ifdef ONEMKL_ENABLE_ROCBLAS_BACKEND +#ifdef ONEMATH_ENABLE_ROCBLAS_BACKEND #define TEST_RUN_AMDGPU_ROCBLAS_SELECT(q, func, ...) \ func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_AMDGPU_ROCBLAS_SELECT(q, func, ...) #endif -#ifdef ONEMKL_ENABLE_CURAND_BACKEND +#ifdef ONEMATH_ENABLE_CURAND_BACKEND #define TEST_RUN_NVIDIAGPU_CURAND_SELECT(q, func, ...) \ func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_NVIDIAGPU_CURAND_SELECT(q, func, ...) #endif -#ifdef ONEMKL_ENABLE_ROCRAND_BACKEND +#ifdef ONEMATH_ENABLE_ROCRAND_BACKEND #define TEST_RUN_AMDGPU_ROCRAND_SELECT(q, func, ...) \ func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_AMDGPU_ROCRAND_SELECT(q, func, ...) #endif -#ifdef ONEMKL_ENABLE_ROCSOLVER_BACKEND +#ifdef ONEMATH_ENABLE_ROCSOLVER_BACKEND #define TEST_RUN_AMDGPU_ROCSOLVER_SELECT(q, func, ...) \ func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_AMDGPU_ROCSOLVER_SELECT(q, func, ...) #endif -#ifdef ONEMKL_ENABLE_PORTBLAS_BACKEND +#ifdef ONEMATH_ENABLE_PORTBLAS_BACKEND #define TEST_RUN_PORTBLAS_SELECT(q, func, ...) \ func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_PORTBLAS_SELECT(q, func, ...) #endif -#ifdef ONEMKL_ENABLE_CUFFT_BACKEND +#ifdef ONEMATH_ENABLE_CUFFT_BACKEND #define TEST_RUN_NVIDIAGPU_CUFFT_SELECT_NO_ARGS(q, func) \ func(oneapi::mkl::backend_selector{ q }) #define TEST_RUN_NVIDIAGPU_CUFFT_SELECT(q, func, ...) \ @@ -156,7 +156,7 @@ #define TEST_RUN_NVIDIAGPU_CUFFT_SELECT(q, func, ...) #endif -#ifdef ONEMKL_ENABLE_ROCFFT_BACKEND +#ifdef ONEMATH_ENABLE_ROCFFT_BACKEND #define TEST_RUN_AMDGPU_ROCFFT_SELECT_NO_ARGS(q, func) \ func(oneapi::mkl::backend_selector{ q }) #define TEST_RUN_AMDGPU_ROCFFT_SELECT(q, func, ...) \ @@ -166,7 +166,7 @@ #define TEST_RUN_AMDGPU_ROCFFT_SELECT(q, func, ...) #endif -#ifdef ONEMKL_ENABLE_PORTFFT_BACKEND +#ifdef ONEMATH_ENABLE_PORTFFT_BACKEND #define TEST_RUN_PORTFFT_SELECT_NO_ARGS(q, func) \ func(oneapi::mkl::backend_selector{ q }) #define TEST_RUN_PORTFFT_SELECT(q, func, ...) \ @@ -330,10 +330,10 @@ static inline void *malloc_shared(size_t align, size_t size, sycl::device dev, s #ifdef _WIN64 return sycl::malloc_shared(size, dev, ctx); #else -#if defined(ONEMKL_ENABLE_CUBLAS_BACKEND) || defined(ONEMKL_ENABLE_ROCBLAS_BACKEND) +#if defined(ONEMATH_ENABLE_CUBLAS_BACKEND) || defined(ONEMATH_ENABLE_ROCBLAS_BACKEND) return sycl::aligned_alloc_shared(align, size, dev, ctx); #endif -#if !defined(ONEMKL_ENABLE_CUBLAS_BACKEND) && !defined(ONEMKL_ENABLE_ROCBLAS_BACKEND) +#if !defined(ONEMATH_ENABLE_CUBLAS_BACKEND) && !defined(ONEMATH_ENABLE_ROCBLAS_BACKEND) return sycl::malloc_shared(size, dev, ctx); #endif #endif @@ -344,10 +344,10 @@ static inline void *malloc_device(size_t align, size_t size, sycl::device dev, s #ifdef _WIN64 return sycl::malloc_device(size, dev, ctx); #else -#if defined(ONEMKL_ENABLE_CUBLAS_BACKEND) || defined(ONEMKL_ENABLE_ROCBLAS_BACKEND) +#if defined(ONEMATH_ENABLE_CUBLAS_BACKEND) || defined(ONEMATH_ENABLE_ROCBLAS_BACKEND) return sycl::aligned_alloc_device(align, size, dev, ctx); #endif -#if !defined(ONEMKL_ENABLE_CUBLAS_BACKEND) && !defined(ONEMKL_ENABLE_ROCBLAS_BACKEND) +#if !defined(ONEMATH_ENABLE_CUBLAS_BACKEND) && !defined(ONEMATH_ENABLE_ROCBLAS_BACKEND) return sycl::malloc_device(size, dev, ctx); #endif #endif diff --git a/tests/unit_tests/lapack/source/CMakeLists.txt b/tests/unit_tests/lapack/source/CMakeLists.txt index ea9d3140a..f1443440a 100644 --- a/tests/unit_tests/lapack/source/CMakeLists.txt +++ b/tests/unit_tests/lapack/source/CMakeLists.txt @@ -86,7 +86,7 @@ if(BUILD_SHARED_LIBS) PUBLIC ${CMAKE_BINARY_DIR}/bin $<$:${LAPACKE_INCLUDE}> ) - target_link_libraries(lapack_source_rt PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(lapack_source_rt PUBLIC ONEMATH::SYCL::SYCL) endif() add_library(lapack_source_ct OBJECT ${LAPACK_SOURCES}) @@ -99,4 +99,4 @@ target_include_directories(lapack_source_ct PUBLIC ${CMAKE_BINARY_DIR}/bin $<$:${LAPACKE_INCLUDE}> ) -target_link_libraries(lapack_source_ct PUBLIC ONEMKL::SYCL::SYCL) +target_link_libraries(lapack_source_ct PUBLIC ONEMATH::SYCL::SYCL) diff --git a/tests/unit_tests/main_test.cpp b/tests/unit_tests/main_test.cpp index da8a38216..18e6b07a7 100644 --- a/tests/unit_tests/main_test.cpp +++ b/tests/unit_tests/main_test.cpp @@ -112,33 +112,33 @@ int main(int argc, char** argv) { if (unique_devices.find(dev.get_info()) == unique_devices.end()) { unique_devices.insert(dev.get_info()); -#if !defined(ONEMKL_ENABLE_MKLCPU_BACKEND) && \ - !defined(ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_CPU) && \ - !defined(ONEMKL_ENABLE_PORTFFT_BACKEND) && \ - !defined(ONEMKL_ENABLE_NETLIB_BACKEND) +#if !defined(ONEMATH_ENABLE_MKLCPU_BACKEND) && \ + !defined(ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_CPU) && \ + !defined(ONEMATH_ENABLE_PORTFFT_BACKEND) && \ + !defined(ONEMATH_ENABLE_NETLIB_BACKEND) if (dev.is_cpu()) continue; #endif -#if !defined(ONEMKL_ENABLE_MKLGPU_BACKEND) && \ - !defined(ONEMKL_ENABLE_PORTBLAS_BACKEND_INTEL_GPU) && \ - !defined(ONEMKL_ENABLE_PORTFFT_BACKEND) +#if !defined(ONEMATH_ENABLE_MKLGPU_BACKEND) && \ + !defined(ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_GPU) && \ + !defined(ONEMATH_ENABLE_PORTFFT_BACKEND) if (dev.is_gpu() && vendor_id == INTEL_ID) continue; #endif -#if !defined(ONEMKL_ENABLE_CUBLAS_BACKEND) && \ - !defined(ONEMKL_ENABLE_CURAND_BACKEND) && \ - !defined(ONEMKL_ENABLE_CUSOLVER_BACKEND) && \ - !defined(ONEMKL_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU) && \ - !defined(ONEMKL_ENABLE_CUFFT_BACKEND) && !defined(ONEMKL_ENABLE_PORTFFT_BACKEND) +#if !defined(ONEMATH_ENABLE_CUBLAS_BACKEND) && \ + !defined(ONEMATH_ENABLE_CURAND_BACKEND) && \ + !defined(ONEMATH_ENABLE_CUSOLVER_BACKEND) && \ + !defined(ONEMATH_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU) && \ + !defined(ONEMATH_ENABLE_CUFFT_BACKEND) && !defined(ONEMATH_ENABLE_PORTFFT_BACKEND) if (dev.is_gpu() && vendor_id == NVIDIA_ID) continue; #endif -#if !defined(ONEMKL_ENABLE_ROCBLAS_BACKEND) && \ - !defined(ONEMKL_ENABLE_ROCRAND_BACKEND) && \ - !defined(ONEMKL_ENABLE_ROCSOLVER_BACKEND) && \ - !defined(ONEMKL_ENABLE_PORTBLAS_BACKEND_AMD_GPU) && \ - !defined(ONEMKL_ENABLE_ROCFFT_BACKEND) && \ - !defined(ONEMKL_ENABLE_PORTFFT_BACKEND) +#if !defined(ONEMATH_ENABLE_ROCBLAS_BACKEND) && \ + !defined(ONEMATH_ENABLE_ROCRAND_BACKEND) && \ + !defined(ONEMATH_ENABLE_ROCSOLVER_BACKEND) && \ + !defined(ONEMATH_ENABLE_PORTBLAS_BACKEND_AMD_GPU) && \ + !defined(ONEMATH_ENABLE_ROCFFT_BACKEND) && \ + !defined(ONEMATH_ENABLE_PORTFFT_BACKEND) if (dev.is_gpu() && vendor_id == AMD_ID) continue; #endif diff --git a/tests/unit_tests/rng/device/moments/CMakeLists.txt b/tests/unit_tests/rng/device/moments/CMakeLists.txt index 2da8033bf..e486955da 100644 --- a/tests/unit_tests/rng/device/moments/CMakeLists.txt +++ b/tests/unit_tests/rng/device/moments/CMakeLists.txt @@ -32,9 +32,9 @@ target_include_directories(rng_device_moments_ct if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET rng_device_moments_ct SOURCES ${MOMENTS_DEVICE_TESTS_SOURCES}) else() - target_link_libraries(rng_device_moments_ct PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(rng_device_moments_ct PUBLIC ONEMATH::SYCL::SYCL) endif() -if(NOT ${ONEMKL_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") +if(NOT ${ONEMATH_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") target_link_options(rng_device_moments_ct PUBLIC -fsycl -fsycl-device-code-split=per_kernel) endif() diff --git a/tests/unit_tests/rng/device/service/CMakeLists.txt b/tests/unit_tests/rng/device/service/CMakeLists.txt index 03d960e1a..f009a5c7e 100644 --- a/tests/unit_tests/rng/device/service/CMakeLists.txt +++ b/tests/unit_tests/rng/device/service/CMakeLists.txt @@ -32,9 +32,9 @@ target_include_directories(rng_device_service_ct if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET rng_device_service_ct SOURCES ${SERVICE_DEVICE_TESTS_SOURCES}) else() - target_link_libraries(rng_device_service_ct PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(rng_device_service_ct PUBLIC ONEMATH::SYCL::SYCL) endif() -if(NOT ${ONEMKL_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") +if(NOT ${ONEMATH_SYCL_IMPLEMENTATION} STREQUAL "hipsycl") target_link_options(rng_device_service_ct PUBLIC -fsycl -fsycl-device-code-split=per_kernel) endif() diff --git a/tests/unit_tests/rng/include/rng_test_common.hpp b/tests/unit_tests/rng/include/rng_test_common.hpp index d2bc4859f..15c32777b 100644 --- a/tests/unit_tests/rng/include/rng_test_common.hpp +++ b/tests/unit_tests/rng/include/rng_test_common.hpp @@ -92,7 +92,7 @@ class rng_test { } }; -#ifdef ONEMKL_ENABLE_CURAND_BACKEND // w/a for cuda backend hangs when there are several queues with different contexts +#ifdef ONEMATH_ENABLE_CURAND_BACKEND // w/a for cuda backend hangs when there are several queues with different contexts static sycl::device* previous_device = nullptr; static sycl::context* context = nullptr; diff --git a/tests/unit_tests/rng/service/CMakeLists.txt b/tests/unit_tests/rng/service/CMakeLists.txt index 8436ce9eb..f00f7ffaa 100644 --- a/tests/unit_tests/rng/service/CMakeLists.txt +++ b/tests/unit_tests/rng/service/CMakeLists.txt @@ -33,7 +33,7 @@ if(BUILD_SHARED_LIBS) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET rng_service_rt SOURCES ${SERVICE_TESTS_SOURCES}) else() - target_link_libraries(rng_service_rt PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(rng_service_rt PUBLIC ONEMATH::SYCL::SYCL) endif() endif() @@ -49,5 +49,5 @@ target_include_directories(rng_service_ct if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET rng_service_ct SOURCES ${SERVICE_TESTS_SOURCES}) else() - target_link_libraries(rng_service_ct PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(rng_service_ct PUBLIC ONEMATH::SYCL::SYCL) endif() diff --git a/tests/unit_tests/rng/statistics_check/CMakeLists.txt b/tests/unit_tests/rng/statistics_check/CMakeLists.txt index 244d33976..90ef555a5 100644 --- a/tests/unit_tests/rng/statistics_check/CMakeLists.txt +++ b/tests/unit_tests/rng/statistics_check/CMakeLists.txt @@ -33,7 +33,7 @@ if(BUILD_SHARED_LIBS) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET rng_statistics_rt SOURCES ${STATS_CHECK_SOURCES}) else() - target_link_libraries(rng_statistics_rt PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(rng_statistics_rt PUBLIC ONEMATH::SYCL::SYCL) endif() endif() @@ -49,5 +49,5 @@ target_include_directories(rng_statistics_ct if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET rng_statistics_ct SOURCES ${STATS_CHECK_SOURCES}) else() - target_link_libraries(rng_statistics_ct PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(rng_statistics_ct PUBLIC ONEMATH::SYCL::SYCL) endif() diff --git a/tests/unit_tests/sparse_blas/source/CMakeLists.txt b/tests/unit_tests/sparse_blas/source/CMakeLists.txt index 53b34e2ce..c0e86cbae 100644 --- a/tests/unit_tests/sparse_blas/source/CMakeLists.txt +++ b/tests/unit_tests/sparse_blas/source/CMakeLists.txt @@ -41,7 +41,7 @@ if (BUILD_SHARED_LIBS) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET spblas_source_rt SOURCES ${SPBLAS_SOURCES}) else () - target_link_libraries(spblas_source_rt PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(spblas_source_rt PUBLIC ONEMATH::SYCL::SYCL) endif () target_link_libraries(spblas_source_rt PRIVATE onemath_warnings) endif () @@ -58,6 +58,6 @@ target_include_directories(spblas_source_ct if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET spblas_source_ct SOURCES ${SPBLAS_SOURCES}) else () - target_link_libraries(spblas_source_ct PUBLIC ONEMKL::SYCL::SYCL) + target_link_libraries(spblas_source_ct PUBLIC ONEMATH::SYCL::SYCL) endif () target_link_libraries(spblas_source_ct PRIVATE onemath_warnings) From fc8129406d4c2912051887fb6c165e8513e44b00 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 8 Oct 2024 12:34:32 +0200 Subject: [PATCH 05/41] Clarify comment BLAS -> CBLAS --- src/blas/backends/netlib/netlib_common.hpp | 2 +- tests/unit_tests/blas/include/onemath_blas_helper.hpp | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/src/blas/backends/netlib/netlib_common.hpp b/src/blas/backends/netlib/netlib_common.hpp index d202d0faf..984f755da 100644 --- a/src/blas/backends/netlib/netlib_common.hpp +++ b/src/blas/backends/netlib/netlib_common.hpp @@ -42,7 +42,7 @@ namespace netlib { typedef enum { CblasFixOffset = 101, CblasColOffset = 102, CblasRowOffset = 103 } CBLAS_OFFSET; /** - * Helper methods for converting between onemath types and their BLAS + * Helper methods for converting between onemath types and their CBLAS * equivalents. */ diff --git a/tests/unit_tests/blas/include/onemath_blas_helper.hpp b/tests/unit_tests/blas/include/onemath_blas_helper.hpp index 081fb2234..3358c1c9c 100644 --- a/tests/unit_tests/blas/include/onemath_blas_helper.hpp +++ b/tests/unit_tests/blas/include/onemath_blas_helper.hpp @@ -27,7 +27,7 @@ typedef enum { CblasRowOffset = 101, CblasColOffset = 102, CblasFixOffset = 103 } CBLAS_OFFSET; /** - * Helper methods for converting between onemath types and their BLAS + * Helper methods for converting between onemath types and their CBLAS * equivalents. */ From 8ad288433a9fed5f0722ec75891e6b04314d59d1 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 8 Oct 2024 12:58:22 +0200 Subject: [PATCH 06/41] oneMKL (Interfaces)? -> oneMath --- .github/ISSUE_TEMPLATE/RFC.md | 4 +- .github/ISSUE_TEMPLATE/bug_report.md | 4 +- .github/pull_request_template.md | 2 +- .github/workflows/pr.yml | 8 +-- CMakeLists.txt | 20 +++---- LICENSE | 2 +- MAINTAINERS.md | 22 ++++---- SECURITY.md | 18 +++---- cmake/WarningsUtils.cmake | 2 +- cmake/oneMKLConfig.cmake | 2 +- docs/building_and_running_tests.rst | 2 +- .../building_the_project_with_adaptivecpp.rst | 8 +-- docs/conf.py.in | 4 +- docs/create_new_backend.rst | 52 +++++++++---------- docs/introduction.rst | 4 +- docs/using_onemath_with_cmake.rst | 42 +++++++-------- .../level3/gemm_usm_mklcpu_cublas.cpp | 2 +- .../complex_fwd_usm_mklcpu_cufft.cpp | 2 +- .../dft/run_time_dispatching/real_fwd_usm.cpp | 2 +- .../getrs_usm_mklcpu_cusolver.cpp | 2 +- .../lapack/run_time_dispatching/getrs_usm.cpp | 2 +- .../uniform_usm_mklcpu_curand.cpp | 2 +- examples/rng/device/uniform.cpp | 2 +- .../rng/run_time_dispatching/uniform_usm.cpp | 2 +- include/oneapi/math/detail/exceptions.hpp | 4 +- .../math/dft/detail/descriptor_impl.hpp | 4 +- include/oneapi/math/exceptions.hpp | 4 +- .../math/rng/device/detail/bits_impl.hpp | 8 +-- .../rng/device/detail/uniform_bits_impl.hpp | 4 +- .../oneapi/math/rng/device/distributions.hpp | 36 ++++++------- include/oneapi/math/sparse_blas/types.hpp | 2 +- scripts/generate_cmake.py | 4 +- src/CMakeLists.txt | 8 +-- src/blas/backends/cublas/CMakeLists.txt | 4 +- src/blas/backends/cublas/cublas_helper.hpp | 4 +- src/blas/backends/mklcpu/CMakeLists.txt | 4 +- src/blas/backends/mklgpu/CMakeLists.txt | 4 +- src/blas/backends/netlib/CMakeLists.txt | 4 +- src/blas/backends/portblas/CMakeLists.txt | 6 +-- .../backends/portblas/portblas_common.hpp | 10 ++-- src/blas/backends/rocblas/CMakeLists.txt | 4 +- src/blas/backends/rocblas/rocblas_helper.hpp | 4 +- src/dft/backends/cufft/CMakeLists.txt | 6 +-- src/dft/backends/cufft/commit.cpp | 2 +- src/dft/backends/mklcpu/CMakeLists.txt | 6 +-- src/dft/backends/mklgpu/CMakeLists.txt | 6 +-- src/dft/backends/mklgpu/backward.cpp | 2 +- src/dft/backends/mklgpu/commit.cpp | 12 ++--- src/dft/backends/mklgpu/forward.cpp | 7 ++- src/dft/backends/portfft/CMakeLists.txt | 6 +-- src/dft/backends/rocfft/CMakeLists.txt | 6 +-- src/dft/backends/rocfft/commit.cpp | 4 +- src/lapack/backends/cusolver/CMakeLists.txt | 4 +- .../backends/cusolver/cusolver_helper.hpp | 4 +- .../cusolver/cusolver_scope_handle.hpp | 2 +- src/lapack/backends/mklcpu/CMakeLists.txt | 6 +-- src/lapack/backends/mklgpu/CMakeLists.txt | 6 +-- src/lapack/backends/rocsolver/CMakeLists.txt | 4 +- .../backends/rocsolver/rocsolver_helper.hpp | 4 +- src/rng/backends/curand/CMakeLists.txt | 4 +- src/rng/backends/curand/philox4x32x10.cpp | 2 +- src/rng/backends/mklcpu/CMakeLists.txt | 6 +-- src/rng/backends/mklgpu/CMakeLists.txt | 6 +-- src/rng/backends/rocrand/CMakeLists.txt | 4 +- src/rng/backends/rocrand/philox4x32x10.cpp | 2 +- .../backends/mkl_common/mkl_handles.cxx | 2 +- .../backends/mkl_common/mkl_handles.hpp | 4 +- .../backends/mkl_common/mkl_spmm.cxx | 6 +-- .../backends/mkl_common/mkl_spmv.cxx | 2 +- .../backends/mkl_common/mkl_spsv.cxx | 2 +- .../backends/mklcpu/CMakeLists.txt | 6 +-- .../backends/mklgpu/CMakeLists.txt | 6 +-- 72 files changed, 234 insertions(+), 235 deletions(-) diff --git a/.github/ISSUE_TEMPLATE/RFC.md b/.github/ISSUE_TEMPLATE/RFC.md index 1fa6e4fb4..31946cd3d 100755 --- a/.github/ISSUE_TEMPLATE/RFC.md +++ b/.github/ISSUE_TEMPLATE/RFC.md @@ -9,7 +9,7 @@ assignees: '' # Summary Include a short summary of the request. Sections below provide guidance on what factors are considered important. Describe how new interface will meet -[library functionality guidelines](https://github.com/oneapi-src/oneMKL/blob/master/CONTRIBUTING.md#library-functionality-guidelines). +[library functionality guidelines](https://github.com/oneapi-src/oneMath/blob/master/CONTRIBUTING.md#library-functionality-guidelines). # Problem statement Describe the problem you are trying to solve with reasonable level of details. @@ -18,4 +18,4 @@ Describe the problem you are trying to solve with reasonable level of details. * The definition of the function including interface and semantics. How this interface will be extendable for different HW implementations. * What existing libraries have implementation of this function and can be used -under oneMKL interface. +under oneMath. diff --git a/.github/ISSUE_TEMPLATE/bug_report.md b/.github/ISSUE_TEMPLATE/bug_report.md index f6b78f782..24eb32da5 100755 --- a/.github/ISSUE_TEMPLATE/bug_report.md +++ b/.github/ISSUE_TEMPLATE/bug_report.md @@ -11,11 +11,11 @@ Provide a short summary of the issue. Sections below provide guidance on what factors are considered important to reproduce an issue. # Version -Report oneMKL version and githash. +Report oneMath version and githash. If it is a regression, report githash for the last known good revision. # Environment -oneMKL works with multiple HW and backend libraries and also depends on the +oneMath works with multiple HW and backend libraries and also depends on the compiler and build environment. Include the following information to help reproduce the issue: * HW you use diff --git a/.github/pull_request_template.md b/.github/pull_request_template.md index 6fda36ac8..976e9b23c 100755 --- a/.github/pull_request_template.md +++ b/.github/pull_request_template.md @@ -2,7 +2,7 @@ Please include a summary of the change. Please also include relevant motivation and context. See -[contribution guidelines](https://github.com/oneapi-src/oneMKL/blob/master/CONTRIBUTING.md) +[contribution guidelines](https://github.com/oneapi-src/oneMath/blob/master/CONTRIBUTING.md) for more details. If the change fixes an issue not documented in the project's Github issue tracker, please document all steps necessary to reproduce it. diff --git a/.github/workflows/pr.yml b/.github/workflows/pr.yml index 0da2153c6..51ca5f43a 100644 --- a/.github/workflows/pr.yml +++ b/.github/workflows/pr.yml @@ -27,15 +27,15 @@ jobs: domain: dft build_options: -DENABLE_PORTFFT_BACKEND=ON -DENABLE_MKLCPU_BACKEND=OFF test_options: -R 'DFT/CT/.*ComputeTests_in_place_COMPLEX.COMPLEX_SINGLE_in_place_buffer.sizes_8_batches_1*' - - config: MKL BLAS + - config: oneMath BLAS domain: blas build_options: -DREF_BLAS_ROOT=${PWD}/lapack/install - - config: MKL DFT + - config: oneMath DFT domain: dft - - config: MKL LAPACK + - config: oneMath LAPACK domain: lapack build_options: -DREF_LAPACK_ROOT=${PWD}/lapack/install - - config: MKL RNG + - config: oneMath RNG domain: rng name: unit tests ${{ matrix.config }} CPU steps: diff --git a/CMakeLists.txt b/CMakeLists.txt index e78525b68..54090c6b0 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -38,7 +38,7 @@ option(BUILD_SHARED_LIBS "Build dynamic libraries" ON) option(ENABLE_MKLCPU_BACKEND "Enable the Intel oneMKL CPU backend for supported interfaces" ON) option(ENABLE_MKLGPU_BACKEND "Enable the Intel oneMKL GPU backend for supported interfaces" ON) if(ENABLE_MKLCPU_BACKEND) - option(ENABLE_MKLCPU_THREAD_TBB "Enable the use of Intel TBB with the oneMKL CPU backend" ON) + option(ENABLE_MKLCPU_THREAD_TBB "Enable the use of Intel TBB with the oneMath CPU backend" ON) endif() # blas @@ -185,7 +185,7 @@ else() endif() endif() -project(oneMKL VERSION 0.5.0 LANGUAGES CXX) +project(oneMath VERSION 0.5.0 LANGUAGES CXX) # Override default CXX compile/link lines for Windows after project if(WIN32 AND ONEMATH_SYCL_IMPLEMENTATION STREQUAL "dpc++") @@ -229,7 +229,7 @@ else() endif() message(STATUS "TARGET_DOMAINS: ${TARGET_DOMAINS}") -# Include Intel oneMKL +# Include Intel(R) oneAPI Math Kernel Library (oneMKL) if(ENABLE_MKLGPU_BACKEND OR ENABLE_MKLCPU_BACKEND) set(MKL_ARCH intel64) set(MKL_INTERFACE ilp64) @@ -326,7 +326,7 @@ install(DIRECTORY include/ include(CMakePackageConfigHelpers) write_basic_package_version_file( - "${CMAKE_CURRENT_BINARY_DIR}/oneMKLConfigVersion.cmake" + "${CMAKE_CURRENT_BINARY_DIR}/oneMathConfigVersion.cmake" VERSION ${PROJECT_VERSION} COMPATIBILITY AnyNewerVersion ) @@ -335,21 +335,21 @@ export(EXPORT oneMathTargets FILE "${CMAKE_CURRENT_BINARY_DIR}/oneMathTargets.cmake" NAMESPACE ONEMATH:: ) -configure_file("${PROJECT_SOURCE_DIR}/cmake/oneMKLConfig.cmake" - "${CMAKE_CURRENT_BINARY_DIR}/oneMKLConfig.cmake" +configure_file("${PROJECT_SOURCE_DIR}/cmake/oneMathConfig.cmake" + "${CMAKE_CURRENT_BINARY_DIR}/oneMathConfig.cmake" COPYONLY ) set(config_package_location "lib/cmake/${PROJECT_NAME}") -install(EXPORT oneMKLTargets - FILE oneMKLTargets.cmake +install(EXPORT oneMathTargets + FILE oneMathTargets.cmake NAMESPACE MKL:: DESTINATION ${config_package_location} ) install( FILES - "${PROJECT_SOURCE_DIR}/cmake/oneMKLConfig.cmake" - "${CMAKE_CURRENT_BINARY_DIR}/oneMKLConfigVersion.cmake" + "${PROJECT_SOURCE_DIR}/cmake/oneMathConfig.cmake" + "${CMAKE_CURRENT_BINARY_DIR}/oneMathConfigVersion.cmake" DESTINATION ${config_package_location} COMPONENT Devel ) diff --git a/LICENSE b/LICENSE index 580630b67..203f8b0dc 100644 --- a/LICENSE +++ b/LICENSE @@ -203,7 +203,7 @@ limitations under the License. This distribution includes third party software ("third party programs"). This -third party software, even if included with the distribution of the oneMKL open +third party software, even if included with the distribution of the oneMath open source project, may be governed by separate license terms, including without limitation, third party license terms, and open source software license terms. These separate license terms govern your use of the third party programs as set diff --git a/MAINTAINERS.md b/MAINTAINERS.md index 98b996522..9594af0ea 100644 --- a/MAINTAINERS.md +++ b/MAINTAINERS.md @@ -1,10 +1,10 @@ # Introduction -This document defines roles in oneMKL Interfaces project. +This document defines roles in oneMath project. # Roles and responsibilities -oneMKL Interfaces project defines three main roles: +oneMath project defines three main roles: * [Contributor](#contributor) * [Domain maintainer](#domain-maintainer) * [Architecture maintainer](#architecture-maintainer) @@ -14,7 +14,7 @@ requirements and the nomination process. ## Contributor -A Contributor invests time and resources to improve oneMKL Interfaces project. +A Contributor invests time and resources to improve oneMath project. Anyone can become a Contributor by bringing value in one of the following ways: * Answer questions from community members. * Submit feedback to design proposals. @@ -72,12 +72,12 @@ The process of becoming a Domain maintainer is: | GitHub team name | Domain maintainers | :-----------|:------------| -| @oneapi-src/onemkl-blas-write | oneMKL Interfaces BLAS maintainers | -| @oneapi-src/onemkl-dft-write | oneMKL Interfaces DFT maintainers | -| @oneapi-src/onemkl-lapack-write) | oneMKL Interfaces LAPACK maintainers | -| @oneapi-src/onemkl-rng-write | oneMKL Interfaces RNG maintainers | -| @oneapi-src/onemkl-sparse-write | oneMKL Interfaces Sparse Algebra maintainers | -| @oneapi-src/onemkl-vm-write | oneMKL Interfaces Vector Math maintainers | +| @oneapi-src/onemath-blas-write | oneMath BLAS maintainers | +| @oneapi-src/onemath-dft-write | oneMath DFT maintainers | +| @oneapi-src/onemath-lapack-write | oneMath LAPACK maintainers | +| @oneapi-src/onemath-rng-write | oneMath RNG maintainers | +| @oneapi-src/onemath-sparse-write | oneMath Sparse Algebra maintainers | +| @oneapi-src/onemath-vm-write | oneMath Vector Math maintainers | ## Architecture Maintainer Architecture maintainers are the most established contributors who are responsible for the @@ -107,7 +107,7 @@ Privileges: * Can recommend Contributor or Domain maintainer to become Architecture maintainers. Process of becoming a maintainer: -1. A Contributor or Domain maintainer requests to join oneMKL Interfaces Architecture maintainers GitHub team -(@oneapi-src/onemkl-arch-write). +1. A Contributor or Domain maintainer requests to join oneMath Architecture maintainers GitHub team +(@oneapi-src/onemath-arch-write). 2. At least one of Architecture maintainers approves the request. diff --git a/SECURITY.md b/SECURITY.md index 480361d12..acaa1152e 100644 --- a/SECURITY.md +++ b/SECURITY.md @@ -2,7 +2,7 @@ As an open-source project, we understand the importance of and responsibility for security. This Security policy outlines our guidelines and procedures for ensuring the highest level of Security and trust for our users who consume -oneMKL Interfaces. +oneMath. ## Supported Versions We provide support for the [latest version][1] only. @@ -27,22 +27,22 @@ Along with the report, please include the following info: please provide details. ### When Should I Report a Vulnerability? -* You think you discovered a potential security vulnerability in oneMKL Interfaces. -* You are unsure how the potential vulnerability affects oneMKL Interfaces. +* You think you discovered a potential security vulnerability in oneMath. +* You are unsure how the potential vulnerability affects oneMath. * You think you discovered a vulnerability in another project or 3rd party -component on which oneMKL Interfaces depends. If the issue is not fixed in the 3rd party +component on which oneMath depends. If the issue is not fixed in the 3rd party component, try to report directly there first. ### When Should I NOT Report a Vulnerability? * You got an automated scan hit and are unable to provide details. -* You need help using oneMKL Interfaces for security. +* You need help using oneMath for security. * You need help applying security-related updates. * Your issue is not security-related. ## Security Reports Review Process Our goal is to respond quickly to your inquiry, and to coordinate a fix and disclosure with you. All confirmed security vulnerabilities will be addressed -according to severity level and impact on oneMKL Interfaces. Normally, security issues +according to severity level and impact on oneMath. Normally, security issues are fixed in the next planned release. ## Disclosure Policy @@ -60,6 +60,6 @@ If you have any suggestions on how this Policy could be improved, please submit an issue or a pull request to this repository. Please **do not** report potential vulnerabilities or security flaws via a pull request. -[1]: https://github.com/oneapi-src/oneMKL/releases/latest -[2]: https://github.com/oneapi-src/oneMKL/security/advisories/new -[3]: https://github.com/oneapi-src/oneMKL/security/advisories \ No newline at end of file +[1]: https://github.com/oneapi-src/oneMath/releases/latest +[2]: https://github.com/oneapi-src/oneMath/security/advisories/new +[3]: https://github.com/oneapi-src/oneMath/security/advisories \ No newline at end of file diff --git a/cmake/WarningsUtils.cmake b/cmake/WarningsUtils.cmake index 43e19d8ef..14eb274f5 100644 --- a/cmake/WarningsUtils.cmake +++ b/cmake/WarningsUtils.cmake @@ -45,4 +45,4 @@ message(VERBOSE "Domains with warnings enabled use: ${ONEMATH_WARNINGS}") target_compile_options(onemath_warnings INTERFACE ${ONEMATH_WARNINGS}) # Add the library to install package -install(TARGETS onemath_warnings EXPORT oneMKLTargets) +install(TARGETS onemath_warnings EXPORT oneMathTargets) diff --git a/cmake/oneMKLConfig.cmake b/cmake/oneMKLConfig.cmake index 5baf9024b..b274c611a 100644 --- a/cmake/oneMKLConfig.cmake +++ b/cmake/oneMKLConfig.cmake @@ -27,4 +27,4 @@ if(NOT ${SYCL_FOUND}) find_package(Compiler REQUIRED) endif() -include("${CMAKE_CURRENT_LIST_DIR}/oneMKLTargets.cmake") +include("${CMAKE_CURRENT_LIST_DIR}/oneMathTargets.cmake") diff --git a/docs/building_and_running_tests.rst b/docs/building_and_running_tests.rst index 4cf5c9187..da7ff30fd 100644 --- a/docs/building_and_running_tests.rst +++ b/docs/building_and_running_tests.rst @@ -28,7 +28,7 @@ following: cmake -DCMAKE_INSTALL_PREFIX=~/lapack -DCBLAS=True -DLAPACK=True -DLAPACKE=True -DBUILD_INDEX64=False -DBUILD_SHARED_LIBS=True .. cmake --build . -j --target install -and then used in oneMKL by setting ``-REF_BLAS_ROOT=/path/to/lapack/install`` +and then used in oneMath by setting ``-REF_BLAS_ROOT=/path/to/lapack/install`` and ``-DREF_LAPACK_ROOT=/path/to/lapack/install``. You can re-run tests without re-building the entire project. diff --git a/docs/building_the_project_with_adaptivecpp.rst b/docs/building_the_project_with_adaptivecpp.rst index 5be18fb5b..073652d01 100644 --- a/docs/building_the_project_with_adaptivecpp.rst +++ b/docs/building_the_project_with_adaptivecpp.rst @@ -20,13 +20,13 @@ Environment Setup #. Download and install the `required dependencies - `_ + `_ manually. Build Commands ############### -In most cases, building oneMKL Interfaces is as simple as setting the compiler and +In most cases, building oneMath is as simple as setting the compiler and selecting the desired backends to build with. On Linux (other OSes are not supported with the AdaptiveCpp compiler): @@ -48,9 +48,9 @@ On Linux (other OSes are not supported with the AdaptiveCpp compiler): Backends should be enabled by setting ``-DENABLE__BACKEND=True`` for each desired backend. By default, the ``MKLGPU`` and ``MKLCPU`` backends are enabled, but ``MKLGPU`` must be disabled with AdaptiveCpp. The supported -backends for the compilers are given in the table at `oneMKL supported +backends for the compilers are given in the table at `oneMath supported configurations table -`_, +`_, and the CMake option names are given in the table below. Some backends may require additional parameters to be set. See the relevant section below for additional guidance. The target architectures must be specified with diff --git a/docs/conf.py.in b/docs/conf.py.in index d874dbab7..2038b3f4f 100644 --- a/docs/conf.py.in +++ b/docs/conf.py.in @@ -19,7 +19,7 @@ # -- Project information ----------------------------------------------------- -project = 'oneAPI Math Kernel Library Interfaces' +project = 'oneMath' copyright = '2020-2022, Intel Corporation' author = 'Intel Corporation' @@ -90,7 +90,7 @@ html_favicon = f'{static_dir}/favicons.png' # Theme options html_theme_options = { -'repository_url': 'https://github.com/oneapi-src/oneMKL', +'repository_url': 'https://github.com/oneapi-src/oneMath', 'path_to_docs': 'docs', 'use_issues_button': True, 'use_edit_page_button': True, diff --git a/docs/create_new_backend.rst b/docs/create_new_backend.rst index 98c9d8877..6cdebe394 100644 --- a/docs/create_new_backend.rst +++ b/docs/create_new_backend.rst @@ -3,12 +3,12 @@ .. _create_backend_wrappers: -Integrating a Third-Party Library to oneAPI Math Kernel Library (oneMKL) Interfaces +Integrating a Third-Party Library to oneAPI Math Kernel Library (oneMath) Interfaces ==================================================================================== -This step-by-step tutorial provides examples for enabling new third-party libraries in oneMKL. +This step-by-step tutorial provides examples for enabling new third-party libraries in oneMath. -oneMKL has a header-based implementation of the interface layer (``include`` directory) and a source-based implementation of the backend layer for each third-party library (``src`` directory). To enable a third-party library, you must update both parts of oneMKL and integrate the new third-party library to the oneMKL build and test systems. +oneMath has a header-based implementation of the interface layer (``include`` directory) and a source-based implementation of the backend layer for each third-party library (``src`` directory). To enable a third-party library, you must update both parts of oneMath and integrate the new third-party library to the oneMath build and test systems. For the new backend library and header naming please use the following template: @@ -38,7 +38,7 @@ If there is no need for multiple wrappers only ```` and ``<3rd-party lib For each new backend library, you should create the following two header files: * Header file with a declaration of entry points to the new third-party library wrappers -* Compiler-time dispatching interface (see `oneMKL Usage Models <../README.md#supported-usage-models>`_) for new third-party libraries +* Compiler-time dispatching interface (see `oneMath Usage Models <../README.md#supported-usage-models>`_) for new third-party libraries **Header File Example**: command to generate the header file with a declaration of BLAS entry points in the oneapi::mkl::newlib namespace @@ -95,37 +95,37 @@ Code snippet of the generated header file ``include/oneapi/math/blas/detail/newl 2. Integrate Header Files ------------------------- -Below you can see structure of oneMKL top-level include directory: +Below you can see structure of oneMath top-level include directory: :: include/ oneapi/ math/ - math.hpp -> oneMKL spec APIs - types.hpp -> oneMKL spec types - blas.hpp -> oneMKL BLAS APIs w/ pre-check/dispatching/post-check + math.hpp -> oneMath spec APIs + types.hpp -> oneMath spec types + blas.hpp -> oneMath BLAS APIs w/ pre-check/dispatching/post-check detail/ -> implementation specific header files - exceptions.hpp -> oneMKL exception classes - backends.hpp -> list of oneMKL backends + exceptions.hpp -> oneMath exception classes + backends.hpp -> list of oneMath backends backends_table.hpp -> table of backend libraries for each domain and device get_device_id.hpp -> function to query device information from queue for Run-time dispatching blas/ - predicates.hpp -> oneMKL BLAS pre-check post-check + predicates.hpp -> oneMath BLAS pre-check post-check detail/ -> BLAS domain specific implementation details - blas_loader.hpp -> oneMKL Run-time BLAS API - blas_ct_templates.hpp -> oneMKL Compile-time BLAS API general templates + blas_loader.hpp -> oneMath Run-time BLAS API + blas_ct_templates.hpp -> oneMath Compile-time BLAS API general templates cublas/ - blas_ct.hpp -> oneMKL Compile-time BLAS API template instantiations for + blas_ct.hpp -> oneMath Compile-time BLAS API template instantiations for onemath_blas_cublas.hpp -> backend wrappers library API mklcpu/ - blas_ct.hpp -> oneMKL Compile-time BLAS API template instantiations for + blas_ct.hpp -> oneMath Compile-time BLAS API template instantiations for onemath_blas_mklcpu.hpp -> backend wrappers library API / / -To integrate the new third-party library to a oneMKL header-based part, following files from this structure should be updated: +To integrate the new third-party library to a oneMath header-based part, following files from this structure should be updated: * ``include/oneapi/math/detail/backends.hpp``: add the new backend @@ -179,7 +179,7 @@ To integrate the new third-party library to a oneMKL header-based part, followin + if (queue.is_host()) + device_id=device::newdevice; -* ``include/oneapi/math/blas.hpp``: include the generated header file for the compile-time dispatching interface (see `oneMKL Usage Models <../README.md#supported-usage-models>`_) +* ``include/oneapi/math/blas.hpp``: include the generated header file for the compile-time dispatching interface (see `oneMath Usage Models <../README.md#supported-usage-models>`_) **Example**: add ``include/oneapi/math/blas/detail/newlib/blas_ct.hpp`` generated at the `1. Create Header Files`_ step @@ -197,20 +197,20 @@ The new files generated at the `1. Create Header Files`_ step result in the foll include/ oneapi/ math/ - blas.hpp -> oneMKL BLAS APIs w/ pre-check/dispatching/post-check + blas.hpp -> oneMath BLAS APIs w/ pre-check/dispatching/post-check blas/ - predicates.hpp -> oneMKL BLAS pre-check post-check + predicates.hpp -> oneMath BLAS pre-check post-check detail/ -> BLAS domain specific implementation details - blas_loader.hpp -> oneMKL Run-time BLAS API - blas_ct_templates.hpp -> oneMKL Compile-time BLAS API general templates + blas_loader.hpp -> oneMath Run-time BLAS API + blas_ct_templates.hpp -> oneMath Compile-time BLAS API general templates cublas/ - blas_ct.hpp -> oneMKL Compile-time BLAS API template instantiations for + blas_ct.hpp -> oneMath Compile-time BLAS API template instantiations for onemath_blas_cublas.hpp -> backend wrappers library API mklcpu/ - blas_ct.hpp -> oneMKL Compile-time BLAS API template instantiations for + blas_ct.hpp -> oneMath Compile-time BLAS API template instantiations for onemath_blas_mklcpu.hpp -> backend wrappers library API + newlib/ - + blas_ct.hpp -> oneMKL Compile-time BLAS API template instantiations for + + blas_ct.hpp -> oneMath Compile-time BLAS API template instantiations for + onemath_blas_newlib.hpp -> backend wrappers library API / / @@ -219,7 +219,7 @@ The new files generated at the `1. Create Header Files`_ step result in the foll 3. Create Wrappers ------------------ -Wrappers convert Data Parallel C++ (DPC++) input data types to third-party library data types and call corresponding implementation from the third-party library. Wrappers for each third-party library are built to separate oneMKL backend libraries. The ``libonemkl.so`` dispatcher library loads the wrappers at run-time if you are using the interface for run-time dispatching, or you will link with them directly in case you are using the interface for compile-time dispatching (for more information see `oneMKL Usage Models <../README.md#supported-usage-models>`_). +Wrappers convert Data Parallel C++ (DPC++) input data types to third-party library data types and call corresponding implementation from the third-party library. Wrappers for each third-party library are built to separate oneMath backend libraries. The ``libonemath.so`` dispatcher library loads the wrappers at run-time if you are using the interface for run-time dispatching, or you will link with them directly in case you are using the interface for compile-time dispatching (for more information see `oneMath Usage Models <../README.md#supported-usage-models>`_). All wrappers and dispatcher library implementations are in the ``src`` directory: @@ -327,7 +327,7 @@ Updated structure of the ``src`` folder with the ``newlib`` wrappers: 4. Integrate Wrappers to the Build System ----------------------------------------- -Here is the list of files that should be created/updated to integrate the new wrappers for the third-party library to the oneMKL build system: +Here is the list of files that should be created/updated to integrate the new wrappers for the third-party library to the oneMath build system: * Add the new option ``ENABLE_XXX_BACKEND`` for the new third-party library to the top of the ``CMakeList.txt`` file. diff --git a/docs/introduction.rst b/docs/introduction.rst index 6bf86fc56..b055f8519 100644 --- a/docs/introduction.rst +++ b/docs/introduction.rst @@ -3,7 +3,7 @@ Introduction ============ -oneMKL Interfaces is an open-source implementation of oneMKL Data Parallel C++ -(DPC++) interfaces according to the `oneMKL specification `_ +oneMath is an open-source implementation of oneMath Data Parallel C++ +(DPC++) interfaces according to the `oneMath specification `_ that can work with multiple devices (backends) using device-specific libraries underneath. diff --git a/docs/using_onemath_with_cmake.rst b/docs/using_onemath_with_cmake.rst index b1c80b7a8..acc7a0a67 100644 --- a/docs/using_onemath_with_cmake.rst +++ b/docs/using_onemath_with_cmake.rst @@ -1,39 +1,39 @@ -.. _using_onemath_interface_library_with_cmake: +.. _using_onemath_library_with_cmake: -Using the oneMKL Interfaces in your project with CMake -============================================================= +Using oneMath in your project with CMake +======================================== -The CMake build tool can help you use oneMKL Interfaces in your own project. -Instead of manually linking and including directories, you can use the CMake targets -exported by the oneMKL Interfaces project. You can use oneMKL in one of two -forms, with the target names depending on the approach taken: +The CMake build tool can help you use oneMath in your own project. Instead of +manually linking and including directories, you can use the CMake targets +exported by the oneMath project. You can use oneMath in one of two forms, with +the target names depending on the approach taken: * you can use a previously installed copy, either from a binary distribution or built from source. This can be imported using CMake's ``find_package`` command. See the section `using_from_installed_binary`_. -* or you can have CMake automatically download and build oneMKL as part of the +* or you can have CMake automatically download and build oneMath as part of the build process using CMake's FetchContent_ functionality. See the section `using_with_fetchcontent`_. .. _using_from_installed_binary: -Using an installed oneMKL Interfaces -#################################### +Using an installed oneMath +########################## -If the oneMKL Interfaces have been previously installed, either by building from -source or as a distributed binary, they can be consumed using CMake using -``find_package(oneMKL REQUIRED)``. The compiler used for the target library or -application should match that used to build oneMKL Interfaces. +If oneMath has been previously installed, either by building from source or as a +distributed binary, they can be consumed using CMake using +``find_package(oneMath REQUIRED)``. The compiler used for the target library or +application should match that used to build oneMath. For example: .. code-block:: cmake - find_package(oneMKL REQUIRED) + find_package(oneMath REQUIRED) target_link_libraries(myTarget PRIVATE MKL::onemath) -Different targets can be used depending on the requirements of oneMKL. +Different targets can be used depending on the requirements of oneMath. To link against the entire library, the ``MKL::onemath`` target should be used. For specific domains, ``MKL::onemath_`` should be used. And for specific backends, ``MKL::onemath__`` should be used. @@ -47,7 +47,7 @@ during the build. To check for the existence of backends, CMake's ``if(TARGET if(TARGET MKL::onemath_dft_cufft) target_link_libraries(myTarget PRIVATE MKL::onemath_dft_cufft) else() - message(FATAL_ERROR "oneMKL Interfaces was not built with CuFFT backend") + message(FATAL_ERROR "oneMath was not built with CuFFT backend") endif() .. _using_with_fetchcontent: @@ -57,7 +57,7 @@ Using CMake's FetchContent The FetchContent_ functionality of CMake can be used to download, build and -install oneMKL Interfaces as part of the build. +install oneMath as part of the build. For example: @@ -68,11 +68,11 @@ For example: set(BUILD_EXAMPLES False) set(ENABLE__BACKEND True) FetchContent_Declare( - onemath_interface_library - GIT_REPOSITORY https://github.com/oneapi-src/oneMKL.git + onemath_library + GIT_REPOSITORY https://github.com/oneapi-src/oneMath.git GIT_TAG develop ) - FetchContent_MakeAvailable(onemath_interface_library) + FetchContent_MakeAvailable(onemath_library) target_link_libraries(myTarget PRIVATE onemath) diff --git a/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp b/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp index a8e5058e6..d16821288 100644 --- a/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp +++ b/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp @@ -41,7 +41,7 @@ #include #include -// oneMKL/SYCL includes +// oneMath/SYCL includes #if __has_include() #include #else diff --git a/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp b/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp index c7cc624b0..ce030e6c8 100644 --- a/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp +++ b/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp @@ -20,7 +20,7 @@ // STL includes #include -// oneMKL/SYCL includes +// oneMath/SYCL includes #if __has_include() #include #else diff --git a/examples/dft/run_time_dispatching/real_fwd_usm.cpp b/examples/dft/run_time_dispatching/real_fwd_usm.cpp index c5a8d50f2..f14eb077c 100644 --- a/examples/dft/run_time_dispatching/real_fwd_usm.cpp +++ b/examples/dft/run_time_dispatching/real_fwd_usm.cpp @@ -21,7 +21,7 @@ #include #include -// oneMKL/SYCL includes +// oneMath/SYCL includes #if __has_include() #include #else diff --git a/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp b/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp index 9bb8cc646..e824789c0 100644 --- a/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp +++ b/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp @@ -35,7 +35,7 @@ #include #include -// oneMKL/SYCL includes +// oneMath/SYCL includes #if __has_include() #include #else diff --git a/examples/lapack/run_time_dispatching/getrs_usm.cpp b/examples/lapack/run_time_dispatching/getrs_usm.cpp index dde358f16..ded21cc22 100644 --- a/examples/lapack/run_time_dispatching/getrs_usm.cpp +++ b/examples/lapack/run_time_dispatching/getrs_usm.cpp @@ -36,7 +36,7 @@ #include #include -// oneMKL/SYCL includes +// oneMath/SYCL includes #if __has_include() #include #else diff --git a/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp b/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp index 5698931d8..305e10da0 100644 --- a/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp +++ b/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp @@ -36,7 +36,7 @@ #include #include -// oneMKL/SYCL includes +// oneMath/SYCL includes #if __has_include() #include #else diff --git a/examples/rng/device/uniform.cpp b/examples/rng/device/uniform.cpp index 9ccd37534..156afc1ff 100644 --- a/examples/rng/device/uniform.cpp +++ b/examples/rng/device/uniform.cpp @@ -30,7 +30,7 @@ #include #include -// oneMKL/SYCL includes +// oneMath/SYCL includes #if __has_include() #include #else diff --git a/examples/rng/run_time_dispatching/uniform_usm.cpp b/examples/rng/run_time_dispatching/uniform_usm.cpp index 942ad5f11..d179273ee 100644 --- a/examples/rng/run_time_dispatching/uniform_usm.cpp +++ b/examples/rng/run_time_dispatching/uniform_usm.cpp @@ -36,7 +36,7 @@ #include #include -// oneMKL/SYCL includes +// oneMath/SYCL includes #if __has_include() #include #else diff --git a/include/oneapi/math/detail/exceptions.hpp b/include/oneapi/math/detail/exceptions.hpp index 7d77d1f43..f70321a8d 100644 --- a/include/oneapi/math/detail/exceptions.hpp +++ b/include/oneapi/math/detail/exceptions.hpp @@ -17,7 +17,7 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -// These are oneAPI Math Kernel Library (oneMKL) Interfaces specific exceptions +// These are oneMath specific exceptions with no equivalent in Intel(R) oneMKL #ifndef _ONEMATH_DETAIL_EXCEPTIONS_HPP_ #define _ONEMATH_DETAIL_EXCEPTIONS_HPP_ @@ -50,7 +50,7 @@ class specification_mismatch : public oneapi::mkl::exception { specification_mismatch(const std::string &info = "") : oneapi::mkl::exception( "", "", - ((info.length() != 0) ? info : "Loaded oneMKL specification version mismatch")) {} + ((info.length() != 0) ? info : "Loaded oneMath specification version mismatch")) {} }; } // namespace mkl diff --git a/include/oneapi/math/dft/detail/descriptor_impl.hpp b/include/oneapi/math/dft/detail/descriptor_impl.hpp index 174c1a22d..4b1841989 100644 --- a/include/oneapi/math/dft/detail/descriptor_impl.hpp +++ b/include/oneapi/math/dft/detail/descriptor_impl.hpp @@ -55,9 +55,9 @@ class descriptor { // Syntax for d-dimensional DFT descriptor(std::vector dimensions); - // Copy operations are included in the oneAPI oneMKL specification, but not yet + // Copy operations are included in the oneAPI oneMath specification, but not yet // implemented here. If you need copies, please open an issue at - // https://github.com/oneapi-src/oneMKL/issues + // https://github.com/oneapi-src/oneMath/issues descriptor(descriptor&&); diff --git a/include/oneapi/math/exceptions.hpp b/include/oneapi/math/exceptions.hpp index 1a5a7addf..3a89ba56c 100644 --- a/include/oneapi/math/exceptions.hpp +++ b/include/oneapi/math/exceptions.hpp @@ -30,7 +30,7 @@ #include "oneapi/math/types.hpp" -// These are oneAPI oneMKL Specification exceptions +// These are oneAPI oneMath Specification exceptions namespace oneapi { namespace mkl { @@ -40,7 +40,7 @@ class exception : public std::exception { public: exception(const std::string &domain, const std::string &function, const std::string &info = "") : std::exception() { - msg_ = std::string("oneMKL: ") + domain + + msg_ = std::string("oneMath: ") + domain + ((domain.length() != 0 && function.length() != 0) ? "/" : "") + function + ((info.length() != 0) ? (((domain.length() + function.length() != 0) ? ": " : "") + info) diff --git a/include/oneapi/math/rng/device/detail/bits_impl.hpp b/include/oneapi/math/rng/device/detail/bits_impl.hpp index aa68956d6..83d6831a1 100644 --- a/include/oneapi/math/rng/device/detail/bits_impl.hpp +++ b/include/oneapi/math/rng/device/detail/bits_impl.hpp @@ -33,7 +33,7 @@ class distribution_base> { typename std::conditional>::type>::type { static_assert(std::is_same::value, - "oneMKL: bits works only with std::uint32_t"); + "oneMath: bits works only with std::uint32_t"); return engine.generate(); } @@ -43,7 +43,7 @@ class distribution_base> { typename std::conditional>::type>::type { static_assert(std::is_same::value, - "oneMKL: bits for mcg59 works only with std::uint64_t"); + "oneMath: bits for mcg59 works only with std::uint64_t"); return engine.generate_bits(); } @@ -52,7 +52,7 @@ class distribution_base> { UIntType>::type generate_single(EngineType& engine) { static_assert(std::is_same::value, - "oneMKL: bits works only with std::uint32_t"); + "oneMath: bits works only with std::uint32_t"); return engine.generate_single(); } @@ -61,7 +61,7 @@ class distribution_base> { UIntType>::type generate_single(EngineType& engine) { static_assert(std::is_same::value, - "oneMKL: bits for mcg59 works only with std::uint64_t"); + "oneMath: bits for mcg59 works only with std::uint64_t"); return engine.generate_single(); } }; diff --git a/include/oneapi/math/rng/device/detail/uniform_bits_impl.hpp b/include/oneapi/math/rng/device/detail/uniform_bits_impl.hpp index cd3cd2eed..1f9c8a717 100644 --- a/include/oneapi/math/rng/device/detail/uniform_bits_impl.hpp +++ b/include/oneapi/math/rng/device/detail/uniform_bits_impl.hpp @@ -33,7 +33,7 @@ class distribution_base> { sycl::vec>::type { static_assert(std::is_same>::value || std::is_same>::value, - "oneMKL: uniform_bits works only with philox4x32x10/mcg59 engines"); + "oneMath: uniform_bits works only with philox4x32x10/mcg59 engines"); return engine.template generate_uniform_bits(); } @@ -41,7 +41,7 @@ class distribution_base> { UIntType generate_single(EngineType& engine) { static_assert(std::is_same>::value || std::is_same>::value, - "oneMKL: uniform_bits works only with philox4x32x10/mcg59 engines"); + "oneMath: uniform_bits works only with philox4x32x10/mcg59 engines"); return engine.template generate_single_uniform_bits(); } }; diff --git a/include/oneapi/math/rng/device/distributions.hpp b/include/oneapi/math/rng/device/distributions.hpp index 28bb439f1..2f5082886 100644 --- a/include/oneapi/math/rng/device/distributions.hpp +++ b/include/oneapi/math/rng/device/distributions.hpp @@ -58,14 +58,14 @@ class uniform : detail::distribution_base> { public: static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/uniform: method is incorrect"); + "oneMath: rng/uniform: method is incorrect"); static_assert(std::is_same::value || std::is_same::value || std::is_same::value || std::is_same::value || std::is_same::value || std::is_same::value, - "oneMKL: rng/uniform: type is not supported"); + "oneMath: rng/uniform: type is not supported"); using method_type = Method; using result_type = Type; @@ -136,12 +136,12 @@ class gaussian : detail::distribution_base> { || std::is_same::value #endif , - "oneMKL: rng/gaussian: method is incorrect"); + "oneMath: rng/gaussian: method is incorrect"); #if !MKL_RNG_USE_BINARY_CODE static_assert(!std::is_same::value, "icdf method not supported"); #endif static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/gaussian: type is not supported"); + "oneMath: rng/gaussian: type is not supported"); using method_type = Method; using result_type = RealType; @@ -200,10 +200,10 @@ template class lognormal : detail::distribution_base> { public: static_assert(std::is_same::value, - "oneMKL: rng/lognormal: method is incorrect"); + "oneMath: rng/lognormal: method is incorrect"); static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/lognormal: type is not supported"); + "oneMath: rng/lognormal: type is not supported"); using method_type = Method; using result_type = RealType; @@ -275,10 +275,10 @@ class beta : detail::distribution_base> { public: static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/beta: method is incorrect"); + "oneMath: rng/beta: method is incorrect"); static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/beta: type is not supported"); + "oneMath: rng/beta: type is not supported"); using method_type = Method; using result_type = RealType; @@ -352,10 +352,10 @@ class gamma : detail::distribution_base> { public: static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/gamma: method is incorrect"); + "oneMath: rng/gamma: method is incorrect"); static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/gamma: type is not supported"); + "oneMath: rng/gamma: type is not supported"); using method_type = Method; using result_type = RealType; @@ -416,7 +416,7 @@ class uniform_bits : detail::distribution_base> { public: static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/uniform_bits: type is not supported"); + "oneMath: rng/uniform_bits: type is not supported"); using result_type = UIntType; private: @@ -442,7 +442,7 @@ class bits : detail::distribution_base> { public: static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/bits: type is not supported"); + "oneMath: rng/bits: type is not supported"); using result_type = UIntType; private: @@ -476,10 +476,10 @@ class exponential : detail::distribution_base> { public: static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/exponential: method is incorrect"); + "oneMath: rng/exponential: method is incorrect"); static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/exponential: type is not supported"); + "oneMath: rng/exponential: type is not supported"); using method_type = Method; using result_type = RealType; @@ -538,11 +538,11 @@ template class poisson : detail::distribution_base> { public: static_assert(std::is_same::value, - "oneMKL: rng/poisson: method is incorrect"); + "oneMath: rng/poisson: method is incorrect"); static_assert(std::is_same::value || std::is_same::value, - "oneMKL: rng/poisson: type is not supported"); + "oneMath: rng/poisson: type is not supported"); using method_type = Method; using result_type = IntType; @@ -592,7 +592,7 @@ template class bernoulli : detail::distribution_base> { public: static_assert(std::is_same::value, - "oneMKL: rng/bernoulli: method is incorrect"); + "oneMath: rng/bernoulli: method is incorrect"); static_assert(std::is_same::value || std::is_same::value || @@ -600,7 +600,7 @@ class bernoulli : detail::distribution_base> { std::is_same::value || std::is_same::value || std::is_same::value, - "oneMKL: rng/bernoulli: type is not supported"); + "oneMath: rng/bernoulli: type is not supported"); using method_type = Method; using result_type = IntType; diff --git a/include/oneapi/math/sparse_blas/types.hpp b/include/oneapi/math/sparse_blas/types.hpp index 986d447c9..2812a0702 100644 --- a/include/oneapi/math/sparse_blas/types.hpp +++ b/include/oneapi/math/sparse_blas/types.hpp @@ -26,7 +26,7 @@ #include "detail/operation_types.hpp" /** - * @file Include and define the sparse types that are common between Intel(R) oneMKL API and oneMKL interfaces API. + * @file Include and define the sparse types that are common between Intel(R) oneMKL API and oneMath API. */ namespace oneapi { diff --git a/scripts/generate_cmake.py b/scripts/generate_cmake.py index 17bc3a279..506cf3529 100644 --- a/scripts/generate_cmake.py +++ b/scripts/generate_cmake.py @@ -118,8 +118,8 @@ def usage(err = None): list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${{LIB_OBJ}} EXPORT oneMKLTargets) -install(TARGETS ${{LIB_NAME}} EXPORT oneMKLTargets +install(TARGETS ${{LIB_OBJ}} EXPORT oneMathTargets) +install(TARGETS ${{LIB_NAME}} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index 0b59cd968..6ccb6578b 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -17,7 +17,7 @@ # SPDX-License-Identifier: Apache-2.0 #=============================================================================== -# Define common build flags for oneMKL libraries +# Define common build flags for oneMath libraries set(ONEMATH_BUILD_COPT "") if(WIN32 AND BUILD_SHARED_LIBS) list(APPEND ONEMATH_BUILD_COPT "-Donemath_EXPORTS") @@ -89,7 +89,7 @@ if(BUILD_SHARED_LIBS) set_target_properties(onemath PROPERTIES SOVERSION ${PROJECT_VERSION_MAJOR} ) - # w/a for setting oneMKL Interfaces installed headers as -I instead of -isystem for cmake >= 3.25 for workable find_package(MKL) combination + # w/a for setting oneMath installed headers as -I instead of -isystem for cmake >= 3.25 for workable find_package(MKL) combination if(CMAKE_VERSION VERSION_GREATER_EQUAL "3.25.0") set_target_properties(onemath PROPERTIES EXPORT_NO_SYSTEM true) endif() @@ -105,9 +105,9 @@ if(BUILD_SHARED_LIBS) # Add the library to install package foreach(domain_lib ${ONEMATH_LIBS}) - install(TARGETS ${domain_lib} EXPORT oneMKLTargets) + install(TARGETS ${domain_lib} EXPORT oneMathTargets) endforeach() - install(TARGETS onemath EXPORT oneMKLTargets + install(TARGETS onemath EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/blas/backends/cublas/CMakeLists.txt b/src/blas/backends/cublas/CMakeLists.txt index d9270c124..2eb20fd87 100644 --- a/src/blas/backends/cublas/CMakeLists.txt +++ b/src/blas/backends/cublas/CMakeLists.txt @@ -64,8 +64,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/blas/backends/cublas/cublas_helper.hpp b/src/blas/backends/cublas/cublas_helper.hpp index d8779d65f..a8bc388d7 100644 --- a/src/blas/backends/cublas/cublas_helper.hpp +++ b/src/blas/backends/cublas/cublas_helper.hpp @@ -43,7 +43,7 @@ namespace blas { namespace cublas { // The static assert to make sure that all index types used in -// src/oneMKL/backend/cublas/blas.hpp interface are int64_t +// src/oneMath/backend/cublas/blas.hpp interface are int64_t template struct is_int64 : std::false_type {}; @@ -75,7 +75,7 @@ struct Overflow { template void overflow_check(Index index, Next... indices) { - static_assert(is_int64::value, "oneMKL index type must be 64 bit integer."); + static_assert(is_int64::value, "oneMath index type must be 64 bit integer."); Overflow::check(index, indices...); } diff --git a/src/blas/backends/mklcpu/CMakeLists.txt b/src/blas/backends/mklcpu/CMakeLists.txt index 05e4e4a3a..3db5a2fb3 100644 --- a/src/blas/backends/mklcpu/CMakeLists.txt +++ b/src/blas/backends/mklcpu/CMakeLists.txt @@ -67,8 +67,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/blas/backends/mklgpu/CMakeLists.txt b/src/blas/backends/mklgpu/CMakeLists.txt index a5cc7abf3..6d69df3dc 100644 --- a/src/blas/backends/mklgpu/CMakeLists.txt +++ b/src/blas/backends/mklgpu/CMakeLists.txt @@ -64,8 +64,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/blas/backends/netlib/CMakeLists.txt b/src/blas/backends/netlib/CMakeLists.txt index 6ecab7591..f6dadaac1 100644 --- a/src/blas/backends/netlib/CMakeLists.txt +++ b/src/blas/backends/netlib/CMakeLists.txt @@ -68,8 +68,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/blas/backends/portblas/CMakeLists.txt b/src/blas/backends/portblas/CMakeLists.txt index 0809ceed3..1aa6c965b 100644 --- a/src/blas/backends/portblas/CMakeLists.txt +++ b/src/blas/backends/portblas/CMakeLists.txt @@ -135,7 +135,7 @@ else() endif() # If find_package doesn't work, download portBLAS from Github. This is -# intended to make OneMKL easier to use. +# intended to make oneMath easier to use. message(STATUS "Looking for portBLAS") find_package(PORTBLAS QUIET) if (NOT PORTBLAS_FOUND) @@ -214,8 +214,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/blas/backends/portblas/portblas_common.hpp b/src/blas/backends/portblas/portblas_common.hpp index 58ef28d3c..b66759d4a 100644 --- a/src/blas/backends/portblas/portblas_common.hpp +++ b/src/blas/backends/portblas/portblas_common.hpp @@ -44,10 +44,10 @@ using buffer_iterator_t = ::blas::BufferIterator; template using sycl_complex_t = sycl::ext::oneapi::experimental::complex; -/** A trait for obtaining equivalent portBLAS API types from oneMKL API +/** A trait for obtaining equivalent portBLAS API types from oneMath API * types. * - * @tparam InputT is the oneMKL type. + * @tparam InputT is the oneMath type. * portblas_type::type should be the equivalent portBLAS type. **/ template @@ -101,10 +101,10 @@ struct portblas_type> { using type = std::vector; }; -/** Convert a OneMKL argument to the type required for portBLAS. +/** Convert a oneMath argument to the type required for portBLAS. * - * @tparam InputT The OneMKL type. - * @param input The value of the oneMKL type. + * @tparam InputT The oneMath type. + * @param input The value of the oneMath type. * @return The portBLAS value with appropriate type. **/ template diff --git a/src/blas/backends/rocblas/CMakeLists.txt b/src/blas/backends/rocblas/CMakeLists.txt index 6fc87fc5d..59427458b 100644 --- a/src/blas/backends/rocblas/CMakeLists.txt +++ b/src/blas/backends/rocblas/CMakeLists.txt @@ -78,8 +78,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/blas/backends/rocblas/rocblas_helper.hpp b/src/blas/backends/rocblas/rocblas_helper.hpp index 2f61d92d5..261518951 100644 --- a/src/blas/backends/rocblas/rocblas_helper.hpp +++ b/src/blas/backends/rocblas/rocblas_helper.hpp @@ -39,7 +39,7 @@ namespace blas { namespace rocblas { // The static assert to make sure that all index types used in -// src/oneMKL/backend/rocblas/blas.hpp interface are int64_t +// src/oneMath/backend/rocblas/blas.hpp interface are int64_t template struct is_int64 : std::false_type {}; @@ -71,7 +71,7 @@ struct Overflow { template void overflow_check(Index index, Next... indices) { - static_assert(is_int64::value, "oneMKL index type must be 64 bit integer."); + static_assert(is_int64::value, "oneMath index type must be 64 bit integer."); Overflow::check(index, indices...); } diff --git a/src/dft/backends/cufft/CMakeLists.txt b/src/dft/backends/cufft/CMakeLists.txt index 2f9e908c7..a4309fd18 100644 --- a/src/dft/backends/cufft/CMakeLists.txt +++ b/src/dft/backends/cufft/CMakeLists.txt @@ -61,7 +61,7 @@ set_target_properties(${LIB_OBJ} PROPERTIES ) target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) -#Set oneMKL libraries as not transitive for dynamic +#Set oneMath libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL @@ -77,8 +77,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/dft/backends/cufft/commit.cpp b/src/dft/backends/cufft/commit.cpp index ad6f4abe4..382a97f7a 100644 --- a/src/dft/backends/cufft/commit.cpp +++ b/src/dft/backends/cufft/commit.cpp @@ -49,7 +49,7 @@ class cufft_commit final : public dft::detail::commit_impl { using scalar_type = typename dft::detail::commit_impl::scalar_type; // For real to complex transforms, the "type" arg also encodes the direction (e.g. CUFFT_R2C vs CUFFT_C2R) in the plan so we must have one for each direction. - // We also need this because oneMKL uses a directionless "FWD_DISTANCE" and "BWD_DISTANCE" while cuFFT uses a directional "idist" and "odist". + // We also need this because oneMath uses a directionless "FWD_DISTANCE" and "BWD_DISTANCE" while cuFFT uses a directional "idist" and "odist". // plans[0] is forward, plans[1] is backward std::array, 2> plans = { std::nullopt, std::nullopt }; std::int64_t offset_fwd_in, offset_fwd_out, offset_bwd_in, offset_bwd_out; diff --git a/src/dft/backends/mklcpu/CMakeLists.txt b/src/dft/backends/mklcpu/CMakeLists.txt index f1bcdd9d5..682bdb929 100644 --- a/src/dft/backends/mklcpu/CMakeLists.txt +++ b/src/dft/backends/mklcpu/CMakeLists.txt @@ -69,7 +69,7 @@ set_target_properties(${LIB_OBJ} PROPERTIES ) target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) -#Set oneMKL libraries as not transitive for dynamic +#Set oneMath libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL @@ -85,8 +85,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/dft/backends/mklgpu/CMakeLists.txt b/src/dft/backends/mklgpu/CMakeLists.txt index d0f21c8a7..2d9b3b4e7 100644 --- a/src/dft/backends/mklgpu/CMakeLists.txt +++ b/src/dft/backends/mklgpu/CMakeLists.txt @@ -71,7 +71,7 @@ set_target_properties(${LIB_OBJ} PROPERTIES ) target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) -#Set oneMKL libraries as not transitive for dynamic +#Set oneMath libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL @@ -87,8 +87,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/dft/backends/mklgpu/backward.cpp b/src/dft/backends/mklgpu/backward.cpp index 4a8bb1671..17afddeaf 100644 --- a/src/dft/backends/mklgpu/backward.cpp +++ b/src/dft/backends/mklgpu/backward.cpp @@ -61,7 +61,7 @@ inline auto compute_backward(dft::detail::descriptor &desc, ArgTs &&. throw mkl::invalid_argument("DFT", "compute_backward", "MKLGPU DFT descriptor was not successfully committed."); } - // The MKLGPU backend's iterface contains fewer function signatures than in this + // The MKLGPU backend's interface contains fewer function signatures than in this // open-source library. Consequently, it is not required to forward template arguments // to resolve to the correct function. return dft::compute_backward(*mklgpu_desc, std::forward(args)...); diff --git a/src/dft/backends/mklgpu/commit.cpp b/src/dft/backends/mklgpu/commit.cpp index 7bbdc7bec..7192b15f1 100644 --- a/src/dft/backends/mklgpu/commit.cpp +++ b/src/dft/backends/mklgpu/commit.cpp @@ -45,13 +45,13 @@ // MKL 2024.1 deprecates input/output strides. #if INTEL_MKL_VERSION < 20240001 -#error MKLGPU requires oneMKL 2024.1 or later +#error MKLGPU requires oneMath 2024.1 or later #endif /** -Note that in this file, the Intel oneMKL closed-source library's interface mirrors the interface -of this OneMKL open-source library. Consequently, the types under dft::TYPE are closed-source oneMKL types, -and types under dft::detail::TYPE are from this library. +Note that in this file, the Intel oneMKL-GPU library's interface mirrors the +interface of this oneMath library. Consequently, the types under dft::TYPE are +Intel oneMKL types, and types under dft::detail::TYPE are from this library. **/ namespace oneapi::mkl::dft::mklgpu { @@ -61,12 +61,12 @@ namespace detail { template class mklgpu_commit final : public dft::detail::commit_impl { private: - // Equivalent MKLGPU precision and domain from OneMKL's precision / domain. + // Equivalent MKLGPU precision and domain from oneMath's precision / domain. static constexpr dft::precision mklgpu_prec = to_mklgpu(prec); static constexpr dft::domain mklgpu_dom = to_mklgpu(dom); // A pair of descriptors are needed because of the [[deprecated]]IN/OUTPUT_STRIDES vs F/BWD_STRIDES API. - // Of the pair [0] is fwd DFT, [1] is backward DFT. If possible, the pointers refer to the same desciptor. + // Of the pair [0] is fwd DFT, [1] is backward DFT. If possible, the pointers refer to the same descriptor. // Both pointers must be valid. using mklgpu_descriptor_t = dft::descriptor; using descriptor_shptr_t = std::shared_ptr; diff --git a/src/dft/backends/mklgpu/forward.cpp b/src/dft/backends/mklgpu/forward.cpp index 5037241c9..d17183e92 100644 --- a/src/dft/backends/mklgpu/forward.cpp +++ b/src/dft/backends/mklgpu/forward.cpp @@ -41,9 +41,8 @@ /** Note that in this file, the Intel oneMKL-GPU library's interface mirrors the -interface of this OneMKL library. Consequently, the types under dft::TYPE are -closed-source Intel oneMKL types, and types under dft::detail::TYPE are from -this library. +interface of this oneMath library. Consequently, the types under dft::TYPE are +Intel oneMKL types, and types under dft::detail::TYPE are from this library. **/ namespace oneapi::mkl::dft::mklgpu { @@ -68,7 +67,7 @@ inline auto compute_forward(dft::detail::descriptor &desc, ArgTs &&.. throw mkl::invalid_argument("DFT", "compute_forward", "MKLGPU DFT descriptor was not successfully committed."); } - // The MKLGPU backend's iterface contains fewer function signatures than in this + // The MKLGPU backend's interface contains fewer function signatures than in this // open-source library. Consequently, it is not required to forward template arguments // to resolve to the correct function. return dft::compute_forward(*mklgpu_desc, std::forward(args)...); diff --git a/src/dft/backends/portfft/CMakeLists.txt b/src/dft/backends/portfft/CMakeLists.txt index da351deb2..fcabcc19b 100644 --- a/src/dft/backends/portfft/CMakeLists.txt +++ b/src/dft/backends/portfft/CMakeLists.txt @@ -110,7 +110,7 @@ set_target_properties(${LIB_OBJ} PROPERTIES ) target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) -#Set oneMKL libraries as not transitive for dynamic +#Set oneMath libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL @@ -126,8 +126,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/dft/backends/rocfft/CMakeLists.txt b/src/dft/backends/rocfft/CMakeLists.txt index 0e2548d95..a02877398 100644 --- a/src/dft/backends/rocfft/CMakeLists.txt +++ b/src/dft/backends/rocfft/CMakeLists.txt @@ -71,7 +71,7 @@ set_target_properties(${LIB_OBJ} PROPERTIES ) target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) -# Set oneMKL libraries as not transitive for dynamic +# Set oneMath libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL @@ -87,8 +87,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/dft/backends/rocfft/commit.cpp b/src/dft/backends/rocfft/commit.cpp index 2ce871447..e87e9d714 100644 --- a/src/dft/backends/rocfft/commit.cpp +++ b/src/dft/backends/rocfft/commit.cpp @@ -85,7 +85,7 @@ class rocfft_commit final : public dft::detail::commit_impl { using scalar_type = typename dft::detail::commit_impl::scalar_type; // For real to complex transforms, the "transform_type" arg also encodes the direction (e.g. rocfft_transform_type_*_forward vs rocfft_transform_type_*_backward) // in the plan so we must have one for each direction. - // We also need this because oneMKL uses a directionless "FWD_DISTANCE" and "BWD_DISTANCE" while rocFFT uses a directional "in_distance" and "out_distance". + // We also need this because oneMath uses a directionless "FWD_DISTANCE" and "BWD_DISTANCE" while rocFFT uses a directional "in_distance" and "out_distance". // The same is also true for "FORWARD_SCALE" and "BACKWARD_SCALE". // handles[0] is forward, handles[1] is backward std::array handles{}; @@ -180,7 +180,7 @@ class rocfft_commit final : public dft::detail::commit_impl { constexpr std::size_t max_supported_dims = 3; std::array lengths; - // rocfft does dimensions in the reverse order to oneMKL + // rocfft does dimensions in the reverse order to oneMath std::copy(config_values.dimensions.crbegin(), config_values.dimensions.crend(), lengths.data()); diff --git a/src/lapack/backends/cusolver/CMakeLists.txt b/src/lapack/backends/cusolver/CMakeLists.txt index 1828c64df..7340e6f7e 100644 --- a/src/lapack/backends/cusolver/CMakeLists.txt +++ b/src/lapack/backends/cusolver/CMakeLists.txt @@ -59,8 +59,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/lapack/backends/cusolver/cusolver_helper.hpp b/src/lapack/backends/cusolver/cusolver_helper.hpp index 861e45ef8..4163fe893 100644 --- a/src/lapack/backends/cusolver/cusolver_helper.hpp +++ b/src/lapack/backends/cusolver/cusolver_helper.hpp @@ -44,7 +44,7 @@ namespace lapack { namespace cusolver { // The static assert to make sure that all index types used in -// oneMKL/include/oneapi/math/lapack.hpp interface are int64_t +// oneMath/include/oneapi/math/lapack.hpp interface are int64_t template struct is_int64 : std::false_type {}; @@ -76,7 +76,7 @@ struct Overflow { template void overflow_check(Index index, Next... indices) { - static_assert(is_int64::value, "oneMKL index type must be 64 bit integer."); + static_assert(is_int64::value, "oneMath index type must be 64 bit integer."); Overflow::check(index, indices...); } diff --git a/src/lapack/backends/cusolver/cusolver_scope_handle.hpp b/src/lapack/backends/cusolver/cusolver_scope_handle.hpp index c25cc2c35..9ee4afa40 100644 --- a/src/lapack/backends/cusolver/cusolver_scope_handle.hpp +++ b/src/lapack/backends/cusolver/cusolver_scope_handle.hpp @@ -71,7 +71,7 @@ According to NVIDIA: using the same host thread. The advice above is for using cublas with the cuda runtime API. Given that cusolver is based on cublas the advice is -transferable. The cusolver_scope_handle is based on the oneMKL cublas_scope_handle. The NVIDIA runtime API creates a +transferable. The cusolver_scope_handle is based on the oneMath cublas_scope_handle. The NVIDIA runtime API creates a default context for users. The cusolverDnCreate function in uses the context located on top of the stack for each thread. Then, the cuSolver routine uses this context for resource allocation/access. Calling a cuSolver function with a handle created for context A and memories/queue created for context B results in a segmentation fault. Thus we need to create diff --git a/src/lapack/backends/mklcpu/CMakeLists.txt b/src/lapack/backends/mklcpu/CMakeLists.txt index 688dba71d..0821098cc 100644 --- a/src/lapack/backends/mklcpu/CMakeLists.txt +++ b/src/lapack/backends/mklcpu/CMakeLists.txt @@ -47,7 +47,7 @@ set_target_properties(${LIB_OBJ} PROPERTIES ) target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) -#Set oneMKL libraries as not transitive for dynamic +#Set oneMath libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL @@ -63,8 +63,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/lapack/backends/mklgpu/CMakeLists.txt b/src/lapack/backends/mklgpu/CMakeLists.txt index 2be7539e9..24801ebc8 100644 --- a/src/lapack/backends/mklgpu/CMakeLists.txt +++ b/src/lapack/backends/mklgpu/CMakeLists.txt @@ -47,7 +47,7 @@ set_target_properties(${LIB_OBJ} PROPERTIES ) target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) -#Set oneMKL libraries as not transitive for dynamic +#Set oneMath libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL @@ -63,8 +63,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/lapack/backends/rocsolver/CMakeLists.txt b/src/lapack/backends/rocsolver/CMakeLists.txt index a7dde150c..8488cc6ad 100644 --- a/src/lapack/backends/rocsolver/CMakeLists.txt +++ b/src/lapack/backends/rocsolver/CMakeLists.txt @@ -60,8 +60,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/lapack/backends/rocsolver/rocsolver_helper.hpp b/src/lapack/backends/rocsolver/rocsolver_helper.hpp index d51c240a9..36f78d00d 100644 --- a/src/lapack/backends/rocsolver/rocsolver_helper.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_helper.hpp @@ -43,7 +43,7 @@ namespace lapack { namespace rocsolver { // The static assert to make sure that all index types used in -// oneMKL/include/oneapi/math/lapack.hpp interface are int64_t +// oneMath/include/oneapi/math/lapack.hpp interface are int64_t template struct is_int64 : std::false_type {}; @@ -75,7 +75,7 @@ struct Overflow { template void overflow_check(Index index, Next... indices) { - static_assert(is_int64::value, "oneMKL index type must be 64 bit integer."); + static_assert(is_int64::value, "oneMath index type must be 64 bit integer."); Overflow::check(index, indices...); } diff --git a/src/rng/backends/curand/CMakeLists.txt b/src/rng/backends/curand/CMakeLists.txt index 1eee395e7..f8415c4a2 100644 --- a/src/rng/backends/curand/CMakeLists.txt +++ b/src/rng/backends/curand/CMakeLists.txt @@ -95,8 +95,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/rng/backends/curand/philox4x32x10.cpp b/src/rng/backends/curand/philox4x32x10.cpp index 53fef8f60..89ca45853 100644 --- a/src/rng/backends/curand/philox4x32x10.cpp +++ b/src/rng/backends/curand/philox4x32x10.cpp @@ -103,7 +103,7 @@ namespace curand { * consumed in other kernels without requiring the random numbers to be written * to, and read from, global memory. * - * Here we utilize the host API since this is most aligned with how oneMKL + * Here we utilize the host API since this is most aligned with how oneMath * generates random numbers. * */ diff --git a/src/rng/backends/mklcpu/CMakeLists.txt b/src/rng/backends/mklcpu/CMakeLists.txt index b96b8ae27..49aff884b 100644 --- a/src/rng/backends/mklcpu/CMakeLists.txt +++ b/src/rng/backends/mklcpu/CMakeLists.txt @@ -47,7 +47,7 @@ set_target_properties(${LIB_OBJ} PROPERTIES ) target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) -# Set oneMKL libraries as not transitive for dynamic +# Set oneMath libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL @@ -63,8 +63,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/rng/backends/mklgpu/CMakeLists.txt b/src/rng/backends/mklgpu/CMakeLists.txt index 891aff8be..92572d91b 100644 --- a/src/rng/backends/mklgpu/CMakeLists.txt +++ b/src/rng/backends/mklgpu/CMakeLists.txt @@ -49,7 +49,7 @@ set_target_properties(${LIB_OBJ} PROPERTIES ) target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) -# Set oneMKL libraries as not transitive for dynamic +# Set oneMath libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL @@ -65,8 +65,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/rng/backends/rocrand/CMakeLists.txt b/src/rng/backends/rocrand/CMakeLists.txt index 50211fe02..49de4054b 100644 --- a/src/rng/backends/rocrand/CMakeLists.txt +++ b/src/rng/backends/rocrand/CMakeLists.txt @@ -85,10 +85,10 @@ set_target_properties(${LIB_NAME} PROPERTIES SOVERSION ${PROJECT_VERSION_MAJOR}) list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) install( TARGETS ${LIB_NAME} - EXPORT oneMKLTargets + EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib) diff --git a/src/rng/backends/rocrand/philox4x32x10.cpp b/src/rng/backends/rocrand/philox4x32x10.cpp index 3bacaa46d..e3656a43d 100644 --- a/src/rng/backends/rocrand/philox4x32x10.cpp +++ b/src/rng/backends/rocrand/philox4x32x10.cpp @@ -103,7 +103,7 @@ namespace rocrand { * consumed in other kernels without requiring the random numbers to be written * to, and read from, global memory. * - * Here we utilize the host API since this is most aligned with how oneMKL + * Here we utilize the host API since this is most aligned with how oneMath * generates random numbers. * */ diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.cxx b/src/sparse_blas/backends/mkl_common/mkl_handles.cxx index 3ae84ca64..85e973fe6 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.cxx @@ -378,7 +378,7 @@ bool set_matrix_property(sycl::queue & /*queue*/, oneapi::mkl::sparse::matrix_ha // Store the matrix property internally for better error checking internal_smhandle->set_matrix_property(property); // Set the matrix property on the backend handle - // Backend and oneMKL interface types for the property don't match + // Backend and oneMath types for the property don't match switch (property) { case oneapi::mkl::sparse::matrix_property::symmetric: oneapi::mkl::sparse::set_matrix_property(internal_smhandle->backend_handle, diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp index 117faa630..e459c9b8c 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp @@ -30,7 +30,7 @@ namespace oneapi::mkl::sparse { // Complete the definition of incomplete types dense_vector_handle and -// dense_matrix_handle as they don't exist in oneMKL backends yet. +// dense_matrix_handle as they don't exist in oneMath backends yet. struct dense_vector_handle : public detail::generic_dense_vector_handle { template @@ -70,7 +70,7 @@ namespace oneapi::mkl::sparse::detail { */ using sparse_matrix_handle = detail::generic_sparse_handle; -/// Cast to oneMKL's interface handle type +/// Cast to oneMath's internal handle type inline auto get_internal_handle(matrix_handle_t handle) { return reinterpret_cast(handle); } diff --git a/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx b/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx index dad611252..dce464228 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx @@ -103,7 +103,7 @@ void spmm_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::sparse::dense_matrix_handle_t C_handle, oneapi::mkl::sparse::spmm_alg /*alg*/, oneapi::mkl::sparse::spmm_descr_t spmm_descr, std::size_t &temp_buffer_size) { - // TODO: Add support for external workspace once the close-source oneMKL backend supports it. + // TODO: Add support for external workspace once the close-source oneMath backend supports it. bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); check_valid_spmm(__func__, opA, A_view, A_handle, B_handle, C_handle, is_alpha_host_accessible, @@ -153,7 +153,7 @@ void spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl:: return; } internal_A_handle->can_be_reset = false; - // TODO: Add support for spmm_optimize once the close-source oneMKL backend supports it. + // TODO: Add support for spmm_optimize once the close-source oneMath backend supports it. } sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, @@ -175,7 +175,7 @@ sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, return detail::collapse_dependencies(queue, dependencies); } internal_A_handle->can_be_reset = false; - // TODO: Add support for spmm_optimize once the close-source oneMKL backend supports it. + // TODO: Add support for spmm_optimize once the close-source oneMath backend supports it. return detail::collapse_dependencies(queue, dependencies); } diff --git a/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx b/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx index d2332286b..a959f1f8c 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx @@ -92,7 +92,7 @@ void spmv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void oneapi::mkl::sparse::dense_vector_handle_t y_handle, oneapi::mkl::sparse::spmv_alg /*alg*/, oneapi::mkl::sparse::spmv_descr_t spmv_descr, std::size_t &temp_buffer_size) { - // TODO: Add support for external workspace once the close-source oneMKL backend supports it. + // TODO: Add support for external workspace once the close-source oneMath backend supports it. bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); check_valid_spmv(__func__, opA, A_view, A_handle, x_handle, y_handle, is_alpha_host_accessible, diff --git a/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx b/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx index 7ef5b3c39..6a7f50eb1 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx @@ -91,7 +91,7 @@ void spsv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void oneapi::mkl::sparse::dense_vector_handle_t y_handle, oneapi::mkl::sparse::spsv_alg alg, oneapi::mkl::sparse::spsv_descr_t spsv_descr, std::size_t &temp_buffer_size) { - // TODO: Add support for external workspace once the close-source oneMKL backend supports it. + // TODO: Add support for external workspace once the close-source oneMath backend supports it. bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); check_valid_spsv(__func__, opA, A_view, A_handle, x_handle, y_handle, is_alpha_host_accessible, alg); diff --git a/src/sparse_blas/backends/mklcpu/CMakeLists.txt b/src/sparse_blas/backends/mklcpu/CMakeLists.txt index de1cbf6a8..111d3f7ca 100644 --- a/src/sparse_blas/backends/mklcpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklcpu/CMakeLists.txt @@ -58,7 +58,7 @@ set_target_properties(${LIB_OBJ} PROPERTIES ) target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) -#Set oneMKL libraries as not transitive for dynamic +#Set oneMath libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL @@ -74,8 +74,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib diff --git a/src/sparse_blas/backends/mklgpu/CMakeLists.txt b/src/sparse_blas/backends/mklgpu/CMakeLists.txt index 5d86d4f65..4f4a9f82d 100644 --- a/src/sparse_blas/backends/mklgpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklgpu/CMakeLists.txt @@ -58,7 +58,7 @@ set_target_properties(${LIB_OBJ} PROPERTIES ) target_link_libraries(${LIB_NAME} PUBLIC ${LIB_OBJ}) -#Set oneMKL libraries as not transitive for dynamic +#Set oneMath libraries as not transitive for dynamic if(BUILD_SHARED_LIBS) set_target_properties(${LIB_NAME} PROPERTIES INTERFACE_LINK_LIBRARIES ONEMATH::SYCL::SYCL @@ -74,8 +74,8 @@ set_target_properties(${LIB_NAME} PROPERTIES list(APPEND CMAKE_BUILD_RPATH $) # Add the library to install package -install(TARGETS ${LIB_OBJ} EXPORT oneMKLTargets) -install(TARGETS ${LIB_NAME} EXPORT oneMKLTargets +install(TARGETS ${LIB_OBJ} EXPORT oneMathTargets) +install(TARGETS ${LIB_NAME} EXPORT oneMathTargets RUNTIME DESTINATION bin ARCHIVE DESTINATION lib LIBRARY DESTINATION lib From eda55ffa594a5446cddaf3e27bda9102866e894d Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 8 Oct 2024 12:59:50 +0200 Subject: [PATCH 07/41] Reword introduction --- docs/introduction.rst | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/docs/introduction.rst b/docs/introduction.rst index b055f8519..693e85f61 100644 --- a/docs/introduction.rst +++ b/docs/introduction.rst @@ -3,7 +3,7 @@ Introduction ============ -oneMath is an open-source implementation of oneMath Data Parallel C++ -(DPC++) interfaces according to the `oneMath specification `_ -that can work with multiple devices (backends) using device-specific -libraries underneath. +oneMath is an open-source implementation of the `oneMath specification +`_ +that can work with multiple devices using multiple libraries (backends) +underneath. From 44dc387543e7230dc3d4a7366ddde1b2ba79eaf9 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 8 Oct 2024 17:10:52 +0200 Subject: [PATCH 08/41] Clarify comments referring to MKL --- src/dft/backends/mklgpu/commit.cpp | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/src/dft/backends/mklgpu/commit.cpp b/src/dft/backends/mklgpu/commit.cpp index 7192b15f1..87efa22a2 100644 --- a/src/dft/backends/mklgpu/commit.cpp +++ b/src/dft/backends/mklgpu/commit.cpp @@ -43,7 +43,8 @@ #include #endif -// MKL 2024.1 deprecates input/output strides. +// Intel oneMKL 2024.1 deprecates input/output strides. +#include "mkl_version.h" #if INTEL_MKL_VERSION < 20240001 #error MKLGPU requires oneMath 2024.1 or later #endif From 760d3976fa0df0a99a2bcdf2674e1d0a3d016ca6 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 8 Oct 2024 17:17:15 +0200 Subject: [PATCH 09/41] Rename project CMake namespace MKL:: to ONEMATH:: --- CMakeLists.txt | 2 +- docs/using_onemath_with_cmake.rst | 12 ++++++------ 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/CMakeLists.txt b/CMakeLists.txt index 54090c6b0..cd5ac0a2c 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -343,7 +343,7 @@ configure_file("${PROJECT_SOURCE_DIR}/cmake/oneMathConfig.cmake" set(config_package_location "lib/cmake/${PROJECT_NAME}") install(EXPORT oneMathTargets FILE oneMathTargets.cmake - NAMESPACE MKL:: + NAMESPACE ONEMATH:: DESTINATION ${config_package_location} ) install( diff --git a/docs/using_onemath_with_cmake.rst b/docs/using_onemath_with_cmake.rst index acc7a0a67..9ada0d937 100644 --- a/docs/using_onemath_with_cmake.rst +++ b/docs/using_onemath_with_cmake.rst @@ -31,12 +31,12 @@ For example: .. code-block:: cmake find_package(oneMath REQUIRED) - target_link_libraries(myTarget PRIVATE MKL::onemath) + target_link_libraries(myTarget PRIVATE ONEMATH::onemath) Different targets can be used depending on the requirements of oneMath. -To link against the entire library, the ``MKL::onemath`` target should be used. -For specific domains, ``MKL::onemath_`` should be used. -And for specific backends, ``MKL::onemath__`` should be used. +To link against the entire library, the ``ONEMATH::onemath`` target should be used. +For specific domains, ``ONEMATH::onemath_`` should be used. +And for specific backends, ``ONEMATH::onemath__`` should be used. When using a binary, it may be useful to know the backends that were enabled during the build. To check for the existence of backends, CMake's ``if(TARGET @@ -44,8 +44,8 @@ during the build. To check for the existence of backends, CMake's ``if(TARGET .. code-block:: cmake - if(TARGET MKL::onemath_dft_cufft) - target_link_libraries(myTarget PRIVATE MKL::onemath_dft_cufft) + if(TARGET ONEMATH::onemath_dft_cufft) + target_link_libraries(myTarget PRIVATE ONEMATH::onemath_dft_cufft) else() message(FATAL_ERROR "oneMath was not built with CuFFT backend") endif() From d03aea4367aef1cb57e5391ae07cd90019b889e8 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 8 Oct 2024 18:31:07 +0200 Subject: [PATCH 10/41] Rename occurrences in third-party-programs/THIRD-PARTY-PROGRAMS --- third-party-programs/THIRD-PARTY-PROGRAMS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/third-party-programs/THIRD-PARTY-PROGRAMS b/third-party-programs/THIRD-PARTY-PROGRAMS index fd462fa83..1de26e969 100644 --- a/third-party-programs/THIRD-PARTY-PROGRAMS +++ b/third-party-programs/THIRD-PARTY-PROGRAMS @@ -1,4 +1,4 @@ -Intel® oneAPI Math Kernel Library (oneMKL) interfaces +oneMath This file contains the list of third party software (“third party programs”) contained in the Intel software and their required notices and/or license terms. From ac41b314c430f3f3616c86deb4d64d012a9aa14b Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 8 Oct 2024 18:40:34 +0200 Subject: [PATCH 11/41] oneapi-src -> uxlfoundation --- .github/ISSUE_TEMPLATE/RFC.md | 2 +- .github/pull_request_template.md | 2 +- CONTRIBUTING.md | 22 +++++++++---------- MAINTAINERS.md | 14 ++++++------ README.md | 22 +++++++++---------- SECURITY.md | 6 ++--- docs/_templates/layout.html | 2 +- .../building_the_project_with_adaptivecpp.rst | 4 ++-- docs/building_the_project_with_dpcpp.rst | 8 +++---- docs/conf.py.in | 2 +- docs/using_onemath_with_cmake.rst | 2 +- .../math/dft/detail/descriptor_impl.hpp | 2 +- tests/README.md | 2 +- 13 files changed, 45 insertions(+), 45 deletions(-) diff --git a/.github/ISSUE_TEMPLATE/RFC.md b/.github/ISSUE_TEMPLATE/RFC.md index 31946cd3d..ad7a100a2 100755 --- a/.github/ISSUE_TEMPLATE/RFC.md +++ b/.github/ISSUE_TEMPLATE/RFC.md @@ -9,7 +9,7 @@ assignees: '' # Summary Include a short summary of the request. Sections below provide guidance on what factors are considered important. Describe how new interface will meet -[library functionality guidelines](https://github.com/oneapi-src/oneMath/blob/master/CONTRIBUTING.md#library-functionality-guidelines). +[library functionality guidelines](https://github.com/uxlfoundation/oneMath/blob/master/CONTRIBUTING.md#library-functionality-guidelines). # Problem statement Describe the problem you are trying to solve with reasonable level of details. diff --git a/.github/pull_request_template.md b/.github/pull_request_template.md index 976e9b23c..70c7a35e2 100755 --- a/.github/pull_request_template.md +++ b/.github/pull_request_template.md @@ -2,7 +2,7 @@ Please include a summary of the change. Please also include relevant motivation and context. See -[contribution guidelines](https://github.com/oneapi-src/oneMath/blob/master/CONTRIBUTING.md) +[contribution guidelines](https://github.com/uxlfoundation/oneMath/blob/master/CONTRIBUTING.md) for more details. If the change fixes an issue not documented in the project's Github issue tracker, please document all steps necessary to reproduce it. diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index f29d70bd6..342bbca5d 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -37,7 +37,7 @@ For the new API to become a part of the open source project, it should be accept ### Request for Comments Process For changes impacting the public API or any significant changes in the library, such as adding new backend or changes to the architecture, -please follow the [RFC process](https://github.com/oneapi-src/oneMKL/tree/rfcs). +please follow the [RFC process](https://github.com/uxlfoundation/oneMKL/tree/rfcs). Please also provide the following details as part of the RFC: @@ -51,7 +51,7 @@ Please also provide the following details as part of the RFC: ## Bug Reporting -If you find a bug or problem, please open a request under [Issues](https://github.com/oneapi-src/oneMKL/issues). +If you find a bug or problem, please open a request under [Issues](https://github.com/uxlfoundation/oneMKL/issues). ## Security Issues @@ -64,14 +64,14 @@ For GitHub questions, issues, RFCs, or PRs you can contact maintainers via one o | GitHub team name | Description | :-----------|:------------| -| @oneapi-src/onemkl-maintain | All oneMKL Interfaces maintainers | -| @oneapi-src/onemkl-arch-write | oneMKL Interfaces Architecture maintainers | -| @oneapi-src/onemkl-blas-write | oneMKL Interfaces BLAS maintainers | -| @oneapi-src/onemkl-dft-write | oneMKL Interfaces DFT maintainers | -| @oneapi-src/onemkl-lapack-write | oneMKL Interfaces LAPACK maintainers | -| @oneapi-src/onemkl-rng-write | oneMKL Interfaces RNG maintainers | -| @oneapi-src/onemkl-sparse-write | oneMKL Interfaces Sparse Algebra maintainers | -| @oneapi-src/onemkl-vm-write | oneMKL Interfaces Vector Math maintainers | +| @uxlfoundation/onemath-maintain | All oneMKL Interfaces maintainers | +| @uxlfoundation/onemath-arch-write | oneMKL Interfaces Architecture maintainers | +| @uxlfoundation/onemath-blas-write | oneMKL Interfaces BLAS maintainers | +| @uxlfoundation/onemath-dft-write | oneMKL Interfaces DFT maintainers | +| @uxlfoundation/onemath-lapack-write | oneMKL Interfaces LAPACK maintainers | +| @uxlfoundation/onemath-rng-write | oneMKL Interfaces RNG maintainers | +| @uxlfoundation/onemath-sparse-write | oneMKL Interfaces Sparse Algebra maintainers | +| @uxlfoundation/onemath-vm-write | oneMKL Interfaces Vector Math maintainers | Please read [MAINTAINERS page](MAINTAINERS.md) for more information about maintainer roles, responsibilities, and how to become one of them. @@ -342,6 +342,6 @@ for (int i = 0; i < loop_size; i++) ...; ## Unit Tests -oneMKL uses GoogleTest for functional testing. For more information about how to build and run Unit Tests please see [Building and Running Tests](https://oneapi-src.github.io/oneMKL/building_and_running_tests.html). +oneMKL uses GoogleTest for functional testing. For more information about how to build and run Unit Tests please see [Building and Running Tests](https://uxlfoundation.github.io/oneMKL/building_and_running_tests.html). Be sure to extend the existing tests when fixing an issue, adding a new interface or new implementation under existing interfaces. diff --git a/MAINTAINERS.md b/MAINTAINERS.md index 9594af0ea..1889851cd 100644 --- a/MAINTAINERS.md +++ b/MAINTAINERS.md @@ -72,12 +72,12 @@ The process of becoming a Domain maintainer is: | GitHub team name | Domain maintainers | :-----------|:------------| -| @oneapi-src/onemath-blas-write | oneMath BLAS maintainers | -| @oneapi-src/onemath-dft-write | oneMath DFT maintainers | -| @oneapi-src/onemath-lapack-write | oneMath LAPACK maintainers | -| @oneapi-src/onemath-rng-write | oneMath RNG maintainers | -| @oneapi-src/onemath-sparse-write | oneMath Sparse Algebra maintainers | -| @oneapi-src/onemath-vm-write | oneMath Vector Math maintainers | +| @uxlfoundation/onemath-blas-write | oneMath BLAS maintainers | +| @uxlfoundation/onemath-dft-write | oneMath DFT maintainers | +| @uxlfoundation/onemath-lapack-write | oneMath LAPACK maintainers | +| @uxlfoundation/onemath-rng-write | oneMath RNG maintainers | +| @uxlfoundation/onemath-sparse-write | oneMath Sparse Algebra maintainers | +| @uxlfoundation/onemath-vm-write | oneMath Vector Math maintainers | ## Architecture Maintainer Architecture maintainers are the most established contributors who are responsible for the @@ -108,6 +108,6 @@ Privileges: Process of becoming a maintainer: 1. A Contributor or Domain maintainer requests to join oneMath Architecture maintainers GitHub team -(@oneapi-src/onemath-arch-write). +(@uxlfoundation/onemath-arch-write). 2. At least one of Architecture maintainers approves the request. diff --git a/README.md b/README.md index d8f228b96..f79c40913 100644 --- a/README.md +++ b/README.md @@ -141,7 +141,7 @@ There are two oneMKL selector layer implementations: $> clang++ -fsycl app.o –L$ONEMKL/lib –lonemath_blas_mklcpu –lonemath_blas_cublas ``` -*Refer to [Selecting a Compiler](https://oneapi-src.github.io/oneMKL/selecting_a_compiler.html) for the choice between `icpx/icx` and `clang++` compilers.* +*Refer to [Selecting a Compiler](https://uxlfoundation.github.io/oneMKL/selecting_a_compiler.html) for the choice between `icpx/icx` and `clang++` compilers.* #### Device API @@ -534,16 +534,16 @@ Product | Supported Version | License --- ## Documentation -- [Contents](https://oneapi-src.github.io/oneMKL/) -- [About](https://oneapi-src.github.io/oneMKL/introduction.html) +- [Contents](https://uxlfoundation.github.io/oneMKL/) +- [About](https://uxlfoundation.github.io/oneMKL/introduction.html) - Get Started - - [Selecting a Compiler](https://oneapi-src.github.io/oneMKL/selecting_a_compiler.html) - - [Building the Project with DPC++](https://oneapi-src.github.io/oneMKL/building_the_project_with_dpcpp.html) - - [Building the Project with AdaptiveCpp](https://oneapi-src.github.io/oneMKL/building_the_project_with_adaptivecpp.html) + - [Selecting a Compiler](https://uxlfoundation.github.io/oneMKL/selecting_a_compiler.html) + - [Building the Project with DPC++](https://uxlfoundation.github.io/oneMKL/building_the_project_with_dpcpp.html) + - [Building the Project with AdaptiveCpp](https://uxlfoundation.github.io/oneMKL/building_the_project_with_adaptivecpp.html) - Developer Reference - - [oneMKL Defined Datatypes](https://oneapi-src.github.io/oneMKL/onemkl-datatypes.html) - - [Dense Linear Algebra](https://oneapi-src.github.io/oneMKL/domains/dense_linear_algebra.html) -- [Integrating a Third-Party Library](https://oneapi-src.github.io/oneMKL/create_new_backend.html) + - [oneMKL Defined Datatypes](https://uxlfoundation.github.io/oneMKL/onemkl-datatypes.html) + - [Dense Linear Algebra](https://uxlfoundation.github.io/oneMKL/domains/dense_linear_algebra.html) +- [Integrating a Third-Party Library](https://uxlfoundation.github.io/oneMKL/create_new_backend.html) --- @@ -573,13 +573,13 @@ Distributed under the Apache license 2.0. See [LICENSE](LICENSE) for more inform **Q: What is the difference between the following oneMKL items?** - The [oneAPI Specification for oneMKL](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemkl/source/) - - The [oneAPI Math Kernel Library (oneMKL) Interfaces](https://github.com/oneapi-src/oneMKL) Project + - The [oneAPI Math Kernel Library (oneMKL) Interfaces](https://github.com/uxlfoundation/oneMKL) Project - The [Intel(R) oneAPI Math Kernel Library (oneMKL)](https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onemkl.html) Product **A:** - The [oneAPI Specification for oneMKL](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemkl/source/) defines the DPC++ interfaces for performance math library functions. The oneMKL specification can evolve faster and more frequently than implementations of the specification. -- The [oneAPI Math Kernel Library (oneMKL) Interfaces](https://github.com/oneapi-src/oneMKL) Project is an open source implementation of the specification. The project goal is to demonstrate how the DPC++ interfaces documented in the oneMKL specification can be implemented for any math library and work for any target hardware. While the implementation provided here may not yet be the full implementation of the specification, the goal is to build it out over time. We encourage the community to contribute to this project and help to extend support to multiple hardware targets and other math libraries. +- The [oneAPI Math Kernel Library (oneMKL) Interfaces](https://github.com/uxlfoundation/oneMKL) Project is an open source implementation of the specification. The project goal is to demonstrate how the DPC++ interfaces documented in the oneMKL specification can be implemented for any math library and work for any target hardware. While the implementation provided here may not yet be the full implementation of the specification, the goal is to build it out over time. We encourage the community to contribute to this project and help to extend support to multiple hardware targets and other math libraries. - The [Intel(R) oneAPI Math Kernel Library (oneMKL)](https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onemkl.html) product is the Intel product implementation of the specification (with DPC++ interfaces) as well as similar functionality with C and Fortran interfaces, and is provided as part of Intel® oneAPI Base Toolkit. It is highly optimized for Intel CPU and Intel GPU hardware. diff --git a/SECURITY.md b/SECURITY.md index acaa1152e..7c8af680f 100644 --- a/SECURITY.md +++ b/SECURITY.md @@ -60,6 +60,6 @@ If you have any suggestions on how this Policy could be improved, please submit an issue or a pull request to this repository. Please **do not** report potential vulnerabilities or security flaws via a pull request. -[1]: https://github.com/oneapi-src/oneMath/releases/latest -[2]: https://github.com/oneapi-src/oneMath/security/advisories/new -[3]: https://github.com/oneapi-src/oneMath/security/advisories \ No newline at end of file +[1]: https://github.com/uxlfoundation/oneMath/releases/latest +[2]: https://github.com/uxlfoundation/oneMath/security/advisories/new +[3]: https://github.com/uxlfoundation/oneMath/security/advisories \ No newline at end of file diff --git a/docs/_templates/layout.html b/docs/_templates/layout.html index cae88b2cc..d6145c75e 100644 --- a/docs/_templates/layout.html +++ b/docs/_templates/layout.html @@ -5,7 +5,7 @@ var wapLocalCode = 'us-en'; // Dynamically set per localized site; see mapping table for values var wapSection = "oneapi-mkl"; // WAP team will give you a unique section for your site // Load TMS - if (document.location.href.includes("oneapi-src.github.io")) { + if (document.location.href.includes("uxlfoundation.github.io")) { (function () { var url = 'https://www.intel.com/content/dam/www/global/wap/tms-loader.js'; // WAP file URL var po = document.createElement('script'); po.type = 'text/javascript'; po.async = true; po.src = url; diff --git a/docs/building_the_project_with_adaptivecpp.rst b/docs/building_the_project_with_adaptivecpp.rst index 073652d01..b907285a9 100644 --- a/docs/building_the_project_with_adaptivecpp.rst +++ b/docs/building_the_project_with_adaptivecpp.rst @@ -20,7 +20,7 @@ Environment Setup #. Download and install the `required dependencies - `_ + `_ manually. Build Commands @@ -50,7 +50,7 @@ each desired backend. By default, the ``MKLGPU`` and ``MKLCPU`` backends are enabled, but ``MKLGPU`` must be disabled with AdaptiveCpp. The supported backends for the compilers are given in the table at `oneMath supported configurations table -`_, +`_, and the CMake option names are given in the table below. Some backends may require additional parameters to be set. See the relevant section below for additional guidance. The target architectures must be specified with diff --git a/docs/building_the_project_with_dpcpp.rst b/docs/building_the_project_with_dpcpp.rst index 6076117f7..8eb473596 100644 --- a/docs/building_the_project_with_dpcpp.rst +++ b/docs/building_the_project_with_dpcpp.rst @@ -23,7 +23,7 @@ Environment Setup #. Build and install all `required dependencies - `_. + `_. .. _build_introduction_with_dpcpp: @@ -63,7 +63,7 @@ are enabled. Multiple backends for multiple device vendors can be enabled at once (albeit with limitations when using portBLAS and portFFT). The supported backends for the compilers are given in the table at `oneMKL supported configurations table -`_, +`_, and the CMake option names are given in the table below. Some backends may require additional parameters to be set. See the relevant section below for additional guidance. @@ -348,7 +348,7 @@ The following table provides details of CMake options and their default values: .. note:: When building with ``BUILD_FUNCTIONAL_TESTS=True`` (default option) only single CUDA backend can be built - (`#270 `_). + (`#270 `_). .. _build_invocation_examples_dpcpp: @@ -479,7 +479,7 @@ Building for Windows #################### The Windows build is similar to the Linux build, albeit that `fewer backends are -supported `_. +supported `_. Additionally, the Ninja build system must be used. For example: .. code-block:: bash diff --git a/docs/conf.py.in b/docs/conf.py.in index 2038b3f4f..6167f8f7f 100644 --- a/docs/conf.py.in +++ b/docs/conf.py.in @@ -90,7 +90,7 @@ html_favicon = f'{static_dir}/favicons.png' # Theme options html_theme_options = { -'repository_url': 'https://github.com/oneapi-src/oneMath', +'repository_url': 'https://github.com/uxlfoundation/oneMath', 'path_to_docs': 'docs', 'use_issues_button': True, 'use_edit_page_button': True, diff --git a/docs/using_onemath_with_cmake.rst b/docs/using_onemath_with_cmake.rst index 9ada0d937..8dac6e092 100644 --- a/docs/using_onemath_with_cmake.rst +++ b/docs/using_onemath_with_cmake.rst @@ -69,7 +69,7 @@ For example: set(ENABLE__BACKEND True) FetchContent_Declare( onemath_library - GIT_REPOSITORY https://github.com/oneapi-src/oneMath.git + GIT_REPOSITORY https://github.com/uxlfoundation/oneMath.git GIT_TAG develop ) FetchContent_MakeAvailable(onemath_library) diff --git a/include/oneapi/math/dft/detail/descriptor_impl.hpp b/include/oneapi/math/dft/detail/descriptor_impl.hpp index 4b1841989..e5d7e3087 100644 --- a/include/oneapi/math/dft/detail/descriptor_impl.hpp +++ b/include/oneapi/math/dft/detail/descriptor_impl.hpp @@ -57,7 +57,7 @@ class descriptor { // Copy operations are included in the oneAPI oneMath specification, but not yet // implemented here. If you need copies, please open an issue at - // https://github.com/oneapi-src/oneMath/issues + // https://github.com/uxlfoundation/oneMath/issues descriptor(descriptor&&); diff --git a/tests/README.md b/tests/README.md index 71511711a..cbb87e8ff 100644 --- a/tests/README.md +++ b/tests/README.md @@ -3,7 +3,7 @@ ## Overview Inside the `unit_tests` directory, there are domain-level directories which contain domain-specific tests, usually per function or per configuration. -See [Building and Running Tests](https://oneapi-src.github.io/oneMKL/building_and_running_tests.html) documentation for more information about how to build and run the tests. +See [Building and Running Tests](https://uxlfoundation.github.io/oneMKL/building_and_running_tests.html) documentation for more information about how to build and run the tests. [GoogleTest](https://github.com/google/googletest) is used as the unit-testing framework. From 70af1c93969c6ef55e9100ac76f515abb9e44954 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 8 Oct 2024 18:48:31 +0200 Subject: [PATCH 12/41] Rename oneMKL in docs/building_the_project_with_dpcpp.rst --- docs/building_the_project_with_dpcpp.rst | 99 ++++++++++++------------ 1 file changed, 49 insertions(+), 50 deletions(-) diff --git a/docs/building_the_project_with_dpcpp.rst b/docs/building_the_project_with_dpcpp.rst index 8eb473596..780e0c6b3 100644 --- a/docs/building_the_project_with_dpcpp.rst +++ b/docs/building_the_project_with_dpcpp.rst @@ -3,15 +3,14 @@ Building the Project with DPC++ =============================== -This page describes building the oneMKL Interfaces with either the Intel(R) -oneAPI DPC++ Compiler or open-source oneAPI DPC++ Compiler. For guidance on -building the project with AdaptiveCpp, see -:ref:`building_the_project_with_adaptivecpp`. +This page describes building the oneMath with either the Intel(R) oneAPI DPC++ +Compiler or open-source oneAPI DPC++ Compiler. For guidance on building the +project with AdaptiveCpp, see :ref:`building_the_project_with_adaptivecpp`. .. _build_setup_with_dpcpp: Environment Setup -################## +################# #. Install the required DPC++ compiler (Intel(R) DPC++ or Open DPC++ - see @@ -19,16 +18,16 @@ Environment Setup #. Clone this project. The root directory of the cloned repository will be - referred to as ````. + referred to as ````. #. Build and install all `required dependencies - `_. + `_. .. _build_introduction_with_dpcpp: Build Commands -############### +############## The build commands for various compilers and backends differ mostly in setting the values of CMake options for compiler and backend. In this section, we @@ -40,7 +39,7 @@ for Windows`_ for building on Windows): .. code-block:: bash - # Inside + # Inside mkdir build && cd build cmake .. -DCMAKE_CXX_COMPILER=$CXX_COMPILER \ # Should be icpx or clang++ -DCMAKE_C_COMPILER=$C_COMPILER \ # Should be icx or clang @@ -57,13 +56,13 @@ In the above, the ``$CXX_COMPILER`` and ``$C_COMPILER`` should be set to ``icpx`` and ``icx`` respectively when using the Intel(R) oneAPI DPC++ Compiler, or ``clang++`` and ``clang`` respectively when using the Open DPC++ Compiler. -Backends should be enabled by setting ``-DENABLE__BACKEND=True`` for -each desired backend. By default, only the ``MKLGPU`` and ``MKLCPU`` backends -are enabled. Multiple backends for multiple device vendors can be enabled at -once (albeit with limitations when using portBLAS and portFFT). The supported -backends for the compilers are given in the table at `oneMKL supported -configurations table -`_, +Backends should be enabled by setting ``-DENABLE__BACKEND=True`` +for each desired backend. By default, only the ``MKLGPU`` and ``MKLCPU`` +backends are enabled. Multiple backends for multiple device vendors can be +enabled at once (albeit with limitations when using portBLAS and portFFT). The +supported backends for the compilers are given in the table at `oneMath +supported configurations table +`_, and the CMake option names are given in the table below. Some backends may require additional parameters to be set. See the relevant section below for additional guidance. @@ -145,13 +144,13 @@ Some additional build options are given in the section `Additional build options TARGET_DOMAINS ^^^^^^^^^^^^^^ -oneMKL supports multiple domains: BLAS, DFT, LAPACK, RNG and sparse BLAS. The -domains built by oneMKL can be selected using the ``TARGET_DOMAINS`` parameter. +oneMath supports multiple domains: BLAS, DFT, LAPACK, RNG and sparse BLAS. The +domains built by oneMath can be selected using the ``TARGET_DOMAINS`` parameter. In most cases, ``TARGET_DOMAINS`` is set automatically according to the domains supported by the backend libraries enabled. However, while most backend libraries support only one of these domains, but some may support multiple. For example, the ``MKLCPU`` backend supports every domain. To enable support for -only the BLAS domain in the oneMKL Interfaces whilst compiling with ``MKLCPU``, +only the BLAS domain in oneMath whilst compiling with ``MKLCPU``, ``TARGET_DOMAINS`` could be set to ``blas``. To enable BLAS and DFT, ``-DTARGET_DOMAINS="blas dft"`` would be used. @@ -159,12 +158,12 @@ only the BLAS domain in the oneMKL Interfaces whilst compiling with ``MKLCPU``, Backends ######### -.. _build_for_intel_onemkl_dpcpp: +.. _build_for_intel_onemath_dpcpp: Building for Intel(R) oneMKL ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ -The Intel(R) oneMKL backend supports multiple domains on both x86 CPUs and Intel +The Intel(R) oneMKL backends support multiple domains on both x86 CPUs and Intel GPUs. The MKLCPU backend using Intel(R) oneMKL for x86 CPU is enabled by default, and controlled with the parameter ``ENABLE_MKLCPU_BACKEND``. The MKLGPU backend using Intel(R) oneMKL for Intel GPU is enabled by default, and @@ -231,8 +230,8 @@ Building for other SYCL devices ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ SYCL enables portable heterogeneous computing on a wide range of accelerators. -Consequently, it is possible to use oneMKL Interfaces with accelerators not -anticipated by the oneMKL Interfaces team. +Consequently, it is possible to use oneMath with accelerators not anticipated by +the project. For generic SYCL devices, only portBLAS and portFFT backend are enabled. The user must set the appropriate ``-fsycl-targets`` for their device, and also @@ -243,14 +242,14 @@ unsupported configurations. .. _build_for_portlibs_dpcpp: Pure SYCL backends: portBLAS and portFFT -^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ `portBLAS `_ and `portFFT `_ are experimental pure-SYCL backends that work on all SYCL targets supported by the DPC++ compiler. Since they support multiple targets, they cannot be enabled with other backends in the -same domain, or the MKLCPU or MKLGPU backends. Both libraries are experimental -and currently only support a subset of operations and features. +same domain, or the ``MKLCPU`` or ``MKLGPU`` backends. Both libraries are +experimental and currently only support a subset of operations and features. For best performance, both libraries must be tuned. See the individual sections for more details. @@ -312,10 +311,10 @@ specified. See `DPC++ User Manual .. _build_additional_options_dpcpp: Additional Build Options -########################## +######################## -When building oneMKL the SYCL implementation can be specified by setting the -``ONEMKL_SYCL_IMPLEMENTATION`` option. Possible values are: +When building oneMath the SYCL implementation can be specified by setting the +``ONEMATH_SYCL_IMPLEMENTATION`` option. Possible values are: * ``dpc++`` (default) for the `Intel(R) oneAPI DPC++ Compiler `_ and for the `oneAPI @@ -348,20 +347,20 @@ The following table provides details of CMake options and their default values: .. note:: When building with ``BUILD_FUNCTIONAL_TESTS=True`` (default option) only single CUDA backend can be built - (`#270 `_). + (`#270 `_). .. _build_invocation_examples_dpcpp: CMake invocation examples -########################## +######################### -Build oneMKL with support for Nvidia GPUs with tests +Build oneMath with support for Nvidia GPUs with tests disabled using the Ninja build system: .. code-block:: bash - cmake $ONEMKL_DIR \ + cmake $ONEMATH_DIR \ -GNinja \ -DCMAKE_CXX_COMPILER=clang++ \ -DCMAKE_C_COMPILER=clang \ @@ -373,17 +372,17 @@ disabled using the Ninja build system: -DENABLE_CURAND_BACKEND=True \ -DBUILD_FUNCTIONAL_TESTS=False -``$ONEMKL_DIR`` points at the oneMKL source directly. The x86 CPU (``MKLCPU``) +``$ONEMATH_DIR`` points at the oneMath source directly. The x86 CPU (``MKLCPU``) and Intel GPU (``MKLGPU``) backends are enabled by default, but are disabled -here. The backends for Nvidia GPUs must all be explicilty enabled. The tests are +here. The backends for Nvidia GPUs must all be explicitly enabled. The tests are disabled, but the examples will still be built. -Building oneMKL with support for AMD GPUs with tests +Building oneMath with support for AMD GPUs with tests disabled: .. code-block:: bash - cmake $ONEMKL_DIR \ + cmake $ONEMATH_DIR \ -DCMAKE_CXX_COMPILER=clang++ \ -DCMAKE_C_COMPILER=clang \ -DENABLE_MKLCPU_BACKEND=False \ @@ -394,18 +393,18 @@ disabled: -DHIP_TARGETS=gfx90a \ -DBUILD_FUNCTIONAL_TESTS=False -``$ONEMKL_DIR`` points at the oneMKL source directly. The x86 CPU (``MKLCPU``) +``$ONEMATH_DIR`` points at the oneMath source directly. The x86 CPU (``MKLCPU``) and Intel GPU (``MKLGPU``) backends are enabled by default, but are disabled -here. The backends for AMD GPUs must all be explicilty enabled. The tests are +here. The backends for AMD GPUs must all be explicitly enabled. The tests are disabled, but the examples will still be built. -Build oneMKL for the DFT domain only with support for x86 CPU, Intel GPU, AMD +Build oneMath for the DFT domain only with support for x86 CPU, Intel GPU, AMD GPU and Nvidia GPU with testing enabled: .. code-block:: bash - cmake $ONEMKL_DIR \ + cmake $ONEMATH_DIR \ -DCMAKE_CXX_COMPILER=icpx \ -DCMAKE_C_COMPILER=icx \ -DENABLE_ROCFFT_BACKEND=True \ @@ -423,20 +422,20 @@ set, the backend libraries to enable the use of BLAS, LAPACK and RNG with MKLGPU and MKLCPU would also be enabled. The build of examples is disabled. Since functional testing was not disabled, tests would be built. -Build oneMKL for the BLAS domain on a generic SYCL device: +Build oneMath for the BLAS domain on a generic SYCL device: .. code-block:: bash - cmake $ONEMKL_DIR \ + cmake $ONEMATH_DIR \ -DCMAKE_CXX_COMPILER=clang++ \ -DCMAKE_C_COMPILER=clang \ -DENABLE_MKLCPU_BACKEND=False \ -DENABLE_MKLGPU_BACKEND=False \ -DENABLE_PORTBLAS_BACKEND=True -Note that this is not a tested configuration. This builds oneMKL Interfaces -with the portBLAS backend only, for a generic SYCL device supported by the -Open DPC++ project. +Note that this is not a tested configuration. This builds oneMath with the +portBLAS backend only, for a generic SYCL device supported by the Open DPC++ +project. Build oneMKL for the DFT domain on a generic SYCL device: @@ -479,12 +478,12 @@ Building for Windows #################### The Windows build is similar to the Linux build, albeit that `fewer backends are -supported `_. +supported `_. Additionally, the Ninja build system must be used. For example: .. code-block:: bash - # Inside + # Inside md build && cd build cmake .. -G Ninja [-DCMAKE_CXX_COMPILER=\bin\icx] # required only if icx is not found in environment variable PATH [-DCMAKE_C_COMPILER=\bin\icx] # required only if icx is not found in environment variable PATH @@ -501,14 +500,14 @@ Build FAQ ######### clangrt builtins lib not found - Encountered when trying to build oneMKL with some ROCm libraries. There are + Encountered when trying to build oneMath with some ROCm libraries. There are several possible solutions: * If building Open DPC++ from source, add ``compiler-rt`` to the external projects compile option: ``--llvm-external-projects compiler-rt``. * Manually set the variable ``HIP_CXX_COMPILER`` to HIP's toolkit ``clang++`` path, for instance ``-DHIP_CXX_COMPILER=/opt/rocm/6.1.0/llvm/bin/clang++``. - oneMKL may fail to link if the clang versions of ``icpx`` and ``rocm`` are + oneMath may fail to link if the clang versions of ``icpx`` and ``rocm`` are not compatible. Could NOT find CBLAS (missing: CBLAS file) From 1814a09e943f3e50395581aedd0e5a526a201979 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Wed, 9 Oct 2024 15:04:42 +0200 Subject: [PATCH 13/41] Rename mkl in CONTRIBUTING.md --- CONTRIBUTING.md | 52 ++++++++++++++++++++++++------------------------- 1 file changed, 26 insertions(+), 26 deletions(-) diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 342bbca5d..8f90769d9 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -1,5 +1,5 @@ # Contributing Guidelines -If you have improvements, new libraries integrated under oneAPI Math Kernel Library (oneMKL) Interfaces, or new interfaces to contribute to the oneMKL Specification, please send us your pull requests! For getting started, see GitHub [howto](https://help.github.com/en/articles/about-pull-requests). +If you have improvements, new libraries integrated under oneMath, or new interfaces to contribute to the oneMath Specification, please send us your pull requests! For getting started, see GitHub [howto](https://help.github.com/en/articles/about-pull-requests). For how to enable a new third-party library, see the [guidelines](docs/create_new_backend.rst). @@ -15,11 +15,11 @@ Before sending your pull requests, ensure that you follow this checklist: * Ensure that [unit tests](CONTRIBUTING.md#unit-tests) pass. Include logs from tests as attachments to the pull request. -* Ensure that corresponding [maintainer GitHub team](#onemkl-interfaces-maintainers) is assigned to the PR review. +* Ensure that corresponding [maintainer GitHub team](#onemath-maintainers) is assigned to the PR review. ## Library Functionality Guidelines -oneMKL focuses on the following criteria: +oneMath focuses on the following criteria: 1. *Performance*: Functionality that has highly optimized and extensively parallelized routines for applications that require maximum performance. @@ -31,13 +31,13 @@ oneMKL focuses on the following criteria: 3. *Complexity*: Functionality that is not trivial to implement directly or by combining existing primitives. -For the new API to become a part of the open source project, it should be accepted as part of [oneMKL spec](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemkl/source/). +For the new API to become a part of the open source project, it should be accepted as part of [oneMath spec](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemath/source/). ### Request for Comments Process For changes impacting the public API or any significant changes in the library, such as adding new backend or changes to the architecture, -please follow the [RFC process](https://github.com/uxlfoundation/oneMKL/tree/rfcs). +please follow the [RFC process](https://github.com/uxlfoundation/oneMath/tree/rfcs). Please also provide the following details as part of the RFC: @@ -45,33 +45,33 @@ Please also provide the following details as part of the RFC: * The definition of the function including the interface and semantics, and how this interface will be extendable for different HW implementations. -* What existing libraries have implementations of this function and can be used under the oneMKL interface. +* What existing libraries have implementations of this function and can be used under oneMath. -* Ensure that corresponding [maintainer GitHub team](#onemkl-interfaces-maintainers) is assigned to the RFC review. +* Ensure that corresponding [maintainer GitHub team](#onemath-maintainers) is assigned to the RFC review. ## Bug Reporting -If you find a bug or problem, please open a request under [Issues](https://github.com/uxlfoundation/oneMKL/issues). +If you find a bug or problem, please open a request under [Issues](https://github.com/uxlfoundation/oneMath/issues). ## Security Issues Report security issues to onemkl.maintainers@intel.com. -## oneMKL Interfaces Maintainers +## oneMath Maintainers For GitHub questions, issues, RFCs, or PRs you can contact maintainers via one of the following GitHub teams based on the topic: | GitHub team name | Description | :-----------|:------------| -| @uxlfoundation/onemath-maintain | All oneMKL Interfaces maintainers | -| @uxlfoundation/onemath-arch-write | oneMKL Interfaces Architecture maintainers | -| @uxlfoundation/onemath-blas-write | oneMKL Interfaces BLAS maintainers | -| @uxlfoundation/onemath-dft-write | oneMKL Interfaces DFT maintainers | -| @uxlfoundation/onemath-lapack-write | oneMKL Interfaces LAPACK maintainers | -| @uxlfoundation/onemath-rng-write | oneMKL Interfaces RNG maintainers | -| @uxlfoundation/onemath-sparse-write | oneMKL Interfaces Sparse Algebra maintainers | -| @uxlfoundation/onemath-vm-write | oneMKL Interfaces Vector Math maintainers | +| @uxlfoundation/onemath-maintain | All oneMath maintainers | +| @uxlfoundation/onemath-arch-write | oneMath Architecture maintainers | +| @uxlfoundation/onemath-blas-write | oneMath BLAS maintainers | +| @uxlfoundation/onemath-dft-write | oneMath DFT maintainers | +| @uxlfoundation/onemath-lapack-write | oneMath LAPACK maintainers | +| @uxlfoundation/onemath-rng-write | oneMath RNG maintainers | +| @uxlfoundation/onemath-sparse-write | oneMath Sparse Algebra maintainers | +| @uxlfoundation/onemath-vm-write | oneMath Vector Math maintainers | Please read [MAINTAINERS page](MAINTAINERS.md) for more information about maintainer roles, responsibilities, and how to become one of them. @@ -81,7 +81,7 @@ The general principle is to follow the style of existing/surrounding code. If yo ```sh clang-format -style=file -i foo.cpp ``` -This formats code using the `_clang_format` file found in the oneMKL top-level directory. +This formats code using the `_clang_format` file found in the oneMath top-level directory. ### GN: General Naming @@ -168,27 +168,27 @@ files in the following order: ### NS: Namespaces * **NS1:** Use snake_case: all lowercase, with underscores "_" between words for all namespaces. -* **NS2:** The name of a top-level namespace must be the name of the project (oneMKL). +* **NS2:** The name of a top-level namespace must be the name of the project (oneMath). * **NS3:** Do not indent content inside a namespace scope. ```c // Wrong! Do not indent namespace oneapi { -namespace mkl { +namespace math { class table { }; -} // namespace mkl +} // namespace math } // namespace oneapi // Right namespace oneapi { -namespace mkl { +namespace math { class table { }; -} // namespace mkl +} // namespace math } // namespace oneapi ``` @@ -198,13 +198,13 @@ class table { }; #include "oneapi/math/blas/path_to_some_header.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { /* ... */ } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi ``` @@ -342,6 +342,6 @@ for (int i = 0; i < loop_size; i++) ...; ## Unit Tests -oneMKL uses GoogleTest for functional testing. For more information about how to build and run Unit Tests please see [Building and Running Tests](https://uxlfoundation.github.io/oneMKL/building_and_running_tests.html). +oneMath uses GoogleTest for functional testing. For more information about how to build and run Unit Tests please see [Building and Running Tests](https://uxlfoundation.github.io/oneMath/building_and_running_tests.html). Be sure to extend the existing tests when fixing an issue, adding a new interface or new implementation under existing interfaces. From 85fd50d6f86e628e4082f1d80f5041179ecb60e5 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Wed, 9 Oct 2024 17:44:50 +0200 Subject: [PATCH 14/41] Rename oneMKL in READMEs --- README.md | 82 +++++++++++++++++++++++----------------------- docs/README.md | 4 +-- examples/README.md | 14 ++++---- tests/README.md | 4 +-- 4 files changed, 51 insertions(+), 53 deletions(-) diff --git a/README.md b/README.md index f79c40913..ad334072a 100644 --- a/README.md +++ b/README.md @@ -1,25 +1,25 @@ UXL Foundation Logo -# oneAPI Math Kernel Library (oneMKL) Interfaces +# oneMath -oneMKL Interfaces is an open-source implementation of the oneMKL Data Parallel C++ (DPC++) interface according to the [oneMKL specification](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemkl/source/). It works with multiple devices (backends) using device-specific libraries underneath. +oneMath is an open-source implementation of the [oneMath specification](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemath/source/). It can work with multiple devices using multiple libraries (backends) underneath. The oneMath project was previously referred to as oneMKL Interface. -oneMKL is part of the [UXL Foundation](http://www.uxlfoundation.org). +oneMath is part of the [UXL Foundation](http://www.uxlfoundation.org).

- + - - + + @@ -87,9 +87,9 @@ oneMKL is part of the [UXL Foundation](http://www.uxlfoundation.org). #### Host API -There are two oneMKL selector layer implementations: +There are two oneMath selector layer implementations: -- **Run-time dispatching**: The application is linked with the oneMKL library and the required backend is loaded at run-time based on device vendor (all libraries should be dynamic). +- **Run-time dispatching**: The application is linked with the oneMath library and the required backend is loaded at run-time based on device vendor (all libraries should be dynamic). Example of app.cpp with run-time dispatching: @@ -103,19 +103,19 @@ There are two oneMKL selector layer implementations: sycl::queue cpu_queue(cpu_dev); sycl::queue gpu_queue(gpu_dev); - oneapi::mkl::blas::column_major::gemm(cpu_queue, transA, transB, m, ...); - oneapi::mkl::blas::column_major::gemm(gpu_queue, transA, transB, m, ...); + oneapi::math::blas::column_major::gemm(cpu_queue, transA, transB, m, ...); + oneapi::math::blas::column_major::gemm(gpu_queue, transA, transB, m, ...); ``` How to build an application with run-time dispatching: if OS is Linux, use icpx compiler. If OS is Windows, use icx compiler. Linux example: ```cmd - $> icpx -fsycl –I$ONEMKL/include app.cpp - $> icpx -fsycl app.o –L$ONEMKL/lib –lonemkl + $> icpx -fsycl –I$ONEMATH/include app.cpp + $> icpx -fsycl app.o –L$ONEMATH/lib –lonemath ``` -- **Compile-time dispatching**: The application uses a templated backend selector API where the template parameters specify the required backends and third-party libraries and the application is linked with the required oneMKL backend wrapper libraries (libraries can be static or dynamic). +- **Compile-time dispatching**: The application uses a templated backend selector API where the template parameters specify the required backends and third-party libraries and the application is linked with the required oneMath backend wrapper libraries (libraries can be static or dynamic). Example of app.cpp with compile-time dispatching: @@ -129,23 +129,23 @@ There are two oneMKL selector layer implementations: sycl::queue cpu_queue(cpu_dev); sycl::queue gpu_queue(gpu_dev); - oneapi::mkl::backend_selector cpu_selector(cpu_queue); + oneapi::math::backend_selector cpu_selector(cpu_queue); - oneapi::mkl::blas::column_major::gemm(cpu_selector, transA, transB, m, ...); - oneapi::mkl::blas::column_major::gemm(oneapi::mkl::backend_selector {gpu_queue}, transA, transB, m, ...); + oneapi::math::blas::column_major::gemm(cpu_selector, transA, transB, m, ...); + oneapi::math::blas::column_major::gemm(oneapi::math::backend_selector {gpu_queue}, transA, transB, m, ...); ``` How to build an application with compile-time dispatching: ```cmd - $> clang++ -fsycl –I$ONEMKL/include app.cpp - $> clang++ -fsycl app.o –L$ONEMKL/lib –lonemath_blas_mklcpu –lonemath_blas_cublas + $> clang++ -fsycl –I$ONEMATH/include app.cpp + $> clang++ -fsycl app.o –L$ONEMATH/lib –lonemath_blas_mklcpu –lonemath_blas_cublas ``` -*Refer to [Selecting a Compiler](https://uxlfoundation.github.io/oneMKL/selecting_a_compiler.html) for the choice between `icpx/icx` and `clang++` compilers.* +*Refer to [Selecting a Compiler](https://uxlfoundation.github.io/oneMath/selecting_a_compiler.html) for the choice between `icpx/icx` and `clang++` compilers.* #### Device API -Header-based and backend-independent Device API can be called within ```sycl kernel``` or work from Host code ([device-rng-usage-model-example](https://spec.oneapi.io/versions/latest/elements/oneMKL/source/domains/rng/device_api/device-rng-usage-model.html#id2)). Currently, the following domains support the Device API: +Header-based and backend-independent Device API can be called within ```sycl kernel``` or work from Host code ([device-rng-usage-model-example](https://spec.oneapi.io/versions/latest/elements/oneMath/source/domains/rng/device_api/device-rng-usage-model.html#id2)). Currently, the following domains support the Device API: - **RNG**. To use RNG Device API functionality it's required to include ```oneapi/math/rng/device.hpp``` header file. @@ -534,22 +534,22 @@ Product | Supported Version | License --- ## Documentation -- [Contents](https://uxlfoundation.github.io/oneMKL/) -- [About](https://uxlfoundation.github.io/oneMKL/introduction.html) +- [Contents](https://uxlfoundation.github.io/oneMath/) +- [About](https://uxlfoundation.github.io/oneMath/introduction.html) - Get Started - - [Selecting a Compiler](https://uxlfoundation.github.io/oneMKL/selecting_a_compiler.html) - - [Building the Project with DPC++](https://uxlfoundation.github.io/oneMKL/building_the_project_with_dpcpp.html) - - [Building the Project with AdaptiveCpp](https://uxlfoundation.github.io/oneMKL/building_the_project_with_adaptivecpp.html) + - [Selecting a Compiler](https://uxlfoundation.github.io/oneMath/selecting_a_compiler.html) + - [Building the Project with DPC++](https://uxlfoundation.github.io/oneMath/building_the_project_with_dpcpp.html) + - [Building the Project with AdaptiveCpp](https://uxlfoundation.github.io/oneMath/building_the_project_with_adaptivecpp.html) - Developer Reference - - [oneMKL Defined Datatypes](https://uxlfoundation.github.io/oneMKL/onemkl-datatypes.html) - - [Dense Linear Algebra](https://uxlfoundation.github.io/oneMKL/domains/dense_linear_algebra.html) -- [Integrating a Third-Party Library](https://uxlfoundation.github.io/oneMKL/create_new_backend.html) + - [oneMath Defined Datatypes](https://uxlfoundation.github.io/oneMath/onemath-datatypes.html) + - [Dense Linear Algebra](https://uxlfoundation.github.io/oneMath/domains/dense_linear_algebra.html) +- [Integrating a Third-Party Library](https://uxlfoundation.github.io/oneMath/create_new_backend.html) --- ## Governance -The oneMKL Interfaces project is governed by the UXL Foundation and you can get involved in this project in multiple ways. It is possible to join the [Math Special Interest Group (SIG)](https://github.com/uxlfoundation/foundation/tree/main/math) meetings where the group discusses and demonstrates work using this project. Members can also join the Open Source and Specification Working Group meetings. +The oneMath project is governed by the UXL Foundation and you can get involved in this project in multiple ways. It is possible to join the [Math Special Interest Group (SIG)](https://github.com/uxlfoundation/foundation/tree/main/math) meetings where the group discusses and demonstrates work using this project. Members can also join the Open Source and Specification Working Group meetings. You can also join the mailing lists for the [UXL Foundation](https://lists.uxlfoundation.org/g/main/subgroups) to be informed of when meetings are happening and receive the latest information and discussions. @@ -557,7 +557,7 @@ You can also join the mailing lists for the [UXL Foundation](https://lists.uxlfo ## Contributing -You can contribute to this project and also contribute to [the specification for this project](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemkl/source/). Please read the [CONTRIBUTING](CONTRIBUTING.md) page for more information. You can also contact oneMKL developers and maintainers via [UXL Foundation Slack](https://slack-invite.uxlfoundation.org/) using [#onemkl](https://uxlfoundation.slack.com/archives/onemkl) channel. +You can contribute to this project and also contribute to [the specification for this project](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemath/source/). Please read the [CONTRIBUTING](CONTRIBUTING.md) page for more information. You can also contact oneMath developers and maintainers via [UXL Foundation Slack](https://slack-invite.uxlfoundation.org/) using [#onemath](https://uxlfoundation.slack.com/archives/onemath) channel. --- @@ -569,31 +569,31 @@ Distributed under the Apache license 2.0. See [LICENSE](LICENSE) for more inform ## FAQs -### oneMKL +### oneMath -**Q: What is the difference between the following oneMKL items?** - - The [oneAPI Specification for oneMKL](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemkl/source/) - - The [oneAPI Math Kernel Library (oneMKL) Interfaces](https://github.com/uxlfoundation/oneMKL) Project +**Q: What is the difference between the following items?** + - The [oneAPI Specification for oneMath](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemath/source/) + - The [oneMath](https://github.com/uxlfoundation/oneMath) Project - The [Intel(R) oneAPI Math Kernel Library (oneMKL)](https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onemkl.html) Product **A:** -- The [oneAPI Specification for oneMKL](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemkl/source/) defines the DPC++ interfaces for performance math library functions. The oneMKL specification can evolve faster and more frequently than implementations of the specification. +- The [oneAPI Specification for oneMath](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemath/source/) defines the SYCL interfaces for performance math library functions. The oneMath specification can evolve faster and more frequently than implementations of the specification. -- The [oneAPI Math Kernel Library (oneMKL) Interfaces](https://github.com/uxlfoundation/oneMKL) Project is an open source implementation of the specification. The project goal is to demonstrate how the DPC++ interfaces documented in the oneMKL specification can be implemented for any math library and work for any target hardware. While the implementation provided here may not yet be the full implementation of the specification, the goal is to build it out over time. We encourage the community to contribute to this project and help to extend support to multiple hardware targets and other math libraries. +- The [oneMath](https://github.com/uxlfoundation/oneMath) project is an open source implementation of the specification. The project goal is to demonstrate how the SYCL interfaces documented in the oneMath specification can be implemented for any math library and work for any target hardware. While the implementation provided here may not yet be the full implementation of the specification, the goal is to build it out over time. We encourage the community to contribute to this project and help to extend support to multiple hardware targets and other math libraries. -- The [Intel(R) oneAPI Math Kernel Library (oneMKL)](https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onemkl.html) product is the Intel product implementation of the specification (with DPC++ interfaces) as well as similar functionality with C and Fortran interfaces, and is provided as part of Intel® oneAPI Base Toolkit. It is highly optimized for Intel CPU and Intel GPU hardware. +- The [Intel(R) oneAPI Math Kernel Library (oneMKL)](https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onemkl.html) project is an Intel product provided as part of the Intel(R) oneAPI Base Toolkit. It is used for the Intel backends of oneMath. Its C++ API is very similar to the oneMath specification. It is highly optimized for Intel CPU and Intel GPU hardware. -**Q: I'm trying to use oneMKL Interfaces in my project using `FetchContent`**, but I keep running into `ONEMKL::SYCL::SYCL target was not found` problem when I try to build the project. What should I do? +**Q: I'm trying to use oneMath in my project using `FetchContent`**, but I keep running into `ONEMATH::SYCL::SYCL target was not found` problem when I try to build the project. What should I do? **A:** Make sure you set the compiler when you configure your project. E.g. `cmake -Bbuild . -DCMAKE_CXX_COMPILER=icpx`. -**Q: I'm trying to use oneMKL Interfaces in my project using `find_package(oneMKL)`.** I set oneMKL/oneTBB and Compiler environment first, then I built and installed oneMKL Interfaces, and finally I tried to build my project using installed oneMKL Interfaces (e.g. like this `cmake -Bbuild -GNinja -DCMAKE_CXX_COMPILER=icpx -DoneMKL_ROOT= .`) and I noticed that cmake includes installed oneMKL Interfaces headers as a system include which ends up as a lower priority than the installed oneMKL package includes which I set before for building oneMKL Interfaces. As a result, I get conflicts between oneMKL and installed oneMKL Interfaces headers. What should I do? +**Q: I'm trying to use oneMath in my project using `find_package(oneMath)`.** I set oneMath/oneTBB and Compiler environment first, then I built and installed oneMath, and finally I tried to build my project using installed oneMath (e.g. like this `cmake -Bbuild -GNinja -DCMAKE_CXX_COMPILER=icpx -DoneMath_ROOT= .`) and I noticed that cmake includes installed oneMath headers as a system include which ends up as a lower priority than the installed Intel(R) oneAPI Math Kernel Library package includes which I set before for building oneMath. As a result, I get conflicts between Intel(R) oneAPI Math Kernel Library and installed oneMath headers. What should I do? **A:** -Having installed oneMKL Interfaces headers as `-I` instead on system includes (as `-isystem`) helps to resolve this problem. We use `INTERFACE_INCLUDE_DIRECTORIES` to add paths to installed oneMKL Interfaces headers (check `oneMKLTargets.cmake` in `lib/cmake` to find it). It's a known limitation that `INTERFACE_INCLUDE_DIRECTORIES` puts headers paths as system headers. To avoid that: -- Option 1: Use CMake >=3.25. In this case oneMKL Interfaces will be built with `EXPORT_NO_SYSTEM` property set to `true` and you won't see the issue. +Having installed oneMath headers as `-I` instead on system includes (as `-isystem`) helps to resolve this problem. We use `INTERFACE_INCLUDE_DIRECTORIES` to add paths to installed oneMath headers (check `oneMathTargets.cmake` in `lib/cmake` to find it). It's a known limitation that `INTERFACE_INCLUDE_DIRECTORIES` puts headers paths as system headers. To avoid that: +- Option 1: Use CMake >=3.25. In this case oneMath will be built with `EXPORT_NO_SYSTEM` property set to `true` and you won't see the issue. - Option 2: If you use CMake < 3.25, set `PROPERTIES NO_SYSTEM_FROM_IMPORTED true` for your target. E.g: `set_target_properties(test PROPERTIES NO_SYSTEM_FROM_IMPORTED true)`. --- diff --git a/docs/README.md b/docs/README.md index 040d22aac..0e1d70187 100644 --- a/docs/README.md +++ b/docs/README.md @@ -1,6 +1,6 @@ -### oneMKL documentation +### oneMath documentation -This folder contains oneMKL documentation in reStructuredText (rST) format. +This folder contains oneMath documentation in reStructuredText (rST) format. The documentation build step is skipped by default. To enable building documentation from the main build, set `-DBUILD_DOC=ON`. diff --git a/examples/README.md b/examples/README.md index 0dad8772d..46730b900 100644 --- a/examples/README.md +++ b/examples/README.md @@ -1,20 +1,18 @@ -# oneAPI Math Kernel Library (oneMKL) Interfaces Examples -oneAPI Math Kernel Library (oneMKL) Interfaces offers examples with the following routines: -- blas: level3/gemm_usm -- rng: uniform_usm +# oneMath Examples +oneMath offers examples with the following routines: +- blas: level3/gemm_usm +- rng: uniform_usm - lapack: getrs_usm - dft: complex_fwd_usm, real_fwd_usm - sparse_blas: sparse_gemv_usm Each routine has one run-time dispatching example and one compile-time dispatching example (which uses both mklcpu and cuda backends), located in `example/<$domain>/run_time_dispatching` and `example/<$domain>/compile_time_dispatching` subfolders, respectively. -To build examples, use cmake build option `-DBUILD_EXAMPLES=true`. +To build examples, use cmake build option `-DBUILD_EXAMPLES=true`. Compile_time_dispatching will be built if `-DBUILD_EXAMPLES=true` and cuda backend is enabled, because the compile-time dispatching example runs on both mklcpu and cuda backends. Run_time_dispatching will be built if `-DBUILD_EXAMPLES=true` and `-DBUILD_SHARED_LIBS=true`. -The example executable naming convention follows `example_<$domain>_<$routine>_<$backend>` for compile-time dispatching examples - or `example_<$domain>_<$routine>` for run-time dispatching examples. - E.g. `example_blas_gemm_usm_mklcpu_cublas ` `example_blas_gemm_usm` +The example executable naming convention follows `example_<$domain>_<$routine>_<$backend>` for compile-time dispatching examples or `example_<$domain>_<$routine>` for run-time dispatching examples. E.g. `example_blas_gemm_usm_mklcpu_cublas ` `example_blas_gemm_usm` ## Example outputs (blas, rng, lapack, dft, sparse_blas) diff --git a/tests/README.md b/tests/README.md index cbb87e8ff..6c3ee1584 100644 --- a/tests/README.md +++ b/tests/README.md @@ -1,9 +1,9 @@ -# oneMKL Interfaces Testing +# oneMath Testing ## Overview Inside the `unit_tests` directory, there are domain-level directories which contain domain-specific tests, usually per function or per configuration. -See [Building and Running Tests](https://uxlfoundation.github.io/oneMKL/building_and_running_tests.html) documentation for more information about how to build and run the tests. +See [Building and Running Tests](https://uxlfoundation.github.io/oneMath/building_and_running_tests.html) documentation for more information about how to build and run the tests. [GoogleTest](https://github.com/google/googletest) is used as the unit-testing framework. From 41a1cb86ede9d8b2d764ee1c455746e68fb82252 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Wed, 9 Oct 2024 18:05:41 +0200 Subject: [PATCH 15/41] oneMath interface -> oneMath --- README.md | 2 +- docs/create_new_backend.rst | 4 ++-- src/CMakeLists.txt | 4 ++-- src/dft/backends/cufft/CMakeLists.txt | 4 ++-- src/dft/backends/mklcpu/CMakeLists.txt | 4 ++-- src/dft/backends/mklgpu/CMakeLists.txt | 2 +- src/dft/backends/portfft/CMakeLists.txt | 4 ++-- src/dft/backends/rocfft/CMakeLists.txt | 4 ++-- 8 files changed, 14 insertions(+), 14 deletions(-) diff --git a/README.md b/README.md index ad334072a..7ab3d1c0e 100644 --- a/README.md +++ b/README.md @@ -18,7 +18,7 @@ oneMath is part of the [UXL Foundation](http://www.uxlfoundation.org). - + diff --git a/docs/create_new_backend.rst b/docs/create_new_backend.rst index 6cdebe394..ecc79a631 100644 --- a/docs/create_new_backend.rst +++ b/docs/create_new_backend.rst @@ -3,8 +3,8 @@ .. _create_backend_wrappers: -Integrating a Third-Party Library to oneAPI Math Kernel Library (oneMath) Interfaces -==================================================================================== +Integrating a Third-Party Library to oneMath +============================================ This step-by-step tutorial provides examples for enabling new third-party libraries in oneMath. diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index 6ccb6578b..1e4cf917d 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -27,7 +27,7 @@ endif() set(ONEMATH_GENERATED_INCLUDE_PATH ${CMAKE_CURRENT_BINARY_DIR}) -set(ONEMATH_INTERFACE_INCLUDE_DIRS +set(ONEMATH_INCLUDE_DIRS $ $ $ @@ -84,7 +84,7 @@ if(BUILD_SHARED_LIBS) add_dependencies(onemath onemath_backend_libs) target_include_directories(onemath - PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INCLUDE_DIRS} ) set_target_properties(onemath PROPERTIES SOVERSION ${PROJECT_VERSION_MAJOR} diff --git a/src/dft/backends/cufft/CMakeLists.txt b/src/dft/backends/cufft/CMakeLists.txt index a4309fd18..dbbe92316 100644 --- a/src/dft/backends/cufft/CMakeLists.txt +++ b/src/dft/backends/cufft/CMakeLists.txt @@ -32,10 +32,10 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INCLUDE_DIRS} ) target_include_directories(${LIB_NAME} - PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INCLUDE_DIRS} ) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src diff --git a/src/dft/backends/mklcpu/CMakeLists.txt b/src/dft/backends/mklcpu/CMakeLists.txt index 682bdb929..b6eb2a674 100644 --- a/src/dft/backends/mklcpu/CMakeLists.txt +++ b/src/dft/backends/mklcpu/CMakeLists.txt @@ -33,10 +33,10 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INCLUDE_DIRS} ) target_include_directories(${LIB_NAME} - PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INCLUDE_DIRS} ) target_include_directories(${LIB_OBJ} diff --git a/src/dft/backends/mklgpu/CMakeLists.txt b/src/dft/backends/mklgpu/CMakeLists.txt index 2d9b3b4e7..3617290a1 100644 --- a/src/dft/backends/mklgpu/CMakeLists.txt +++ b/src/dft/backends/mklgpu/CMakeLists.txt @@ -33,7 +33,7 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_NAME} - PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INCLUDE_DIRS} ) # Due to using the same file name for different headers in this library and in diff --git a/src/dft/backends/portfft/CMakeLists.txt b/src/dft/backends/portfft/CMakeLists.txt index fcabcc19b..0c3d59b49 100644 --- a/src/dft/backends/portfft/CMakeLists.txt +++ b/src/dft/backends/portfft/CMakeLists.txt @@ -91,10 +91,10 @@ endif() target_link_libraries(${LIB_OBJ} PRIVATE onemath_warnings) target_include_directories(${LIB_OBJ} - PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INCLUDE_DIRS} ) target_include_directories(${LIB_NAME} - PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INCLUDE_DIRS} ) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src diff --git a/src/dft/backends/rocfft/CMakeLists.txt b/src/dft/backends/rocfft/CMakeLists.txt index a02877398..6e0bf6034 100644 --- a/src/dft/backends/rocfft/CMakeLists.txt +++ b/src/dft/backends/rocfft/CMakeLists.txt @@ -32,10 +32,10 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_dft ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INCLUDE_DIRS} ) target_include_directories(${LIB_NAME} - PUBLIC ${ONEMATH_INTERFACE_INCLUDE_DIRS} + PUBLIC ${ONEMATH_INCLUDE_DIRS} ) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src From 27503c13eed70855a5c6f7a29ac0e852211d2c6a Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 8 Oct 2024 18:12:45 +0200 Subject: [PATCH 16/41] Rename mkl namespace to math --- docs/create_new_backend.rst | 22 +- docs/domains/sparse_linear_algebra.rst | 26 +- .../level3/gemm_usm_mklcpu_cublas.cpp | 32 +- .../run_time_dispatching/level3/gemm_usm.cpp | 26 +- .../complex_fwd_usm_mklcpu_cufft.cpp | 18 +- .../dft/run_time_dispatching/real_fwd_usm.cpp | 14 +- examples/include/example_helper.hpp | 4 +- .../getrs_usm_mklcpu_cusolver.cpp | 48 +- .../lapack/run_time_dispatching/getrs_usm.cpp | 22 +- .../uniform_usm_mklcpu_curand.cpp | 18 +- examples/rng/device/uniform.cpp | 14 +- .../rng/run_time_dispatching/uniform_usm.cpp | 10 +- .../sparse_blas_spmv_usm_mklcpu.cpp | 50 +- .../sparse_blas_spmv_usm.cpp | 48 +- include/oneapi/math/bfloat16.hpp | 4 +- include/oneapi/math/blas.hpp | 4 +- .../math/blas/detail/blas_ct_backends.hpp | 4 +- .../oneapi/math/blas/detail/blas_loader.hpp | 4 +- .../oneapi/math/blas/detail/blas_loader.hxx | 980 ++++---- .../math/blas/detail/cublas/blas_ct.hpp | 4 +- .../math/blas/detail/cublas/blas_ct.hxx | 980 ++++---- .../detail/cublas/onemath_blas_cublas.hpp | 14 +- .../math/blas/detail/mklcpu/blas_ct.hpp | 4 +- .../math/blas/detail/mklcpu/blas_ct.hxx | 980 ++++---- .../detail/mklcpu/onemath_blas_mklcpu.hpp | 14 +- .../math/blas/detail/mklgpu/blas_ct.hpp | 4 +- .../math/blas/detail/mklgpu/blas_ct.hxx | 980 ++++---- .../detail/mklgpu/onemath_blas_mklgpu.hpp | 4 +- .../math/blas/detail/netlib/blas_ct.hpp | 4 +- .../math/blas/detail/netlib/blas_ct.hxx | 980 ++++---- .../detail/netlib/onemath_blas_netlib.hpp | 14 +- .../blas/detail/onemath_blas_backends.hxx | 1088 ++++----- .../math/blas/detail/portblas/blas_ct.hpp | 4 +- .../math/blas/detail/portblas/blas_ct.hxx | 980 ++++---- .../detail/portblas/onemath_blas_portblas.hpp | 14 +- .../math/blas/detail/rocblas/blas_ct.hpp | 4 +- .../math/blas/detail/rocblas/blas_ct.hxx | 980 ++++---- .../detail/rocblas/onemath_blas_rocblas.hpp | 14 +- .../oneapi/math/detail/backend_selector.hpp | 4 +- .../detail/backend_selector_predicates.hpp | 4 +- include/oneapi/math/detail/backends.hpp | 4 +- include/oneapi/math/detail/backends_table.hpp | 4 +- include/oneapi/math/detail/exceptions.hpp | 16 +- include/oneapi/math/detail/get_device_id.hpp | 8 +- include/oneapi/math/dft/backward.hpp | 4 +- include/oneapi/math/dft/descriptor.hpp | 4 +- .../oneapi/math/dft/detail/commit_impl.hpp | 14 +- .../dft/detail/cufft/onemath_dft_cufft.hpp | 4 +- .../math/dft/detail/descriptor_impl.hpp | 4 +- include/oneapi/math/dft/detail/dft_loader.hpp | 4 +- .../dft/detail/external_workspace_helper.hpp | 16 +- .../dft/detail/mklcpu/onemath_dft_mklcpu.hpp | 4 +- .../dft/detail/mklgpu/onemath_dft_mklgpu.hpp | 4 +- .../detail/portfft/onemath_dft_portfft.hpp | 4 +- .../dft/detail/rocfft/onemath_dft_rocfft.hpp | 4 +- include/oneapi/math/dft/detail/types_impl.hpp | 4 +- include/oneapi/math/dft/forward.hpp | 4 +- include/oneapi/math/dft/types.hpp | 4 +- include/oneapi/math/exceptions.hpp | 40 +- .../math/lapack/detail/cusolver/lapack_ct.hpp | 4 +- .../math/lapack/detail/cusolver/lapack_ct.hxx | 1064 ++++----- .../cusolver/onemath_lapack_cusolver.hpp | 4 +- .../cusolver/onemath_lapack_cusolver.hxx | 478 ++-- .../math/lapack/detail/lapack_loader.hpp | 1478 ++++++------ .../oneapi/math/lapack/detail/lapack_rt.hpp | 542 ++--- .../lapack/detail/mkl_common/lapack_ct.hxx | 1114 ++++----- .../mkl_common/onemath_lapack_backends.hxx | 758 +++--- .../math/lapack/detail/mklcpu/lapack_ct.hpp | 4 +- .../detail/mklcpu/onemath_lapack_mklcpu.hpp | 4 +- .../math/lapack/detail/mklgpu/lapack_ct.hpp | 4 +- .../detail/mklgpu/onemath_lapack_mklgpu.hpp | 4 +- .../lapack/detail/rocsolver/lapack_ct.hpp | 4 +- .../lapack/detail/rocsolver/lapack_ct.hxx | 1064 ++++----- .../rocsolver/onemath_lapack_rocsolver.hpp | 4 +- .../rocsolver/onemath_lapack_rocsolver.hxx | 478 ++-- include/oneapi/math/lapack/exceptions.hpp | 36 +- include/oneapi/math/lapack/types.hpp | 4 +- .../rng/detail/curand/onemath_rng_curand.hpp | 12 +- .../oneapi/math/rng/detail/engine_impl.hpp | 4 +- .../rng/detail/mklcpu/onemath_rng_mklcpu.hpp | 12 +- .../rng/detail/mklgpu/onemath_rng_mklgpu.hpp | 12 +- include/oneapi/math/rng/detail/rng_loader.hpp | 12 +- .../detail/rocrand/onemath_rng_rocrand.hpp | 12 +- .../math/rng/device/detail/bernoulli_impl.hpp | 10 +- .../math/rng/device/detail/beta_impl.hpp | 18 +- .../math/rng/device/detail/bits_impl.hpp | 6 +- .../rng/device/detail/distribution_base.hpp | 4 +- .../math/rng/device/detail/engine_base.hpp | 4 +- .../rng/device/detail/exponential_impl.hpp | 22 +- .../math/rng/device/detail/gamma_impl.hpp | 18 +- .../math/rng/device/detail/gaussian_impl.hpp | 22 +- .../math/rng/device/detail/lognormal_impl.hpp | 12 +- .../math/rng/device/detail/mcg31m1_impl.hpp | 18 +- .../math/rng/device/detail/mcg59_impl.hpp | 18 +- .../math/rng/device/detail/mrg32k3a_impl.hpp | 22 +- .../detail/mrg32k3a_skip_ahead_matrix.hpp | 4 +- .../rng/device/detail/philox4x32x10_impl.hpp | 24 +- .../math/rng/device/detail/poisson_impl.hpp | 14 +- .../rng/device/detail/uniform_bits_impl.hpp | 6 +- .../math/rng/device/detail/uniform_impl.hpp | 10 +- .../math/rng/device/detail/vm_wrappers.hpp | 4 +- .../oneapi/math/rng/device/distributions.hpp | 50 +- include/oneapi/math/rng/device/engines.hpp | 12 +- include/oneapi/math/rng/device/functions.hpp | 4 +- include/oneapi/math/rng/device/types.hpp | 4 +- include/oneapi/math/rng/distributions.hpp | 46 +- include/oneapi/math/rng/engines.hpp | 8 +- include/oneapi/math/rng/functions.hpp | 14 +- include/oneapi/math/rng/predicates.hpp | 10 +- .../math/sparse_blas/detail/handles.hpp | 4 +- .../math/sparse_blas/detail/helper_types.hpp | 4 +- .../mklcpu/onemath_sparse_blas_mklcpu.hpp | 6 +- .../detail/mklcpu/sparse_blas_ct.hpp | 4 +- .../mklgpu/onemath_sparse_blas_mklgpu.hpp | 6 +- .../detail/mklgpu/sparse_blas_ct.hpp | 4 +- .../detail/onemath_sparse_blas_backends.hxx | 32 +- .../sparse_blas/detail/operation_types.hpp | 4 +- .../sparse_blas/detail/sparse_blas_ct.hxx | 28 +- .../sparse_blas/detail/sparse_blas_rt.hpp | 30 +- .../oneapi/math/sparse_blas/matrix_view.hpp | 4 +- include/oneapi/math/sparse_blas/types.hpp | 4 +- include/oneapi/math/types.hpp | 4 +- scripts/generate_backend_api.py | 4 +- scripts/generate_ct_instant.py | 4 +- scripts/generate_ct_templates.py | 2 +- src/blas/backends/backend_wrappers.cxx | 980 ++++---- src/blas/backends/cublas/cublas_batch.cpp | 10 +- .../backends/cublas/cublas_extensions.cpp | 20 +- src/blas/backends/cublas/cublas_handle.hpp | 4 +- src/blas/backends/cublas/cublas_helper.hpp | 30 +- src/blas/backends/cublas/cublas_level1.cpp | 8 +- src/blas/backends/cublas/cublas_level2.cpp | 4 +- src/blas/backends/cublas/cublas_level3.cpp | 6 +- .../backends/cublas/cublas_scope_handle.cpp | 4 +- .../backends/cublas/cublas_scope_handle.hpp | 4 +- .../cublas/cublas_scope_handle_hipsycl.cpp | 4 +- .../cublas/cublas_scope_handle_hipsycl.hpp | 4 +- src/blas/backends/cublas/cublas_task.hpp | 4 +- src/blas/backends/cublas/cublas_wrappers.cpp | 1960 +++++++-------- .../backends/mkl_common/mkl_blas_backend.hpp | 4 +- src/blas/backends/mklcpu/mklcpu_batch.cpp | 8 +- .../backends/mklcpu/mklcpu_extensions.cpp | 8 +- src/blas/backends/mklcpu/mklcpu_level1.cpp | 8 +- src/blas/backends/mklcpu/mklcpu_level2.cpp | 8 +- src/blas/backends/mklcpu/mklcpu_level3.cpp | 8 +- src/blas/backends/mklgpu/mklgpu_batch.cpp | 8 +- .../backends/mklgpu/mklgpu_extensions.cpp | 8 +- src/blas/backends/mklgpu/mklgpu_level1.cpp | 8 +- src/blas/backends/mklgpu/mklgpu_level2.cpp | 8 +- src/blas/backends/mklgpu/mklgpu_level3.cpp | 8 +- src/blas/backends/netlib/netlib_batch.cpp | 4 +- src/blas/backends/netlib/netlib_common.hpp | 4 +- .../backends/netlib/netlib_extensions.cpp | 4 +- src/blas/backends/netlib/netlib_level1.cpp | 4 +- src/blas/backends/netlib/netlib_level2.cpp | 4 +- src/blas/backends/netlib/netlib_level3.cpp | 4 +- src/blas/backends/portblas/portblas_batch.cpp | 4 +- src/blas/backends/portblas/portblas_batch.cxx | 288 +-- .../backends/portblas/portblas_common.hpp | 34 +- .../backends/portblas/portblas_gemm_bias.cxx | 32 +- .../portblas/portblas_level1_double.cpp | 4 +- .../portblas/portblas_level1_float.cpp | 4 +- .../backends/portblas/portblas_level2.cxx | 168 +- .../portblas/portblas_level2_double.cpp | 4 +- .../portblas/portblas_level2_float.cpp | 4 +- .../backends/portblas/portblas_level3.cxx | 100 +- .../portblas/portblas_level3_bfloat16.cpp | 24 +- .../portblas/portblas_level3_double.cpp | 4 +- .../portblas/portblas_level3_float.cpp | 4 +- .../portblas/portblas_level3_half.cpp | 20 +- src/blas/backends/rocblas/rocblas_batch.cpp | 110 +- .../backends/rocblas/rocblas_extensions.cpp | 12 +- src/blas/backends/rocblas/rocblas_handle.hpp | 4 +- src/blas/backends/rocblas/rocblas_helper.hpp | 30 +- src/blas/backends/rocblas/rocblas_level1.cpp | 4 +- src/blas/backends/rocblas/rocblas_level2.cpp | 404 ++-- src/blas/backends/rocblas/rocblas_level3.cpp | 116 +- .../backends/rocblas/rocblas_scope_handle.cpp | 4 +- .../backends/rocblas/rocblas_scope_handle.hpp | 4 +- .../rocblas/rocblas_scope_handle_hipsycl.cpp | 4 +- .../rocblas/rocblas_scope_handle_hipsycl.hpp | 4 +- src/blas/backends/rocblas/rocblas_task.hpp | 4 +- .../backends/rocblas/rocblas_wrappers.cpp | 1960 +++++++-------- src/blas/blas_loader.cpp | 1968 +++++++-------- src/blas/function_table.hpp | 2118 ++++++++--------- .../backends/backend_compute_signature.cxx | 32 +- src/dft/backends/backend_wrappers.cxx | 8 +- src/dft/backends/cufft/backward.cpp | 20 +- src/dft/backends/cufft/commit.cpp | 44 +- src/dft/backends/cufft/descriptor.cpp | 4 +- src/dft/backends/cufft/execute_helper.hpp | 26 +- src/dft/backends/cufft/forward.cpp | 20 +- src/dft/backends/descriptor.cpp | 4 +- src/dft/backends/mklcpu/backward.cpp | 28 +- src/dft/backends/mklcpu/commit.cpp | 22 +- .../backends/mklcpu/commit_derived_impl.hpp | 4 +- src/dft/backends/mklcpu/descriptor.cpp | 4 +- src/dft/backends/mklcpu/forward.cpp | 28 +- src/dft/backends/mklcpu/mklcpu_helpers.hpp | 16 +- src/dft/backends/mklgpu/backward.cpp | 20 +- src/dft/backends/mklgpu/commit.cpp | 26 +- src/dft/backends/mklgpu/descriptor.cpp | 4 +- src/dft/backends/mklgpu/forward.cpp | 20 +- src/dft/backends/mklgpu/mklgpu_helpers.hpp | 16 +- src/dft/backends/portfft/commit.cpp | 36 +- src/dft/backends/portfft/descriptor.cpp | 4 +- src/dft/backends/portfft/portfft_helper.hpp | 6 +- src/dft/backends/rocfft/backward.cpp | 12 +- src/dft/backends/rocfft/commit.cpp | 74 +- src/dft/backends/rocfft/descriptor.cpp | 4 +- src/dft/backends/rocfft/execute_helper.hpp | 20 +- src/dft/backends/rocfft/forward.cpp | 12 +- src/dft/backends/stride_helper.hpp | 10 +- src/dft/descriptor.cxx | 30 +- src/dft/descriptor_config_helper.hpp | 38 +- src/dft/dft_loader.cpp | 10 +- src/dft/function_table.hpp | 32 +- src/include/allocator_helper.hpp | 4 +- src/include/exceptions_helper.hpp | 4 +- src/include/function_table_initializer.hpp | 24 +- .../backends/cusolver/cusolver_batch.cpp | 52 +- .../backends/cusolver/cusolver_handle.hpp | 4 +- .../backends/cusolver/cusolver_helper.hpp | 48 +- .../backends/cusolver/cusolver_lapack.cpp | 382 +-- .../cusolver/cusolver_scope_handle.cpp | 4 +- .../cusolver/cusolver_scope_handle.hpp | 4 +- .../backends/cusolver/cusolver_task.hpp | 4 +- .../backends/cusolver/cusolver_wrappers.cpp | 796 +++---- .../backends/mkl_common/lapack_wrappers.cxx | 624 ++--- .../mkl_common/mkl_lapack_backend.hpp | 510 ++-- src/lapack/backends/mklcpu/mkl_lapack.cpp | 4 +- src/lapack/backends/mklgpu/mkl_lapack.cpp | 4 +- .../backends/rocsolver/rocsolver_batch.cpp | 104 +- .../backends/rocsolver/rocsolver_handle.hpp | 4 +- .../backends/rocsolver/rocsolver_helper.hpp | 48 +- .../backends/rocsolver/rocsolver_lapack.cpp | 326 +-- .../rocsolver/rocsolver_scope_handle.cpp | 4 +- .../rocsolver/rocsolver_scope_handle.hpp | 4 +- .../backends/rocsolver/rocsolver_task.hpp | 4 +- .../backends/rocsolver/rocsolver_wrappers.cpp | 796 +++---- src/lapack/function_table.hpp | 636 ++--- src/lapack/lapack_loader.cpp | 1126 ++++----- src/rng/backends/curand/curand_helper.hpp | 10 +- src/rng/backends/curand/curand_task.hpp | 4 +- src/rng/backends/curand/curand_wrappers.cpp | 6 +- src/rng/backends/curand/mrg32k3a.cpp | 278 +-- src/rng/backends/curand/philox4x32x10.cpp | 278 +-- src/rng/backends/mklcpu/cpu_common.hpp | 4 +- .../backends/mklcpu/mkl_rng_cpu_wrappers.cpp | 6 +- src/rng/backends/mklcpu/mrg32k3a.cpp | 20 +- src/rng/backends/mklcpu/philox4x32x10.cpp | 20 +- .../backends/mklgpu/mkl_internal_rng_gpu.hpp | 4 +- .../backends/mklgpu/mkl_rng_gpu_wrappers.cpp | 6 +- src/rng/backends/mklgpu/mrg32k3a.cpp | 168 +- src/rng/backends/mklgpu/philox4x32x10.cpp | 168 +- src/rng/backends/rocrand/mrg32k3a.cpp | 252 +- src/rng/backends/rocrand/philox4x32x10.cpp | 252 +- src/rng/backends/rocrand/rocrand_helper.hpp | 10 +- src/rng/backends/rocrand/rocrand_task.hpp | 4 +- src/rng/backends/rocrand/rocrand_wrappers.cpp | 6 +- src/rng/function_table.hpp | 8 +- src/rng/rng_loader.cpp | 14 +- src/sparse_blas/backends/backend_wrappers.cxx | 76 +- .../backends/mkl_common/mkl_handles.cxx | 150 +- .../backends/mkl_common/mkl_handles.hpp | 12 +- .../backends/mkl_common/mkl_helper.hpp | 10 +- .../backends/mkl_common/mkl_spmm.cxx | 138 +- .../backends/mkl_common/mkl_spmv.cxx | 142 +- .../backends/mkl_common/mkl_spsv.cxx | 136 +- .../backends/mklcpu/mklcpu_handles.cpp | 4 +- .../backends/mklgpu/mklgpu_handles.cpp | 4 +- src/sparse_blas/enum_data_types.hpp | 4 +- src/sparse_blas/function_table.hpp | 248 +- src/sparse_blas/generic_container.hpp | 30 +- src/sparse_blas/macros.hpp | 2 +- src/sparse_blas/matrix_view_comparison.hpp | 8 +- src/sparse_blas/sparse_blas_loader.cpp | 32 +- .../blas/batch/axpy_batch_stride.cpp | 26 +- .../blas/batch/axpy_batch_stride_usm.cpp | 26 +- .../unit_tests/blas/batch/axpy_batch_usm.cpp | 108 +- .../blas/batch/copy_batch_stride.cpp | 26 +- .../blas/batch/copy_batch_stride_usm.cpp | 26 +- .../unit_tests/blas/batch/copy_batch_usm.cpp | 100 +- .../blas/batch/dgmm_batch_stride.cpp | 80 +- .../blas/batch/dgmm_batch_stride_usm.cpp | 80 +- .../unit_tests/blas/batch/dgmm_batch_usm.cpp | 132 +- .../blas/batch/gemm_batch_stride.cpp | 56 +- .../blas/batch/gemm_batch_stride_usm.cpp | 98 +- .../unit_tests/blas/batch/gemm_batch_usm.cpp | 188 +- .../blas/batch/gemv_batch_stride.cpp | 36 +- .../blas/batch/gemv_batch_stride_usm.cpp | 36 +- .../unit_tests/blas/batch/gemv_batch_usm.cpp | 134 +- .../blas/batch/imatcopy_batch_stride.cpp | 42 +- .../blas/batch/imatcopy_batch_stride_usm.cpp | 58 +- .../blas/batch/imatcopy_batch_usm.cpp | 68 +- .../blas/batch/omatadd_batch_stride.cpp | 50 +- .../blas/batch/omatadd_batch_stride_usm.cpp | 82 +- .../blas/batch/omatcopy_batch_stride.cpp | 40 +- .../blas/batch/omatcopy_batch_stride_usm.cpp | 68 +- .../blas/batch/omatcopy_batch_usm.cpp | 78 +- .../blas/batch/syrk_batch_stride.cpp | 54 +- .../blas/batch/syrk_batch_stride_usm.cpp | 82 +- .../unit_tests/blas/batch/syrk_batch_usm.cpp | 138 +- .../blas/batch/trsm_batch_stride.cpp | 58 +- .../blas/batch/trsm_batch_stride_usm.cpp | 58 +- .../unit_tests/blas/batch/trsm_batch_usm.cpp | 160 +- .../unit_tests/blas/extensions/gemm_bias.cpp | 234 +- .../blas/extensions/gemm_bias_usm.cpp | 234 +- tests/unit_tests/blas/extensions/gemmt.cpp | 238 +- .../unit_tests/blas/extensions/gemmt_usm.cpp | 238 +- tests/unit_tests/blas/extensions/imatcopy.cpp | 42 +- .../blas/extensions/imatcopy_usm.cpp | 42 +- tests/unit_tests/blas/extensions/omatadd.cpp | 50 +- .../blas/extensions/omatadd_usm.cpp | 50 +- tests/unit_tests/blas/extensions/omatcopy.cpp | 46 +- .../unit_tests/blas/extensions/omatcopy2.cpp | 42 +- .../blas/extensions/omatcopy2_usm.cpp | 44 +- .../blas/extensions/omatcopy_usm.cpp | 44 +- .../blas/include/allocator_helper.hpp | 4 +- .../blas/include/onemath_blas_helper.hpp | 28 +- .../blas/include/reference_blas_templates.hpp | 138 +- .../blas/include/reference_blas_wrappers.hpp | 4 +- tests/unit_tests/blas/include/test_common.hpp | 142 +- tests/unit_tests/blas/level1/asum.cpp | 26 +- tests/unit_tests/blas/level1/asum_usm.cpp | 32 +- tests/unit_tests/blas/level1/axpby.cpp | 26 +- tests/unit_tests/blas/level1/axpby_usm.cpp | 26 +- tests/unit_tests/blas/level1/axpy.cpp | 26 +- tests/unit_tests/blas/level1/axpy_usm.cpp | 26 +- tests/unit_tests/blas/level1/copy.cpp | 26 +- tests/unit_tests/blas/level1/copy_usm.cpp | 26 +- tests/unit_tests/blas/level1/dot.cpp | 26 +- tests/unit_tests/blas/level1/dot_usm.cpp | 32 +- tests/unit_tests/blas/level1/dotc.cpp | 26 +- tests/unit_tests/blas/level1/dotc_usm.cpp | 30 +- tests/unit_tests/blas/level1/dotu.cpp | 26 +- tests/unit_tests/blas/level1/dotu_usm.cpp | 30 +- tests/unit_tests/blas/level1/iamax.cpp | 26 +- tests/unit_tests/blas/level1/iamax_usm.cpp | 32 +- tests/unit_tests/blas/level1/iamin.cpp | 26 +- tests/unit_tests/blas/level1/iamin_usm.cpp | 32 +- tests/unit_tests/blas/level1/nrm2.cpp | 26 +- tests/unit_tests/blas/level1/nrm2_usm.cpp | 32 +- tests/unit_tests/blas/level1/rot.cpp | 26 +- tests/unit_tests/blas/level1/rot_usm.cpp | 26 +- tests/unit_tests/blas/level1/rotg.cpp | 26 +- tests/unit_tests/blas/level1/rotg_usm.cpp | 50 +- tests/unit_tests/blas/level1/rotm.cpp | 26 +- tests/unit_tests/blas/level1/rotm_usm.cpp | 26 +- tests/unit_tests/blas/level1/rotmg.cpp | 26 +- tests/unit_tests/blas/level1/rotmg_usm.cpp | 44 +- tests/unit_tests/blas/level1/scal.cpp | 26 +- tests/unit_tests/blas/level1/scal_usm.cpp | 26 +- tests/unit_tests/blas/level1/sdsdot.cpp | 26 +- tests/unit_tests/blas/level1/sdsdot_usm.cpp | 30 +- tests/unit_tests/blas/level1/swap.cpp | 26 +- tests/unit_tests/blas/level1/swap_usm.cpp | 26 +- tests/unit_tests/blas/level2/gbmv.cpp | 88 +- tests/unit_tests/blas/level2/gbmv_usm.cpp | 88 +- tests/unit_tests/blas/level2/gemv.cpp | 88 +- tests/unit_tests/blas/level2/gemv_usm.cpp | 88 +- tests/unit_tests/blas/level2/ger.cpp | 28 +- tests/unit_tests/blas/level2/ger_usm.cpp | 28 +- tests/unit_tests/blas/level2/gerc.cpp | 28 +- tests/unit_tests/blas/level2/gerc_usm.cpp | 28 +- tests/unit_tests/blas/level2/geru.cpp | 28 +- tests/unit_tests/blas/level2/geru_usm.cpp | 28 +- tests/unit_tests/blas/level2/hbmv.cpp | 52 +- tests/unit_tests/blas/level2/hbmv_usm.cpp | 52 +- tests/unit_tests/blas/level2/hemv.cpp | 52 +- tests/unit_tests/blas/level2/hemv_usm.cpp | 52 +- tests/unit_tests/blas/level2/her.cpp | 52 +- tests/unit_tests/blas/level2/her2.cpp | 52 +- tests/unit_tests/blas/level2/her2_usm.cpp | 52 +- tests/unit_tests/blas/level2/her_usm.cpp | 52 +- tests/unit_tests/blas/level2/hpmv.cpp | 52 +- tests/unit_tests/blas/level2/hpmv_usm.cpp | 52 +- tests/unit_tests/blas/level2/hpr.cpp | 52 +- tests/unit_tests/blas/level2/hpr2.cpp | 52 +- tests/unit_tests/blas/level2/hpr2_usm.cpp | 52 +- tests/unit_tests/blas/level2/hpr_usm.cpp | 52 +- tests/unit_tests/blas/level2/sbmv.cpp | 52 +- tests/unit_tests/blas/level2/sbmv_usm.cpp | 52 +- tests/unit_tests/blas/level2/spmv.cpp | 52 +- tests/unit_tests/blas/level2/spmv_usm.cpp | 52 +- tests/unit_tests/blas/level2/spr.cpp | 52 +- tests/unit_tests/blas/level2/spr2.cpp | 52 +- tests/unit_tests/blas/level2/spr2_usm.cpp | 52 +- tests/unit_tests/blas/level2/spr_usm.cpp | 52 +- tests/unit_tests/blas/level2/symv.cpp | 52 +- tests/unit_tests/blas/level2/symv_usm.cpp | 52 +- tests/unit_tests/blas/level2/syr.cpp | 52 +- tests/unit_tests/blas/level2/syr2.cpp | 52 +- tests/unit_tests/blas/level2/syr2_usm.cpp | 52 +- tests/unit_tests/blas/level2/syr_usm.cpp | 52 +- tests/unit_tests/blas/level2/tbmv.cpp | 188 +- tests/unit_tests/blas/level2/tbmv_usm.cpp | 188 +- tests/unit_tests/blas/level2/tbsv.cpp | 188 +- tests/unit_tests/blas/level2/tbsv_usm.cpp | 188 +- tests/unit_tests/blas/level2/tpmv.cpp | 188 +- tests/unit_tests/blas/level2/tpmv_usm.cpp | 188 +- tests/unit_tests/blas/level2/tpsv.cpp | 188 +- tests/unit_tests/blas/level2/tpsv_usm.cpp | 188 +- tests/unit_tests/blas/level2/trmv.cpp | 188 +- tests/unit_tests/blas/level2/trmv_usm.cpp | 188 +- tests/unit_tests/blas/level2/trsv.cpp | 188 +- tests/unit_tests/blas/level2/trsv_usm.cpp | 188 +- tests/unit_tests/blas/level3/gemm.cpp | 194 +- tests/unit_tests/blas/level3/gemm_usm.cpp | 190 +- tests/unit_tests/blas/level3/hemm.cpp | 54 +- tests/unit_tests/blas/level3/hemm_usm.cpp | 54 +- tests/unit_tests/blas/level3/her2k.cpp | 62 +- tests/unit_tests/blas/level3/her2k_usm.cpp | 62 +- tests/unit_tests/blas/level3/herk.cpp | 62 +- tests/unit_tests/blas/level3/herk_usm.cpp | 62 +- tests/unit_tests/blas/level3/symm.cpp | 70 +- tests/unit_tests/blas/level3/symm_usm.cpp | 70 +- tests/unit_tests/blas/level3/syr2k.cpp | 78 +- tests/unit_tests/blas/level3/syr2k_usm.cpp | 78 +- tests/unit_tests/blas/level3/syrk.cpp | 78 +- tests/unit_tests/blas/level3/syrk_usm.cpp | 78 +- tests/unit_tests/blas/level3/trmm.cpp | 234 +- tests/unit_tests/blas/level3/trmm_usm.cpp | 234 +- tests/unit_tests/blas/level3/trsm.cpp | 402 ++-- tests/unit_tests/blas/level3/trsm_usm.cpp | 402 ++-- .../dft/include/compute_inplace.hpp | 76 +- .../dft/include/compute_inplace_real_real.hpp | 48 +- .../dft/include/compute_out_of_place.hpp | 92 +- .../compute_out_of_place_real_real.hpp | 48 +- .../unit_tests/dft/include/compute_tester.hpp | 12 +- .../unit_tests/dft/include/reference_dft.hpp | 2 +- tests/unit_tests/dft/include/test_common.hpp | 20 +- tests/unit_tests/dft/source/compute_tests.cpp | 6 +- .../dft/source/descriptor_tests.cpp | 460 ++-- .../dft/source/workspace_external_tests.cpp | 76 +- tests/unit_tests/include/test_helper.hpp | 46 +- .../lapack/include/lapack_accuracy_checks.hpp | 88 +- .../lapack/include/lapack_common.hpp | 24 +- .../include/lapack_reference_wrappers.hpp | 286 +-- .../lapack/include/lapack_test_controller.hpp | 18 +- tests/unit_tests/lapack/source/gebrd.cpp | 20 +- tests/unit_tests/lapack/source/geqrf.cpp | 20 +- .../lapack/source/geqrf_batch_group.cpp | 20 +- .../lapack/source/geqrf_batch_stride.cpp | 20 +- tests/unit_tests/lapack/source/gerqf.cpp | 20 +- tests/unit_tests/lapack/source/gesvd.cpp | 52 +- tests/unit_tests/lapack/source/getrf.cpp | 20 +- .../lapack/source/getrf_batch_group.cpp | 20 +- .../lapack/source/getrf_batch_stride.cpp | 20 +- tests/unit_tests/lapack/source/getri.cpp | 20 +- .../lapack/source/getri_batch_group.cpp | 20 +- .../lapack/source/getri_batch_stride.cpp | 20 +- tests/unit_tests/lapack/source/getrs.cpp | 28 +- .../lapack/source/getrs_batch_group.cpp | 30 +- .../lapack/source/getrs_batch_stride.cpp | 28 +- tests/unit_tests/lapack/source/heevd.cpp | 20 +- tests/unit_tests/lapack/source/hegvd.cpp | 42 +- tests/unit_tests/lapack/source/hetrd.cpp | 30 +- tests/unit_tests/lapack/source/hetrf.cpp | 34 +- tests/unit_tests/lapack/source/orgbr.cpp | 36 +- tests/unit_tests/lapack/source/orgqr.cpp | 20 +- .../lapack/source/orgqr_batch_group.cpp | 20 +- .../lapack/source/orgqr_batch_stride.cpp | 20 +- tests/unit_tests/lapack/source/orgtr.cpp | 24 +- tests/unit_tests/lapack/source/ormqr.cpp | 34 +- tests/unit_tests/lapack/source/ormrq.cpp | 34 +- tests/unit_tests/lapack/source/ormtr.cpp | 36 +- tests/unit_tests/lapack/source/potrf.cpp | 20 +- .../lapack/source/potrf_batch_group.cpp | 22 +- .../lapack/source/potrf_batch_stride.cpp | 20 +- tests/unit_tests/lapack/source/potri.cpp | 22 +- tests/unit_tests/lapack/source/potrs.cpp | 24 +- .../lapack/source/potrs_batch_group.cpp | 26 +- .../lapack/source/potrs_batch_stride.cpp | 24 +- tests/unit_tests/lapack/source/syevd.cpp | 20 +- tests/unit_tests/lapack/source/sygvd.cpp | 42 +- tests/unit_tests/lapack/source/sytrd.cpp | 30 +- tests/unit_tests/lapack/source/sytrf.cpp | 34 +- tests/unit_tests/lapack/source/trtrs.cpp | 32 +- tests/unit_tests/lapack/source/ungbr.cpp | 36 +- tests/unit_tests/lapack/source/ungqr.cpp | 20 +- .../lapack/source/ungqr_batch_group.cpp | 20 +- .../lapack/source/ungqr_batch_stride.cpp | 20 +- tests/unit_tests/lapack/source/ungtr.cpp | 24 +- tests/unit_tests/lapack/source/unmqr.cpp | 34 +- tests/unit_tests/lapack/source/unmrq.cpp | 34 +- tests/unit_tests/lapack/source/unmtr.cpp | 36 +- .../unit_tests/rng/device/include/moments.hpp | 26 +- .../device/include/rng_device_test_common.hpp | 58 +- .../rng/device/include/skip_ahead_test.hpp | 28 +- .../unit_tests/rng/device/moments/moments.cpp | 1086 ++++----- .../rng/device/service/skip_ahead.cpp | 36 +- .../rng/include/engines_api_tests.hpp | 26 +- .../rng/include/skip_ahead_test.hpp | 22 +- .../rng/include/statistics_check.hpp | 24 +- .../rng/include/statistics_check_test.hpp | 8 +- .../rng/service/engines_api_test.cpp | 8 +- tests/unit_tests/rng/service/skip_ahead.cpp | 8 +- .../rng/statistics_check/bernoulli.cpp | 16 +- .../rng/statistics_check/bernoulli_usm.cpp | 16 +- .../rng/statistics_check/gaussian.cpp | 32 +- .../rng/statistics_check/gaussian_usm.cpp | 32 +- .../rng/statistics_check/lognormal.cpp | 32 +- .../rng/statistics_check/lognormal_usm.cpp | 32 +- .../rng/statistics_check/poisson.cpp | 24 +- .../rng/statistics_check/poisson_usm.cpp | 24 +- .../rng/statistics_check/uniform.cpp | 40 +- .../rng/statistics_check/uniform_usm.cpp | 40 +- .../include/common_sparse_reference.hpp | 50 +- .../sparse_blas/include/test_common.hpp | 64 +- .../sparse_blas/include/test_spmm.hpp | 58 +- .../sparse_blas/include/test_spmv.hpp | 56 +- .../sparse_blas/include/test_spsv.hpp | 34 +- .../sparse_blas/source/sparse_spmm_buffer.cpp | 52 +- .../sparse_blas/source/sparse_spmm_usm.cpp | 52 +- .../sparse_blas/source/sparse_spmv_buffer.cpp | 48 +- .../sparse_blas/source/sparse_spmv_usm.cpp | 48 +- .../sparse_blas/source/sparse_spsv_buffer.cpp | 70 +- .../sparse_blas/source/sparse_spsv_usm.cpp | 70 +- 519 files changed, 25856 insertions(+), 25856 deletions(-) diff --git a/docs/create_new_backend.rst b/docs/create_new_backend.rst index ecc79a631..8b0d10055 100644 --- a/docs/create_new_backend.rst +++ b/docs/create_new_backend.rst @@ -40,20 +40,20 @@ For each new backend library, you should create the following two header files: * Header file with a declaration of entry points to the new third-party library wrappers * Compiler-time dispatching interface (see `oneMath Usage Models <../README.md#supported-usage-models>`_) for new third-party libraries -**Header File Example**: command to generate the header file with a declaration of BLAS entry points in the oneapi::mkl::newlib namespace +**Header File Example**: command to generate the header file with a declaration of BLAS entry points in the oneapi::math::newlib namespace .. code-block:: bash python scripts/generate_backend_api.py include/oneapi/math/blas.hpp \ # Base header file include/oneapi/math/blas/detail/newlib/onemath_blas_newlib.hpp \ # Output header file - oneapi::mkl::newlib # Wrappers namespace + oneapi::math::newlib # Wrappers namespace Code snippet of the generated header file ``include/oneapi/math/blas/detail/newlib/onemath_blas_newlib.hpp`` .. code-block:: cpp namespace oneapi { - namespace mkl { + namespace math { namespace newlib { void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, @@ -70,14 +70,14 @@ Code snippet of the generated header file ``include/oneapi/math/blas/detail/newl include/oneapi/math/blas/detail/newlib/onemath_blas_newlib.hpp \ # Header file with declaration of entry points to wrappers newlib \ # Library name newdevice \ # Backend name - oneapi::mkl::newlib # Wrappers namespace + oneapi::math::newlib # Wrappers namespace Code snippet of the generated header file ``include/oneapi/math/blas/detail/newlib/blas_ct.hpp`` .. code-block:: cpp namespace oneapi { - namespace mkl { + namespace math { namespace blas { template <> @@ -85,7 +85,7 @@ Code snippet of the generated header file ``include/oneapi/math/blas/detail/newl sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { asum_precondition(queue, n, x, incx, result); - oneapi::mkl::newlib::asum(queue, n, x, incx, result); + oneapi::math::newlib::asum(queue, n, x, incx, result); asum_postcondition(queue, n, x, incx, result); } @@ -174,8 +174,8 @@ To integrate the new third-party library to a oneMath header-based part, followi .. code-block:: diff - inline oneapi::mkl::device get_device_id(sycl::queue &queue) { - oneapi::mkl::device device_id; + inline oneapi::math::device get_device_id(sycl::queue &queue) { + oneapi::math::device device_id; + if (queue.is_host()) + device_id=device::newdevice; @@ -283,7 +283,7 @@ The following code snippet is updated for ``src/blas/backends/newlib/newlib_wrap + #include "newlib.h" namespace oneapi { - namespace mkl { + namespace math { namespace newlib { void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, @@ -466,14 +466,14 @@ Update the following files to enable the new third-party library for unit tests: #ifdef ONEMATH_ENABLE_MKLGPU_BACKEND #define TEST_RUN_INTELGPU(q, func, args) \ - func args + func args #else #define TEST_RUN_INTELGPU(q, func, args) #endif + + #ifdef ONEMATH_ENABLE_NEWLIB_BACKEND + #define TEST_RUN_NEWDEVICE(q, func, args) \ - + func args + + func args + #else + #define TEST_RUN_NEWDEVICE(q, func, args) + #endif diff --git a/docs/domains/sparse_linear_algebra.rst b/docs/domains/sparse_linear_algebra.rst index ba2699c76..568fe229e 100644 --- a/docs/domains/sparse_linear_algebra.rst +++ b/docs/domains/sparse_linear_algebra.rst @@ -17,24 +17,24 @@ Currently known limitations: - All operations' algorithms except ``no_optimize_alg`` map to the default algorithm. - The required external workspace size is always 0 bytes. -- ``oneapi::mkl::sparse::set_csr_data`` and - ``oneapi::mkl::sparse::set_coo_data`` functions cannot be used on a handle +- ``oneapi::math::sparse::set_csr_data`` and + ``oneapi::math::sparse::set_coo_data`` functions cannot be used on a handle that has already been used for an operation or its optimize function. Doing so - will throw an ``oneapi::mkl::unimplemented`` exception. -- Using ``spsv`` with the ``oneapi::mkl::sparse::spsv_alg::no_optimize_alg`` and + will throw an ``oneapi::math::unimplemented`` exception. +- Using ``spsv`` with the ``oneapi::math::sparse::spsv_alg::no_optimize_alg`` and a sparse matrix that does not have the - ``oneapi::mkl::sparse::matrix_property::sorted`` property will throw an - ``oneapi::mkl::unimplemented`` exception. + ``oneapi::math::sparse::matrix_property::sorted`` property will throw an + ``oneapi::math::unimplemented`` exception. - Using ``spmm`` on Intel GPU with a sparse matrix that is - ``oneapi::mkl::transpose::conjtrans`` and has the - ``oneapi::mkl::sparse::matrix_property::symmetric`` property will throw an - ``oneapi::mkl::unimplemented`` exception. + ``oneapi::math::transpose::conjtrans`` and has the + ``oneapi::math::sparse::matrix_property::symmetric`` property will throw an + ``oneapi::math::unimplemented`` exception. - Using ``spmv`` with a sparse matrix that is - ``oneapi::mkl::transpose::conjtrans`` with a ``type_view`` + ``oneapi::math::transpose::conjtrans`` with a ``type_view`` ``matrix_descr::symmetric`` or ``matrix_descr::hermitian`` will throw an - ``oneapi::mkl::unimplemented`` exception. + ``oneapi::math::unimplemented`` exception. - Using ``spsv`` on Intel GPU with a sparse matrix that is - ``oneapi::mkl::transpose::conjtrans`` and will throw an - ``oneapi::mkl::unimplemented`` exception. + ``oneapi::math::transpose::conjtrans`` and will throw an + ``oneapi::math::unimplemented`` exception. - Scalar parameters ``alpha`` and ``beta`` should be host pointers to prevent synchronizations and copies to the host. diff --git a/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp b/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp index d16821288..641ee5ef4 100644 --- a/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp +++ b/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp @@ -20,13 +20,13 @@ /* * * Content: -* This example demonstrates use of DPCPP API oneapi::mkl::blas::gemm +* This example demonstrates use of DPCPP API oneapi::math::blas::gemm * using unified shared memory to perform General Matrix-Matrix * Multiplication on a INTEL CPU SYCL device and an NVIDIA GPU SYCL device * * C = alpha * op(A) * op(B) + beta * C * -* where op() is defined by one of oneapi::mkl::transpose::{nontrans,trans,conjtrans} +* where op() is defined by one of oneapi::math::transpose::{nontrans,trans,conjtrans} * * * This example demonstrates only single precision (float) data type for @@ -67,8 +67,8 @@ void run_gemm_example(const sycl::device &cpu_dev, const sycl::device &gpu_dev) // // C = alpha * op(A) * op(B) + beta * C // - oneapi::mkl::transpose transA = oneapi::mkl::transpose::trans; - oneapi::mkl::transpose transB = oneapi::mkl::transpose::nontrans; + oneapi::math::transpose transA = oneapi::math::transpose::trans; + oneapi::math::transpose transB = oneapi::math::transpose::nontrans; // matrix data sizes int m = 45; @@ -79,8 +79,8 @@ void run_gemm_example(const sycl::device &cpu_dev, const sycl::device &gpu_dev) int ldA = 103; int ldB = 105; int ldC = 106; - int sizea = (transA == oneapi::mkl::transpose::nontrans) ? ldA * k : ldA * m; - int sizeb = (transB == oneapi::mkl::transpose::nontrans) ? ldB * n : ldB * k; + int sizea = (transA == oneapi::math::transpose::nontrans) ? ldA * k : ldA * m; + int sizeb = (transB == oneapi::math::transpose::nontrans) ? ldB * n : ldB * k; int sizec = ldC * n; // set scalar fp values @@ -131,7 +131,7 @@ void run_gemm_example(const sycl::device &cpu_dev, const sycl::device &gpu_dev) rand_matrix(A, transA, m, k, ldA); rand_matrix(B, transB, k, n, ldB); - rand_matrix(C, oneapi::mkl::transpose::nontrans, m, n, ldC); + rand_matrix(C, oneapi::math::transpose::nontrans, m, n, ldC); // // Preparation on CPU @@ -172,12 +172,12 @@ void run_gemm_example(const sycl::device &cpu_dev, const sycl::device &gpu_dev) // // Execute Gemm on CPU and GPU device // - // add oneapi::mkl::blas::gemm to execution queue - cpu_gemm_done = oneapi::mkl::blas::column_major::gemm( - oneapi::mkl::backend_selector{ cpu_queue }, transA, transB, m, + // add oneapi::math::blas::gemm to execution queue + cpu_gemm_done = oneapi::math::blas::column_major::gemm( + oneapi::math::backend_selector{ cpu_queue }, transA, transB, m, n, k, alpha, cpu_A, ldA, cpu_B, ldB, beta, cpu_C, ldC); - gpu_gemm_done = oneapi::mkl::blas::column_major::gemm( - oneapi::mkl::backend_selector{ gpu_queue }, transA, transB, m, + gpu_gemm_done = oneapi::math::blas::column_major::gemm( + oneapi::math::backend_selector{ gpu_queue }, transA, transB, m, n, k, alpha, gpu_A, ldA, gpu_B, ldB, beta, gpu_C, ldC); // Wait until calculations are done @@ -196,13 +196,13 @@ void run_gemm_example(const sycl::device &cpu_dev, const sycl::device &gpu_dev) // print results std::cout << "\n\t\tGEMM parameters:" << std::endl; std::cout << "\t\t\ttransA = " - << (transA == oneapi::mkl::transpose::nontrans + << (transA == oneapi::math::transpose::nontrans ? "nontrans" - : (transA == oneapi::mkl::transpose::trans ? "trans" : "conjtrans")) + : (transA == oneapi::math::transpose::trans ? "trans" : "conjtrans")) << ", transB = " - << (transB == oneapi::mkl::transpose::nontrans + << (transB == oneapi::math::transpose::nontrans ? "nontrans" - : (transB == oneapi::mkl::transpose::trans ? "trans" : "conjtrans")) + : (transB == oneapi::math::transpose::trans ? "trans" : "conjtrans")) << std::endl; std::cout << "\t\t\tm = " << m << ", n = " << n << ", k = " << k << std::endl; std::cout << "\t\t\tlda = " << ldA << ", ldB = " << ldB << ", ldC = " << ldC << std::endl; diff --git a/examples/blas/run_time_dispatching/level3/gemm_usm.cpp b/examples/blas/run_time_dispatching/level3/gemm_usm.cpp index e213e384b..3df710101 100644 --- a/examples/blas/run_time_dispatching/level3/gemm_usm.cpp +++ b/examples/blas/run_time_dispatching/level3/gemm_usm.cpp @@ -20,14 +20,14 @@ /* * * Content: -* This example demonstrates use of DPCPP API oneapi::mkl::blas::gemm +* This example demonstrates use of DPCPP API oneapi::math::blas::gemm * using unified shared memory to perform General Matrix-Matrix * Multiplication on a SYCL device (HOST, CPU, GPU) that is selected * during runtime. * * C = alpha * op(A) * op(B) + beta * C * -* where op() is defined by one of oneapi::mkl::transpose::{nontrans,trans,conjtrans} +* where op() is defined by one of oneapi::math::transpose::{nontrans,trans,conjtrans} * * * This example demonstrates only single precision (float) data type for @@ -67,8 +67,8 @@ void run_gemm_example(const sycl::device& dev) { // C = alpha * op(A) * op(B) + beta * C // - oneapi::mkl::transpose transA = oneapi::mkl::transpose::trans; - oneapi::mkl::transpose transB = oneapi::mkl::transpose::nontrans; + oneapi::math::transpose transA = oneapi::math::transpose::trans; + oneapi::math::transpose transB = oneapi::math::transpose::nontrans; // matrix data sizes int m = 45; @@ -79,8 +79,8 @@ void run_gemm_example(const sycl::device& dev) { int ldA = 103; int ldB = 105; int ldC = 106; - int sizea = (transA == oneapi::mkl::transpose::nontrans) ? ldA * k : ldA * m; - int sizeb = (transB == oneapi::mkl::transpose::nontrans) ? ldB * n : ldB * k; + int sizea = (transA == oneapi::math::transpose::nontrans) ? ldA * k : ldA * m; + int sizeb = (transB == oneapi::math::transpose::nontrans) ? ldB * n : ldB * k; int sizec = ldC * n; // set scalar fp values @@ -116,7 +116,7 @@ void run_gemm_example(const sycl::device& dev) { rand_matrix(A, transA, m, k, ldA); rand_matrix(B, transB, k, n, ldB); - rand_matrix(C, oneapi::mkl::transpose::nontrans, m, n, ldC); + rand_matrix(C, oneapi::math::transpose::nontrans, m, n, ldC); // allocate memory on device auto dev_A = sycl::malloc_device(sizea * sizeof(float), main_queue); @@ -134,8 +134,8 @@ void run_gemm_example(const sycl::device& dev) { // // Execute Gemm // - // add oneapi::mkl::blas::gemm to execution queue - gemm_done = oneapi::mkl::blas::column_major::gemm(main_queue, transA, transB, m, n, k, alpha, + // add oneapi::math::blas::gemm to execution queue + gemm_done = oneapi::math::blas::column_major::gemm(main_queue, transA, transB, m, n, k, alpha, dev_A, ldA, dev_B, ldB, beta, dev_C, ldC); // Wait until calculations are done @@ -149,13 +149,13 @@ void run_gemm_example(const sycl::device& dev) { std::cout << "\n\t\tGEMM parameters:" << std::endl; std::cout << "\t\t\ttransA = " - << (transA == oneapi::mkl::transpose::nontrans + << (transA == oneapi::math::transpose::nontrans ? "nontrans" - : (transA == oneapi::mkl::transpose::trans ? "trans" : "conjtrans")) + : (transA == oneapi::math::transpose::trans ? "trans" : "conjtrans")) << ", transB = " - << (transB == oneapi::mkl::transpose::nontrans + << (transB == oneapi::math::transpose::nontrans ? "nontrans" - : (transB == oneapi::mkl::transpose::trans ? "trans" : "conjtrans")) + : (transB == oneapi::math::transpose::trans ? "trans" : "conjtrans")) << std::endl; std::cout << "\t\t\tm = " << m << ", n = " << n << ", k = " << k << std::endl; std::cout << "\t\t\tlda = " << ldA << ", ldB = " << ldB << ", ldC = " << ldC << std::endl; diff --git a/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp b/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp index ce030e6c8..fdad04de2 100644 --- a/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp +++ b/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp @@ -82,28 +82,28 @@ void run_example(const sycl::device& cpu_device, const sycl::device& gpu_device) // enabling // 1. create descriptors - oneapi::mkl::dft::descriptor + oneapi::math::dft::descriptor desc(static_cast(N)); // 2. variadic set_value - desc.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::NOT_INPLACE); - desc.set_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, + desc.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::NOT_INPLACE); + desc.set_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, static_cast(1)); // 3a. commit_descriptor (compile_time MKLCPU) - desc.commit(oneapi::mkl::backend_selector{ cpu_queue }); + desc.commit(oneapi::math::backend_selector{ cpu_queue }); // 4a. compute_forward / compute_backward (MKLCPU) - oneapi::mkl::dft::compute_forward, std::complex>( + oneapi::math::dft::compute_forward, std::complex>( desc, cpu_input_data, cpu_output_data); // 3b. commit_descriptor (compile_time cuFFT) - desc.commit(oneapi::mkl::backend_selector{ gpu_queue }); + desc.commit(oneapi::math::backend_selector{ gpu_queue }); // 4b. compute_forward / compute_backward (cuFFT) - oneapi::mkl::dft::compute_forward, std::complex>( + oneapi::math::dft::compute_forward, std::complex>( desc, gpu_input_data, gpu_output_data); cpu_queue.wait_and_throw(); diff --git a/examples/dft/run_time_dispatching/real_fwd_usm.cpp b/examples/dft/run_time_dispatching/real_fwd_usm.cpp index f14eb077c..5f1f53acf 100644 --- a/examples/dft/run_time_dispatching/real_fwd_usm.cpp +++ b/examples/dft/run_time_dispatching/real_fwd_usm.cpp @@ -53,21 +53,21 @@ void run_example(const sycl::device& dev) { auto x_usm = sycl::malloc_shared(N * 2, sycl_queue); // 1. create descriptors - oneapi::mkl::dft::descriptor + oneapi::math::dft::descriptor desc(static_cast(N)); // 2. variadic set_value - desc.set_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, + desc.set_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, static_cast(1)); - desc.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::INPLACE); + desc.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::INPLACE); // 3. commit_descriptor (runtime dispatch) desc.commit(sycl_queue); // 4. compute_forward / compute_backward (runtime dispatch) - auto compute_event = oneapi::mkl::dft::compute_forward(desc, x_usm); + auto compute_event = oneapi::math::dft::compute_forward(desc, x_usm); // Do something with transformed data. compute_event.wait(); @@ -122,7 +122,7 @@ int main(int /*argc*/, char** /*argv*/) { run_example(my_dev); std::cout << "DFT example ran OK" << std::endl; } - catch (oneapi::mkl::unimplemented const& e) { + catch (oneapi::math::unimplemented const& e) { std::cerr << "Unsupported Configuration:" << std::endl; std::cerr << "\t" << e.what() << std::endl; return 0; diff --git a/examples/include/example_helper.hpp b/examples/include/example_helper.hpp index 4a89e6fae..9ed99940c 100644 --- a/examples/include/example_helper.hpp +++ b/examples/include/example_helper.hpp @@ -88,10 +88,10 @@ fp rand_scalar() { } template -void rand_matrix(vec &M, oneapi::mkl::transpose trans, int m, int n, int ld) { +void rand_matrix(vec &M, oneapi::math::transpose trans, int m, int n, int ld) { using fp = typename vec::value_type; - if (trans == oneapi::mkl::transpose::nontrans) { + if (trans == oneapi::math::transpose::nontrans) { for (int j = 0; j < n; j++) for (int i = 0; i < m; i++) M.at(i + j * ld) = rand_scalar(); diff --git a/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp b/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp index e824789c0..f8ee084de 100644 --- a/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp +++ b/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp @@ -20,8 +20,8 @@ /* * * Content: -* This example demonstrates use of oneapi::mkl::lapack::getrf and -* oneapi::mkl::lapack::getrs to perform LU factorization and compute +* This example demonstrates use of oneapi::math::lapack::getrf and +* oneapi::math::lapack::getrs to perform LU factorization and compute * the solution on both an Intel CPU device and NVIDIA GPU device. * * This example demonstrates only single precision (float) data type @@ -67,7 +67,7 @@ void run_getrs_example(const sycl::device& cpu_device, const sycl::device& gpu_d std::int64_t A_size = n * lda; std::int64_t B_size = nrhs * ldb; std::int64_t ipiv_size = n; - oneapi::mkl::transpose trans = oneapi::mkl::transpose::nontrans; + oneapi::math::transpose trans = oneapi::math::transpose::nontrans; // Catch asynchronous exceptions for CPU and GPU auto cpu_error_handler = [&](sycl::exception_list exceptions) { @@ -75,7 +75,7 @@ void run_getrs_example(const sycl::device& cpu_device, const sycl::device& gpu_d try { std::rethrow_exception(e); } - catch (oneapi::mkl::lapack::exception const& e) { + catch (oneapi::math::lapack::exception const& e) { // Handle LAPACK related exceptions that happened during asynchronous call std::cerr << "Caught asynchronous LAPACK exception on CPU device during GETRF or GETRS:" @@ -98,7 +98,7 @@ void run_getrs_example(const sycl::device& cpu_device, const sycl::device& gpu_d try { std::rethrow_exception(e); } - catch (oneapi::mkl::lapack::exception const& e) { + catch (oneapi::math::lapack::exception const& e) { // Handle LAPACK related exceptions that happened during asynchronous call std::cerr << "Caught asynchronous LAPACK exception on GPU device during GETRF or GETRS:" @@ -145,10 +145,10 @@ void run_getrs_example(const sycl::device& cpu_device, const sycl::device& gpu_d std::int64_t* cpu_ipiv = sycl::malloc_device(ipiv_size * sizeof(std::int64_t), cpu_queue); - std::int64_t cpu_getrf_scratchpad_size = oneapi::mkl::lapack::getrf_scratchpad_size( - oneapi::mkl::backend_selector{ cpu_queue }, m, n, lda); - std::int64_t cpu_getrs_scratchpad_size = oneapi::mkl::lapack::getrs_scratchpad_size( - oneapi::mkl::backend_selector{ cpu_queue }, trans, n, nrhs, + std::int64_t cpu_getrf_scratchpad_size = oneapi::math::lapack::getrf_scratchpad_size( + oneapi::math::backend_selector{ cpu_queue }, m, n, lda); + std::int64_t cpu_getrs_scratchpad_size = oneapi::math::lapack::getrs_scratchpad_size( + oneapi::math::backend_selector{ cpu_queue }, trans, n, nrhs, lda, ldb); float* cpu_getrf_scratchpad = sycl::malloc_device( cpu_getrf_scratchpad_size * sizeof(float), cpu_device, cpu_context); @@ -175,10 +175,10 @@ void run_getrs_example(const sycl::device& cpu_device, const sycl::device& gpu_d std::int64_t* gpu_ipiv = sycl::malloc_device(ipiv_size * sizeof(std::int64_t), gpu_queue); - std::int64_t gpu_getrf_scratchpad_size = oneapi::mkl::lapack::getrf_scratchpad_size( - oneapi::mkl::backend_selector{ gpu_queue }, m, n, lda); - std::int64_t gpu_getrs_scratchpad_size = oneapi::mkl::lapack::getrs_scratchpad_size( - oneapi::mkl::backend_selector{ gpu_queue }, trans, n, nrhs, + std::int64_t gpu_getrf_scratchpad_size = oneapi::math::lapack::getrf_scratchpad_size( + oneapi::math::backend_selector{ gpu_queue }, m, n, lda); + std::int64_t gpu_getrs_scratchpad_size = oneapi::math::lapack::getrs_scratchpad_size( + oneapi::math::backend_selector{ gpu_queue }, trans, n, nrhs, lda, ldb); float* gpu_getrf_scratchpad = sycl::malloc_device( gpu_getrf_scratchpad_size * sizeof(float), gpu_device, gpu_context); @@ -196,18 +196,18 @@ void run_getrs_example(const sycl::device& cpu_device, const sycl::device& gpu_d // Execute on CPU and GPU devices // - cpu_getrf_done = oneapi::mkl::lapack::getrf( - oneapi::mkl::backend_selector{ cpu_queue }, m, n, cpu_A, lda, + cpu_getrf_done = oneapi::math::lapack::getrf( + oneapi::math::backend_selector{ cpu_queue }, m, n, cpu_A, lda, cpu_ipiv, cpu_getrf_scratchpad, cpu_getrf_scratchpad_size); - cpu_getrs_done = oneapi::mkl::lapack::getrs( - oneapi::mkl::backend_selector{ cpu_queue }, trans, n, nrhs, + cpu_getrs_done = oneapi::math::lapack::getrs( + oneapi::math::backend_selector{ cpu_queue }, trans, n, nrhs, cpu_A, lda, cpu_ipiv, cpu_B, ldb, cpu_getrs_scratchpad, cpu_getrs_scratchpad_size, { cpu_getrf_done }); - gpu_getrf_done = oneapi::mkl::lapack::getrf( - oneapi::mkl::backend_selector{ gpu_queue }, m, n, gpu_A, + gpu_getrf_done = oneapi::math::lapack::getrf( + oneapi::math::backend_selector{ gpu_queue }, m, n, gpu_A, lda, gpu_ipiv, gpu_getrf_scratchpad, gpu_getrf_scratchpad_size); - gpu_getrs_done = oneapi::mkl::lapack::getrs( - oneapi::mkl::backend_selector{ gpu_queue }, trans, n, nrhs, + gpu_getrs_done = oneapi::math::lapack::getrs( + oneapi::math::backend_selector{ gpu_queue }, trans, n, nrhs, gpu_A, lda, gpu_ipiv, gpu_B, ldb, gpu_getrs_scratchpad, gpu_getrs_scratchpad_size, { gpu_getrf_done }); @@ -227,9 +227,9 @@ void run_getrs_example(const sycl::device& cpu_device, const sycl::device& gpu_d // Print results std::cout << "\n\t\tGETRF and GETRS parameters:" << std::endl; std::cout << "\t\t\ttrans = " - << (trans == oneapi::mkl::transpose::nontrans + << (trans == oneapi::math::transpose::nontrans ? "nontrans" - : (trans == oneapi::mkl::transpose::trans ? "trans" : "conjtrans")) + : (trans == oneapi::math::transpose::trans ? "trans" : "conjtrans")) << std::endl; std::cout << "\t\t\tm = " << m << ", n = " << n << ", nrhs = " << nrhs << std::endl; std::cout << "\t\t\tlda = " << lda << ", ldb = " << ldb << std::endl; @@ -308,7 +308,7 @@ int main(int argc, char** argv) { run_getrs_example(cpu_dev, gpu_dev); std::cout << "LAPACK GETRS USM example ran OK on MKLCPU and CUSOLVER" << std::endl; } - catch (oneapi::mkl::lapack::exception const& e) { + catch (oneapi::math::lapack::exception const& e) { // Handle LAPACK related exceptions that happened during synchronous call std::cerr << "Caught synchronous LAPACK exception:" << std::endl; std::cerr << "\t" << e.what() << std::endl; diff --git a/examples/lapack/run_time_dispatching/getrs_usm.cpp b/examples/lapack/run_time_dispatching/getrs_usm.cpp index ded21cc22..68a7836e0 100644 --- a/examples/lapack/run_time_dispatching/getrs_usm.cpp +++ b/examples/lapack/run_time_dispatching/getrs_usm.cpp @@ -20,8 +20,8 @@ /* * * Content: -* This example demonstrates use of oneapi::mkl::lapack::getrf and -* oneapi::mkl::lapack::getrs to perform LU factorization and compute +* This example demonstrates use of oneapi::math::lapack::getrf and +* oneapi::math::lapack::getrs to perform LU factorization and compute * the solution on a SYCL device (HOST, CPU, GPU) that is selected * during runtime. * @@ -68,7 +68,7 @@ void run_getrs_example(const sycl::device& device) { std::int64_t A_size = n * lda; std::int64_t B_size = nrhs * ldb; std::int64_t ipiv_size = n; - oneapi::mkl::transpose trans = oneapi::mkl::transpose::nontrans; + oneapi::math::transpose trans = oneapi::math::transpose::nontrans; // Asynchronous error handler auto error_handler = [&](sycl::exception_list exceptions) { @@ -76,7 +76,7 @@ void run_getrs_example(const sycl::device& device) { try { std::rethrow_exception(e); } - catch (oneapi::mkl::lapack::exception const& e) { + catch (oneapi::math::lapack::exception const& e) { // Handle LAPACK related exceptions that happened during asynchronous call std::cerr << "Caught asynchronous LAPACK exception during GETRF or GETRS:" << std::endl; @@ -114,9 +114,9 @@ void run_getrs_example(const sycl::device& device) { sycl::malloc_device(ipiv_size * sizeof(std::int64_t), queue); std::int64_t getrf_scratchpad_size = - oneapi::mkl::lapack::getrf_scratchpad_size(queue, m, n, lda); + oneapi::math::lapack::getrf_scratchpad_size(queue, m, n, lda); std::int64_t getrs_scratchpad_size = - oneapi::mkl::lapack::getrs_scratchpad_size(queue, trans, n, nrhs, lda, ldb); + oneapi::math::lapack::getrs_scratchpad_size(queue, trans, n, nrhs, lda, ldb); float* getrf_scratchpad = sycl::malloc_shared(getrf_scratchpad_size * sizeof(float), device, context); float* getrs_scratchpad = @@ -145,10 +145,10 @@ void run_getrs_example(const sycl::device& device) { queue.memcpy(dev_B, B.data(), B_size * sizeof(float)).wait(); // Execute on device - getrf_done = oneapi::mkl::lapack::getrf(queue, m, n, dev_A, lda, dev_ipiv, getrf_scratchpad, + getrf_done = oneapi::math::lapack::getrf(queue, m, n, dev_A, lda, dev_ipiv, getrf_scratchpad, getrf_scratchpad_size); getrs_done = - oneapi::mkl::lapack::getrs(queue, trans, n, nrhs, dev_A, lda, dev_ipiv, dev_B, ldb, + oneapi::math::lapack::getrs(queue, trans, n, nrhs, dev_A, lda, dev_ipiv, dev_B, ldb, getrs_scratchpad, getrs_scratchpad_size, { getrf_done }); // Wait until calculations are done @@ -160,9 +160,9 @@ void run_getrs_example(const sycl::device& device) { // Print results std::cout << "\n\t\tGETRF and GETRS parameters:" << std::endl; std::cout << "\t\t\ttrans = " - << (trans == oneapi::mkl::transpose::nontrans + << (trans == oneapi::math::transpose::nontrans ? "nontrans" - : (trans == oneapi::mkl::transpose::trans ? "trans" : "conjtrans")) + : (trans == oneapi::math::transpose::trans ? "trans" : "conjtrans")) << std::endl; std::cout << "\t\t\tm = " << m << ", n = " << n << ", nrhs = " << nrhs << std::endl; std::cout << "\t\t\tlda = " << lda << ", ldb = " << ldb << std::endl; @@ -235,7 +235,7 @@ int main(int argc, char** argv) { run_getrs_example(dev); std::cout << "LAPACK GETRS USM example ran OK" << std::endl; } - catch (oneapi::mkl::lapack::exception const& e) { + catch (oneapi::math::lapack::exception const& e) { // Handle LAPACK related exceptions that happened during synchronous call std::cerr << "Caught synchronous LAPACK exception:" << std::endl; std::cerr << "\t" << e.what() << std::endl; diff --git a/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp b/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp index 305e10da0..2d15397dd 100644 --- a/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp +++ b/examples/rng/compile_time_dispatching/uniform_usm_mklcpu_curand.cpp @@ -20,8 +20,8 @@ /* * * Content: -* This example demonstrates use of DPC++ API oneapi::mkl::rng::uniform distribution -* with oneapi::mkl::rng::philox4x32x10 random number generator to produce +* This example demonstrates use of DPC++ API oneapi::math::rng::uniform distribution +* with oneapi::math::rng::philox4x32x10 random number generator to produce * random numbers on a INTEL CPU SYCL device and an NVIDIA GPU SYCL device * with Unified Shared Memory(USM) API. * @@ -98,12 +98,12 @@ void run_uniform_example(const sycl::device& cpu_dev, const sycl::device& gpu_de // preparation on CPU device and GPU device sycl::queue cpu_queue(cpu_dev, cpu_exception_handler); sycl::queue gpu_queue(gpu_dev, gpu_exception_handler); - oneapi::mkl::rng::default_engine cpu_engine( - oneapi::mkl::backend_selector{ cpu_queue }, seed); - oneapi::mkl::rng::default_engine gpu_engine( - oneapi::mkl::backend_selector{ gpu_queue }, seed); + oneapi::math::rng::default_engine cpu_engine( + oneapi::math::backend_selector{ cpu_queue }, seed); + oneapi::math::rng::default_engine gpu_engine( + oneapi::math::backend_selector{ gpu_queue }, seed); - oneapi::mkl::rng::uniform distribution(a, b); + oneapi::math::rng::uniform distribution(a, b); // // Data preparation on host: prepare array for random numbers @@ -127,8 +127,8 @@ void run_uniform_example(const sycl::device& cpu_dev, const sycl::device& gpu_de // sycl::event event_out_cpu; sycl::event event_out_gpu; - event_out_cpu = oneapi::mkl::rng::generate(distribution, cpu_engine, n, dev_cpu); - event_out_gpu = oneapi::mkl::rng::generate(distribution, gpu_engine, n, dev_gpu); + event_out_cpu = oneapi::math::rng::generate(distribution, cpu_engine, n, dev_cpu); + event_out_gpu = oneapi::math::rng::generate(distribution, gpu_engine, n, dev_gpu); event_out_cpu.wait_and_throw(); event_out_gpu.wait_and_throw(); diff --git a/examples/rng/device/uniform.cpp b/examples/rng/device/uniform.cpp index 156afc1ff..d9ed83dad 100644 --- a/examples/rng/device/uniform.cpp +++ b/examples/rng/device/uniform.cpp @@ -20,7 +20,7 @@ /* * * Content: -* This example demonstrates usage of oneapi::mkl::rng::device::mcg59 +* This example demonstrates usage of oneapi::math::rng::device::mcg59 * random number generator to produce random * numbers using unifrom distribution on a SYCL device (CPU, GPU). * @@ -74,10 +74,10 @@ int run_example(sycl::queue& queue) { sycl::accessor r_acc(r_buf, cgh, sycl::write_only); cgh.parallel_for(sycl::range<1>(n / VecSize), [=](sycl::item<1> item) { size_t item_id = item.get_id(0); - oneapi::mkl::rng::device::mcg59 engine(seed, item_id * VecSize); - oneapi::mkl::rng::device::uniform distr; + oneapi::math::rng::device::mcg59 engine(seed, item_id * VecSize); + oneapi::math::rng::device::uniform distr; - auto res = oneapi::mkl::rng::device::generate(distr, engine); + auto res = oneapi::math::rng::device::generate(distr, engine); if constexpr (VecSize == 1) { r_acc[item_id] = res; } @@ -104,13 +104,13 @@ int run_example(sycl::queue& queue) { } // buffer life-time ends // compare results with host-side generation - oneapi::mkl::rng::device::mcg59<1> engine(seed); - oneapi::mkl::rng::device::uniform distr; + oneapi::math::rng::device::mcg59<1> engine(seed); + oneapi::math::rng::device::uniform distr; int err = 0; Type res_host; for (int i = 0; i < n; i++) { - res_host = oneapi::mkl::rng::device::generate(distr, engine); + res_host = oneapi::math::rng::device::generate(distr, engine); if (res_host != r_dev[i]) { std::cout << "error in " << i << " element " << res_host << " " << r_dev[i] << std::endl; diff --git a/examples/rng/run_time_dispatching/uniform_usm.cpp b/examples/rng/run_time_dispatching/uniform_usm.cpp index d179273ee..34a3d4c99 100644 --- a/examples/rng/run_time_dispatching/uniform_usm.cpp +++ b/examples/rng/run_time_dispatching/uniform_usm.cpp @@ -20,8 +20,8 @@ /* * * Content: -* This example demonstrates use of DPC++ API oneapi::mkl::rng::uniform distribution -* with oneapi::mkl::rng::philox4x32x10 random number generator to produce +* This example demonstrates use of DPC++ API oneapi::math::rng::uniform distribution +* with oneapi::math::rng::philox4x32x10 random number generator to produce * random numbers on a SYCL device (HOST, CPU, GPU) that is selected * during runtime with Unified Shared Memory(USM) API. * @@ -83,8 +83,8 @@ void run_uniform_example(const sycl::device& dev) { float a(0.0); float b(10.0); - oneapi::mkl::rng::default_engine engine(queue, seed); - oneapi::mkl::rng::uniform distribution(a, b); + oneapi::math::rng::default_engine engine(queue, seed); + oneapi::math::rng::uniform distribution(a, b); // // Data preparation on host: prepare array for random numbers @@ -101,7 +101,7 @@ void run_uniform_example(const sycl::device& dev) { // Perform generation on device // sycl::event event_out; - event_out = oneapi::mkl::rng::generate(distribution, engine, n, dev_r); + event_out = oneapi::math::rng::generate(distribution, engine, n, dev_r); event_out.wait_and_throw(); // diff --git a/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp b/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp index 0f8956ad5..a023c0a69 100644 --- a/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp +++ b/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp @@ -20,7 +20,7 @@ /* * * Content: -* This example demonstrates use of DPCPP API oneapi::mkl::sparse::spmv +* This example demonstrates use of DPCPP API oneapi::math::sparse::spmv * using unified shared memory to perform general sparse matrix-vector * multiplication on a INTEL CPU SYCL device. * @@ -28,7 +28,7 @@ * * where op() is defined by one of * -* oneapi::mkl::transpose::{nontrans,trans,conjtrans} +* oneapi::math::transpose::{nontrans,trans,conjtrans} * * * This example demonstrates only single precision (float) data type for @@ -85,7 +85,7 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &cpu_dev) { // create execution queue and buffers of matrix data sycl::queue cpu_queue(cpu_dev, exception_handler); - oneapi::mkl::backend_selector cpu_selector{ cpu_queue }; + oneapi::math::backend_selector cpu_selector{ cpu_queue }; intType *ia, *ja; fp *a, *x, *y, *z; @@ -127,59 +127,59 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &cpu_dev) { // Execute Matrix Multiply // - oneapi::mkl::transpose transA = oneapi::mkl::transpose::nontrans; - oneapi::mkl::sparse::spmv_alg alg = oneapi::mkl::sparse::spmv_alg::default_alg; - oneapi::mkl::sparse::matrix_view A_view; + oneapi::math::transpose transA = oneapi::math::transpose::nontrans; + oneapi::math::sparse::spmv_alg alg = oneapi::math::sparse::spmv_alg::default_alg; + oneapi::math::sparse::matrix_view A_view; std::cout << "\n\t\tsparse::spmv parameters:\n"; std::cout << "\t\t\ttransA = " - << (transA == oneapi::mkl::transpose::nontrans + << (transA == oneapi::math::transpose::nontrans ? "nontrans" - : (transA == oneapi::mkl::transpose::trans ? "trans" : "conjtrans")) + : (transA == oneapi::math::transpose::trans ? "trans" : "conjtrans")) << std::endl; std::cout << "\t\t\tnrows = " << nrows << std::endl; std::cout << "\t\t\talpha = " << alpha << ", beta = " << beta << std::endl; // Create and initialize handle for a Sparse Matrix in CSR format - oneapi::mkl::sparse::matrix_handle_t A_handle = nullptr; - oneapi::mkl::sparse::init_csr_matrix(cpu_selector, &A_handle, nrows, nrows, nnz, - oneapi::mkl::index_base::zero, ia, ja, a); + oneapi::math::sparse::matrix_handle_t A_handle = nullptr; + oneapi::math::sparse::init_csr_matrix(cpu_selector, &A_handle, nrows, nrows, nnz, + oneapi::math::index_base::zero, ia, ja, a); // Create and initialize dense vector handles - oneapi::mkl::sparse::dense_vector_handle_t x_handle = nullptr; - oneapi::mkl::sparse::dense_vector_handle_t y_handle = nullptr; - oneapi::mkl::sparse::init_dense_vector(cpu_selector, &x_handle, sizevec, x); - oneapi::mkl::sparse::init_dense_vector(cpu_selector, &y_handle, sizevec, y); + oneapi::math::sparse::dense_vector_handle_t x_handle = nullptr; + oneapi::math::sparse::dense_vector_handle_t y_handle = nullptr; + oneapi::math::sparse::init_dense_vector(cpu_selector, &x_handle, sizevec, x); + oneapi::math::sparse::init_dense_vector(cpu_selector, &y_handle, sizevec, y); // Create operation descriptor - oneapi::mkl::sparse::spmv_descr_t descr = nullptr; - oneapi::mkl::sparse::init_spmv_descr(cpu_selector, &descr); + oneapi::math::sparse::spmv_descr_t descr = nullptr; + oneapi::math::sparse::init_spmv_descr(cpu_selector, &descr); // Allocate external workspace std::size_t workspace_size = 0; - oneapi::mkl::sparse::spmv_buffer_size(cpu_selector, transA, &alpha, A_view, A_handle, x_handle, + oneapi::math::sparse::spmv_buffer_size(cpu_selector, transA, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr, workspace_size); void *workspace = sycl::malloc_device(workspace_size, cpu_queue); // Optimize spmv auto ev_opt = - oneapi::mkl::sparse::spmv_optimize(cpu_selector, transA, &alpha, A_view, A_handle, x_handle, + oneapi::math::sparse::spmv_optimize(cpu_selector, transA, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr, workspace); // Run spmv - auto ev_spmv = oneapi::mkl::sparse::spmv(cpu_selector, transA, &alpha, A_view, A_handle, + auto ev_spmv = oneapi::math::sparse::spmv(cpu_selector, transA, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr, { ev_opt }); // Release handles and descriptor std::vector release_events; release_events.push_back( - oneapi::mkl::sparse::release_dense_vector(cpu_selector, x_handle, { ev_spmv })); + oneapi::math::sparse::release_dense_vector(cpu_selector, x_handle, { ev_spmv })); release_events.push_back( - oneapi::mkl::sparse::release_dense_vector(cpu_selector, y_handle, { ev_spmv })); + oneapi::math::sparse::release_dense_vector(cpu_selector, y_handle, { ev_spmv })); release_events.push_back( - oneapi::mkl::sparse::release_sparse_matrix(cpu_selector, A_handle, { ev_spmv })); + oneapi::math::sparse::release_sparse_matrix(cpu_selector, A_handle, { ev_spmv })); release_events.push_back( - oneapi::mkl::sparse::release_spmv_descr(cpu_selector, descr, { ev_spmv })); + oneapi::math::sparse::release_spmv_descr(cpu_selector, descr, { ev_spmv })); for (auto event : release_events) { event.wait_and_throw(); } @@ -189,7 +189,7 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &cpu_dev) { // fp *res = y; - const bool isConj = (transA == oneapi::mkl::transpose::conjtrans); + const bool isConj = (transA == oneapi::math::transpose::conjtrans); for (intType row = 0; row < nrows; row++) { z[row] *= beta; } diff --git a/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp b/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp index cc6908aae..ba51a08e8 100644 --- a/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp +++ b/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp @@ -20,7 +20,7 @@ /* * * Content: -* This example demonstrates use of DPCPP API oneapi::mkl::sparse::spmv +* This example demonstrates use of DPCPP API oneapi::math::sparse::spmv * using unified shared memory to perform general sparse matrix-vector * multiplication on a SYCL device (HOST, CPU, GPU) that is selected * during runtime. @@ -29,7 +29,7 @@ * * where op() is defined by one of * -* oneapi::mkl::transpose::{nontrans,trans,conjtrans} +* oneapi::math::transpose::{nontrans,trans,conjtrans} * * * This example demonstrates only single precision (float) data type for @@ -128,59 +128,59 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &dev) { // Execute Matrix Multiply // - oneapi::mkl::transpose transA = oneapi::mkl::transpose::nontrans; - oneapi::mkl::sparse::spmv_alg alg = oneapi::mkl::sparse::spmv_alg::default_alg; - oneapi::mkl::sparse::matrix_view A_view; + oneapi::math::transpose transA = oneapi::math::transpose::nontrans; + oneapi::math::sparse::spmv_alg alg = oneapi::math::sparse::spmv_alg::default_alg; + oneapi::math::sparse::matrix_view A_view; std::cout << "\n\t\tsparse::spmv parameters:\n"; std::cout << "\t\t\ttransA = " - << (transA == oneapi::mkl::transpose::nontrans + << (transA == oneapi::math::transpose::nontrans ? "nontrans" - : (transA == oneapi::mkl::transpose::trans ? "trans" : "conjtrans")) + : (transA == oneapi::math::transpose::trans ? "trans" : "conjtrans")) << std::endl; std::cout << "\t\t\tnrows = " << nrows << std::endl; std::cout << "\t\t\talpha = " << alpha << ", beta = " << beta << std::endl; // Create and initialize handle for a Sparse Matrix in CSR format - oneapi::mkl::sparse::matrix_handle_t A_handle = nullptr; - oneapi::mkl::sparse::init_csr_matrix(main_queue, &A_handle, nrows, nrows, nnz, - oneapi::mkl::index_base::zero, ia, ja, a); + oneapi::math::sparse::matrix_handle_t A_handle = nullptr; + oneapi::math::sparse::init_csr_matrix(main_queue, &A_handle, nrows, nrows, nnz, + oneapi::math::index_base::zero, ia, ja, a); // Create and initialize dense vector handles - oneapi::mkl::sparse::dense_vector_handle_t x_handle = nullptr; - oneapi::mkl::sparse::dense_vector_handle_t y_handle = nullptr; - oneapi::mkl::sparse::init_dense_vector(main_queue, &x_handle, sizevec_i64, x); - oneapi::mkl::sparse::init_dense_vector(main_queue, &y_handle, sizevec_i64, y); + oneapi::math::sparse::dense_vector_handle_t x_handle = nullptr; + oneapi::math::sparse::dense_vector_handle_t y_handle = nullptr; + oneapi::math::sparse::init_dense_vector(main_queue, &x_handle, sizevec_i64, x); + oneapi::math::sparse::init_dense_vector(main_queue, &y_handle, sizevec_i64, y); // Create operation descriptor - oneapi::mkl::sparse::spmv_descr_t descr = nullptr; - oneapi::mkl::sparse::init_spmv_descr(main_queue, &descr); + oneapi::math::sparse::spmv_descr_t descr = nullptr; + oneapi::math::sparse::init_spmv_descr(main_queue, &descr); // Allocate external workspace std::size_t workspace_size = 0; - oneapi::mkl::sparse::spmv_buffer_size(main_queue, transA, &alpha, A_view, A_handle, x_handle, + oneapi::math::sparse::spmv_buffer_size(main_queue, transA, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr, workspace_size); void *workspace = sycl::malloc_device(workspace_size, main_queue); // Optimize spmv auto ev_opt = - oneapi::mkl::sparse::spmv_optimize(main_queue, transA, &alpha, A_view, A_handle, x_handle, + oneapi::math::sparse::spmv_optimize(main_queue, transA, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr, workspace); // Run spmv - auto ev_spmv = oneapi::mkl::sparse::spmv(main_queue, transA, &alpha, A_view, A_handle, x_handle, + auto ev_spmv = oneapi::math::sparse::spmv(main_queue, transA, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr, { ev_opt }); // Release handles and descriptor std::vector release_events; release_events.push_back( - oneapi::mkl::sparse::release_dense_vector(main_queue, x_handle, { ev_spmv })); + oneapi::math::sparse::release_dense_vector(main_queue, x_handle, { ev_spmv })); release_events.push_back( - oneapi::mkl::sparse::release_dense_vector(main_queue, y_handle, { ev_spmv })); + oneapi::math::sparse::release_dense_vector(main_queue, y_handle, { ev_spmv })); release_events.push_back( - oneapi::mkl::sparse::release_sparse_matrix(main_queue, A_handle, { ev_spmv })); + oneapi::math::sparse::release_sparse_matrix(main_queue, A_handle, { ev_spmv })); release_events.push_back( - oneapi::mkl::sparse::release_spmv_descr(main_queue, descr, { ev_spmv })); + oneapi::math::sparse::release_spmv_descr(main_queue, descr, { ev_spmv })); for (auto event : release_events) { event.wait_and_throw(); } @@ -190,7 +190,7 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &dev) { // fp *res = y; - const bool isConj = (transA == oneapi::mkl::transpose::conjtrans); + const bool isConj = (transA == oneapi::math::transpose::conjtrans); for (intType row = 0; row < nrows; row++) { z[row] *= beta; } diff --git a/include/oneapi/math/bfloat16.hpp b/include/oneapi/math/bfloat16.hpp index afa155b1a..ebf267283 100644 --- a/include/oneapi/math/bfloat16.hpp +++ b/include/oneapi/math/bfloat16.hpp @@ -25,7 +25,7 @@ #include namespace oneapi { -namespace mkl { +namespace math { namespace bfloat16_impl { @@ -222,7 +222,7 @@ inline bfloat16::operator float() const { return bfloat16_impl::raw_to_float(raw << 16); } -} /* namespace mkl */ +} /* namespace math */ } // namespace oneapi #endif /* _BFLOAT16_HPP__ */ diff --git a/include/oneapi/math/blas.hpp b/include/oneapi/math/blas.hpp index 66f31c4d4..a58f72fb4 100644 --- a/include/oneapi/math/blas.hpp +++ b/include/oneapi/math/blas.hpp @@ -54,7 +54,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace column_major { @@ -67,7 +67,7 @@ namespace row_major { } //namespace row_major } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_ONEMATH_BLAS_LOADER_HPP_ diff --git a/include/oneapi/math/blas/detail/blas_ct_backends.hpp b/include/oneapi/math/blas/detail/blas_ct_backends.hpp index a12e6778a..ee2631fba 100644 --- a/include/oneapi/math/blas/detail/blas_ct_backends.hpp +++ b/include/oneapi/math/blas/detail/blas_ct_backends.hpp @@ -32,7 +32,7 @@ #include "oneapi/math/detail/backend_selector.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace column_major { @@ -79,7 +79,7 @@ namespace row_major { } //namespace row_major } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_BLAS_CT_BACKENDS_HPP__ diff --git a/include/oneapi/math/blas/detail/blas_loader.hpp b/include/oneapi/math/blas/detail/blas_loader.hpp index 078334154..611add52a 100644 --- a/include/oneapi/math/blas/detail/blas_loader.hpp +++ b/include/oneapi/math/blas/detail/blas_loader.hpp @@ -34,7 +34,7 @@ #include "oneapi/math/detail/get_device_id.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace column_major { namespace detail { @@ -51,7 +51,7 @@ namespace detail { } //namespace detail } //namespace row_major } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_ONEMATH_BLAS_LOADER_HPP_ diff --git a/include/oneapi/math/blas/detail/blas_loader.hxx b/include/oneapi/math/blas/detail/blas_loader.hxx index bebe32c02..9732218de 100644 --- a/include/oneapi/math/blas/detail/blas_loader.hxx +++ b/include/oneapi/math/blas/detail/blas_loader.hxx @@ -19,88 +19,88 @@ // Buffer APIs -ONEMATH_EXPORT void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); -ONEMATH_EXPORT void spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); -ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -108,7 +108,7 @@ ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, t std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -116,7 +116,7 @@ ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, t std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, @@ -124,21 +124,21 @@ ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, t std::int64_t stride_b, sycl::half beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, @@ -146,43 +146,43 @@ ONEMATH_EXPORT void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, t sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -190,158 +190,158 @@ ONEMATH_EXPORT void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, u sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, float s); -ONEMATH_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, double s); -ONEMATH_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s); -ONEMATH_EXPORT void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s); -ONEMATH_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, @@ -349,7 +349,7 @@ ONEMATH_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, t sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -357,7 +357,7 @@ ONEMATH_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, t sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -365,21 +365,21 @@ ONEMATH_EXPORT void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, t sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMATH_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMATH_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -387,7 +387,7 @@ ONEMATH_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, s sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMATH_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -395,44 +395,44 @@ ONEMATH_EXPORT void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, s sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMATH_EXPORT void her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); -ONEMATH_EXPORT void hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); -ONEMATH_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMATH_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMATH_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMATH_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, @@ -440,606 +440,606 @@ ONEMATH_EXPORT void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, tr sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMATH_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void rotmg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotmg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m); -ONEMATH_EXPORT void rotmg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotmg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m); -ONEMATH_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMATH_EXPORT void dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMATH_EXPORT void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); -ONEMATH_EXPORT void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); -ONEMATH_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMATH_EXPORT void dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result); -ONEMATH_EXPORT void syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result); -ONEMATH_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); -ONEMATH_EXPORT void spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); -ONEMATH_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m); -ONEMATH_EXPORT void rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m); -ONEMATH_EXPORT void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMATH_EXPORT void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMATH_EXPORT void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMATH_EXPORT void sdsdot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void sdsdot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float sb, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result); -ONEMATH_EXPORT void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s); -ONEMATH_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s); -ONEMATH_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s); -ONEMATH_EXPORT void rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s); -ONEMATH_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, @@ -1048,73 +1048,73 @@ ONEMATH_EXPORT void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, @@ -1123,92 +1123,92 @@ ONEMATH_EXPORT void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, tran // USM APIs -ONEMATH_EXPORT sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, const float **b, @@ -1216,7 +1216,7 @@ ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, const double **b, @@ -1224,7 +1224,7 @@ ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, @@ -1234,13 +1234,13 @@ ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemm_batch( - oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, + oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, const sycl::half **a, std::int64_t *lda, @@ -1248,14 +1248,14 @@ ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &q sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, @@ -1263,7 +1263,7 @@ ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &q float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, @@ -1271,7 +1271,7 @@ ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &q std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, @@ -1279,7 +1279,7 @@ ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &q float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, @@ -1288,26 +1288,26 @@ ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemm_batch( - oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, + oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemm_batch( - oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, + oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemm_batch( - oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, + oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, @@ -1315,7 +1315,7 @@ ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &q float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, @@ -1323,7 +1323,7 @@ ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &q float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, @@ -1332,24 +1332,24 @@ ONEMATH_EXPORT sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1357,21 +1357,21 @@ ONEMATH_EXPORT sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, @@ -1379,7 +1379,7 @@ ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, @@ -1388,21 +1388,21 @@ ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1410,7 +1410,7 @@ ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &q std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1419,27 +1419,27 @@ ONEMATH_EXPORT sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, @@ -1447,84 +1447,84 @@ ONEMATH_EXPORT sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, double c, double s, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float c, float s, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double c, double s, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, float *alpha, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, double *alpha, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::complex *alpha, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::complex *alpha, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, @@ -1532,54 +1532,54 @@ ONEMATH_EXPORT sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1587,7 +1587,7 @@ ONEMATH_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1596,24 +1596,24 @@ ONEMATH_EXPORT sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, @@ -1621,7 +1621,7 @@ ONEMATH_EXPORT sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, @@ -1629,7 +1629,7 @@ ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, @@ -1638,32 +1638,32 @@ ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t batch_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemv_batch( - oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, + oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemv_batch( - oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, + oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, @@ -1672,27 +1672,27 @@ ONEMATH_EXPORT sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemv_batch( - oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, + oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, @@ -1700,7 +1700,7 @@ ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &q std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, @@ -1708,26 +1708,26 @@ ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &q std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, @@ -1735,135 +1735,135 @@ ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &q std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotmg(oneapi::mkl::device libkey, sycl::queue &queue, float *d1, +ONEMATH_EXPORT sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotmg(oneapi::mkl::device libkey, sycl::queue &queue, double *d1, +ONEMATH_EXPORT sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1871,7 +1871,7 @@ ONEMATH_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1880,19 +1880,19 @@ ONEMATH_EXPORT sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1900,7 +1900,7 @@ ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1908,27 +1908,27 @@ ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, const std::int8_t *a, int64_t lda, std::int8_t ao, @@ -1937,11 +1937,11 @@ ONEMATH_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &qu const std::int32_t *co, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemm_bias( - oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, + oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, const std::int8_t *a, int64_t lda, std::int8_t ao, const std::int8_t *b, int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, int64_t ldc, const std::int32_t *co, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, const std::uint8_t *a, int64_t lda, std::uint8_t ao, @@ -1949,7 +1949,7 @@ ONEMATH_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &qu float beta, std::int32_t *c, int64_t ldc, const std::int32_t *co, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, const std::uint8_t *a, int64_t lda, std::uint8_t ao, @@ -1958,97 +1958,97 @@ ONEMATH_EXPORT sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &qu const std::int32_t *co, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, float *alpha, const float **a, int64_t *lda, float **b, int64_t *ldb, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, double *alpha, const double **a, int64_t *lda, double **b, int64_t *ldb, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, std::complex *alpha, const std::complex **a, int64_t *lda, std::complex **b, int64_t *ldb, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, std::complex *alpha, @@ -2057,18 +2057,18 @@ ONEMATH_EXPORT sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &q int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2076,7 +2076,7 @@ ONEMATH_EXPORT sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2085,32 +2085,32 @@ ONEMATH_EXPORT sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2118,7 +2118,7 @@ ONEMATH_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2127,40 +2127,40 @@ ONEMATH_EXPORT sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2168,7 +2168,7 @@ ONEMATH_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2177,171 +2177,171 @@ ONEMATH_EXPORT sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -2349,115 +2349,115 @@ ONEMATH_EXPORT sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spr2(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float *param, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotm(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double *param, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, double *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sdsdot(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event sdsdot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float sb, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -2465,41 +2465,41 @@ ONEMATH_EXPORT sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, float *a, +ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, float *a, float *b, float *c, float *s, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, double *a, +ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, double *a, double *b, double *c, double *s, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, std::complex *a, std::complex *b, float *c, std::complex *s, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, std::complex *a, std::complex *b, double *c, std::complex *s, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, @@ -2507,31 +2507,31 @@ ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queu std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, @@ -2539,7 +2539,7 @@ ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, @@ -2548,97 +2548,97 @@ ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue std::int64_t batch_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event omatadd_batch( - oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, + oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event omatadd_batch( - oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, + oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, @@ -2646,26 +2646,26 @@ ONEMATH_EXPORT sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queu std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, @@ -2673,25 +2673,25 @@ ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queu std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, float* alpha, float** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, double* alpha, double** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue& queue, +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, std::complex* alpha, std::complex** ab, std::int64_t* lda, std::int64_t* ldb, diff --git a/include/oneapi/math/blas/detail/cublas/blas_ct.hpp b/include/oneapi/math/blas/detail/cublas/blas_ct.hpp index 8b53854f4..6f3384ecd 100644 --- a/include/oneapi/math/blas/detail/cublas/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/cublas/blas_ct.hpp @@ -34,7 +34,7 @@ #include "oneapi/math/blas/detail/blas_ct_backends.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace column_major { @@ -51,7 +51,7 @@ namespace row_major { } //namespace row_major } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_DETAIL_CUBLAS_BLAS_CT_HPP_ diff --git a/include/oneapi/math/blas/detail/cublas/blas_ct.hxx b/include/oneapi/math/blas/detail/cublas/blas_ct.hxx index 9483a66c1..c9ac759ab 100644 --- a/include/oneapi/math/blas/detail/cublas/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/cublas/blas_ct.hxx @@ -21,7 +21,7 @@ void herk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -29,104 +29,104 @@ void herk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void scal(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, @@ -135,7 +135,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -146,7 +146,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -158,7 +158,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -170,7 +170,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -181,7 +181,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -192,7 +192,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -203,7 +203,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -214,7 +214,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -222,14 +222,14 @@ void gemm_batch(backend_selector selector, transpose transa, tr void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -237,7 +237,7 @@ void syrk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -245,7 +245,7 @@ void syrk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -253,7 +253,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -263,7 +263,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -274,7 +274,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -285,7 +285,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -294,7 +294,7 @@ void her2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -302,7 +302,7 @@ void her2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -311,7 +311,7 @@ void hbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -320,61 +320,61 @@ void hbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void rot(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { - oneapi::mkl::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { - oneapi::mkl::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { - oneapi::mkl::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { - oneapi::mkl::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -382,7 +382,7 @@ void axpy_batch(backend_selector selector, std::int64_t n, doub sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -391,7 +391,7 @@ void axpy_batch(backend_selector selector, std::int64_t n, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -400,46 +400,46 @@ void axpy_batch(backend_selector selector, std::int64_t n, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } void axpby(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void sdsdot(backend_selector selector, std::int64_t n, float sb, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); + oneapi::math::blas::cublas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -447,7 +447,7 @@ void gerc(backend_selector selector, std::int64_t m, std::int64 std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -455,7 +455,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -463,7 +463,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -472,7 +472,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -482,7 +482,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -490,7 +490,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -498,7 +498,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -507,7 +507,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -516,7 +516,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -525,7 +525,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -535,7 +535,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -547,7 +547,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -559,7 +559,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t incx, std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -569,7 +569,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -579,7 +579,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -590,7 +590,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -601,7 +601,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -609,58 +609,58 @@ void dgmm_batch(backend_selector selector, side left_right, std void her(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -669,21 +669,21 @@ void hpmv(backend_selector selector, uplo upper_lower, std::int sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -693,7 +693,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -703,7 +703,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -713,7 +713,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -723,38 +723,38 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -762,37 +762,37 @@ void geru(backend_selector selector, std::int64_t m, std::int64 std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -800,7 +800,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -809,7 +809,7 @@ void gemm(backend_selector selector, transpose transa, transpos sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -818,7 +818,7 @@ void gemm(backend_selector selector, transpose transa, transpos sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -826,7 +826,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -834,7 +834,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -842,14 +842,14 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -857,21 +857,21 @@ void syr2(backend_selector selector, uplo upper_lower, std::int double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -879,7 +879,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -887,7 +887,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -895,7 +895,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -903,7 +903,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -911,14 +911,14 @@ void dotu(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, @@ -926,7 +926,7 @@ void hemm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -935,7 +935,7 @@ void hemm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -943,7 +943,7 @@ void hpr2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -951,7 +951,7 @@ void hpr2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -959,7 +959,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -967,7 +967,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -976,7 +976,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -986,21 +986,21 @@ void gbmv(backend_selector selector, transpose trans, std::int6 sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1008,7 +1008,7 @@ void tbmv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1016,7 +1016,7 @@ void tbmv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1024,7 +1024,7 @@ void symm(backend_selector selector, side left_right, uplo uppe std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1032,7 +1032,7 @@ void symm(backend_selector selector, side left_right, uplo uppe std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1041,7 +1041,7 @@ void symm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1050,7 +1050,7 @@ void symm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1058,27 +1058,27 @@ void dotc(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } @@ -1086,7 +1086,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1094,7 +1094,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1102,7 +1102,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1110,106 +1110,106 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m) { - oneapi::mkl::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m) { - oneapi::mkl::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1217,7 +1217,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1225,7 +1225,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1233,7 +1233,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1241,7 +1241,7 @@ void hemv(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::cublas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1250,7 +1250,7 @@ void hemv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::cublas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1258,7 +1258,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1266,7 +1266,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1275,7 +1275,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1285,37 +1285,37 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -1323,21 +1323,21 @@ void sbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1345,7 +1345,7 @@ void tbsv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1353,76 +1353,76 @@ void tbsv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, @@ -1430,7 +1430,7 @@ void trsm_batch(backend_selector selector, side left_right, upl sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1440,7 +1440,7 @@ void trsm_batch(backend_selector selector, side left_right, upl sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1451,7 +1451,7 @@ void trsm_batch(backend_selector selector, side left_right, upl std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1462,7 +1462,7 @@ void trsm_batch(backend_selector selector, side left_right, upl std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1472,7 +1472,7 @@ void her2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1481,38 +1481,38 @@ void her2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::cublas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1520,7 +1520,7 @@ void symv(backend_selector selector, uplo upper_lower, std::int double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::cublas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::cublas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1528,7 +1528,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1536,7 +1536,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1545,7 +1545,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1554,21 +1554,21 @@ void omatcopy_batch(backend_selector selector, transpose trans, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1576,7 +1576,7 @@ void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1584,7 +1584,7 @@ void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1593,7 +1593,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1603,7 +1603,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1614,7 +1614,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1626,7 +1626,7 @@ void omatadd_batch(backend_selector selector, transpose transa, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1634,28 +1634,28 @@ void omatadd_batch(backend_selector selector, transpose transa, void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } @@ -1663,7 +1663,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1671,7 +1671,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1679,7 +1679,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1687,35 +1687,35 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } @@ -1723,7 +1723,7 @@ void omatadd(backend_selector selector, transpose transa, trans std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1731,7 +1731,7 @@ void omatadd(backend_selector selector, transpose transa, trans std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1740,7 +1740,7 @@ void omatadd(backend_selector selector, transpose transa, trans sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1749,7 +1749,7 @@ void omatadd(backend_selector selector, transpose transa, trans sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1759,7 +1759,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1768,7 +1768,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1776,7 +1776,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s sycl::event scal(backend_selector selector, std::int64_t n, float alpha, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1784,7 +1784,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, flo sycl::event scal(backend_selector selector, std::int64_t n, double alpha, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1792,7 +1792,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, dou sycl::event scal(backend_selector selector, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1800,7 +1800,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, sycl::event scal(backend_selector selector, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1808,7 +1808,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, sycl::event scal(backend_selector selector, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1816,7 +1816,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, flo sycl::event scal(backend_selector selector, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1824,7 +1824,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, dou sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1832,7 +1832,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1841,7 +1841,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1850,7 +1850,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1858,7 +1858,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1866,7 +1866,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1875,7 +1875,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1884,7 +1884,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1892,7 +1892,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1900,7 +1900,7 @@ sycl::event spr(backend_selector selector, uplo upper_lower, st sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1910,7 +1910,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1920,7 +1920,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1929,7 +1929,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk( + auto done = oneapi::math::blas::cublas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1938,7 +1938,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk( + auto done = oneapi::math::blas::cublas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1948,7 +1948,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk( + auto done = oneapi::math::blas::cublas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1958,7 +1958,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk( + auto done = oneapi::math::blas::cublas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1968,7 +1968,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1979,7 +1979,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1991,7 +1991,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2003,7 +2003,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2015,7 +2015,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2027,7 +2027,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2040,7 +2040,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2053,7 +2053,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2063,7 +2063,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2072,7 +2072,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2083,7 +2083,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, s std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2094,7 +2094,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, s std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2103,7 +2103,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2112,7 +2112,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2120,7 +2120,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, sycl::event rot(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2128,7 +2128,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, floa sycl::event rot(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2136,7 +2136,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, doub sycl::event axpy(backend_selector selector, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2144,7 +2144,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, flo sycl::event axpy(backend_selector selector, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2153,7 +2153,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2162,7 +2162,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2171,7 +2171,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t float *alpha, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2180,7 +2180,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t double *alpha, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2190,7 +2190,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2200,7 +2200,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2209,7 +2209,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2219,7 +2219,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2230,7 +2230,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2241,7 +2241,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2250,7 +2250,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t sycl::event axpby(backend_selector selector, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2258,7 +2258,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, fl sycl::event axpby(backend_selector selector, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2267,7 +2267,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2276,7 +2276,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2285,7 +2285,7 @@ sycl::event gerc(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2294,7 +2294,7 @@ sycl::event gerc(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2304,7 +2304,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2314,7 +2314,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2325,7 +2325,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2337,7 +2337,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2346,7 +2346,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2355,7 +2355,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2365,7 +2365,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2375,7 +2375,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2386,7 +2386,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::int64_t incx, std::int64_t stridex, float beta, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2398,7 +2398,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::int64_t incx, std::int64_t stridex, double beta, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2411,7 +2411,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2424,7 +2424,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2436,7 +2436,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2448,7 +2448,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2461,7 +2461,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2474,7 +2474,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2486,7 +2486,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2498,7 +2498,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2510,7 +2510,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2522,7 +2522,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2533,7 +2533,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2544,7 +2544,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2556,7 +2556,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2568,7 +2568,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2578,7 +2578,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2587,7 +2587,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2595,7 +2595,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -2603,7 +2603,7 @@ sycl::event hpr(backend_selector selector, uplo upper_lower, st sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -2611,7 +2611,7 @@ sycl::event hpr(backend_selector selector, uplo upper_lower, st sycl::event iamin(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2619,7 +2619,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, co sycl::event iamin(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2627,7 +2627,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, co sycl::event iamin(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2635,7 +2635,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, sycl::event iamin(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2646,7 +2646,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2658,7 +2658,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2671,7 +2671,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2684,7 +2684,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2697,7 +2697,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2709,7 +2709,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2721,7 +2721,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2733,7 +2733,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2746,7 +2746,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2759,7 +2759,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2773,7 +2773,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2787,7 +2787,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2800,7 +2800,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2812,7 +2812,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2824,7 +2824,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2836,7 +2836,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2846,7 +2846,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -2855,7 +2855,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -2863,7 +2863,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s sycl::event swap(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2871,7 +2871,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, flo sycl::event swap(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2879,7 +2879,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, dou sycl::event swap(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2887,7 +2887,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, sycl::event swap(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2896,7 +2896,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2905,7 +2905,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2913,7 +2913,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std sycl::event nrm2(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2921,7 +2921,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, sycl::event nrm2(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2929,7 +2929,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, sycl::event nrm2(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2937,7 +2937,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, con sycl::event nrm2(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2947,7 +2947,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2957,7 +2957,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2968,7 +2968,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2979,7 +2979,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2990,7 +2990,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3001,7 +3001,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3011,7 +3011,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3022,7 +3022,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3034,7 +3034,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3046,7 +3046,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3058,7 +3058,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::cublas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3068,7 +3068,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::herk( + auto done = oneapi::math::blas::cublas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -3077,7 +3077,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::herk( + auto done = oneapi::math::blas::cublas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -3086,7 +3086,7 @@ sycl::event ger(backend_selector selector, std::int64_t m, std: float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, + auto done = oneapi::math::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -3095,7 +3095,7 @@ sycl::event ger(backend_selector selector, std::int64_t m, std: double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, + auto done = oneapi::math::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -3104,7 +3104,7 @@ sycl::event trsm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3114,7 +3114,7 @@ sycl::event trsm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3125,7 +3125,7 @@ sycl::event trsm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3136,7 +3136,7 @@ sycl::event trsm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3148,7 +3148,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3160,7 +3160,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3172,7 +3172,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3185,7 +3185,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3197,7 +3197,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3209,7 +3209,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3222,7 +3222,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3235,7 +3235,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3245,7 +3245,7 @@ sycl::event dotu(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3254,7 +3254,7 @@ sycl::event dotu(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3265,7 +3265,7 @@ sycl::event hemm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3276,7 +3276,7 @@ sycl::event hemm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3285,7 +3285,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3294,7 +3294,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3304,7 +3304,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3315,7 +3315,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st double beta, double *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3326,7 +3326,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3337,7 +3337,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3346,7 +3346,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tbmv( + auto done = oneapi::math::blas::cublas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3355,7 +3355,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tbmv( + auto done = oneapi::math::blas::cublas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3364,7 +3364,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tbmv( + auto done = oneapi::math::blas::cublas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3373,7 +3373,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tbmv( + auto done = oneapi::math::blas::cublas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3383,7 +3383,7 @@ sycl::event symm(backend_selector selector, side left_right, up const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3393,7 +3393,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3404,7 +3404,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3415,7 +3415,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3424,7 +3424,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3433,7 +3433,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3441,7 +3441,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3449,7 +3449,7 @@ sycl::event syr(backend_selector selector, uplo upper_lower, st sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3458,7 +3458,7 @@ sycl::event trmm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3468,7 +3468,7 @@ sycl::event trmm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3479,7 +3479,7 @@ sycl::event trmm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3490,7 +3490,7 @@ sycl::event trmm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3498,7 +3498,7 @@ sycl::event trmm(backend_selector selector, side left_right, up sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, + auto done = oneapi::math::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } @@ -3506,7 +3506,7 @@ sycl::event rotmg(backend_selector selector, float *d1, float * sycl::event rotmg(backend_selector selector, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, + auto done = oneapi::math::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } @@ -3514,7 +3514,7 @@ sycl::event rotmg(backend_selector selector, double *d1, double sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3522,7 +3522,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3531,7 +3531,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3540,7 +3540,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3548,7 +3548,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3556,7 +3556,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3565,7 +3565,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3574,7 +3574,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3582,7 +3582,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t sycl::event copy(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3590,7 +3590,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, con sycl::event copy(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3598,7 +3598,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, con sycl::event copy(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3606,7 +3606,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, sycl::event copy(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3615,7 +3615,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3624,7 +3624,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3634,7 +3634,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3644,7 +3644,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3653,7 +3653,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3662,7 +3662,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3672,7 +3672,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3682,7 +3682,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::copy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3692,7 +3692,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::hemv( + auto done = oneapi::math::blas::cublas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3702,7 +3702,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::hemv( + auto done = oneapi::math::blas::cublas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3712,7 +3712,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3723,7 +3723,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3735,7 +3735,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3747,7 +3747,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3758,7 +3758,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3768,7 +3768,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3776,7 +3776,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s sycl::event asum(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3784,7 +3784,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, sycl::event asum(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3792,7 +3792,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, sycl::event asum(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3800,7 +3800,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, con sycl::event asum(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3809,7 +3809,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tbsv( + auto done = oneapi::math::blas::cublas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3818,7 +3818,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tbsv( + auto done = oneapi::math::blas::cublas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3827,7 +3827,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tbsv( + auto done = oneapi::math::blas::cublas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3836,7 +3836,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::tbsv( + auto done = oneapi::math::blas::cublas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3845,7 +3845,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3854,7 +3854,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3862,7 +3862,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s sycl::event iamax(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3870,7 +3870,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, co sycl::event iamax(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3878,7 +3878,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, co sycl::event iamax(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3886,7 +3886,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event iamax(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3894,7 +3894,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event rotm(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3902,7 +3902,7 @@ sycl::event rotm(backend_selector selector, std::int64_t n, flo sycl::event rotm(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3910,14 +3910,14 @@ sycl::event rotm(backend_selector selector, std::int64_t n, dou sycl::event rotg(backend_selector selector, float *a, float *b, float *c, float *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event rotg(backend_selector selector, double *a, double *b, double *c, double *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3925,7 +3925,7 @@ sycl::event rotg(backend_selector selector, std::complex std::complex *b, float *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3933,14 +3933,14 @@ sycl::event rotg(backend_selector selector, std::complex *b, double *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event sdsdot(backend_selector selector, std::int64_t n, float sb, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, + auto done = oneapi::math::blas::cublas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result, dependencies); return done; } @@ -3951,7 +3951,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3963,7 +3963,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3971,7 +3971,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, sycl::event dot(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3979,7 +3979,7 @@ sycl::event dot(backend_selector selector, std::int64_t n, cons sycl::event dot(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3987,7 +3987,7 @@ sycl::event dot(backend_selector selector, std::int64_t n, cons sycl::event dot(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3996,7 +3996,7 @@ sycl::event symv(backend_selector selector, uplo upper_lower, s float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::symv( + auto done = oneapi::math::blas::cublas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -4005,7 +4005,7 @@ sycl::event symv(backend_selector selector, uplo upper_lower, s double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::symv( + auto done = oneapi::math::blas::cublas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -4015,7 +4015,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4026,7 +4026,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4037,7 +4037,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4048,7 +4048,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4058,7 +4058,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4067,7 +4067,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4077,7 +4077,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4087,7 +4087,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4098,7 +4098,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::cublas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4110,7 +4110,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::cublas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4123,7 +4123,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::cublas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4136,7 +4136,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::cublas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4145,7 +4145,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4153,7 +4153,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4162,7 +4162,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4171,7 +4171,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4180,7 +4180,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4189,7 +4189,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4199,7 +4199,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4209,7 +4209,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4217,7 +4217,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4225,7 +4225,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4234,7 +4234,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4243,7 +4243,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4253,7 +4253,7 @@ sycl::event omatadd(backend_selector selector, transpose transa float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4263,7 +4263,7 @@ sycl::event omatadd(backend_selector selector, transpose transa double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4274,7 +4274,7 @@ sycl::event omatadd(backend_selector selector, transpose transa const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4285,7 +4285,7 @@ sycl::event omatadd(backend_selector selector, transpose transa const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4295,7 +4295,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4306,7 +4306,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4318,7 +4318,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4330,7 +4330,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4341,7 +4341,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4352,7 +4352,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4363,7 +4363,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4374,7 +4374,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; diff --git a/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp b/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp index 234782ca4..1b86c5bb6 100644 --- a/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp +++ b/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp @@ -29,12 +29,12 @@ #include "oneapi/math/types.hpp" namespace oneapi { -namespace mkl { -using oneapi::mkl::diag; -using oneapi::mkl::offset; -using oneapi::mkl::side; -using oneapi::mkl::transpose; -using oneapi::mkl::uplo; +namespace math { +using oneapi::math::diag; +using oneapi::math::offset; +using oneapi::math::side; +using oneapi::math::transpose; +using oneapi::math::uplo; namespace blas { namespace cublas { namespace column_major { @@ -49,7 +49,7 @@ namespace row_major { } //namespace row_major } //namespace cublas } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_ONEMATH_BLAS_CUBLAS_HPP_ diff --git a/include/oneapi/math/blas/detail/mklcpu/blas_ct.hpp b/include/oneapi/math/blas/detail/mklcpu/blas_ct.hpp index 6f614fcda..fc9367ed1 100644 --- a/include/oneapi/math/blas/detail/mklcpu/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/mklcpu/blas_ct.hpp @@ -35,7 +35,7 @@ #include "oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace column_major { @@ -52,7 +52,7 @@ namespace row_major { } //namespace row_major } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_DETAIL_MKLCPU_BLAS_CT_HPP_ diff --git a/include/oneapi/math/blas/detail/mklcpu/blas_ct.hxx b/include/oneapi/math/blas/detail/mklcpu/blas_ct.hxx index 1724bf5c7..0d34723db 100644 --- a/include/oneapi/math/blas/detail/mklcpu/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/mklcpu/blas_ct.hxx @@ -23,7 +23,7 @@ void herk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -31,104 +31,104 @@ void herk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void scal(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, @@ -137,7 +137,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -148,7 +148,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -160,7 +160,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -172,7 +172,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -183,7 +183,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -194,7 +194,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -205,7 +205,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -216,7 +216,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -224,14 +224,14 @@ void gemm_batch(backend_selector selector, transpose transa, tr void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -239,7 +239,7 @@ void syrk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -247,7 +247,7 @@ void syrk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -255,7 +255,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -265,7 +265,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -276,7 +276,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -287,7 +287,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -296,7 +296,7 @@ void her2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -304,7 +304,7 @@ void her2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -313,7 +313,7 @@ void hbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -322,61 +322,61 @@ void hbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void rot(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { - oneapi::mkl::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { - oneapi::mkl::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { - oneapi::mkl::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { - oneapi::mkl::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -384,7 +384,7 @@ void axpy_batch(backend_selector selector, std::int64_t n, doub sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -393,7 +393,7 @@ void axpy_batch(backend_selector selector, std::int64_t n, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -402,46 +402,46 @@ void axpy_batch(backend_selector selector, std::int64_t n, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } void axpby(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void sdsdot(backend_selector selector, std::int64_t n, float sb, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); + oneapi::math::blas::mklcpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -449,7 +449,7 @@ void gerc(backend_selector selector, std::int64_t m, std::int64 std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -457,7 +457,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -465,7 +465,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -474,7 +474,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -484,7 +484,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -492,7 +492,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -500,7 +500,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -509,7 +509,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -518,7 +518,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -527,7 +527,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -537,7 +537,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -549,7 +549,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -561,7 +561,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t incx, std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -571,7 +571,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -581,7 +581,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -592,7 +592,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -603,7 +603,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -611,58 +611,58 @@ void dgmm_batch(backend_selector selector, side left_right, std void her(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -671,21 +671,21 @@ void hpmv(backend_selector selector, uplo upper_lower, std::int sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -695,7 +695,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -705,7 +705,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -715,7 +715,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -725,38 +725,38 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -764,37 +764,37 @@ void geru(backend_selector selector, std::int64_t m, std::int64 std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -802,7 +802,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -811,7 +811,7 @@ void gemm(backend_selector selector, transpose transa, transpos sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -820,7 +820,7 @@ void gemm(backend_selector selector, transpose transa, transpos sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -828,7 +828,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -836,7 +836,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -844,14 +844,14 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -859,21 +859,21 @@ void syr2(backend_selector selector, uplo upper_lower, std::int double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -881,7 +881,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -889,7 +889,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -897,7 +897,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -905,7 +905,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -913,14 +913,14 @@ void dotu(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, @@ -928,7 +928,7 @@ void hemm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -937,7 +937,7 @@ void hemm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -945,7 +945,7 @@ void hpr2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -953,7 +953,7 @@ void hpr2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -961,7 +961,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -969,7 +969,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -978,7 +978,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -988,21 +988,21 @@ void gbmv(backend_selector selector, transpose trans, std::int6 sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1010,7 +1010,7 @@ void tbmv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1018,7 +1018,7 @@ void tbmv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1026,7 +1026,7 @@ void symm(backend_selector selector, side left_right, uplo uppe std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1034,7 +1034,7 @@ void symm(backend_selector selector, side left_right, uplo uppe std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1043,7 +1043,7 @@ void symm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1052,7 +1052,7 @@ void symm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1060,27 +1060,27 @@ void dotc(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } @@ -1088,7 +1088,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1096,7 +1096,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1104,7 +1104,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1112,106 +1112,106 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m) { - oneapi::mkl::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m) { - oneapi::mkl::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1219,7 +1219,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1227,7 +1227,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1235,7 +1235,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1243,7 +1243,7 @@ void hemv(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::mklcpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1252,7 +1252,7 @@ void hemv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::mklcpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1260,7 +1260,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1268,7 +1268,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1277,7 +1277,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1287,37 +1287,37 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -1325,21 +1325,21 @@ void sbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1347,7 +1347,7 @@ void tbsv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1355,76 +1355,76 @@ void tbsv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, @@ -1432,7 +1432,7 @@ void trsm_batch(backend_selector selector, side left_right, upl sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1442,7 +1442,7 @@ void trsm_batch(backend_selector selector, side left_right, upl sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1453,7 +1453,7 @@ void trsm_batch(backend_selector selector, side left_right, upl std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1464,7 +1464,7 @@ void trsm_batch(backend_selector selector, side left_right, upl std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1474,7 +1474,7 @@ void her2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1483,38 +1483,38 @@ void her2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::mklcpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1522,7 +1522,7 @@ void symv(backend_selector selector, uplo upper_lower, std::int double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklcpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::mklcpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1530,7 +1530,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1538,7 +1538,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1547,7 +1547,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1556,21 +1556,21 @@ void omatcopy_batch(backend_selector selector, transpose trans, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1578,7 +1578,7 @@ void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1586,7 +1586,7 @@ void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1595,7 +1595,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1605,7 +1605,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1616,7 +1616,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1628,7 +1628,7 @@ void omatadd_batch(backend_selector selector, transpose transa, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1636,28 +1636,28 @@ void omatadd_batch(backend_selector selector, transpose transa, void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } @@ -1665,7 +1665,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1673,7 +1673,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1681,7 +1681,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1689,35 +1689,35 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } @@ -1725,7 +1725,7 @@ void omatadd(backend_selector selector, transpose transa, trans std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1733,7 +1733,7 @@ void omatadd(backend_selector selector, transpose transa, trans std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1742,7 +1742,7 @@ void omatadd(backend_selector selector, transpose transa, trans sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1751,7 +1751,7 @@ void omatadd(backend_selector selector, transpose transa, trans sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1761,7 +1761,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1770,7 +1770,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1778,7 +1778,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s sycl::event scal(backend_selector selector, std::int64_t n, float alpha, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1786,7 +1786,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, flo sycl::event scal(backend_selector selector, std::int64_t n, double alpha, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1794,7 +1794,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, dou sycl::event scal(backend_selector selector, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1802,7 +1802,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, sycl::event scal(backend_selector selector, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1810,7 +1810,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, sycl::event scal(backend_selector selector, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1818,7 +1818,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, flo sycl::event scal(backend_selector selector, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1826,7 +1826,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, dou sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1834,7 +1834,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1843,7 +1843,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1852,7 +1852,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1860,7 +1860,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1868,7 +1868,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1877,7 +1877,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1886,7 +1886,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1894,7 +1894,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1902,7 +1902,7 @@ sycl::event spr(backend_selector selector, uplo upper_lower, st sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1912,7 +1912,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1922,7 +1922,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1931,7 +1931,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1940,7 +1940,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1950,7 +1950,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1960,7 +1960,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1970,7 +1970,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1981,7 +1981,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1993,7 +1993,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2005,7 +2005,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2017,7 +2017,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2029,7 +2029,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2042,7 +2042,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2055,7 +2055,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2065,7 +2065,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2074,7 +2074,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2085,7 +2085,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, s std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2096,7 +2096,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, s std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2105,7 +2105,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2114,7 +2114,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2122,7 +2122,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, sycl::event rot(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2130,7 +2130,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, floa sycl::event rot(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2138,7 +2138,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, doub sycl::event axpy(backend_selector selector, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2146,7 +2146,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, flo sycl::event axpy(backend_selector selector, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2155,7 +2155,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2164,7 +2164,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2173,7 +2173,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t float *alpha, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2182,7 +2182,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t double *alpha, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2192,7 +2192,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2202,7 +2202,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2211,7 +2211,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2221,7 +2221,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2232,7 +2232,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2243,7 +2243,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2252,7 +2252,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t sycl::event axpby(backend_selector selector, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2260,7 +2260,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, fl sycl::event axpby(backend_selector selector, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2269,7 +2269,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2278,7 +2278,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2287,7 +2287,7 @@ sycl::event gerc(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2296,7 +2296,7 @@ sycl::event gerc(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2306,7 +2306,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2316,7 +2316,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2327,7 +2327,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2339,7 +2339,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2348,7 +2348,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2357,7 +2357,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2367,7 +2367,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2377,7 +2377,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2388,7 +2388,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::int64_t incx, std::int64_t stridex, float beta, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2400,7 +2400,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::int64_t incx, std::int64_t stridex, double beta, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2413,7 +2413,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2426,7 +2426,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2438,7 +2438,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2450,7 +2450,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2463,7 +2463,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2476,7 +2476,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2488,7 +2488,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2500,7 +2500,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2512,7 +2512,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2524,7 +2524,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2535,7 +2535,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2546,7 +2546,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2558,7 +2558,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2570,7 +2570,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2580,7 +2580,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2589,7 +2589,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2597,7 +2597,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -2605,7 +2605,7 @@ sycl::event hpr(backend_selector selector, uplo upper_lower, st sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -2613,7 +2613,7 @@ sycl::event hpr(backend_selector selector, uplo upper_lower, st sycl::event iamin(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2621,7 +2621,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, co sycl::event iamin(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2629,7 +2629,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, co sycl::event iamin(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2637,7 +2637,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, sycl::event iamin(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2648,7 +2648,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2660,7 +2660,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2673,7 +2673,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2686,7 +2686,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2699,7 +2699,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2711,7 +2711,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2723,7 +2723,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2735,7 +2735,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2748,7 +2748,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2761,7 +2761,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2775,7 +2775,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2789,7 +2789,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2802,7 +2802,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2814,7 +2814,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2826,7 +2826,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2838,7 +2838,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2848,7 +2848,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -2857,7 +2857,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -2865,7 +2865,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s sycl::event swap(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2873,7 +2873,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, flo sycl::event swap(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2881,7 +2881,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, dou sycl::event swap(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2889,7 +2889,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, sycl::event swap(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2898,7 +2898,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2907,7 +2907,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2915,7 +2915,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std sycl::event nrm2(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2923,7 +2923,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, sycl::event nrm2(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2931,7 +2931,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, sycl::event nrm2(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2939,7 +2939,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, con sycl::event nrm2(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2949,7 +2949,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2959,7 +2959,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2970,7 +2970,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2981,7 +2981,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2992,7 +2992,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3003,7 +3003,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3013,7 +3013,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3024,7 +3024,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_bias( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3036,7 +3036,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_bias( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3048,7 +3048,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_bias( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3060,7 +3060,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemm_bias( + auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3070,7 +3070,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::herk( + auto done = oneapi::math::blas::mklcpu::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -3079,7 +3079,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::herk( + auto done = oneapi::math::blas::mklcpu::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -3088,7 +3088,7 @@ sycl::event ger(backend_selector selector, std::int64_t m, std: float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -3097,7 +3097,7 @@ sycl::event ger(backend_selector selector, std::int64_t m, std: double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -3106,7 +3106,7 @@ sycl::event trsm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3116,7 +3116,7 @@ sycl::event trsm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3127,7 +3127,7 @@ sycl::event trsm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3138,7 +3138,7 @@ sycl::event trsm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3150,7 +3150,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3162,7 +3162,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3174,7 +3174,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3187,7 +3187,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3199,7 +3199,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3211,7 +3211,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3224,7 +3224,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3237,7 +3237,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3247,7 +3247,7 @@ sycl::event dotu(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3256,7 +3256,7 @@ sycl::event dotu(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3267,7 +3267,7 @@ sycl::event hemm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3278,7 +3278,7 @@ sycl::event hemm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3287,7 +3287,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3296,7 +3296,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3306,7 +3306,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3317,7 +3317,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st double beta, double *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3328,7 +3328,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3339,7 +3339,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3348,7 +3348,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tbmv( + auto done = oneapi::math::blas::mklcpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3357,7 +3357,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tbmv( + auto done = oneapi::math::blas::mklcpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3366,7 +3366,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tbmv( + auto done = oneapi::math::blas::mklcpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3375,7 +3375,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tbmv( + auto done = oneapi::math::blas::mklcpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3385,7 +3385,7 @@ sycl::event symm(backend_selector selector, side left_right, up const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3395,7 +3395,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3406,7 +3406,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3417,7 +3417,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3426,7 +3426,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3435,7 +3435,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3443,7 +3443,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3451,7 +3451,7 @@ sycl::event syr(backend_selector selector, uplo upper_lower, st sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3460,7 +3460,7 @@ sycl::event trmm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3470,7 +3470,7 @@ sycl::event trmm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3481,7 +3481,7 @@ sycl::event trmm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3492,7 +3492,7 @@ sycl::event trmm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3500,7 +3500,7 @@ sycl::event trmm(backend_selector selector, side left_right, up sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, + auto done = oneapi::math::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } @@ -3508,7 +3508,7 @@ sycl::event rotmg(backend_selector selector, float *d1, float * sycl::event rotmg(backend_selector selector, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, + auto done = oneapi::math::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } @@ -3516,7 +3516,7 @@ sycl::event rotmg(backend_selector selector, double *d1, double sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3524,7 +3524,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3533,7 +3533,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3542,7 +3542,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3550,7 +3550,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3558,7 +3558,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3567,7 +3567,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3576,7 +3576,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3584,7 +3584,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t sycl::event copy(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3592,7 +3592,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, con sycl::event copy(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3600,7 +3600,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, con sycl::event copy(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3608,7 +3608,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, sycl::event copy(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3617,7 +3617,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3626,7 +3626,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3636,7 +3636,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3646,7 +3646,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3655,7 +3655,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3664,7 +3664,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3674,7 +3674,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3684,7 +3684,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3694,7 +3694,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::hemv( + auto done = oneapi::math::blas::mklcpu::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3704,7 +3704,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::hemv( + auto done = oneapi::math::blas::mklcpu::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3714,7 +3714,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3725,7 +3725,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3737,7 +3737,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3749,7 +3749,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3760,7 +3760,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3770,7 +3770,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3778,7 +3778,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s sycl::event asum(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3786,7 +3786,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, sycl::event asum(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3794,7 +3794,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, sycl::event asum(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3802,7 +3802,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, con sycl::event asum(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3811,7 +3811,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tbsv( + auto done = oneapi::math::blas::mklcpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3820,7 +3820,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tbsv( + auto done = oneapi::math::blas::mklcpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3829,7 +3829,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tbsv( + auto done = oneapi::math::blas::mklcpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3838,7 +3838,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::tbsv( + auto done = oneapi::math::blas::mklcpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3847,7 +3847,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3856,7 +3856,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3864,7 +3864,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s sycl::event iamax(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3872,7 +3872,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, co sycl::event iamax(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3880,7 +3880,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, co sycl::event iamax(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3888,7 +3888,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event iamax(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3896,7 +3896,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event rotm(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3904,7 +3904,7 @@ sycl::event rotm(backend_selector selector, std::int64_t n, flo sycl::event rotm(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3912,14 +3912,14 @@ sycl::event rotm(backend_selector selector, std::int64_t n, dou sycl::event rotg(backend_selector selector, float *a, float *b, float *c, float *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event rotg(backend_selector selector, double *a, double *b, double *c, double *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3927,7 +3927,7 @@ sycl::event rotg(backend_selector selector, std::complex std::complex *b, float *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3935,14 +3935,14 @@ sycl::event rotg(backend_selector selector, std::complex *b, double *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event sdsdot(backend_selector selector, std::int64_t n, float sb, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, + auto done = oneapi::math::blas::mklcpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result, dependencies); return done; } @@ -3953,7 +3953,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3965,7 +3965,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3973,7 +3973,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, sycl::event dot(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3981,7 +3981,7 @@ sycl::event dot(backend_selector selector, std::int64_t n, cons sycl::event dot(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3989,7 +3989,7 @@ sycl::event dot(backend_selector selector, std::int64_t n, cons sycl::event dot(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3998,7 +3998,7 @@ sycl::event symv(backend_selector selector, uplo upper_lower, s float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::symv( + auto done = oneapi::math::blas::mklcpu::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -4007,7 +4007,7 @@ sycl::event symv(backend_selector selector, uplo upper_lower, s double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::symv( + auto done = oneapi::math::blas::mklcpu::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -4017,7 +4017,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4028,7 +4028,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4039,7 +4039,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4050,7 +4050,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4060,7 +4060,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4069,7 +4069,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4079,7 +4079,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4089,7 +4089,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4100,7 +4100,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4112,7 +4112,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4125,7 +4125,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4138,7 +4138,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4147,7 +4147,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4155,7 +4155,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4164,7 +4164,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4173,7 +4173,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4182,7 +4182,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy2( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4191,7 +4191,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy2( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4201,7 +4201,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy2( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4211,7 +4211,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy2( + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4219,7 +4219,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4227,7 +4227,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4236,7 +4236,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4245,7 +4245,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4255,7 +4255,7 @@ sycl::event omatadd(backend_selector selector, transpose transa float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4265,7 +4265,7 @@ sycl::event omatadd(backend_selector selector, transpose transa double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4276,7 +4276,7 @@ sycl::event omatadd(backend_selector selector, transpose transa const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4287,7 +4287,7 @@ sycl::event omatadd(backend_selector selector, transpose transa const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4297,7 +4297,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4308,7 +4308,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4320,7 +4320,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4332,7 +4332,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4343,7 +4343,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4354,7 +4354,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4365,7 +4365,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4376,7 +4376,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; diff --git a/include/oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp b/include/oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp index 1c6eefe36..bb82eaa81 100644 --- a/include/oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp +++ b/include/oneapi/math/blas/detail/mklcpu/onemath_blas_mklcpu.hpp @@ -30,12 +30,12 @@ #include "oneapi/math/detail/export.hpp" namespace oneapi { -namespace mkl { -using oneapi::mkl::transpose; -using oneapi::mkl::uplo; -using oneapi::mkl::side; -using oneapi::mkl::diag; -using oneapi::mkl::offset; +namespace math { +using oneapi::math::transpose; +using oneapi::math::uplo; +using oneapi::math::side; +using oneapi::math::diag; +using oneapi::math::offset; namespace blas { namespace mklcpu { namespace column_major { @@ -50,7 +50,7 @@ namespace row_major { } // namespace row_major } // namespace mklcpu } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_BLAS_MKLCPU_HPP_ diff --git a/include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp b/include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp index f767a1678..dac4fba96 100644 --- a/include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp @@ -35,7 +35,7 @@ #include "oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace column_major { @@ -52,7 +52,7 @@ namespace row_major { } //namespace row_major } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_DETAIL_MKLGPU_BLAS_CT_HPP_ diff --git a/include/oneapi/math/blas/detail/mklgpu/blas_ct.hxx b/include/oneapi/math/blas/detail/mklgpu/blas_ct.hxx index c69257e9c..b00317329 100644 --- a/include/oneapi/math/blas/detail/mklgpu/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/mklgpu/blas_ct.hxx @@ -23,7 +23,7 @@ void herk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -31,104 +31,104 @@ void herk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void scal(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, @@ -137,7 +137,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -148,7 +148,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -160,7 +160,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -172,7 +172,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -183,7 +183,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -194,7 +194,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -205,7 +205,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -216,7 +216,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -224,14 +224,14 @@ void gemm_batch(backend_selector selector, transpose transa, tr void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -239,7 +239,7 @@ void syrk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -247,7 +247,7 @@ void syrk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -255,7 +255,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -265,7 +265,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -276,7 +276,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -287,7 +287,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -296,7 +296,7 @@ void her2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -304,7 +304,7 @@ void her2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -313,7 +313,7 @@ void hbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -322,61 +322,61 @@ void hbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void rot(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { - oneapi::mkl::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { - oneapi::mkl::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { - oneapi::mkl::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { - oneapi::mkl::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -384,7 +384,7 @@ void axpy_batch(backend_selector selector, std::int64_t n, doub sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -393,7 +393,7 @@ void axpy_batch(backend_selector selector, std::int64_t n, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -402,46 +402,46 @@ void axpy_batch(backend_selector selector, std::int64_t n, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } void axpby(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void sdsdot(backend_selector selector, std::int64_t n, float sb, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); + oneapi::math::blas::mklgpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -449,7 +449,7 @@ void gerc(backend_selector selector, std::int64_t m, std::int64 std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -457,7 +457,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -465,7 +465,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -474,7 +474,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -484,7 +484,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -492,7 +492,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -500,7 +500,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -509,7 +509,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -518,7 +518,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -527,7 +527,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -537,7 +537,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -549,7 +549,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -561,7 +561,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t incx, std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -571,7 +571,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -581,7 +581,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -592,7 +592,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -603,7 +603,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -611,58 +611,58 @@ void dgmm_batch(backend_selector selector, side left_right, std void her(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -671,21 +671,21 @@ void hpmv(backend_selector selector, uplo upper_lower, std::int sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -695,7 +695,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -705,7 +705,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -715,7 +715,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -725,38 +725,38 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -764,37 +764,37 @@ void geru(backend_selector selector, std::int64_t m, std::int64 std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -802,7 +802,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -811,7 +811,7 @@ void gemm(backend_selector selector, transpose transa, transpos sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -820,7 +820,7 @@ void gemm(backend_selector selector, transpose transa, transpos sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -828,7 +828,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -836,7 +836,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -844,14 +844,14 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -859,21 +859,21 @@ void syr2(backend_selector selector, uplo upper_lower, std::int double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -881,7 +881,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -889,7 +889,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -897,7 +897,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -905,7 +905,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -913,14 +913,14 @@ void dotu(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, @@ -928,7 +928,7 @@ void hemm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -937,7 +937,7 @@ void hemm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -945,7 +945,7 @@ void hpr2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -953,7 +953,7 @@ void hpr2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -961,7 +961,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -969,7 +969,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -978,7 +978,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -988,21 +988,21 @@ void gbmv(backend_selector selector, transpose trans, std::int6 sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1010,7 +1010,7 @@ void tbmv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1018,7 +1018,7 @@ void tbmv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1026,7 +1026,7 @@ void symm(backend_selector selector, side left_right, uplo uppe std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1034,7 +1034,7 @@ void symm(backend_selector selector, side left_right, uplo uppe std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1043,7 +1043,7 @@ void symm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1052,7 +1052,7 @@ void symm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1060,27 +1060,27 @@ void dotc(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } @@ -1088,7 +1088,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1096,7 +1096,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1104,7 +1104,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1112,106 +1112,106 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m) { - oneapi::mkl::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m) { - oneapi::mkl::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1219,7 +1219,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1227,7 +1227,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1235,7 +1235,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1243,7 +1243,7 @@ void hemv(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::mklgpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1252,7 +1252,7 @@ void hemv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::mklgpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1260,7 +1260,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1268,7 +1268,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1277,7 +1277,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1287,37 +1287,37 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -1325,21 +1325,21 @@ void sbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1347,7 +1347,7 @@ void tbsv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1355,76 +1355,76 @@ void tbsv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, @@ -1432,7 +1432,7 @@ void trsm_batch(backend_selector selector, side left_right, upl sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1442,7 +1442,7 @@ void trsm_batch(backend_selector selector, side left_right, upl sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1453,7 +1453,7 @@ void trsm_batch(backend_selector selector, side left_right, upl std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1464,7 +1464,7 @@ void trsm_batch(backend_selector selector, side left_right, upl std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1474,7 +1474,7 @@ void her2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1483,38 +1483,38 @@ void her2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::mklgpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1522,7 +1522,7 @@ void symv(backend_selector selector, uplo upper_lower, std::int double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::mklgpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::mklgpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1530,7 +1530,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1538,7 +1538,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1547,7 +1547,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1556,21 +1556,21 @@ void omatcopy_batch(backend_selector selector, transpose trans, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1578,7 +1578,7 @@ void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1586,7 +1586,7 @@ void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1595,7 +1595,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1605,7 +1605,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1616,7 +1616,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1628,7 +1628,7 @@ void omatadd_batch(backend_selector selector, transpose transa, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1636,28 +1636,28 @@ void omatadd_batch(backend_selector selector, transpose transa, void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } @@ -1665,7 +1665,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1673,7 +1673,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1681,7 +1681,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1689,35 +1689,35 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } @@ -1725,7 +1725,7 @@ void omatadd(backend_selector selector, transpose transa, trans std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1733,7 +1733,7 @@ void omatadd(backend_selector selector, transpose transa, trans std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1742,7 +1742,7 @@ void omatadd(backend_selector selector, transpose transa, trans sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1751,7 +1751,7 @@ void omatadd(backend_selector selector, transpose transa, trans sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1761,7 +1761,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1770,7 +1770,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1778,7 +1778,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s sycl::event scal(backend_selector selector, std::int64_t n, float alpha, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1786,7 +1786,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, flo sycl::event scal(backend_selector selector, std::int64_t n, double alpha, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1794,7 +1794,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, dou sycl::event scal(backend_selector selector, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1802,7 +1802,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, sycl::event scal(backend_selector selector, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1810,7 +1810,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, sycl::event scal(backend_selector selector, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1818,7 +1818,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, flo sycl::event scal(backend_selector selector, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1826,7 +1826,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, dou sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1834,7 +1834,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1843,7 +1843,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1852,7 +1852,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1860,7 +1860,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1868,7 +1868,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1877,7 +1877,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1886,7 +1886,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1894,7 +1894,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1902,7 +1902,7 @@ sycl::event spr(backend_selector selector, uplo upper_lower, st sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1912,7 +1912,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1922,7 +1922,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1931,7 +1931,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1940,7 +1940,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1950,7 +1950,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1960,7 +1960,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1970,7 +1970,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1981,7 +1981,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1993,7 +1993,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2005,7 +2005,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2017,7 +2017,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2029,7 +2029,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2042,7 +2042,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2055,7 +2055,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syrk_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2065,7 +2065,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2074,7 +2074,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2085,7 +2085,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, s std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2096,7 +2096,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, s std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2105,7 +2105,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2114,7 +2114,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2122,7 +2122,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, sycl::event rot(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2130,7 +2130,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, floa sycl::event rot(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2138,7 +2138,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, doub sycl::event axpy(backend_selector selector, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2146,7 +2146,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, flo sycl::event axpy(backend_selector selector, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2155,7 +2155,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2164,7 +2164,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2173,7 +2173,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t float *alpha, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2182,7 +2182,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t double *alpha, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2192,7 +2192,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2202,7 +2202,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2211,7 +2211,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2221,7 +2221,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2232,7 +2232,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2243,7 +2243,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2252,7 +2252,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t sycl::event axpby(backend_selector selector, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2260,7 +2260,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, fl sycl::event axpby(backend_selector selector, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2269,7 +2269,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2278,7 +2278,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2287,7 +2287,7 @@ sycl::event gerc(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2296,7 +2296,7 @@ sycl::event gerc(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2306,7 +2306,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2316,7 +2316,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2327,7 +2327,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2339,7 +2339,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2348,7 +2348,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2357,7 +2357,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2367,7 +2367,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2377,7 +2377,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2388,7 +2388,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::int64_t incx, std::int64_t stridex, float beta, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2400,7 +2400,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::int64_t incx, std::int64_t stridex, double beta, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2413,7 +2413,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2426,7 +2426,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2438,7 +2438,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2450,7 +2450,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2463,7 +2463,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2476,7 +2476,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2488,7 +2488,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2500,7 +2500,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2512,7 +2512,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2524,7 +2524,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2535,7 +2535,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2546,7 +2546,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2558,7 +2558,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2570,7 +2570,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2580,7 +2580,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2589,7 +2589,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2597,7 +2597,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -2605,7 +2605,7 @@ sycl::event hpr(backend_selector selector, uplo upper_lower, st sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -2613,7 +2613,7 @@ sycl::event hpr(backend_selector selector, uplo upper_lower, st sycl::event iamin(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2621,7 +2621,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, co sycl::event iamin(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2629,7 +2629,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, co sycl::event iamin(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2637,7 +2637,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, sycl::event iamin(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2649,7 +2649,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2661,7 +2661,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2673,7 +2673,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2685,7 +2685,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2697,7 +2697,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2709,7 +2709,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2722,7 +2722,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2735,7 +2735,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2748,7 +2748,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2760,7 +2760,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2772,7 +2772,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2784,7 +2784,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2797,7 +2797,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2810,7 +2810,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2824,7 +2824,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2838,7 +2838,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2848,7 +2848,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -2857,7 +2857,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -2865,7 +2865,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s sycl::event swap(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2873,7 +2873,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, flo sycl::event swap(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2881,7 +2881,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, dou sycl::event swap(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2889,7 +2889,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, sycl::event swap(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2898,7 +2898,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2907,7 +2907,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2915,7 +2915,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std sycl::event nrm2(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2923,7 +2923,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, sycl::event nrm2(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2931,7 +2931,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, sycl::event nrm2(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2939,7 +2939,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, con sycl::event nrm2(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2949,7 +2949,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2959,7 +2959,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2970,7 +2970,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2981,7 +2981,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2992,7 +2992,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3003,7 +3003,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3013,7 +3013,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3024,7 +3024,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_bias( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3036,7 +3036,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_bias( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3048,7 +3048,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_bias( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3060,7 +3060,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemm_bias( + auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3070,7 +3070,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::herk( + auto done = oneapi::math::blas::mklgpu::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -3079,7 +3079,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::herk( + auto done = oneapi::math::blas::mklgpu::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -3088,7 +3088,7 @@ sycl::event ger(backend_selector selector, std::int64_t m, std: float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -3097,7 +3097,7 @@ sycl::event ger(backend_selector selector, std::int64_t m, std: double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, + auto done = oneapi::math::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -3106,7 +3106,7 @@ sycl::event trsm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3116,7 +3116,7 @@ sycl::event trsm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3127,7 +3127,7 @@ sycl::event trsm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3138,7 +3138,7 @@ sycl::event trsm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3150,7 +3150,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3162,7 +3162,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3174,7 +3174,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3187,7 +3187,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3199,7 +3199,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3211,7 +3211,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3224,7 +3224,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3237,7 +3237,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsm_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3247,7 +3247,7 @@ sycl::event dotu(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3256,7 +3256,7 @@ sycl::event dotu(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3267,7 +3267,7 @@ sycl::event hemm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3278,7 +3278,7 @@ sycl::event hemm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3287,7 +3287,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3296,7 +3296,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3306,7 +3306,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3317,7 +3317,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st double beta, double *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3328,7 +3328,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3339,7 +3339,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3348,7 +3348,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tbmv( + auto done = oneapi::math::blas::mklgpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3357,7 +3357,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tbmv( + auto done = oneapi::math::blas::mklgpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3366,7 +3366,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tbmv( + auto done = oneapi::math::blas::mklgpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3375,7 +3375,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tbmv( + auto done = oneapi::math::blas::mklgpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3385,7 +3385,7 @@ sycl::event symm(backend_selector selector, side left_right, up const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3395,7 +3395,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3406,7 +3406,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3417,7 +3417,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3426,7 +3426,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3435,7 +3435,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3443,7 +3443,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3451,7 +3451,7 @@ sycl::event syr(backend_selector selector, uplo upper_lower, st sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3460,7 +3460,7 @@ sycl::event trmm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3470,7 +3470,7 @@ sycl::event trmm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3481,7 +3481,7 @@ sycl::event trmm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3492,7 +3492,7 @@ sycl::event trmm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3500,7 +3500,7 @@ sycl::event trmm(backend_selector selector, side left_right, up sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, + auto done = oneapi::math::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } @@ -3508,7 +3508,7 @@ sycl::event rotmg(backend_selector selector, float *d1, float * sycl::event rotmg(backend_selector selector, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, + auto done = oneapi::math::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } @@ -3516,7 +3516,7 @@ sycl::event rotmg(backend_selector selector, double *d1, double sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3524,7 +3524,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3533,7 +3533,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3542,7 +3542,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3550,7 +3550,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3558,7 +3558,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3567,7 +3567,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3576,7 +3576,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3584,7 +3584,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t sycl::event copy(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3592,7 +3592,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, con sycl::event copy(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3600,7 +3600,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, con sycl::event copy(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3608,7 +3608,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, sycl::event copy(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3617,7 +3617,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3626,7 +3626,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3636,7 +3636,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3646,7 +3646,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3655,7 +3655,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3664,7 +3664,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3674,7 +3674,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3684,7 +3684,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::copy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3694,7 +3694,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::hemv( + auto done = oneapi::math::blas::mklgpu::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3704,7 +3704,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::hemv( + auto done = oneapi::math::blas::mklgpu::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3714,7 +3714,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3725,7 +3725,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3737,7 +3737,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3749,7 +3749,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3760,7 +3760,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3770,7 +3770,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3778,7 +3778,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s sycl::event asum(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3786,7 +3786,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, sycl::event asum(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3794,7 +3794,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, sycl::event asum(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3802,7 +3802,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, con sycl::event asum(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3811,7 +3811,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tbsv( + auto done = oneapi::math::blas::mklgpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3820,7 +3820,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tbsv( + auto done = oneapi::math::blas::mklgpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3829,7 +3829,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tbsv( + auto done = oneapi::math::blas::mklgpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3838,7 +3838,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::tbsv( + auto done = oneapi::math::blas::mklgpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3847,7 +3847,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3856,7 +3856,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3864,7 +3864,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s sycl::event iamax(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3872,7 +3872,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, co sycl::event iamax(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3880,7 +3880,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, co sycl::event iamax(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3888,7 +3888,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event iamax(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3896,7 +3896,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event rotm(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3904,7 +3904,7 @@ sycl::event rotm(backend_selector selector, std::int64_t n, flo sycl::event rotm(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3912,14 +3912,14 @@ sycl::event rotm(backend_selector selector, std::int64_t n, dou sycl::event rotg(backend_selector selector, float *a, float *b, float *c, float *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event rotg(backend_selector selector, double *a, double *b, double *c, double *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3927,7 +3927,7 @@ sycl::event rotg(backend_selector selector, std::complex std::complex *b, float *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3935,14 +3935,14 @@ sycl::event rotg(backend_selector selector, std::complex *b, double *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event sdsdot(backend_selector selector, std::int64_t n, float sb, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, + auto done = oneapi::math::blas::mklgpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result, dependencies); return done; } @@ -3953,7 +3953,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3965,7 +3965,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3973,7 +3973,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, sycl::event dot(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3981,7 +3981,7 @@ sycl::event dot(backend_selector selector, std::int64_t n, cons sycl::event dot(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3989,7 +3989,7 @@ sycl::event dot(backend_selector selector, std::int64_t n, cons sycl::event dot(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3998,7 +3998,7 @@ sycl::event symv(backend_selector selector, uplo upper_lower, s float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::symv( + auto done = oneapi::math::blas::mklgpu::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -4007,7 +4007,7 @@ sycl::event symv(backend_selector selector, uplo upper_lower, s double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::symv( + auto done = oneapi::math::blas::mklgpu::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -4017,7 +4017,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4028,7 +4028,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4039,7 +4039,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4050,7 +4050,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4060,7 +4060,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4069,7 +4069,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4079,7 +4079,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4089,7 +4089,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4100,7 +4100,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4112,7 +4112,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4125,7 +4125,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4138,7 +4138,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4147,7 +4147,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4155,7 +4155,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4164,7 +4164,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4173,7 +4173,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4182,7 +4182,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy2( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4191,7 +4191,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy2( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4201,7 +4201,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy2( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4211,7 +4211,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy2( + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4219,7 +4219,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4227,7 +4227,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4236,7 +4236,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4245,7 +4245,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4255,7 +4255,7 @@ sycl::event omatadd(backend_selector selector, transpose transa float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4265,7 +4265,7 @@ sycl::event omatadd(backend_selector selector, transpose transa double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4276,7 +4276,7 @@ sycl::event omatadd(backend_selector selector, transpose transa const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4287,7 +4287,7 @@ sycl::event omatadd(backend_selector selector, transpose transa const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4297,7 +4297,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4308,7 +4308,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4320,7 +4320,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4332,7 +4332,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4343,7 +4343,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4354,7 +4354,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4365,7 +4365,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4376,7 +4376,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; diff --git a/include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp b/include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp index 6624a79a3..636e89d4b 100644 --- a/include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp +++ b/include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp @@ -33,7 +33,7 @@ #include "oneapi/math/detail/export.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace mklgpu { namespace column_major { @@ -48,7 +48,7 @@ namespace row_major { } //namespace row_major } //namespace mklgpu } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_ONEMATH_BLAS_MKLGPU_HPP_ diff --git a/include/oneapi/math/blas/detail/netlib/blas_ct.hpp b/include/oneapi/math/blas/detail/netlib/blas_ct.hpp index 7c8b62405..5e4b4d727 100644 --- a/include/oneapi/math/blas/detail/netlib/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/netlib/blas_ct.hpp @@ -34,7 +34,7 @@ #include "oneapi/math/blas/detail/blas_ct_backends.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace column_major { @@ -51,7 +51,7 @@ namespace row_major { } //namespace row_major } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_DETAIL_NETLIB_BLAS_CT_HPP_ diff --git a/include/oneapi/math/blas/detail/netlib/blas_ct.hxx b/include/oneapi/math/blas/detail/netlib/blas_ct.hxx index 404d79ae0..1e55eaa94 100644 --- a/include/oneapi/math/blas/detail/netlib/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/netlib/blas_ct.hxx @@ -23,7 +23,7 @@ void herk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -31,104 +31,104 @@ void herk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void scal(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, @@ -137,7 +137,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -148,7 +148,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -160,7 +160,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -172,7 +172,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -183,7 +183,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -194,7 +194,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -205,7 +205,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -216,7 +216,7 @@ void gemm_batch(backend_selector selector, transpose transa, tr sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -224,14 +224,14 @@ void gemm_batch(backend_selector selector, transpose transa, tr void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -239,7 +239,7 @@ void syrk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -247,7 +247,7 @@ void syrk(backend_selector selector, uplo upper_lower, transpos std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -255,7 +255,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -265,7 +265,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -276,7 +276,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -287,7 +287,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, tr std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -296,7 +296,7 @@ void her2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -304,7 +304,7 @@ void her2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -313,7 +313,7 @@ void hbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -322,61 +322,61 @@ void hbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void rot(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { - oneapi::mkl::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { - oneapi::mkl::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { - oneapi::mkl::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { - oneapi::mkl::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -384,7 +384,7 @@ void axpy_batch(backend_selector selector, std::int64_t n, doub sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -393,7 +393,7 @@ void axpy_batch(backend_selector selector, std::int64_t n, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -402,46 +402,46 @@ void axpy_batch(backend_selector selector, std::int64_t n, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } void axpby(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void sdsdot(backend_selector selector, std::int64_t n, float sb, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); + oneapi::math::blas::netlib::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -449,7 +449,7 @@ void gerc(backend_selector selector, std::int64_t m, std::int64 std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -457,7 +457,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -465,7 +465,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -474,7 +474,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -484,7 +484,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -492,7 +492,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -500,7 +500,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -509,7 +509,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -518,7 +518,7 @@ void gemv(backend_selector selector, transpose trans, std::int6 std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -527,7 +527,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -537,7 +537,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -549,7 +549,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -561,7 +561,7 @@ void gemv_batch(backend_selector selector, transpose trans, std std::int64_t incx, std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -571,7 +571,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -581,7 +581,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -592,7 +592,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -603,7 +603,7 @@ void dgmm_batch(backend_selector selector, side left_right, std std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -611,58 +611,58 @@ void dgmm_batch(backend_selector selector, side left_right, std void her(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -671,21 +671,21 @@ void hpmv(backend_selector selector, uplo upper_lower, std::int sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -695,7 +695,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -705,7 +705,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -715,7 +715,7 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -725,38 +725,38 @@ void gemm_bias(backend_selector selector, transpose transa, tra sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, + oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -764,37 +764,37 @@ void geru(backend_selector selector, std::int64_t m, std::int64 std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -802,7 +802,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -811,7 +811,7 @@ void gemm(backend_selector selector, transpose transa, transpos sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -820,7 +820,7 @@ void gemm(backend_selector selector, transpose transa, transpos sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -828,7 +828,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -836,7 +836,7 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -844,14 +844,14 @@ void gemm(backend_selector selector, transpose transa, transpos std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -859,21 +859,21 @@ void syr2(backend_selector selector, uplo upper_lower, std::int double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -881,7 +881,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -889,7 +889,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -897,7 +897,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -905,7 +905,7 @@ void trsm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -913,14 +913,14 @@ void dotu(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, @@ -928,7 +928,7 @@ void hemm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -937,7 +937,7 @@ void hemm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -945,7 +945,7 @@ void hpr2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -953,7 +953,7 @@ void hpr2(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -961,7 +961,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -969,7 +969,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -978,7 +978,7 @@ void gbmv(backend_selector selector, transpose trans, std::int6 sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -988,21 +988,21 @@ void gbmv(backend_selector selector, transpose trans, std::int6 sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1010,7 +1010,7 @@ void tbmv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1018,7 +1018,7 @@ void tbmv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1026,7 +1026,7 @@ void symm(backend_selector selector, side left_right, uplo uppe std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1034,7 +1034,7 @@ void symm(backend_selector selector, side left_right, uplo uppe std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1043,7 +1043,7 @@ void symm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1052,7 +1052,7 @@ void symm(backend_selector selector, side left_right, uplo uppe sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1060,27 +1060,27 @@ void dotc(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } @@ -1088,7 +1088,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1096,7 +1096,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1104,7 +1104,7 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1112,106 +1112,106 @@ void trmm(backend_selector selector, side left_right, uplo uppe transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m) { - oneapi::mkl::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m) { - oneapi::mkl::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1219,7 +1219,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1227,7 +1227,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1235,7 +1235,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, + oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1243,7 +1243,7 @@ void hemv(backend_selector selector, uplo upper_lower, std::int std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::netlib::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1252,7 +1252,7 @@ void hemv(backend_selector selector, uplo upper_lower, std::int std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::netlib::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1260,7 +1260,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1268,7 +1268,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1277,7 +1277,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1287,37 +1287,37 @@ void gemmt(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, + oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -1325,21 +1325,21 @@ void sbmv(backend_selector selector, uplo upper_lower, std::int std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1347,7 +1347,7 @@ void tbsv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1355,76 +1355,76 @@ void tbsv(backend_selector selector, uplo upper_lower, transpos diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, @@ -1432,7 +1432,7 @@ void trsm_batch(backend_selector selector, side left_right, upl sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1442,7 +1442,7 @@ void trsm_batch(backend_selector selector, side left_right, upl sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1453,7 +1453,7 @@ void trsm_batch(backend_selector selector, side left_right, upl std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1464,7 +1464,7 @@ void trsm_batch(backend_selector selector, side left_right, upl std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1474,7 +1474,7 @@ void her2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1483,38 +1483,38 @@ void her2k(backend_selector selector, uplo upper_lower, transpo sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::netlib::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1522,7 +1522,7 @@ void symv(backend_selector selector, uplo upper_lower, std::int double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::netlib::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::netlib::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1530,7 +1530,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1538,7 +1538,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1547,7 +1547,7 @@ void omatcopy_batch(backend_selector selector, transpose trans, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1556,21 +1556,21 @@ void omatcopy_batch(backend_selector selector, transpose trans, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1578,7 +1578,7 @@ void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1586,7 +1586,7 @@ void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1595,7 +1595,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1605,7 +1605,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1616,7 +1616,7 @@ void omatadd_batch(backend_selector selector, transpose transa, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1628,7 +1628,7 @@ void omatadd_batch(backend_selector selector, transpose transa, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1636,28 +1636,28 @@ void omatadd_batch(backend_selector selector, transpose transa, void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } @@ -1665,7 +1665,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1673,7 +1673,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1681,7 +1681,7 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1689,35 +1689,35 @@ void omatcopy2(backend_selector selector, transpose trans, std: std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } @@ -1725,7 +1725,7 @@ void omatadd(backend_selector selector, transpose transa, trans std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1733,7 +1733,7 @@ void omatadd(backend_selector selector, transpose transa, trans std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1742,7 +1742,7 @@ void omatadd(backend_selector selector, transpose transa, trans sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1751,7 +1751,7 @@ void omatadd(backend_selector selector, transpose transa, trans sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1761,7 +1761,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1770,7 +1770,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1778,7 +1778,7 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, s sycl::event scal(backend_selector selector, std::int64_t n, float alpha, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1786,7 +1786,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, flo sycl::event scal(backend_selector selector, std::int64_t n, double alpha, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1794,7 +1794,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, dou sycl::event scal(backend_selector selector, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1802,7 +1802,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, sycl::event scal(backend_selector selector, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1810,7 +1810,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, sycl::event scal(backend_selector selector, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1818,7 +1818,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, flo sycl::event scal(backend_selector selector, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1826,7 +1826,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, dou sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1834,7 +1834,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1843,7 +1843,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1852,7 +1852,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1860,7 +1860,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, t sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1868,7 +1868,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1877,7 +1877,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1886,7 +1886,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1894,7 +1894,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, t sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1902,7 +1902,7 @@ sycl::event spr(backend_selector selector, uplo upper_lower, st sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1912,7 +1912,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1922,7 +1922,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1931,7 +1931,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk( + auto done = oneapi::math::blas::netlib::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1940,7 +1940,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk( + auto done = oneapi::math::blas::netlib::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1950,7 +1950,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk( + auto done = oneapi::math::blas::netlib::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1960,7 +1960,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, t const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk( + auto done = oneapi::math::blas::netlib::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1970,7 +1970,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk_batch( + auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1981,7 +1981,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk_batch( + auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1993,7 +1993,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk_batch( + auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2005,7 +2005,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk_batch( + auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2017,7 +2017,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk_batch( + auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2029,7 +2029,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk_batch( + auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2042,7 +2042,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk_batch( + auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2055,7 +2055,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syrk_batch( + auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2065,7 +2065,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2074,7 +2074,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2085,7 +2085,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, s std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2096,7 +2096,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, s std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2105,7 +2105,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2114,7 +2114,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2122,7 +2122,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, sycl::event rot(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2130,7 +2130,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, floa sycl::event rot(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2138,7 +2138,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, doub sycl::event axpy(backend_selector selector, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2146,7 +2146,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, flo sycl::event axpy(backend_selector selector, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2155,7 +2155,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2164,7 +2164,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2173,7 +2173,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t float *alpha, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2182,7 +2182,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t double *alpha, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2192,7 +2192,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2202,7 +2202,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2211,7 +2211,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2221,7 +2221,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2232,7 +2232,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2243,7 +2243,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2252,7 +2252,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_t sycl::event axpby(backend_selector selector, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2260,7 +2260,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, fl sycl::event axpby(backend_selector selector, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2269,7 +2269,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2278,7 +2278,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2287,7 +2287,7 @@ sycl::event gerc(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2296,7 +2296,7 @@ sycl::event gerc(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2306,7 +2306,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2316,7 +2316,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2327,7 +2327,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2339,7 +2339,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2348,7 +2348,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2357,7 +2357,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2367,7 +2367,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2377,7 +2377,7 @@ sycl::event gemv(backend_selector selector, transpose trans, st std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2388,7 +2388,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::int64_t incx, std::int64_t stridex, float beta, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2400,7 +2400,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::int64_t incx, std::int64_t stridex, double beta, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2413,7 +2413,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2426,7 +2426,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tra std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2438,7 +2438,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2450,7 +2450,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2463,7 +2463,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2476,7 +2476,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2488,7 +2488,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2500,7 +2500,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2512,7 +2512,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2524,7 +2524,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_rig std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2535,7 +2535,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2546,7 +2546,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2558,7 +2558,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2570,7 +2570,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ri std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2580,7 +2580,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2589,7 +2589,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2597,7 +2597,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, st sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -2605,7 +2605,7 @@ sycl::event hpr(backend_selector selector, uplo upper_lower, st sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -2613,7 +2613,7 @@ sycl::event hpr(backend_selector selector, uplo upper_lower, st sycl::event iamin(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2621,7 +2621,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, co sycl::event iamin(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2629,7 +2629,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, co sycl::event iamin(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2637,7 +2637,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, sycl::event iamin(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2648,7 +2648,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2660,7 +2660,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2673,7 +2673,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2686,7 +2686,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2699,7 +2699,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2711,7 +2711,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2723,7 +2723,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2735,7 +2735,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2748,7 +2748,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2761,7 +2761,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2775,7 +2775,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2789,7 +2789,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2802,7 +2802,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2814,7 +2814,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2826,7 +2826,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2838,7 +2838,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tra const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2848,7 +2848,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -2857,7 +2857,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -2865,7 +2865,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, s sycl::event swap(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2873,7 +2873,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, flo sycl::event swap(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2881,7 +2881,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, dou sycl::event swap(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2889,7 +2889,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, sycl::event swap(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2898,7 +2898,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2907,7 +2907,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2915,7 +2915,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, std sycl::event nrm2(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2923,7 +2923,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, sycl::event nrm2(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2931,7 +2931,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, sycl::event nrm2(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2939,7 +2939,7 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, con sycl::event nrm2(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2949,7 +2949,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2959,7 +2959,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2970,7 +2970,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2981,7 +2981,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2992,7 +2992,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3003,7 +3003,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3013,7 +3013,7 @@ sycl::event gemm(backend_selector selector, transpose transa, t std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3024,7 +3024,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_bias( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3036,7 +3036,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_bias( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3048,7 +3048,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_bias( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3060,7 +3060,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tran std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemm_bias( + auto done = oneapi::math::blas::netlib::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3070,7 +3070,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::herk( + auto done = oneapi::math::blas::netlib::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -3079,7 +3079,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, t std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::herk( + auto done = oneapi::math::blas::netlib::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -3088,7 +3088,7 @@ sycl::event ger(backend_selector selector, std::int64_t m, std: float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, + auto done = oneapi::math::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -3097,7 +3097,7 @@ sycl::event ger(backend_selector selector, std::int64_t m, std: double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, + auto done = oneapi::math::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -3106,7 +3106,7 @@ sycl::event trsm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3116,7 +3116,7 @@ sycl::event trsm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3127,7 +3127,7 @@ sycl::event trsm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3138,7 +3138,7 @@ sycl::event trsm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3150,7 +3150,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3162,7 +3162,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3174,7 +3174,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3187,7 +3187,7 @@ sycl::event trsm_batch(backend_selector selector, side left_rig std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3199,7 +3199,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3211,7 +3211,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3224,7 +3224,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3237,7 +3237,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsm_batch( + auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3247,7 +3247,7 @@ sycl::event dotu(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3256,7 +3256,7 @@ sycl::event dotu(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3267,7 +3267,7 @@ sycl::event hemm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3278,7 +3278,7 @@ sycl::event hemm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3287,7 +3287,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3296,7 +3296,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3306,7 +3306,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3317,7 +3317,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st double beta, double *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3328,7 +3328,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3339,7 +3339,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, st std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3348,7 +3348,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tbmv( + auto done = oneapi::math::blas::netlib::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3357,7 +3357,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tbmv( + auto done = oneapi::math::blas::netlib::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3366,7 +3366,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tbmv( + auto done = oneapi::math::blas::netlib::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3375,7 +3375,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tbmv( + auto done = oneapi::math::blas::netlib::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3385,7 +3385,7 @@ sycl::event symm(backend_selector selector, side left_right, up const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3395,7 +3395,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3406,7 +3406,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3417,7 +3417,7 @@ sycl::event symm(backend_selector selector, side left_right, up std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3426,7 +3426,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3435,7 +3435,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3443,7 +3443,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3451,7 +3451,7 @@ sycl::event syr(backend_selector selector, uplo upper_lower, st sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3460,7 +3460,7 @@ sycl::event trmm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3470,7 +3470,7 @@ sycl::event trmm(backend_selector selector, side left_right, up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3481,7 +3481,7 @@ sycl::event trmm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3492,7 +3492,7 @@ sycl::event trmm(backend_selector selector, side left_right, up std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3500,7 +3500,7 @@ sycl::event trmm(backend_selector selector, side left_right, up sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, + auto done = oneapi::math::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } @@ -3508,7 +3508,7 @@ sycl::event rotmg(backend_selector selector, float *d1, float * sycl::event rotmg(backend_selector selector, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, + auto done = oneapi::math::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } @@ -3516,7 +3516,7 @@ sycl::event rotmg(backend_selector selector, double *d1, double sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3524,7 +3524,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3533,7 +3533,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3542,7 +3542,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3550,7 +3550,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, t sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3558,7 +3558,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3567,7 +3567,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3576,7 +3576,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3584,7 +3584,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, t sycl::event copy(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3592,7 +3592,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, con sycl::event copy(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3600,7 +3600,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, con sycl::event copy(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3608,7 +3608,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, sycl::event copy(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3617,7 +3617,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3626,7 +3626,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3636,7 +3636,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3646,7 +3646,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3655,7 +3655,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3664,7 +3664,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3674,7 +3674,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3684,7 +3684,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_t std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::copy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3694,7 +3694,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::hemv( + auto done = oneapi::math::blas::netlib::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3704,7 +3704,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, s const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::hemv( + auto done = oneapi::math::blas::netlib::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3714,7 +3714,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3725,7 +3725,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3737,7 +3737,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3749,7 +3749,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3760,7 +3760,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3770,7 +3770,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3778,7 +3778,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, s sycl::event asum(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3786,7 +3786,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, sycl::event asum(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3794,7 +3794,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, sycl::event asum(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3802,7 +3802,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, con sycl::event asum(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3811,7 +3811,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tbsv( + auto done = oneapi::math::blas::netlib::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3820,7 +3820,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tbsv( + auto done = oneapi::math::blas::netlib::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3829,7 +3829,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tbsv( + auto done = oneapi::math::blas::netlib::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3838,7 +3838,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, t diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::tbsv( + auto done = oneapi::math::blas::netlib::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3847,7 +3847,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3856,7 +3856,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3864,7 +3864,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, s sycl::event iamax(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3872,7 +3872,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, co sycl::event iamax(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3880,7 +3880,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, co sycl::event iamax(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3888,7 +3888,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event iamax(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3896,7 +3896,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event rotm(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3904,7 +3904,7 @@ sycl::event rotm(backend_selector selector, std::int64_t n, flo sycl::event rotm(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3912,14 +3912,14 @@ sycl::event rotm(backend_selector selector, std::int64_t n, dou sycl::event rotg(backend_selector selector, float *a, float *b, float *c, float *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event rotg(backend_selector selector, double *a, double *b, double *c, double *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3927,7 +3927,7 @@ sycl::event rotg(backend_selector selector, std::complex std::complex *b, float *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3935,14 +3935,14 @@ sycl::event rotg(backend_selector selector, std::complex *b, double *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event sdsdot(backend_selector selector, std::int64_t n, float sb, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, + auto done = oneapi::math::blas::netlib::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result, dependencies); return done; } @@ -3953,7 +3953,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3965,7 +3965,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3973,7 +3973,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, sycl::event dot(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3981,7 +3981,7 @@ sycl::event dot(backend_selector selector, std::int64_t n, cons sycl::event dot(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3989,7 +3989,7 @@ sycl::event dot(backend_selector selector, std::int64_t n, cons sycl::event dot(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3998,7 +3998,7 @@ sycl::event symv(backend_selector selector, uplo upper_lower, s float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::symv( + auto done = oneapi::math::blas::netlib::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -4007,7 +4007,7 @@ sycl::event symv(backend_selector selector, uplo upper_lower, s double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::symv( + auto done = oneapi::math::blas::netlib::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -4017,7 +4017,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4028,7 +4028,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4040,7 +4040,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4052,7 +4052,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -4062,7 +4062,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4072,7 +4072,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4082,7 +4082,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4092,7 +4092,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4103,7 +4103,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::netlib::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4115,7 +4115,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::netlib::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4129,7 +4129,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::netlib::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4143,7 +4143,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::netlib::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4152,7 +4152,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4160,7 +4160,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4169,7 +4169,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4178,7 +4178,7 @@ sycl::event omatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4187,7 +4187,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy2( + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4196,7 +4196,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy2( + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4206,7 +4206,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy2( + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4216,7 +4216,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy2( + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4224,7 +4224,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tran sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4232,7 +4232,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4241,7 +4241,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4250,7 +4250,7 @@ sycl::event imatcopy(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4260,7 +4260,7 @@ sycl::event omatadd(backend_selector selector, transpose transa float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4270,7 +4270,7 @@ sycl::event omatadd(backend_selector selector, transpose transa double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4281,7 +4281,7 @@ sycl::event omatadd(backend_selector selector, transpose transa const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4292,7 +4292,7 @@ sycl::event omatadd(backend_selector selector, transpose transa const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } @@ -4302,7 +4302,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4313,7 +4313,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4325,7 +4325,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4337,7 +4337,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4348,7 +4348,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4359,7 +4359,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4370,7 +4370,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4381,7 +4381,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; diff --git a/include/oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp b/include/oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp index 404355d33..4c6a5f0ea 100644 --- a/include/oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp +++ b/include/oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp @@ -34,13 +34,13 @@ #include "oneapi/math/detail/export.hpp" namespace oneapi { -namespace mkl { +namespace math { -using oneapi::mkl::transpose; -using oneapi::mkl::uplo; -using oneapi::mkl::side; -using oneapi::mkl::diag; -using oneapi::mkl::offset; +using oneapi::math::transpose; +using oneapi::math::uplo; +using oneapi::math::side; +using oneapi::math::diag; +using oneapi::math::offset; namespace blas { namespace netlib { @@ -56,7 +56,7 @@ namespace row_major { } //namespace row_major } //namespace netlib } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_ONEMATH_BLAS_NETLIB_HPP_ diff --git a/include/oneapi/math/blas/detail/onemath_blas_backends.hxx b/include/oneapi/math/blas/detail/onemath_blas_backends.hxx index 28ac984c6..e9fcbbf66 100644 --- a/include/oneapi/math/blas/detail/onemath_blas_backends.hxx +++ b/include/oneapi/math/blas/detail/onemath_blas_backends.hxx @@ -19,283 +19,283 @@ // Buffer APIs -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, +ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, +ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void hemm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void hemm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void hemm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void hemm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, +ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, +ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, +ONEMATH_EXPORT void herk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, +ONEMATH_EXPORT void herk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, +ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, +ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT void her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, +ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, +ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, +ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, +ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -303,7 +303,7 @@ ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, @@ -311,54 +311,54 @@ ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, @@ -399,270 +399,270 @@ ONEMATH_EXPORT void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void her(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void her(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void her2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void her2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); -ONEMATH_EXPORT void hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); -ONEMATH_EXPORT void hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); -ONEMATH_EXPORT void hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); -ONEMATH_EXPORT void sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void symv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void symv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void syr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void syr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); -ONEMATH_EXPORT void spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); -ONEMATH_EXPORT void spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); -ONEMATH_EXPORT void spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +ONEMATH_EXPORT void spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); -ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t n, +ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); @@ -930,24 +930,24 @@ ONEMATH_EXPORT void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -955,8 +955,8 @@ ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -964,8 +964,8 @@ ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, @@ -973,190 +973,190 @@ ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, +ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, +ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, +ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, +ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); -ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -1164,85 +1164,85 @@ ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose tra std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, @@ -1250,22 +1250,22 @@ ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, // USM APIs -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, @@ -1273,8 +1273,8 @@ ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose trans std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, @@ -1282,29 +1282,29 @@ ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose trans std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, @@ -1312,8 +1312,8 @@ ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose const std::int32_t *co, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, @@ -1321,8 +1321,8 @@ ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose const std::int32_t *co, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, @@ -1330,8 +1330,8 @@ ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose const std::int32_t *co, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, @@ -1339,94 +1339,94 @@ ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose const std::int32_t *co, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, std::int64_t *n, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, std::int64_t *n, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, std::int64_t *n, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, @@ -1434,8 +1434,8 @@ ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upp std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, std::int64_t *n, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, @@ -1443,24 +1443,24 @@ ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upp std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -1468,8 +1468,8 @@ ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo uppe std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -1477,131 +1477,131 @@ ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo uppe std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event herk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event herk(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +ONEMATH_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, @@ -1609,64 +1609,64 @@ ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event trsm_batch( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event trsm_batch( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, - oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, - oneapi::mkl::diag *unit_diag, std::int64_t *m, +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, + oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, + oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, - oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, - oneapi::mkl::diag *unit_diag, std::int64_t *m, +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, + oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, + oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event trsm_batch( - sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event trsm_batch( - sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1674,7 +1674,7 @@ ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1682,7 +1682,7 @@ ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, @@ -1690,7 +1690,7 @@ ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, @@ -1699,20 +1699,20 @@ ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemv_batch( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemv_batch( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, float **y, @@ -1720,7 +1720,7 @@ ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, double **y, @@ -1728,7 +1728,7 @@ ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, @@ -1738,13 +1738,13 @@ ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemv_batch( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float *c, @@ -1752,7 +1752,7 @@ ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double *c, @@ -1760,7 +1760,7 @@ ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, @@ -1769,7 +1769,7 @@ ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, @@ -1778,21 +1778,21 @@ ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, @@ -1800,7 +1800,7 @@ ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *lef std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, @@ -1808,19 +1808,19 @@ ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *lef std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, @@ -1828,7 +1828,7 @@ ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, @@ -1870,7 +1870,7 @@ ONEMATH_EXPORT sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1878,7 +1878,7 @@ ONEMATH_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lowe std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1886,7 +1886,7 @@ ONEMATH_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lowe std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1894,7 +1894,7 @@ ONEMATH_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lowe std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, @@ -1902,276 +1902,276 @@ ONEMATH_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lowe std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event her(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event her(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event her2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event her2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event symv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event symv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, +ONEMATH_EXPORT sycl::event spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, +ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies = {}); @@ -2537,8 +2537,8 @@ ONEMATH_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, const float **b, std::int64_t *ldb, float *beta, float **c, @@ -2546,8 +2546,8 @@ ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, const double **b, std::int64_t *ldb, double *beta, double **c, @@ -2555,8 +2555,8 @@ ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, @@ -2566,15 +2566,15 @@ ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemm_batch( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, sycl::half *beta, @@ -2582,16 +2582,16 @@ ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, @@ -2599,8 +2599,8 @@ ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, @@ -2608,8 +2608,8 @@ ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t *group_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, const float *b, std::int64_t ldb, std::int64_t stride_b, @@ -2617,8 +2617,8 @@ ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, const double *b, std::int64_t ldb, std::int64_t stride_b, @@ -2627,7 +2627,7 @@ ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemm_batch( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, @@ -2635,7 +2635,7 @@ ONEMATH_EXPORT sycl::event gemm_batch( const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemm_batch( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -2643,14 +2643,14 @@ ONEMATH_EXPORT sycl::event gemm_batch( std::int64_t batch_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event gemm_batch( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, @@ -2658,8 +2658,8 @@ ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, @@ -2667,8 +2667,8 @@ ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -2676,22 +2676,22 @@ ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, @@ -2699,8 +2699,8 @@ ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_low std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, @@ -2708,21 +2708,21 @@ ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_low std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, @@ -2730,7 +2730,7 @@ ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::trans std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, @@ -2738,34 +2738,34 @@ ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::trans std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, +ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, const float *b, std::int64_t ldb, std::int64_t stride_b, @@ -2773,8 +2773,8 @@ ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transp std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, +ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, const double *b, std::int64_t ldb, std::int64_t stride_b, @@ -2783,7 +2783,7 @@ ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transp const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event omatadd_batch( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, @@ -2791,102 +2791,102 @@ ONEMATH_EXPORT sycl::event omatadd_batch( const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event omatadd_batch( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, diff --git a/include/oneapi/math/blas/detail/portblas/blas_ct.hpp b/include/oneapi/math/blas/detail/portblas/blas_ct.hpp index a7ffbe877..c79dc2dbd 100644 --- a/include/oneapi/math/blas/detail/portblas/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/portblas/blas_ct.hpp @@ -34,7 +34,7 @@ #include "oneapi/math/blas/detail/blas_ct_backends.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace column_major { @@ -51,7 +51,7 @@ namespace row_major { } //namespace row_major } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_DETAIL_PORTBLAS_BLAS_CT_HPP_ diff --git a/include/oneapi/math/blas/detail/portblas/blas_ct.hxx b/include/oneapi/math/blas/detail/portblas/blas_ct.hxx index 8a66ed707..8f505ef8c 100644 --- a/include/oneapi/math/blas/detail/portblas/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/portblas/blas_ct.hxx @@ -22,7 +22,7 @@ void herk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -30,105 +30,105 @@ void herk(backend_selector selector, uplo upper_lower, transp std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void scal(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } @@ -138,7 +138,7 @@ void gemm_batch(backend_selector selector, transpose transa, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -149,7 +149,7 @@ void gemm_batch(backend_selector selector, transpose transa, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -160,7 +160,7 @@ void gemm_batch(backend_selector selector, transpose transa, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -171,7 +171,7 @@ void gemm_batch(backend_selector selector, transpose transa, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -182,7 +182,7 @@ void gemm_batch(backend_selector selector, transpose transa, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -193,7 +193,7 @@ void gemm_batch(backend_selector selector, transpose transa, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -204,7 +204,7 @@ void gemm_batch(backend_selector selector, transpose transa, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -215,7 +215,7 @@ void gemm_batch(backend_selector selector, transpose transa, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -223,14 +223,14 @@ void gemm_batch(backend_selector selector, transpose transa, void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -238,7 +238,7 @@ void syrk(backend_selector selector, uplo upper_lower, transp std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -246,7 +246,7 @@ void syrk(backend_selector selector, uplo upper_lower, transp std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -254,7 +254,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -263,7 +263,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -273,7 +273,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -283,7 +283,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -292,7 +292,7 @@ void her2(backend_selector selector, uplo upper_lower, std::i std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -300,7 +300,7 @@ void her2(backend_selector selector, uplo upper_lower, std::i std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -308,7 +308,7 @@ void hbmv(backend_selector selector, uplo upper_lower, std::i std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -316,61 +316,61 @@ void hbmv(backend_selector selector, uplo upper_lower, std::i std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void rot(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { - oneapi::mkl::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { - oneapi::mkl::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { - oneapi::mkl::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { - oneapi::mkl::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -378,7 +378,7 @@ void axpy_batch(backend_selector selector, std::int64_t n, do sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -386,7 +386,7 @@ void axpy_batch(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -394,42 +394,42 @@ void axpy_batch(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } void axpby(backend_selector selector, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void sdsdot(backend_selector selector, std::int64_t n, float sb, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, + oneapi::math::blas::portblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); } @@ -437,7 +437,7 @@ void gerc(backend_selector selector, std::int64_t m, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -445,7 +445,7 @@ void gerc(backend_selector selector, std::int64_t m, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -453,7 +453,7 @@ void syr2k(backend_selector selector, uplo upper_lower, trans std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -461,7 +461,7 @@ void syr2k(backend_selector selector, uplo upper_lower, trans std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -470,7 +470,7 @@ void syr2k(backend_selector selector, uplo upper_lower, trans sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -479,7 +479,7 @@ void syr2k(backend_selector selector, uplo upper_lower, trans sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -487,7 +487,7 @@ void gemv(backend_selector selector, transpose trans, std::in std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -495,7 +495,7 @@ void gemv(backend_selector selector, transpose trans, std::in std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -503,7 +503,7 @@ void gemv(backend_selector selector, transpose trans, std::in std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -511,7 +511,7 @@ void gemv(backend_selector selector, transpose trans, std::in std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -520,7 +520,7 @@ void gemv_batch(backend_selector selector, transpose trans, s std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -530,7 +530,7 @@ void gemv_batch(backend_selector selector, transpose trans, s std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -541,7 +541,7 @@ void gemv_batch(backend_selector selector, transpose trans, s std::int64_t incx, std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -552,7 +552,7 @@ void gemv_batch(backend_selector selector, transpose trans, s sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -562,7 +562,7 @@ void dgmm_batch(backend_selector selector, side left_right, s sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -572,7 +572,7 @@ void dgmm_batch(backend_selector selector, side left_right, s sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -582,7 +582,7 @@ void dgmm_batch(backend_selector selector, side left_right, s std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -592,7 +592,7 @@ void dgmm_batch(backend_selector selector, side left_right, s std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -600,58 +600,58 @@ void dgmm_batch(backend_selector selector, side left_right, s void her(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, + oneapi::math::blas::portblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -659,21 +659,21 @@ void hpmv(backend_selector selector, uplo upper_lower, std::i std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, + oneapi::math::blas::portblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, + oneapi::math::blas::portblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, + oneapi::math::blas::portblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -682,7 +682,7 @@ void gemm_bias(backend_selector selector, transpose transa, t sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + oneapi::math::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -692,7 +692,7 @@ void gemm_bias(backend_selector selector, transpose transa, t sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + oneapi::math::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -702,7 +702,7 @@ void gemm_bias(backend_selector selector, transpose transa, t sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + oneapi::math::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -712,38 +712,38 @@ void gemm_bias(backend_selector selector, transpose transa, t sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + oneapi::math::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -751,37 +751,37 @@ void geru(backend_selector selector, std::int64_t m, std::int std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -789,7 +789,7 @@ void gemm(backend_selector selector, transpose transa, transp std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -798,7 +798,7 @@ void gemm(backend_selector selector, transpose transa, transp sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -807,7 +807,7 @@ void gemm(backend_selector selector, transpose transa, transp sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -815,7 +815,7 @@ void gemm(backend_selector selector, transpose transa, transp std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -823,7 +823,7 @@ void gemm(backend_selector selector, transpose transa, transp std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -831,35 +831,35 @@ void gemm(backend_selector selector, transpose transa, transp std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -867,7 +867,7 @@ void trsm(backend_selector selector, side left_right, uplo up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -875,7 +875,7 @@ void trsm(backend_selector selector, side left_right, uplo up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -883,7 +883,7 @@ void trsm(backend_selector selector, side left_right, uplo up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -891,7 +891,7 @@ void trsm(backend_selector selector, side left_right, uplo up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -899,14 +899,14 @@ void dotu(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, @@ -914,7 +914,7 @@ void hemm(backend_selector selector, side left_right, uplo up sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -923,7 +923,7 @@ void hemm(backend_selector selector, side left_right, uplo up sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -931,7 +931,7 @@ void hpr2(backend_selector selector, uplo upper_lower, std::i std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -939,7 +939,7 @@ void hpr2(backend_selector selector, uplo upper_lower, std::i std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -947,7 +947,7 @@ void gbmv(backend_selector selector, transpose trans, std::in std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -955,7 +955,7 @@ void gbmv(backend_selector selector, transpose trans, std::in std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -964,7 +964,7 @@ void gbmv(backend_selector selector, transpose trans, std::in sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -973,35 +973,35 @@ void gbmv(backend_selector selector, transpose trans, std::in sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } @@ -1009,7 +1009,7 @@ void symm(backend_selector selector, side left_right, uplo up std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1017,7 +1017,7 @@ void symm(backend_selector selector, side left_right, uplo up std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1026,7 +1026,7 @@ void symm(backend_selector selector, side left_right, uplo up sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1035,7 +1035,7 @@ void symm(backend_selector selector, side left_right, uplo up sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1043,27 +1043,27 @@ void dotc(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - oneapi::mkl::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } @@ -1071,7 +1071,7 @@ void trmm(backend_selector selector, side left_right, uplo up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1079,7 +1079,7 @@ void trmm(backend_selector selector, side left_right, uplo up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1087,7 +1087,7 @@ void trmm(backend_selector selector, side left_right, uplo up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1095,105 +1095,105 @@ void trmm(backend_selector selector, side left_right, uplo up transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m) { - oneapi::mkl::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m) { - oneapi::mkl::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + oneapi::math::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1201,7 +1201,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + oneapi::math::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1209,7 +1209,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + oneapi::math::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1217,7 +1217,7 @@ void copy_batch(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + oneapi::math::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1225,7 +1225,7 @@ void hemv(backend_selector selector, uplo upper_lower, std::i std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::portblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1233,7 +1233,7 @@ void hemv(backend_selector selector, uplo upper_lower, std::i std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::portblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1241,7 +1241,7 @@ void gemmt(backend_selector selector, uplo upper_lower, trans transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1249,7 +1249,7 @@ void gemmt(backend_selector selector, uplo upper_lower, trans transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1258,7 +1258,7 @@ void gemmt(backend_selector selector, uplo upper_lower, trans sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1267,37 +1267,37 @@ void gemmt(backend_selector selector, uplo upper_lower, trans sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -1305,102 +1305,102 @@ void sbmv(backend_selector selector, uplo upper_lower, std::i std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::mkl::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + oneapi::math::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, @@ -1408,7 +1408,7 @@ void trsm_batch(backend_selector selector, side left_right, u sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1418,7 +1418,7 @@ void trsm_batch(backend_selector selector, side left_right, u sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1428,7 +1428,7 @@ void trsm_batch(backend_selector selector, side left_right, u std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1438,7 +1438,7 @@ void trsm_batch(backend_selector selector, side left_right, u std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1448,7 +1448,7 @@ void her2k(backend_selector selector, uplo upper_lower, trans sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1457,43 +1457,43 @@ void her2k(backend_selector selector, uplo upper_lower, trans sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::portblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::mkl::blas::portblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::portblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1501,7 +1501,7 @@ void omatcopy_batch(backend_selector selector, transpose tran std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1509,7 +1509,7 @@ void omatcopy_batch(backend_selector selector, transpose tran std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1518,7 +1518,7 @@ void omatcopy_batch(backend_selector selector, transpose tran sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1527,21 +1527,21 @@ void omatcopy_batch(backend_selector selector, transpose tran sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1549,7 +1549,7 @@ void imatcopy_batch(backend_selector selector, transpose tran std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1557,7 +1557,7 @@ void imatcopy_batch(backend_selector selector, transpose tran std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1566,7 +1566,7 @@ void omatadd_batch(backend_selector selector, transpose trans std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1576,7 +1576,7 @@ void omatadd_batch(backend_selector selector, transpose trans std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1587,7 +1587,7 @@ void omatadd_batch(backend_selector selector, transpose trans std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1599,7 +1599,7 @@ void omatadd_batch(backend_selector selector, transpose trans sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1607,28 +1607,28 @@ void omatadd_batch(backend_selector selector, transpose trans void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } @@ -1636,7 +1636,7 @@ void omatcopy2(backend_selector selector, transpose trans, st std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1644,7 +1644,7 @@ void omatcopy2(backend_selector selector, transpose trans, st std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1652,7 +1652,7 @@ void omatcopy2(backend_selector selector, transpose trans, st std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1660,35 +1660,35 @@ void omatcopy2(backend_selector selector, transpose trans, st std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } @@ -1696,7 +1696,7 @@ void omatadd(backend_selector selector, transpose transa, tra std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1704,7 +1704,7 @@ void omatadd(backend_selector selector, transpose transa, tra std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1713,7 +1713,7 @@ void omatadd(backend_selector selector, transpose transa, tra sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1722,7 +1722,7 @@ void omatadd(backend_selector selector, transpose transa, tra sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1731,7 +1731,7 @@ void omatadd(backend_selector selector, transpose transa, tra sycl::event syr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syr2( + auto done = oneapi::math::blas::portblas::MAJOR::syr2( selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1740,21 +1740,21 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syr2( + auto done = oneapi::math::blas::portblas::MAJOR::syr2( selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, float alpha, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, double alpha, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1762,7 +1762,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, d sycl::event scal(backend_selector selector, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1770,7 +1770,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, sycl::event scal(backend_selector selector, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1778,7 +1778,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, sycl::event scal(backend_selector selector, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1786,7 +1786,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, f sycl::event scal(backend_selector selector, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1794,7 +1794,7 @@ sycl::event scal(backend_selector selector, std::int64_t n, d sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trmv( + auto done = oneapi::math::blas::portblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1802,7 +1802,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trmv( + auto done = oneapi::math::blas::portblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1811,7 +1811,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trmv( + auto done = oneapi::math::blas::portblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1820,7 +1820,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trmv( + auto done = oneapi::math::blas::portblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1828,7 +1828,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1836,7 +1836,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1845,7 +1845,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1854,7 +1854,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1862,7 +1862,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1870,7 +1870,7 @@ sycl::event spr(backend_selector selector, uplo upper_lower, sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1880,7 +1880,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::hpmv( + auto done = oneapi::math::blas::portblas::MAJOR::hpmv( selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1890,7 +1890,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::hpmv( + auto done = oneapi::math::blas::portblas::MAJOR::hpmv( selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1899,7 +1899,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk( + auto done = oneapi::math::blas::portblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1908,7 +1908,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk( + auto done = oneapi::math::blas::portblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1918,7 +1918,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk( + auto done = oneapi::math::blas::portblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1928,7 +1928,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk( + auto done = oneapi::math::blas::portblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1938,7 +1938,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1949,7 +1949,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1961,7 +1961,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1973,7 +1973,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1984,7 +1984,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_ std::int64_t lda, std::int64_t stride_a, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -1995,7 +1995,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_ const double *a, std::int64_t lda, std::int64_t stride_a, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2007,7 +2007,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_ std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2019,7 +2019,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_ std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2029,7 +2029,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::her2( + auto done = oneapi::math::blas::portblas::MAJOR::her2( selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2038,7 +2038,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::her2( + auto done = oneapi::math::blas::portblas::MAJOR::her2( selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2049,7 +2049,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2060,7 +2060,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2068,7 +2068,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, sycl::event rot(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2077,7 +2077,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2085,7 +2085,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, sycl::event rot(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2093,7 +2093,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, fl sycl::event rot(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2101,7 +2101,7 @@ sycl::event rot(backend_selector selector, std::int64_t n, do sycl::event axpy(backend_selector selector, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2109,7 +2109,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, f sycl::event axpy(backend_selector selector, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2118,7 +2118,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2127,7 +2127,7 @@ sycl::event axpy(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2136,7 +2136,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_ const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2145,7 +2145,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_ const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2155,7 +2155,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_ std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2165,7 +2165,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_ std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2174,7 +2174,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_ const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2184,7 +2184,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_ const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2195,7 +2195,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_ std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2206,7 +2206,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_ std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2215,7 +2215,7 @@ sycl::event axpy_batch(backend_selector selector, std::int64_ sycl::event axpby(backend_selector selector, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2223,7 +2223,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, sycl::event axpby(backend_selector selector, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2232,7 +2232,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2241,7 +2241,7 @@ sycl::event axpby(backend_selector selector, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2250,7 +2250,7 @@ sycl::event gerc(backend_selector selector, std::int64_t m, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2259,7 +2259,7 @@ sycl::event gerc(backend_selector selector, std::int64_t m, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2268,7 +2268,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -2278,7 +2278,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -2289,7 +2289,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -2300,7 +2300,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -2310,7 +2310,7 @@ sycl::event gemv(backend_selector selector, transpose trans, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv( + auto done = oneapi::math::blas::portblas::MAJOR::gemv( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2319,7 +2319,7 @@ sycl::event gemv(backend_selector selector, transpose trans, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv( + auto done = oneapi::math::blas::portblas::MAJOR::gemv( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2329,7 +2329,7 @@ sycl::event gemv(backend_selector selector, transpose trans, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv( + auto done = oneapi::math::blas::portblas::MAJOR::gemv( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2339,7 +2339,7 @@ sycl::event gemv(backend_selector selector, transpose trans, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv( + auto done = oneapi::math::blas::portblas::MAJOR::gemv( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2350,7 +2350,7 @@ sycl::event gemv_batch(backend_selector selector, transpose t std::int64_t stridex, float beta, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2362,7 +2362,7 @@ sycl::event gemv_batch(backend_selector selector, transpose t std::int64_t stridex, double beta, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2375,7 +2375,7 @@ sycl::event gemv_batch(backend_selector selector, transpose t std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2388,7 +2388,7 @@ sycl::event gemv_batch(backend_selector selector, transpose t std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2399,7 +2399,7 @@ sycl::event gemv_batch(backend_selector selector, transpose * std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2410,7 +2410,7 @@ sycl::event gemv_batch(backend_selector selector, transpose * std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2422,7 +2422,7 @@ sycl::event gemv_batch(backend_selector selector, transpose * const std::complex **x, std::int64_t *incx, std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2435,7 +2435,7 @@ sycl::event gemv_batch(backend_selector selector, transpose * std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2446,7 +2446,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_r std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2457,7 +2457,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_r std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2469,7 +2469,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_r std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2481,7 +2481,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_r std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2492,7 +2492,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2503,7 +2503,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2514,7 +2514,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2525,7 +2525,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2535,7 +2535,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2544,7 +2544,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2552,7 +2552,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -2560,7 +2560,7 @@ sycl::event hpr(backend_selector selector, uplo upper_lower, sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -2568,7 +2568,7 @@ sycl::event hpr(backend_selector selector, uplo upper_lower, sycl::event iamin(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2576,7 +2576,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, sycl::event iamin(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2584,7 +2584,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, sycl::event iamin(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2592,7 +2592,7 @@ sycl::event iamin(backend_selector selector, std::int64_t n, sycl::event iamin(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2603,7 +2603,7 @@ sycl::event gemm_batch(backend_selector selector, transpose * const sycl::half **b, std::int64_t *ldb, sycl::half *beta, sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2615,7 +2615,7 @@ sycl::event gemm_batch(backend_selector selector, transpose * std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2627,7 +2627,7 @@ sycl::event gemm_batch(backend_selector selector, transpose * const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2639,7 +2639,7 @@ sycl::event gemm_batch(backend_selector selector, transpose * const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2651,7 +2651,7 @@ sycl::event gemm_batch(backend_selector selector, transpose * std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2663,7 +2663,7 @@ sycl::event gemm_batch(backend_selector selector, transpose * std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2675,7 +2675,7 @@ sycl::event gemm_batch(backend_selector selector, transpose * const std::complex **b, std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2688,7 +2688,7 @@ sycl::event gemm_batch(backend_selector selector, transpose * std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2701,7 +2701,7 @@ sycl::event gemm_batch(backend_selector selector, transpose t std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2713,7 +2713,7 @@ sycl::event gemm_batch(backend_selector selector, transpose t const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2725,7 +2725,7 @@ sycl::event gemm_batch(backend_selector selector, transpose t const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2737,7 +2737,7 @@ sycl::event gemm_batch(backend_selector selector, transpose t const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2749,7 +2749,7 @@ sycl::event gemm_batch(backend_selector selector, transpose t const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2761,7 +2761,7 @@ sycl::event gemm_batch(backend_selector selector, transpose t const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2774,7 +2774,7 @@ sycl::event gemm_batch(backend_selector selector, transpose t std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2787,7 +2787,7 @@ sycl::event gemm_batch(backend_selector selector, transpose t std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2796,7 +2796,7 @@ sycl::event gemm_batch(backend_selector selector, transpose t sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::spmv( + auto done = oneapi::math::blas::portblas::MAJOR::spmv( selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -2804,7 +2804,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::spmv( + auto done = oneapi::math::blas::portblas::MAJOR::spmv( selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -2812,7 +2812,7 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, sycl::event swap(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2820,7 +2820,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, f sycl::event swap(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2828,7 +2828,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, d sycl::event swap(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2836,7 +2836,7 @@ sycl::event swap(backend_selector selector, std::int64_t n, sycl::event swap(backend_selector selector, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2845,7 +2845,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2854,7 +2854,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, s std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2862,7 +2862,7 @@ sycl::event geru(backend_selector selector, std::int64_t m, s sycl::event nrm2(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2870,21 +2870,21 @@ sycl::event nrm2(backend_selector selector, std::int64_t n, sycl::event nrm2(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2894,7 +2894,7 @@ sycl::event gemm(backend_selector selector, transpose transa, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2904,7 +2904,7 @@ sycl::event gemm(backend_selector selector, transpose transa, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2915,7 +2915,7 @@ sycl::event gemm(backend_selector selector, transpose transa, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2926,7 +2926,7 @@ sycl::event gemm(backend_selector selector, transpose transa, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2937,7 +2937,7 @@ sycl::event gemm(backend_selector selector, transpose transa, sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2947,7 +2947,7 @@ sycl::event gemm(backend_selector selector, transpose transa, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2957,7 +2957,7 @@ sycl::event gemm(backend_selector selector, transpose transa, std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2968,7 +2968,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tr std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -2980,7 +2980,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tr std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -2992,7 +2992,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tr std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3004,7 +3004,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tr std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::portblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -3014,7 +3014,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::herk( + auto done = oneapi::math::blas::portblas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -3023,7 +3023,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::herk( + auto done = oneapi::math::blas::portblas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -3031,7 +3031,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, sycl::event ger(backend_selector selector, std::int64_t m, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -3040,7 +3040,7 @@ sycl::event ger(backend_selector selector, std::int64_t m, st double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -3049,7 +3049,7 @@ sycl::event trsm(backend_selector selector, side left_right, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3059,7 +3059,7 @@ sycl::event trsm(backend_selector selector, side left_right, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3070,7 +3070,7 @@ sycl::event trsm(backend_selector selector, side left_right, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3081,7 +3081,7 @@ sycl::event trsm(backend_selector selector, side left_right, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3092,7 +3092,7 @@ sycl::event trsm_batch(backend_selector selector, side left_r std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3103,7 +3103,7 @@ sycl::event trsm_batch(backend_selector selector, side left_r std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3115,7 +3115,7 @@ sycl::event trsm_batch(backend_selector selector, side left_r std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3127,7 +3127,7 @@ sycl::event trsm_batch(backend_selector selector, side left_r std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3138,7 +3138,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3149,7 +3149,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3161,7 +3161,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3173,7 +3173,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_ std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3183,7 +3183,7 @@ sycl::event dotu(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3192,7 +3192,7 @@ sycl::event dotu(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3202,7 +3202,7 @@ sycl::event hemm(backend_selector selector, side left_right, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3213,7 +3213,7 @@ sycl::event hemm(backend_selector selector, side left_right, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3223,7 +3223,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, + auto done = oneapi::math::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3232,7 +3232,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, + auto done = oneapi::math::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3242,7 +3242,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, + oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3252,7 +3252,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, + oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3263,7 +3263,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, + oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3274,7 +3274,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, + oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3282,7 +3282,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tbmv( + auto done = oneapi::math::blas::portblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3290,7 +3290,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tbmv( + auto done = oneapi::math::blas::portblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3299,7 +3299,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tbmv( + auto done = oneapi::math::blas::portblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3308,7 +3308,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tbmv( + auto done = oneapi::math::blas::portblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3317,7 +3317,7 @@ sycl::event symm(backend_selector selector, side left_right, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3327,7 +3327,7 @@ sycl::event symm(backend_selector selector, side left_right, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3338,7 +3338,7 @@ sycl::event symm(backend_selector selector, side left_right, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3349,7 +3349,7 @@ sycl::event symm(backend_selector selector, side left_right, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3359,7 +3359,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3368,7 +3368,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3376,7 +3376,7 @@ sycl::event dotc(backend_selector selector, std::int64_t n, sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3384,7 +3384,7 @@ sycl::event syr(backend_selector selector, uplo upper_lower, sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3393,7 +3393,7 @@ sycl::event trmm(backend_selector selector, side left_right, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3403,7 +3403,7 @@ sycl::event trmm(backend_selector selector, side left_right, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3414,7 +3414,7 @@ sycl::event trmm(backend_selector selector, side left_right, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3425,7 +3425,7 @@ sycl::event trmm(backend_selector selector, side left_right, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3433,14 +3433,14 @@ sycl::event trmm(backend_selector selector, side left_right, sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + auto done = oneapi::math::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } sycl::event rotmg(backend_selector selector, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + auto done = oneapi::math::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } @@ -3448,7 +3448,7 @@ sycl::event rotmg(backend_selector selector, double *d1, doub sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3456,7 +3456,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3465,7 +3465,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3474,7 +3474,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3482,7 +3482,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsv( + auto done = oneapi::math::blas::portblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3490,7 +3490,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsv( + auto done = oneapi::math::blas::portblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3499,7 +3499,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsv( + auto done = oneapi::math::blas::portblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3508,7 +3508,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::trsv( + auto done = oneapi::math::blas::portblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3516,7 +3516,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, sycl::event copy(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3524,7 +3524,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, c sycl::event copy(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3532,7 +3532,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, c sycl::event copy(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3540,7 +3540,7 @@ sycl::event copy(backend_selector selector, std::int64_t n, sycl::event copy(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3549,7 +3549,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_ const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3558,7 +3558,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_ const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3567,7 +3567,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_ const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3576,7 +3576,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_ const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3585,7 +3585,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_ std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3594,7 +3594,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_ const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3603,7 +3603,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_ const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3612,7 +3612,7 @@ sycl::event copy_batch(backend_selector selector, std::int64_ const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3622,7 +3622,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::hemv( + auto done = oneapi::math::blas::portblas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3632,7 +3632,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::hemv( + auto done = oneapi::math::blas::portblas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3641,7 +3641,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower transpose transb, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3651,7 +3651,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower transpose transb, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3662,7 +3662,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3673,7 +3673,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3684,7 +3684,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3694,7 +3694,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3702,7 +3702,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, sycl::event asum(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3710,21 +3710,21 @@ sycl::event asum(backend_selector selector, std::int64_t n, sycl::event asum(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3732,7 +3732,7 @@ sycl::event asum(backend_selector selector, std::int64_t n, c sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tbsv( + auto done = oneapi::math::blas::portblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3740,7 +3740,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tbsv( + auto done = oneapi::math::blas::portblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3749,7 +3749,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tbsv( + auto done = oneapi::math::blas::portblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3758,7 +3758,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::tbsv( + auto done = oneapi::math::blas::portblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3766,7 +3766,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, + auto done = oneapi::math::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3774,7 +3774,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, + auto done = oneapi::math::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3782,7 +3782,7 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, sycl::event iamax(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3790,7 +3790,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event iamax(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3798,7 +3798,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event iamax(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3806,7 +3806,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event iamax(backend_selector selector, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3814,7 +3814,7 @@ sycl::event iamax(backend_selector selector, std::int64_t n, sycl::event rotm(backend_selector selector, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3822,7 +3822,7 @@ sycl::event rotm(backend_selector selector, std::int64_t n, f sycl::event rotm(backend_selector selector, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3830,14 +3830,14 @@ sycl::event rotm(backend_selector selector, std::int64_t n, d sycl::event rotg(backend_selector selector, float *a, float *b, float *c, float *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event rotg(backend_selector selector, double *a, double *b, double *c, double *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3845,7 +3845,7 @@ sycl::event rotg(backend_selector selector, std::complex *b, float *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3853,14 +3853,14 @@ sycl::event rotg(backend_selector selector, std::complex *b, double *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event sdsdot(backend_selector selector, std::int64_t n, float sb, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, + auto done = oneapi::math::blas::portblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result, dependencies); return done; } @@ -3870,7 +3870,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3881,7 +3881,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3890,7 +3890,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower sycl::event dot(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3898,7 +3898,7 @@ sycl::event dot(backend_selector selector, std::int64_t n, co sycl::event dot(backend_selector selector, std::int64_t n, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3906,7 +3906,7 @@ sycl::event dot(backend_selector selector, std::int64_t n, co sycl::event dot(backend_selector selector, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3915,7 +3915,7 @@ sycl::event symv(backend_selector selector, uplo upper_lower, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::symv( + auto done = oneapi::math::blas::portblas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3924,7 +3924,7 @@ sycl::event symv(backend_selector selector, uplo upper_lower, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::symv( + auto done = oneapi::math::blas::portblas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3934,7 +3934,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpo std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3945,7 +3945,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpo std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3956,7 +3956,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpo const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3967,7 +3967,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpo const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3977,7 +3977,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpo std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -3986,7 +3986,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpo std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -3996,7 +3996,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpo std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4006,7 +4006,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpo std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -4017,7 +4017,7 @@ sycl::event omatadd_batch(backend_selector selector, transpos const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4029,7 +4029,7 @@ sycl::event omatadd_batch(backend_selector selector, transpos const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4042,7 +4042,7 @@ sycl::event omatadd_batch(backend_selector selector, transpos const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4055,7 +4055,7 @@ sycl::event omatadd_batch(backend_selector selector, transpos const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -4064,7 +4064,7 @@ sycl::event omatadd_batch(backend_selector selector, transpos sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4072,7 +4072,7 @@ sycl::event omatcopy(backend_selector selector, transpose tra sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4081,7 +4081,7 @@ sycl::event omatcopy(backend_selector selector, transpose tra std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4090,7 +4090,7 @@ sycl::event omatcopy(backend_selector selector, transpose tra std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -4099,7 +4099,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tr std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4108,7 +4108,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tr std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4118,7 +4118,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tr std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4128,7 +4128,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tr std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4136,7 +4136,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tr sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4144,7 +4144,7 @@ sycl::event imatcopy(backend_selector selector, transpose tra sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4153,7 +4153,7 @@ sycl::event imatcopy(backend_selector selector, transpose tra std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4162,7 +4162,7 @@ sycl::event imatcopy(backend_selector selector, transpose tra std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4171,7 +4171,7 @@ sycl::event omatadd(backend_selector selector, transpose tran transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; @@ -4181,7 +4181,7 @@ sycl::event omatadd(backend_selector selector, transpose tran transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; @@ -4192,7 +4192,7 @@ sycl::event omatadd(backend_selector selector, transpose tran const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; @@ -4203,7 +4203,7 @@ sycl::event omatadd(backend_selector selector, transpose tran const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; @@ -4214,7 +4214,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpo std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4225,7 +4225,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpo std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4236,7 +4236,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpo const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4247,7 +4247,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpo const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4257,7 +4257,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpo std::int64_t *m, std::int64_t *n, float *alpha, float **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4267,7 +4267,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpo std::int64_t *m, std::int64_t *n, double *alpha, double **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4278,7 +4278,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpo std::complex **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4289,7 +4289,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpo std::complex **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::portblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; diff --git a/include/oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp b/include/oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp index 2c7cf0905..52420dc85 100644 --- a/include/oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp +++ b/include/oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp @@ -33,13 +33,13 @@ #include "oneapi/math/detail/export.hpp" namespace oneapi { -namespace mkl { +namespace math { -using oneapi::mkl::transpose; -using oneapi::mkl::uplo; -using oneapi::mkl::side; -using oneapi::mkl::diag; -using oneapi::mkl::offset; +using oneapi::math::transpose; +using oneapi::math::uplo; +using oneapi::math::side; +using oneapi::math::diag; +using oneapi::math::offset; namespace blas { namespace portblas { @@ -55,7 +55,7 @@ namespace row_major { } //namespace row_major } // namespace portblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ONEMATH_BLAS_PORTBLAS_HPP_ diff --git a/include/oneapi/math/blas/detail/rocblas/blas_ct.hpp b/include/oneapi/math/blas/detail/rocblas/blas_ct.hpp index e06bfb335..73f2bc775 100644 --- a/include/oneapi/math/blas/detail/rocblas/blas_ct.hpp +++ b/include/oneapi/math/blas/detail/rocblas/blas_ct.hpp @@ -37,7 +37,7 @@ #include "oneapi/math/blas/detail/blas_ct_backends.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace column_major { @@ -54,7 +54,7 @@ namespace row_major { } //namespace row_major } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_DETAIL_ROCBLAS_BLAS_CT_HPP_ diff --git a/include/oneapi/math/blas/detail/rocblas/blas_ct.hxx b/include/oneapi/math/blas/detail/rocblas/blas_ct.hxx index bc86929b0..cdfeae51b 100644 --- a/include/oneapi/math/blas/detail/rocblas/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/rocblas/blas_ct.hxx @@ -22,111 +22,111 @@ void herk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, sycl::buffer, 1> &a, int64_t lda, float beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void herk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, sycl::buffer, 1> &a, int64_t lda, double beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void scal(backend_selector selector, int64_t n, float alpha, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, int64_t n, double alpha, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, int64_t n, float alpha, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, int64_t n, double alpha, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); + oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void spr(backend_selector selector, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void spr(backend_selector selector, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a) { - oneapi::mkl::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, @@ -134,7 +134,7 @@ void gemm_batch(backend_selector selector, transpose transa, t int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -144,7 +144,7 @@ void gemm_batch(backend_selector selector, transpose transa, t int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -155,7 +155,7 @@ void gemm_batch(backend_selector selector, transpose transa, t sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -166,7 +166,7 @@ void gemm_batch(backend_selector selector, transpose transa, t sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -176,7 +176,7 @@ void gemm_batch(backend_selector selector, transpose transa, t int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, sycl::half beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -186,7 +186,7 @@ void gemm_batch(backend_selector selector, transpose transa, t int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -196,7 +196,7 @@ void gemm_batch(backend_selector selector, transpose transa, t int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -206,7 +206,7 @@ void gemm_batch(backend_selector selector, transpose transa, t int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -214,14 +214,14 @@ void gemm_batch(backend_selector selector, transpose transa, t void syrk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, float beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, double beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -229,7 +229,7 @@ void syrk(backend_selector selector, uplo upper_lower, transpo int64_t k, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -237,7 +237,7 @@ void syrk(backend_selector selector, uplo upper_lower, transpo int64_t k, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -245,7 +245,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, t int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -254,7 +254,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, t int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -264,7 +264,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, t sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -274,7 +274,7 @@ void syrk_batch(backend_selector selector, uplo upper_lower, t sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -283,7 +283,7 @@ void her2(backend_selector selector, uplo upper_lower, int64_t std::complex alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -291,7 +291,7 @@ void her2(backend_selector selector, uplo upper_lower, int64_t std::complex alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } @@ -299,7 +299,7 @@ void hbmv(backend_selector selector, uplo upper_lower, int64_t std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -307,65 +307,65 @@ void hbmv(backend_selector selector, uplo upper_lower, int64_t std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void rot(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, float c, float s) { - oneapi::mkl::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, double c, double s) { - oneapi::mkl::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, float c, float s) { - oneapi::mkl::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, double c, double s) { - oneapi::mkl::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); + oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, int64_t n, float alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, int64_t n, double alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, int64_t n, float alpha, sycl::buffer &x, int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, int64_t n, double alpha, sycl::buffer &x, int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -373,7 +373,7 @@ void axpy_batch(backend_selector selector, int64_t n, std::com sycl::buffer, 1> &x, int64_t incx, int64_t stridex, sycl::buffer, 1> &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } @@ -381,42 +381,42 @@ void axpy_batch(backend_selector selector, int64_t n, std::com sycl::buffer, 1> &x, int64_t incx, int64_t stridex, sycl::buffer, 1> &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, + oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } void axpby(backend_selector selector, int64_t n, float alpha, sycl::buffer &x, int64_t incx, float beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, int64_t n, double alpha, sycl::buffer &x, int64_t incx, double beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void axpby(backend_selector selector, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); } void sdsdot(backend_selector selector, int64_t n, float sb, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, + oneapi::math::blas::rocblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); } @@ -424,7 +424,7 @@ void gerc(backend_selector selector, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -432,7 +432,7 @@ void gerc(backend_selector selector, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -440,7 +440,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transp int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -448,7 +448,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transp int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -456,7 +456,7 @@ void syr2k(backend_selector selector, uplo upper_lower, transp int64_t n, int64_t k, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -465,21 +465,21 @@ void syr2k(backend_selector selector, uplo upper_lower, transp sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemv(backend_selector selector, transpose trans, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, float beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, double beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -487,7 +487,7 @@ void gemv(backend_selector selector, transpose trans, int64_t std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -495,7 +495,7 @@ void gemv(backend_selector selector, transpose trans, int64_t std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, + oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -503,7 +503,7 @@ void gemv_batch(backend_selector selector, transpose trans, in float alpha, sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &x, int64_t incx, int64_t stridex, float beta, sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -512,7 +512,7 @@ void gemv_batch(backend_selector selector, transpose trans, in double alpha, sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &x, int64_t incx, int64_t stridex, double beta, sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -522,7 +522,7 @@ void gemv_batch(backend_selector selector, transpose trans, in int64_t stridea, sycl::buffer, 1> &x, int64_t incx, int64_t stridex, std::complex beta, sycl::buffer, 1> &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -533,7 +533,7 @@ void gemv_batch(backend_selector selector, transpose trans, in int64_t stridex, std::complex beta, sycl::buffer, 1> &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } @@ -542,7 +542,7 @@ void dgmm_batch(backend_selector selector, side left_right, in sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &x, int64_t incx, int64_t stridex, sycl::buffer &c, int64_t ldc, int64_t stridec, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -551,7 +551,7 @@ void dgmm_batch(backend_selector selector, side left_right, in sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &x, int64_t incx, int64_t stridex, sycl::buffer &c, int64_t ldc, int64_t stridec, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -561,7 +561,7 @@ void dgmm_batch(backend_selector selector, side left_right, in sycl::buffer, 1> &x, int64_t incx, int64_t stridex, sycl::buffer, 1> &c, int64_t ldc, int64_t stridec, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -571,7 +571,7 @@ void dgmm_batch(backend_selector selector, side left_right, in sycl::buffer, 1> &x, int64_t incx, int64_t stridex, sycl::buffer, 1> &c, int64_t ldc, int64_t stridec, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, + oneapi::math::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } @@ -579,56 +579,56 @@ void dgmm_batch(backend_selector selector, side left_right, in void her(backend_selector selector, uplo upper_lower, int64_t n, float alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void her(backend_selector selector, uplo upper_lower, int64_t n, double alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void hpr(backend_selector selector, uplo upper_lower, int64_t n, float alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void hpr(backend_selector selector, uplo upper_lower, int64_t n, double alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &a) { - oneapi::mkl::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + oneapi::math::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void iamin(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -636,21 +636,21 @@ void hpmv(backend_selector selector, uplo upper_lower, int64_t std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, int64_t incx, float beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, int64_t incx, double beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, + oneapi::math::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy); } @@ -659,7 +659,7 @@ void gemm_bias(backend_selector selector, transpose transa, tr sycl::buffer &a, int64_t lda, int8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + oneapi::math::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -669,7 +669,7 @@ void gemm_bias(backend_selector selector, transpose transa, tr sycl::buffer &a, int64_t lda, int8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + oneapi::math::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -679,7 +679,7 @@ void gemm_bias(backend_selector selector, transpose transa, tr sycl::buffer &a, int64_t lda, uint8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + oneapi::math::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -689,38 +689,38 @@ void gemm_bias(backend_selector selector, transpose transa, tr sycl::buffer &a, int64_t lda, uint8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - oneapi::mkl::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + oneapi::math::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } void swap(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } @@ -728,35 +728,35 @@ void geru(backend_selector selector, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void nrm2(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -764,7 +764,7 @@ void gemm(backend_selector selector, transpose transa, transpo int64_t m, int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -773,7 +773,7 @@ void gemm(backend_selector selector, transpose transa, transpo sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -782,7 +782,7 @@ void gemm(backend_selector selector, transpose transa, transpo sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -790,7 +790,7 @@ void gemm(backend_selector selector, transpose transa, transpo int64_t m, int64_t n, int64_t k, sycl::half alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, sycl::half beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -798,7 +798,7 @@ void gemm(backend_selector selector, transpose transa, transpo int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -806,49 +806,49 @@ void gemm(backend_selector selector, transpose transa, transpo int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } void syr2(backend_selector selector, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, int64_t m, int64_t n, float alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void ger(backend_selector selector, int64_t m, int64_t n, double alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, + oneapi::math::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -856,7 +856,7 @@ void trsm(backend_selector selector, side left_right, uplo upp transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -864,7 +864,7 @@ void trsm(backend_selector selector, side left_right, uplo upp transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -872,21 +872,21 @@ void dotu(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -894,7 +894,7 @@ void hemm(backend_selector selector, side left_right, uplo upp int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -902,7 +902,7 @@ void hpr2(backend_selector selector, uplo upper_lower, int64_t std::complex alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -910,7 +910,7 @@ void hpr2(backend_selector selector, uplo upper_lower, int64_t std::complex alpha, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &a) { - oneapi::mkl::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } @@ -918,7 +918,7 @@ void gbmv(backend_selector selector, transpose trans, int64_t int64_t kl, int64_t ku, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, float beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -926,7 +926,7 @@ void gbmv(backend_selector selector, transpose trans, int64_t int64_t kl, int64_t ku, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, double beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -935,7 +935,7 @@ void gbmv(backend_selector selector, transpose trans, int64_t sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } @@ -944,42 +944,42 @@ void gbmv(backend_selector selector, transpose trans, int64_t sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void symm(backend_selector selector, side left_right, uplo upper_lower, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -987,7 +987,7 @@ void symm(backend_selector selector, side left_right, uplo upp int64_t n, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -995,7 +995,7 @@ void symm(backend_selector selector, side left_right, uplo upp int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1003,7 +1003,7 @@ void symm(backend_selector selector, side left_right, uplo upp int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1011,39 +1011,39 @@ void dotc(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy, sycl::buffer, 1> &result) { - oneapi::mkl::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void syr(backend_selector selector, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { - oneapi::mkl::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, + oneapi::math::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1051,7 +1051,7 @@ void trmm(backend_selector selector, side left_right, uplo upp transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } @@ -1059,111 +1059,111 @@ void trmm(backend_selector selector, side left_right, uplo upp transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, + oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m) { - oneapi::mkl::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void rotmg(backend_selector selector, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m) { - oneapi::mkl::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); + oneapi::math::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx); } void copy(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); + oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy_batch(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + oneapi::math::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } void copy_batch(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + oneapi::math::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1171,7 +1171,7 @@ void copy_batch(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, int64_t stridex, sycl::buffer, 1> &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + oneapi::math::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1179,7 +1179,7 @@ void copy_batch(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, int64_t stridex, sycl::buffer, 1> &y, int64_t incy, int64_t stridey, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + oneapi::math::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size); } @@ -1187,7 +1187,7 @@ void hemv(backend_selector selector, uplo upper_lower, int64_t std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::rocblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1195,7 +1195,7 @@ void hemv(backend_selector selector, uplo upper_lower, int64_t std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::rocblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1203,7 +1203,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transp transpose transb, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1211,7 +1211,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transp transpose transb, int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1220,7 +1220,7 @@ void gemmt(backend_selector selector, uplo upper_lower, transp sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1229,138 +1229,138 @@ void gemmt(backend_selector selector, uplo upper_lower, transp sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, float beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void sbmv(backend_selector selector, uplo upper_lower, int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, double beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - oneapi::mkl::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, + oneapi::math::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void spr2(backend_selector selector, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a) { - oneapi::mkl::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, + oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a); } void iamax(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, int64_t n, sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); + oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void rotm(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void rotm(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { - oneapi::mkl::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); + oneapi::math::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } void dot(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void dot(backend_selector selector, int64_t n, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &result) { - oneapi::mkl::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); + oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1369,7 +1369,7 @@ void trsm_batch(backend_selector selector, side left_right, up transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1379,7 +1379,7 @@ void trsm_batch(backend_selector selector, side left_right, up sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1389,7 +1389,7 @@ void trsm_batch(backend_selector selector, side left_right, up sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, + oneapi::math::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1398,7 +1398,7 @@ void her2k(backend_selector selector, uplo upper_lower, transp int64_t n, int64_t k, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, float beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1407,43 +1407,43 @@ void her2k(backend_selector selector, uplo upper_lower, transp sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, double beta, sycl::buffer, 1> &c, int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, + oneapi::math::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - oneapi::mkl::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void rotg(backend_selector selector, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - oneapi::mkl::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); + oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, int64_t n, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, float beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::rocblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } void symv(backend_selector selector, uplo upper_lower, int64_t n, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, double beta, sycl::buffer &y, int64_t incy) { - oneapi::mkl::blas::rocblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, + oneapi::math::blas::rocblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -1451,7 +1451,7 @@ void omatcopy_batch(backend_selector selector, transpose trans std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1459,7 +1459,7 @@ void omatcopy_batch(backend_selector selector, transpose trans std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1468,7 +1468,7 @@ void omatcopy_batch(backend_selector selector, transpose trans sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -1477,21 +1477,21 @@ void omatcopy_batch(backend_selector selector, transpose trans sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, + oneapi::math::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1499,7 +1499,7 @@ void imatcopy_batch(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1507,7 +1507,7 @@ void imatcopy_batch(backend_selector selector, transpose trans std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, + oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } @@ -1516,7 +1516,7 @@ void omatadd_batch(backend_selector selector, transpose transa std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1526,7 +1526,7 @@ void omatadd_batch(backend_selector selector, transpose transa std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1537,7 +1537,7 @@ void omatadd_batch(backend_selector selector, transpose transa std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1549,7 +1549,7 @@ void omatadd_batch(backend_selector selector, transpose transa sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { - oneapi::mkl::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, + oneapi::math::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -1557,28 +1557,28 @@ void omatadd_batch(backend_selector selector, transpose transa void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, + oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb); } @@ -1586,7 +1586,7 @@ void omatcopy2(backend_selector selector, transpose trans, std std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1594,7 +1594,7 @@ void omatcopy2(backend_selector selector, transpose trans, std std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1602,7 +1602,7 @@ void omatcopy2(backend_selector selector, transpose trans, std std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } @@ -1610,35 +1610,35 @@ void omatcopy2(backend_selector selector, transpose trans, std std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { - oneapi::mkl::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, + oneapi::math::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { - oneapi::mkl::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, + oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb); } @@ -1646,7 +1646,7 @@ void omatadd(backend_selector selector, transpose transa, tran std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1654,7 +1654,7 @@ void omatadd(backend_selector selector, transpose transa, tran std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1663,7 +1663,7 @@ void omatadd(backend_selector selector, transpose transa, tran sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1672,7 +1672,7 @@ void omatadd(backend_selector selector, transpose transa, tran sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::mkl::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, + oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } @@ -1681,7 +1681,7 @@ void omatadd(backend_selector selector, transpose transa, tran sycl::event syr2(backend_selector selector, uplo upper_lower, int64_t n, float alpha, const float *x, int64_t incx, const float *y, int64_t incy, float *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1689,21 +1689,21 @@ sycl::event syr2(backend_selector selector, uplo upper_lower, sycl::event syr2(backend_selector selector, uplo upper_lower, int64_t n, double alpha, const double *x, int64_t incx, const double *y, int64_t incy, double *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } sycl::event scal(backend_selector selector, int64_t n, float alpha, float *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } sycl::event scal(backend_selector selector, int64_t n, double alpha, double *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1711,7 +1711,7 @@ sycl::event scal(backend_selector selector, int64_t n, double sycl::event scal(backend_selector selector, int64_t n, std::complex alpha, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1719,7 +1719,7 @@ sycl::event scal(backend_selector selector, int64_t n, std::co sycl::event scal(backend_selector selector, int64_t n, std::complex alpha, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1727,7 +1727,7 @@ sycl::event scal(backend_selector selector, int64_t n, std::co sycl::event scal(backend_selector selector, int64_t n, float alpha, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1735,7 +1735,7 @@ sycl::event scal(backend_selector selector, int64_t n, float a sycl::event scal(backend_selector selector, int64_t n, double alpha, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, dependencies); return done; } @@ -1743,7 +1743,7 @@ sycl::event scal(backend_selector selector, int64_t n, double sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const float *a, int64_t lda, float *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trmv( + auto done = oneapi::math::blas::rocblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1751,7 +1751,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const double *a, int64_t lda, double *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trmv( + auto done = oneapi::math::blas::rocblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1760,7 +1760,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, diag unit_diag, int64_t n, const std::complex *a, int64_t lda, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trmv( + auto done = oneapi::math::blas::rocblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1769,7 +1769,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, diag unit_diag, int64_t n, const std::complex *a, int64_t lda, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trmv( + auto done = oneapi::math::blas::rocblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -1777,7 +1777,7 @@ sycl::event trmv(backend_selector selector, uplo upper_lower, sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const float *a, float *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1785,7 +1785,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const double *a, double *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1793,7 +1793,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const std::complex *a, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1801,7 +1801,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const std::complex *a, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -1809,7 +1809,7 @@ sycl::event tpmv(backend_selector selector, uplo upper_lower, sycl::event spr(backend_selector selector, uplo upper_lower, int64_t n, float alpha, const float *x, int64_t incx, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1817,7 +1817,7 @@ sycl::event spr(backend_selector selector, uplo upper_lower, i sycl::event spr(backend_selector selector, uplo upper_lower, int64_t n, double alpha, const double *x, int64_t incx, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -1827,7 +1827,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, const std::complex *x, int64_t incx, std::complex beta, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1837,7 +1837,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, const std::complex *x, int64_t incx, std::complex beta, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -1845,7 +1845,7 @@ sycl::event hpmv(backend_selector selector, uplo upper_lower, sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, const float *a, int64_t lda, float beta, float *c, int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1853,7 +1853,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, const double *a, int64_t lda, double beta, double *c, int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1862,7 +1862,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, int64_t n, int64_t k, std::complex alpha, const std::complex *a, int64_t lda, std::complex beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1871,7 +1871,7 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, int64_t n, int64_t k, std::complex alpha, const std::complex *a, int64_t lda, std::complex beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -1880,7 +1880,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_ transpose *trans, int64_t *n, int64_t *k, float *alpha, const float **a, int64_t *lda, float *beta, float **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1890,7 +1890,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_ transpose *trans, int64_t *n, int64_t *k, double *alpha, const double **a, int64_t *lda, double *beta, double **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1901,7 +1901,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_ const std::complex **a, int64_t *lda, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1912,7 +1912,7 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_ const std::complex **a, int64_t *lda, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -1923,7 +1923,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_l int64_t lda, int64_t stride_a, float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -1934,7 +1934,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_l int64_t lda, int64_t stride_a, double beta, double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -1946,7 +1946,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_l std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -1958,7 +1958,7 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_l std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syrk_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -1968,7 +1968,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, std::complex alpha, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1977,7 +1977,7 @@ sycl::event her2(backend_selector selector, uplo upper_lower, std::complex alpha, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -1988,7 +1988,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, std::complex *y, int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -1999,7 +1999,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, std::complex *y, int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2007,7 +2007,7 @@ sycl::event hbmv(backend_selector selector, uplo upper_lower, sycl::event rot(backend_selector selector, int64_t n, std::complex *x, int64_t incx, std::complex *y, int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2015,7 +2015,7 @@ sycl::event rot(backend_selector selector, int64_t n, std::com sycl::event rot(backend_selector selector, int64_t n, std::complex *x, int64_t incx, std::complex *y, int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2023,7 +2023,7 @@ sycl::event rot(backend_selector selector, int64_t n, std::com sycl::event rot(backend_selector selector, int64_t n, float *x, int64_t incx, float *y, int64_t incy, float c, float s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2031,7 +2031,7 @@ sycl::event rot(backend_selector selector, int64_t n, float *x sycl::event rot(backend_selector selector, int64_t n, double *x, int64_t incx, double *y, int64_t incy, double c, double s, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, + auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s, dependencies); return done; } @@ -2039,7 +2039,7 @@ sycl::event rot(backend_selector selector, int64_t n, double * sycl::event axpy(backend_selector selector, int64_t n, float alpha, const float *x, int64_t incx, float *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2047,7 +2047,7 @@ sycl::event axpy(backend_selector selector, int64_t n, float a sycl::event axpy(backend_selector selector, int64_t n, double alpha, const double *x, int64_t incx, double *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2055,7 +2055,7 @@ sycl::event axpy(backend_selector selector, int64_t n, double sycl::event axpy(backend_selector selector, int64_t n, std::complex alpha, const std::complex *x, int64_t incx, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2063,7 +2063,7 @@ sycl::event axpy(backend_selector selector, int64_t n, std::co sycl::event axpy(backend_selector selector, int64_t n, std::complex alpha, const std::complex *x, int64_t incx, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, + auto done = oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy, dependencies); return done; } @@ -2072,7 +2072,7 @@ sycl::event axpy_batch(backend_selector selector, int64_t *n, const float **x, int64_t *incx, float **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2081,7 +2081,7 @@ sycl::event axpy_batch(backend_selector selector, int64_t *n, const double **x, int64_t *incx, double **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2090,7 +2090,7 @@ sycl::event axpy_batch(backend_selector selector, int64_t *n, std::complex *alpha, const std::complex **x, int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2099,7 +2099,7 @@ sycl::event axpy_batch(backend_selector selector, int64_t *n, std::complex *alpha, const std::complex **x, int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -2108,7 +2108,7 @@ sycl::event axpy_batch(backend_selector selector, int64_t n, f const float *x, int64_t incx, int64_t stridex, float *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2118,7 +2118,7 @@ sycl::event axpy_batch(backend_selector selector, int64_t n, d const double *x, int64_t incx, int64_t stridex, double *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2128,7 +2128,7 @@ sycl::event axpy_batch(backend_selector selector, int64_t n, std::complex alpha, const std::complex *x, int64_t incx, int64_t stridex, std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2138,7 +2138,7 @@ sycl::event axpy_batch(backend_selector selector, int64_t n, std::complex alpha, const std::complex *x, int64_t incx, int64_t stridex, std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, + auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; @@ -2147,7 +2147,7 @@ sycl::event axpy_batch(backend_selector selector, int64_t n, sycl::event axpby(backend_selector selector, int64_t n, float alpha, const float *x, int64_t incx, const float beta, float *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2155,7 +2155,7 @@ sycl::event axpby(backend_selector selector, int64_t n, float sycl::event axpby(backend_selector selector, int64_t n, double alpha, const double *x, int64_t incx, const double beta, double *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2164,7 +2164,7 @@ sycl::event axpby(backend_selector selector, int64_t n, std::c const std::complex *x, int64_t incx, const std::complex beta, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2173,7 +2173,7 @@ sycl::event axpby(backend_selector selector, int64_t n, std::complex alpha, const std::complex *x, int64_t incx, const std::complex beta, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy, dependencies); return done; } @@ -2182,7 +2182,7 @@ sycl::event gerc(backend_selector selector, int64_t m, int64_t std::complex alpha, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2191,7 +2191,7 @@ sycl::event gerc(backend_selector selector, int64_t m, int64_t std::complex alpha, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2201,7 +2201,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, int64_t ldb, float beta, float *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2211,7 +2211,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, int64_t ldb, double beta, double *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2222,7 +2222,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::complex *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2233,7 +2233,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, std::complex beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2241,7 +2241,7 @@ sycl::event syr2k(backend_selector selector, uplo upper_lower, sycl::event gemv(backend_selector selector, transpose trans, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, const float *x, int64_t incx, float beta, float *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2250,7 +2250,7 @@ sycl::event gemv(backend_selector selector, transpose trans, i double alpha, const double *a, int64_t lda, const double *x, int64_t incx, double beta, double *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2260,7 +2260,7 @@ sycl::event gemv(backend_selector selector, transpose trans, i const std::complex *x, int64_t incx, std::complex beta, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2270,7 +2270,7 @@ sycl::event gemv(backend_selector selector, transpose trans, i const std::complex *x, int64_t incx, std::complex beta, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, + auto done = oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -2280,7 +2280,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tr const float *x, int64_t incx, int64_t stridex, float beta, float *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2291,7 +2291,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tr const double *x, int64_t incx, int64_t stridex, double beta, double *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2303,7 +2303,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tr int64_t stridex, std::complex beta, std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2315,7 +2315,7 @@ sycl::event gemv_batch(backend_selector selector, transpose tr int64_t stridex, std::complex beta, std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; @@ -2325,7 +2325,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *t int64_t *n, float *alpha, const float **a, int64_t *lda, const float **x, int64_t *incx, float *beta, float **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2335,7 +2335,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *t int64_t *n, double *alpha, const double **a, int64_t *lda, const double **x, int64_t *incx, double *beta, double **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2347,7 +2347,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *t std::complex *beta, std::complex **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2359,7 +2359,7 @@ sycl::event gemv_batch(backend_selector selector, transpose *t std::complex *beta, std::complex **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemv_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; @@ -2369,7 +2369,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_ri int64_t n, const float *a, int64_t lda, int64_t stridea, const float *x, int64_t incx, int64_t stridex, float *c, int64_t ldc, int64_t stridec, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2379,7 +2379,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_ri int64_t n, const double *a, int64_t lda, int64_t stridea, const double *x, int64_t incx, int64_t stridex, double *c, int64_t ldc, int64_t stridec, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2390,7 +2390,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_ri const std::complex *x, int64_t incx, int64_t stridex, std::complex *c, int64_t ldc, int64_t stridec, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2401,7 +2401,7 @@ sycl::event dgmm_batch(backend_selector selector, side left_ri const std::complex *x, int64_t incx, int64_t stridex, std::complex *c, int64_t ldc, int64_t stridec, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; @@ -2411,7 +2411,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_r int64_t *n, const float **a, int64_t *lda, const float **x, int64_t *incx, float **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2421,7 +2421,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_r int64_t *n, const double **a, int64_t *lda, const double **x, int64_t *incx, double **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2432,7 +2432,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_r const std::complex **x, int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2443,7 +2443,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_r const std::complex **x, int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dgmm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; @@ -2452,7 +2452,7 @@ sycl::event dgmm_batch(backend_selector selector, side *left_r sycl::event her(backend_selector selector, uplo upper_lower, int64_t n, float alpha, const std::complex *x, int64_t incx, std::complex *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2460,7 +2460,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, i sycl::event her(backend_selector selector, uplo upper_lower, int64_t n, double alpha, const std::complex *x, int64_t incx, std::complex *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -2468,7 +2468,7 @@ sycl::event her(backend_selector selector, uplo upper_lower, i sycl::event hpr(backend_selector selector, uplo upper_lower, int64_t n, float alpha, const std::complex *x, int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } @@ -2476,21 +2476,21 @@ sycl::event hpr(backend_selector selector, uplo upper_lower, i sycl::event hpr(backend_selector selector, uplo upper_lower, int64_t n, double alpha, const std::complex *x, int64_t incx, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, dependencies); return done; } sycl::event iamin(backend_selector selector, int64_t n, const float *x, int64_t incx, int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } sycl::event iamin(backend_selector selector, int64_t n, const double *x, int64_t incx, int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2498,7 +2498,7 @@ sycl::event iamin(backend_selector selector, int64_t n, const sycl::event iamin(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2506,7 +2506,7 @@ sycl::event iamin(backend_selector selector, int64_t n, sycl::event iamin(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2516,7 +2516,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *t const float **a, int64_t *lda, const float **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2527,7 +2527,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *t const double **a, int64_t *lda, const double **b, int64_t *ldb, double *beta, double **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2539,7 +2539,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *t const std::complex **b, int64_t *ldb, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2551,7 +2551,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *t const std::complex **b, int64_t *ldb, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2562,7 +2562,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *t const sycl::half **a, int64_t *lda, const sycl::half **b, int64_t *ldb, sycl::half *beta, sycl::half **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2573,7 +2573,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *t const sycl::half **a, int64_t *lda, const sycl::half **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2584,7 +2584,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *t const std::int8_t **a, int64_t *lda, const std::int8_t **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2595,7 +2595,7 @@ sycl::event gemm_batch(backend_selector selector, transpose *t const std::int8_t **a, int64_t *lda, const std::int8_t **b, int64_t *ldb, float *beta, std::int32_t **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; @@ -2606,7 +2606,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tr const float *a, int64_t lda, int64_t stride_a, const float *b, int64_t ldb, int64_t stride_b, float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2617,7 +2617,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tr const double *a, int64_t lda, int64_t stride_a, const double *b, int64_t ldb, int64_t stride_b, double beta, double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2630,7 +2630,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tr std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2643,7 +2643,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tr int64_t stride_b, std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2655,7 +2655,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tr int64_t ldb, int64_t stride_b, sycl::half beta, sycl::half *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2667,7 +2667,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tr int64_t ldb, int64_t stride_b, float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2679,7 +2679,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tr int64_t ldb, int64_t stride_b, float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2691,7 +2691,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tr int64_t ldb, int64_t stride_b, float beta, std::int32_t *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; @@ -2700,7 +2700,7 @@ sycl::event gemm_batch(backend_selector selector, transpose tr sycl::event spmv(backend_selector selector, uplo upper_lower, int64_t n, float alpha, const float *a, const float *x, int64_t incx, float beta, float *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } @@ -2708,21 +2708,21 @@ sycl::event spmv(backend_selector selector, uplo upper_lower, sycl::event spmv(backend_selector selector, uplo upper_lower, int64_t n, double alpha, const double *a, const double *x, int64_t incx, double beta, double *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } sycl::event swap(backend_selector selector, int64_t n, float *x, int64_t incx, float *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } sycl::event swap(backend_selector selector, int64_t n, double *x, int64_t incx, double *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2730,7 +2730,7 @@ sycl::event swap(backend_selector selector, int64_t n, double sycl::event swap(backend_selector selector, int64_t n, std::complex *x, int64_t incx, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2738,7 +2738,7 @@ sycl::event swap(backend_selector selector, int64_t n, std::co sycl::event swap(backend_selector selector, int64_t n, std::complex *x, int64_t incx, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -2747,7 +2747,7 @@ sycl::event geru(backend_selector selector, int64_t m, int64_t std::complex alpha, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2756,7 +2756,7 @@ sycl::event geru(backend_selector selector, int64_t m, int64_t std::complex alpha, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2764,7 +2764,7 @@ sycl::event geru(backend_selector selector, int64_t m, int64_t sycl::event nrm2(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2772,21 +2772,21 @@ sycl::event nrm2(backend_selector selector, int64_t n, sycl::event nrm2(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } sycl::event nrm2(backend_selector selector, int64_t n, const float *x, int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } sycl::event nrm2(backend_selector selector, int64_t n, const double *x, int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -2796,7 +2796,7 @@ sycl::event gemm(backend_selector selector, transpose transa, const float *b, int64_t ldb, float beta, float *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2806,7 +2806,7 @@ sycl::event gemm(backend_selector selector, transpose transa, const double *b, int64_t ldb, double beta, double *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2817,7 +2817,7 @@ sycl::event gemm(backend_selector selector, transpose transa, int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2828,7 +2828,7 @@ sycl::event gemm(backend_selector selector, transpose transa, int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2838,7 +2838,7 @@ sycl::event gemm(backend_selector selector, transpose transa, int64_t lda, const sycl::half *b, int64_t ldb, sycl::half beta, sycl::half *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2848,7 +2848,7 @@ sycl::event gemm(backend_selector selector, transpose transa, const sycl::half *b, int64_t ldb, float beta, float *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2858,7 +2858,7 @@ sycl::event gemm(backend_selector selector, transpose transa, const bfloat16 *b, int64_t ldb, float beta, float *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -2869,7 +2869,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tra const std::uint8_t *b, int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -2881,7 +2881,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tra const std::int8_t *b, int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -2893,7 +2893,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tra const std::int8_t *b, int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -2905,7 +2905,7 @@ sycl::event gemm_bias(backend_selector selector, transpose tra const std::uint8_t *b, int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, int64_t ldc, const std::int32_t *co, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemm_bias( + auto done = oneapi::math::blas::rocblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; @@ -2915,7 +2915,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, int64_t n, int64_t k, float alpha, const std::complex *a, int64_t lda, float beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::herk( + auto done = oneapi::math::blas::rocblas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -2924,7 +2924,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, int64_t n, int64_t k, double alpha, const std::complex *a, int64_t lda, double beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::herk( + auto done = oneapi::math::blas::rocblas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } @@ -2932,7 +2932,7 @@ sycl::event herk(backend_selector selector, uplo upper_lower, sycl::event ger(backend_selector selector, int64_t m, int64_t n, float alpha, const float *x, int64_t incx, const float *y, int64_t incy, float *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2940,7 +2940,7 @@ sycl::event ger(backend_selector selector, int64_t m, int64_t sycl::event ger(backend_selector selector, int64_t m, int64_t n, double alpha, const double *x, int64_t incx, const double *y, int64_t incy, double *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + auto done = oneapi::math::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } @@ -2948,7 +2948,7 @@ sycl::event ger(backend_selector selector, int64_t m, int64_t sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, float *b, int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -2958,7 +2958,7 @@ sycl::event trsm(backend_selector selector, side left_right, u transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, double *b, int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -2968,7 +2968,7 @@ sycl::event trsm(backend_selector selector, side left_right, u transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -2978,7 +2978,7 @@ sycl::event trsm(backend_selector selector, side left_right, u transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -2989,7 +2989,7 @@ sycl::event trsm_batch(backend_selector selector, side left_ri float alpha, const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3000,7 +3000,7 @@ sycl::event trsm_batch(backend_selector selector, side left_ri double alpha, const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3011,7 +3011,7 @@ sycl::event trsm_batch(backend_selector selector, side left_ri std::complex alpha, const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3022,7 +3022,7 @@ sycl::event trsm_batch(backend_selector selector, side left_ri std::complex alpha, const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3033,7 +3033,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_r float *alpha, const float **a, int64_t *lda, float **b, int64_t *ldb, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3044,7 +3044,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_r double *alpha, const double **a, int64_t *lda, double **b, int64_t *ldb, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3055,7 +3055,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_r std::complex *alpha, const std::complex **a, int64_t *lda, std::complex **b, int64_t *ldb, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3066,7 +3066,7 @@ sycl::event trsm_batch(backend_selector selector, side *left_r std::complex *alpha, const std::complex **a, int64_t *lda, std::complex **b, int64_t *ldb, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsm_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; @@ -3076,7 +3076,7 @@ sycl::event dotu(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3085,7 +3085,7 @@ sycl::event dotu(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3096,7 +3096,7 @@ sycl::event hemm(backend_selector selector, side left_right, u std::complex *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3107,7 +3107,7 @@ sycl::event hemm(backend_selector selector, side left_right, u std::complex *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3116,7 +3116,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, std::complex alpha, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3125,7 +3125,7 @@ sycl::event hpr2(backend_selector selector, uplo upper_lower, std::complex alpha, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3135,7 +3135,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, i int64_t incx, float beta, float *y, int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3145,7 +3145,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, i const double *x, int64_t incx, double beta, double *y, int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3156,7 +3156,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, i std::complex *y, int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3167,7 +3167,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, i std::complex beta, std::complex *y, int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3175,7 +3175,7 @@ sycl::event gbmv(backend_selector selector, transpose trans, i sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, const float *a, int64_t lda, float *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tbmv( + auto done = oneapi::math::blas::rocblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3183,7 +3183,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, const double *a, int64_t lda, double *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tbmv( + auto done = oneapi::math::blas::rocblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3192,7 +3192,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, diag unit_diag, int64_t n, int64_t k, const std::complex *a, int64_t lda, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tbmv( + auto done = oneapi::math::blas::rocblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3201,7 +3201,7 @@ sycl::event tbmv(backend_selector selector, uplo upper_lower, diag unit_diag, int64_t n, int64_t k, const std::complex *a, int64_t lda, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tbmv( + auto done = oneapi::math::blas::rocblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3211,7 +3211,7 @@ sycl::event symm(backend_selector selector, side left_right, u int64_t ldb, float beta, float *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3221,7 +3221,7 @@ sycl::event symm(backend_selector selector, side left_right, u int64_t ldb, double beta, double *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3232,7 +3232,7 @@ sycl::event symm(backend_selector selector, side left_right, u std::complex *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3243,7 +3243,7 @@ sycl::event symm(backend_selector selector, side left_right, u std::complex *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, + oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3252,7 +3252,7 @@ sycl::event dotc(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3261,7 +3261,7 @@ sycl::event dotc(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, const std::complex *y, int64_t incy, std::complex *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3269,7 +3269,7 @@ sycl::event dotc(backend_selector selector, int64_t n, sycl::event syr(backend_selector selector, uplo upper_lower, int64_t n, float alpha, const float *x, int64_t incx, float *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3277,7 +3277,7 @@ sycl::event syr(backend_selector selector, uplo upper_lower, i sycl::event syr(backend_selector selector, uplo upper_lower, int64_t n, double alpha, const double *x, int64_t incx, double *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } @@ -3285,7 +3285,7 @@ sycl::event syr(backend_selector selector, uplo upper_lower, i sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, float *b, int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3295,7 +3295,7 @@ sycl::event trmm(backend_selector selector, side left_right, u transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, double *b, int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3305,7 +3305,7 @@ sycl::event trmm(backend_selector selector, side left_right, u transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3315,7 +3315,7 @@ sycl::event trmm(backend_selector selector, side left_right, u transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, + auto done = oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; @@ -3323,14 +3323,14 @@ sycl::event trmm(backend_selector selector, side left_right, u sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + auto done = oneapi::math::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } sycl::event rotmg(backend_selector selector, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + auto done = oneapi::math::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, dependencies); return done; } @@ -3338,7 +3338,7 @@ sycl::event rotmg(backend_selector selector, double *d1, doubl sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const float *a, float *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3346,7 +3346,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const double *a, double *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3354,7 +3354,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const std::complex *a, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3362,7 +3362,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const std::complex *a, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, + auto done = oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } @@ -3370,7 +3370,7 @@ sycl::event tpsv(backend_selector selector, uplo upper_lower, sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const float *a, int64_t lda, float *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsv( + auto done = oneapi::math::blas::rocblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3378,7 +3378,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const double *a, int64_t lda, double *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsv( + auto done = oneapi::math::blas::rocblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3387,7 +3387,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, diag unit_diag, int64_t n, const std::complex *a, int64_t lda, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsv( + auto done = oneapi::math::blas::rocblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3396,7 +3396,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, diag unit_diag, int64_t n, const std::complex *a, int64_t lda, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::trsv( + auto done = oneapi::math::blas::rocblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } @@ -3404,7 +3404,7 @@ sycl::event trsv(backend_selector selector, uplo upper_lower, sycl::event copy(backend_selector selector, int64_t n, const float *x, int64_t incx, float *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3412,7 +3412,7 @@ sycl::event copy(backend_selector selector, int64_t n, const f sycl::event copy(backend_selector selector, int64_t n, const double *x, int64_t incx, double *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3420,7 +3420,7 @@ sycl::event copy(backend_selector selector, int64_t n, const d sycl::event copy(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3428,7 +3428,7 @@ sycl::event copy(backend_selector selector, int64_t n, sycl::event copy(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, dependencies); return done; } @@ -3436,7 +3436,7 @@ sycl::event copy(backend_selector selector, int64_t n, sycl::event copy_batch(backend_selector selector, int64_t *n, const float **x, int64_t *incx, float **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3444,7 +3444,7 @@ sycl::event copy_batch(backend_selector selector, int64_t *n, sycl::event copy_batch(backend_selector selector, int64_t *n, const double **x, int64_t *incx, double **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3453,7 +3453,7 @@ sycl::event copy_batch(backend_selector selector, int64_t *n, const std::complex **x, int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3462,7 +3462,7 @@ sycl::event copy_batch(backend_selector selector, int64_t *n, const std::complex **x, int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } @@ -3470,7 +3470,7 @@ sycl::event copy_batch(backend_selector selector, int64_t *n, sycl::event copy_batch(backend_selector selector, int64_t n, const float *x, int64_t incx, int64_t stridex, float *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3478,7 +3478,7 @@ sycl::event copy_batch(backend_selector selector, int64_t n, c sycl::event copy_batch(backend_selector selector, int64_t n, const double *x, int64_t incx, int64_t stridex, double *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3487,7 +3487,7 @@ sycl::event copy_batch(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, int64_t stridex, std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3496,7 +3496,7 @@ sycl::event copy_batch(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, int64_t stridex, std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::copy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } @@ -3506,7 +3506,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, const std::complex *x, int64_t incx, std::complex beta, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::hemv( + auto done = oneapi::math::blas::rocblas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3516,7 +3516,7 @@ sycl::event hemv(backend_selector selector, uplo upper_lower, const std::complex *x, int64_t incx, std::complex beta, std::complex *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::hemv( + auto done = oneapi::math::blas::rocblas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3525,7 +3525,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transb, int64_t n, int64_t k, float alpha, const float *a, int64_t lda, const float *b, int64_t ldb, float beta, float *c, int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3535,7 +3535,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transb, int64_t n, int64_t k, double alpha, const double *a, int64_t lda, const double *b, int64_t ldb, double beta, double *c, int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3546,7 +3546,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, const std::complex *a, int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3557,7 +3557,7 @@ sycl::event gemmt(backend_selector selector, uplo upper_lower, const std::complex *a, int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, + auto done = oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; @@ -3567,7 +3567,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, int64_t k, float alpha, const float *a, int64_t lda, const float *x, int64_t incx, float beta, float *y, int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3577,7 +3577,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, int64_t incx, double beta, double *y, int64_t incy, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + oneapi::math::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3585,7 +3585,7 @@ sycl::event sbmv(backend_selector selector, uplo upper_lower, sycl::event asum(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3593,21 +3593,21 @@ sycl::event asum(backend_selector selector, int64_t n, sycl::event asum(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } sycl::event asum(backend_selector selector, int64_t n, const float *x, int64_t incx, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } sycl::event asum(backend_selector selector, int64_t n, const double *x, int64_t incx, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3615,7 +3615,7 @@ sycl::event asum(backend_selector selector, int64_t n, const d sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, const float *a, int64_t lda, float *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tbsv( + auto done = oneapi::math::blas::rocblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3623,7 +3623,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, const double *a, int64_t lda, double *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tbsv( + auto done = oneapi::math::blas::rocblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3632,7 +3632,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, diag unit_diag, int64_t n, int64_t k, const std::complex *a, int64_t lda, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tbsv( + auto done = oneapi::math::blas::rocblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3641,7 +3641,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, diag unit_diag, int64_t n, int64_t k, const std::complex *a, int64_t lda, std::complex *x, int64_t incx, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::tbsv( + auto done = oneapi::math::blas::rocblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } @@ -3649,7 +3649,7 @@ sycl::event tbsv(backend_selector selector, uplo upper_lower, sycl::event spr2(backend_selector selector, uplo upper_lower, int64_t n, float alpha, const float *x, int64_t incx, const float *y, int64_t incy, float *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } @@ -3657,21 +3657,21 @@ sycl::event spr2(backend_selector selector, uplo upper_lower, sycl::event spr2(backend_selector selector, uplo upper_lower, int64_t n, double alpha, const double *x, int64_t incx, const double *y, int64_t incy, double *a, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, + auto done = oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } sycl::event iamax(backend_selector selector, int64_t n, const float *x, int64_t incx, int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } sycl::event iamax(backend_selector selector, int64_t n, const double *x, int64_t incx, int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3679,7 +3679,7 @@ sycl::event iamax(backend_selector selector, int64_t n, const sycl::event iamax(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3687,7 +3687,7 @@ sycl::event iamax(backend_selector selector, int64_t n, sycl::event iamax(backend_selector selector, int64_t n, const std::complex *x, int64_t incx, int64_t *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, + auto done = oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, dependencies); return done; } @@ -3695,7 +3695,7 @@ sycl::event iamax(backend_selector selector, int64_t n, sycl::event rotm(backend_selector selector, int64_t n, float *x, int64_t incx, float *y, int64_t incy, float *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3703,7 +3703,7 @@ sycl::event rotm(backend_selector selector, int64_t n, float * sycl::event rotm(backend_selector selector, int64_t n, double *x, int64_t incx, double *y, int64_t incy, double *param, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param, dependencies); return done; } @@ -3711,14 +3711,14 @@ sycl::event rotm(backend_selector selector, int64_t n, double sycl::event rotg(backend_selector selector, float *a, float *b, float *c, float *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event rotg(backend_selector selector, double *a, double *b, double *c, double *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3726,7 +3726,7 @@ sycl::event rotg(backend_selector selector, std::complex *b, float *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } @@ -3734,14 +3734,14 @@ sycl::event rotg(backend_selector selector, std::complex *b, double *c, std::complex *s, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); + oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event sdsdot(backend_selector selector, int64_t n, float sb, const float *x, int64_t incx, const float *y, int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, + auto done = oneapi::math::blas::rocblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result, dependencies); return done; } @@ -3752,7 +3752,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, std::complex *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3763,7 +3763,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, std::complex *c, int64_t ldc, const std::vector &dependencies) { auto done = - oneapi::mkl::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + oneapi::math::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } @@ -3771,7 +3771,7 @@ sycl::event her2k(backend_selector selector, uplo upper_lower, sycl::event dot(backend_selector selector, int64_t n, const float *x, int64_t incx, const float *y, int64_t incy, float *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3779,7 +3779,7 @@ sycl::event dot(backend_selector selector, int64_t n, const fl sycl::event dot(backend_selector selector, int64_t n, const double *x, int64_t incx, const double *y, int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3787,7 +3787,7 @@ sycl::event dot(backend_selector selector, int64_t n, const do sycl::event dot(backend_selector selector, int64_t n, const float *x, int64_t incx, const float *y, int64_t incy, double *result, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, + auto done = oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result, dependencies); return done; } @@ -3795,7 +3795,7 @@ sycl::event dot(backend_selector selector, int64_t n, const fl sycl::event symv(backend_selector selector, uplo upper_lower, int64_t n, float alpha, const float *a, int64_t lda, const float *x, int64_t incx, float beta, float *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::symv( + auto done = oneapi::math::blas::rocblas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3804,7 +3804,7 @@ sycl::event symv(backend_selector selector, uplo upper_lower, double alpha, const double *a, int64_t lda, const double *x, int64_t incx, double beta, double *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::symv( + auto done = oneapi::math::blas::rocblas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } @@ -3814,7 +3814,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpos std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3825,7 +3825,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpos std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3836,7 +3836,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpos const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3847,7 +3847,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpos const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; @@ -3857,7 +3857,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpos std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -3866,7 +3866,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpos std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -3876,7 +3876,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpos std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -3886,7 +3886,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpos std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } @@ -3897,7 +3897,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -3909,7 +3909,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -3922,7 +3922,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -3935,7 +3935,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatadd_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; @@ -3944,7 +3944,7 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -3952,7 +3952,7 @@ sycl::event omatcopy(backend_selector selector, transpose tran sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -3961,7 +3961,7 @@ sycl::event omatcopy(backend_selector selector, transpose tran std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -3970,7 +3970,7 @@ sycl::event omatcopy(backend_selector selector, transpose tran std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } @@ -3979,7 +3979,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tra std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -3988,7 +3988,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tra std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -3998,7 +3998,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tra std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4008,7 +4008,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tra std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy2( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } @@ -4016,7 +4016,7 @@ sycl::event omatcopy2(backend_selector selector, transpose tra sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4024,7 +4024,7 @@ sycl::event imatcopy(backend_selector selector, transpose tran sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4033,7 +4033,7 @@ sycl::event imatcopy(backend_selector selector, transpose tran std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4042,7 +4042,7 @@ sycl::event imatcopy(backend_selector selector, transpose tran std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } @@ -4051,7 +4051,7 @@ sycl::event omatadd(backend_selector selector, transpose trans std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + auto done = oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; @@ -4061,7 +4061,7 @@ sycl::event omatadd(backend_selector selector, transpose trans std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + auto done = oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; @@ -4072,7 +4072,7 @@ sycl::event omatadd(backend_selector selector, transpose trans const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + auto done = oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; @@ -4083,7 +4083,7 @@ sycl::event omatadd(backend_selector selector, transpose trans const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + auto done = oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; @@ -4094,7 +4094,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpos std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4105,7 +4105,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpos std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4117,7 +4117,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpos std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4129,7 +4129,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpos std::complex** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::omatcopy_batch( + auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; @@ -4140,7 +4140,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpos std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4151,7 +4151,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpos std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4162,7 +4162,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpos std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; @@ -4173,7 +4173,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpos std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::mkl::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; diff --git a/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp b/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp index e2e8c8f97..45f4f888d 100644 --- a/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp +++ b/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp @@ -32,12 +32,12 @@ #include "oneapi/math/detail/config.hpp" namespace oneapi { -namespace mkl { -using oneapi::mkl::diag; -using oneapi::mkl::offset; -using oneapi::mkl::side; -using oneapi::mkl::transpose; -using oneapi::mkl::uplo; +namespace math { +using oneapi::math::diag; +using oneapi::math::offset; +using oneapi::math::side; +using oneapi::math::transpose; +using oneapi::math::uplo; namespace blas { namespace rocblas { namespace column_major { @@ -52,7 +52,7 @@ namespace row_major { } //namespace row_major } //namespace rocblas } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_ONEMATH_BLAS_ROCBLAS_HPP_ diff --git a/include/oneapi/math/detail/backend_selector.hpp b/include/oneapi/math/detail/backend_selector.hpp index 688e8934c..905d3ac01 100644 --- a/include/oneapi/math/detail/backend_selector.hpp +++ b/include/oneapi/math/detail/backend_selector.hpp @@ -25,7 +25,7 @@ #include "oneapi/math/detail/backend_selector_predicates.hpp" namespace oneapi { -namespace mkl { +namespace math { template class backend_selector { @@ -41,7 +41,7 @@ class backend_selector { sycl::queue queue_; }; -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_BACKEND_SELECTOR_HPP_ diff --git a/include/oneapi/math/detail/backend_selector_predicates.hpp b/include/oneapi/math/detail/backend_selector_predicates.hpp index ec995121e..3a08f980e 100644 --- a/include/oneapi/math/detail/backend_selector_predicates.hpp +++ b/include/oneapi/math/detail/backend_selector_predicates.hpp @@ -32,7 +32,7 @@ #include "oneapi/math/detail/get_device_id.hpp" namespace oneapi { -namespace mkl { +namespace math { template inline void backend_selector_precondition(sycl::queue&) {} @@ -144,7 +144,7 @@ inline void backend_selector_precondition(sycl::queue& queue } #endif } -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_BACKEND_SELECTOR_PREDICATES_HPP_ diff --git a/include/oneapi/math/detail/backends.hpp b/include/oneapi/math/detail/backends.hpp index 0925149d1..2c5c08ed1 100644 --- a/include/oneapi/math/detail/backends.hpp +++ b/include/oneapi/math/detail/backends.hpp @@ -24,7 +24,7 @@ #include namespace oneapi { -namespace mkl { +namespace math { enum class backend { mklcpu, @@ -55,7 +55,7 @@ static backendmap backend_map = { { backend::portfft, "portfft" }, { backend::unsupported, "unsupported" } }; -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_ONEMATH_BACKENDS_HPP_ diff --git a/include/oneapi/math/detail/backends_table.hpp b/include/oneapi/math/detail/backends_table.hpp index 4ed59b679..2b931527e 100644 --- a/include/oneapi/math/detail/backends_table.hpp +++ b/include/oneapi/math/detail/backends_table.hpp @@ -38,7 +38,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { enum class device : uint16_t { x86cpu, intelgpu, nvidiagpu, amdgpu, generic_device }; enum class domain : uint16_t { blas, dft, lapack, rng, sparse_blas }; @@ -209,7 +209,7 @@ static std::map table_names = { { domain::blas, "mkl_blas_t { domain::sparse_blas, "mkl_sparse_blas_table" } }; -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_ONEMATH_BACKENDS_TABLE_HPP_ diff --git a/include/oneapi/math/detail/exceptions.hpp b/include/oneapi/math/detail/exceptions.hpp index f70321a8d..0a1fc0a73 100644 --- a/include/oneapi/math/detail/exceptions.hpp +++ b/include/oneapi/math/detail/exceptions.hpp @@ -27,33 +27,33 @@ #include "oneapi/math/exceptions.hpp" namespace oneapi { -namespace mkl { +namespace math { -class backend_not_found : public oneapi::mkl::exception { +class backend_not_found : public oneapi::math::exception { public: backend_not_found(const std::string &info = "") - : oneapi::mkl::exception( + : oneapi::math::exception( "", "", ((info.length() != 0) ? info : "Couldn't load selected backend")) {} }; -class function_not_found : public oneapi::mkl::exception { +class function_not_found : public oneapi::math::exception { public: function_not_found(const std::string &info = "") - : oneapi::mkl::exception( + : oneapi::math::exception( "", "", ((info.length() != 0) ? info : "Couldn't load functions from selected backend")) { } }; -class specification_mismatch : public oneapi::mkl::exception { +class specification_mismatch : public oneapi::math::exception { public: specification_mismatch(const std::string &info = "") - : oneapi::mkl::exception( + : oneapi::math::exception( "", "", ((info.length() != 0) ? info : "Loaded oneMath specification version mismatch")) {} }; -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ONEMATH_DETAIL_EXCEPTIONS_HPP_ diff --git a/include/oneapi/math/detail/get_device_id.hpp b/include/oneapi/math/detail/get_device_id.hpp index 8521fb5b8..c81f96c13 100644 --- a/include/oneapi/math/detail/get_device_id.hpp +++ b/include/oneapi/math/detail/get_device_id.hpp @@ -38,10 +38,10 @@ #endif namespace oneapi { -namespace mkl { +namespace math { -inline oneapi::mkl::device get_device_id(sycl::queue &queue) { - oneapi::mkl::device device_id; +inline oneapi::math::device get_device_id(sycl::queue &queue) { + oneapi::math::device device_id; if (queue.get_device().is_cpu()) device_id = device::x86cpu; #ifdef __HIPSYCL__ @@ -67,7 +67,7 @@ inline oneapi::mkl::device get_device_id(sycl::queue &queue) { return device_id; } -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_ONEMATH_GET_DEVICE_ID_HPP_ diff --git a/include/oneapi/math/dft/backward.hpp b/include/oneapi/math/dft/backward.hpp index 2cff87055..50c9f4234 100644 --- a/include/oneapi/math/dft/backward.hpp +++ b/include/oneapi/math/dft/backward.hpp @@ -28,7 +28,7 @@ #include "detail/types_impl.hpp" -namespace oneapi::mkl::dft { +namespace oneapi::math::dft { //Buffer version //In-place transform @@ -159,6 +159,6 @@ sycl::event compute_backward(descriptor_type &desc, input_type *in_re, input_typ reinterpret_cast(out_re), reinterpret_cast(out_im), dependencies); } -} // namespace oneapi::mkl::dft +} // namespace oneapi::math::dft #endif // _ONEMATH_DFT_BACKWARD_HPP_ diff --git a/include/oneapi/math/dft/descriptor.hpp b/include/oneapi/math/dft/descriptor.hpp index 3d1fa1996..e12d4ced1 100644 --- a/include/oneapi/math/dft/descriptor.hpp +++ b/include/oneapi/math/dft/descriptor.hpp @@ -24,7 +24,7 @@ #include "types.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace dft { /** The detail namespace is required since the MKLGPU backend uses identical names and function signatures in many places. **/ @@ -32,7 +32,7 @@ names and function signatures in many places. **/ template using descriptor = detail::descriptor; } // namespace dft -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ONEMATH_DFT_DESCRIPTOR_HPP_ diff --git a/include/oneapi/math/dft/detail/commit_impl.hpp b/include/oneapi/math/dft/detail/commit_impl.hpp index 1a2c013bc..d3f7d7d8d 100644 --- a/include/oneapi/math/dft/detail/commit_impl.hpp +++ b/include/oneapi/math/dft/detail/commit_impl.hpp @@ -29,11 +29,11 @@ #include "descriptor_impl.hpp" #include "external_workspace_helper.hpp" -namespace oneapi::mkl { +namespace oneapi::math { enum class backend; } -namespace oneapi::mkl::dft::detail { +namespace oneapi::math::dft::detail { template class dft_values; @@ -41,10 +41,10 @@ class dft_values; template class commit_impl { sycl::queue queue_; - mkl::backend backend_; + math::backend backend_; public: - using descriptor_type = typename oneapi::mkl::dft::detail::descriptor; + using descriptor_type = typename oneapi::math::dft::detail::descriptor; using fwd_type = typename descriptor_info::forward_type; using bwd_type = typename descriptor_info::backward_type; using scalar_type = typename descriptor_info::scalar_type; @@ -53,7 +53,7 @@ class commit_impl { external_workspace_helper external_workspace_helper_; public: - commit_impl(sycl::queue queue, mkl::backend backend, + commit_impl(sycl::queue queue, math::backend backend, const dft::detail::dft_values &config_values) : queue_(queue), backend_(backend), @@ -69,7 +69,7 @@ class commit_impl { return queue_; } - mkl::backend get_backend() const noexcept { + math::backend get_backend() const noexcept { return backend_; } @@ -179,6 +179,6 @@ class commit_impl { } }; -} // namespace oneapi::mkl::dft::detail +} // namespace oneapi::math::dft::detail #endif //_ONEMATH_DFT_COMMIT_IMPL_HPP_ diff --git a/include/oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp b/include/oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp index d053f7f55..c8b463d76 100644 --- a/include/oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp +++ b/include/oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp @@ -29,10 +29,10 @@ #include "oneapi/math/detail/export.hpp" #include "oneapi/math/dft/detail/types_impl.hpp" -namespace oneapi::mkl::dft::cufft { +namespace oneapi::math::dft::cufft { #include "oneapi/math/dft/detail/dft_ct.hxx" -} // namespace oneapi::mkl::dft::cufft +} // namespace oneapi::math::dft::cufft #endif // _ONEMATH_DFT_CUFFT_HPP_ diff --git a/include/oneapi/math/dft/detail/descriptor_impl.hpp b/include/oneapi/math/dft/detail/descriptor_impl.hpp index e5d7e3087..6267eb86d 100644 --- a/include/oneapi/math/dft/detail/descriptor_impl.hpp +++ b/include/oneapi/math/dft/detail/descriptor_impl.hpp @@ -33,7 +33,7 @@ #include "oneapi/math/dft/detail/commit_impl.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace dft { namespace detail { // Forward declaration: @@ -118,7 +118,7 @@ inline commit_impl* get_commit(descriptor& desc) { } // namespace detail } // namespace dft -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ONEMATH_DFT_DETAIL_DESCRIPTOR_IMPL_HPP_ diff --git a/include/oneapi/math/dft/detail/dft_loader.hpp b/include/oneapi/math/dft/detail/dft_loader.hpp index a04f3e3cf..41def11db 100644 --- a/include/oneapi/math/dft/detail/dft_loader.hpp +++ b/include/oneapi/math/dft/detail/dft_loader.hpp @@ -30,7 +30,7 @@ #include "oneapi/math/dft/detail/types_impl.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace dft { namespace detail { @@ -46,7 +46,7 @@ ONEMATH_EXPORT commit_impl* create_commit(const descriptor } // namespace detail } // namespace dft -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_DFT_LOADER_HPP_ diff --git a/include/oneapi/math/dft/detail/external_workspace_helper.hpp b/include/oneapi/math/dft/detail/external_workspace_helper.hpp index b0ae40966..24bdf93c5 100644 --- a/include/oneapi/math/dft/detail/external_workspace_helper.hpp +++ b/include/oneapi/math/dft/detail/external_workspace_helper.hpp @@ -31,7 +31,7 @@ #include "oneapi/math/dft/detail/commit_impl.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace dft { namespace detail { @@ -90,7 +90,7 @@ class external_workspace_helper { */ void set_workspace_throw(commit_impl_t& committed_desc, scalar_t* usm_workspace) { if (get_rqd_workspace_bytes(committed_desc) > 0 && usm_workspace == nullptr) { - throw mkl::invalid_argument("DFT", "set_workspace", + throw math::invalid_argument("DFT", "set_workspace", "Backend expected a non-null workspace pointer."); } m_ext_workspace_rqd = true; @@ -105,11 +105,11 @@ class external_workspace_helper { sycl::buffer& buffer_workspace) { if (static_cast(get_rqd_workspace_bytes(committed_desc)) / sizeof(scalar_t) > buffer_workspace.size()) { - throw mkl::invalid_argument("DFT", "set_workspace", "Provided workspace is too small"); + throw math::invalid_argument("DFT", "set_workspace", "Provided workspace is too small"); return; } if (buffer_workspace.is_sub_buffer()) { - throw mkl::invalid_argument("DFT", "set_workspace", + throw math::invalid_argument("DFT", "set_workspace", "Cannot use sub-buffers for workspace"); return; } @@ -137,7 +137,7 @@ class external_workspace_helper { } } else { - throw mkl::invalid_argument( + throw math::invalid_argument( "DFT", function_name, "Buffer external workspace must be used with buffer compute calls"); } @@ -169,7 +169,7 @@ class external_workspace_helper { */ void usm_compute_call_throw(const char* function_name) const { if (m_ext_workspace_rqd && m_workspace_type != ext_workspace_type::usm) { - throw mkl::invalid_argument( + throw math::invalid_argument( "DFT", function_name, "USM external workspace must be used with usm compute calls"); } } @@ -179,7 +179,7 @@ class external_workspace_helper { */ void buffer_compute_call_throw(const char* function_name) const { if (m_ext_workspace_rqd && m_workspace_type != ext_workspace_type::buffer) { - throw mkl::invalid_argument( + throw math::invalid_argument( "DFT", function_name, "Buffer external workspace must be used with buffer compute calls"); } @@ -188,7 +188,7 @@ class external_workspace_helper { } // namespace detail } // namespace dft -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_DFT_EXTERNAL_WORKSPACE_HELPER_HPP_ diff --git a/include/oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp b/include/oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp index 466bb59db..9cc8ef8c1 100644 --- a/include/oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp +++ b/include/oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp @@ -29,10 +29,10 @@ #include "oneapi/math/detail/export.hpp" #include "oneapi/math/dft/detail/types_impl.hpp" -namespace oneapi::mkl::dft::mklcpu { +namespace oneapi::math::dft::mklcpu { #include "oneapi/math/dft/detail/dft_ct.hxx" -} // namespace oneapi::mkl::dft::mklcpu +} // namespace oneapi::math::dft::mklcpu #endif // _ONEMATH_DFT_MKLCPU_HPP_ diff --git a/include/oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp b/include/oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp index 97dfd188a..3a2325178 100644 --- a/include/oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp +++ b/include/oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp @@ -29,10 +29,10 @@ #include "oneapi/math/detail/export.hpp" #include "oneapi/math/dft/detail/types_impl.hpp" -namespace oneapi::mkl::dft::mklgpu { +namespace oneapi::math::dft::mklgpu { #include "oneapi/math/dft/detail/dft_ct.hxx" -} // namespace oneapi::mkl::dft::mklgpu +} // namespace oneapi::math::dft::mklgpu #endif // _ONEMATH_DFT_MKLGPU_HPP_ diff --git a/include/oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp b/include/oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp index 9ad6d762f..3739bc1aa 100644 --- a/include/oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp +++ b/include/oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp @@ -29,11 +29,11 @@ #include "oneapi/math/detail/export.hpp" #include "oneapi/math/dft/detail/types_impl.hpp" -namespace oneapi::mkl::dft::portfft { +namespace oneapi::math::dft::portfft { // We don't need the forward declarations of compute_xxxward templates (just need the create_commit template), but it doesn't hurt and keeps things simple. #include "oneapi/math/dft/detail/dft_ct.hxx" -} // namespace oneapi::mkl::dft::portfft +} // namespace oneapi::math::dft::portfft #endif // _ONEMATH_DFT_PORTFFT_HPP_ diff --git a/include/oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp b/include/oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp index 0c3f79182..fb5b4b276 100644 --- a/include/oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp +++ b/include/oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp @@ -29,10 +29,10 @@ #include "oneapi/math/detail/export.hpp" #include "oneapi/math/dft/detail/types_impl.hpp" -namespace oneapi::mkl::dft::rocfft { +namespace oneapi::math::dft::rocfft { #include "oneapi/math/dft/detail/dft_ct.hxx" -} // namespace oneapi::mkl::dft::rocfft +} // namespace oneapi::math::dft::rocfft #endif // _ONEMATH_DFT_ROCFFT_HPP_ diff --git a/include/oneapi/math/dft/detail/types_impl.hpp b/include/oneapi/math/dft/detail/types_impl.hpp index aa9096301..da17c762c 100644 --- a/include/oneapi/math/dft/detail/types_impl.hpp +++ b/include/oneapi/math/dft/detail/types_impl.hpp @@ -32,7 +32,7 @@ #include namespace oneapi { -namespace mkl { +namespace math { namespace dft { namespace detail { @@ -228,7 +228,7 @@ class dft_values { } // namespace detail } // namespace dft -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_DETAIL_TYPES_IMPL_HPP_ diff --git a/include/oneapi/math/dft/forward.hpp b/include/oneapi/math/dft/forward.hpp index 457e7c723..67f106923 100644 --- a/include/oneapi/math/dft/forward.hpp +++ b/include/oneapi/math/dft/forward.hpp @@ -28,7 +28,7 @@ #include "detail/types_impl.hpp" -namespace oneapi::mkl::dft { +namespace oneapi::math::dft { //Buffer version @@ -156,6 +156,6 @@ sycl::event compute_forward(descriptor_type &desc, input_type *in_re, input_type reinterpret_cast(out_re), reinterpret_cast(out_im), dependencies); } -} // namespace oneapi::mkl::dft +} // namespace oneapi::math::dft #endif // _ONEMATH_DFT_FORWARD_HPP_ diff --git a/include/oneapi/math/dft/types.hpp b/include/oneapi/math/dft/types.hpp index 106faeb33..c7529fe76 100644 --- a/include/oneapi/math/dft/types.hpp +++ b/include/oneapi/math/dft/types.hpp @@ -23,7 +23,7 @@ #include "detail/types_impl.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace dft { /** The detail namespace is required since the MKLGPU backend uses identical @@ -36,7 +36,7 @@ using config_value = detail::config_value; using DFT_ERROR = detail::DFT_ERROR; } // namespace dft -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_TYPES_HPP_ diff --git a/include/oneapi/math/exceptions.hpp b/include/oneapi/math/exceptions.hpp index 3a89ba56c..babb3af0a 100644 --- a/include/oneapi/math/exceptions.hpp +++ b/include/oneapi/math/exceptions.hpp @@ -33,7 +33,7 @@ // These are oneAPI oneMath Specification exceptions namespace oneapi { -namespace mkl { +namespace math { class exception : public std::exception { std::string msg_; @@ -52,79 +52,79 @@ class exception : public std::exception { } }; -class unsupported_device : public oneapi::mkl::exception { +class unsupported_device : public oneapi::math::exception { public: unsupported_device(const std::string &domain, const std::string &function, const sycl::device &device) - : oneapi::mkl::exception( + : oneapi::math::exception( domain, function, device.get_info() + " is not supported") {} }; -class host_bad_alloc : public oneapi::mkl::exception { +class host_bad_alloc : public oneapi::math::exception { public: host_bad_alloc(const std::string &domain, const std::string &function) - : oneapi::mkl::exception(domain, function, "cannot allocate memory on host") {} + : oneapi::math::exception(domain, function, "cannot allocate memory on host") {} }; -class device_bad_alloc : public oneapi::mkl::exception { +class device_bad_alloc : public oneapi::math::exception { public: device_bad_alloc(const std::string &domain, const std::string &function, const sycl::device &device) - : oneapi::mkl::exception( + : oneapi::math::exception( domain, function, "cannot allocate memory on " + device.get_info()) {} }; -class unimplemented : public oneapi::mkl::exception { +class unimplemented : public oneapi::math::exception { public: unimplemented(const std::string &domain, const std::string &function, const std::string &info = "") - : oneapi::mkl::exception(domain, function, "function is not implemented " + info) {} + : oneapi::math::exception(domain, function, "function is not implemented " + info) {} }; -class invalid_argument : public oneapi::mkl::exception { +class invalid_argument : public oneapi::math::exception { public: invalid_argument(const std::string &domain, const std::string &function, const std::string &info = "") - : oneapi::mkl::exception(domain, function, "invalid argument " + info) {} + : oneapi::math::exception(domain, function, "invalid argument " + info) {} }; -class uninitialized : public oneapi::mkl::exception { +class uninitialized : public oneapi::math::exception { public: uninitialized(const std::string &domain, const std::string &function, const std::string &info = "") - : oneapi::mkl::exception(domain, function, + : oneapi::math::exception(domain, function, "handle/descriptor is not initialized " + info) {} }; -class computation_error : public oneapi::mkl::exception { +class computation_error : public oneapi::math::exception { public: computation_error(const std::string &domain, const std::string &function, const std::string &info = "") - : oneapi::mkl::exception( + : oneapi::math::exception( domain, function, "computation error" + ((info.length() != 0) ? (": " + info) : "")) {} }; -class batch_error : public oneapi::mkl::exception { +class batch_error : public oneapi::math::exception { public: batch_error(const std::string &domain, const std::string &function, const std::string &info = "") - : oneapi::mkl::exception(domain, function, + : oneapi::math::exception(domain, function, "batch error" + ((info.length() != 0) ? (": " + info) : "")) {} }; -class library_not_found : public oneapi::mkl::exception { +class library_not_found : public oneapi::math::exception { public: library_not_found(const std::string &domain, const std::string &function, const std::string &info = "") - : oneapi::mkl::exception( + : oneapi::math::exception( domain, function, "library not found" + ((info.length() != 0) ? (": " + info) : "")) {} }; -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ONEMATH_EXCEPTIONS_HPP_ diff --git a/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hpp b/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hpp index b25ec1c9e..d0ed74581 100644 --- a/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hpp +++ b/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hpp @@ -34,7 +34,7 @@ #include "oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { #define LAPACK_BACKEND cusolver @@ -42,7 +42,7 @@ namespace lapack { #undef LAPACK_BACKEND } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_DETAIL_CUSOLVER_LAPACK_CT_HPP_ diff --git a/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hxx b/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hxx index cd1d76765..175e4500e 100644 --- a/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hxx +++ b/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hxx @@ -26,7 +26,7 @@ static inline void gebrd(backend_selector selector, std::int6 sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, @@ -34,7 +34,7 @@ static inline void gebrd(backend_selector selector, std::int6 sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, @@ -42,7 +42,7 @@ static inline void gebrd(backend_selector selector, std::int6 sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, @@ -52,21 +52,21 @@ static inline void gebrd(backend_selector selector, std::int6 sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, @@ -74,7 +74,7 @@ static inline void gerqf(backend_selector selector, std::int6 sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, @@ -82,7 +82,7 @@ static inline void gerqf(backend_selector selector, std::int6 sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, @@ -90,21 +90,21 @@ static inline void geqrf(backend_selector selector, std::int6 sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, @@ -112,7 +112,7 @@ static inline void geqrf(backend_selector selector, std::int6 sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, @@ -120,21 +120,21 @@ static inline void getrf(backend_selector selector, std::int6 sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, @@ -142,7 +142,7 @@ static inline void getrf(backend_selector selector, std::int6 sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, @@ -150,20 +150,20 @@ static inline void getri(backend_selector selector, std::int6 sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, @@ -171,446 +171,446 @@ static inline void getri(backend_selector selector, std::int6 sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void getrs(backend_selector selector, oneapi::mkl::transpose trans, +static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, + oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(backend_selector selector, oneapi::mkl::transpose trans, +static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, + oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(backend_selector selector, oneapi::mkl::transpose trans, +static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, + oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(backend_selector selector, oneapi::mkl::transpose trans, +static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, + oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void gesvd(backend_selector selector, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, + oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(backend_selector selector, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, + oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(backend_selector selector, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, + oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(backend_selector selector, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, + oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void heevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +static inline void heevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, + oneapi::math::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void heevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +static inline void heevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, + oneapi::math::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, + oneapi::math::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, + oneapi::math::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void hetrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void hetrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void hetrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void hetrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void orgbr(backend_selector selector, oneapi::mkl::generate vec, +static inline void orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgbr(backend_selector selector, oneapi::mkl::generate vec, +static inline void orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ormtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void ormtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, + oneapi::math::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void ormtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, + oneapi::math::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void syevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +static inline void syevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, + oneapi::math::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void syevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +static inline void syevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, + oneapi::math::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, + oneapi::math::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, + oneapi::math::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void sytrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void sytrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void ungbr(backend_selector selector, oneapi::mkl::generate vec, +static inline void ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungbr(backend_selector selector, oneapi::mkl::generate vec, +static inline void ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, @@ -618,7 +618,7 @@ static inline void ungqr(backend_selector selector, std::int6 std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, @@ -626,83 +626,83 @@ static inline void ungqr(backend_selector selector, std::int6 std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, + oneapi::math::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void unmrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void unmtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, + oneapi::math::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void unmtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, + oneapi::math::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, @@ -710,7 +710,7 @@ static inline void geqrf_batch(backend_selector selector, std std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, + oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, @@ -718,7 +718,7 @@ static inline void geqrf_batch(backend_selector selector, std std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, + oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, @@ -728,7 +728,7 @@ static inline void geqrf_batch(backend_selector selector, std std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, + oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, @@ -738,7 +738,7 @@ static inline void geqrf_batch(backend_selector selector, std std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, + oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, @@ -746,7 +746,7 @@ static inline void getri_batch(backend_selector selector, std sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, + oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -755,7 +755,7 @@ static inline void getri_batch(backend_selector selector, std sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, + oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -765,7 +765,7 @@ static inline void getri_batch(backend_selector selector, std std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, + oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -775,53 +775,53 @@ static inline void getri_batch(backend_selector selector, std std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, + oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } @@ -830,7 +830,7 @@ static inline void getrf_batch(backend_selector selector, std std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, + oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -839,7 +839,7 @@ static inline void getrf_batch(backend_selector selector, std std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, + oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -850,7 +850,7 @@ static inline void getrf_batch(backend_selector selector, std std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, + oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -861,7 +861,7 @@ static inline void getrf_batch(backend_selector selector, std std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, + oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -870,7 +870,7 @@ static inline void orgqr_batch(backend_selector selector, std std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, + oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void orgqr_batch(backend_selector selector, std::int64_t m, @@ -878,76 +878,76 @@ static inline void orgqr_batch(backend_selector selector, std std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, + oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, + oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, + oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, + oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, + oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } @@ -958,7 +958,7 @@ static inline void ungqr_batch(backend_selector selector, std std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, + oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void ungqr_batch(backend_selector selector, std::int64_t m, @@ -968,7 +968,7 @@ static inline void ungqr_batch(backend_selector selector, std std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, + oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } @@ -979,7 +979,7 @@ static inline sycl::event gebrd(backend_selector selector, st float *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + return oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, @@ -987,7 +987,7 @@ static inline sycl::event gebrd(backend_selector selector, st double *tauq, double *taup, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + return oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, @@ -995,7 +995,7 @@ static inline sycl::event gebrd(backend_selector selector, st float *tauq, float *taup, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + return oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, @@ -1004,21 +1004,21 @@ static inline sycl::event gebrd(backend_selector selector, st std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + return oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, @@ -1026,7 +1026,7 @@ static inline sycl::event gerqf(backend_selector selector, st std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, @@ -1034,7 +1034,7 @@ static inline sycl::event gerqf(backend_selector selector, st std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, @@ -1042,21 +1042,21 @@ static inline sycl::event geqrf(backend_selector selector, st std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, @@ -1064,7 +1064,7 @@ static inline sycl::event geqrf(backend_selector selector, st std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, @@ -1072,21 +1072,21 @@ static inline sycl::event getrf(backend_selector selector, st std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, @@ -1094,516 +1094,516 @@ static inline sycl::event getrf(backend_selector selector, st std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + return oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + return oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + return oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + return oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + return oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + return oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + return oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + return oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event heevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +static inline sycl::event heevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + return oneapi::math::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event heevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +static inline sycl::event heevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + return oneapi::math::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hegvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + return oneapi::math::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hegvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + return oneapi::math::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + return oneapi::math::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + return oneapi::math::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, + return oneapi::math::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, + return oneapi::math::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event syevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, +static inline sycl::event syevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + return oneapi::math::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event syevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, +static inline sycl::event syevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + return oneapi::math::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sygvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + return oneapi::math::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sygvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + return oneapi::math::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + return oneapi::math::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + return oneapi::math::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, @@ -1611,7 +1611,7 @@ static inline sycl::event ungqr(backend_selector selector, st std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, @@ -1619,91 +1619,91 @@ static inline sycl::event ungqr(backend_selector selector, st std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, + return oneapi::math::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, + return oneapi::math::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } @@ -1713,7 +1713,7 @@ static inline sycl::event geqrf_batch(backend_selector select std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1723,7 +1723,7 @@ static inline sycl::event geqrf_batch(backend_selector select std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1733,7 +1733,7 @@ static inline sycl::event geqrf_batch(backend_selector select std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1744,7 +1744,7 @@ static inline sycl::event geqrf_batch(backend_selector select std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1753,7 +1753,7 @@ static inline sycl::event geqrf_batch(backend_selector select std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1762,7 +1762,7 @@ static inline sycl::event geqrf_batch(backend_selector select std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1772,7 +1772,7 @@ static inline sycl::event geqrf_batch(backend_selector select std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1782,7 +1782,7 @@ static inline sycl::event geqrf_batch(backend_selector select std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1792,7 +1792,7 @@ static inline sycl::event getrf_batch(backend_selector select std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1802,7 +1802,7 @@ static inline sycl::event getrf_batch(backend_selector select std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1812,7 +1812,7 @@ static inline sycl::event getrf_batch(backend_selector select std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1823,7 +1823,7 @@ static inline sycl::event getrf_batch(backend_selector select std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1833,7 +1833,7 @@ static inline sycl::event getrf_batch(backend_selector select std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1843,7 +1843,7 @@ static inline sycl::event getrf_batch(backend_selector select std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1853,7 +1853,7 @@ static inline sycl::event getrf_batch(backend_selector select std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1863,7 +1863,7 @@ static inline sycl::event getrf_batch(backend_selector select std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1873,7 +1873,7 @@ static inline sycl::event getri_batch(backend_selector select std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1883,7 +1883,7 @@ static inline sycl::event getri_batch(backend_selector select std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1893,7 +1893,7 @@ static inline sycl::event getri_batch(backend_selector select std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1904,7 +1904,7 @@ static inline sycl::event getri_batch(backend_selector select std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, + return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1913,7 +1913,7 @@ static inline sycl::event getri_batch(backend_selector select std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1922,7 +1922,7 @@ static inline sycl::event getri_batch(backend_selector select std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1932,7 +1932,7 @@ static inline sycl::event getri_batch(backend_selector select std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1942,95 +1942,95 @@ static inline sycl::event getri_batch(backend_selector select std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs_batch( + return oneapi::math::lapack::cusolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs_batch( + return oneapi::math::lapack::cusolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( - backend_selector selector, oneapi::mkl::transpose trans, std::int64_t n, + backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs_batch( + return oneapi::math::lapack::cusolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( - backend_selector selector, oneapi::mkl::transpose trans, std::int64_t n, + backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs_batch( + return oneapi::math::lapack::cusolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( - backend_selector selector, oneapi::mkl::transpose *trans, std::int64_t *n, + backend_selector selector, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2040,7 +2040,7 @@ static inline sycl::event orgqr_batch(backend_selector select std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, + return oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2050,7 +2050,7 @@ static inline sycl::event orgqr_batch(backend_selector select std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, + return oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2060,7 +2060,7 @@ static inline sycl::event orgqr_batch(backend_selector select std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2070,178 +2070,178 @@ static inline sycl::event orgqr_batch(backend_selector select std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, float **a, + oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, double **a, + oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2252,7 +2252,7 @@ static inline sycl::event ungqr_batch(backend_selector select std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, + return oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2263,7 +2263,7 @@ static inline sycl::event ungqr_batch(backend_selector select std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, + return oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2273,7 +2273,7 @@ static inline sycl::event ungqr_batch(backend_selector select std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2284,7 +2284,7 @@ static inline sycl::event ungqr_batch(backend_selector select std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2293,211 +2293,211 @@ static inline sycl::event ungqr_batch(backend_selector select template std::int64_t gebrd_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::gebrd_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::cusolver::gebrd_scratchpad_size(selector.get_queue(), m, n, lda); } template std::int64_t gerqf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::gerqf_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::cusolver::gerqf_scratchpad_size(selector.get_queue(), m, n, lda); } template std::int64_t geqrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::geqrf_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::cusolver::geqrf_scratchpad_size(selector.get_queue(), m, n, lda); } template std::int64_t gesvd_scratchpad_size(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return oneapi::mkl::lapack::cusolver::gesvd_scratchpad_size( + return oneapi::math::lapack::cusolver::gesvd_scratchpad_size( selector.get_queue(), jobu, jobvt, m, n, lda, ldu, ldvt); } template std::int64_t getrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::getrf_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::cusolver::getrf_scratchpad_size(selector.get_queue(), m, n, lda); } template std::int64_t getri_scratchpad_size(backend_selector selector, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::getri_scratchpad_size(selector.get_queue(), n, + return oneapi::math::lapack::cusolver::getri_scratchpad_size(selector.get_queue(), n, lda); } template std::int64_t getrs_scratchpad_size(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::cusolver::getrs_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::cusolver::getrs_scratchpad_size(selector.get_queue(), trans, n, nrhs, lda, ldb); } template std::int64_t heevd_scratchpad_size(backend_selector selector, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::heevd_scratchpad_size(selector.get_queue(), jobz, + return oneapi::math::lapack::cusolver::heevd_scratchpad_size(selector.get_queue(), jobz, uplo, n, lda); } template std::int64_t hegvd_scratchpad_size(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::cusolver::hegvd_scratchpad_size( + return oneapi::math::lapack::cusolver::hegvd_scratchpad_size( selector.get_queue(), itype, jobz, uplo, n, lda, ldb); } template std::int64_t hetrd_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::hetrd_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::cusolver::hetrd_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t hetrf_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::hetrf_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::cusolver::hetrf_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t orgbr_scratchpad_size(backend_selector selector, - oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::orgbr_scratchpad_size(selector.get_queue(), vect, + return oneapi::math::lapack::cusolver::orgbr_scratchpad_size(selector.get_queue(), vect, m, n, k, lda); } template std::int64_t orgtr_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::orgtr_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::cusolver::orgtr_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t orgqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::orgqr_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::cusolver::orgqr_scratchpad_size(selector.get_queue(), m, n, k, lda); } template std::int64_t ormrq_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::cusolver::ormrq_scratchpad_size(selector.get_queue(), side, + return oneapi::math::lapack::cusolver::ormrq_scratchpad_size(selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t ormqr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::cusolver::ormqr_scratchpad_size(selector.get_queue(), side, + return oneapi::math::lapack::cusolver::ormqr_scratchpad_size(selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t ormtr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::cusolver::ormtr_scratchpad_size( + return oneapi::math::lapack::cusolver::ormtr_scratchpad_size( selector.get_queue(), side, uplo, trans, m, n, lda, ldc); } template std::int64_t potrf_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::potrf_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::cusolver::potrf_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t potrs_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::cusolver::potrs_scratchpad_size(selector.get_queue(), uplo, + return oneapi::math::lapack::cusolver::potrs_scratchpad_size(selector.get_queue(), uplo, n, nrhs, lda, ldb); } template std::int64_t potri_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::potri_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::cusolver::potri_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t sytrf_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::sytrf_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::cusolver::sytrf_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t syevd_scratchpad_size(backend_selector selector, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::syevd_scratchpad_size(selector.get_queue(), jobz, + return oneapi::math::lapack::cusolver::syevd_scratchpad_size(selector.get_queue(), jobz, uplo, n, lda); } template std::int64_t sygvd_scratchpad_size(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::cusolver::sygvd_scratchpad_size( + return oneapi::math::lapack::cusolver::sygvd_scratchpad_size( selector.get_queue(), itype, jobz, uplo, n, lda, ldb); } template std::int64_t sytrd_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::sytrd_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::cusolver::sytrd_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t trtrs_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::cusolver::trtrs_scratchpad_size( + return oneapi::math::lapack::cusolver::trtrs_scratchpad_size( selector.get_queue(), uplo, trans, diag, n, nrhs, lda, ldb); } template std::int64_t ungbr_scratchpad_size(backend_selector selector, - oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::ungbr_scratchpad_size(selector.get_queue(), vect, + return oneapi::math::lapack::cusolver::ungbr_scratchpad_size(selector.get_queue(), vect, m, n, k, lda); } template std::int64_t ungqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::ungqr_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::cusolver::ungqr_scratchpad_size(selector.get_queue(), m, n, k, lda); } template std::int64_t ungtr_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::cusolver::ungtr_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::cusolver::ungtr_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t unmrq_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::cusolver::unmrq_scratchpad_size(selector.get_queue(), side, + return oneapi::math::lapack::cusolver::unmrq_scratchpad_size(selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t unmqr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::cusolver::unmqr_scratchpad_size(selector.get_queue(), side, + return oneapi::math::lapack::cusolver::unmqr_scratchpad_size(selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t unmtr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::cusolver::unmtr_scratchpad_size( + return oneapi::math::lapack::cusolver::unmtr_scratchpad_size( selector.get_queue(), side, uplo, trans, m, n, lda, ldc); } template @@ -2505,23 +2505,23 @@ std::int64_t getrf_batch_scratchpad_size(backend_selector sel std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return oneapi::mkl::lapack::cusolver::getrf_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::getrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, stride_a, stride_ipiv, batch_size); } template std::int64_t getri_batch_scratchpad_size(backend_selector selector, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return oneapi::mkl::lapack::cusolver::getri_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::getri_batch_scratchpad_size( selector.get_queue(), n, lda, stride_a, stride_ipiv, batch_size); } template std::int64_t getrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return oneapi::mkl::lapack::cusolver::getrs_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::getrs_batch_scratchpad_size( selector.get_queue(), trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } @@ -2530,22 +2530,22 @@ std::int64_t geqrf_batch_scratchpad_size(backend_selector sel std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return oneapi::mkl::lapack::cusolver::geqrf_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::geqrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, stride_a, stride_tau, batch_size); } template std::int64_t potrf_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - return oneapi::mkl::lapack::cusolver::potrf_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::potrf_batch_scratchpad_size( selector.get_queue(), uplo, n, lda, stride_a, batch_size); } template std::int64_t potrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return oneapi::mkl::lapack::cusolver::potrs_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::potrs_batch_scratchpad_size( selector.get_queue(), uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template @@ -2553,7 +2553,7 @@ std::int64_t orgqr_batch_scratchpad_size(backend_selector sel std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return oneapi::mkl::lapack::cusolver::orgqr_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::orgqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, stride_a, stride_tau, batch_size); } template @@ -2561,36 +2561,36 @@ std::int64_t ungqr_batch_scratchpad_size(backend_selector sel std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return oneapi::mkl::lapack::cusolver::ungqr_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::ungqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, stride_a, stride_tau, batch_size); } template std::int64_t getrf_batch_scratchpad_size(backend_selector selector, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::cusolver::getrf_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::getrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } template std::int64_t getri_batch_scratchpad_size(backend_selector selector, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::cusolver::getri_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::getri_batch_scratchpad_size( selector.get_queue(), n, lda, group_count, group_sizes); } template std::int64_t getrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::cusolver::getrs_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::getrs_batch_scratchpad_size( selector.get_queue(), trans, n, nrhs, lda, ldb, group_count, group_sizes); } template std::int64_t geqrf_batch_scratchpad_size(backend_selector selector, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::cusolver::geqrf_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::geqrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } template @@ -2598,23 +2598,23 @@ std::int64_t orgqr_batch_scratchpad_size(backend_selector sel std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::cusolver::orgqr_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::orgqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } template std::int64_t potrf_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::cusolver::potrf_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::potrf_batch_scratchpad_size( selector.get_queue(), uplo, n, lda, group_count, group_sizes); } template std::int64_t potrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::cusolver::potrs_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::potrs_batch_scratchpad_size( selector.get_queue(), uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template @@ -2622,6 +2622,6 @@ std::int64_t ungqr_batch_scratchpad_size(backend_selector sel std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::cusolver::ungqr_batch_scratchpad_size( + return oneapi::math::lapack::cusolver::ungqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } diff --git a/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp index 7c27483e8..40192b082 100644 --- a/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp +++ b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp @@ -31,7 +31,7 @@ #include "oneapi/math/detail/export.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace cusolver { @@ -39,7 +39,7 @@ namespace cusolver { } // namespace cusolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_LAPACK_CUSOLVER_HPP_ diff --git a/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx index 26c3323b1..ae8e8a147 100644 --- a/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx +++ b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx @@ -125,42 +125,42 @@ ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, @@ -168,7 +168,7 @@ ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi:: sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, @@ -176,62 +176,62 @@ ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi:: sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); @@ -244,194 +244,194 @@ ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, st sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, @@ -449,19 +449,19 @@ ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, st sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -469,7 +469,7 @@ ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mk sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -477,7 +477,7 @@ ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mk sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -485,7 +485,7 @@ ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mk sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -493,16 +493,16 @@ ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mk sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, @@ -561,21 +561,21 @@ ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -584,7 +584,7 @@ ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -631,41 +631,41 @@ ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -673,7 +673,7 @@ ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std: sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -800,112 +800,112 @@ ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::comple std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); @@ -920,201 +920,201 @@ ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_ std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1131,56 +1131,56 @@ ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_ std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, @@ -1343,7 +1343,7 @@ ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std: std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, @@ -1351,7 +1351,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, @@ -1359,7 +1359,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1368,7 +1368,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1377,7 +1377,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, @@ -1385,7 +1385,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, @@ -1393,7 +1393,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1401,7 +1401,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1436,71 +1436,71 @@ ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std: double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -1508,7 +1508,7 @@ ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo upl std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -1516,28 +1516,28 @@ ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo upl std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, @@ -1589,8 +1589,8 @@ ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64 std::int64_t lda); template -ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); @@ -1603,36 +1603,36 @@ ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64 std::int64_t lda); template -ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template @@ -1640,65 +1640,65 @@ ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64 std::int64_t k, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, +ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); @@ -1707,25 +1707,25 @@ ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64 std::int64_t k, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); @@ -1744,7 +1744,7 @@ ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std: template ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); @@ -1756,13 +1756,13 @@ ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std: std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, @@ -1795,7 +1795,7 @@ ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std: template ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template @@ -1811,13 +1811,13 @@ ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std: std::int64_t *group_sizes); template -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, diff --git a/include/oneapi/math/lapack/detail/lapack_loader.hpp b/include/oneapi/math/lapack/detail/lapack_loader.hpp index 88d92688c..7e0da894b 100644 --- a/include/oneapi/math/lapack/detail/lapack_loader.hpp +++ b/include/oneapi/math/lapack/detail/lapack_loader.hpp @@ -34,1222 +34,1222 @@ #include "oneapi/math/detail/get_device_id.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace detail { -ONEMATH_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +ONEMATH_EXPORT void orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +ONEMATH_EXPORT void orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +ONEMATH_EXPORT void ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +ONEMATH_EXPORT void ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT void unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT void ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tauq, double *taup, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tauq, float *taup, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event heevd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event heevd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event hetrd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event hetrd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event hetrf(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event hetrf(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event orgbr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event orgbr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event orgtr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event orgtr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormtr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormtr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event ormrq(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event ormrq(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event ormqr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event ormqr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, +ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event syevd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event syevd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event sytrd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event sytrd(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, +ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ungbr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ungbr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT sycl::event ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event ungtr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event ungtr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event unmrq(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event unmrq(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event unmqr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event unmqr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmtr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmtr(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, @@ -1257,33 +1257,33 @@ ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue & double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event getrs_batch( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event getrs_batch( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose *trans, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose *trans, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose *trans, std::int64_t *n, +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1291,143 +1291,143 @@ ONEMATH_EXPORT sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue & std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event getrs_batch( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, float **a, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, double **a, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, @@ -1435,18 +1435,18 @@ ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue & std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); ONEMATH_EXPORT sycl::event ungqr_batch( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, +ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, @@ -1454,929 +1454,929 @@ ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue & std::int64_t scratchpad_size, const std::vector &dependencies = {}); -template = nullptr> -std::int64_t gebrd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +template = nullptr> +std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t gerqf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +template = nullptr> +std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t geqrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +template = nullptr> +std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t gesvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template = nullptr> -std::int64_t gesvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); -template = nullptr> -std::int64_t getrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +template = nullptr> +std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t getri_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +template = nullptr> +std::int64_t getri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t getrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +template = nullptr> +std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t heevd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t heevd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hegvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t hegvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t hetrd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t hetrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hetrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t hetrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t orgbr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t orgtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t orgqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ormrq_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); -template = nullptr> -std::int64_t potrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t potrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +template = nullptr> +std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); +template = nullptr> +std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); -template = nullptr> -std::int64_t potri_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t sytrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); +template = nullptr> +std::int64_t potri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); +template = nullptr> +std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t syevd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t sygvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t sytrd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t trtrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); +template = nullptr> +std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t ungbr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t ungbr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ungqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t ungqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ungtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t ungtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t unmrq_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t unmrq_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t unmqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t unmtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); -template = nullptr> -std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +template = nullptr> +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); -template = nullptr> -std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +template = nullptr> +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); -template = nullptr> -std::int64_t getrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, +template = nullptr> +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +template = nullptr> +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); -template = nullptr> -std::int64_t potrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, +template = nullptr> +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); -template = nullptr> -std::int64_t potrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +template = nullptr> +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t ungqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); -template = nullptr> -std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +template = nullptr> +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> -std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +template = nullptr> +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> -std::int64_t getrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose *trans, std::int64_t *n, +template = nullptr> +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +template = nullptr> +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> -std::int64_t potrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, +template = nullptr> +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> -std::int64_t potrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, +template = nullptr> +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t ungqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> -ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> -ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getri_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getri_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); + oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); + oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t potri_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potri_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> -ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); } //namespace detail } //namespace lapack -} //namespace mkl +} //namespace math } //namespace oneapi diff --git a/include/oneapi/math/lapack/detail/lapack_rt.hpp b/include/oneapi/math/lapack/detail/lapack_rt.hpp index ef03694fd..8f7aa92f6 100644 --- a/include/oneapi/math/lapack/detail/lapack_rt.hpp +++ b/include/oneapi/math/lapack/detail/lapack_rt.hpp @@ -35,7 +35,7 @@ #include "oneapi/math/lapack/detail/lapack_loader.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { static inline void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, @@ -168,7 +168,7 @@ static inline void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, @@ -176,7 +176,7 @@ static inline void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std:: detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +static inline void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, @@ -184,14 +184,14 @@ static inline void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std:: detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +static inline void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +static inline void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, @@ -199,7 +199,7 @@ static inline void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std:: detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +static inline void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, @@ -207,7 +207,7 @@ static inline void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::m detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +static inline void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, @@ -215,7 +215,7 @@ static inline void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::m detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +static inline void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, @@ -225,7 +225,7 @@ static inline void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::m detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +static inline void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, @@ -235,22 +235,22 @@ static inline void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::m detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +static inline void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::heevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +static inline void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::heevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +static inline void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, @@ -258,8 +258,8 @@ static inline void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::jo detail::hegvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +static inline void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, @@ -267,7 +267,7 @@ static inline void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::jo detail::hegvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, @@ -276,7 +276,7 @@ static inline void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_ detail::hetrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, @@ -285,28 +285,28 @@ static inline void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_ detail::hetrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::hetrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::hetrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +static inline void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::orgbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +static inline void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -323,33 +323,33 @@ static inline void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::orgqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::orgtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::orgtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::ormtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::ormtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +static inline void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, @@ -357,7 +357,7 @@ static inline void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl detail::ormrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +static inline void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, @@ -365,7 +365,7 @@ static inline void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl detail::ormrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +static inline void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, @@ -373,7 +373,7 @@ static inline void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl detail::ormqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +static inline void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, @@ -381,65 +381,65 @@ static inline void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl detail::ormqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, @@ -447,7 +447,7 @@ static inline void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_ detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, @@ -455,77 +455,77 @@ static inline void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_ detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +static inline void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::syevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +static inline void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::syevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +static inline void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::sygvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +static inline void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::sygvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::sytrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::sytrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +static inline void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, @@ -533,24 +533,24 @@ static inline void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +static inline void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +static inline void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +static inline void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, @@ -558,7 +558,7 @@ static inline void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +static inline void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, @@ -566,7 +566,7 @@ static inline void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int detail::ungbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +static inline void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, @@ -588,21 +588,21 @@ static inline void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std std::int64_t scratchpad_size) { detail::ungqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::ungtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { detail::ungtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +static inline void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -612,7 +612,7 @@ static inline void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl detail::unmrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +static inline void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -622,7 +622,7 @@ static inline void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl detail::unmrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +static inline void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -632,7 +632,7 @@ static inline void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl detail::unmqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +static inline void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -642,8 +642,8 @@ static inline void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl detail::unmqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, @@ -652,8 +652,8 @@ static inline void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl detail::unmtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, @@ -730,7 +730,7 @@ static inline void getri_batch(sycl::queue &queue, std::int64_t n, detail::getri_batch(get_device_id(queue), queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -static inline void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +static inline void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, @@ -739,7 +739,7 @@ static inline void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +static inline void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, @@ -748,7 +748,7 @@ static inline void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +static inline void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -759,7 +759,7 @@ static inline void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +static inline void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -820,21 +820,21 @@ static inline void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t detail::orgqr_batch(get_device_id(queue), queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, @@ -842,7 +842,7 @@ static inline void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std:: detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, @@ -850,7 +850,7 @@ static inline void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std:: detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -858,7 +858,7 @@ static inline void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std:: detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -866,7 +866,7 @@ static inline void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std:: detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -876,7 +876,7 @@ static inline void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std:: detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -1047,7 +1047,7 @@ static inline sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex return detail::getri(get_device_id(queue), queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +static inline sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1055,7 +1055,7 @@ static inline sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans return detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +static inline sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, @@ -1063,7 +1063,7 @@ static inline sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans return detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +static inline sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, @@ -1071,7 +1071,7 @@ static inline sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans return detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +static inline sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1079,8 +1079,8 @@ static inline sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans return detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +static inline sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, @@ -1088,16 +1088,16 @@ static inline sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, return detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, +static inline sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +static inline sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, @@ -1106,8 +1106,8 @@ static inline sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, return detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +static inline sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, @@ -1116,14 +1116,14 @@ static inline sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, return detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +static inline sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::heevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +static inline sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1131,8 +1131,8 @@ static inline sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneap return detail::heevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +static inline sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1140,8 +1140,8 @@ static inline sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi:: return detail::hegvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +static inline sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1149,7 +1149,7 @@ static inline sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi:: return detail::hegvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1157,7 +1157,7 @@ static inline sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std: return detail::hetrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1165,28 +1165,28 @@ static inline sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std: return detail::hetrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::hetrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::hetrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +static inline sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::orgbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +static inline sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { @@ -1207,139 +1207,139 @@ static inline sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t return detail::orgqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::orgtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::orgtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::ormtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::ormtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::ormrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::ormrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::ormqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::ormqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1347,7 +1347,7 @@ static inline sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std: return detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1355,80 +1355,80 @@ static inline sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std: return detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +static inline sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::syevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +static inline sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::syevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, +static inline sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::sygvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, +static inline sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::sygvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::sytrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::sytrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +static inline sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1436,8 +1436,8 @@ static inline sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, return detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +static inline sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, @@ -1445,8 +1445,8 @@ static inline sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, return detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +static inline sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, @@ -1454,8 +1454,8 @@ static inline sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, return detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +static inline sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1463,7 +1463,7 @@ static inline sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, return detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +static inline sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1471,7 +1471,7 @@ static inline sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, s return detail::ungbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +static inline sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1494,14 +1494,14 @@ static inline sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t return detail::ungqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { return detail::ungtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1509,8 +1509,8 @@ static inline sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std: return detail::ungtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1518,8 +1518,8 @@ static inline sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, return detail::unmrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, @@ -1528,8 +1528,8 @@ static inline sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, return detail::unmrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1537,8 +1537,8 @@ static inline sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, return detail::unmqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, @@ -1547,8 +1547,8 @@ static inline sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, return detail::unmqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1556,8 +1556,8 @@ static inline sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, onea return detail::unmtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, @@ -1770,7 +1770,7 @@ static inline sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std:: return detail::getri_batch(get_device_id(queue), queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, @@ -1781,7 +1781,7 @@ static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, @@ -1792,7 +1792,7 @@ static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1804,7 +1804,7 @@ static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1816,7 +1816,7 @@ static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, @@ -1826,7 +1826,7 @@ static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose return detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, @@ -1836,7 +1836,7 @@ static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose return detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1846,7 +1846,7 @@ static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose return detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1891,7 +1891,7 @@ static inline sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std:: return detail::orgqr_batch(get_device_id(queue), queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, @@ -1899,7 +1899,7 @@ static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, @@ -1907,7 +1907,7 @@ static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1915,7 +1915,7 @@ static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, @@ -1924,7 +1924,7 @@ static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, @@ -1932,7 +1932,7 @@ static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *upl return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, @@ -1940,7 +1940,7 @@ static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *upl return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1948,7 +1948,7 @@ static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *upl return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, @@ -1957,7 +1957,7 @@ static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *upl return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -1966,7 +1966,7 @@ static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -1975,7 +1975,7 @@ static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -1985,7 +1985,7 @@ static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -1995,7 +1995,7 @@ static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, @@ -2004,7 +2004,7 @@ static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *upl return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, @@ -2013,7 +2013,7 @@ static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *upl return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, @@ -2022,7 +2022,7 @@ static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *upl return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, @@ -2075,7 +2075,7 @@ std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int6 std::int64_t lda) { return detail::gebrd_scratchpad_size(get_device_id(queue), queue, m, n, lda); } -template = nullptr> +template = nullptr> std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return detail::gerqf_scratchpad_size(get_device_id(queue), queue, m, n, lda); @@ -2086,15 +2086,15 @@ std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int6 return detail::geqrf_scratchpad_size(get_device_id(queue), queue, m, n, lda); } template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { return detail::gesvd_scratchpad_size(get_device_id(queue), queue, jobu, jobvt, m, n, lda, ldu, ldvt); } template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { return detail::gesvd_scratchpad_size(get_device_id(queue), queue, jobu, jobvt, m, n, lda, ldu, ldvt); @@ -2109,43 +2109,43 @@ std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int6 return detail::getri_scratchpad_size(get_device_id(queue), queue, n, lda); } template = nullptr> -std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return detail::getrs_scratchpad_size(get_device_id(queue), queue, trans, n, nrhs, lda, ldb); } template = nullptr> -std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { +std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::heevd_scratchpad_size(get_device_id(queue), queue, jobz, uplo, n, lda); } template = nullptr> -std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, +std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { return detail::hegvd_scratchpad_size(get_device_id(queue), queue, itype, jobz, uplo, n, lda, ldb); } template = nullptr> -std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::hetrd_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::hetrf_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return detail::orgbr_scratchpad_size(get_device_id(queue), queue, vect, m, n, k, lda); } template = nullptr> -std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::orgtr_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } @@ -2155,78 +2155,78 @@ std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int6 return detail::orgqr_scratchpad_size(get_device_id(queue), queue, m, n, k, lda); } template = nullptr> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return detail::ormrq_scratchpad_size(get_device_id(queue), queue, side, trans, m, n, k, lda, ldc); } template = nullptr> -std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return detail::ormqr_scratchpad_size(get_device_id(queue), queue, side, trans, m, n, k, lda, ldc); } template = nullptr> -std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { return detail::ormtr_scratchpad_size(get_device_id(queue), queue, side, uplo, trans, m, n, lda, ldc); } template = nullptr> -std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::potrf_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } -template = nullptr> -std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return detail::potrs_scratchpad_size(get_device_id(queue), queue, uplo, n, nrhs, lda, ldb); } -template = nullptr> -std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::potri_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } -template = nullptr> -std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::sytrf_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { +std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::syevd_scratchpad_size(get_device_id(queue), queue, jobz, uplo, n, lda); } template = nullptr> -std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, +std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { return detail::sygvd_scratchpad_size(get_device_id(queue), queue, itype, jobz, uplo, n, lda, ldb); } template = nullptr> -std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::sytrd_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return detail::trtrs_scratchpad_size(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, lda, ldb); } template = nullptr> -std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return detail::ungbr_scratchpad_size(get_device_id(queue), queue, vect, m, n, k, lda); } @@ -2236,51 +2236,51 @@ std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int6 return detail::ungqr_scratchpad_size(get_device_id(queue), queue, m, n, k, lda); } template = nullptr> -std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::ungtr_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return detail::unmrq_scratchpad_size(get_device_id(queue), queue, side, trans, m, n, k, lda, ldc); } template = nullptr> -std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return detail::unmqr_scratchpad_size(get_device_id(queue), queue, side, trans, m, n, k, lda, ldc); } template = nullptr> -std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { return detail::unmtr_scratchpad_size(get_device_id(queue), queue, side, uplo, trans, m, n, lda, ldc); } -template = nullptr> +template = nullptr> std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return detail::getrf_batch_scratchpad_size(get_device_id(queue), queue, m, n, lda, stride_a, stride_ipiv, batch_size); } -template = nullptr> +template = nullptr> std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return detail::getri_batch_scratchpad_size(get_device_id(queue), queue, n, lda, stride_a, stride_ipiv, batch_size); } -template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +template = nullptr> +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, @@ -2289,22 +2289,22 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transp lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } -template = nullptr> +template = nullptr> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return detail::geqrf_batch_scratchpad_size(get_device_id(queue), queue, m, n, lda, stride_a, stride_tau, batch_size); } -template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { return detail::potrf_batch_scratchpad_size(get_device_id(queue), queue, uplo, n, lda, stride_a, batch_size); } -template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { @@ -2312,7 +2312,7 @@ std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo u lda, stride_a, ldb, stride_b, batch_size); } template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { @@ -2320,35 +2320,35 @@ std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std stride_a, stride_tau, batch_size); } template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return detail::ungqr_batch_scratchpad_size(get_device_id(queue), queue, m, n, k, lda, stride_a, stride_tau, batch_size); } -template = nullptr> +template = nullptr> std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return detail::getrf_batch_scratchpad_size(get_device_id(queue), queue, m, n, lda, group_count, group_sizes); } -template = nullptr> +template = nullptr> std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return detail::getri_batch_scratchpad_size(get_device_id(queue), queue, n, lda, group_count, group_sizes); } -template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose *trans, +template = nullptr> +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { return detail::getrs_batch_scratchpad_size(get_device_id(queue), queue, trans, n, nrhs, lda, ldb, group_count, group_sizes); } -template = nullptr> +template = nullptr> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2356,22 +2356,22 @@ std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, st group_count, group_sizes); } template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return detail::orgqr_batch_scratchpad_size(get_device_id(queue), queue, m, n, k, lda, group_count, group_sizes); } -template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +template = nullptr> +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return detail::potrf_batch_scratchpad_size(get_device_id(queue), queue, uplo, n, lda, group_count, group_sizes); } -template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +template = nullptr> +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2379,7 +2379,7 @@ std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo * lda, ldb, group_count, group_sizes); } template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2388,5 +2388,5 @@ std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, st } } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/include/oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx b/include/oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx index 1ebe97527..41e7e8753 100644 --- a/include/oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx +++ b/include/oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx @@ -24,7 +24,7 @@ static inline void gebrd(backend_selector selector, std sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, @@ -32,7 +32,7 @@ static inline void gebrd(backend_selector selector, std sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, @@ -40,7 +40,7 @@ static inline void gebrd(backend_selector selector, std sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, @@ -50,21 +50,21 @@ static inline void gebrd(backend_selector selector, std sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, @@ -72,7 +72,7 @@ static inline void gerqf(backend_selector selector, std sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, @@ -80,7 +80,7 @@ static inline void gerqf(backend_selector selector, std sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, @@ -88,21 +88,21 @@ static inline void geqrf(backend_selector selector, std sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, @@ -110,7 +110,7 @@ static inline void geqrf(backend_selector selector, std sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, @@ -118,21 +118,21 @@ static inline void getrf(backend_selector selector, std sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, @@ -140,7 +140,7 @@ static inline void getrf(backend_selector selector, std sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, @@ -148,20 +148,20 @@ static inline void getri(backend_selector selector, std sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, @@ -169,447 +169,447 @@ static inline void getri(backend_selector selector, std sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } static inline void getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } static inline void getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } static inline void getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void heevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +static inline void heevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + oneapi::math::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void heevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +static inline void heevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + oneapi::math::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void hetrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void hetrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void hetrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, + oneapi::math::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void hetrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, + oneapi::math::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void orgbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void orgbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ormtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void ormtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, + oneapi::math::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void ormtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, + oneapi::math::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void syevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +static inline void syevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + oneapi::math::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void syevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +static inline void syevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + oneapi::math::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void sytrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void sytrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } static inline void ungbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void ungbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, @@ -617,7 +617,7 @@ static inline void ungqr(backend_selector selector, std std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, @@ -625,83 +625,83 @@ static inline void ungqr(backend_selector selector, std std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, + oneapi::math::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void unmrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void unmtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, + oneapi::math::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void unmtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, + oneapi::math::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, @@ -709,7 +709,7 @@ static inline void geqrf_batch(backend_selector selecto std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } @@ -718,7 +718,7 @@ static inline void geqrf_batch(backend_selector selecto std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } @@ -729,7 +729,7 @@ static inline void geqrf_batch(backend_selector selecto std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } @@ -740,7 +740,7 @@ static inline void geqrf_batch(backend_selector selecto std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } @@ -749,7 +749,7 @@ static inline void getri_batch(backend_selector selecto sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -758,7 +758,7 @@ static inline void getri_batch(backend_selector selecto sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -768,7 +768,7 @@ static inline void getri_batch(backend_selector selecto std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -778,53 +778,53 @@ static inline void getri_batch(backend_selector selecto std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } @@ -833,7 +833,7 @@ static inline void getrf_batch(backend_selector selecto std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -842,7 +842,7 @@ static inline void getrf_batch(backend_selector selecto std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -853,7 +853,7 @@ static inline void getrf_batch(backend_selector selecto std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -864,7 +864,7 @@ static inline void getrf_batch(backend_selector selecto std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -873,7 +873,7 @@ static inline void orgqr_batch(backend_selector selecto std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch(selector.get_queue(), m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } @@ -882,81 +882,81 @@ static inline void orgqr_batch(backend_selector selecto std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch(selector.get_queue(), m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch( + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch( selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } static inline void potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch( + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch( selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } static inline void potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch( + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch( selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } static inline void potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch( + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch( selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } static inline void potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } @@ -967,7 +967,7 @@ static inline void ungqr_batch(backend_selector selecto std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch(selector.get_queue(), m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } @@ -978,7 +978,7 @@ static inline void ungqr_batch(backend_selector selecto std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch(selector.get_queue(), m, n, k, a, lda, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } @@ -987,7 +987,7 @@ static inline sycl::event gebrd(backend_selector select float *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, + return oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } @@ -996,7 +996,7 @@ static inline sycl::event gebrd(backend_selector select double *tauq, double *taup, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, + return oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } @@ -1005,7 +1005,7 @@ static inline sycl::event gebrd(backend_selector select float *tauq, float *taup, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, + return oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } @@ -1015,7 +1015,7 @@ static inline sycl::event gebrd(backend_selector select std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, + return oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } @@ -1023,14 +1023,14 @@ static inline sycl::event gerqf(backend_selector select std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, @@ -1038,7 +1038,7 @@ static inline sycl::event gerqf(backend_selector select std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, @@ -1046,7 +1046,7 @@ static inline sycl::event gerqf(backend_selector select std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, @@ -1054,21 +1054,21 @@ static inline sycl::event geqrf(backend_selector select std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, @@ -1076,7 +1076,7 @@ static inline sycl::event geqrf(backend_selector select std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, @@ -1084,21 +1084,21 @@ static inline sycl::event getrf(backend_selector select std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, @@ -1106,532 +1106,532 @@ static inline sycl::event getrf(backend_selector select std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, + return oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, + return oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, + return oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, + return oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } static inline sycl::event heevd(backend_selector selector, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event heevd(backend_selector selector, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hegvd(backend_selector selector, - std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, + return oneapi::math::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hegvd(backend_selector selector, - std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, + return oneapi::math::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, + return oneapi::math::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, + return oneapi::math::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgbr( + return oneapi::math::lapack::LAPACK_BACKEND::orgbr( selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgbr( + return oneapi::math::lapack::LAPACK_BACKEND::orgbr( selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, + return oneapi::math::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, + return oneapi::math::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, + return oneapi::math::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, + return oneapi::math::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, + return oneapi::math::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, + return oneapi::math::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event syevd(backend_selector selector, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event syevd(backend_selector selector, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sygvd(backend_selector selector, - std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, + return oneapi::math::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sygvd(backend_selector selector, - std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, + return oneapi::math::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, + return oneapi::math::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, + return oneapi::math::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, + return oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, + return oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, + return oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, + return oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungbr( + return oneapi::math::lapack::LAPACK_BACKEND::ungbr( selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungbr( + return oneapi::math::lapack::LAPACK_BACKEND::ungbr( selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, @@ -1639,7 +1639,7 @@ static inline sycl::event ungqr(backend_selector select std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, @@ -1647,91 +1647,91 @@ static inline sycl::event ungqr(backend_selector select std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, + return oneapi::math::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, + return oneapi::math::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, + return oneapi::math::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, + return oneapi::math::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, + return oneapi::math::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, + return oneapi::math::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } @@ -1741,7 +1741,7 @@ static inline sycl::event geqrf_batch(backend_selector std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1751,7 +1751,7 @@ static inline sycl::event geqrf_batch(backend_selector std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1762,7 +1762,7 @@ static inline sycl::event geqrf_batch(backend_selector std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1773,7 +1773,7 @@ static inline sycl::event geqrf_batch(backend_selector std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1783,7 +1783,7 @@ static inline sycl::event geqrf_batch(backend_selector std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1793,7 +1793,7 @@ static inline sycl::event geqrf_batch(backend_selector std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1803,7 +1803,7 @@ static inline sycl::event geqrf_batch(backend_selector std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1814,7 +1814,7 @@ static inline sycl::event geqrf_batch(backend_selector std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1824,7 +1824,7 @@ static inline sycl::event getrf_batch(backend_selector std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1834,7 +1834,7 @@ static inline sycl::event getrf_batch(backend_selector std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1844,7 +1844,7 @@ static inline sycl::event getrf_batch(backend_selector std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1855,7 +1855,7 @@ static inline sycl::event getrf_batch(backend_selector std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1865,7 +1865,7 @@ static inline sycl::event getrf_batch(backend_selector std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1875,7 +1875,7 @@ static inline sycl::event getrf_batch(backend_selector std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1885,7 +1885,7 @@ static inline sycl::event getrf_batch(backend_selector std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1896,7 +1896,7 @@ static inline sycl::event getrf_batch(backend_selector std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1906,7 +1906,7 @@ static inline sycl::event getri_batch(backend_selector std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getri_batch( selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1916,7 +1916,7 @@ static inline sycl::event getri_batch(backend_selector std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getri_batch( selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1926,7 +1926,7 @@ static inline sycl::event getri_batch(backend_selector std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getri_batch( selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1937,7 +1937,7 @@ static inline sycl::event getri_batch(backend_selector std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getri_batch( selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1947,7 +1947,7 @@ static inline sycl::event getri_batch(backend_selector std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1957,7 +1957,7 @@ static inline sycl::event getri_batch(backend_selector std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1967,7 +1967,7 @@ static inline sycl::event getri_batch(backend_selector std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1977,36 +1977,36 @@ static inline sycl::event getri_batch(backend_selector std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -2014,12 +2014,12 @@ static inline sycl::event getrs_batch(backend_selector std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -2027,51 +2027,51 @@ static inline sycl::event getrs_batch(backend_selector std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( - backend_selector selector, oneapi::mkl::transpose *trans, + backend_selector selector, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2081,7 +2081,7 @@ static inline sycl::event orgqr_batch(backend_selector std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch( + return oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch( selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2091,7 +2091,7 @@ static inline sycl::event orgqr_batch(backend_selector std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch( + return oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch( selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2101,7 +2101,7 @@ static inline sycl::event orgqr_batch(backend_selector std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch( + return oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch( selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2111,178 +2111,178 @@ static inline sycl::event orgqr_batch(backend_selector std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch( + return oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch( selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, float **a, + oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, double **a, + oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch( + return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2293,7 +2293,7 @@ static inline sycl::event ungqr_batch(backend_selector std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch( + return oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch( selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2305,7 +2305,7 @@ static inline sycl::event ungqr_batch(backend_selector std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch( + return oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch( selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2316,7 +2316,7 @@ static inline sycl::event ungqr_batch(backend_selector std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch( + return oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch( selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2327,7 +2327,7 @@ static inline sycl::event ungqr_batch(backend_selector std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch( + return oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch( selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2335,360 +2335,360 @@ static inline sycl::event ungqr_batch(backend_selector template = nullptr> std::int64_t gebrd_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gebrd_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::LAPACK_BACKEND::gebrd_scratchpad_size(selector.get_queue(), m, n, lda); } -template = nullptr> +template = nullptr> std::int64_t gerqf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gerqf_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::LAPACK_BACKEND::gerqf_scratchpad_size(selector.get_queue(), m, n, lda); } template = nullptr> std::int64_t geqrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_scratchpad_size(selector.get_queue(), m, n, lda); } template = nullptr> std::int64_t gesvd_scratchpad_size(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gesvd_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::gesvd_scratchpad_size( selector.get_queue(), jobu, jobvt, m, n, lda, ldu, ldvt); } template = nullptr> std::int64_t gesvd_scratchpad_size(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return oneapi::mkl::lapack::LAPACK_BACKEND::gesvd_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::gesvd_scratchpad_size( selector.get_queue(), jobu, jobvt, m, n, lda, ldu, ldvt); } template = nullptr> std::int64_t getrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::LAPACK_BACKEND::getrf_scratchpad_size(selector.get_queue(), m, n, lda); } template = nullptr> std::int64_t getri_scratchpad_size(backend_selector selector, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::LAPACK_BACKEND::getri_scratchpad_size(selector.get_queue(), n, lda); } template = nullptr> std::int64_t getrs_scratchpad_size(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::getrs_scratchpad_size( selector.get_queue(), trans, n, nrhs, lda, ldb); } template = nullptr> std::int64_t heevd_scratchpad_size(backend_selector selector, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::heevd_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::LAPACK_BACKEND::heevd_scratchpad_size(selector.get_queue(), jobz, uplo, n, lda); } template = nullptr> std::int64_t hegvd_scratchpad_size(backend_selector selector, - std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::LAPACK_BACKEND::hegvd_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::hegvd_scratchpad_size( selector.get_queue(), itype, jobz, uplo, n, lda, ldb); } template = nullptr> std::int64_t hetrd_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::hetrd_scratchpad_size(selector.get_queue(), + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::LAPACK_BACKEND::hetrd_scratchpad_size(selector.get_queue(), uplo, n, lda); } template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t hetrf_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::hetrf_scratchpad_size(selector.get_queue(), + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::LAPACK_BACKEND::hetrf_scratchpad_size(selector.get_queue(), uplo, n, lda); } template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t orgbr_scratchpad_size(backend_selector selector, - oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgbr_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::LAPACK_BACKEND::orgbr_scratchpad_size(selector.get_queue(), vect, m, n, k, lda); } template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t orgtr_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgtr_scratchpad_size(selector.get_queue(), + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::LAPACK_BACKEND::orgtr_scratchpad_size(selector.get_queue(), uplo, n, lda); } template = nullptr> std::int64_t orgqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::LAPACK_BACKEND::orgqr_scratchpad_size(selector.get_queue(), m, n, k, lda); } template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t ormrq_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ormrq_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::ormrq_scratchpad_size( selector.get_queue(), side, trans, m, n, k, lda, ldc); } template = nullptr> std::int64_t ormqr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ormqr_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::ormqr_scratchpad_size( selector.get_queue(), side, trans, m, n, k, lda, ldc); } template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t ormtr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ormtr_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::ormtr_scratchpad_size( selector.get_queue(), side, uplo, trans, m, n, lda, ldc); } template = nullptr> std::int64_t potrf_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf_scratchpad_size(selector.get_queue(), + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::LAPACK_BACKEND::potrf_scratchpad_size(selector.get_queue(), uplo, n, lda); } -template = nullptr> +template = nullptr> std::int64_t potrs_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::potrs_scratchpad_size( selector.get_queue(), uplo, n, nrhs, lda, ldb); } -template = nullptr> +template = nullptr> std::int64_t potri_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potri_scratchpad_size(selector.get_queue(), + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::LAPACK_BACKEND::potri_scratchpad_size(selector.get_queue(), uplo, n, lda); } -template = nullptr> +template = nullptr> std::int64_t sytrf_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::sytrf_scratchpad_size(selector.get_queue(), + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::LAPACK_BACKEND::sytrf_scratchpad_size(selector.get_queue(), uplo, n, lda); } template = nullptr> std::int64_t syevd_scratchpad_size(backend_selector selector, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::syevd_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::LAPACK_BACKEND::syevd_scratchpad_size(selector.get_queue(), jobz, uplo, n, lda); } template = nullptr> std::int64_t sygvd_scratchpad_size(backend_selector selector, - std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::LAPACK_BACKEND::sygvd_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::sygvd_scratchpad_size( selector.get_queue(), itype, jobz, uplo, n, lda, ldb); } template = nullptr> std::int64_t sytrd_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::sytrd_scratchpad_size(selector.get_queue(), + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::LAPACK_BACKEND::sytrd_scratchpad_size(selector.get_queue(), uplo, n, lda); } template = nullptr> std::int64_t trtrs_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::LAPACK_BACKEND::trtrs_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::trtrs_scratchpad_size( selector.get_queue(), uplo, trans, diag, n, nrhs, lda, ldb); } template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t ungbr_scratchpad_size(backend_selector selector, - oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungbr_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::LAPACK_BACKEND::ungbr_scratchpad_size(selector.get_queue(), vect, m, n, k, lda); } template = nullptr> std::int64_t ungqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::LAPACK_BACKEND::ungqr_scratchpad_size(selector.get_queue(), m, n, k, lda); } template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t ungtr_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungtr_scratchpad_size(selector.get_queue(), + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::LAPACK_BACKEND::ungtr_scratchpad_size(selector.get_queue(), uplo, n, lda); } template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t unmrq_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::LAPACK_BACKEND::unmrq_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::unmrq_scratchpad_size( selector.get_queue(), side, trans, m, n, k, lda, ldc); } template = nullptr> std::int64_t unmqr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::LAPACK_BACKEND::unmqr_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::unmqr_scratchpad_size( selector.get_queue(), side, trans, m, n, k, lda, ldc); } template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t unmtr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::LAPACK_BACKEND::unmtr_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::unmtr_scratchpad_size( selector.get_queue(), side, uplo, trans, m, n, lda, ldc); } -template = nullptr> +template = nullptr> std::int64_t getrf_batch_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, stride_a, stride_ipiv, batch_size); } -template = nullptr> +template = nullptr> std::int64_t getri_batch_scratchpad_size(backend_selector selector, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size( selector.get_queue(), n, lda, stride_a, stride_ipiv, batch_size); } -template = nullptr> +template = nullptr> std::int64_t getrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size( selector.get_queue(), trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } -template = nullptr> +template = nullptr> std::int64_t geqrf_batch_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, stride_a, stride_tau, batch_size); } -template = nullptr> +template = nullptr> std::int64_t potrf_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size( selector.get_queue(), uplo, n, lda, stride_a, batch_size); } -template = nullptr> +template = nullptr> std::int64_t potrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size( selector.get_queue(), uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t orgqr_batch_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, stride_a, stride_tau, batch_size); } template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t ungqr_batch_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, stride_a, stride_tau, batch_size); } -template = nullptr> +template = nullptr> std::int64_t getrf_batch_scratchpad_size(backend_selector selector, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } -template = nullptr> +template = nullptr> std::int64_t getri_batch_scratchpad_size(backend_selector selector, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size( selector.get_queue(), n, lda, group_count, group_sizes); } -template = nullptr> +template = nullptr> std::int64_t getrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size( selector.get_queue(), trans, n, nrhs, lda, ldb, group_count, group_sizes); } -template = nullptr> +template = nullptr> std::int64_t geqrf_batch_scratchpad_size(backend_selector selector, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t orgqr_batch_scratchpad_size(backend_selector selector, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } -template = nullptr> +template = nullptr> std::int64_t potrf_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size( selector.get_queue(), uplo, n, lda, group_count, group_sizes); } -template = nullptr> +template = nullptr> std::int64_t potrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size( selector.get_queue(), uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t ungqr_batch_scratchpad_size(backend_selector selector, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size( + return oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } diff --git a/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx b/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx index ef6908cc5..5c1bdb095 100644 --- a/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx +++ b/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx @@ -103,96 +103,96 @@ ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); @@ -202,159 +202,159 @@ ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, st ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, @@ -369,53 +369,53 @@ ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, st sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, @@ -465,19 +465,19 @@ ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -485,7 +485,7 @@ ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -525,42 +525,42 @@ ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -662,95 +662,95 @@ ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::comple std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); @@ -762,166 +762,166 @@ ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_ float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -935,49 +935,49 @@ ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_ std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, @@ -1115,21 +1115,21 @@ ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std: std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1137,7 +1137,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1145,28 +1145,28 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1196,93 +1196,93 @@ ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std: std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, @@ -1318,19 +1318,19 @@ ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std: template = nullptr> std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); -template = nullptr> +template = nullptr> std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template = nullptr> std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, @@ -1338,170 +1338,170 @@ std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int6 template = nullptr> std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); +std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, +std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); -template = nullptr> -std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); +std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, +std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); -template = nullptr> +template = nullptr> std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); -template = nullptr> +template = nullptr> std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); -template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +template = nullptr> +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -template = nullptr> +template = nullptr> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); -template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); -template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); -template = nullptr> +template = nullptr> std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> +template = nullptr> std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose *trans, +template = nullptr> +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> +template = nullptr> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +template = nullptr> +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +template = nullptr> +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); @@ -1555,23 +1555,23 @@ ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(sycl::qu std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, - oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, - oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, @@ -1605,77 +1605,77 @@ ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(sycl::qu std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, - oneapi::mkl::transpose trans, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, - oneapi::mkl::transpose trans, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( - sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( - sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, @@ -1686,156 +1686,156 @@ ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, st std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); + sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); + sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( @@ -1845,39 +1845,39 @@ ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, + sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, + sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, + sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, + sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, + sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, + sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, @@ -1921,22 +1921,22 @@ ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> @@ -1961,39 +1961,39 @@ ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( @@ -2049,19 +2049,19 @@ ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, @@ -2095,39 +2095,39 @@ ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &que std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, - oneapi::mkl::uplo *uplo, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, - oneapi::mkl::uplo *uplo, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *lda, + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( diff --git a/include/oneapi/math/lapack/detail/mklcpu/lapack_ct.hpp b/include/oneapi/math/lapack/detail/mklcpu/lapack_ct.hpp index 580ea72d1..21aa41248 100644 --- a/include/oneapi/math/lapack/detail/mklcpu/lapack_ct.hpp +++ b/include/oneapi/math/lapack/detail/mklcpu/lapack_ct.hpp @@ -34,7 +34,7 @@ #include "oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { #define LAPACK_BACKEND mklcpu @@ -42,5 +42,5 @@ namespace lapack { #undef LAPACK_BACKEND } //namespace lapack -} //namespace mkl +} //namespace math } //namespace oneapi diff --git a/include/oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp b/include/oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp index a90c8eefe..3341b9e30 100644 --- a/include/oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp +++ b/include/oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp @@ -33,7 +33,7 @@ #include "oneapi/math/detail/export.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace mklcpu { @@ -41,5 +41,5 @@ namespace mklcpu { } //namespace mklcpu } //namespace lapack -} //namespace mkl +} //namespace math } //namespace oneapi diff --git a/include/oneapi/math/lapack/detail/mklgpu/lapack_ct.hpp b/include/oneapi/math/lapack/detail/mklgpu/lapack_ct.hpp index 23dc557e1..ece801387 100644 --- a/include/oneapi/math/lapack/detail/mklgpu/lapack_ct.hpp +++ b/include/oneapi/math/lapack/detail/mklgpu/lapack_ct.hpp @@ -33,7 +33,7 @@ #include "oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { #define LAPACK_BACKEND mklgpu @@ -41,5 +41,5 @@ namespace lapack { #undef LAPACK_BACKEND } //namespace lapack -} //namespace mkl +} //namespace math } //namespace oneapi diff --git a/include/oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp b/include/oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp index e0fda1d97..4eb216ad2 100644 --- a/include/oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp +++ b/include/oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp @@ -33,7 +33,7 @@ #include "oneapi/math/detail/export.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace mklgpu { @@ -41,5 +41,5 @@ namespace mklgpu { } //namespace mklgpu } //namespace lapack -} //namespace mkl +} //namespace math } //namespace oneapi diff --git a/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hpp b/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hpp index 888863167..126d62ca6 100644 --- a/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hpp +++ b/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hpp @@ -36,7 +36,7 @@ #include "oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { #define LAPACK_BACKEND rocsolver @@ -44,7 +44,7 @@ namespace lapack { #undef LAPACK_BACKEND } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_DETAIL_ROCSOLVER_LAPACK_CT_HPP_ diff --git a/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hxx b/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hxx index 774441409..40468215c 100644 --- a/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hxx +++ b/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hxx @@ -28,7 +28,7 @@ static inline void gebrd(backend_selector selector, std::int sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, @@ -36,7 +36,7 @@ static inline void gebrd(backend_selector selector, std::int sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, @@ -44,7 +44,7 @@ static inline void gebrd(backend_selector selector, std::int sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, @@ -54,21 +54,21 @@ static inline void gebrd(backend_selector selector, std::int sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, + oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, @@ -76,7 +76,7 @@ static inline void gerqf(backend_selector selector, std::int sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, @@ -84,7 +84,7 @@ static inline void gerqf(backend_selector selector, std::int sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, @@ -92,21 +92,21 @@ static inline void geqrf(backend_selector selector, std::int sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, @@ -114,7 +114,7 @@ static inline void geqrf(backend_selector selector, std::int sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, @@ -122,21 +122,21 @@ static inline void getrf(backend_selector selector, std::int sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, @@ -144,7 +144,7 @@ static inline void getrf(backend_selector selector, std::int sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, @@ -152,20 +152,20 @@ static inline void getri(backend_selector selector, std::int sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, @@ -173,446 +173,446 @@ static inline void getri(backend_selector selector, std::int sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void getrs(backend_selector selector, oneapi::mkl::transpose trans, +static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, + oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(backend_selector selector, oneapi::mkl::transpose trans, +static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, + oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(backend_selector selector, oneapi::mkl::transpose trans, +static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, + oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(backend_selector selector, oneapi::mkl::transpose trans, +static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, + oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void gesvd(backend_selector selector, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, + oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(backend_selector selector, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, + oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(backend_selector selector, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, + oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(backend_selector selector, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, + oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void heevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +static inline void heevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, + oneapi::math::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void heevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +static inline void heevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, + oneapi::math::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, + oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, + oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void hetrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void hetrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void hetrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void hetrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void orgbr(backend_selector selector, oneapi::mkl::generate vec, +static inline void orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgbr(backend_selector selector, oneapi::mkl::generate vec, +static inline void orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ormtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void ormtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, + oneapi::math::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void ormtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, + oneapi::math::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void ormqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potri(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void syevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +static inline void syevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, + oneapi::math::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void syevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +static inline void syevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, + oneapi::math::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, + oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, + oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void sytrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void sytrd(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + oneapi::math::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, + oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void ungbr(backend_selector selector, oneapi::mkl::generate vec, +static inline void ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungbr(backend_selector selector, oneapi::mkl::generate vec, +static inline void ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + oneapi::math::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, @@ -620,7 +620,7 @@ static inline void ungqr(backend_selector selector, std::int std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, @@ -628,83 +628,83 @@ static inline void ungqr(backend_selector selector, std::int std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungtr(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, + oneapi::math::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void unmrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmrq(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmrq(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +static inline void unmqr(backend_selector selector, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, + oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void unmtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, + oneapi::math::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(backend_selector selector, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +static inline void unmtr(backend_selector selector, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, + oneapi::math::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, @@ -712,7 +712,7 @@ static inline void geqrf_batch(backend_selector selector, st std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, + oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, @@ -720,7 +720,7 @@ static inline void geqrf_batch(backend_selector selector, st std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, + oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, @@ -730,7 +730,7 @@ static inline void geqrf_batch(backend_selector selector, st std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, + oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, @@ -740,7 +740,7 @@ static inline void geqrf_batch(backend_selector selector, st std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, + oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, @@ -748,7 +748,7 @@ static inline void getri_batch(backend_selector selector, st sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, + oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -757,7 +757,7 @@ static inline void getri_batch(backend_selector selector, st sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, + oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -767,7 +767,7 @@ static inline void getri_batch(backend_selector selector, st std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, + oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -777,53 +777,53 @@ static inline void getri_batch(backend_selector selector, st std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, + oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } @@ -832,7 +832,7 @@ static inline void getrf_batch(backend_selector selector, st std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, + oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -841,7 +841,7 @@ static inline void getrf_batch(backend_selector selector, st std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, + oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -852,7 +852,7 @@ static inline void getrf_batch(backend_selector selector, st std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, + oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -863,7 +863,7 @@ static inline void getrf_batch(backend_selector selector, st std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, + oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } @@ -872,7 +872,7 @@ static inline void orgqr_batch(backend_selector selector, st std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, + oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void orgqr_batch(backend_selector selector, std::int64_t m, @@ -880,76 +880,76 @@ static inline void orgqr_batch(backend_selector selector, st std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, + oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, + oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, + oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, + oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, + oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::mkl::uplo uplo, +static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } @@ -960,7 +960,7 @@ static inline void ungqr_batch(backend_selector selector, st std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, + oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } static inline void ungqr_batch(backend_selector selector, std::int64_t m, @@ -970,7 +970,7 @@ static inline void ungqr_batch(backend_selector selector, st std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - oneapi::mkl::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, + oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } @@ -981,7 +981,7 @@ static inline sycl::event gebrd(backend_selector selector, s float *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + return oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, @@ -989,7 +989,7 @@ static inline sycl::event gebrd(backend_selector selector, s double *tauq, double *taup, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + return oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, @@ -997,7 +997,7 @@ static inline sycl::event gebrd(backend_selector selector, s float *tauq, float *taup, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + return oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, @@ -1006,21 +1006,21 @@ static inline sycl::event gebrd(backend_selector selector, s std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + return oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, @@ -1028,7 +1028,7 @@ static inline sycl::event gerqf(backend_selector selector, s std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, @@ -1036,7 +1036,7 @@ static inline sycl::event gerqf(backend_selector selector, s std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, @@ -1044,21 +1044,21 @@ static inline sycl::event geqrf(backend_selector selector, s std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, @@ -1066,7 +1066,7 @@ static inline sycl::event geqrf(backend_selector selector, s std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, + return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, @@ -1074,21 +1074,21 @@ static inline sycl::event getrf(backend_selector selector, s std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, @@ -1096,516 +1096,516 @@ static inline sycl::event getrf(backend_selector selector, s std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + return oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + return oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + return oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, + return oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, + return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event heevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +static inline sycl::event heevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + return oneapi::math::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event heevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +static inline sycl::event heevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + return oneapi::math::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hegvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hegvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + return oneapi::math::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + return oneapi::math::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, + return oneapi::math::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, + return oneapi::math::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, + return oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, + return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event syevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, +static inline sycl::event syevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + return oneapi::math::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event syevd(backend_selector selector, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, +static inline sycl::event syevd(backend_selector selector, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + return oneapi::math::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sygvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sygvd(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + return oneapi::math::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, + return oneapi::math::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, + return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungbr(backend_selector selector, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, @@ -1613,7 +1613,7 @@ static inline sycl::event ungqr(backend_selector selector, s std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, @@ -1621,91 +1621,91 @@ static inline sycl::event ungqr(backend_selector selector, s std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, + return oneapi::math::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmrq(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmqr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, + return oneapi::math::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmtr(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, + return oneapi::math::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } @@ -1715,7 +1715,7 @@ static inline sycl::event geqrf_batch(backend_selector selec std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1725,7 +1725,7 @@ static inline sycl::event geqrf_batch(backend_selector selec std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1735,7 +1735,7 @@ static inline sycl::event geqrf_batch(backend_selector selec std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1746,7 +1746,7 @@ static inline sycl::event geqrf_batch(backend_selector selec std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1755,7 +1755,7 @@ static inline sycl::event geqrf_batch(backend_selector selec std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1764,7 +1764,7 @@ static inline sycl::event geqrf_batch(backend_selector selec std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1774,7 +1774,7 @@ static inline sycl::event geqrf_batch(backend_selector selec std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1784,7 +1784,7 @@ static inline sycl::event geqrf_batch(backend_selector selec std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1794,7 +1794,7 @@ static inline sycl::event getrf_batch(backend_selector selec std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1804,7 +1804,7 @@ static inline sycl::event getrf_batch(backend_selector selec std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1814,7 +1814,7 @@ static inline sycl::event getrf_batch(backend_selector selec std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1825,7 +1825,7 @@ static inline sycl::event getrf_batch(backend_selector selec std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1835,7 +1835,7 @@ static inline sycl::event getrf_batch(backend_selector selec std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1845,7 +1845,7 @@ static inline sycl::event getrf_batch(backend_selector selec std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1855,7 +1855,7 @@ static inline sycl::event getrf_batch(backend_selector selec std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1865,7 +1865,7 @@ static inline sycl::event getrf_batch(backend_selector selec std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1875,7 +1875,7 @@ static inline sycl::event getri_batch(backend_selector selec std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1885,7 +1885,7 @@ static inline sycl::event getri_batch(backend_selector selec std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1895,7 +1895,7 @@ static inline sycl::event getri_batch(backend_selector selec std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1906,7 +1906,7 @@ static inline sycl::event getri_batch(backend_selector selec std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, + return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -1915,7 +1915,7 @@ static inline sycl::event getri_batch(backend_selector selec std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1924,7 +1924,7 @@ static inline sycl::event getri_batch(backend_selector selec std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1934,7 +1934,7 @@ static inline sycl::event getri_batch(backend_selector selec std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1944,95 +1944,95 @@ static inline sycl::event getri_batch(backend_selector selec std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, + return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs_batch( + return oneapi::math::lapack::rocsolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs_batch( + return oneapi::math::lapack::rocsolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( - backend_selector selector, oneapi::mkl::transpose trans, std::int64_t n, + backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs_batch( + return oneapi::math::lapack::rocsolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( - backend_selector selector, oneapi::mkl::transpose trans, std::int64_t n, + backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs_batch( + return oneapi::math::lapack::rocsolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( - backend_selector selector, oneapi::mkl::transpose *trans, std::int64_t *n, + backend_selector selector, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2042,7 +2042,7 @@ static inline sycl::event orgqr_batch(backend_selector selec std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2052,7 +2052,7 @@ static inline sycl::event orgqr_batch(backend_selector selec std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2062,7 +2062,7 @@ static inline sycl::event orgqr_batch(backend_selector selec std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2072,178 +2072,178 @@ static inline sycl::event orgqr_batch(backend_selector selec std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, float **a, + oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, double **a, + oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, + return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, + return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2254,7 +2254,7 @@ static inline sycl::event ungqr_batch(backend_selector selec std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2265,7 +2265,7 @@ static inline sycl::event ungqr_batch(backend_selector selec std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, + return oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } @@ -2275,7 +2275,7 @@ static inline sycl::event ungqr_batch(backend_selector selec std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2286,7 +2286,7 @@ static inline sycl::event ungqr_batch(backend_selector selec std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}) { - return oneapi::mkl::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, + return oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -2295,211 +2295,211 @@ static inline sycl::event ungqr_batch(backend_selector selec template std::int64_t gebrd_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::gebrd_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::rocsolver::gebrd_scratchpad_size(selector.get_queue(), m, n, lda); } template std::int64_t gerqf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::gerqf_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::rocsolver::gerqf_scratchpad_size(selector.get_queue(), m, n, lda); } template std::int64_t geqrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::geqrf_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::rocsolver::geqrf_scratchpad_size(selector.get_queue(), m, n, lda); } template std::int64_t gesvd_scratchpad_size(backend_selector selector, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return oneapi::mkl::lapack::rocsolver::gesvd_scratchpad_size( + return oneapi::math::lapack::rocsolver::gesvd_scratchpad_size( selector.get_queue(), jobu, jobvt, m, n, lda, ldu, ldvt); } template std::int64_t getrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::getrf_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::rocsolver::getrf_scratchpad_size(selector.get_queue(), m, n, lda); } template std::int64_t getri_scratchpad_size(backend_selector selector, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::getri_scratchpad_size(selector.get_queue(), n, + return oneapi::math::lapack::rocsolver::getri_scratchpad_size(selector.get_queue(), n, lda); } template std::int64_t getrs_scratchpad_size(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::rocsolver::getrs_scratchpad_size(selector.get_queue(), + return oneapi::math::lapack::rocsolver::getrs_scratchpad_size(selector.get_queue(), trans, n, nrhs, lda, ldb); } template std::int64_t heevd_scratchpad_size(backend_selector selector, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::heevd_scratchpad_size(selector.get_queue(), jobz, + return oneapi::math::lapack::rocsolver::heevd_scratchpad_size(selector.get_queue(), jobz, uplo, n, lda); } template std::int64_t hegvd_scratchpad_size(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::rocsolver::hegvd_scratchpad_size( + return oneapi::math::lapack::rocsolver::hegvd_scratchpad_size( selector.get_queue(), itype, jobz, uplo, n, lda, ldb); } template std::int64_t hetrd_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::hetrd_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::rocsolver::hetrd_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t hetrf_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::hetrf_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::rocsolver::hetrf_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t orgbr_scratchpad_size(backend_selector selector, - oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::orgbr_scratchpad_size(selector.get_queue(), vect, + return oneapi::math::lapack::rocsolver::orgbr_scratchpad_size(selector.get_queue(), vect, m, n, k, lda); } template std::int64_t orgtr_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::orgtr_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::rocsolver::orgtr_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t orgqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::orgqr_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::rocsolver::orgqr_scratchpad_size(selector.get_queue(), m, n, k, lda); } template std::int64_t ormrq_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::rocsolver::ormrq_scratchpad_size(selector.get_queue(), side, + return oneapi::math::lapack::rocsolver::ormrq_scratchpad_size(selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t ormqr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::rocsolver::ormqr_scratchpad_size(selector.get_queue(), side, + return oneapi::math::lapack::rocsolver::ormqr_scratchpad_size(selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t ormtr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::rocsolver::ormtr_scratchpad_size( + return oneapi::math::lapack::rocsolver::ormtr_scratchpad_size( selector.get_queue(), side, uplo, trans, m, n, lda, ldc); } template std::int64_t potrf_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::potrf_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::rocsolver::potrf_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t potrs_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::rocsolver::potrs_scratchpad_size(selector.get_queue(), uplo, + return oneapi::math::lapack::rocsolver::potrs_scratchpad_size(selector.get_queue(), uplo, n, nrhs, lda, ldb); } template std::int64_t potri_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::potri_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::rocsolver::potri_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t sytrf_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::sytrf_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::rocsolver::sytrf_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t syevd_scratchpad_size(backend_selector selector, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::syevd_scratchpad_size(selector.get_queue(), jobz, + return oneapi::math::lapack::rocsolver::syevd_scratchpad_size(selector.get_queue(), jobz, uplo, n, lda); } template std::int64_t sygvd_scratchpad_size(backend_selector selector, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::rocsolver::sygvd_scratchpad_size( + return oneapi::math::lapack::rocsolver::sygvd_scratchpad_size( selector.get_queue(), itype, jobz, uplo, n, lda, ldb); } template std::int64_t sytrd_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::sytrd_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::rocsolver::sytrd_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t trtrs_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::mkl::lapack::rocsolver::trtrs_scratchpad_size( + return oneapi::math::lapack::rocsolver::trtrs_scratchpad_size( selector.get_queue(), uplo, trans, diag, n, nrhs, lda, ldb); } template std::int64_t ungbr_scratchpad_size(backend_selector selector, - oneapi::mkl::generate vect, std::int64_t m, std::int64_t n, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::ungbr_scratchpad_size(selector.get_queue(), vect, + return oneapi::math::lapack::rocsolver::ungbr_scratchpad_size(selector.get_queue(), vect, m, n, k, lda); } template std::int64_t ungqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::ungqr_scratchpad_size(selector.get_queue(), m, n, + return oneapi::math::lapack::rocsolver::ungqr_scratchpad_size(selector.get_queue(), m, n, k, lda); } template std::int64_t ungtr_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::mkl::lapack::rocsolver::ungtr_scratchpad_size(selector.get_queue(), uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { + return oneapi::math::lapack::rocsolver::ungtr_scratchpad_size(selector.get_queue(), uplo, n, lda); } template std::int64_t unmrq_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::rocsolver::unmrq_scratchpad_size(selector.get_queue(), side, + return oneapi::math::lapack::rocsolver::unmrq_scratchpad_size(selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t unmqr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::rocsolver::unmqr_scratchpad_size(selector.get_queue(), side, + return oneapi::math::lapack::rocsolver::unmqr_scratchpad_size(selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t unmtr_scratchpad_size(backend_selector selector, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return oneapi::mkl::lapack::rocsolver::unmtr_scratchpad_size( + return oneapi::math::lapack::rocsolver::unmtr_scratchpad_size( selector.get_queue(), side, uplo, trans, m, n, lda, ldc); } template @@ -2507,23 +2507,23 @@ std::int64_t getrf_batch_scratchpad_size(backend_selector se std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return oneapi::mkl::lapack::rocsolver::getrf_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::getrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, stride_a, stride_ipiv, batch_size); } template std::int64_t getri_batch_scratchpad_size(backend_selector selector, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return oneapi::mkl::lapack::rocsolver::getri_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::getri_batch_scratchpad_size( selector.get_queue(), n, lda, stride_a, stride_ipiv, batch_size); } template std::int64_t getrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return oneapi::mkl::lapack::rocsolver::getrs_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::getrs_batch_scratchpad_size( selector.get_queue(), trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } @@ -2532,22 +2532,22 @@ std::int64_t geqrf_batch_scratchpad_size(backend_selector se std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return oneapi::mkl::lapack::rocsolver::geqrf_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::geqrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, stride_a, stride_tau, batch_size); } template std::int64_t potrf_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - return oneapi::mkl::lapack::rocsolver::potrf_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::potrf_batch_scratchpad_size( selector.get_queue(), uplo, n, lda, stride_a, batch_size); } template std::int64_t potrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return oneapi::mkl::lapack::rocsolver::potrs_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::potrs_batch_scratchpad_size( selector.get_queue(), uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template @@ -2555,7 +2555,7 @@ std::int64_t orgqr_batch_scratchpad_size(backend_selector se std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return oneapi::mkl::lapack::rocsolver::orgqr_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::orgqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, stride_a, stride_tau, batch_size); } template @@ -2563,36 +2563,36 @@ std::int64_t ungqr_batch_scratchpad_size(backend_selector se std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return oneapi::mkl::lapack::rocsolver::ungqr_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::ungqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, stride_a, stride_tau, batch_size); } template std::int64_t getrf_batch_scratchpad_size(backend_selector selector, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::rocsolver::getrf_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::getrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } template std::int64_t getri_batch_scratchpad_size(backend_selector selector, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::rocsolver::getri_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::getri_batch_scratchpad_size( selector.get_queue(), n, lda, group_count, group_sizes); } template std::int64_t getrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::rocsolver::getrs_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::getrs_batch_scratchpad_size( selector.get_queue(), trans, n, nrhs, lda, ldb, group_count, group_sizes); } template std::int64_t geqrf_batch_scratchpad_size(backend_selector selector, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::rocsolver::geqrf_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::geqrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } template @@ -2600,23 +2600,23 @@ std::int64_t orgqr_batch_scratchpad_size(backend_selector se std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::rocsolver::orgqr_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::orgqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } template std::int64_t potrf_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::rocsolver::potrf_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::potrf_batch_scratchpad_size( selector.get_queue(), uplo, n, lda, group_count, group_sizes); } template std::int64_t potrs_batch_scratchpad_size(backend_selector selector, - oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::rocsolver::potrs_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::potrs_batch_scratchpad_size( selector.get_queue(), uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template @@ -2624,6 +2624,6 @@ std::int64_t ungqr_batch_scratchpad_size(backend_selector se std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return oneapi::mkl::lapack::rocsolver::ungqr_batch_scratchpad_size( + return oneapi::math::lapack::rocsolver::ungqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } diff --git a/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp index 28d02a542..905c46085 100644 --- a/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp +++ b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp @@ -35,7 +35,7 @@ #include "oneapi/math/detail/export.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace rocsolver { @@ -43,7 +43,7 @@ namespace rocsolver { } // namespace rocsolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_LAPACK_ROCSOLVER_HPP_ diff --git a/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx index 819bb1cd6..11a9fe326 100644 --- a/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx +++ b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx @@ -127,42 +127,42 @@ ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, @@ -170,7 +170,7 @@ ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi:: sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, @@ -178,62 +178,62 @@ ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi:: sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); @@ -246,194 +246,194 @@ ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, st sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, @@ -451,19 +451,19 @@ ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, st sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -471,7 +471,7 @@ ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mk sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -479,7 +479,7 @@ ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mk sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -487,7 +487,7 @@ ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mk sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, @@ -495,16 +495,16 @@ ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mk sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, @@ -563,21 +563,21 @@ ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -586,7 +586,7 @@ ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -633,41 +633,41 @@ ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -675,7 +675,7 @@ ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std: sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -802,112 +802,112 @@ ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::comple std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, +ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); @@ -922,201 +922,201 @@ ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_ std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, +ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, +ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1133,56 +1133,56 @@ ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_ std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, @@ -1345,7 +1345,7 @@ ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std: std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, @@ -1353,7 +1353,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, @@ -1361,7 +1361,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1370,7 +1370,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, @@ -1379,7 +1379,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, @@ -1387,7 +1387,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, @@ -1395,7 +1395,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1403,7 +1403,7 @@ ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpos std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, @@ -1438,71 +1438,71 @@ ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std: double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -1510,7 +1510,7 @@ ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo upl std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -1518,28 +1518,28 @@ ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo upl std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, @@ -1591,8 +1591,8 @@ ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64 std::int64_t lda); template -ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); @@ -1605,36 +1605,36 @@ ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64 std::int64_t lda); template -ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template @@ -1642,65 +1642,65 @@ ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64 std::int64_t k, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, +ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); @@ -1709,25 +1709,25 @@ ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64 std::int64_t k, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, +ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); @@ -1746,7 +1746,7 @@ ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std: template ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, - oneapi::mkl::transpose trans, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, @@ -1761,13 +1761,13 @@ ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std: std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, @@ -1800,7 +1800,7 @@ ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std: template ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template @@ -1816,13 +1816,13 @@ ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std: std::int64_t *group_sizes); template -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template -ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, diff --git a/include/oneapi/math/lapack/exceptions.hpp b/include/oneapi/math/lapack/exceptions.hpp index da205cc1a..33eddea4d 100644 --- a/include/oneapi/math/lapack/exceptions.hpp +++ b/include/oneapi/math/lapack/exceptions.hpp @@ -20,12 +20,12 @@ #pragma once namespace oneapi { -namespace mkl { +namespace math { namespace lapack { class exception { public: - exception(oneapi::mkl::exception *_ex, std::int64_t info, std::int64_t detail = 0) + exception(oneapi::math::exception *_ex, std::int64_t info, std::int64_t detail = 0) : _info(info), _detail(detail), _ex(_ex) {} @@ -42,27 +42,27 @@ class exception { private: std::int64_t _info; std::int64_t _detail; - mkl::exception *_ex; + math::exception *_ex; }; -class computation_error : public oneapi::mkl::computation_error, - public oneapi::mkl::lapack::exception { +class computation_error : public oneapi::math::computation_error, + public oneapi::math::lapack::exception { public: computation_error(const std::string &function, const std::string &info, std::int64_t code) - : oneapi::mkl::computation_error("LAPACK", function, info), - oneapi::mkl::lapack::exception(this, code) {} - using oneapi::mkl::computation_error::what; + : oneapi::math::computation_error("LAPACK", function, info), + oneapi::math::lapack::exception(this, code) {} + using oneapi::math::computation_error::what; }; -class batch_error : public oneapi::mkl::batch_error, public oneapi::mkl::lapack::exception { +class batch_error : public oneapi::math::batch_error, public oneapi::math::lapack::exception { public: batch_error(const std::string &function, const std::string &info, std::int64_t num_errors, std::vector ids = {}, std::vector exceptions = {}) - : oneapi::mkl::batch_error("LAPACK", function, info), - oneapi::mkl::lapack::exception(this, num_errors), + : oneapi::math::batch_error("LAPACK", function, info), + oneapi::math::lapack::exception(this, num_errors), _ids(ids), _exceptions(exceptions) {} - using oneapi::mkl::batch_error::what; + using oneapi::math::batch_error::what; const std::vector &ids() const { return _ids; } @@ -75,16 +75,16 @@ class batch_error : public oneapi::mkl::batch_error, public oneapi::mkl::lapack: std::vector _exceptions; }; -class invalid_argument : public oneapi::mkl::invalid_argument, - public oneapi::mkl::lapack::exception { +class invalid_argument : public oneapi::math::invalid_argument, + public oneapi::math::lapack::exception { public: invalid_argument(const std::string &function, const std::string &info, std::int64_t arg_position = 0, std::int64_t detail = 0) - : oneapi::mkl::invalid_argument("LAPACK", function, info), - oneapi::mkl::lapack::exception(this, arg_position, detail) {} - using oneapi::mkl::invalid_argument::what; + : oneapi::math::invalid_argument("LAPACK", function, info), + oneapi::math::lapack::exception(this, arg_position, detail) {} + using oneapi::math::invalid_argument::what; }; } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/include/oneapi/math/lapack/types.hpp b/include/oneapi/math/lapack/types.hpp index 8dbe19e2e..e43721a50 100644 --- a/include/oneapi/math/lapack/types.hpp +++ b/include/oneapi/math/lapack/types.hpp @@ -29,7 +29,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace internal { @@ -93,5 +93,5 @@ using is_complex_floating_point = typename enable_if::value>::type*; } // namespace internal } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp b/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp index 7148f8a24..874432e19 100644 --- a/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp +++ b/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp @@ -70,25 +70,25 @@ #include "oneapi/math/rng/detail/engine_impl.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace curand { -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); } // namespace curand } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_RNG_CURAND_HPP_ diff --git a/include/oneapi/math/rng/detail/engine_impl.hpp b/include/oneapi/math/rng/detail/engine_impl.hpp index 36429b066..2d3c2b562 100644 --- a/include/oneapi/math/rng/detail/engine_impl.hpp +++ b/include/oneapi/math/rng/detail/engine_impl.hpp @@ -34,7 +34,7 @@ #include "oneapi/math/types.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace detail { @@ -191,7 +191,7 @@ class engine_impl { } // namespace detail } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_RNG_ENGINE_IMPL_HPP_ diff --git a/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp b/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp index 0e97d6003..ff875be2f 100644 --- a/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp +++ b/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp @@ -31,25 +31,25 @@ #include "oneapi/math/rng/detail/engine_impl.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace mklcpu { -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); } // namespace mklcpu } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_RNG_MKLCPU_HPP_ diff --git a/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp b/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp index 0de6f9b45..9fc40dfcd 100644 --- a/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp +++ b/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp @@ -31,25 +31,25 @@ #include "oneapi/math/rng/detail/engine_impl.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace mklgpu { -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); } // namespace mklgpu } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_RNG_MKLGPU_HPP_ diff --git a/include/oneapi/math/rng/detail/rng_loader.hpp b/include/oneapi/math/rng/detail/rng_loader.hpp index dcb92ea84..19de23529 100644 --- a/include/oneapi/math/rng/detail/rng_loader.hpp +++ b/include/oneapi/math/rng/detail/rng_loader.hpp @@ -33,25 +33,25 @@ #include "oneapi/math/rng/detail/engine_impl.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace detail { -ONEMATH_EXPORT engine_impl* create_philox4x32x10(oneapi::mkl::device libkey, sycl::queue queue, +ONEMATH_EXPORT engine_impl* create_philox4x32x10(oneapi::math::device libkey, sycl::queue queue, std::uint64_t seed); -ONEMATH_EXPORT engine_impl* create_philox4x32x10(oneapi::mkl::device libkey, sycl::queue queue, +ONEMATH_EXPORT engine_impl* create_philox4x32x10(oneapi::math::device libkey, sycl::queue queue, std::initializer_list seed); -ONEMATH_EXPORT engine_impl* create_mrg32k3a(oneapi::mkl::device libkey, sycl::queue queue, +ONEMATH_EXPORT engine_impl* create_mrg32k3a(oneapi::math::device libkey, sycl::queue queue, std::uint32_t seed); -ONEMATH_EXPORT engine_impl* create_mrg32k3a(oneapi::mkl::device libkey, sycl::queue queue, +ONEMATH_EXPORT engine_impl* create_mrg32k3a(oneapi::math::device libkey, sycl::queue queue, std::initializer_list seed); } // namespace detail } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_RNG_LOADER_HPP_ diff --git a/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp b/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp index 00b0865af..4e2538800 100644 --- a/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp +++ b/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp @@ -72,25 +72,25 @@ #include "oneapi/math/rng/detail/engine_impl.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace rocrand { -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed); -ONEMATH_EXPORT oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a( +ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); } // namespace rocrand } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_RNG_ROCRAND_HPP_ diff --git a/include/oneapi/math/rng/device/detail/bernoulli_impl.hpp b/include/oneapi/math/rng/device/detail/bernoulli_impl.hpp index 83bb92f2d..9fb3d0559 100644 --- a/include/oneapi/math/rng/device/detail/bernoulli_impl.hpp +++ b/include/oneapi/math/rng/device/detail/bernoulli_impl.hpp @@ -20,10 +20,10 @@ #ifndef _MKL_RNG_DEVICE_BERNOULLI_IMPL_HPP_ #define _MKL_RNG_DEVICE_BERNOULLI_IMPL_HPP_ -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { template -class distribution_base> { +class distribution_base> { public: struct param_type { param_type(float p) : p_(p) {} @@ -33,7 +33,7 @@ class distribution_base> { distribution_base(float p) : p_(p) { #ifndef __SYCL_DEVICE_ONLY__ if ((p > 1.0f) || (p < 0.0f)) { - throw oneapi::mkl::invalid_argument("rng", "bernoulli", "p < 0 || p > 1"); + throw oneapi::math::invalid_argument("rng", "bernoulli", "p < 0 || p > 1"); } #endif } @@ -49,7 +49,7 @@ class distribution_base> { void param(const param_type& pt) { #ifndef __SYCL_DEVICE_ONLY__ if ((pt.p_ > 1.0f) || (pt.p_ < 0.0f)) { - throw oneapi::mkl::invalid_argument("rng", "bernoulli", "p < 0 || p > 1"); + throw oneapi::math::invalid_argument("rng", "bernoulli", "p < 0 || p > 1"); } #endif p_ = pt.p_; @@ -84,6 +84,6 @@ class distribution_base> { float p_; }; -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #endif // _MKL_RNG_DEVICE_BERNOULLI_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/beta_impl.hpp b/include/oneapi/math/rng/device/detail/beta_impl.hpp index e412ee157..fcb9db459 100644 --- a/include/oneapi/math/rng/device/detail/beta_impl.hpp +++ b/include/oneapi/math/rng/device/detail/beta_impl.hpp @@ -22,7 +22,7 @@ #include "vm_wrappers.hpp" -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { enum class beta_algorithm { Johnk = 0, Atkinson1, Atkinson2, Atkinson3, Cheng, p1, q1, p1q1 }; @@ -54,7 +54,7 @@ inline DataType beta_c() { } template -class distribution_base> { +class distribution_base> { public: struct param_type { param_type(RealType p, RealType q, RealType a, RealType b) : p_(p), q_(q), a_(a), b_(b) {} @@ -73,13 +73,13 @@ class distribution_base> { set_algorithm(); #ifndef __SYCL_DEVICE_ONLY__ if (p <= RealType(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "beta", "p <= 0"); + throw oneapi::math::invalid_argument("rng", "beta", "p <= 0"); } else if (q <= RealType(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "beta", "q <= 0"); + throw oneapi::math::invalid_argument("rng", "beta", "q <= 0"); } else if (b <= RealType(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "beta", "b <= 0"); + throw oneapi::math::invalid_argument("rng", "beta", "b <= 0"); } #endif } @@ -111,13 +111,13 @@ class distribution_base> { void param(const param_type& pt) { #ifndef __SYCL_DEVICE_ONLY__ if (pt.p_ <= RealType(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "beta", "p <= 0"); + throw oneapi::math::invalid_argument("rng", "beta", "p <= 0"); } else if (pt.q_ <= RealType(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "beta", "q <= 0"); + throw oneapi::math::invalid_argument("rng", "beta", "q <= 0"); } else if (pt.b_ <= RealType(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "beta", "b <= 0"); + throw oneapi::math::invalid_argument("rng", "beta", "b <= 0"); } #endif p_ = pt.p_; @@ -463,6 +463,6 @@ class distribution_base> { beta_algorithm algorithm_; }; -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #endif // _MKL_RNG_DEVICE_BETA_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/bits_impl.hpp b/include/oneapi/math/rng/device/detail/bits_impl.hpp index 83d6831a1..f48364638 100644 --- a/include/oneapi/math/rng/device/detail/bits_impl.hpp +++ b/include/oneapi/math/rng/device/detail/bits_impl.hpp @@ -22,10 +22,10 @@ #include "engine_base.hpp" -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { template -class distribution_base> { +class distribution_base> { protected: template auto generate(EngineType& engine) -> typename std::enable_if< @@ -66,6 +66,6 @@ class distribution_base> { } }; -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #endif // _MKL_RNG_DEVICE_BITS_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/distribution_base.hpp b/include/oneapi/math/rng/device/detail/distribution_base.hpp index 643d2fa6b..b511e20b6 100644 --- a/include/oneapi/math/rng/device/detail/distribution_base.hpp +++ b/include/oneapi/math/rng/device/detail/distribution_base.hpp @@ -25,7 +25,7 @@ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/rng/device/types.hpp" -namespace oneapi::mkl::rng::device { +namespace oneapi::math::rng::device { namespace detail { @@ -65,7 +65,7 @@ class poisson; template class bernoulli; -} // namespace oneapi::mkl::rng::device +} // namespace oneapi::math::rng::device #include "oneapi/math/rng/device/detail/uniform_impl.hpp" #include "oneapi/math/rng/device/detail/gaussian_impl.hpp" diff --git a/include/oneapi/math/rng/device/detail/engine_base.hpp b/include/oneapi/math/rng/device/detail/engine_base.hpp index 6c754dcd4..e104620fd 100644 --- a/include/oneapi/math/rng/device/detail/engine_base.hpp +++ b/include/oneapi/math/rng/device/detail/engine_base.hpp @@ -24,7 +24,7 @@ #include -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { // internal structure to specify state of engine template @@ -33,7 +33,7 @@ struct engine_state {}; template class engine_base {}; -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #include "oneapi/math/rng/device/detail/philox4x32x10_impl.hpp" #include "oneapi/math/rng/device/detail/mrg32k3a_impl.hpp" diff --git a/include/oneapi/math/rng/device/detail/exponential_impl.hpp b/include/oneapi/math/rng/device/detail/exponential_impl.hpp index 9419fc154..3db32fc13 100644 --- a/include/oneapi/math/rng/device/detail/exponential_impl.hpp +++ b/include/oneapi/math/rng/device/detail/exponential_impl.hpp @@ -22,10 +22,10 @@ #include "vm_wrappers.hpp" -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { template -class distribution_base> { +class distribution_base> { public: struct param_type { param_type(RealType a, RealType beta) : a_(a), beta_(beta) {} @@ -36,7 +36,7 @@ class distribution_base> distribution_base(RealType a, RealType beta) : a_(a), beta_(beta) { #ifndef __SYCL_DEVICE_ONLY__ if (beta <= static_cast(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "exponential", "beta <= 0"); + throw oneapi::math::invalid_argument("rng", "exponential", "beta <= 0"); } #endif } @@ -56,7 +56,7 @@ class distribution_base> void param(const param_type& pt) { #ifndef __SYCL_DEVICE_ONLY__ if (pt.beta_ <= static_cast(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "exponential", "beta <= 0"); + throw oneapi::math::invalid_argument("rng", "exponential", "beta <= 0"); } #endif a_ = pt.a_; @@ -123,18 +123,18 @@ class distribution_base> RealType beta_; friend class distribution_base< - oneapi::mkl::rng::device::poisson>; + oneapi::math::rng::device::poisson>; friend class distribution_base< - oneapi::mkl::rng::device::poisson>; - friend class distribution_base>; + oneapi::math::rng::device::poisson>; + friend class distribution_base>; friend class distribution_base< - oneapi::mkl::rng::device::gamma>; + oneapi::math::rng::device::gamma>; friend class distribution_base< - oneapi::mkl::rng::device::gamma>; + oneapi::math::rng::device::gamma>; friend class distribution_base< - oneapi::mkl::rng::device::gamma>; + oneapi::math::rng::device::gamma>; }; -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #endif // _MKL_RNG_DEVICE_EXPONENTIAL_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/gamma_impl.hpp b/include/oneapi/math/rng/device/detail/gamma_impl.hpp index 11397a69d..7371f8a8a 100644 --- a/include/oneapi/math/rng/device/detail/gamma_impl.hpp +++ b/include/oneapi/math/rng/device/detail/gamma_impl.hpp @@ -22,7 +22,7 @@ #include "vm_wrappers.hpp" -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { enum class gamma_algorithm { Exponential = 0, Vaduva, EPD_Transform, Marsaglia }; @@ -54,7 +54,7 @@ inline DataType gamma_c06() { } template -class distribution_base> { +class distribution_base> { public: struct param_type { param_type(RealType alpha, RealType a, RealType beta) : alpha_(alpha), a_(a), beta_(beta) {} @@ -71,10 +71,10 @@ class distribution_base> { set_algorithm(); #ifndef __SYCL_DEVICE_ONLY__ if (alpha <= RealType(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "gamma", "alpha <= 0"); + throw oneapi::math::invalid_argument("rng", "gamma", "alpha <= 0"); } else if (beta <= RealType(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "gamma", "beta <= 0"); + throw oneapi::math::invalid_argument("rng", "gamma", "beta <= 0"); } #endif } @@ -102,10 +102,10 @@ class distribution_base> { void param(const param_type& pt) { #ifndef __SYCL_DEVICE_ONLY__ if (pt.alpha_ <= RealType(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "gamma", "alpha <= 0"); + throw oneapi::math::invalid_argument("rng", "gamma", "alpha <= 0"); } else if (pt.beta_ <= RealType(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "gamma", "beta <= 0"); + throw oneapi::math::invalid_argument("rng", "gamma", "beta <= 0"); } #endif alpha_ = pt.alpha_; @@ -250,7 +250,7 @@ class distribution_base> { typename std::conditional>::type { if (algorithm_ == gamma_algorithm::Exponential) { - distribution_base> distr_exp(a_, beta_); + distribution_base> distr_exp(a_, beta_); return distr_exp.generate(engine); } sycl::vec res{}; @@ -262,7 +262,7 @@ class distribution_base> { template RealType generate_single(EngineType& engine) { if (algorithm_ == gamma_algorithm::Exponential) { - distribution_base> distr_exp(a_, beta_); + distribution_base> distr_exp(a_, beta_); RealType z = distr_exp.generate_single(engine); return z; } @@ -282,6 +282,6 @@ class distribution_base> { gamma_algorithm algorithm_; }; -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #endif // _MKL_RNG_DEVICE_GAMMA_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/gaussian_impl.hpp b/include/oneapi/math/rng/device/detail/gaussian_impl.hpp index 4588aea97..ff0b36471 100644 --- a/include/oneapi/math/rng/device/detail/gaussian_impl.hpp +++ b/include/oneapi/math/rng/device/detail/gaussian_impl.hpp @@ -22,7 +22,7 @@ #include "vm_wrappers.hpp" -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { // sqrt(2) template @@ -37,7 +37,7 @@ constexpr inline double sqrt2() { template class distribution_base< - oneapi::mkl::rng::device::gaussian> { + oneapi::math::rng::device::gaussian> { public: struct param_type { param_type(RealType mean, RealType stddev) : mean_(mean), stddev_(stddev) {} @@ -49,7 +49,7 @@ class distribution_base< flag_ = false; #ifndef __SYCL_DEVICE_ONLY__ if (stddev <= RealType(0)) { - throw oneapi::mkl::invalid_argument("rng", "gaussian", "stddev <= 0"); + throw oneapi::math::invalid_argument("rng", "gaussian", "stddev <= 0"); } #endif } @@ -69,7 +69,7 @@ class distribution_base< void param(const param_type& pt) { #ifndef __SYCL_DEVICE_ONLY__ if (pt.stddev_ <= RealType(0)) { - throw oneapi::mkl::invalid_argument("rng", "gaussian", "stddev <= 0"); + throw oneapi::math::invalid_argument("rng", "gaussian", "stddev <= 0"); } #endif mean_ = pt.mean_; @@ -185,17 +185,17 @@ class distribution_base< RealType u2_; friend class distribution_base< - oneapi::mkl::rng::device::lognormal>; + oneapi::math::rng::device::lognormal>; friend class distribution_base< - oneapi::mkl::rng::device::poisson>; + oneapi::math::rng::device::poisson>; friend class distribution_base< - oneapi::mkl::rng::device::poisson>; + oneapi::math::rng::device::poisson>; }; #if MKL_RNG_USE_BINARY_CODE template -class distribution_base> { +class distribution_base> { public: struct param_type { param_type(RealType mean, RealType stddev) : mean_(mean), stddev_(stddev) {} @@ -206,7 +206,7 @@ class distribution_base -class distribution_base> { +class distribution_base> { public: struct param_type { param_type(RealType m, RealType s, RealType displ, RealType scale) @@ -43,7 +43,7 @@ class distribution_base> { scale_(scale) { #ifndef __SYCL_DEVICE_ONLY__ if (scale <= static_cast(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "lognormal", "scale <= 0"); + throw oneapi::math::invalid_argument("rng", "lognormal", "scale <= 0"); } #endif } @@ -71,7 +71,7 @@ class distribution_base> { void param(const param_type& pt) { #ifndef __SYCL_DEVICE_ONLY__ if (pt.scale_ <= static_cast(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "lognormal", "scale <= 0"); + throw oneapi::math::invalid_argument("rng", "lognormal", "scale <= 0"); } #endif gaussian_.param({ pt.m_, pt.s_ }); @@ -94,12 +94,12 @@ class distribution_base> { return sycl::exp(res) * scale_ + displ_; } - distribution_base> + distribution_base> gaussian_; RealType displ_; RealType scale_; }; -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #endif // _MKL_RNG_DEVICE_LOGNORMAL_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/mcg31m1_impl.hpp b/include/oneapi/math/rng/device/detail/mcg31m1_impl.hpp index 72447bc5d..96a52c120 100644 --- a/include/oneapi/math/rng/device/detail/mcg31m1_impl.hpp +++ b/include/oneapi/math/rng/device/detail/mcg31m1_impl.hpp @@ -20,7 +20,7 @@ #ifndef _MKL_RNG_DEVICE_MCG31M1_IMPL_HPP_ #define _MKL_RNG_DEVICE_MCG31M1_IMPL_HPP_ -namespace oneapi::mkl::rng::device { +namespace oneapi::math::rng::device { template class mcg31m1; @@ -71,7 +71,7 @@ struct mcg31m1_param { }; template -struct engine_state> { +struct engine_state> { std::uint32_t s; }; @@ -134,14 +134,14 @@ static inline std::uint64_t power(std::uint64_t a, std::uint64_t n) { } template -static inline void skip_ahead(engine_state>& state, +static inline void skip_ahead(engine_state>& state, std::uint64_t num_to_skip) { std::uint64_t loc_A = power(static_cast(mcg31m1_param::a), num_to_skip); state.s = custom_mod(loc_A * static_cast(state.s)); } template -static inline void init(engine_state>& state, +static inline void init(engine_state>& state, std::uint32_t seed, std::uint64_t offset) { state.s = custom_mod(seed); if (state.s == 0) @@ -151,7 +151,7 @@ static inline void init(engine_state> template static inline sycl::vec generate( - engine_state>& state) { + engine_state>& state) { sycl::vec x(state.s); sycl::vec res; #ifndef __HIPSYCL__ @@ -167,7 +167,7 @@ static inline sycl::vec generate( template static inline std::uint32_t generate_single( - engine_state>& state) { + engine_state>& state) { std::uint32_t x = state.s; state.s = custom_mod(mcg31m1_param::a * static_cast(state.s)); return x; @@ -176,7 +176,7 @@ static inline std::uint32_t generate_single( } // namespace mcg31m1_impl template -class engine_base> { +class engine_base> { protected: engine_base(std::uint32_t seed, std::uint64_t offset = 0) { mcg31m1_impl::init(this->state_, seed, offset); @@ -223,11 +223,11 @@ class engine_base> { detail::mcg31m1_impl::skip_ahead(this->state_, num_to_skip); } - engine_state> state_; + engine_state> state_; }; } // namespace detail -} // namespace oneapi::mkl::rng::device +} // namespace oneapi::math::rng::device #endif // _MKL_RNG_DEVICE_MCG31M1_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/mcg59_impl.hpp b/include/oneapi/math/rng/device/detail/mcg59_impl.hpp index a70bb323d..176330e1e 100644 --- a/include/oneapi/math/rng/device/detail/mcg59_impl.hpp +++ b/include/oneapi/math/rng/device/detail/mcg59_impl.hpp @@ -20,7 +20,7 @@ #ifndef _MKL_RNG_DEVICE_MCG59_IMPL_HPP_ #define _MKL_RNG_DEVICE_MCG59_IMPL_HPP_ -namespace oneapi::mkl::rng::device { +namespace oneapi::math::rng::device { template class mcg59; @@ -71,7 +71,7 @@ struct mcg59_param { }; template -struct engine_state> { +struct engine_state> { std::uint64_t s; }; @@ -103,14 +103,14 @@ static inline std::uint64_t power(std::uint64_t a, std::uint64_t n) { } template -static inline void skip_ahead(engine_state>& state, +static inline void skip_ahead(engine_state>& state, std::uint64_t num_to_skip) { std::uint64_t loc_A = power(mcg59_param::a, num_to_skip); state.s = custom_mod(loc_A * state.s); } template -static inline void init(engine_state>& state, +static inline void init(engine_state>& state, std::uint64_t seed, std::uint64_t offset) { state.s = seed & mcg59_param::m_64; if (state.s == 0) @@ -121,7 +121,7 @@ static inline void init(engine_state>& template static inline sycl::vec generate( - engine_state>& state) { + engine_state>& state) { sycl::vec res(state.s); #ifndef __HIPSYCL__ res = custom_mod(mcg59_vector_a::vector_a * res); @@ -135,7 +135,7 @@ static inline sycl::vec generate( template static inline std::uint64_t generate_single( - engine_state>& state) { + engine_state>& state) { std::uint64_t x = state.s; state.s = custom_mod(mcg59_param::a * x); return x; @@ -144,7 +144,7 @@ static inline std::uint64_t generate_single( } // namespace mcg59_impl template -class engine_base> { +class engine_base> { protected: engine_base(std::uint64_t seed, std::uint64_t offset = 0) { mcg59_impl::init(this->state_, seed, offset); @@ -266,10 +266,10 @@ class engine_base> { detail::mcg59_impl::skip_ahead(this->state_, num_to_skip); } - engine_state> state_; + engine_state> state_; }; } // namespace detail -} // namespace oneapi::mkl::rng::device +} // namespace oneapi::math::rng::device #endif // _MKL_RNG_DEVICE_MCG59_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp b/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp index f285c7683..4ca6abdae 100644 --- a/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp +++ b/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp @@ -27,7 +27,7 @@ #include "oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp" -namespace oneapi::mkl::rng::device { +namespace oneapi::math::rng::device { template class mrg32k3a; @@ -35,7 +35,7 @@ class mrg32k3a; namespace detail { template -struct engine_state> { +struct engine_state> { std::uint32_t s[6]; }; @@ -176,12 +176,12 @@ static inline void vec3_pow_mod( } template -static inline void skip_ahead(engine_state>& state, +static inline void skip_ahead(engine_state>& state, std::uint64_t n, const std::uint64_t* num_to_skip_ptr) { if (n > 3) { n = 3; #ifndef __SYCL_DEVICE_ONLY__ - throw oneapi::mkl::invalid_argument("rng", "mrg32k3a", + throw oneapi::math::invalid_argument("rng", "mrg32k3a", "period is 2 ^ 191, skip on more than 2^192"); #endif } @@ -190,7 +190,7 @@ static inline void skip_ahead(engine_state -static inline void validate_seed(engine_state>& state) { +static inline void validate_seed(engine_state>& state) { int i; for (i = 0; i < 3; i++) { if (state.s[i] >= mrg32k3a_params::m1) { @@ -212,7 +212,7 @@ static inline void validate_seed(engine_state -static inline void init(engine_state>& state, +static inline void init(engine_state>& state, std::uint64_t n, const std::uint32_t* seed_ptr, std::uint64_t n_offset, const std::uint64_t* offset_ptr) { std::uint64_t i; @@ -231,7 +231,7 @@ static inline void init(engine_state template static inline sycl::vec generate( - engine_state>& state) { + engine_state>& state) { const std::int32_t num_elements = VecSize; sycl::vec res; std::int64_t x, y; @@ -270,7 +270,7 @@ static inline sycl::vec generate( template static inline std::uint32_t generate_single( - engine_state>& state) { + engine_state>& state) { std::uint32_t res; std::int64_t x, y; x = mrg32k3a_params::a12 * static_cast(state.s[1]) - @@ -307,7 +307,7 @@ static inline std::uint32_t generate_single( } // namespace mrg32k3a_impl template -class engine_base> { +class engine_base> { protected: engine_base(std::uint32_t seed, std::uint64_t offset = 0) { mrg32k3a_impl::init(this->state_, 1, &seed, 1, &offset); @@ -375,10 +375,10 @@ class engine_base> { detail::mrg32k3a_impl::skip_ahead(this->state_, num_to_skip.size(), num_to_skip.begin()); } - engine_state> state_; + engine_state> state_; }; } // namespace detail -} // namespace oneapi::mkl::rng::device +} // namespace oneapi::math::rng::device #endif // _MKL_RNG_DEVICE_MRG32K3A_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp b/include/oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp index d1ea8c263..2c69d2949 100644 --- a/include/oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp +++ b/include/oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp @@ -20,7 +20,7 @@ #ifndef _MKL_RNG_DEVICE_MRG32K3A_SKIP_AHEAD_MATRIX_HPP_ #define _MKL_RNG_DEVICE_MRG32K3A_SKIP_AHEAD_MATRIX_HPP_ -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { namespace mrg32k3a_impl { constexpr std::size_t quantity_of_3x3_matrices = 455; // number of 3x3 matrices for skipping @@ -3663,6 +3663,6 @@ static const std::uint32_t skip_ahead_matrix[2][quantity_of_3x3_matrices][3][3] }; } // namespace mrg32k3a_impl -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #endif // _MKL_RNG_DEVICE_MRG32K3A_SKIP_AHEAD_MATRIX_HPP_ diff --git a/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp b/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp index f061bb754..e3b0db9aa 100644 --- a/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp +++ b/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp @@ -22,7 +22,7 @@ #include // std::pair -namespace oneapi::mkl::rng::device { +namespace oneapi::math::rng::device { template class philox4x32x10; @@ -30,7 +30,7 @@ class philox4x32x10; namespace detail { template -struct engine_state> { +struct engine_state> { std::uint32_t key[2]; std::uint32_t counter[4]; std::uint32_t part; @@ -119,7 +119,7 @@ static inline void round_10(std::uint32_t* cnt, std::uint32_t* k) { } template -static inline void skip_ahead(engine_state>& state, +static inline void skip_ahead(engine_state>& state, std::uint64_t num_to_skip) { std::uint64_t num_to_skip_tmp = num_to_skip; std::uint64_t c_inc; @@ -157,7 +157,7 @@ static inline void skip_ahead(engine_state -static inline void skip_ahead(engine_state>& state, +static inline void skip_ahead(engine_state>& state, std::uint64_t n, const std::uint64_t* num_to_skip_ptr) { constexpr std::uint64_t uint_max = 0xFFFFFFFFFFFFFFFF; std::uint64_t post_buffer, pre_buffer; @@ -250,7 +250,7 @@ static inline void skip_ahead(engine_state -static inline void init(engine_state>& state, +static inline void init(engine_state>& state, std::uint64_t n, const std::uint64_t* seed_ptr, std::uint64_t offset) { state.key[0] = static_cast(seed_ptr[0]); state.key[1] = static_cast(seed_ptr[0] >> 32); @@ -270,7 +270,7 @@ static inline void init(engine_state -static inline void init(engine_state>& state, +static inline void init(engine_state>& state, std::uint64_t n, const std::uint64_t* seed_ptr, std::uint64_t n_offset, const std::uint64_t* offset_ptr) { state.key[0] = static_cast(seed_ptr[0]); @@ -293,7 +293,7 @@ static inline void init(engine_state 4 template __attribute__((always_inline)) static inline sycl::vec generate_full( - engine_state>& state) { + engine_state>& state) { const std::int32_t num_elements = VecSize; sycl::vec res; @@ -347,7 +347,7 @@ __attribute__((always_inline)) static inline sycl::vec g // for VecSize <= 4 template __attribute__((always_inline)) static inline sycl::vec generate_small( - engine_state>& state) { + engine_state>& state) { const std::int32_t num_elements = VecSize; sycl::vec res; @@ -383,7 +383,7 @@ __attribute__((always_inline)) static inline sycl::vec g template __attribute__((always_inline)) static inline std::uint32_t generate_single( - engine_state>& state) { + engine_state>& state) { std::uint32_t res; std::uint32_t counter[4]; @@ -413,7 +413,7 @@ __attribute__((always_inline)) static inline std::uint32_t generate_single( } // namespace philox4x32x10_impl template -class engine_base> { +class engine_base> { protected: engine_base(std::uint64_t seed, std::uint64_t offset = 0) { philox4x32x10_impl::init(this->state_, 1, &seed, offset); @@ -543,10 +543,10 @@ class engine_base> { num_to_skip.begin()); } - engine_state> state_; + engine_state> state_; }; } // namespace detail -} // namespace oneapi::mkl::rng::device +} // namespace oneapi::math::rng::device #endif // _MKL_RNG_DEVICE_PHILOX4X32X10_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/poisson_impl.hpp b/include/oneapi/math/rng/device/detail/poisson_impl.hpp index 9fa9b26ec..bbcba7335 100644 --- a/include/oneapi/math/rng/device/detail/poisson_impl.hpp +++ b/include/oneapi/math/rng/device/detail/poisson_impl.hpp @@ -22,7 +22,7 @@ #include -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { // Implementation of Poisson distribution uses 3 methods depending on lambda parameter: // - table-lookup method [1] for small lambdas (lambda < 60) @@ -126,7 +126,7 @@ struct poisson_parameters { }; template -class distribution_base> { +class distribution_base> { public: struct param_type { param_type(double lambda) : lambda_(lambda) {} @@ -136,7 +136,7 @@ class distribution_base> gaussian_ = { 0.0, 1.0 }; - distribution_base> exponential_ = { 0.0, 1.0 }; + distribution_base> gaussian_ = { 0.0, 1.0 }; + distribution_base> exponential_ = { 0.0, 1.0 }; poisson_parameters params_; double lambda_; }; -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #endif // _MKL_RNG_DEVICE_POISSON_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/uniform_bits_impl.hpp b/include/oneapi/math/rng/device/detail/uniform_bits_impl.hpp index 1f9c8a717..2e8f0cdf2 100644 --- a/include/oneapi/math/rng/device/detail/uniform_bits_impl.hpp +++ b/include/oneapi/math/rng/device/detail/uniform_bits_impl.hpp @@ -22,10 +22,10 @@ #include "engine_base.hpp" -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { template -class distribution_base> { +class distribution_base> { protected: template auto generate(EngineType& engine) -> @@ -46,6 +46,6 @@ class distribution_base> { } }; -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #endif // _MKL_RNG_DEVICE_UNIFORM_BITS_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/uniform_impl.hpp b/include/oneapi/math/rng/device/detail/uniform_impl.hpp index ec50eb8fc..8e8c55ad4 100644 --- a/include/oneapi/math/rng/device/detail/uniform_impl.hpp +++ b/include/oneapi/math/rng/device/detail/uniform_impl.hpp @@ -24,7 +24,7 @@ #include #include "engine_base.hpp" -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { static inline std::uint64_t umul_hi_64(const std::uint64_t a, const std::uint64_t b) { const std::uint64_t a_lo = a & 0xFFFFFFFFULL; @@ -63,7 +63,7 @@ static inline void generate_leftover(std::uint64_t range, Generator generate, st } template -class distribution_base> { +class distribution_base> { public: struct param_type { param_type(Type a, Type b) : a_(a), b_(b) {} @@ -74,7 +74,7 @@ class distribution_base> { distribution_base(Type a, Type b) : a_(a), b_(b) { #ifndef __SYCL_DEVICE_ONLY__ if (a >= b) { - throw oneapi::mkl::invalid_argument("rng", "uniform", "a >= b"); + throw oneapi::math::invalid_argument("rng", "uniform", "a >= b"); } #endif } @@ -94,7 +94,7 @@ class distribution_base> { void param(const param_type& pt) { #ifndef __SYCL_DEVICE_ONLY__ if (pt.a_ >= pt.b_) { - throw oneapi::mkl::invalid_argument("rng", "uniform", "a >= b"); + throw oneapi::math::invalid_argument("rng", "uniform", "a >= b"); } #endif a_ = pt.a_; @@ -318,6 +318,6 @@ class distribution_base> { Type b_; }; -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #endif // _MKL_RNG_DEVICE_UNIFORM_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/vm_wrappers.hpp b/include/oneapi/math/rng/device/detail/vm_wrappers.hpp index 850945a4c..acd730b66 100644 --- a/include/oneapi/math/rng/device/detail/vm_wrappers.hpp +++ b/include/oneapi/math/rng/device/detail/vm_wrappers.hpp @@ -22,7 +22,7 @@ #include -namespace oneapi::mkl::rng::device::detail { +namespace oneapi::math::rng::device::detail { template static inline DataType sqrt_wrapper(DataType a) { @@ -70,6 +70,6 @@ template static inline DataType exp_wrapper(DataType a) { return sycl::exp(a); } -} // namespace oneapi::mkl::rng::device::detail +} // namespace oneapi::math::rng::device::detail #endif // _MKL_RNG_DEVICE_VM_WRAPPERS_HPP_ diff --git a/include/oneapi/math/rng/device/distributions.hpp b/include/oneapi/math/rng/device/distributions.hpp index 2f5082886..6db4f70a7 100644 --- a/include/oneapi/math/rng/device/distributions.hpp +++ b/include/oneapi/math/rng/device/distributions.hpp @@ -25,11 +25,11 @@ #include "oneapi/math/rng/device/detail/distribution_base.hpp" #include "oneapi/math/rng/device/functions.hpp" -namespace oneapi::mkl::rng::device { +namespace oneapi::math::rng::device { // CONTINUOUS AND DISCRETE RANDOM NUMBER DISTRIBUTIONS -// Class template oneapi::mkl::rng::device::uniform +// Class template oneapi::math::rng::device::uniform // // Represents continuous and discrete uniform random number distribution // @@ -40,8 +40,8 @@ namespace oneapi::mkl::rng::device { // std::uint32_t // // Supported methods: -// oneapi::mkl::rng::device::uniform_method::standard -// oneapi::mkl::rng::device::uniform_method::accurate +// oneapi::math::rng::device::uniform_method::standard +// oneapi::math::rng::device::uniform_method::accurate // // Input arguments: // a - left bound. 0.0 by default @@ -112,7 +112,7 @@ class uniform : detail::distribution_base> { friend typename Distr::result_type generate_single(Distr& distr, Engine& engine); }; -// Class template oneapi::mkl::rng::device::gaussian +// Class template oneapi::math::rng::device::gaussian // // Represents continuous normal random number distribution // @@ -121,8 +121,8 @@ class uniform : detail::distribution_base> { // double // // Supported methods: -// oneapi::mkl::rng::device::gaussian_method::box_muller2 -// oneapi::mkl::rng::device::gaussian_method::icdf +// oneapi::math::rng::device::gaussian_method::box_muller2 +// oneapi::math::rng::device::gaussian_method::icdf // // Input arguments: // mean - mean. 0 by default @@ -179,7 +179,7 @@ class gaussian : detail::distribution_base> { friend typename Distr::result_type generate_single(Distr& distr, Engine& engine); }; -// Class template oneapi::mkl::rng::device::lognormal +// Class template oneapi::math::rng::device::lognormal // // Represents continuous lognormal random number distribution // @@ -188,7 +188,7 @@ class gaussian : detail::distribution_base> { // double // // Supported methods: -// oneapi::mkl::rng::device::lognormal_method::box_muller2 +// oneapi::math::rng::device::lognormal_method::box_muller2 // // Input arguments: // m - mean of the subject normal distribution. 0.0 by default @@ -252,7 +252,7 @@ class lognormal : detail::distribution_base> { friend typename Distr::result_type generate_single(Distr& distr, Engine& engine); }; -// Class template oneapi::mkl::rng::device::beta +// Class template oneapi::math::rng::device::beta // // Represents continuous beta random number distribution // @@ -261,8 +261,8 @@ class lognormal : detail::distribution_base> { // double // // Supported methods: -// oneapi::mkl::rng::device::beta_method::cja -// oneapi::mkl::rng::device::beta_method::cja_accurate +// oneapi::math::rng::device::beta_method::cja +// oneapi::math::rng::device::beta_method::cja_accurate // // Input arguments: // p - shape. 1.0 by default @@ -330,7 +330,7 @@ class beta : detail::distribution_base> { friend typename Distr::result_type generate_single(Distr& distr, Engine& engine); }; -// Class template oneapi::mkl::rng::device::gamma +// Class template oneapi::math::rng::device::gamma // // Represents continuous gamma random number distribution // @@ -339,8 +339,8 @@ class beta : detail::distribution_base> { // double // // Supported methods: -// oneapi::mkl::rng::device::gamma_method::marsaglia -// oneapi::mkl::rng::device::gamma_method::marsaglia_accurate +// oneapi::math::rng::device::gamma_method::marsaglia +// oneapi::math::rng::device::gamma_method::marsaglia_accurate // // Input arguments: // alpha - shape. 1.0 by default @@ -403,7 +403,7 @@ class gamma : detail::distribution_base> { friend typename Distr::result_type generate_single(Distr& distr, Engine& engine); }; -// Class template oneapi::mkl::rng::device::uniform_bits +// Class template oneapi::math::rng::device::uniform_bits // // Represents discrete uniform bits random number distribution // @@ -429,7 +429,7 @@ class uniform_bits : detail::distribution_base> { friend typename Distr::result_type generate_single(Distr& distr, Engine& engine); }; -// Class template oneapi::mkl::rng::device::bits +// Class template oneapi::math::rng::device::bits // // Represents bits of underlying random number engine // @@ -455,7 +455,7 @@ class bits : detail::distribution_base> { friend typename Distr::result_type generate_single(Distr& distr, Engine& engine); }; -// Class template oneapi::mkl::rng::device::exponential +// Class template oneapi::math::rng::device::exponential // // Represents continuous exponential random number distribution // @@ -464,8 +464,8 @@ class bits : detail::distribution_base> { // double // // Supported methods: -// oneapi::mkl::rng::device::exponential_method::icdf -// oneapi::mkl::rng::device::exponential_method::icdf_accurate +// oneapi::math::rng::device::exponential_method::icdf +// oneapi::math::rng::device::exponential_method::icdf_accurate // // Input arguments: // displ - displacement. 0.0 by default @@ -520,7 +520,7 @@ class exponential : detail::distribution_base> { friend typename Distr::result_type generate_single(Distr& distr, Engine& engine); }; -// Class template oneapi::mkl::rng::device::poisson +// Class template oneapi::math::rng::device::poisson // // Represents discrete poisson random number distribution // @@ -529,7 +529,7 @@ class exponential : detail::distribution_base> { // std::uint32_t // // Supported methods: -// oneapi::mkl::rng::device::poisson_method::devroye +// oneapi::math::rng::device::poisson_method::devroye // // Input arguments: // lambda - mean value. 1.0 by default @@ -574,7 +574,7 @@ class poisson : detail::distribution_base> { friend typename Distr::result_type generate_single(Distr& distr, Engine& engine); }; -// Class template oneapi::mkl::rng::device::bernoulli +// Class template oneapi::math::rng::device::bernoulli // // Represents discrete Bernoulli random number distribution // @@ -583,7 +583,7 @@ class poisson : detail::distribution_base> { // std::int32_t // // Supported methods: -// oneapi::mkl::rng::bernoulli_method::icdf; +// oneapi::math::rng::bernoulli_method::icdf; // // Input arguments: // p - success probablity of a trial. 0.5 by default @@ -632,6 +632,6 @@ class bernoulli : detail::distribution_base> { friend typename Distr::result_type generate_single(Distr& distr, Engine& engine); }; -} // namespace oneapi::mkl::rng::device +} // namespace oneapi::math::rng::device #endif // _MKL_RNG_DEVICE_DISTRIBUTIONS_HPP_ diff --git a/include/oneapi/math/rng/device/engines.hpp b/include/oneapi/math/rng/device/engines.hpp index 3eda19793..2bad9b692 100644 --- a/include/oneapi/math/rng/device/engines.hpp +++ b/include/oneapi/math/rng/device/engines.hpp @@ -26,11 +26,11 @@ #include "oneapi/math/rng/device/functions.hpp" #include "oneapi/math/rng/device/detail/engine_base.hpp" -namespace oneapi::mkl::rng::device { +namespace oneapi::math::rng::device { // PSEUDO-RANDOM NUMBER DEVICE-SIDE ENGINES -// Class template oneapi::mkl::rng::device::philox4x32x10 +// Class template oneapi::math::rng::device::philox4x32x10 // // Represents Philox4x32-10 counter-based pseudorandom number generator // @@ -71,7 +71,7 @@ class philox4x32x10 : detail::engine_base> { friend class detail::distribution_base; }; -// Class oneapi::mkl::rng::device::mrg32k3a +// Class oneapi::math::rng::device::mrg32k3a // // Represents the combined recurcive pseudorandom number generator // @@ -111,7 +111,7 @@ class mrg32k3a : detail::engine_base> { friend class detail::distribution_base; }; -// Class oneapi::mkl::rng::device::mcg31m1 +// Class oneapi::math::rng::device::mcg31m1 // // // @@ -138,7 +138,7 @@ class mcg31m1 : detail::engine_base> { friend class detail::distribution_base; }; -// Class oneapi::mkl::rng::device::mcg59 +// Class oneapi::math::rng::device::mcg59 // // // @@ -165,6 +165,6 @@ class mcg59 : detail::engine_base> { friend class detail::distribution_base; }; -} // namespace oneapi::mkl::rng::device +} // namespace oneapi::math::rng::device #endif // _MKL_RNG_DEVICE_ENGINES_HPP_ diff --git a/include/oneapi/math/rng/device/functions.hpp b/include/oneapi/math/rng/device/functions.hpp index 89f322740..d8ed65337 100644 --- a/include/oneapi/math/rng/device/functions.hpp +++ b/include/oneapi/math/rng/device/functions.hpp @@ -24,7 +24,7 @@ #include "oneapi/math/rng/device/detail/distribution_base.hpp" -namespace oneapi::mkl::rng::device { +namespace oneapi::math::rng::device { // GENERATE FUNCTIONS @@ -47,6 +47,6 @@ void skip_ahead(Engine& engine, std::initializer_list num_to_skip engine.skip_ahead(num_to_skip); } -} // namespace oneapi::mkl::rng::device +} // namespace oneapi::math::rng::device #endif // _MKL_RNG_DEVICE_FUNCTIONS_HPP_ diff --git a/include/oneapi/math/rng/device/types.hpp b/include/oneapi/math/rng/device/types.hpp index 6f87917f8..1fecbc5b5 100644 --- a/include/oneapi/math/rng/device/types.hpp +++ b/include/oneapi/math/rng/device/types.hpp @@ -20,7 +20,7 @@ #ifndef _MKL_RNG_DEVICE_TYPES_HPP_ #define _MKL_RNG_DEVICE_TYPES_HPP_ -namespace oneapi::mkl::rng::device { +namespace oneapi::math::rng::device { // METHODS FOR DISTRIBUTIONS @@ -69,6 +69,6 @@ struct marsaglia_accurate {}; using by_default = marsaglia; } // namespace gamma_method -} // namespace oneapi::mkl::rng::device +} // namespace oneapi::math::rng::device #endif // _MKL_RNG_DEVICE_TYPES_HPP_ diff --git a/include/oneapi/math/rng/distributions.hpp b/include/oneapi/math/rng/distributions.hpp index 1167b62a8..0af2b80e0 100644 --- a/include/oneapi/math/rng/distributions.hpp +++ b/include/oneapi/math/rng/distributions.hpp @@ -31,10 +31,10 @@ #include "oneapi/math/exceptions.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { -// Class template oneapi::mkl::rng::uniform +// Class template oneapi::math::rng::uniform // // Represents continuous and discrete uniform random number distribution // @@ -44,8 +44,8 @@ namespace rng { // std::int32_t // // Supported methods: -// oneapi::mkl::rng::uniform_method::standard -// oneapi::mkl::rng::uniform_method::accurate - for float and double types only +// oneapi::math::rng::uniform_method::standard +// oneapi::math::rng::uniform_method::accurate - for float and double types only // // Input arguments: // a - left bound. 0.0 by default @@ -75,7 +75,7 @@ class uniform { explicit uniform(Type a, Type b) : a_(a), b_(b) { if (a >= b) { - throw oneapi::mkl::invalid_argument("rng", "uniform", + throw oneapi::math::invalid_argument("rng", "uniform", "parameters are incorrect, a >= b"); } } @@ -103,7 +103,7 @@ class uniform { explicit uniform(std::int32_t a, std::int32_t b) : a_(a), b_(b) { if (a >= b) { - throw oneapi::mkl::invalid_argument("rng", "uniform", + throw oneapi::math::invalid_argument("rng", "uniform", "parameters are incorrect, a >= b"); } } @@ -121,7 +121,7 @@ class uniform { std::int32_t b_; }; -// Class template oneapi::mkl::rng::gaussian +// Class template oneapi::math::rng::gaussian // // Represents continuous normal random number distribution // @@ -130,8 +130,8 @@ class uniform { // double // // Supported methods: -// oneapi::mkl::rng::gaussian_method::box_muller2 -// oneapi::mkl::rng::gaussian_method::icdf +// oneapi::math::rng::gaussian_method::box_muller2 +// oneapi::math::rng::gaussian_method::icdf // // Input arguments: // mean - mean. 0 by default @@ -160,7 +160,7 @@ class gaussian { explicit gaussian(RealType mean, RealType stddev) : mean_(mean), stddev_(stddev) { if (stddev <= static_cast(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "gaussian", + throw oneapi::math::invalid_argument("rng", "gaussian", "stddev parameter is incorrect, stddev <= 0.0"); } } @@ -178,7 +178,7 @@ class gaussian { RealType stddev_; }; -// Class template oneapi::mkl::rng::lognormal +// Class template oneapi::math::rng::lognormal // // Represents continuous lognormal random number distribution // @@ -187,8 +187,8 @@ class gaussian { // double // // Supported methods: -// oneapi::mkl::rng::lognormal_method::box_muller2 -// oneapi::mkl::rng::lognormal_method::icdf +// oneapi::math::rng::lognormal_method::box_muller2 +// oneapi::math::rng::lognormal_method::icdf // // Input arguments: // m - mean of the subject normal distribution. 0.0 by default @@ -226,10 +226,10 @@ class lognormal { displ_(displ), scale_(scale) { if (s <= static_cast(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "lognormal", "s <= 0"); + throw oneapi::math::invalid_argument("rng", "lognormal", "s <= 0"); } if (scale <= static_cast(0.0)) { - throw oneapi::mkl::invalid_argument("rng", "lognormal", "scale <= 0"); + throw oneapi::math::invalid_argument("rng", "lognormal", "scale <= 0"); } } @@ -256,7 +256,7 @@ class lognormal { RealType scale_; }; -// Class template oneapi::mkl::rng::bernoulli +// Class template oneapi::math::rng::bernoulli // // Represents discrete Bernoulli random number distribution // @@ -265,7 +265,7 @@ class lognormal { // std::int32_t // // Supported methods: -// oneapi::mkl::rng::bernoulli_method::icdf; +// oneapi::math::rng::bernoulli_method::icdf; // // Input arguments: // p - success probability of a trial. 0.5 by default @@ -292,7 +292,7 @@ class bernoulli { explicit bernoulli(float p) : p_(p) { if ((p > 1.0f) || (p < 0.0f)) { - throw oneapi::mkl::invalid_argument("rng", "bernoulli", "p > 1 or p < 0"); + throw oneapi::math::invalid_argument("rng", "bernoulli", "p > 1 or p < 0"); } } @@ -304,7 +304,7 @@ class bernoulli { float p_; }; -// Class template oneapi::mkl::rng::poisson +// Class template oneapi::math::rng::poisson // // Represents discrete Poisson random number distribution // @@ -312,7 +312,7 @@ class bernoulli { // std::int32_t // // Supported methods: -// oneapi::mkl::rng::poisson_method::gaussian_icdf_based +// oneapi::math::rng::poisson_method::gaussian_icdf_based // // Input arguments: // lambda - distribution parameter. 0.5 by default @@ -339,7 +339,7 @@ class poisson { explicit poisson(double lambda) : lambda_(lambda) { if ((lambda <= 0.0)) { - throw oneapi::mkl::invalid_argument("rng", "poisson", "lamdba < 0"); + throw oneapi::math::invalid_argument("rng", "poisson", "lamdba < 0"); } } @@ -351,7 +351,7 @@ class poisson { double lambda_; }; -// Class template oneapi::mkl::rng::bits +// Class template oneapi::math::rng::bits // // Represents bits of underlying random number engine // @@ -367,7 +367,7 @@ class bits { }; } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_RNG_DISTRIBUTIONS_HPP_ diff --git a/include/oneapi/math/rng/engines.hpp b/include/oneapi/math/rng/engines.hpp index e72948ae1..f09243459 100644 --- a/include/oneapi/math/rng/engines.hpp +++ b/include/oneapi/math/rng/engines.hpp @@ -49,10 +49,10 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace rng { -// Class oneapi::mkl::rng::philox4x32x10 +// Class oneapi::math::rng::philox4x32x10 // // Represents Philox4x32-10 counter-based pseudorandom number generator // @@ -144,7 +144,7 @@ class philox4x32x10 { const std::vector& dependencies); }; -// Class oneapi::mkl::rng::mrg32k3a +// Class oneapi::math::rng::mrg32k3a // // Represents the combined recurcive pseudorandom number generator // @@ -237,7 +237,7 @@ class mrg32k3a { using default_engine = philox4x32x10; } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_RNG_ENGINES_HPP_ diff --git a/include/oneapi/math/rng/functions.hpp b/include/oneapi/math/rng/functions.hpp index e84762b4e..72a370551 100644 --- a/include/oneapi/math/rng/functions.hpp +++ b/include/oneapi/math/rng/functions.hpp @@ -31,10 +31,10 @@ #include "oneapi/math/rng/predicates.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { -// Function oneapi::mkl::rng::generate().Buffer API +// Function oneapi::math::rng::generate().Buffer API // Provides random numbers from a given engine with a given statistics // // Input parameters: @@ -51,7 +51,7 @@ static inline void generate(const Distr& distr, Engine& engine, std::int64_t n, engine.pimpl_->generate(distr, n, r); } -// Function oneapi::mkl::rng::generate(). USM API +// Function oneapi::math::rng::generate(). USM API // Provides random numbers from a given engine with a given statistics // // Input parameters: @@ -76,7 +76,7 @@ static inline sycl::event generate(const Distr& distr, Engine& engine, std::int6 // SERVICE FUNCTIONS -// Function oneapi::mkl::rng::skip_ahead(). Common interface +// Function oneapi::math::rng::skip_ahead(). Common interface // // Proceeds state of engine using the skip-ahead method // @@ -88,7 +88,7 @@ static inline void skip_ahead(Engine& engine, std::uint64_t num_to_skip) { engine.pimpl_->skip_ahead(num_to_skip); } -// Function oneapi::mkl::rng::skip_ahead(). Interface with partitioned number of skipped elements +// Function oneapi::math::rng::skip_ahead(). Interface with partitioned number of skipped elements // // Proceeds state of engine using the skip-ahead method // @@ -100,7 +100,7 @@ static inline void skip_ahead(Engine& engine, std::initializer_listskip_ahead(num_to_skip); } -// Function oneapi::mkl::rng::leapfrog() +// Function oneapi::math::rng::leapfrog() // // Proceeds state of engine using the leapfrog method // @@ -114,7 +114,7 @@ static inline void leapfrog(Engine& engine, std::uint64_t idx, std::uint64_t str } } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_RNG_FUNCTIONS_HPP_ diff --git a/include/oneapi/math/rng/predicates.hpp b/include/oneapi/math/rng/predicates.hpp index 798825bbd..07c4e4591 100644 --- a/include/oneapi/math/rng/predicates.hpp +++ b/include/oneapi/math/rng/predicates.hpp @@ -31,7 +31,7 @@ #include "oneapi/math/types.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { // Buffer APIs @@ -41,7 +41,7 @@ inline void generate_precondition(const Distr& /*distr*/, Engine& /*engine*/, st sycl::buffer& r) { #ifndef ONEMATH_DISABLE_PREDICATES if (n < 0 || n > r.size()) { - throw oneapi::mkl::invalid_argument("rng", "generate", "n"); + throw oneapi::math::invalid_argument("rng", "generate", "n"); } #endif } @@ -54,16 +54,16 @@ inline void generate_precondition(const Distr& /*distr*/, Engine& /*engine*/, st const std::vector& /*dependencies*/) { #ifndef ONEMATH_DISABLE_PREDICATES if (n < 0) { - throw oneapi::mkl::invalid_argument("rng", "generate", "n"); + throw oneapi::math::invalid_argument("rng", "generate", "n"); } if (r == nullptr) { - throw oneapi::mkl::invalid_argument("rng", "generate", "r is nullptr"); + throw oneapi::math::invalid_argument("rng", "generate", "r is nullptr"); } #endif } } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_RNG_PREDICATES_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/handles.hpp b/include/oneapi/math/sparse_blas/detail/handles.hpp index 792cfe3a1..15153339c 100644 --- a/include/oneapi/math/sparse_blas/detail/handles.hpp +++ b/include/oneapi/math/sparse_blas/detail/handles.hpp @@ -20,7 +20,7 @@ #ifndef _ONEMATH_SPARSE_BLAS_DETAIL_HANDLES_HPP_ #define _ONEMATH_SPARSE_BLAS_DETAIL_HANDLES_HPP_ -namespace oneapi::mkl::sparse { +namespace oneapi::math::sparse { // Each backend can create its own handle type or re-use the native handle types that will be reinterpret_cast'ed to the types below @@ -33,6 +33,6 @@ using dense_vector_handle_t = dense_vector_handle*; struct matrix_handle; using matrix_handle_t = matrix_handle*; -} // namespace oneapi::mkl::sparse +} // namespace oneapi::math::sparse #endif // _ONEMATH_SPARSE_BLAS_DETAIL_HANDLES_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/helper_types.hpp b/include/oneapi/math/sparse_blas/detail/helper_types.hpp index 95b5a6d48..b3082b736 100644 --- a/include/oneapi/math/sparse_blas/detail/helper_types.hpp +++ b/include/oneapi/math/sparse_blas/detail/helper_types.hpp @@ -25,7 +25,7 @@ #include namespace oneapi { -namespace mkl { +namespace math { namespace sparse { namespace detail { @@ -44,7 +44,7 @@ inline constexpr bool are_fp_int_supported_v = } // namespace detail } // namespace sparse -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ONEMATH_SPARSE_BLAS_DETAIL_HELPER_TYPES_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp b/include/oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp index 710c93d2a..59544a503 100644 --- a/include/oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklcpu/onemath_sparse_blas_mklcpu.hpp @@ -24,12 +24,12 @@ #include "oneapi/math/sparse_blas/detail/helper_types.hpp" #include "oneapi/math/sparse_blas/types.hpp" -namespace oneapi::mkl::sparse::mklcpu { +namespace oneapi::math::sparse::mklcpu { -namespace detail = oneapi::mkl::sparse::detail; +namespace detail = oneapi::math::sparse::detail; #include "oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx" -} // namespace oneapi::mkl::sparse::mklcpu +} // namespace oneapi::math::sparse::mklcpu #endif // _ONEMATH_SPARSE_BLAS_DETAIL_MKLCPU_ONEMATH_SPARSE_BLAS_MKLCPU_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp b/include/oneapi/math/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp index d4678ce1e..fc8077eff 100644 --- a/include/oneapi/math/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklcpu/sparse_blas_ct.hpp @@ -26,7 +26,7 @@ #include "onemath_sparse_blas_mklcpu.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace sparse { #define BACKEND mklcpu @@ -34,7 +34,7 @@ namespace sparse { #undef BACKEND } //namespace sparse -} //namespace mkl +} //namespace math } //namespace oneapi #endif // _ONEMATH_SPARSE_BLAS_DETAIL_MKLCPU_SPARSE_BLAS_CT_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp b/include/oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp index 03c3592b2..651ee7731 100644 --- a/include/oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklgpu/onemath_sparse_blas_mklgpu.hpp @@ -24,12 +24,12 @@ #include "oneapi/math/sparse_blas/detail/helper_types.hpp" #include "oneapi/math/sparse_blas/types.hpp" -namespace oneapi::mkl::sparse::mklgpu { +namespace oneapi::math::sparse::mklgpu { -namespace detail = oneapi::mkl::sparse::detail; +namespace detail = oneapi::math::sparse::detail; #include "oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx" -} // namespace oneapi::mkl::sparse::mklgpu +} // namespace oneapi::math::sparse::mklgpu #endif // _ONEMATH_SPARSE_BLAS_DETAIL_MKLGPU_ONEMATH_SPARSE_BLAS_MKLGPU_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp b/include/oneapi/math/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp index 19b4805c9..7e49e6305 100644 --- a/include/oneapi/math/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp +++ b/include/oneapi/math/sparse_blas/detail/mklgpu/sparse_blas_ct.hpp @@ -26,7 +26,7 @@ #include "onemath_sparse_blas_mklgpu.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace sparse { #define BACKEND mklgpu @@ -34,7 +34,7 @@ namespace sparse { #undef BACKEND } //namespace sparse -} //namespace mkl +} //namespace math } //namespace oneapi #endif // _ONEMATH_SPARSE_BLAS_DETAIL_MKLGPU_SPARSE_BLAS_CT_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx b/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx index 65c30750c..310a9a4e8 100644 --- a/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx +++ b/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx @@ -125,29 +125,29 @@ ONEMATH_EXPORT void init_spmm_descr(sycl::queue &queue, spmm_descr_t *p_spmm_des ONEMATH_EXPORT sycl::event release_spmm_descr(sycl::queue &queue, spmm_descr_t spmm_descr, const std::vector &dependencies = {}); -ONEMATH_EXPORT void spmm_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, - oneapi::mkl::transpose opB, const void *alpha, +ONEMATH_EXPORT void spmm_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, std::size_t &temp_buffer_size); -ONEMATH_EXPORT void spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, - oneapi::mkl::transpose opB, const void *alpha, matrix_view A_view, +ONEMATH_EXPORT void spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, sycl::buffer workspace); -ONEMATH_EXPORT sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, - oneapi::mkl::transpose opB, const void *alpha, +ONEMATH_EXPORT sycl::event spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, void *workspace, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spmm(sycl::queue &queue, oneapi::mkl::transpose opA, - oneapi::mkl::transpose opB, const void *alpha, matrix_view A_view, +ONEMATH_EXPORT sycl::event spmm(sycl::queue &queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, @@ -159,26 +159,26 @@ ONEMATH_EXPORT void init_spmv_descr(sycl::queue &queue, spmv_descr_t *p_spmv_des ONEMATH_EXPORT sycl::event release_spmv_descr(sycl::queue &queue, spmv_descr_t spmv_descr, const std::vector &dependencies = {}); -ONEMATH_EXPORT void spmv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT void spmv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, std::size_t &temp_buffer_size); -ONEMATH_EXPORT void spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +ONEMATH_EXPORT void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, sycl::buffer workspace); -ONEMATH_EXPORT sycl::event spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, void *workspace, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, @@ -191,26 +191,26 @@ ONEMATH_EXPORT void init_spsv_descr(sycl::queue &queue, spsv_descr_t *p_spsv_des ONEMATH_EXPORT sycl::event release_spsv_descr(sycl::queue &queue, spsv_descr_t spsv_descr, const std::vector &dependencies = {}); -ONEMATH_EXPORT void spsv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT void spsv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, std::size_t &temp_buffer_size); -ONEMATH_EXPORT void spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +ONEMATH_EXPORT void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, sycl::buffer workspace); -ONEMATH_EXPORT sycl::event spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, +ONEMATH_EXPORT sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, void *workspace, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spsv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +ONEMATH_EXPORT sycl::event spsv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, diff --git a/include/oneapi/math/sparse_blas/detail/operation_types.hpp b/include/oneapi/math/sparse_blas/detail/operation_types.hpp index 51611070a..495cfdc70 100644 --- a/include/oneapi/math/sparse_blas/detail/operation_types.hpp +++ b/include/oneapi/math/sparse_blas/detail/operation_types.hpp @@ -20,7 +20,7 @@ #ifndef _ONEMATH_SPARSE_BLAS_DETAIL_OPERATION_TYPES_HPP_ #define _ONEMATH_SPARSE_BLAS_DETAIL_OPERATION_TYPES_HPP_ -namespace oneapi::mkl::sparse { +namespace oneapi::math::sparse { // Each backend can create its own descriptor type or re-use the native descriptor types that will be reinterpret_cast'ed to the types below @@ -33,6 +33,6 @@ using spmv_descr_t = spmv_descr*; struct spsv_descr; using spsv_descr_t = spsv_descr*; -} // namespace oneapi::mkl::sparse +} // namespace oneapi::math::sparse #endif // _ONEMATH_SPARSE_BLAS_DETAIL_OPERATION_TYPES_HPP_ diff --git a/include/oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx b/include/oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx index 8200964d2..5f28b1ecb 100644 --- a/include/oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx +++ b/include/oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx @@ -193,7 +193,7 @@ inline sycl::event release_spmm_descr(backend_selector selecto } inline void spmm_buffer_size(backend_selector selector, - oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, + oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, @@ -202,8 +202,8 @@ inline void spmm_buffer_size(backend_selector selector, beta, C_handle, alg, spmm_descr, temp_buffer_size); } -inline void spmm_optimize(backend_selector selector, oneapi::mkl::transpose opA, - oneapi::mkl::transpose opB, const void *alpha, matrix_view A_view, +inline void spmm_optimize(backend_selector selector, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, sycl::buffer workspace) { @@ -212,7 +212,7 @@ inline void spmm_optimize(backend_selector selector, oneapi::m } inline sycl::event spmm_optimize(backend_selector selector, - oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, + oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, @@ -222,8 +222,8 @@ inline sycl::event spmm_optimize(backend_selector selector, beta, C_handle, alg, spmm_descr, workspace, dependencies); } -inline sycl::event spmm(backend_selector selector, oneapi::mkl::transpose opA, - oneapi::mkl::transpose opB, const void *alpha, matrix_view A_view, +inline sycl::event spmm(backend_selector selector, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, const std::vector &dependencies = {}) { @@ -244,7 +244,7 @@ inline sycl::event release_spmv_descr(backend_selector selecto } inline void spmv_buffer_size(backend_selector selector, - oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, + oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, std::size_t &temp_buffer_size) { @@ -252,7 +252,7 @@ inline void spmv_buffer_size(backend_selector selector, y_handle, alg, spmv_descr, temp_buffer_size); } -inline void spmv_optimize(backend_selector selector, oneapi::mkl::transpose opA, +inline void spmv_optimize(backend_selector selector, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, @@ -262,7 +262,7 @@ inline void spmv_optimize(backend_selector selector, oneapi::m } inline sycl::event spmv_optimize(backend_selector selector, - oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, + oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, void *workspace, @@ -271,7 +271,7 @@ inline sycl::event spmv_optimize(backend_selector selector, beta, y_handle, alg, spmv_descr, workspace, dependencies); } -inline sycl::event spmv(backend_selector selector, oneapi::mkl::transpose opA, +inline sycl::event spmv(backend_selector selector, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, @@ -293,7 +293,7 @@ inline sycl::event release_spsv_descr(backend_selector selecto } inline void spsv_buffer_size(backend_selector selector, - oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, + oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, std::size_t &temp_buffer_size) { @@ -301,7 +301,7 @@ inline void spsv_buffer_size(backend_selector selector, y_handle, alg, spsv_descr, temp_buffer_size); } -inline void spsv_optimize(backend_selector selector, oneapi::mkl::transpose opA, +inline void spsv_optimize(backend_selector selector, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, @@ -311,7 +311,7 @@ inline void spsv_optimize(backend_selector selector, oneapi::m } inline sycl::event spsv_optimize(backend_selector selector, - oneapi::mkl::transpose opA, const void *alpha, matrix_view A_view, + oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, void *workspace, @@ -320,7 +320,7 @@ inline sycl::event spsv_optimize(backend_selector selector, y_handle, alg, spsv_descr, workspace, dependencies); } -inline sycl::event spsv(backend_selector selector, oneapi::mkl::transpose opA, +inline sycl::event spsv(backend_selector selector, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, diff --git a/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp b/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp index d220979ce..df3341814 100644 --- a/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp +++ b/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp @@ -24,7 +24,7 @@ #include "oneapi/math/sparse_blas/types.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace sparse { // Dense vector @@ -123,25 +123,25 @@ void init_spmm_descr(sycl::queue &queue, spmm_descr_t *p_spmm_descr); sycl::event release_spmm_descr(sycl::queue &queue, spmm_descr_t spmm_descr, const std::vector &dependencies = {}); -void spmm_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, +void spmm_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, std::size_t &temp_buffer_size); -void spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, +void spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, sycl::buffer workspace); -sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, - oneapi::mkl::transpose opB, const void *alpha, matrix_view A_view, +sycl::event spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, void *workspace, const std::vector &dependencies = {}); -sycl::event spmm(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, +sycl::event spmm(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, @@ -153,23 +153,23 @@ void init_spmv_descr(sycl::queue &queue, spmv_descr_t *p_spmv_descr); sycl::event release_spmv_descr(sycl::queue &queue, spmv_descr_t spmv_descr, const std::vector &dependencies = {}); -void spmv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +void spmv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, std::size_t &temp_buffer_size); -void spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, sycl::buffer workspace); -sycl::event spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, void *workspace, const std::vector &dependencies = {}); -sycl::event spmv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +sycl::event spmv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, const std::vector &dependencies = {}); @@ -180,29 +180,29 @@ void init_spsv_descr(sycl::queue &queue, spsv_descr_t *p_spsv_descr); sycl::event release_spsv_descr(sycl::queue &queue, spsv_descr_t spsv_descr, const std::vector &dependencies = {}); -void spsv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +void spsv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, std::size_t &temp_buffer_size); -void spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, sycl::buffer workspace); -sycl::event spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, void *workspace, const std::vector &dependencies = {}); -sycl::event spsv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +sycl::event spsv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, const std::vector &dependencies = {}); } // namespace sparse -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ONEMATH_SPARSE_BLAS_DETAIL_SPARSE_BLAS_RT_HPP_ diff --git a/include/oneapi/math/sparse_blas/matrix_view.hpp b/include/oneapi/math/sparse_blas/matrix_view.hpp index 8625b2ab9..8f817b83f 100644 --- a/include/oneapi/math/sparse_blas/matrix_view.hpp +++ b/include/oneapi/math/sparse_blas/matrix_view.hpp @@ -23,7 +23,7 @@ #include "oneapi/math/types.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace sparse { enum class matrix_descr { @@ -45,7 +45,7 @@ struct matrix_view { }; } // namespace sparse -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ONEMATH_SPARSE_BLAS_MATRIX_VIEW_HPP_ diff --git a/include/oneapi/math/sparse_blas/types.hpp b/include/oneapi/math/sparse_blas/types.hpp index 2812a0702..5c79244af 100644 --- a/include/oneapi/math/sparse_blas/types.hpp +++ b/include/oneapi/math/sparse_blas/types.hpp @@ -30,7 +30,7 @@ */ namespace oneapi { -namespace mkl { +namespace math { namespace sparse { enum class matrix_property { @@ -66,7 +66,7 @@ enum class spsv_alg { }; } // namespace sparse -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ONEMATH_SPARSE_BLAS_TYPES_HPP_ diff --git a/include/oneapi/math/types.hpp b/include/oneapi/math/types.hpp index bb58ff647..b95a1be0b 100644 --- a/include/oneapi/math/types.hpp +++ b/include/oneapi/math/types.hpp @@ -31,7 +31,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { #ifndef __HIPSYCL__ using bfloat16 = sycl::ext::oneapi::bfloat16; @@ -116,7 +116,7 @@ enum class order : char { E = 1, }; -} //namespace mkl +} //namespace math } //namespace oneapi #endif //_ONEMATH_TYPES_HPP_ diff --git a/scripts/generate_backend_api.py b/scripts/generate_backend_api.py index ab392230c..2f597f001 100755 --- a/scripts/generate_backend_api.py +++ b/scripts/generate_backend_api.py @@ -41,9 +41,9 @@ def usage(err = None): Example: The command below will generate: "onemath_blas_mklgpu.hpp" header with declaration of all backend library APIs. -API from backend library will be called from "oneapi::mkl::mklgpu::blas" namespace. +API from backend library will be called from "oneapi::math::mklgpu::blas" namespace. -{script} include/oneapi/math/blas.hpp include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp oneapi::mkl::mklgpu::blas +{script} include/oneapi/math/blas.hpp include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp oneapi::math::mklgpu::blas '''.format(script = argv[0])) if len(argv) < 3: diff --git a/scripts/generate_ct_instant.py b/scripts/generate_ct_instant.py index 875d66533..ea5b64ea2 100755 --- a/scripts/generate_ct_instant.py +++ b/scripts/generate_ct_instant.py @@ -41,9 +41,9 @@ def usage(err = None): Example: The command below will generate: "blas_ct.hpp" header with compile-time BLAS API based on "blas_ct_templates.hpp" for "mklgpu" backend. -API from the backend library will be called from "oneapi::mkl::mklgpu::blas" namespace. +API from the backend library will be called from "oneapi::math::mklgpu::blas" namespace. -{script} include/oneapi/math/blas/detail/blas_ct_templates.hpp include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp mklgpu oneapi::mkl::mklgpu::blas +{script} include/oneapi/math/blas/detail/blas_ct_templates.hpp include/oneapi/math/blas/detail/mklgpu/blas_ct.hpp include/oneapi/math/blas/detail/mklgpu/onemath_blas_mklgpu.hpp mklgpu oneapi::math::mklgpu::blas '''.format(script = argv[0])) if len(argv) < 6: diff --git a/scripts/generate_ct_templates.py b/scripts/generate_ct_templates.py index 9d7768727..796737d6a 100755 --- a/scripts/generate_ct_templates.py +++ b/scripts/generate_ct_templates.py @@ -65,7 +65,7 @@ def print_funcs(func_list): code="" for data in func_list: code +=""" -template static inline {ret_type} {name}{par_str}; +template static inline {ret_type} {name}{par_str}; """.format(**data) return code diff --git a/src/blas/backends/backend_wrappers.cxx b/src/blas/backends/backend_wrappers.cxx index 62f6ced13..b0c8c9c70 100644 --- a/src/blas/backends/backend_wrappers.cxx +++ b/src/blas/backends/backend_wrappers.cxx @@ -18,494 +18,494 @@ *******************************************************************************/ // clang-format off -oneapi::mkl::blas::BACKEND::MAJOR::asum, -oneapi::mkl::blas::BACKEND::MAJOR::asum, -oneapi::mkl::blas::BACKEND::MAJOR::asum, -oneapi::mkl::blas::BACKEND::MAJOR::asum, -oneapi::mkl::blas::BACKEND::MAJOR::axpy, -oneapi::mkl::blas::BACKEND::MAJOR::axpy, -oneapi::mkl::blas::BACKEND::MAJOR::axpy, -oneapi::mkl::blas::BACKEND::MAJOR::axpy, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpby, -oneapi::mkl::blas::BACKEND::MAJOR::axpby, -oneapi::mkl::blas::BACKEND::MAJOR::axpby, -oneapi::mkl::blas::BACKEND::MAJOR::axpby, -oneapi::mkl::blas::BACKEND::MAJOR::copy, -oneapi::mkl::blas::BACKEND::MAJOR::copy, -oneapi::mkl::blas::BACKEND::MAJOR::copy, -oneapi::mkl::blas::BACKEND::MAJOR::copy, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dot, -oneapi::mkl::blas::BACKEND::MAJOR::dot, -oneapi::mkl::blas::BACKEND::MAJOR::dot, -oneapi::mkl::blas::BACKEND::MAJOR::dotc, -oneapi::mkl::blas::BACKEND::MAJOR::dotc, -oneapi::mkl::blas::BACKEND::MAJOR::dotu, -oneapi::mkl::blas::BACKEND::MAJOR::dotu, -oneapi::mkl::blas::BACKEND::MAJOR::iamin, -oneapi::mkl::blas::BACKEND::MAJOR::iamin, -oneapi::mkl::blas::BACKEND::MAJOR::iamin, -oneapi::mkl::blas::BACKEND::MAJOR::iamin, -oneapi::mkl::blas::BACKEND::MAJOR::iamax, -oneapi::mkl::blas::BACKEND::MAJOR::iamax, -oneapi::mkl::blas::BACKEND::MAJOR::iamax, -oneapi::mkl::blas::BACKEND::MAJOR::iamax, -oneapi::mkl::blas::BACKEND::MAJOR::nrm2, -oneapi::mkl::blas::BACKEND::MAJOR::nrm2, -oneapi::mkl::blas::BACKEND::MAJOR::nrm2, -oneapi::mkl::blas::BACKEND::MAJOR::nrm2, -oneapi::mkl::blas::BACKEND::MAJOR::rot, -oneapi::mkl::blas::BACKEND::MAJOR::rot, -oneapi::mkl::blas::BACKEND::MAJOR::rot, -oneapi::mkl::blas::BACKEND::MAJOR::rot, -oneapi::mkl::blas::BACKEND::MAJOR::rotg, -oneapi::mkl::blas::BACKEND::MAJOR::rotg, -oneapi::mkl::blas::BACKEND::MAJOR::rotg, -oneapi::mkl::blas::BACKEND::MAJOR::rotg, -oneapi::mkl::blas::BACKEND::MAJOR::rotm, -oneapi::mkl::blas::BACKEND::MAJOR::rotm, -oneapi::mkl::blas::BACKEND::MAJOR::rotmg, -oneapi::mkl::blas::BACKEND::MAJOR::rotmg, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::sdsdot, -oneapi::mkl::blas::BACKEND::MAJOR::swap, -oneapi::mkl::blas::BACKEND::MAJOR::swap, -oneapi::mkl::blas::BACKEND::MAJOR::swap, -oneapi::mkl::blas::BACKEND::MAJOR::swap, -oneapi::mkl::blas::BACKEND::MAJOR::gbmv, -oneapi::mkl::blas::BACKEND::MAJOR::gbmv, -oneapi::mkl::blas::BACKEND::MAJOR::gbmv, -oneapi::mkl::blas::BACKEND::MAJOR::gbmv, -oneapi::mkl::blas::BACKEND::MAJOR::gemv, -oneapi::mkl::blas::BACKEND::MAJOR::gemv, -oneapi::mkl::blas::BACKEND::MAJOR::gemv, -oneapi::mkl::blas::BACKEND::MAJOR::gemv, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::ger, -oneapi::mkl::blas::BACKEND::MAJOR::ger, -oneapi::mkl::blas::BACKEND::MAJOR::gerc, -oneapi::mkl::blas::BACKEND::MAJOR::gerc, -oneapi::mkl::blas::BACKEND::MAJOR::geru, -oneapi::mkl::blas::BACKEND::MAJOR::geru, -oneapi::mkl::blas::BACKEND::MAJOR::hbmv, -oneapi::mkl::blas::BACKEND::MAJOR::hbmv, -oneapi::mkl::blas::BACKEND::MAJOR::hemv, -oneapi::mkl::blas::BACKEND::MAJOR::hemv, -oneapi::mkl::blas::BACKEND::MAJOR::her, -oneapi::mkl::blas::BACKEND::MAJOR::her, -oneapi::mkl::blas::BACKEND::MAJOR::her2, -oneapi::mkl::blas::BACKEND::MAJOR::her2, -oneapi::mkl::blas::BACKEND::MAJOR::hpmv, -oneapi::mkl::blas::BACKEND::MAJOR::hpmv, -oneapi::mkl::blas::BACKEND::MAJOR::hpr, -oneapi::mkl::blas::BACKEND::MAJOR::hpr, -oneapi::mkl::blas::BACKEND::MAJOR::hpr2, -oneapi::mkl::blas::BACKEND::MAJOR::hpr2, -oneapi::mkl::blas::BACKEND::MAJOR::sbmv, -oneapi::mkl::blas::BACKEND::MAJOR::sbmv, -oneapi::mkl::blas::BACKEND::MAJOR::spmv, -oneapi::mkl::blas::BACKEND::MAJOR::spmv, -oneapi::mkl::blas::BACKEND::MAJOR::spr, -oneapi::mkl::blas::BACKEND::MAJOR::spr, -oneapi::mkl::blas::BACKEND::MAJOR::spr2, -oneapi::mkl::blas::BACKEND::MAJOR::spr2, -oneapi::mkl::blas::BACKEND::MAJOR::symv, -oneapi::mkl::blas::BACKEND::MAJOR::symv, -oneapi::mkl::blas::BACKEND::MAJOR::syr, -oneapi::mkl::blas::BACKEND::MAJOR::syr, -oneapi::mkl::blas::BACKEND::MAJOR::syr2, -oneapi::mkl::blas::BACKEND::MAJOR::syr2, -oneapi::mkl::blas::BACKEND::MAJOR::tbmv, -oneapi::mkl::blas::BACKEND::MAJOR::tbmv, -oneapi::mkl::blas::BACKEND::MAJOR::tbmv, -oneapi::mkl::blas::BACKEND::MAJOR::tbmv, -oneapi::mkl::blas::BACKEND::MAJOR::tbsv, -oneapi::mkl::blas::BACKEND::MAJOR::tbsv, -oneapi::mkl::blas::BACKEND::MAJOR::tbsv, -oneapi::mkl::blas::BACKEND::MAJOR::tbsv, -oneapi::mkl::blas::BACKEND::MAJOR::tpmv, -oneapi::mkl::blas::BACKEND::MAJOR::tpmv, -oneapi::mkl::blas::BACKEND::MAJOR::tpmv, -oneapi::mkl::blas::BACKEND::MAJOR::tpmv, -oneapi::mkl::blas::BACKEND::MAJOR::tpsv, -oneapi::mkl::blas::BACKEND::MAJOR::tpsv, -oneapi::mkl::blas::BACKEND::MAJOR::tpsv, -oneapi::mkl::blas::BACKEND::MAJOR::tpsv, -oneapi::mkl::blas::BACKEND::MAJOR::trmv, -oneapi::mkl::blas::BACKEND::MAJOR::trmv, -oneapi::mkl::blas::BACKEND::MAJOR::trmv, -oneapi::mkl::blas::BACKEND::MAJOR::trmv, -oneapi::mkl::blas::BACKEND::MAJOR::trsv, -oneapi::mkl::blas::BACKEND::MAJOR::trsv, -oneapi::mkl::blas::BACKEND::MAJOR::trsv, -oneapi::mkl::blas::BACKEND::MAJOR::trsv, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::hemm, -oneapi::mkl::blas::BACKEND::MAJOR::hemm, -oneapi::mkl::blas::BACKEND::MAJOR::herk, -oneapi::mkl::blas::BACKEND::MAJOR::herk, -oneapi::mkl::blas::BACKEND::MAJOR::her2k, -oneapi::mkl::blas::BACKEND::MAJOR::her2k, -oneapi::mkl::blas::BACKEND::MAJOR::symm, -oneapi::mkl::blas::BACKEND::MAJOR::symm, -oneapi::mkl::blas::BACKEND::MAJOR::symm, -oneapi::mkl::blas::BACKEND::MAJOR::symm, -oneapi::mkl::blas::BACKEND::MAJOR::syrk, -oneapi::mkl::blas::BACKEND::MAJOR::syrk, -oneapi::mkl::blas::BACKEND::MAJOR::syrk, -oneapi::mkl::blas::BACKEND::MAJOR::syrk, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syr2k, -oneapi::mkl::blas::BACKEND::MAJOR::syr2k, -oneapi::mkl::blas::BACKEND::MAJOR::syr2k, -oneapi::mkl::blas::BACKEND::MAJOR::syr2k, -oneapi::mkl::blas::BACKEND::MAJOR::trmm, -oneapi::mkl::blas::BACKEND::MAJOR::trmm, -oneapi::mkl::blas::BACKEND::MAJOR::trmm, -oneapi::mkl::blas::BACKEND::MAJOR::trmm, -oneapi::mkl::blas::BACKEND::MAJOR::trsm, -oneapi::mkl::blas::BACKEND::MAJOR::trsm, -oneapi::mkl::blas::BACKEND::MAJOR::trsm, -oneapi::mkl::blas::BACKEND::MAJOR::trsm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemmt, -oneapi::mkl::blas::BACKEND::MAJOR::gemmt, -oneapi::mkl::blas::BACKEND::MAJOR::gemmt, -oneapi::mkl::blas::BACKEND::MAJOR::gemmt, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_bias, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_bias, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_bias, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_bias, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy2, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy2, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy2, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy2, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd, -oneapi::mkl::blas::BACKEND::MAJOR::asum, -oneapi::mkl::blas::BACKEND::MAJOR::asum, -oneapi::mkl::blas::BACKEND::MAJOR::asum, -oneapi::mkl::blas::BACKEND::MAJOR::asum, -oneapi::mkl::blas::BACKEND::MAJOR::axpy, -oneapi::mkl::blas::BACKEND::MAJOR::axpy, -oneapi::mkl::blas::BACKEND::MAJOR::axpy, -oneapi::mkl::blas::BACKEND::MAJOR::axpy, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::axpby, -oneapi::mkl::blas::BACKEND::MAJOR::axpby, -oneapi::mkl::blas::BACKEND::MAJOR::axpby, -oneapi::mkl::blas::BACKEND::MAJOR::axpby, -oneapi::mkl::blas::BACKEND::MAJOR::copy, -oneapi::mkl::blas::BACKEND::MAJOR::copy, -oneapi::mkl::blas::BACKEND::MAJOR::copy, -oneapi::mkl::blas::BACKEND::MAJOR::copy, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::copy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dot, -oneapi::mkl::blas::BACKEND::MAJOR::dot, -oneapi::mkl::blas::BACKEND::MAJOR::dot, -oneapi::mkl::blas::BACKEND::MAJOR::dotc, -oneapi::mkl::blas::BACKEND::MAJOR::dotc, -oneapi::mkl::blas::BACKEND::MAJOR::dotu, -oneapi::mkl::blas::BACKEND::MAJOR::dotu, -oneapi::mkl::blas::BACKEND::MAJOR::iamin, -oneapi::mkl::blas::BACKEND::MAJOR::iamin, -oneapi::mkl::blas::BACKEND::MAJOR::iamin, -oneapi::mkl::blas::BACKEND::MAJOR::iamin, -oneapi::mkl::blas::BACKEND::MAJOR::iamax, -oneapi::mkl::blas::BACKEND::MAJOR::iamax, -oneapi::mkl::blas::BACKEND::MAJOR::iamax, -oneapi::mkl::blas::BACKEND::MAJOR::iamax, -oneapi::mkl::blas::BACKEND::MAJOR::nrm2, -oneapi::mkl::blas::BACKEND::MAJOR::nrm2, -oneapi::mkl::blas::BACKEND::MAJOR::nrm2, -oneapi::mkl::blas::BACKEND::MAJOR::nrm2, -oneapi::mkl::blas::BACKEND::MAJOR::rot, -oneapi::mkl::blas::BACKEND::MAJOR::rot, -oneapi::mkl::blas::BACKEND::MAJOR::rot, -oneapi::mkl::blas::BACKEND::MAJOR::rot, -oneapi::mkl::blas::BACKEND::MAJOR::rotg, -oneapi::mkl::blas::BACKEND::MAJOR::rotg, -oneapi::mkl::blas::BACKEND::MAJOR::rotg, -oneapi::mkl::blas::BACKEND::MAJOR::rotg, -oneapi::mkl::blas::BACKEND::MAJOR::rotm, -oneapi::mkl::blas::BACKEND::MAJOR::rotm, -oneapi::mkl::blas::BACKEND::MAJOR::rotmg, -oneapi::mkl::blas::BACKEND::MAJOR::rotmg, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::scal, -oneapi::mkl::blas::BACKEND::MAJOR::sdsdot, -oneapi::mkl::blas::BACKEND::MAJOR::swap, -oneapi::mkl::blas::BACKEND::MAJOR::swap, -oneapi::mkl::blas::BACKEND::MAJOR::swap, -oneapi::mkl::blas::BACKEND::MAJOR::swap, -oneapi::mkl::blas::BACKEND::MAJOR::gbmv, -oneapi::mkl::blas::BACKEND::MAJOR::gbmv, -oneapi::mkl::blas::BACKEND::MAJOR::gbmv, -oneapi::mkl::blas::BACKEND::MAJOR::gbmv, -oneapi::mkl::blas::BACKEND::MAJOR::gemv, -oneapi::mkl::blas::BACKEND::MAJOR::gemv, -oneapi::mkl::blas::BACKEND::MAJOR::gemv, -oneapi::mkl::blas::BACKEND::MAJOR::gemv, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemv_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::dgmm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::ger, -oneapi::mkl::blas::BACKEND::MAJOR::ger, -oneapi::mkl::blas::BACKEND::MAJOR::gerc, -oneapi::mkl::blas::BACKEND::MAJOR::gerc, -oneapi::mkl::blas::BACKEND::MAJOR::geru, -oneapi::mkl::blas::BACKEND::MAJOR::geru, -oneapi::mkl::blas::BACKEND::MAJOR::hbmv, -oneapi::mkl::blas::BACKEND::MAJOR::hbmv, -oneapi::mkl::blas::BACKEND::MAJOR::hemv, -oneapi::mkl::blas::BACKEND::MAJOR::hemv, -oneapi::mkl::blas::BACKEND::MAJOR::her, -oneapi::mkl::blas::BACKEND::MAJOR::her, -oneapi::mkl::blas::BACKEND::MAJOR::her2, -oneapi::mkl::blas::BACKEND::MAJOR::her2, -oneapi::mkl::blas::BACKEND::MAJOR::hpmv, -oneapi::mkl::blas::BACKEND::MAJOR::hpmv, -oneapi::mkl::blas::BACKEND::MAJOR::hpr, -oneapi::mkl::blas::BACKEND::MAJOR::hpr, -oneapi::mkl::blas::BACKEND::MAJOR::hpr2, -oneapi::mkl::blas::BACKEND::MAJOR::hpr2, -oneapi::mkl::blas::BACKEND::MAJOR::sbmv, -oneapi::mkl::blas::BACKEND::MAJOR::sbmv, -oneapi::mkl::blas::BACKEND::MAJOR::spmv, -oneapi::mkl::blas::BACKEND::MAJOR::spmv, -oneapi::mkl::blas::BACKEND::MAJOR::spr, -oneapi::mkl::blas::BACKEND::MAJOR::spr, -oneapi::mkl::blas::BACKEND::MAJOR::spr2, -oneapi::mkl::blas::BACKEND::MAJOR::spr2, -oneapi::mkl::blas::BACKEND::MAJOR::symv, -oneapi::mkl::blas::BACKEND::MAJOR::symv, -oneapi::mkl::blas::BACKEND::MAJOR::syr, -oneapi::mkl::blas::BACKEND::MAJOR::syr, -oneapi::mkl::blas::BACKEND::MAJOR::syr2, -oneapi::mkl::blas::BACKEND::MAJOR::syr2, -oneapi::mkl::blas::BACKEND::MAJOR::tbmv, -oneapi::mkl::blas::BACKEND::MAJOR::tbmv, -oneapi::mkl::blas::BACKEND::MAJOR::tbmv, -oneapi::mkl::blas::BACKEND::MAJOR::tbmv, -oneapi::mkl::blas::BACKEND::MAJOR::tbsv, -oneapi::mkl::blas::BACKEND::MAJOR::tbsv, -oneapi::mkl::blas::BACKEND::MAJOR::tbsv, -oneapi::mkl::blas::BACKEND::MAJOR::tbsv, -oneapi::mkl::blas::BACKEND::MAJOR::tpmv, -oneapi::mkl::blas::BACKEND::MAJOR::tpmv, -oneapi::mkl::blas::BACKEND::MAJOR::tpmv, -oneapi::mkl::blas::BACKEND::MAJOR::tpmv, -oneapi::mkl::blas::BACKEND::MAJOR::tpsv, -oneapi::mkl::blas::BACKEND::MAJOR::tpsv, -oneapi::mkl::blas::BACKEND::MAJOR::tpsv, -oneapi::mkl::blas::BACKEND::MAJOR::tpsv, -oneapi::mkl::blas::BACKEND::MAJOR::trmv, -oneapi::mkl::blas::BACKEND::MAJOR::trmv, -oneapi::mkl::blas::BACKEND::MAJOR::trmv, -oneapi::mkl::blas::BACKEND::MAJOR::trmv, -oneapi::mkl::blas::BACKEND::MAJOR::trsv, -oneapi::mkl::blas::BACKEND::MAJOR::trsv, -oneapi::mkl::blas::BACKEND::MAJOR::trsv, -oneapi::mkl::blas::BACKEND::MAJOR::trsv, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::gemm, -oneapi::mkl::blas::BACKEND::MAJOR::hemm, -oneapi::mkl::blas::BACKEND::MAJOR::hemm, -oneapi::mkl::blas::BACKEND::MAJOR::herk, -oneapi::mkl::blas::BACKEND::MAJOR::herk, -oneapi::mkl::blas::BACKEND::MAJOR::her2k, -oneapi::mkl::blas::BACKEND::MAJOR::her2k, -oneapi::mkl::blas::BACKEND::MAJOR::symm, -oneapi::mkl::blas::BACKEND::MAJOR::symm, -oneapi::mkl::blas::BACKEND::MAJOR::symm, -oneapi::mkl::blas::BACKEND::MAJOR::symm, -oneapi::mkl::blas::BACKEND::MAJOR::syrk, -oneapi::mkl::blas::BACKEND::MAJOR::syrk, -oneapi::mkl::blas::BACKEND::MAJOR::syrk, -oneapi::mkl::blas::BACKEND::MAJOR::syrk, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syrk_batch, -oneapi::mkl::blas::BACKEND::MAJOR::syr2k, -oneapi::mkl::blas::BACKEND::MAJOR::syr2k, -oneapi::mkl::blas::BACKEND::MAJOR::syr2k, -oneapi::mkl::blas::BACKEND::MAJOR::syr2k, -oneapi::mkl::blas::BACKEND::MAJOR::trmm, -oneapi::mkl::blas::BACKEND::MAJOR::trmm, -oneapi::mkl::blas::BACKEND::MAJOR::trmm, -oneapi::mkl::blas::BACKEND::MAJOR::trmm, -oneapi::mkl::blas::BACKEND::MAJOR::trsm, -oneapi::mkl::blas::BACKEND::MAJOR::trsm, -oneapi::mkl::blas::BACKEND::MAJOR::trsm, -oneapi::mkl::blas::BACKEND::MAJOR::trsm, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::trsm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_batch, -oneapi::mkl::blas::BACKEND::MAJOR::gemmt, -oneapi::mkl::blas::BACKEND::MAJOR::gemmt, -oneapi::mkl::blas::BACKEND::MAJOR::gemmt, -oneapi::mkl::blas::BACKEND::MAJOR::gemmt, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_bias, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_bias, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_bias, -oneapi::mkl::blas::BACKEND::MAJOR::gemm_bias, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy2, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy2, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy2, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy2, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd, -oneapi::mkl::blas::BACKEND::MAJOR::omatadd, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::omatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, -oneapi::mkl::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::asum, +oneapi::math::blas::BACKEND::MAJOR::asum, +oneapi::math::blas::BACKEND::MAJOR::asum, +oneapi::math::blas::BACKEND::MAJOR::asum, +oneapi::math::blas::BACKEND::MAJOR::axpy, +oneapi::math::blas::BACKEND::MAJOR::axpy, +oneapi::math::blas::BACKEND::MAJOR::axpy, +oneapi::math::blas::BACKEND::MAJOR::axpy, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpby, +oneapi::math::blas::BACKEND::MAJOR::axpby, +oneapi::math::blas::BACKEND::MAJOR::axpby, +oneapi::math::blas::BACKEND::MAJOR::axpby, +oneapi::math::blas::BACKEND::MAJOR::copy, +oneapi::math::blas::BACKEND::MAJOR::copy, +oneapi::math::blas::BACKEND::MAJOR::copy, +oneapi::math::blas::BACKEND::MAJOR::copy, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::dot, +oneapi::math::blas::BACKEND::MAJOR::dot, +oneapi::math::blas::BACKEND::MAJOR::dot, +oneapi::math::blas::BACKEND::MAJOR::dotc, +oneapi::math::blas::BACKEND::MAJOR::dotc, +oneapi::math::blas::BACKEND::MAJOR::dotu, +oneapi::math::blas::BACKEND::MAJOR::dotu, +oneapi::math::blas::BACKEND::MAJOR::iamin, +oneapi::math::blas::BACKEND::MAJOR::iamin, +oneapi::math::blas::BACKEND::MAJOR::iamin, +oneapi::math::blas::BACKEND::MAJOR::iamin, +oneapi::math::blas::BACKEND::MAJOR::iamax, +oneapi::math::blas::BACKEND::MAJOR::iamax, +oneapi::math::blas::BACKEND::MAJOR::iamax, +oneapi::math::blas::BACKEND::MAJOR::iamax, +oneapi::math::blas::BACKEND::MAJOR::nrm2, +oneapi::math::blas::BACKEND::MAJOR::nrm2, +oneapi::math::blas::BACKEND::MAJOR::nrm2, +oneapi::math::blas::BACKEND::MAJOR::nrm2, +oneapi::math::blas::BACKEND::MAJOR::rot, +oneapi::math::blas::BACKEND::MAJOR::rot, +oneapi::math::blas::BACKEND::MAJOR::rot, +oneapi::math::blas::BACKEND::MAJOR::rot, +oneapi::math::blas::BACKEND::MAJOR::rotg, +oneapi::math::blas::BACKEND::MAJOR::rotg, +oneapi::math::blas::BACKEND::MAJOR::rotg, +oneapi::math::blas::BACKEND::MAJOR::rotg, +oneapi::math::blas::BACKEND::MAJOR::rotm, +oneapi::math::blas::BACKEND::MAJOR::rotm, +oneapi::math::blas::BACKEND::MAJOR::rotmg, +oneapi::math::blas::BACKEND::MAJOR::rotmg, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::sdsdot, +oneapi::math::blas::BACKEND::MAJOR::swap, +oneapi::math::blas::BACKEND::MAJOR::swap, +oneapi::math::blas::BACKEND::MAJOR::swap, +oneapi::math::blas::BACKEND::MAJOR::swap, +oneapi::math::blas::BACKEND::MAJOR::gbmv, +oneapi::math::blas::BACKEND::MAJOR::gbmv, +oneapi::math::blas::BACKEND::MAJOR::gbmv, +oneapi::math::blas::BACKEND::MAJOR::gbmv, +oneapi::math::blas::BACKEND::MAJOR::gemv, +oneapi::math::blas::BACKEND::MAJOR::gemv, +oneapi::math::blas::BACKEND::MAJOR::gemv, +oneapi::math::blas::BACKEND::MAJOR::gemv, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::ger, +oneapi::math::blas::BACKEND::MAJOR::ger, +oneapi::math::blas::BACKEND::MAJOR::gerc, +oneapi::math::blas::BACKEND::MAJOR::gerc, +oneapi::math::blas::BACKEND::MAJOR::geru, +oneapi::math::blas::BACKEND::MAJOR::geru, +oneapi::math::blas::BACKEND::MAJOR::hbmv, +oneapi::math::blas::BACKEND::MAJOR::hbmv, +oneapi::math::blas::BACKEND::MAJOR::hemv, +oneapi::math::blas::BACKEND::MAJOR::hemv, +oneapi::math::blas::BACKEND::MAJOR::her, +oneapi::math::blas::BACKEND::MAJOR::her, +oneapi::math::blas::BACKEND::MAJOR::her2, +oneapi::math::blas::BACKEND::MAJOR::her2, +oneapi::math::blas::BACKEND::MAJOR::hpmv, +oneapi::math::blas::BACKEND::MAJOR::hpmv, +oneapi::math::blas::BACKEND::MAJOR::hpr, +oneapi::math::blas::BACKEND::MAJOR::hpr, +oneapi::math::blas::BACKEND::MAJOR::hpr2, +oneapi::math::blas::BACKEND::MAJOR::hpr2, +oneapi::math::blas::BACKEND::MAJOR::sbmv, +oneapi::math::blas::BACKEND::MAJOR::sbmv, +oneapi::math::blas::BACKEND::MAJOR::spmv, +oneapi::math::blas::BACKEND::MAJOR::spmv, +oneapi::math::blas::BACKEND::MAJOR::spr, +oneapi::math::blas::BACKEND::MAJOR::spr, +oneapi::math::blas::BACKEND::MAJOR::spr2, +oneapi::math::blas::BACKEND::MAJOR::spr2, +oneapi::math::blas::BACKEND::MAJOR::symv, +oneapi::math::blas::BACKEND::MAJOR::symv, +oneapi::math::blas::BACKEND::MAJOR::syr, +oneapi::math::blas::BACKEND::MAJOR::syr, +oneapi::math::blas::BACKEND::MAJOR::syr2, +oneapi::math::blas::BACKEND::MAJOR::syr2, +oneapi::math::blas::BACKEND::MAJOR::tbmv, +oneapi::math::blas::BACKEND::MAJOR::tbmv, +oneapi::math::blas::BACKEND::MAJOR::tbmv, +oneapi::math::blas::BACKEND::MAJOR::tbmv, +oneapi::math::blas::BACKEND::MAJOR::tbsv, +oneapi::math::blas::BACKEND::MAJOR::tbsv, +oneapi::math::blas::BACKEND::MAJOR::tbsv, +oneapi::math::blas::BACKEND::MAJOR::tbsv, +oneapi::math::blas::BACKEND::MAJOR::tpmv, +oneapi::math::blas::BACKEND::MAJOR::tpmv, +oneapi::math::blas::BACKEND::MAJOR::tpmv, +oneapi::math::blas::BACKEND::MAJOR::tpmv, +oneapi::math::blas::BACKEND::MAJOR::tpsv, +oneapi::math::blas::BACKEND::MAJOR::tpsv, +oneapi::math::blas::BACKEND::MAJOR::tpsv, +oneapi::math::blas::BACKEND::MAJOR::tpsv, +oneapi::math::blas::BACKEND::MAJOR::trmv, +oneapi::math::blas::BACKEND::MAJOR::trmv, +oneapi::math::blas::BACKEND::MAJOR::trmv, +oneapi::math::blas::BACKEND::MAJOR::trmv, +oneapi::math::blas::BACKEND::MAJOR::trsv, +oneapi::math::blas::BACKEND::MAJOR::trsv, +oneapi::math::blas::BACKEND::MAJOR::trsv, +oneapi::math::blas::BACKEND::MAJOR::trsv, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::hemm, +oneapi::math::blas::BACKEND::MAJOR::hemm, +oneapi::math::blas::BACKEND::MAJOR::herk, +oneapi::math::blas::BACKEND::MAJOR::herk, +oneapi::math::blas::BACKEND::MAJOR::her2k, +oneapi::math::blas::BACKEND::MAJOR::her2k, +oneapi::math::blas::BACKEND::MAJOR::symm, +oneapi::math::blas::BACKEND::MAJOR::symm, +oneapi::math::blas::BACKEND::MAJOR::symm, +oneapi::math::blas::BACKEND::MAJOR::symm, +oneapi::math::blas::BACKEND::MAJOR::syrk, +oneapi::math::blas::BACKEND::MAJOR::syrk, +oneapi::math::blas::BACKEND::MAJOR::syrk, +oneapi::math::blas::BACKEND::MAJOR::syrk, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syr2k, +oneapi::math::blas::BACKEND::MAJOR::syr2k, +oneapi::math::blas::BACKEND::MAJOR::syr2k, +oneapi::math::blas::BACKEND::MAJOR::syr2k, +oneapi::math::blas::BACKEND::MAJOR::trmm, +oneapi::math::blas::BACKEND::MAJOR::trmm, +oneapi::math::blas::BACKEND::MAJOR::trmm, +oneapi::math::blas::BACKEND::MAJOR::trmm, +oneapi::math::blas::BACKEND::MAJOR::trsm, +oneapi::math::blas::BACKEND::MAJOR::trsm, +oneapi::math::blas::BACKEND::MAJOR::trsm, +oneapi::math::blas::BACKEND::MAJOR::trsm, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemmt, +oneapi::math::blas::BACKEND::MAJOR::gemmt, +oneapi::math::blas::BACKEND::MAJOR::gemmt, +oneapi::math::blas::BACKEND::MAJOR::gemmt, +oneapi::math::blas::BACKEND::MAJOR::gemm_bias, +oneapi::math::blas::BACKEND::MAJOR::gemm_bias, +oneapi::math::blas::BACKEND::MAJOR::gemm_bias, +oneapi::math::blas::BACKEND::MAJOR::gemm_bias, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::omatadd_batch, +oneapi::math::blas::BACKEND::MAJOR::omatadd_batch, +oneapi::math::blas::BACKEND::MAJOR::omatadd_batch, +oneapi::math::blas::BACKEND::MAJOR::omatadd_batch, +oneapi::math::blas::BACKEND::MAJOR::omatcopy, +oneapi::math::blas::BACKEND::MAJOR::omatcopy, +oneapi::math::blas::BACKEND::MAJOR::omatcopy, +oneapi::math::blas::BACKEND::MAJOR::omatcopy, +oneapi::math::blas::BACKEND::MAJOR::omatcopy2, +oneapi::math::blas::BACKEND::MAJOR::omatcopy2, +oneapi::math::blas::BACKEND::MAJOR::omatcopy2, +oneapi::math::blas::BACKEND::MAJOR::omatcopy2, +oneapi::math::blas::BACKEND::MAJOR::imatcopy, +oneapi::math::blas::BACKEND::MAJOR::imatcopy, +oneapi::math::blas::BACKEND::MAJOR::imatcopy, +oneapi::math::blas::BACKEND::MAJOR::imatcopy, +oneapi::math::blas::BACKEND::MAJOR::omatadd, +oneapi::math::blas::BACKEND::MAJOR::omatadd, +oneapi::math::blas::BACKEND::MAJOR::omatadd, +oneapi::math::blas::BACKEND::MAJOR::omatadd, +oneapi::math::blas::BACKEND::MAJOR::asum, +oneapi::math::blas::BACKEND::MAJOR::asum, +oneapi::math::blas::BACKEND::MAJOR::asum, +oneapi::math::blas::BACKEND::MAJOR::asum, +oneapi::math::blas::BACKEND::MAJOR::axpy, +oneapi::math::blas::BACKEND::MAJOR::axpy, +oneapi::math::blas::BACKEND::MAJOR::axpy, +oneapi::math::blas::BACKEND::MAJOR::axpy, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpy_batch, +oneapi::math::blas::BACKEND::MAJOR::axpby, +oneapi::math::blas::BACKEND::MAJOR::axpby, +oneapi::math::blas::BACKEND::MAJOR::axpby, +oneapi::math::blas::BACKEND::MAJOR::axpby, +oneapi::math::blas::BACKEND::MAJOR::copy, +oneapi::math::blas::BACKEND::MAJOR::copy, +oneapi::math::blas::BACKEND::MAJOR::copy, +oneapi::math::blas::BACKEND::MAJOR::copy, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::copy_batch, +oneapi::math::blas::BACKEND::MAJOR::dot, +oneapi::math::blas::BACKEND::MAJOR::dot, +oneapi::math::blas::BACKEND::MAJOR::dot, +oneapi::math::blas::BACKEND::MAJOR::dotc, +oneapi::math::blas::BACKEND::MAJOR::dotc, +oneapi::math::blas::BACKEND::MAJOR::dotu, +oneapi::math::blas::BACKEND::MAJOR::dotu, +oneapi::math::blas::BACKEND::MAJOR::iamin, +oneapi::math::blas::BACKEND::MAJOR::iamin, +oneapi::math::blas::BACKEND::MAJOR::iamin, +oneapi::math::blas::BACKEND::MAJOR::iamin, +oneapi::math::blas::BACKEND::MAJOR::iamax, +oneapi::math::blas::BACKEND::MAJOR::iamax, +oneapi::math::blas::BACKEND::MAJOR::iamax, +oneapi::math::blas::BACKEND::MAJOR::iamax, +oneapi::math::blas::BACKEND::MAJOR::nrm2, +oneapi::math::blas::BACKEND::MAJOR::nrm2, +oneapi::math::blas::BACKEND::MAJOR::nrm2, +oneapi::math::blas::BACKEND::MAJOR::nrm2, +oneapi::math::blas::BACKEND::MAJOR::rot, +oneapi::math::blas::BACKEND::MAJOR::rot, +oneapi::math::blas::BACKEND::MAJOR::rot, +oneapi::math::blas::BACKEND::MAJOR::rot, +oneapi::math::blas::BACKEND::MAJOR::rotg, +oneapi::math::blas::BACKEND::MAJOR::rotg, +oneapi::math::blas::BACKEND::MAJOR::rotg, +oneapi::math::blas::BACKEND::MAJOR::rotg, +oneapi::math::blas::BACKEND::MAJOR::rotm, +oneapi::math::blas::BACKEND::MAJOR::rotm, +oneapi::math::blas::BACKEND::MAJOR::rotmg, +oneapi::math::blas::BACKEND::MAJOR::rotmg, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::scal, +oneapi::math::blas::BACKEND::MAJOR::sdsdot, +oneapi::math::blas::BACKEND::MAJOR::swap, +oneapi::math::blas::BACKEND::MAJOR::swap, +oneapi::math::blas::BACKEND::MAJOR::swap, +oneapi::math::blas::BACKEND::MAJOR::swap, +oneapi::math::blas::BACKEND::MAJOR::gbmv, +oneapi::math::blas::BACKEND::MAJOR::gbmv, +oneapi::math::blas::BACKEND::MAJOR::gbmv, +oneapi::math::blas::BACKEND::MAJOR::gbmv, +oneapi::math::blas::BACKEND::MAJOR::gemv, +oneapi::math::blas::BACKEND::MAJOR::gemv, +oneapi::math::blas::BACKEND::MAJOR::gemv, +oneapi::math::blas::BACKEND::MAJOR::gemv, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::gemv_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::dgmm_batch, +oneapi::math::blas::BACKEND::MAJOR::ger, +oneapi::math::blas::BACKEND::MAJOR::ger, +oneapi::math::blas::BACKEND::MAJOR::gerc, +oneapi::math::blas::BACKEND::MAJOR::gerc, +oneapi::math::blas::BACKEND::MAJOR::geru, +oneapi::math::blas::BACKEND::MAJOR::geru, +oneapi::math::blas::BACKEND::MAJOR::hbmv, +oneapi::math::blas::BACKEND::MAJOR::hbmv, +oneapi::math::blas::BACKEND::MAJOR::hemv, +oneapi::math::blas::BACKEND::MAJOR::hemv, +oneapi::math::blas::BACKEND::MAJOR::her, +oneapi::math::blas::BACKEND::MAJOR::her, +oneapi::math::blas::BACKEND::MAJOR::her2, +oneapi::math::blas::BACKEND::MAJOR::her2, +oneapi::math::blas::BACKEND::MAJOR::hpmv, +oneapi::math::blas::BACKEND::MAJOR::hpmv, +oneapi::math::blas::BACKEND::MAJOR::hpr, +oneapi::math::blas::BACKEND::MAJOR::hpr, +oneapi::math::blas::BACKEND::MAJOR::hpr2, +oneapi::math::blas::BACKEND::MAJOR::hpr2, +oneapi::math::blas::BACKEND::MAJOR::sbmv, +oneapi::math::blas::BACKEND::MAJOR::sbmv, +oneapi::math::blas::BACKEND::MAJOR::spmv, +oneapi::math::blas::BACKEND::MAJOR::spmv, +oneapi::math::blas::BACKEND::MAJOR::spr, +oneapi::math::blas::BACKEND::MAJOR::spr, +oneapi::math::blas::BACKEND::MAJOR::spr2, +oneapi::math::blas::BACKEND::MAJOR::spr2, +oneapi::math::blas::BACKEND::MAJOR::symv, +oneapi::math::blas::BACKEND::MAJOR::symv, +oneapi::math::blas::BACKEND::MAJOR::syr, +oneapi::math::blas::BACKEND::MAJOR::syr, +oneapi::math::blas::BACKEND::MAJOR::syr2, +oneapi::math::blas::BACKEND::MAJOR::syr2, +oneapi::math::blas::BACKEND::MAJOR::tbmv, +oneapi::math::blas::BACKEND::MAJOR::tbmv, +oneapi::math::blas::BACKEND::MAJOR::tbmv, +oneapi::math::blas::BACKEND::MAJOR::tbmv, +oneapi::math::blas::BACKEND::MAJOR::tbsv, +oneapi::math::blas::BACKEND::MAJOR::tbsv, +oneapi::math::blas::BACKEND::MAJOR::tbsv, +oneapi::math::blas::BACKEND::MAJOR::tbsv, +oneapi::math::blas::BACKEND::MAJOR::tpmv, +oneapi::math::blas::BACKEND::MAJOR::tpmv, +oneapi::math::blas::BACKEND::MAJOR::tpmv, +oneapi::math::blas::BACKEND::MAJOR::tpmv, +oneapi::math::blas::BACKEND::MAJOR::tpsv, +oneapi::math::blas::BACKEND::MAJOR::tpsv, +oneapi::math::blas::BACKEND::MAJOR::tpsv, +oneapi::math::blas::BACKEND::MAJOR::tpsv, +oneapi::math::blas::BACKEND::MAJOR::trmv, +oneapi::math::blas::BACKEND::MAJOR::trmv, +oneapi::math::blas::BACKEND::MAJOR::trmv, +oneapi::math::blas::BACKEND::MAJOR::trmv, +oneapi::math::blas::BACKEND::MAJOR::trsv, +oneapi::math::blas::BACKEND::MAJOR::trsv, +oneapi::math::blas::BACKEND::MAJOR::trsv, +oneapi::math::blas::BACKEND::MAJOR::trsv, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::gemm, +oneapi::math::blas::BACKEND::MAJOR::hemm, +oneapi::math::blas::BACKEND::MAJOR::hemm, +oneapi::math::blas::BACKEND::MAJOR::herk, +oneapi::math::blas::BACKEND::MAJOR::herk, +oneapi::math::blas::BACKEND::MAJOR::her2k, +oneapi::math::blas::BACKEND::MAJOR::her2k, +oneapi::math::blas::BACKEND::MAJOR::symm, +oneapi::math::blas::BACKEND::MAJOR::symm, +oneapi::math::blas::BACKEND::MAJOR::symm, +oneapi::math::blas::BACKEND::MAJOR::symm, +oneapi::math::blas::BACKEND::MAJOR::syrk, +oneapi::math::blas::BACKEND::MAJOR::syrk, +oneapi::math::blas::BACKEND::MAJOR::syrk, +oneapi::math::blas::BACKEND::MAJOR::syrk, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syrk_batch, +oneapi::math::blas::BACKEND::MAJOR::syr2k, +oneapi::math::blas::BACKEND::MAJOR::syr2k, +oneapi::math::blas::BACKEND::MAJOR::syr2k, +oneapi::math::blas::BACKEND::MAJOR::syr2k, +oneapi::math::blas::BACKEND::MAJOR::trmm, +oneapi::math::blas::BACKEND::MAJOR::trmm, +oneapi::math::blas::BACKEND::MAJOR::trmm, +oneapi::math::blas::BACKEND::MAJOR::trmm, +oneapi::math::blas::BACKEND::MAJOR::trsm, +oneapi::math::blas::BACKEND::MAJOR::trsm, +oneapi::math::blas::BACKEND::MAJOR::trsm, +oneapi::math::blas::BACKEND::MAJOR::trsm, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::trsm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemm_batch, +oneapi::math::blas::BACKEND::MAJOR::gemmt, +oneapi::math::blas::BACKEND::MAJOR::gemmt, +oneapi::math::blas::BACKEND::MAJOR::gemmt, +oneapi::math::blas::BACKEND::MAJOR::gemmt, +oneapi::math::blas::BACKEND::MAJOR::gemm_bias, +oneapi::math::blas::BACKEND::MAJOR::gemm_bias, +oneapi::math::blas::BACKEND::MAJOR::gemm_bias, +oneapi::math::blas::BACKEND::MAJOR::gemm_bias, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::omatadd_batch, +oneapi::math::blas::BACKEND::MAJOR::omatadd_batch, +oneapi::math::blas::BACKEND::MAJOR::omatadd_batch, +oneapi::math::blas::BACKEND::MAJOR::omatadd_batch, +oneapi::math::blas::BACKEND::MAJOR::omatcopy, +oneapi::math::blas::BACKEND::MAJOR::omatcopy, +oneapi::math::blas::BACKEND::MAJOR::omatcopy, +oneapi::math::blas::BACKEND::MAJOR::omatcopy, +oneapi::math::blas::BACKEND::MAJOR::omatcopy2, +oneapi::math::blas::BACKEND::MAJOR::omatcopy2, +oneapi::math::blas::BACKEND::MAJOR::omatcopy2, +oneapi::math::blas::BACKEND::MAJOR::omatcopy2, +oneapi::math::blas::BACKEND::MAJOR::imatcopy, +oneapi::math::blas::BACKEND::MAJOR::imatcopy, +oneapi::math::blas::BACKEND::MAJOR::imatcopy, +oneapi::math::blas::BACKEND::MAJOR::imatcopy, +oneapi::math::blas::BACKEND::MAJOR::omatadd, +oneapi::math::blas::BACKEND::MAJOR::omatadd, +oneapi::math::blas::BACKEND::MAJOR::omatadd, +oneapi::math::blas::BACKEND::MAJOR::omatadd, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::omatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, +oneapi::math::blas::BACKEND::MAJOR::imatcopy_batch, // clang-format on diff --git a/src/blas/backends/cublas/cublas_batch.cpp b/src/blas/backends/cublas/cublas_batch.cpp index b8d768a1d..dc53e0679 100644 --- a/src/blas/backends/cublas/cublas_batch.cpp +++ b/src/blas/backends/cublas/cublas_batch.cpp @@ -22,7 +22,7 @@ #include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { namespace column_major { @@ -155,7 +155,7 @@ inline void gemm_batch_impl(sycl::queue &queue, transpose transa, transpose tran cublasGemmAlgo_t cublas_gemm_algo = CUBLAS_GEMM_DEFAULT; queue.submit([&](sycl::handler &cgh) { if (!verify_support(queue, sycl::aspect::fp16)) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "blas", "sycl::half", "half is not supported by the device or the sycl compiler"); } auto a_acc = a.template get_access(cgh); @@ -625,7 +625,7 @@ inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose tra cublasGemmAlgo_t cublas_gemm_algo = CUBLAS_GEMM_DEFAULT; auto done = queue.submit([&](sycl::handler &cgh) { if (!verify_support(queue, sycl::aspect::fp16)) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "blas", "sycl::half", "half is not supported by the device or the sycl compiler"); } int64_t num_events = dependencies.size(); @@ -711,7 +711,7 @@ inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, tr cublasGemmAlgo_t cublas_gemm_algo = CUBLAS_GEMM_DEFAULT; auto done = queue.submit([&](sycl::handler &cgh) { if (!verify_support(queue, sycl::aspect::fp16)) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "blas", "sycl::half", "half is not supported by the device or the sycl compiler"); } int64_t num_events = dependencies.size(); @@ -1888,5 +1888,5 @@ sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int } // namespace row_major } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/cublas/cublas_extensions.cpp b/src/blas/backends/cublas/cublas_extensions.cpp index d6e358cbc..db6c0aa05 100644 --- a/src/blas/backends/cublas/cublas_extensions.cpp +++ b/src/blas/backends/cublas/cublas_extensions.cpp @@ -22,7 +22,7 @@ #include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { namespace column_major { @@ -94,8 +94,8 @@ void omatcopy(const char *func_name, Func func, sycl::queue &queue, transpose tr queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - const int64_t logical_m = (trans == oneapi::mkl::transpose::nontrans ? m : n); - const int64_t logical_n = (trans == oneapi::mkl::transpose::nontrans ? n : m); + const int64_t logical_m = (trans == oneapi::math::transpose::nontrans ? m : n); + const int64_t logical_n = (trans == oneapi::math::transpose::nontrans ? n : m); onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -272,8 +272,8 @@ sycl::event omatcopy(const char *func_name, Func func, sycl::queue &queue, trans overflow_check(m, n, lda, ldb); auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - const int64_t logical_m = (trans == oneapi::mkl::transpose::nontrans ? m : n); - const int64_t logical_n = (trans == oneapi::mkl::transpose::nontrans ? n : m); + const int64_t logical_m = (trans == oneapi::math::transpose::nontrans ? m : n); + const int64_t logical_n = (trans == oneapi::math::transpose::nontrans ? n : m); onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -457,8 +457,8 @@ void omatcopy(const char *func_name, Func func, sycl::queue &queue, transpose tr queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - const int64_t logical_m = (trans == oneapi::mkl::transpose::nontrans ? n : m); - const int64_t logical_n = (trans == oneapi::mkl::transpose::nontrans ? m : n); + const int64_t logical_m = (trans == oneapi::math::transpose::nontrans ? n : m); + const int64_t logical_n = (trans == oneapi::math::transpose::nontrans ? m : n); onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = sc.get_mem(a_acc); @@ -635,8 +635,8 @@ sycl::event omatcopy(const char *func_name, Func func, sycl::queue &queue, trans overflow_check(m, n, lda, ldb); auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); - const int64_t logical_m = (trans == oneapi::mkl::transpose::nontrans ? n : m); - const int64_t logical_n = (trans == oneapi::mkl::transpose::nontrans ? m : n); + const int64_t logical_m = (trans == oneapi::math::transpose::nontrans ? n : m); + const int64_t logical_n = (trans == oneapi::math::transpose::nontrans ? m : n); onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { auto handle = sc.get_handle(queue); auto a_ = reinterpret_cast(a); @@ -752,5 +752,5 @@ OMATADD_LAUNCHER_USM(std::complex, cublasZgeam) } // namespace row_major } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/cublas/cublas_handle.hpp b/src/blas/backends/cublas/cublas_handle.hpp index db9df5584..254c1c160 100644 --- a/src/blas/backends/cublas/cublas_handle.hpp +++ b/src/blas/backends/cublas/cublas_handle.hpp @@ -22,7 +22,7 @@ #include namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { @@ -55,7 +55,7 @@ struct cublas_handle { } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif // CUBLAS_HANDLE_HPP diff --git a/src/blas/backends/cublas/cublas_helper.hpp b/src/blas/backends/cublas/cublas_helper.hpp index a8bc388d7..27b4a6bf9 100644 --- a/src/blas/backends/cublas/cublas_helper.hpp +++ b/src/blas/backends/cublas/cublas_helper.hpp @@ -38,7 +38,7 @@ #include "dtype_string.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { @@ -226,35 +226,35 @@ inline void cublas_native_named_func(const char *func_name, Func func, #endif }; -inline cublasOperation_t get_cublas_operation(oneapi::mkl::transpose trn) { +inline cublasOperation_t get_cublas_operation(oneapi::math::transpose trn) { switch (trn) { - case oneapi::mkl::transpose::nontrans: return CUBLAS_OP_N; - case oneapi::mkl::transpose::trans: return CUBLAS_OP_T; - case oneapi::mkl::transpose::conjtrans: return CUBLAS_OP_C; + case oneapi::math::transpose::nontrans: return CUBLAS_OP_N; + case oneapi::math::transpose::trans: return CUBLAS_OP_T; + case oneapi::math::transpose::conjtrans: return CUBLAS_OP_C; default: throw "Wrong transpose Operation."; } } -inline cublasFillMode_t get_cublas_fill_mode(oneapi::mkl::uplo ul) { +inline cublasFillMode_t get_cublas_fill_mode(oneapi::math::uplo ul) { switch (ul) { - case oneapi::mkl::uplo::upper: return CUBLAS_FILL_MODE_UPPER; - case oneapi::mkl::uplo::lower: return CUBLAS_FILL_MODE_LOWER; + case oneapi::math::uplo::upper: return CUBLAS_FILL_MODE_UPPER; + case oneapi::math::uplo::lower: return CUBLAS_FILL_MODE_LOWER; default: throw "Wrong fill mode."; } } -inline cublasDiagType_t get_cublas_diag_type(oneapi::mkl::diag un) { +inline cublasDiagType_t get_cublas_diag_type(oneapi::math::diag un) { switch (un) { - case oneapi::mkl::diag::unit: return CUBLAS_DIAG_UNIT; - case oneapi::mkl::diag::nonunit: return CUBLAS_DIAG_NON_UNIT; + case oneapi::math::diag::unit: return CUBLAS_DIAG_UNIT; + case oneapi::math::diag::nonunit: return CUBLAS_DIAG_NON_UNIT; default: throw "Wrong diag type."; } } -inline cublasSideMode_t get_cublas_side_mode(oneapi::mkl::side lr) { +inline cublasSideMode_t get_cublas_side_mode(oneapi::math::side lr) { switch (lr) { - case oneapi::mkl::side::left: return CUBLAS_SIDE_LEFT; - case oneapi::mkl::side::right: return CUBLAS_SIDE_RIGHT; + case oneapi::math::side::left: return CUBLAS_SIDE_LEFT; + case oneapi::math::side::right: return CUBLAS_SIDE_RIGHT; default: throw "Wrong side mode."; } } @@ -330,6 +330,6 @@ struct CudaEquivalentType> { } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _CUBLAS_HELPER_HPP_ diff --git a/src/blas/backends/cublas/cublas_level1.cpp b/src/blas/backends/cublas/cublas_level1.cpp index 77dd6476b..5b0e39978 100644 --- a/src/blas/backends/cublas/cublas_level1.cpp +++ b/src/blas/backends/cublas/cublas_level1.cpp @@ -23,7 +23,7 @@ #include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { namespace column_major { @@ -778,7 +778,7 @@ inline sycl::event rotg(const char *func_name, Func func, sycl::queue &queue, T1 sycl::get_pointer_type(b, ctx) == sycl::usm::alloc::unknown || sycl::get_pointer_type(c, ctx) == sycl::usm::alloc::unknown || sycl::get_pointer_type(s, ctx) == sycl::usm::alloc::unknown) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "blas", "rotg", "If any pointer is only device accessible, all must be device accessible"); } @@ -1034,7 +1034,7 @@ inline sycl::event rotmg(const char *func_name, Func func, sycl::queue &queue, T if (sycl::get_pointer_type(d1, ctx) == sycl::usm::alloc::unknown || sycl::get_pointer_type(d2, ctx) == sycl::usm::alloc::unknown || sycl::get_pointer_type(x1, ctx) == sycl::usm::alloc::unknown) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "blas", "rotmg", "If any pointer is only device accessible, all must be device accessible"); } @@ -1849,5 +1849,5 @@ NRM2_LAUNCHER_USM(std::complex, double, cublasDznrm2) } // namespace row_major } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/cublas/cublas_level2.cpp b/src/blas/backends/cublas/cublas_level2.cpp index 0f9992df2..1bbfbc60f 100644 --- a/src/blas/backends/cublas/cublas_level2.cpp +++ b/src/blas/backends/cublas/cublas_level2.cpp @@ -23,7 +23,7 @@ #include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { namespace column_major { @@ -2698,5 +2698,5 @@ TRSV_LAUNCHER_USM(std::complex, cublasZtrsv) } // namespace row_major } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/cublas/cublas_level3.cpp b/src/blas/backends/cublas/cublas_level3.cpp index 5ef3ea9bb..ec6dc5192 100644 --- a/src/blas/backends/cublas/cublas_level3.cpp +++ b/src/blas/backends/cublas/cublas_level3.cpp @@ -23,7 +23,7 @@ #include "oneapi/math/blas/detail/cublas/onemath_blas_cublas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { namespace column_major { @@ -82,7 +82,7 @@ inline void gemm_ex(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, sycl::que overflow_check(m, n, k, lda, ldb, ldc); queue.submit([&](sycl::handler &cgh) { if (!verify_support(queue, sycl::aspect::fp16)) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "blas", "sycl::half", "half is not supported by the device or the sycl compiler"); } auto a_acc = a.template get_access(cgh); @@ -1346,5 +1346,5 @@ TRSM_LAUNCHER_USM(std::complex, cublasZtrsm) } // namespace row_major } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/cublas/cublas_scope_handle.cpp b/src/blas/backends/cublas/cublas_scope_handle.cpp index 947cd1da1..b83185545 100644 --- a/src/blas/backends/cublas/cublas_scope_handle.cpp +++ b/src/blas/backends/cublas/cublas_scope_handle.cpp @@ -24,7 +24,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { @@ -148,5 +148,5 @@ sycl::context CublasScopedContextHandler::get_context(const sycl::queue &queue) } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/cublas/cublas_scope_handle.hpp b/src/blas/backends/cublas/cublas_scope_handle.hpp index 27bcfd495..3e0b4a8af 100644 --- a/src/blas/backends/cublas/cublas_scope_handle.hpp +++ b/src/blas/backends/cublas/cublas_scope_handle.hpp @@ -53,7 +53,7 @@ #include "cublas_handle.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { @@ -123,6 +123,6 @@ class CublasScopedContextHandler { } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_CUBLAS_SCOPED_HANDLE_HPP_ diff --git a/src/blas/backends/cublas/cublas_scope_handle_hipsycl.cpp b/src/blas/backends/cublas/cublas_scope_handle_hipsycl.cpp index 20675c212..82870f0ae 100644 --- a/src/blas/backends/cublas/cublas_scope_handle_hipsycl.cpp +++ b/src/blas/backends/cublas/cublas_scope_handle_hipsycl.cpp @@ -20,7 +20,7 @@ #include "cublas_handle.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { @@ -70,5 +70,5 @@ CUstream CublasScopedContextHandler::get_stream(const sycl::queue &queue) { } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi \ No newline at end of file diff --git a/src/blas/backends/cublas/cublas_scope_handle_hipsycl.hpp b/src/blas/backends/cublas/cublas_scope_handle_hipsycl.hpp index c7ec3e520..bc358d9ca 100644 --- a/src/blas/backends/cublas/cublas_scope_handle_hipsycl.hpp +++ b/src/blas/backends/cublas/cublas_scope_handle_hipsycl.hpp @@ -29,7 +29,7 @@ #include "cublas_helper.hpp" #include "cublas_handle.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { @@ -79,6 +79,6 @@ class CublasScopedContextHandler { } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif //CUBLAS_SCOPED_HANDLE_HIPSYCL_HPP diff --git a/src/blas/backends/cublas/cublas_task.hpp b/src/blas/backends/cublas/cublas_task.hpp index 63afea9c9..e8702fa62 100644 --- a/src/blas/backends/cublas/cublas_task.hpp +++ b/src/blas/backends/cublas/cublas_task.hpp @@ -52,7 +52,7 @@ using interop_handler = sycl::interop_handle; } #endif namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace cublas { @@ -84,6 +84,6 @@ static inline void onemath_cublas_host_task(H &cgh, sycl::queue queue, F f) { } // namespace cublas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _MKL_BLAS_CUBLAS_TASK_HPP_ diff --git a/src/blas/backends/cublas/cublas_wrappers.cpp b/src/blas/backends/cublas/cublas_wrappers.cpp index 0752fb247..ec89f9c58 100644 --- a/src/blas/backends/cublas/cublas_wrappers.cpp +++ b/src/blas/backends/cublas/cublas_wrappers.cpp @@ -23,984 +23,984 @@ extern "C" blas_function_table_t mkl_blas_table = { WRAPPER_VERSION, - oneapi::mkl::blas::cublas::column_major::asum, - oneapi::mkl::blas::cublas::column_major::asum, - oneapi::mkl::blas::cublas::column_major::asum, - oneapi::mkl::blas::cublas::column_major::asum, - oneapi::mkl::blas::cublas::column_major::axpy, - oneapi::mkl::blas::cublas::column_major::axpy, - oneapi::mkl::blas::cublas::column_major::axpy, - oneapi::mkl::blas::cublas::column_major::axpy, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpby, - oneapi::mkl::blas::cublas::column_major::axpby, - oneapi::mkl::blas::cublas::column_major::axpby, - oneapi::mkl::blas::cublas::column_major::axpby, - oneapi::mkl::blas::cublas::column_major::copy, - oneapi::mkl::blas::cublas::column_major::copy, - oneapi::mkl::blas::cublas::column_major::copy, - oneapi::mkl::blas::cublas::column_major::copy, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::dot, - oneapi::mkl::blas::cublas::column_major::dot, - oneapi::mkl::blas::cublas::column_major::dot, - oneapi::mkl::blas::cublas::column_major::dotc, - oneapi::mkl::blas::cublas::column_major::dotc, - oneapi::mkl::blas::cublas::column_major::dotu, - oneapi::mkl::blas::cublas::column_major::dotu, - oneapi::mkl::blas::cublas::column_major::iamin, - oneapi::mkl::blas::cublas::column_major::iamin, - oneapi::mkl::blas::cublas::column_major::iamin, - oneapi::mkl::blas::cublas::column_major::iamin, - oneapi::mkl::blas::cublas::column_major::iamax, - oneapi::mkl::blas::cublas::column_major::iamax, - oneapi::mkl::blas::cublas::column_major::iamax, - oneapi::mkl::blas::cublas::column_major::iamax, - oneapi::mkl::blas::cublas::column_major::nrm2, - oneapi::mkl::blas::cublas::column_major::nrm2, - oneapi::mkl::blas::cublas::column_major::nrm2, - oneapi::mkl::blas::cublas::column_major::nrm2, - oneapi::mkl::blas::cublas::column_major::rot, - oneapi::mkl::blas::cublas::column_major::rot, - oneapi::mkl::blas::cublas::column_major::rot, - oneapi::mkl::blas::cublas::column_major::rot, - oneapi::mkl::blas::cublas::column_major::rotg, - oneapi::mkl::blas::cublas::column_major::rotg, - oneapi::mkl::blas::cublas::column_major::rotg, - oneapi::mkl::blas::cublas::column_major::rotg, - oneapi::mkl::blas::cublas::column_major::rotm, - oneapi::mkl::blas::cublas::column_major::rotm, - oneapi::mkl::blas::cublas::column_major::rotmg, - oneapi::mkl::blas::cublas::column_major::rotmg, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::sdsdot, - oneapi::mkl::blas::cublas::column_major::swap, - oneapi::mkl::blas::cublas::column_major::swap, - oneapi::mkl::blas::cublas::column_major::swap, - oneapi::mkl::blas::cublas::column_major::swap, - oneapi::mkl::blas::cublas::column_major::gbmv, - oneapi::mkl::blas::cublas::column_major::gbmv, - oneapi::mkl::blas::cublas::column_major::gbmv, - oneapi::mkl::blas::cublas::column_major::gbmv, - oneapi::mkl::blas::cublas::column_major::gemv, - oneapi::mkl::blas::cublas::column_major::gemv, - oneapi::mkl::blas::cublas::column_major::gemv, - oneapi::mkl::blas::cublas::column_major::gemv, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::ger, - oneapi::mkl::blas::cublas::column_major::ger, - oneapi::mkl::blas::cublas::column_major::gerc, - oneapi::mkl::blas::cublas::column_major::gerc, - oneapi::mkl::blas::cublas::column_major::geru, - oneapi::mkl::blas::cublas::column_major::geru, - oneapi::mkl::blas::cublas::column_major::hbmv, - oneapi::mkl::blas::cublas::column_major::hbmv, - oneapi::mkl::blas::cublas::column_major::hemv, - oneapi::mkl::blas::cublas::column_major::hemv, - oneapi::mkl::blas::cublas::column_major::her, - oneapi::mkl::blas::cublas::column_major::her, - oneapi::mkl::blas::cublas::column_major::her2, - oneapi::mkl::blas::cublas::column_major::her2, - oneapi::mkl::blas::cublas::column_major::hpmv, - oneapi::mkl::blas::cublas::column_major::hpmv, - oneapi::mkl::blas::cublas::column_major::hpr, - oneapi::mkl::blas::cublas::column_major::hpr, - oneapi::mkl::blas::cublas::column_major::hpr2, - oneapi::mkl::blas::cublas::column_major::hpr2, - oneapi::mkl::blas::cublas::column_major::sbmv, - oneapi::mkl::blas::cublas::column_major::sbmv, - oneapi::mkl::blas::cublas::column_major::spmv, - oneapi::mkl::blas::cublas::column_major::spmv, - oneapi::mkl::blas::cublas::column_major::spr, - oneapi::mkl::blas::cublas::column_major::spr, - oneapi::mkl::blas::cublas::column_major::spr2, - oneapi::mkl::blas::cublas::column_major::spr2, - oneapi::mkl::blas::cublas::column_major::symv, - oneapi::mkl::blas::cublas::column_major::symv, - oneapi::mkl::blas::cublas::column_major::syr, - oneapi::mkl::blas::cublas::column_major::syr, - oneapi::mkl::blas::cublas::column_major::syr2, - oneapi::mkl::blas::cublas::column_major::syr2, - oneapi::mkl::blas::cublas::column_major::tbmv, - oneapi::mkl::blas::cublas::column_major::tbmv, - oneapi::mkl::blas::cublas::column_major::tbmv, - oneapi::mkl::blas::cublas::column_major::tbmv, - oneapi::mkl::blas::cublas::column_major::tbsv, - oneapi::mkl::blas::cublas::column_major::tbsv, - oneapi::mkl::blas::cublas::column_major::tbsv, - oneapi::mkl::blas::cublas::column_major::tbsv, - oneapi::mkl::blas::cublas::column_major::tpmv, - oneapi::mkl::blas::cublas::column_major::tpmv, - oneapi::mkl::blas::cublas::column_major::tpmv, - oneapi::mkl::blas::cublas::column_major::tpmv, - oneapi::mkl::blas::cublas::column_major::tpsv, - oneapi::mkl::blas::cublas::column_major::tpsv, - oneapi::mkl::blas::cublas::column_major::tpsv, - oneapi::mkl::blas::cublas::column_major::tpsv, - oneapi::mkl::blas::cublas::column_major::trmv, - oneapi::mkl::blas::cublas::column_major::trmv, - oneapi::mkl::blas::cublas::column_major::trmv, - oneapi::mkl::blas::cublas::column_major::trmv, - oneapi::mkl::blas::cublas::column_major::trsv, - oneapi::mkl::blas::cublas::column_major::trsv, - oneapi::mkl::blas::cublas::column_major::trsv, - oneapi::mkl::blas::cublas::column_major::trsv, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::hemm, - oneapi::mkl::blas::cublas::column_major::hemm, - oneapi::mkl::blas::cublas::column_major::herk, - oneapi::mkl::blas::cublas::column_major::herk, - oneapi::mkl::blas::cublas::column_major::her2k, - oneapi::mkl::blas::cublas::column_major::her2k, - oneapi::mkl::blas::cublas::column_major::symm, - oneapi::mkl::blas::cublas::column_major::symm, - oneapi::mkl::blas::cublas::column_major::symm, - oneapi::mkl::blas::cublas::column_major::symm, - oneapi::mkl::blas::cublas::column_major::syrk, - oneapi::mkl::blas::cublas::column_major::syrk, - oneapi::mkl::blas::cublas::column_major::syrk, - oneapi::mkl::blas::cublas::column_major::syrk, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syr2k, - oneapi::mkl::blas::cublas::column_major::syr2k, - oneapi::mkl::blas::cublas::column_major::syr2k, - oneapi::mkl::blas::cublas::column_major::syr2k, - oneapi::mkl::blas::cublas::column_major::trmm, - oneapi::mkl::blas::cublas::column_major::trmm, - oneapi::mkl::blas::cublas::column_major::trmm, - oneapi::mkl::blas::cublas::column_major::trmm, - oneapi::mkl::blas::cublas::column_major::trsm, - oneapi::mkl::blas::cublas::column_major::trsm, - oneapi::mkl::blas::cublas::column_major::trsm, - oneapi::mkl::blas::cublas::column_major::trsm, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::gemmt, - oneapi::mkl::blas::cublas::column_major::gemmt, - oneapi::mkl::blas::cublas::column_major::gemmt, - oneapi::mkl::blas::cublas::column_major::gemmt, - oneapi::mkl::blas::cublas::column_major::gemm_bias, - oneapi::mkl::blas::cublas::column_major::gemm_bias, - oneapi::mkl::blas::cublas::column_major::gemm_bias, - oneapi::mkl::blas::cublas::column_major::gemm_bias, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::column_major::omatadd_batch, - oneapi::mkl::blas::cublas::column_major::omatadd_batch, - oneapi::mkl::blas::cublas::column_major::omatadd_batch, - oneapi::mkl::blas::cublas::column_major::omatadd_batch, - oneapi::mkl::blas::cublas::column_major::omatcopy, - oneapi::mkl::blas::cublas::column_major::omatcopy, - oneapi::mkl::blas::cublas::column_major::omatcopy, - oneapi::mkl::blas::cublas::column_major::omatcopy, - oneapi::mkl::blas::cublas::column_major::omatcopy2, - oneapi::mkl::blas::cublas::column_major::omatcopy2, - oneapi::mkl::blas::cublas::column_major::omatcopy2, - oneapi::mkl::blas::cublas::column_major::omatcopy2, - oneapi::mkl::blas::cublas::column_major::imatcopy, - oneapi::mkl::blas::cublas::column_major::imatcopy, - oneapi::mkl::blas::cublas::column_major::imatcopy, - oneapi::mkl::blas::cublas::column_major::imatcopy, - oneapi::mkl::blas::cublas::column_major::omatadd, - oneapi::mkl::blas::cublas::column_major::omatadd, - oneapi::mkl::blas::cublas::column_major::omatadd, - oneapi::mkl::blas::cublas::column_major::omatadd, - oneapi::mkl::blas::cublas::column_major::asum, - oneapi::mkl::blas::cublas::column_major::asum, - oneapi::mkl::blas::cublas::column_major::asum, - oneapi::mkl::blas::cublas::column_major::asum, - oneapi::mkl::blas::cublas::column_major::axpy, - oneapi::mkl::blas::cublas::column_major::axpy, - oneapi::mkl::blas::cublas::column_major::axpy, - oneapi::mkl::blas::cublas::column_major::axpy, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpy_batch, - oneapi::mkl::blas::cublas::column_major::axpby, - oneapi::mkl::blas::cublas::column_major::axpby, - oneapi::mkl::blas::cublas::column_major::axpby, - oneapi::mkl::blas::cublas::column_major::axpby, - oneapi::mkl::blas::cublas::column_major::copy, - oneapi::mkl::blas::cublas::column_major::copy, - oneapi::mkl::blas::cublas::column_major::copy, - oneapi::mkl::blas::cublas::column_major::copy, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::copy_batch, - oneapi::mkl::blas::cublas::column_major::dot, - oneapi::mkl::blas::cublas::column_major::dot, - oneapi::mkl::blas::cublas::column_major::dot, - oneapi::mkl::blas::cublas::column_major::dotc, - oneapi::mkl::blas::cublas::column_major::dotc, - oneapi::mkl::blas::cublas::column_major::dotu, - oneapi::mkl::blas::cublas::column_major::dotu, - oneapi::mkl::blas::cublas::column_major::iamin, - oneapi::mkl::blas::cublas::column_major::iamin, - oneapi::mkl::blas::cublas::column_major::iamin, - oneapi::mkl::blas::cublas::column_major::iamin, - oneapi::mkl::blas::cublas::column_major::iamax, - oneapi::mkl::blas::cublas::column_major::iamax, - oneapi::mkl::blas::cublas::column_major::iamax, - oneapi::mkl::blas::cublas::column_major::iamax, - oneapi::mkl::blas::cublas::column_major::nrm2, - oneapi::mkl::blas::cublas::column_major::nrm2, - oneapi::mkl::blas::cublas::column_major::nrm2, - oneapi::mkl::blas::cublas::column_major::nrm2, - oneapi::mkl::blas::cublas::column_major::rot, - oneapi::mkl::blas::cublas::column_major::rot, - oneapi::mkl::blas::cublas::column_major::rot, - oneapi::mkl::blas::cublas::column_major::rot, - oneapi::mkl::blas::cublas::column_major::rotg, - oneapi::mkl::blas::cublas::column_major::rotg, - oneapi::mkl::blas::cublas::column_major::rotg, - oneapi::mkl::blas::cublas::column_major::rotg, - oneapi::mkl::blas::cublas::column_major::rotm, - oneapi::mkl::blas::cublas::column_major::rotm, - oneapi::mkl::blas::cublas::column_major::rotmg, - oneapi::mkl::blas::cublas::column_major::rotmg, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::scal, - oneapi::mkl::blas::cublas::column_major::sdsdot, - oneapi::mkl::blas::cublas::column_major::swap, - oneapi::mkl::blas::cublas::column_major::swap, - oneapi::mkl::blas::cublas::column_major::swap, - oneapi::mkl::blas::cublas::column_major::swap, - oneapi::mkl::blas::cublas::column_major::gbmv, - oneapi::mkl::blas::cublas::column_major::gbmv, - oneapi::mkl::blas::cublas::column_major::gbmv, - oneapi::mkl::blas::cublas::column_major::gbmv, - oneapi::mkl::blas::cublas::column_major::gemv, - oneapi::mkl::blas::cublas::column_major::gemv, - oneapi::mkl::blas::cublas::column_major::gemv, - oneapi::mkl::blas::cublas::column_major::gemv, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::gemv_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::dgmm_batch, - oneapi::mkl::blas::cublas::column_major::ger, - oneapi::mkl::blas::cublas::column_major::ger, - oneapi::mkl::blas::cublas::column_major::gerc, - oneapi::mkl::blas::cublas::column_major::gerc, - oneapi::mkl::blas::cublas::column_major::geru, - oneapi::mkl::blas::cublas::column_major::geru, - oneapi::mkl::blas::cublas::column_major::hbmv, - oneapi::mkl::blas::cublas::column_major::hbmv, - oneapi::mkl::blas::cublas::column_major::hemv, - oneapi::mkl::blas::cublas::column_major::hemv, - oneapi::mkl::blas::cublas::column_major::her, - oneapi::mkl::blas::cublas::column_major::her, - oneapi::mkl::blas::cublas::column_major::her2, - oneapi::mkl::blas::cublas::column_major::her2, - oneapi::mkl::blas::cublas::column_major::hpmv, - oneapi::mkl::blas::cublas::column_major::hpmv, - oneapi::mkl::blas::cublas::column_major::hpr, - oneapi::mkl::blas::cublas::column_major::hpr, - oneapi::mkl::blas::cublas::column_major::hpr2, - oneapi::mkl::blas::cublas::column_major::hpr2, - oneapi::mkl::blas::cublas::column_major::sbmv, - oneapi::mkl::blas::cublas::column_major::sbmv, - oneapi::mkl::blas::cublas::column_major::spmv, - oneapi::mkl::blas::cublas::column_major::spmv, - oneapi::mkl::blas::cublas::column_major::spr, - oneapi::mkl::blas::cublas::column_major::spr, - oneapi::mkl::blas::cublas::column_major::spr2, - oneapi::mkl::blas::cublas::column_major::spr2, - oneapi::mkl::blas::cublas::column_major::symv, - oneapi::mkl::blas::cublas::column_major::symv, - oneapi::mkl::blas::cublas::column_major::syr, - oneapi::mkl::blas::cublas::column_major::syr, - oneapi::mkl::blas::cublas::column_major::syr2, - oneapi::mkl::blas::cublas::column_major::syr2, - oneapi::mkl::blas::cublas::column_major::tbmv, - oneapi::mkl::blas::cublas::column_major::tbmv, - oneapi::mkl::blas::cublas::column_major::tbmv, - oneapi::mkl::blas::cublas::column_major::tbmv, - oneapi::mkl::blas::cublas::column_major::tbsv, - oneapi::mkl::blas::cublas::column_major::tbsv, - oneapi::mkl::blas::cublas::column_major::tbsv, - oneapi::mkl::blas::cublas::column_major::tbsv, - oneapi::mkl::blas::cublas::column_major::tpmv, - oneapi::mkl::blas::cublas::column_major::tpmv, - oneapi::mkl::blas::cublas::column_major::tpmv, - oneapi::mkl::blas::cublas::column_major::tpmv, - oneapi::mkl::blas::cublas::column_major::tpsv, - oneapi::mkl::blas::cublas::column_major::tpsv, - oneapi::mkl::blas::cublas::column_major::tpsv, - oneapi::mkl::blas::cublas::column_major::tpsv, - oneapi::mkl::blas::cublas::column_major::trmv, - oneapi::mkl::blas::cublas::column_major::trmv, - oneapi::mkl::blas::cublas::column_major::trmv, - oneapi::mkl::blas::cublas::column_major::trmv, - oneapi::mkl::blas::cublas::column_major::trsv, - oneapi::mkl::blas::cublas::column_major::trsv, - oneapi::mkl::blas::cublas::column_major::trsv, - oneapi::mkl::blas::cublas::column_major::trsv, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::gemm, - oneapi::mkl::blas::cublas::column_major::hemm, - oneapi::mkl::blas::cublas::column_major::hemm, - oneapi::mkl::blas::cublas::column_major::herk, - oneapi::mkl::blas::cublas::column_major::herk, - oneapi::mkl::blas::cublas::column_major::her2k, - oneapi::mkl::blas::cublas::column_major::her2k, - oneapi::mkl::blas::cublas::column_major::symm, - oneapi::mkl::blas::cublas::column_major::symm, - oneapi::mkl::blas::cublas::column_major::symm, - oneapi::mkl::blas::cublas::column_major::symm, - oneapi::mkl::blas::cublas::column_major::syrk, - oneapi::mkl::blas::cublas::column_major::syrk, - oneapi::mkl::blas::cublas::column_major::syrk, - oneapi::mkl::blas::cublas::column_major::syrk, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syrk_batch, - oneapi::mkl::blas::cublas::column_major::syr2k, - oneapi::mkl::blas::cublas::column_major::syr2k, - oneapi::mkl::blas::cublas::column_major::syr2k, - oneapi::mkl::blas::cublas::column_major::syr2k, - oneapi::mkl::blas::cublas::column_major::trmm, - oneapi::mkl::blas::cublas::column_major::trmm, - oneapi::mkl::blas::cublas::column_major::trmm, - oneapi::mkl::blas::cublas::column_major::trmm, - oneapi::mkl::blas::cublas::column_major::trsm, - oneapi::mkl::blas::cublas::column_major::trsm, - oneapi::mkl::blas::cublas::column_major::trsm, - oneapi::mkl::blas::cublas::column_major::trsm, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::trsm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemm_batch, - oneapi::mkl::blas::cublas::column_major::gemmt, - oneapi::mkl::blas::cublas::column_major::gemmt, - oneapi::mkl::blas::cublas::column_major::gemmt, - oneapi::mkl::blas::cublas::column_major::gemmt, - oneapi::mkl::blas::cublas::column_major::gemm_bias, - oneapi::mkl::blas::cublas::column_major::gemm_bias, - oneapi::mkl::blas::cublas::column_major::gemm_bias, - oneapi::mkl::blas::cublas::column_major::gemm_bias, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::column_major::omatadd_batch, - oneapi::mkl::blas::cublas::column_major::omatadd_batch, - oneapi::mkl::blas::cublas::column_major::omatadd_batch, - oneapi::mkl::blas::cublas::column_major::omatadd_batch, - oneapi::mkl::blas::cublas::column_major::omatcopy, - oneapi::mkl::blas::cublas::column_major::omatcopy, - oneapi::mkl::blas::cublas::column_major::omatcopy, - oneapi::mkl::blas::cublas::column_major::omatcopy, - oneapi::mkl::blas::cublas::column_major::omatcopy2, - oneapi::mkl::blas::cublas::column_major::omatcopy2, - oneapi::mkl::blas::cublas::column_major::omatcopy2, - oneapi::mkl::blas::cublas::column_major::omatcopy2, - oneapi::mkl::blas::cublas::column_major::imatcopy, - oneapi::mkl::blas::cublas::column_major::imatcopy, - oneapi::mkl::blas::cublas::column_major::imatcopy, - oneapi::mkl::blas::cublas::column_major::imatcopy, - oneapi::mkl::blas::cublas::column_major::omatadd, - oneapi::mkl::blas::cublas::column_major::omatadd, - oneapi::mkl::blas::cublas::column_major::omatadd, - oneapi::mkl::blas::cublas::column_major::omatadd, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::omatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::column_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::asum, - oneapi::mkl::blas::cublas::row_major::asum, - oneapi::mkl::blas::cublas::row_major::asum, - oneapi::mkl::blas::cublas::row_major::asum, - oneapi::mkl::blas::cublas::row_major::axpy, - oneapi::mkl::blas::cublas::row_major::axpy, - oneapi::mkl::blas::cublas::row_major::axpy, - oneapi::mkl::blas::cublas::row_major::axpy, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpby, - oneapi::mkl::blas::cublas::row_major::axpby, - oneapi::mkl::blas::cublas::row_major::axpby, - oneapi::mkl::blas::cublas::row_major::axpby, - oneapi::mkl::blas::cublas::row_major::copy, - oneapi::mkl::blas::cublas::row_major::copy, - oneapi::mkl::blas::cublas::row_major::copy, - oneapi::mkl::blas::cublas::row_major::copy, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::dot, - oneapi::mkl::blas::cublas::row_major::dot, - oneapi::mkl::blas::cublas::row_major::dot, - oneapi::mkl::blas::cublas::row_major::dotc, - oneapi::mkl::blas::cublas::row_major::dotc, - oneapi::mkl::blas::cublas::row_major::dotu, - oneapi::mkl::blas::cublas::row_major::dotu, - oneapi::mkl::blas::cublas::row_major::iamin, - oneapi::mkl::blas::cublas::row_major::iamin, - oneapi::mkl::blas::cublas::row_major::iamin, - oneapi::mkl::blas::cublas::row_major::iamin, - oneapi::mkl::blas::cublas::row_major::iamax, - oneapi::mkl::blas::cublas::row_major::iamax, - oneapi::mkl::blas::cublas::row_major::iamax, - oneapi::mkl::blas::cublas::row_major::iamax, - oneapi::mkl::blas::cublas::row_major::nrm2, - oneapi::mkl::blas::cublas::row_major::nrm2, - oneapi::mkl::blas::cublas::row_major::nrm2, - oneapi::mkl::blas::cublas::row_major::nrm2, - oneapi::mkl::blas::cublas::row_major::rot, - oneapi::mkl::blas::cublas::row_major::rot, - oneapi::mkl::blas::cublas::row_major::rot, - oneapi::mkl::blas::cublas::row_major::rot, - oneapi::mkl::blas::cublas::row_major::rotg, - oneapi::mkl::blas::cublas::row_major::rotg, - oneapi::mkl::blas::cublas::row_major::rotg, - oneapi::mkl::blas::cublas::row_major::rotg, - oneapi::mkl::blas::cublas::row_major::rotm, - oneapi::mkl::blas::cublas::row_major::rotm, - oneapi::mkl::blas::cublas::row_major::rotmg, - oneapi::mkl::blas::cublas::row_major::rotmg, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::sdsdot, - oneapi::mkl::blas::cublas::row_major::swap, - oneapi::mkl::blas::cublas::row_major::swap, - oneapi::mkl::blas::cublas::row_major::swap, - oneapi::mkl::blas::cublas::row_major::swap, - oneapi::mkl::blas::cublas::row_major::gbmv, - oneapi::mkl::blas::cublas::row_major::gbmv, - oneapi::mkl::blas::cublas::row_major::gbmv, - oneapi::mkl::blas::cublas::row_major::gbmv, - oneapi::mkl::blas::cublas::row_major::gemv, - oneapi::mkl::blas::cublas::row_major::gemv, - oneapi::mkl::blas::cublas::row_major::gemv, - oneapi::mkl::blas::cublas::row_major::gemv, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::ger, - oneapi::mkl::blas::cublas::row_major::ger, - oneapi::mkl::blas::cublas::row_major::gerc, - oneapi::mkl::blas::cublas::row_major::gerc, - oneapi::mkl::blas::cublas::row_major::geru, - oneapi::mkl::blas::cublas::row_major::geru, - oneapi::mkl::blas::cublas::row_major::hbmv, - oneapi::mkl::blas::cublas::row_major::hbmv, - oneapi::mkl::blas::cublas::row_major::hemv, - oneapi::mkl::blas::cublas::row_major::hemv, - oneapi::mkl::blas::cublas::row_major::her, - oneapi::mkl::blas::cublas::row_major::her, - oneapi::mkl::blas::cublas::row_major::her2, - oneapi::mkl::blas::cublas::row_major::her2, - oneapi::mkl::blas::cublas::row_major::hpmv, - oneapi::mkl::blas::cublas::row_major::hpmv, - oneapi::mkl::blas::cublas::row_major::hpr, - oneapi::mkl::blas::cublas::row_major::hpr, - oneapi::mkl::blas::cublas::row_major::hpr2, - oneapi::mkl::blas::cublas::row_major::hpr2, - oneapi::mkl::blas::cublas::row_major::sbmv, - oneapi::mkl::blas::cublas::row_major::sbmv, - oneapi::mkl::blas::cublas::row_major::spmv, - oneapi::mkl::blas::cublas::row_major::spmv, - oneapi::mkl::blas::cublas::row_major::spr, - oneapi::mkl::blas::cublas::row_major::spr, - oneapi::mkl::blas::cublas::row_major::spr2, - oneapi::mkl::blas::cublas::row_major::spr2, - oneapi::mkl::blas::cublas::row_major::symv, - oneapi::mkl::blas::cublas::row_major::symv, - oneapi::mkl::blas::cublas::row_major::syr, - oneapi::mkl::blas::cublas::row_major::syr, - oneapi::mkl::blas::cublas::row_major::syr2, - oneapi::mkl::blas::cublas::row_major::syr2, - oneapi::mkl::blas::cublas::row_major::tbmv, - oneapi::mkl::blas::cublas::row_major::tbmv, - oneapi::mkl::blas::cublas::row_major::tbmv, - oneapi::mkl::blas::cublas::row_major::tbmv, - oneapi::mkl::blas::cublas::row_major::tbsv, - oneapi::mkl::blas::cublas::row_major::tbsv, - oneapi::mkl::blas::cublas::row_major::tbsv, - oneapi::mkl::blas::cublas::row_major::tbsv, - oneapi::mkl::blas::cublas::row_major::tpmv, - oneapi::mkl::blas::cublas::row_major::tpmv, - oneapi::mkl::blas::cublas::row_major::tpmv, - oneapi::mkl::blas::cublas::row_major::tpmv, - oneapi::mkl::blas::cublas::row_major::tpsv, - oneapi::mkl::blas::cublas::row_major::tpsv, - oneapi::mkl::blas::cublas::row_major::tpsv, - oneapi::mkl::blas::cublas::row_major::tpsv, - oneapi::mkl::blas::cublas::row_major::trmv, - oneapi::mkl::blas::cublas::row_major::trmv, - oneapi::mkl::blas::cublas::row_major::trmv, - oneapi::mkl::blas::cublas::row_major::trmv, - oneapi::mkl::blas::cublas::row_major::trsv, - oneapi::mkl::blas::cublas::row_major::trsv, - oneapi::mkl::blas::cublas::row_major::trsv, - oneapi::mkl::blas::cublas::row_major::trsv, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::hemm, - oneapi::mkl::blas::cublas::row_major::hemm, - oneapi::mkl::blas::cublas::row_major::herk, - oneapi::mkl::blas::cublas::row_major::herk, - oneapi::mkl::blas::cublas::row_major::her2k, - oneapi::mkl::blas::cublas::row_major::her2k, - oneapi::mkl::blas::cublas::row_major::symm, - oneapi::mkl::blas::cublas::row_major::symm, - oneapi::mkl::blas::cublas::row_major::symm, - oneapi::mkl::blas::cublas::row_major::symm, - oneapi::mkl::blas::cublas::row_major::syrk, - oneapi::mkl::blas::cublas::row_major::syrk, - oneapi::mkl::blas::cublas::row_major::syrk, - oneapi::mkl::blas::cublas::row_major::syrk, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syr2k, - oneapi::mkl::blas::cublas::row_major::syr2k, - oneapi::mkl::blas::cublas::row_major::syr2k, - oneapi::mkl::blas::cublas::row_major::syr2k, - oneapi::mkl::blas::cublas::row_major::trmm, - oneapi::mkl::blas::cublas::row_major::trmm, - oneapi::mkl::blas::cublas::row_major::trmm, - oneapi::mkl::blas::cublas::row_major::trmm, - oneapi::mkl::blas::cublas::row_major::trsm, - oneapi::mkl::blas::cublas::row_major::trsm, - oneapi::mkl::blas::cublas::row_major::trsm, - oneapi::mkl::blas::cublas::row_major::trsm, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::gemmt, - oneapi::mkl::blas::cublas::row_major::gemmt, - oneapi::mkl::blas::cublas::row_major::gemmt, - oneapi::mkl::blas::cublas::row_major::gemmt, - oneapi::mkl::blas::cublas::row_major::gemm_bias, - oneapi::mkl::blas::cublas::row_major::gemm_bias, - oneapi::mkl::blas::cublas::row_major::gemm_bias, - oneapi::mkl::blas::cublas::row_major::gemm_bias, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::omatadd_batch, - oneapi::mkl::blas::cublas::row_major::omatadd_batch, - oneapi::mkl::blas::cublas::row_major::omatadd_batch, - oneapi::mkl::blas::cublas::row_major::omatadd_batch, - oneapi::mkl::blas::cublas::row_major::omatcopy, - oneapi::mkl::blas::cublas::row_major::omatcopy, - oneapi::mkl::blas::cublas::row_major::omatcopy, - oneapi::mkl::blas::cublas::row_major::omatcopy, - oneapi::mkl::blas::cublas::row_major::omatcopy2, - oneapi::mkl::blas::cublas::row_major::omatcopy2, - oneapi::mkl::blas::cublas::row_major::omatcopy2, - oneapi::mkl::blas::cublas::row_major::omatcopy2, - oneapi::mkl::blas::cublas::row_major::imatcopy, - oneapi::mkl::blas::cublas::row_major::imatcopy, - oneapi::mkl::blas::cublas::row_major::imatcopy, - oneapi::mkl::blas::cublas::row_major::imatcopy, - oneapi::mkl::blas::cublas::row_major::omatadd, - oneapi::mkl::blas::cublas::row_major::omatadd, - oneapi::mkl::blas::cublas::row_major::omatadd, - oneapi::mkl::blas::cublas::row_major::omatadd, - oneapi::mkl::blas::cublas::row_major::asum, - oneapi::mkl::blas::cublas::row_major::asum, - oneapi::mkl::blas::cublas::row_major::asum, - oneapi::mkl::blas::cublas::row_major::asum, - oneapi::mkl::blas::cublas::row_major::axpy, - oneapi::mkl::blas::cublas::row_major::axpy, - oneapi::mkl::blas::cublas::row_major::axpy, - oneapi::mkl::blas::cublas::row_major::axpy, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpy_batch, - oneapi::mkl::blas::cublas::row_major::axpby, - oneapi::mkl::blas::cublas::row_major::axpby, - oneapi::mkl::blas::cublas::row_major::axpby, - oneapi::mkl::blas::cublas::row_major::axpby, - oneapi::mkl::blas::cublas::row_major::copy, - oneapi::mkl::blas::cublas::row_major::copy, - oneapi::mkl::blas::cublas::row_major::copy, - oneapi::mkl::blas::cublas::row_major::copy, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::copy_batch, - oneapi::mkl::blas::cublas::row_major::dot, - oneapi::mkl::blas::cublas::row_major::dot, - oneapi::mkl::blas::cublas::row_major::dot, - oneapi::mkl::blas::cublas::row_major::dotc, - oneapi::mkl::blas::cublas::row_major::dotc, - oneapi::mkl::blas::cublas::row_major::dotu, - oneapi::mkl::blas::cublas::row_major::dotu, - oneapi::mkl::blas::cublas::row_major::iamin, - oneapi::mkl::blas::cublas::row_major::iamin, - oneapi::mkl::blas::cublas::row_major::iamin, - oneapi::mkl::blas::cublas::row_major::iamin, - oneapi::mkl::blas::cublas::row_major::iamax, - oneapi::mkl::blas::cublas::row_major::iamax, - oneapi::mkl::blas::cublas::row_major::iamax, - oneapi::mkl::blas::cublas::row_major::iamax, - oneapi::mkl::blas::cublas::row_major::nrm2, - oneapi::mkl::blas::cublas::row_major::nrm2, - oneapi::mkl::blas::cublas::row_major::nrm2, - oneapi::mkl::blas::cublas::row_major::nrm2, - oneapi::mkl::blas::cublas::row_major::rot, - oneapi::mkl::blas::cublas::row_major::rot, - oneapi::mkl::blas::cublas::row_major::rot, - oneapi::mkl::blas::cublas::row_major::rot, - oneapi::mkl::blas::cublas::row_major::rotg, - oneapi::mkl::blas::cublas::row_major::rotg, - oneapi::mkl::blas::cublas::row_major::rotg, - oneapi::mkl::blas::cublas::row_major::rotg, - oneapi::mkl::blas::cublas::row_major::rotm, - oneapi::mkl::blas::cublas::row_major::rotm, - oneapi::mkl::blas::cublas::row_major::rotmg, - oneapi::mkl::blas::cublas::row_major::rotmg, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::scal, - oneapi::mkl::blas::cublas::row_major::sdsdot, - oneapi::mkl::blas::cublas::row_major::swap, - oneapi::mkl::blas::cublas::row_major::swap, - oneapi::mkl::blas::cublas::row_major::swap, - oneapi::mkl::blas::cublas::row_major::swap, - oneapi::mkl::blas::cublas::row_major::gbmv, - oneapi::mkl::blas::cublas::row_major::gbmv, - oneapi::mkl::blas::cublas::row_major::gbmv, - oneapi::mkl::blas::cublas::row_major::gbmv, - oneapi::mkl::blas::cublas::row_major::gemv, - oneapi::mkl::blas::cublas::row_major::gemv, - oneapi::mkl::blas::cublas::row_major::gemv, - oneapi::mkl::blas::cublas::row_major::gemv, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::gemv_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::dgmm_batch, - oneapi::mkl::blas::cublas::row_major::ger, - oneapi::mkl::blas::cublas::row_major::ger, - oneapi::mkl::blas::cublas::row_major::gerc, - oneapi::mkl::blas::cublas::row_major::gerc, - oneapi::mkl::blas::cublas::row_major::geru, - oneapi::mkl::blas::cublas::row_major::geru, - oneapi::mkl::blas::cublas::row_major::hbmv, - oneapi::mkl::blas::cublas::row_major::hbmv, - oneapi::mkl::blas::cublas::row_major::hemv, - oneapi::mkl::blas::cublas::row_major::hemv, - oneapi::mkl::blas::cublas::row_major::her, - oneapi::mkl::blas::cublas::row_major::her, - oneapi::mkl::blas::cublas::row_major::her2, - oneapi::mkl::blas::cublas::row_major::her2, - oneapi::mkl::blas::cublas::row_major::hpmv, - oneapi::mkl::blas::cublas::row_major::hpmv, - oneapi::mkl::blas::cublas::row_major::hpr, - oneapi::mkl::blas::cublas::row_major::hpr, - oneapi::mkl::blas::cublas::row_major::hpr2, - oneapi::mkl::blas::cublas::row_major::hpr2, - oneapi::mkl::blas::cublas::row_major::sbmv, - oneapi::mkl::blas::cublas::row_major::sbmv, - oneapi::mkl::blas::cublas::row_major::spmv, - oneapi::mkl::blas::cublas::row_major::spmv, - oneapi::mkl::blas::cublas::row_major::spr, - oneapi::mkl::blas::cublas::row_major::spr, - oneapi::mkl::blas::cublas::row_major::spr2, - oneapi::mkl::blas::cublas::row_major::spr2, - oneapi::mkl::blas::cublas::row_major::symv, - oneapi::mkl::blas::cublas::row_major::symv, - oneapi::mkl::blas::cublas::row_major::syr, - oneapi::mkl::blas::cublas::row_major::syr, - oneapi::mkl::blas::cublas::row_major::syr2, - oneapi::mkl::blas::cublas::row_major::syr2, - oneapi::mkl::blas::cublas::row_major::tbmv, - oneapi::mkl::blas::cublas::row_major::tbmv, - oneapi::mkl::blas::cublas::row_major::tbmv, - oneapi::mkl::blas::cublas::row_major::tbmv, - oneapi::mkl::blas::cublas::row_major::tbsv, - oneapi::mkl::blas::cublas::row_major::tbsv, - oneapi::mkl::blas::cublas::row_major::tbsv, - oneapi::mkl::blas::cublas::row_major::tbsv, - oneapi::mkl::blas::cublas::row_major::tpmv, - oneapi::mkl::blas::cublas::row_major::tpmv, - oneapi::mkl::blas::cublas::row_major::tpmv, - oneapi::mkl::blas::cublas::row_major::tpmv, - oneapi::mkl::blas::cublas::row_major::tpsv, - oneapi::mkl::blas::cublas::row_major::tpsv, - oneapi::mkl::blas::cublas::row_major::tpsv, - oneapi::mkl::blas::cublas::row_major::tpsv, - oneapi::mkl::blas::cublas::row_major::trmv, - oneapi::mkl::blas::cublas::row_major::trmv, - oneapi::mkl::blas::cublas::row_major::trmv, - oneapi::mkl::blas::cublas::row_major::trmv, - oneapi::mkl::blas::cublas::row_major::trsv, - oneapi::mkl::blas::cublas::row_major::trsv, - oneapi::mkl::blas::cublas::row_major::trsv, - oneapi::mkl::blas::cublas::row_major::trsv, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::gemm, - oneapi::mkl::blas::cublas::row_major::hemm, - oneapi::mkl::blas::cublas::row_major::hemm, - oneapi::mkl::blas::cublas::row_major::herk, - oneapi::mkl::blas::cublas::row_major::herk, - oneapi::mkl::blas::cublas::row_major::her2k, - oneapi::mkl::blas::cublas::row_major::her2k, - oneapi::mkl::blas::cublas::row_major::symm, - oneapi::mkl::blas::cublas::row_major::symm, - oneapi::mkl::blas::cublas::row_major::symm, - oneapi::mkl::blas::cublas::row_major::symm, - oneapi::mkl::blas::cublas::row_major::syrk, - oneapi::mkl::blas::cublas::row_major::syrk, - oneapi::mkl::blas::cublas::row_major::syrk, - oneapi::mkl::blas::cublas::row_major::syrk, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syrk_batch, - oneapi::mkl::blas::cublas::row_major::syr2k, - oneapi::mkl::blas::cublas::row_major::syr2k, - oneapi::mkl::blas::cublas::row_major::syr2k, - oneapi::mkl::blas::cublas::row_major::syr2k, - oneapi::mkl::blas::cublas::row_major::trmm, - oneapi::mkl::blas::cublas::row_major::trmm, - oneapi::mkl::blas::cublas::row_major::trmm, - oneapi::mkl::blas::cublas::row_major::trmm, - oneapi::mkl::blas::cublas::row_major::trsm, - oneapi::mkl::blas::cublas::row_major::trsm, - oneapi::mkl::blas::cublas::row_major::trsm, - oneapi::mkl::blas::cublas::row_major::trsm, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::trsm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemm_batch, - oneapi::mkl::blas::cublas::row_major::gemmt, - oneapi::mkl::blas::cublas::row_major::gemmt, - oneapi::mkl::blas::cublas::row_major::gemmt, - oneapi::mkl::blas::cublas::row_major::gemmt, - oneapi::mkl::blas::cublas::row_major::gemm_bias, - oneapi::mkl::blas::cublas::row_major::gemm_bias, - oneapi::mkl::blas::cublas::row_major::gemm_bias, - oneapi::mkl::blas::cublas::row_major::gemm_bias, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::omatadd_batch, - oneapi::mkl::blas::cublas::row_major::omatadd_batch, - oneapi::mkl::blas::cublas::row_major::omatadd_batch, - oneapi::mkl::blas::cublas::row_major::omatadd_batch, - oneapi::mkl::blas::cublas::row_major::omatcopy, - oneapi::mkl::blas::cublas::row_major::omatcopy, - oneapi::mkl::blas::cublas::row_major::omatcopy, - oneapi::mkl::blas::cublas::row_major::omatcopy, - oneapi::mkl::blas::cublas::row_major::omatcopy2, - oneapi::mkl::blas::cublas::row_major::omatcopy2, - oneapi::mkl::blas::cublas::row_major::omatcopy2, - oneapi::mkl::blas::cublas::row_major::omatcopy2, - oneapi::mkl::blas::cublas::row_major::imatcopy, - oneapi::mkl::blas::cublas::row_major::imatcopy, - oneapi::mkl::blas::cublas::row_major::imatcopy, - oneapi::mkl::blas::cublas::row_major::imatcopy, - oneapi::mkl::blas::cublas::row_major::omatadd, - oneapi::mkl::blas::cublas::row_major::omatadd, - oneapi::mkl::blas::cublas::row_major::omatadd, - oneapi::mkl::blas::cublas::row_major::omatadd, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::omatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, - oneapi::mkl::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::asum, + oneapi::math::blas::cublas::column_major::asum, + oneapi::math::blas::cublas::column_major::asum, + oneapi::math::blas::cublas::column_major::asum, + oneapi::math::blas::cublas::column_major::axpy, + oneapi::math::blas::cublas::column_major::axpy, + oneapi::math::blas::cublas::column_major::axpy, + oneapi::math::blas::cublas::column_major::axpy, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpby, + oneapi::math::blas::cublas::column_major::axpby, + oneapi::math::blas::cublas::column_major::axpby, + oneapi::math::blas::cublas::column_major::axpby, + oneapi::math::blas::cublas::column_major::copy, + oneapi::math::blas::cublas::column_major::copy, + oneapi::math::blas::cublas::column_major::copy, + oneapi::math::blas::cublas::column_major::copy, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::dot, + oneapi::math::blas::cublas::column_major::dot, + oneapi::math::blas::cublas::column_major::dot, + oneapi::math::blas::cublas::column_major::dotc, + oneapi::math::blas::cublas::column_major::dotc, + oneapi::math::blas::cublas::column_major::dotu, + oneapi::math::blas::cublas::column_major::dotu, + oneapi::math::blas::cublas::column_major::iamin, + oneapi::math::blas::cublas::column_major::iamin, + oneapi::math::blas::cublas::column_major::iamin, + oneapi::math::blas::cublas::column_major::iamin, + oneapi::math::blas::cublas::column_major::iamax, + oneapi::math::blas::cublas::column_major::iamax, + oneapi::math::blas::cublas::column_major::iamax, + oneapi::math::blas::cublas::column_major::iamax, + oneapi::math::blas::cublas::column_major::nrm2, + oneapi::math::blas::cublas::column_major::nrm2, + oneapi::math::blas::cublas::column_major::nrm2, + oneapi::math::blas::cublas::column_major::nrm2, + oneapi::math::blas::cublas::column_major::rot, + oneapi::math::blas::cublas::column_major::rot, + oneapi::math::blas::cublas::column_major::rot, + oneapi::math::blas::cublas::column_major::rot, + oneapi::math::blas::cublas::column_major::rotg, + oneapi::math::blas::cublas::column_major::rotg, + oneapi::math::blas::cublas::column_major::rotg, + oneapi::math::blas::cublas::column_major::rotg, + oneapi::math::blas::cublas::column_major::rotm, + oneapi::math::blas::cublas::column_major::rotm, + oneapi::math::blas::cublas::column_major::rotmg, + oneapi::math::blas::cublas::column_major::rotmg, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::sdsdot, + oneapi::math::blas::cublas::column_major::swap, + oneapi::math::blas::cublas::column_major::swap, + oneapi::math::blas::cublas::column_major::swap, + oneapi::math::blas::cublas::column_major::swap, + oneapi::math::blas::cublas::column_major::gbmv, + oneapi::math::blas::cublas::column_major::gbmv, + oneapi::math::blas::cublas::column_major::gbmv, + oneapi::math::blas::cublas::column_major::gbmv, + oneapi::math::blas::cublas::column_major::gemv, + oneapi::math::blas::cublas::column_major::gemv, + oneapi::math::blas::cublas::column_major::gemv, + oneapi::math::blas::cublas::column_major::gemv, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::ger, + oneapi::math::blas::cublas::column_major::ger, + oneapi::math::blas::cublas::column_major::gerc, + oneapi::math::blas::cublas::column_major::gerc, + oneapi::math::blas::cublas::column_major::geru, + oneapi::math::blas::cublas::column_major::geru, + oneapi::math::blas::cublas::column_major::hbmv, + oneapi::math::blas::cublas::column_major::hbmv, + oneapi::math::blas::cublas::column_major::hemv, + oneapi::math::blas::cublas::column_major::hemv, + oneapi::math::blas::cublas::column_major::her, + oneapi::math::blas::cublas::column_major::her, + oneapi::math::blas::cublas::column_major::her2, + oneapi::math::blas::cublas::column_major::her2, + oneapi::math::blas::cublas::column_major::hpmv, + oneapi::math::blas::cublas::column_major::hpmv, + oneapi::math::blas::cublas::column_major::hpr, + oneapi::math::blas::cublas::column_major::hpr, + oneapi::math::blas::cublas::column_major::hpr2, + oneapi::math::blas::cublas::column_major::hpr2, + oneapi::math::blas::cublas::column_major::sbmv, + oneapi::math::blas::cublas::column_major::sbmv, + oneapi::math::blas::cublas::column_major::spmv, + oneapi::math::blas::cublas::column_major::spmv, + oneapi::math::blas::cublas::column_major::spr, + oneapi::math::blas::cublas::column_major::spr, + oneapi::math::blas::cublas::column_major::spr2, + oneapi::math::blas::cublas::column_major::spr2, + oneapi::math::blas::cublas::column_major::symv, + oneapi::math::blas::cublas::column_major::symv, + oneapi::math::blas::cublas::column_major::syr, + oneapi::math::blas::cublas::column_major::syr, + oneapi::math::blas::cublas::column_major::syr2, + oneapi::math::blas::cublas::column_major::syr2, + oneapi::math::blas::cublas::column_major::tbmv, + oneapi::math::blas::cublas::column_major::tbmv, + oneapi::math::blas::cublas::column_major::tbmv, + oneapi::math::blas::cublas::column_major::tbmv, + oneapi::math::blas::cublas::column_major::tbsv, + oneapi::math::blas::cublas::column_major::tbsv, + oneapi::math::blas::cublas::column_major::tbsv, + oneapi::math::blas::cublas::column_major::tbsv, + oneapi::math::blas::cublas::column_major::tpmv, + oneapi::math::blas::cublas::column_major::tpmv, + oneapi::math::blas::cublas::column_major::tpmv, + oneapi::math::blas::cublas::column_major::tpmv, + oneapi::math::blas::cublas::column_major::tpsv, + oneapi::math::blas::cublas::column_major::tpsv, + oneapi::math::blas::cublas::column_major::tpsv, + oneapi::math::blas::cublas::column_major::tpsv, + oneapi::math::blas::cublas::column_major::trmv, + oneapi::math::blas::cublas::column_major::trmv, + oneapi::math::blas::cublas::column_major::trmv, + oneapi::math::blas::cublas::column_major::trmv, + oneapi::math::blas::cublas::column_major::trsv, + oneapi::math::blas::cublas::column_major::trsv, + oneapi::math::blas::cublas::column_major::trsv, + oneapi::math::blas::cublas::column_major::trsv, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::hemm, + oneapi::math::blas::cublas::column_major::hemm, + oneapi::math::blas::cublas::column_major::herk, + oneapi::math::blas::cublas::column_major::herk, + oneapi::math::blas::cublas::column_major::her2k, + oneapi::math::blas::cublas::column_major::her2k, + oneapi::math::blas::cublas::column_major::symm, + oneapi::math::blas::cublas::column_major::symm, + oneapi::math::blas::cublas::column_major::symm, + oneapi::math::blas::cublas::column_major::symm, + oneapi::math::blas::cublas::column_major::syrk, + oneapi::math::blas::cublas::column_major::syrk, + oneapi::math::blas::cublas::column_major::syrk, + oneapi::math::blas::cublas::column_major::syrk, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syr2k, + oneapi::math::blas::cublas::column_major::syr2k, + oneapi::math::blas::cublas::column_major::syr2k, + oneapi::math::blas::cublas::column_major::syr2k, + oneapi::math::blas::cublas::column_major::trmm, + oneapi::math::blas::cublas::column_major::trmm, + oneapi::math::blas::cublas::column_major::trmm, + oneapi::math::blas::cublas::column_major::trmm, + oneapi::math::blas::cublas::column_major::trsm, + oneapi::math::blas::cublas::column_major::trsm, + oneapi::math::blas::cublas::column_major::trsm, + oneapi::math::blas::cublas::column_major::trsm, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::gemmt, + oneapi::math::blas::cublas::column_major::gemmt, + oneapi::math::blas::cublas::column_major::gemmt, + oneapi::math::blas::cublas::column_major::gemmt, + oneapi::math::blas::cublas::column_major::gemm_bias, + oneapi::math::blas::cublas::column_major::gemm_bias, + oneapi::math::blas::cublas::column_major::gemm_bias, + oneapi::math::blas::cublas::column_major::gemm_bias, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::omatadd_batch, + oneapi::math::blas::cublas::column_major::omatadd_batch, + oneapi::math::blas::cublas::column_major::omatadd_batch, + oneapi::math::blas::cublas::column_major::omatadd_batch, + oneapi::math::blas::cublas::column_major::omatcopy, + oneapi::math::blas::cublas::column_major::omatcopy, + oneapi::math::blas::cublas::column_major::omatcopy, + oneapi::math::blas::cublas::column_major::omatcopy, + oneapi::math::blas::cublas::column_major::omatcopy2, + oneapi::math::blas::cublas::column_major::omatcopy2, + oneapi::math::blas::cublas::column_major::omatcopy2, + oneapi::math::blas::cublas::column_major::omatcopy2, + oneapi::math::blas::cublas::column_major::imatcopy, + oneapi::math::blas::cublas::column_major::imatcopy, + oneapi::math::blas::cublas::column_major::imatcopy, + oneapi::math::blas::cublas::column_major::imatcopy, + oneapi::math::blas::cublas::column_major::omatadd, + oneapi::math::blas::cublas::column_major::omatadd, + oneapi::math::blas::cublas::column_major::omatadd, + oneapi::math::blas::cublas::column_major::omatadd, + oneapi::math::blas::cublas::column_major::asum, + oneapi::math::blas::cublas::column_major::asum, + oneapi::math::blas::cublas::column_major::asum, + oneapi::math::blas::cublas::column_major::asum, + oneapi::math::blas::cublas::column_major::axpy, + oneapi::math::blas::cublas::column_major::axpy, + oneapi::math::blas::cublas::column_major::axpy, + oneapi::math::blas::cublas::column_major::axpy, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpy_batch, + oneapi::math::blas::cublas::column_major::axpby, + oneapi::math::blas::cublas::column_major::axpby, + oneapi::math::blas::cublas::column_major::axpby, + oneapi::math::blas::cublas::column_major::axpby, + oneapi::math::blas::cublas::column_major::copy, + oneapi::math::blas::cublas::column_major::copy, + oneapi::math::blas::cublas::column_major::copy, + oneapi::math::blas::cublas::column_major::copy, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::copy_batch, + oneapi::math::blas::cublas::column_major::dot, + oneapi::math::blas::cublas::column_major::dot, + oneapi::math::blas::cublas::column_major::dot, + oneapi::math::blas::cublas::column_major::dotc, + oneapi::math::blas::cublas::column_major::dotc, + oneapi::math::blas::cublas::column_major::dotu, + oneapi::math::blas::cublas::column_major::dotu, + oneapi::math::blas::cublas::column_major::iamin, + oneapi::math::blas::cublas::column_major::iamin, + oneapi::math::blas::cublas::column_major::iamin, + oneapi::math::blas::cublas::column_major::iamin, + oneapi::math::blas::cublas::column_major::iamax, + oneapi::math::blas::cublas::column_major::iamax, + oneapi::math::blas::cublas::column_major::iamax, + oneapi::math::blas::cublas::column_major::iamax, + oneapi::math::blas::cublas::column_major::nrm2, + oneapi::math::blas::cublas::column_major::nrm2, + oneapi::math::blas::cublas::column_major::nrm2, + oneapi::math::blas::cublas::column_major::nrm2, + oneapi::math::blas::cublas::column_major::rot, + oneapi::math::blas::cublas::column_major::rot, + oneapi::math::blas::cublas::column_major::rot, + oneapi::math::blas::cublas::column_major::rot, + oneapi::math::blas::cublas::column_major::rotg, + oneapi::math::blas::cublas::column_major::rotg, + oneapi::math::blas::cublas::column_major::rotg, + oneapi::math::blas::cublas::column_major::rotg, + oneapi::math::blas::cublas::column_major::rotm, + oneapi::math::blas::cublas::column_major::rotm, + oneapi::math::blas::cublas::column_major::rotmg, + oneapi::math::blas::cublas::column_major::rotmg, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::scal, + oneapi::math::blas::cublas::column_major::sdsdot, + oneapi::math::blas::cublas::column_major::swap, + oneapi::math::blas::cublas::column_major::swap, + oneapi::math::blas::cublas::column_major::swap, + oneapi::math::blas::cublas::column_major::swap, + oneapi::math::blas::cublas::column_major::gbmv, + oneapi::math::blas::cublas::column_major::gbmv, + oneapi::math::blas::cublas::column_major::gbmv, + oneapi::math::blas::cublas::column_major::gbmv, + oneapi::math::blas::cublas::column_major::gemv, + oneapi::math::blas::cublas::column_major::gemv, + oneapi::math::blas::cublas::column_major::gemv, + oneapi::math::blas::cublas::column_major::gemv, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::gemv_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::dgmm_batch, + oneapi::math::blas::cublas::column_major::ger, + oneapi::math::blas::cublas::column_major::ger, + oneapi::math::blas::cublas::column_major::gerc, + oneapi::math::blas::cublas::column_major::gerc, + oneapi::math::blas::cublas::column_major::geru, + oneapi::math::blas::cublas::column_major::geru, + oneapi::math::blas::cublas::column_major::hbmv, + oneapi::math::blas::cublas::column_major::hbmv, + oneapi::math::blas::cublas::column_major::hemv, + oneapi::math::blas::cublas::column_major::hemv, + oneapi::math::blas::cublas::column_major::her, + oneapi::math::blas::cublas::column_major::her, + oneapi::math::blas::cublas::column_major::her2, + oneapi::math::blas::cublas::column_major::her2, + oneapi::math::blas::cublas::column_major::hpmv, + oneapi::math::blas::cublas::column_major::hpmv, + oneapi::math::blas::cublas::column_major::hpr, + oneapi::math::blas::cublas::column_major::hpr, + oneapi::math::blas::cublas::column_major::hpr2, + oneapi::math::blas::cublas::column_major::hpr2, + oneapi::math::blas::cublas::column_major::sbmv, + oneapi::math::blas::cublas::column_major::sbmv, + oneapi::math::blas::cublas::column_major::spmv, + oneapi::math::blas::cublas::column_major::spmv, + oneapi::math::blas::cublas::column_major::spr, + oneapi::math::blas::cublas::column_major::spr, + oneapi::math::blas::cublas::column_major::spr2, + oneapi::math::blas::cublas::column_major::spr2, + oneapi::math::blas::cublas::column_major::symv, + oneapi::math::blas::cublas::column_major::symv, + oneapi::math::blas::cublas::column_major::syr, + oneapi::math::blas::cublas::column_major::syr, + oneapi::math::blas::cublas::column_major::syr2, + oneapi::math::blas::cublas::column_major::syr2, + oneapi::math::blas::cublas::column_major::tbmv, + oneapi::math::blas::cublas::column_major::tbmv, + oneapi::math::blas::cublas::column_major::tbmv, + oneapi::math::blas::cublas::column_major::tbmv, + oneapi::math::blas::cublas::column_major::tbsv, + oneapi::math::blas::cublas::column_major::tbsv, + oneapi::math::blas::cublas::column_major::tbsv, + oneapi::math::blas::cublas::column_major::tbsv, + oneapi::math::blas::cublas::column_major::tpmv, + oneapi::math::blas::cublas::column_major::tpmv, + oneapi::math::blas::cublas::column_major::tpmv, + oneapi::math::blas::cublas::column_major::tpmv, + oneapi::math::blas::cublas::column_major::tpsv, + oneapi::math::blas::cublas::column_major::tpsv, + oneapi::math::blas::cublas::column_major::tpsv, + oneapi::math::blas::cublas::column_major::tpsv, + oneapi::math::blas::cublas::column_major::trmv, + oneapi::math::blas::cublas::column_major::trmv, + oneapi::math::blas::cublas::column_major::trmv, + oneapi::math::blas::cublas::column_major::trmv, + oneapi::math::blas::cublas::column_major::trsv, + oneapi::math::blas::cublas::column_major::trsv, + oneapi::math::blas::cublas::column_major::trsv, + oneapi::math::blas::cublas::column_major::trsv, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::gemm, + oneapi::math::blas::cublas::column_major::hemm, + oneapi::math::blas::cublas::column_major::hemm, + oneapi::math::blas::cublas::column_major::herk, + oneapi::math::blas::cublas::column_major::herk, + oneapi::math::blas::cublas::column_major::her2k, + oneapi::math::blas::cublas::column_major::her2k, + oneapi::math::blas::cublas::column_major::symm, + oneapi::math::blas::cublas::column_major::symm, + oneapi::math::blas::cublas::column_major::symm, + oneapi::math::blas::cublas::column_major::symm, + oneapi::math::blas::cublas::column_major::syrk, + oneapi::math::blas::cublas::column_major::syrk, + oneapi::math::blas::cublas::column_major::syrk, + oneapi::math::blas::cublas::column_major::syrk, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syrk_batch, + oneapi::math::blas::cublas::column_major::syr2k, + oneapi::math::blas::cublas::column_major::syr2k, + oneapi::math::blas::cublas::column_major::syr2k, + oneapi::math::blas::cublas::column_major::syr2k, + oneapi::math::blas::cublas::column_major::trmm, + oneapi::math::blas::cublas::column_major::trmm, + oneapi::math::blas::cublas::column_major::trmm, + oneapi::math::blas::cublas::column_major::trmm, + oneapi::math::blas::cublas::column_major::trsm, + oneapi::math::blas::cublas::column_major::trsm, + oneapi::math::blas::cublas::column_major::trsm, + oneapi::math::blas::cublas::column_major::trsm, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::trsm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemm_batch, + oneapi::math::blas::cublas::column_major::gemmt, + oneapi::math::blas::cublas::column_major::gemmt, + oneapi::math::blas::cublas::column_major::gemmt, + oneapi::math::blas::cublas::column_major::gemmt, + oneapi::math::blas::cublas::column_major::gemm_bias, + oneapi::math::blas::cublas::column_major::gemm_bias, + oneapi::math::blas::cublas::column_major::gemm_bias, + oneapi::math::blas::cublas::column_major::gemm_bias, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::omatadd_batch, + oneapi::math::blas::cublas::column_major::omatadd_batch, + oneapi::math::blas::cublas::column_major::omatadd_batch, + oneapi::math::blas::cublas::column_major::omatadd_batch, + oneapi::math::blas::cublas::column_major::omatcopy, + oneapi::math::blas::cublas::column_major::omatcopy, + oneapi::math::blas::cublas::column_major::omatcopy, + oneapi::math::blas::cublas::column_major::omatcopy, + oneapi::math::blas::cublas::column_major::omatcopy2, + oneapi::math::blas::cublas::column_major::omatcopy2, + oneapi::math::blas::cublas::column_major::omatcopy2, + oneapi::math::blas::cublas::column_major::omatcopy2, + oneapi::math::blas::cublas::column_major::imatcopy, + oneapi::math::blas::cublas::column_major::imatcopy, + oneapi::math::blas::cublas::column_major::imatcopy, + oneapi::math::blas::cublas::column_major::imatcopy, + oneapi::math::blas::cublas::column_major::omatadd, + oneapi::math::blas::cublas::column_major::omatadd, + oneapi::math::blas::cublas::column_major::omatadd, + oneapi::math::blas::cublas::column_major::omatadd, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::omatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::column_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::asum, + oneapi::math::blas::cublas::row_major::asum, + oneapi::math::blas::cublas::row_major::asum, + oneapi::math::blas::cublas::row_major::asum, + oneapi::math::blas::cublas::row_major::axpy, + oneapi::math::blas::cublas::row_major::axpy, + oneapi::math::blas::cublas::row_major::axpy, + oneapi::math::blas::cublas::row_major::axpy, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpby, + oneapi::math::blas::cublas::row_major::axpby, + oneapi::math::blas::cublas::row_major::axpby, + oneapi::math::blas::cublas::row_major::axpby, + oneapi::math::blas::cublas::row_major::copy, + oneapi::math::blas::cublas::row_major::copy, + oneapi::math::blas::cublas::row_major::copy, + oneapi::math::blas::cublas::row_major::copy, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::dot, + oneapi::math::blas::cublas::row_major::dot, + oneapi::math::blas::cublas::row_major::dot, + oneapi::math::blas::cublas::row_major::dotc, + oneapi::math::blas::cublas::row_major::dotc, + oneapi::math::blas::cublas::row_major::dotu, + oneapi::math::blas::cublas::row_major::dotu, + oneapi::math::blas::cublas::row_major::iamin, + oneapi::math::blas::cublas::row_major::iamin, + oneapi::math::blas::cublas::row_major::iamin, + oneapi::math::blas::cublas::row_major::iamin, + oneapi::math::blas::cublas::row_major::iamax, + oneapi::math::blas::cublas::row_major::iamax, + oneapi::math::blas::cublas::row_major::iamax, + oneapi::math::blas::cublas::row_major::iamax, + oneapi::math::blas::cublas::row_major::nrm2, + oneapi::math::blas::cublas::row_major::nrm2, + oneapi::math::blas::cublas::row_major::nrm2, + oneapi::math::blas::cublas::row_major::nrm2, + oneapi::math::blas::cublas::row_major::rot, + oneapi::math::blas::cublas::row_major::rot, + oneapi::math::blas::cublas::row_major::rot, + oneapi::math::blas::cublas::row_major::rot, + oneapi::math::blas::cublas::row_major::rotg, + oneapi::math::blas::cublas::row_major::rotg, + oneapi::math::blas::cublas::row_major::rotg, + oneapi::math::blas::cublas::row_major::rotg, + oneapi::math::blas::cublas::row_major::rotm, + oneapi::math::blas::cublas::row_major::rotm, + oneapi::math::blas::cublas::row_major::rotmg, + oneapi::math::blas::cublas::row_major::rotmg, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::sdsdot, + oneapi::math::blas::cublas::row_major::swap, + oneapi::math::blas::cublas::row_major::swap, + oneapi::math::blas::cublas::row_major::swap, + oneapi::math::blas::cublas::row_major::swap, + oneapi::math::blas::cublas::row_major::gbmv, + oneapi::math::blas::cublas::row_major::gbmv, + oneapi::math::blas::cublas::row_major::gbmv, + oneapi::math::blas::cublas::row_major::gbmv, + oneapi::math::blas::cublas::row_major::gemv, + oneapi::math::blas::cublas::row_major::gemv, + oneapi::math::blas::cublas::row_major::gemv, + oneapi::math::blas::cublas::row_major::gemv, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::ger, + oneapi::math::blas::cublas::row_major::ger, + oneapi::math::blas::cublas::row_major::gerc, + oneapi::math::blas::cublas::row_major::gerc, + oneapi::math::blas::cublas::row_major::geru, + oneapi::math::blas::cublas::row_major::geru, + oneapi::math::blas::cublas::row_major::hbmv, + oneapi::math::blas::cublas::row_major::hbmv, + oneapi::math::blas::cublas::row_major::hemv, + oneapi::math::blas::cublas::row_major::hemv, + oneapi::math::blas::cublas::row_major::her, + oneapi::math::blas::cublas::row_major::her, + oneapi::math::blas::cublas::row_major::her2, + oneapi::math::blas::cublas::row_major::her2, + oneapi::math::blas::cublas::row_major::hpmv, + oneapi::math::blas::cublas::row_major::hpmv, + oneapi::math::blas::cublas::row_major::hpr, + oneapi::math::blas::cublas::row_major::hpr, + oneapi::math::blas::cublas::row_major::hpr2, + oneapi::math::blas::cublas::row_major::hpr2, + oneapi::math::blas::cublas::row_major::sbmv, + oneapi::math::blas::cublas::row_major::sbmv, + oneapi::math::blas::cublas::row_major::spmv, + oneapi::math::blas::cublas::row_major::spmv, + oneapi::math::blas::cublas::row_major::spr, + oneapi::math::blas::cublas::row_major::spr, + oneapi::math::blas::cublas::row_major::spr2, + oneapi::math::blas::cublas::row_major::spr2, + oneapi::math::blas::cublas::row_major::symv, + oneapi::math::blas::cublas::row_major::symv, + oneapi::math::blas::cublas::row_major::syr, + oneapi::math::blas::cublas::row_major::syr, + oneapi::math::blas::cublas::row_major::syr2, + oneapi::math::blas::cublas::row_major::syr2, + oneapi::math::blas::cublas::row_major::tbmv, + oneapi::math::blas::cublas::row_major::tbmv, + oneapi::math::blas::cublas::row_major::tbmv, + oneapi::math::blas::cublas::row_major::tbmv, + oneapi::math::blas::cublas::row_major::tbsv, + oneapi::math::blas::cublas::row_major::tbsv, + oneapi::math::blas::cublas::row_major::tbsv, + oneapi::math::blas::cublas::row_major::tbsv, + oneapi::math::blas::cublas::row_major::tpmv, + oneapi::math::blas::cublas::row_major::tpmv, + oneapi::math::blas::cublas::row_major::tpmv, + oneapi::math::blas::cublas::row_major::tpmv, + oneapi::math::blas::cublas::row_major::tpsv, + oneapi::math::blas::cublas::row_major::tpsv, + oneapi::math::blas::cublas::row_major::tpsv, + oneapi::math::blas::cublas::row_major::tpsv, + oneapi::math::blas::cublas::row_major::trmv, + oneapi::math::blas::cublas::row_major::trmv, + oneapi::math::blas::cublas::row_major::trmv, + oneapi::math::blas::cublas::row_major::trmv, + oneapi::math::blas::cublas::row_major::trsv, + oneapi::math::blas::cublas::row_major::trsv, + oneapi::math::blas::cublas::row_major::trsv, + oneapi::math::blas::cublas::row_major::trsv, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::hemm, + oneapi::math::blas::cublas::row_major::hemm, + oneapi::math::blas::cublas::row_major::herk, + oneapi::math::blas::cublas::row_major::herk, + oneapi::math::blas::cublas::row_major::her2k, + oneapi::math::blas::cublas::row_major::her2k, + oneapi::math::blas::cublas::row_major::symm, + oneapi::math::blas::cublas::row_major::symm, + oneapi::math::blas::cublas::row_major::symm, + oneapi::math::blas::cublas::row_major::symm, + oneapi::math::blas::cublas::row_major::syrk, + oneapi::math::blas::cublas::row_major::syrk, + oneapi::math::blas::cublas::row_major::syrk, + oneapi::math::blas::cublas::row_major::syrk, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syr2k, + oneapi::math::blas::cublas::row_major::syr2k, + oneapi::math::blas::cublas::row_major::syr2k, + oneapi::math::blas::cublas::row_major::syr2k, + oneapi::math::blas::cublas::row_major::trmm, + oneapi::math::blas::cublas::row_major::trmm, + oneapi::math::blas::cublas::row_major::trmm, + oneapi::math::blas::cublas::row_major::trmm, + oneapi::math::blas::cublas::row_major::trsm, + oneapi::math::blas::cublas::row_major::trsm, + oneapi::math::blas::cublas::row_major::trsm, + oneapi::math::blas::cublas::row_major::trsm, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::gemmt, + oneapi::math::blas::cublas::row_major::gemmt, + oneapi::math::blas::cublas::row_major::gemmt, + oneapi::math::blas::cublas::row_major::gemmt, + oneapi::math::blas::cublas::row_major::gemm_bias, + oneapi::math::blas::cublas::row_major::gemm_bias, + oneapi::math::blas::cublas::row_major::gemm_bias, + oneapi::math::blas::cublas::row_major::gemm_bias, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::omatadd_batch, + oneapi::math::blas::cublas::row_major::omatadd_batch, + oneapi::math::blas::cublas::row_major::omatadd_batch, + oneapi::math::blas::cublas::row_major::omatadd_batch, + oneapi::math::blas::cublas::row_major::omatcopy, + oneapi::math::blas::cublas::row_major::omatcopy, + oneapi::math::blas::cublas::row_major::omatcopy, + oneapi::math::blas::cublas::row_major::omatcopy, + oneapi::math::blas::cublas::row_major::omatcopy2, + oneapi::math::blas::cublas::row_major::omatcopy2, + oneapi::math::blas::cublas::row_major::omatcopy2, + oneapi::math::blas::cublas::row_major::omatcopy2, + oneapi::math::blas::cublas::row_major::imatcopy, + oneapi::math::blas::cublas::row_major::imatcopy, + oneapi::math::blas::cublas::row_major::imatcopy, + oneapi::math::blas::cublas::row_major::imatcopy, + oneapi::math::blas::cublas::row_major::omatadd, + oneapi::math::blas::cublas::row_major::omatadd, + oneapi::math::blas::cublas::row_major::omatadd, + oneapi::math::blas::cublas::row_major::omatadd, + oneapi::math::blas::cublas::row_major::asum, + oneapi::math::blas::cublas::row_major::asum, + oneapi::math::blas::cublas::row_major::asum, + oneapi::math::blas::cublas::row_major::asum, + oneapi::math::blas::cublas::row_major::axpy, + oneapi::math::blas::cublas::row_major::axpy, + oneapi::math::blas::cublas::row_major::axpy, + oneapi::math::blas::cublas::row_major::axpy, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpy_batch, + oneapi::math::blas::cublas::row_major::axpby, + oneapi::math::blas::cublas::row_major::axpby, + oneapi::math::blas::cublas::row_major::axpby, + oneapi::math::blas::cublas::row_major::axpby, + oneapi::math::blas::cublas::row_major::copy, + oneapi::math::blas::cublas::row_major::copy, + oneapi::math::blas::cublas::row_major::copy, + oneapi::math::blas::cublas::row_major::copy, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::copy_batch, + oneapi::math::blas::cublas::row_major::dot, + oneapi::math::blas::cublas::row_major::dot, + oneapi::math::blas::cublas::row_major::dot, + oneapi::math::blas::cublas::row_major::dotc, + oneapi::math::blas::cublas::row_major::dotc, + oneapi::math::blas::cublas::row_major::dotu, + oneapi::math::blas::cublas::row_major::dotu, + oneapi::math::blas::cublas::row_major::iamin, + oneapi::math::blas::cublas::row_major::iamin, + oneapi::math::blas::cublas::row_major::iamin, + oneapi::math::blas::cublas::row_major::iamin, + oneapi::math::blas::cublas::row_major::iamax, + oneapi::math::blas::cublas::row_major::iamax, + oneapi::math::blas::cublas::row_major::iamax, + oneapi::math::blas::cublas::row_major::iamax, + oneapi::math::blas::cublas::row_major::nrm2, + oneapi::math::blas::cublas::row_major::nrm2, + oneapi::math::blas::cublas::row_major::nrm2, + oneapi::math::blas::cublas::row_major::nrm2, + oneapi::math::blas::cublas::row_major::rot, + oneapi::math::blas::cublas::row_major::rot, + oneapi::math::blas::cublas::row_major::rot, + oneapi::math::blas::cublas::row_major::rot, + oneapi::math::blas::cublas::row_major::rotg, + oneapi::math::blas::cublas::row_major::rotg, + oneapi::math::blas::cublas::row_major::rotg, + oneapi::math::blas::cublas::row_major::rotg, + oneapi::math::blas::cublas::row_major::rotm, + oneapi::math::blas::cublas::row_major::rotm, + oneapi::math::blas::cublas::row_major::rotmg, + oneapi::math::blas::cublas::row_major::rotmg, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::scal, + oneapi::math::blas::cublas::row_major::sdsdot, + oneapi::math::blas::cublas::row_major::swap, + oneapi::math::blas::cublas::row_major::swap, + oneapi::math::blas::cublas::row_major::swap, + oneapi::math::blas::cublas::row_major::swap, + oneapi::math::blas::cublas::row_major::gbmv, + oneapi::math::blas::cublas::row_major::gbmv, + oneapi::math::blas::cublas::row_major::gbmv, + oneapi::math::blas::cublas::row_major::gbmv, + oneapi::math::blas::cublas::row_major::gemv, + oneapi::math::blas::cublas::row_major::gemv, + oneapi::math::blas::cublas::row_major::gemv, + oneapi::math::blas::cublas::row_major::gemv, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::gemv_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::dgmm_batch, + oneapi::math::blas::cublas::row_major::ger, + oneapi::math::blas::cublas::row_major::ger, + oneapi::math::blas::cublas::row_major::gerc, + oneapi::math::blas::cublas::row_major::gerc, + oneapi::math::blas::cublas::row_major::geru, + oneapi::math::blas::cublas::row_major::geru, + oneapi::math::blas::cublas::row_major::hbmv, + oneapi::math::blas::cublas::row_major::hbmv, + oneapi::math::blas::cublas::row_major::hemv, + oneapi::math::blas::cublas::row_major::hemv, + oneapi::math::blas::cublas::row_major::her, + oneapi::math::blas::cublas::row_major::her, + oneapi::math::blas::cublas::row_major::her2, + oneapi::math::blas::cublas::row_major::her2, + oneapi::math::blas::cublas::row_major::hpmv, + oneapi::math::blas::cublas::row_major::hpmv, + oneapi::math::blas::cublas::row_major::hpr, + oneapi::math::blas::cublas::row_major::hpr, + oneapi::math::blas::cublas::row_major::hpr2, + oneapi::math::blas::cublas::row_major::hpr2, + oneapi::math::blas::cublas::row_major::sbmv, + oneapi::math::blas::cublas::row_major::sbmv, + oneapi::math::blas::cublas::row_major::spmv, + oneapi::math::blas::cublas::row_major::spmv, + oneapi::math::blas::cublas::row_major::spr, + oneapi::math::blas::cublas::row_major::spr, + oneapi::math::blas::cublas::row_major::spr2, + oneapi::math::blas::cublas::row_major::spr2, + oneapi::math::blas::cublas::row_major::symv, + oneapi::math::blas::cublas::row_major::symv, + oneapi::math::blas::cublas::row_major::syr, + oneapi::math::blas::cublas::row_major::syr, + oneapi::math::blas::cublas::row_major::syr2, + oneapi::math::blas::cublas::row_major::syr2, + oneapi::math::blas::cublas::row_major::tbmv, + oneapi::math::blas::cublas::row_major::tbmv, + oneapi::math::blas::cublas::row_major::tbmv, + oneapi::math::blas::cublas::row_major::tbmv, + oneapi::math::blas::cublas::row_major::tbsv, + oneapi::math::blas::cublas::row_major::tbsv, + oneapi::math::blas::cublas::row_major::tbsv, + oneapi::math::blas::cublas::row_major::tbsv, + oneapi::math::blas::cublas::row_major::tpmv, + oneapi::math::blas::cublas::row_major::tpmv, + oneapi::math::blas::cublas::row_major::tpmv, + oneapi::math::blas::cublas::row_major::tpmv, + oneapi::math::blas::cublas::row_major::tpsv, + oneapi::math::blas::cublas::row_major::tpsv, + oneapi::math::blas::cublas::row_major::tpsv, + oneapi::math::blas::cublas::row_major::tpsv, + oneapi::math::blas::cublas::row_major::trmv, + oneapi::math::blas::cublas::row_major::trmv, + oneapi::math::blas::cublas::row_major::trmv, + oneapi::math::blas::cublas::row_major::trmv, + oneapi::math::blas::cublas::row_major::trsv, + oneapi::math::blas::cublas::row_major::trsv, + oneapi::math::blas::cublas::row_major::trsv, + oneapi::math::blas::cublas::row_major::trsv, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::gemm, + oneapi::math::blas::cublas::row_major::hemm, + oneapi::math::blas::cublas::row_major::hemm, + oneapi::math::blas::cublas::row_major::herk, + oneapi::math::blas::cublas::row_major::herk, + oneapi::math::blas::cublas::row_major::her2k, + oneapi::math::blas::cublas::row_major::her2k, + oneapi::math::blas::cublas::row_major::symm, + oneapi::math::blas::cublas::row_major::symm, + oneapi::math::blas::cublas::row_major::symm, + oneapi::math::blas::cublas::row_major::symm, + oneapi::math::blas::cublas::row_major::syrk, + oneapi::math::blas::cublas::row_major::syrk, + oneapi::math::blas::cublas::row_major::syrk, + oneapi::math::blas::cublas::row_major::syrk, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syrk_batch, + oneapi::math::blas::cublas::row_major::syr2k, + oneapi::math::blas::cublas::row_major::syr2k, + oneapi::math::blas::cublas::row_major::syr2k, + oneapi::math::blas::cublas::row_major::syr2k, + oneapi::math::blas::cublas::row_major::trmm, + oneapi::math::blas::cublas::row_major::trmm, + oneapi::math::blas::cublas::row_major::trmm, + oneapi::math::blas::cublas::row_major::trmm, + oneapi::math::blas::cublas::row_major::trsm, + oneapi::math::blas::cublas::row_major::trsm, + oneapi::math::blas::cublas::row_major::trsm, + oneapi::math::blas::cublas::row_major::trsm, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::trsm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemm_batch, + oneapi::math::blas::cublas::row_major::gemmt, + oneapi::math::blas::cublas::row_major::gemmt, + oneapi::math::blas::cublas::row_major::gemmt, + oneapi::math::blas::cublas::row_major::gemmt, + oneapi::math::blas::cublas::row_major::gemm_bias, + oneapi::math::blas::cublas::row_major::gemm_bias, + oneapi::math::blas::cublas::row_major::gemm_bias, + oneapi::math::blas::cublas::row_major::gemm_bias, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::omatadd_batch, + oneapi::math::blas::cublas::row_major::omatadd_batch, + oneapi::math::blas::cublas::row_major::omatadd_batch, + oneapi::math::blas::cublas::row_major::omatadd_batch, + oneapi::math::blas::cublas::row_major::omatcopy, + oneapi::math::blas::cublas::row_major::omatcopy, + oneapi::math::blas::cublas::row_major::omatcopy, + oneapi::math::blas::cublas::row_major::omatcopy, + oneapi::math::blas::cublas::row_major::omatcopy2, + oneapi::math::blas::cublas::row_major::omatcopy2, + oneapi::math::blas::cublas::row_major::omatcopy2, + oneapi::math::blas::cublas::row_major::omatcopy2, + oneapi::math::blas::cublas::row_major::imatcopy, + oneapi::math::blas::cublas::row_major::imatcopy, + oneapi::math::blas::cublas::row_major::imatcopy, + oneapi::math::blas::cublas::row_major::imatcopy, + oneapi::math::blas::cublas::row_major::omatadd, + oneapi::math::blas::cublas::row_major::omatadd, + oneapi::math::blas::cublas::row_major::omatadd, + oneapi::math::blas::cublas::row_major::omatadd, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::omatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, + oneapi::math::blas::cublas::row_major::imatcopy_batch, }; diff --git a/src/blas/backends/mkl_common/mkl_blas_backend.hpp b/src/blas/backends/mkl_common/mkl_blas_backend.hpp index e542e221a..e753b4b68 100644 --- a/src/blas/backends/mkl_common/mkl_blas_backend.hpp +++ b/src/blas/backends/mkl_common/mkl_blas_backend.hpp @@ -25,7 +25,7 @@ #include "oneapi/math/types.hpp" namespace oneapi { -namespace mkl { +namespace math { template class value_or_pointer { @@ -80,5 +80,5 @@ namespace row_major { } } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/mklcpu/mklcpu_batch.cpp b/src/blas/backends/mklcpu/mklcpu_batch.cpp index 64ed392ef..541aa80dc 100644 --- a/src/blas/backends/mklcpu/mklcpu_batch.cpp +++ b/src/blas/backends/mklcpu/mklcpu_batch.cpp @@ -29,22 +29,22 @@ #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace mklcpu { namespace column_major { -namespace blas_major = ::oneapi::mkl::blas::column_major; +namespace blas_major = ::oneapi::math::blas::column_major; #include "../mkl_common/mkl_batch.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::mkl::blas::row_major; +namespace blas_major = ::oneapi::math::blas::row_major; #include "../mkl_common/mkl_batch.cxx" } // namespace row_major } // namespace mklcpu } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/mklcpu/mklcpu_extensions.cpp b/src/blas/backends/mklcpu/mklcpu_extensions.cpp index 0daa063e8..40accb82e 100644 --- a/src/blas/backends/mklcpu/mklcpu_extensions.cpp +++ b/src/blas/backends/mklcpu/mklcpu_extensions.cpp @@ -29,22 +29,22 @@ #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace mklcpu { namespace column_major { -namespace blas_major = ::oneapi::mkl::blas::column_major; +namespace blas_major = ::oneapi::math::blas::column_major; #include "../mkl_common/mkl_extensions.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::mkl::blas::row_major; +namespace blas_major = ::oneapi::math::blas::row_major; #include "../mkl_common/mkl_extensions.cxx" } // namespace row_major } // namespace mklcpu } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/mklcpu/mklcpu_level1.cpp b/src/blas/backends/mklcpu/mklcpu_level1.cpp index edddb1afc..02cd4094e 100644 --- a/src/blas/backends/mklcpu/mklcpu_level1.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level1.cpp @@ -28,22 +28,22 @@ #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace mklcpu { namespace column_major { -namespace blas_major = ::oneapi::mkl::blas::column_major; +namespace blas_major = ::oneapi::math::blas::column_major; #include "../mkl_common/mkl_level1.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::mkl::blas::row_major; +namespace blas_major = ::oneapi::math::blas::row_major; #include "../mkl_common/mkl_level1.cxx" } // namespace row_major } // namespace mklcpu } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/mklcpu/mklcpu_level2.cpp b/src/blas/backends/mklcpu/mklcpu_level2.cpp index 8fa321ff9..3eda04ee7 100644 --- a/src/blas/backends/mklcpu/mklcpu_level2.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level2.cpp @@ -28,22 +28,22 @@ #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace mklcpu { namespace column_major { -namespace blas_major = ::oneapi::mkl::blas::column_major; +namespace blas_major = ::oneapi::math::blas::column_major; #include "../mkl_common/mkl_level2.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::mkl::blas::row_major; +namespace blas_major = ::oneapi::math::blas::row_major; #include "../mkl_common/mkl_level2.cxx" } // namespace row_major } // namespace mklcpu } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/mklcpu/mklcpu_level3.cpp b/src/blas/backends/mklcpu/mklcpu_level3.cpp index debdc802f..72450596a 100644 --- a/src/blas/backends/mklcpu/mklcpu_level3.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level3.cpp @@ -28,22 +28,22 @@ #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace mklcpu { namespace column_major { -namespace blas_major = ::oneapi::mkl::blas::column_major; +namespace blas_major = ::oneapi::math::blas::column_major; #include "../mkl_common/mkl_level3.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::mkl::blas::row_major; +namespace blas_major = ::oneapi::math::blas::row_major; #include "../mkl_common/mkl_level3.cxx" } // namespace row_major } // namespace mklcpu } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/mklgpu/mklgpu_batch.cpp b/src/blas/backends/mklgpu/mklgpu_batch.cpp index 5767126a8..b33a2b030 100644 --- a/src/blas/backends/mklgpu/mklgpu_batch.cpp +++ b/src/blas/backends/mklgpu/mklgpu_batch.cpp @@ -29,22 +29,22 @@ #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace mklgpu { namespace column_major { -namespace blas_major = ::oneapi::mkl::blas::column_major; +namespace blas_major = ::oneapi::math::blas::column_major; #include "../mkl_common/mkl_batch.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::mkl::blas::row_major; +namespace blas_major = ::oneapi::math::blas::row_major; #include "../mkl_common/mkl_batch.cxx" } // namespace row_major } // namespace mklgpu } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/mklgpu/mklgpu_extensions.cpp b/src/blas/backends/mklgpu/mklgpu_extensions.cpp index da9e615c4..9ade59f76 100644 --- a/src/blas/backends/mklgpu/mklgpu_extensions.cpp +++ b/src/blas/backends/mklgpu/mklgpu_extensions.cpp @@ -29,22 +29,22 @@ #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace mklgpu { namespace column_major { -namespace blas_major = ::oneapi::mkl::blas::column_major; +namespace blas_major = ::oneapi::math::blas::column_major; #include "../mkl_common/mkl_extensions.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::mkl::blas::row_major; +namespace blas_major = ::oneapi::math::blas::row_major; #include "../mkl_common/mkl_extensions.cxx" } // namespace row_major } // namespace mklgpu } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/mklgpu/mklgpu_level1.cpp b/src/blas/backends/mklgpu/mklgpu_level1.cpp index 928860eb6..ca45cc6de 100644 --- a/src/blas/backends/mklgpu/mklgpu_level1.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level1.cpp @@ -28,23 +28,23 @@ #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace mklgpu { namespace column_major { -namespace blas_major = ::oneapi::mkl::blas::column_major; +namespace blas_major = ::oneapi::math::blas::column_major; #include "../mkl_common/mkl_level1.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::mkl::blas::row_major; +namespace blas_major = ::oneapi::math::blas::row_major; #include "../mkl_common/mkl_level1.cxx" } // namespace row_major } // namespace mklgpu } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/mklgpu/mklgpu_level2.cpp b/src/blas/backends/mklgpu/mklgpu_level2.cpp index 51cd0c119..4bf1c08c5 100644 --- a/src/blas/backends/mklgpu/mklgpu_level2.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level2.cpp @@ -28,22 +28,22 @@ #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace mklgpu { namespace column_major { -namespace blas_major = ::oneapi::mkl::blas::column_major; +namespace blas_major = ::oneapi::math::blas::column_major; #include "../mkl_common/mkl_level2.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::mkl::blas::row_major; +namespace blas_major = ::oneapi::math::blas::row_major; #include "../mkl_common/mkl_level2.cxx" } // namespace row_major } // namespace mklgpu } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/mklgpu/mklgpu_level3.cpp b/src/blas/backends/mklgpu/mklgpu_level3.cpp index 9f35affe5..bbb7d10fe 100644 --- a/src/blas/backends/mklgpu/mklgpu_level3.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level3.cpp @@ -28,22 +28,22 @@ #include "../mkl_common/mkl_blas_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace mklgpu { namespace column_major { -namespace blas_major = ::oneapi::mkl::blas::column_major; +namespace blas_major = ::oneapi::math::blas::column_major; #include "../mkl_common/mkl_level3.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::mkl::blas::row_major; +namespace blas_major = ::oneapi::math::blas::row_major; #include "../mkl_common/mkl_level3.cxx" } // namespace row_major } // namespace mklgpu } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/netlib/netlib_batch.cpp b/src/blas/backends/netlib/netlib_batch.cpp index ae492ee6d..09bf4c171 100644 --- a/src/blas/backends/netlib/netlib_batch.cpp +++ b/src/blas/backends/netlib/netlib_batch.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace netlib { namespace column_major { @@ -47,5 +47,5 @@ namespace row_major { } // namespace row_major } // namespace netlib } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/netlib/netlib_common.hpp b/src/blas/backends/netlib/netlib_common.hpp index 984f755da..51552f2ce 100644 --- a/src/blas/backends/netlib/netlib_common.hpp +++ b/src/blas/backends/netlib/netlib_common.hpp @@ -35,7 +35,7 @@ #define GET_MULTI_PTR template get_multi_ptr().get_raw() namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace netlib { @@ -97,7 +97,7 @@ static inline void host_task(H &cgh, F f) { } // namespace netlib } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_NETLIB_COMMON_HPP_ diff --git a/src/blas/backends/netlib/netlib_extensions.cpp b/src/blas/backends/netlib/netlib_extensions.cpp index 888129e44..c062ba56e 100644 --- a/src/blas/backends/netlib/netlib_extensions.cpp +++ b/src/blas/backends/netlib/netlib_extensions.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/exceptions.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace netlib { namespace column_major { @@ -47,5 +47,5 @@ namespace row_major { } // namespace row_major } // namespace netlib } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/netlib/netlib_level1.cpp b/src/blas/backends/netlib/netlib_level1.cpp index d067c5c15..ec6b1eb09 100644 --- a/src/blas/backends/netlib/netlib_level1.cpp +++ b/src/blas/backends/netlib/netlib_level1.cpp @@ -222,7 +222,7 @@ void cblas_zrotg(std::complex *ca, const std::complex *cb, doubl } namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace netlib { namespace column_major { @@ -241,5 +241,5 @@ namespace row_major { } // namespace row_major } // namespace netlib } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/netlib/netlib_level2.cpp b/src/blas/backends/netlib/netlib_level2.cpp index f455d609f..c33636564 100644 --- a/src/blas/backends/netlib/netlib_level2.cpp +++ b/src/blas/backends/netlib/netlib_level2.cpp @@ -27,7 +27,7 @@ #include "oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace netlib { namespace column_major { @@ -46,5 +46,5 @@ namespace row_major { } // namespace row_major } // namespace netlib } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/netlib/netlib_level3.cpp b/src/blas/backends/netlib/netlib_level3.cpp index 952d16690..1d1fcc027 100644 --- a/src/blas/backends/netlib/netlib_level3.cpp +++ b/src/blas/backends/netlib/netlib_level3.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/blas/detail/netlib/onemath_blas_netlib.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace netlib { namespace column_major { @@ -51,5 +51,5 @@ namespace row_major { } // namespace row_major } // namespace netlib } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/portblas/portblas_batch.cpp b/src/blas/backends/portblas/portblas_batch.cpp index c4d2b427b..d80207c8c 100644 --- a/src/blas/backends/portblas/portblas_batch.cpp +++ b/src/blas/backends/portblas/portblas_batch.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace portblas { namespace column_major { @@ -53,5 +53,5 @@ constexpr bool is_column_major() { } // namespace row_major } // namespace portblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/portblas/portblas_batch.cxx b/src/blas/backends/portblas/portblas_batch.cxx index 28c7ee5dc..e48e60890 100644 --- a/src/blas/backends/portblas/portblas_batch.cxx +++ b/src/blas/backends/portblas/portblas_batch.cxx @@ -19,21 +19,21 @@ // Buffer APIs -void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "syrk_batch", ""); } -void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "syrk_batch", ""); } -void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, @@ -41,7 +41,7 @@ void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl:: throw unimplemented("blas", "syrk_batch", ""); } -void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, @@ -49,7 +49,7 @@ void syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl:: throw unimplemented("blas", "syrk_batch", ""); } -void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, @@ -57,7 +57,7 @@ void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m throw unimplemented("blas", "gemv_batch", ""); } -void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, @@ -65,7 +65,7 @@ void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m throw unimplemented("blas", "gemv_batch", ""); } -void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, std::complex beta, @@ -74,7 +74,7 @@ void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m throw unimplemented("blas", "gemv_batch", ""); } -void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, std::complex beta, @@ -83,7 +83,7 @@ void gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m throw unimplemented("blas", "gemv_batch", ""); } -void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, +void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, @@ -91,7 +91,7 @@ void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m throw unimplemented("blas", "dgmm_batch", ""); } -void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, +void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, @@ -99,7 +99,7 @@ void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m throw unimplemented("blas", "dgmm_batch", ""); } -void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, +void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, @@ -107,7 +107,7 @@ void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m throw unimplemented("blas", "dgmm_batch", ""); } -void dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, +void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, @@ -166,7 +166,7 @@ void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -176,7 +176,7 @@ void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl:: stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, @@ -186,7 +186,7 @@ void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl:: stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, @@ -195,7 +195,7 @@ void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl:: throw unimplemented("blas", "gemm_batch", " for complex"); } -void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, @@ -204,7 +204,7 @@ void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl:: throw unimplemented("blas", "gemm_batch", " for complex"); } -void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -213,7 +213,7 @@ void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl:: throw unimplemented("blas", "gemm_batch", " for complex"); } -void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -222,7 +222,7 @@ void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl:: throw unimplemented("blas", "gemm_batch", " for unsupported dtype"); } -void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -231,7 +231,7 @@ void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl:: throw unimplemented("blas", "gemm_batch", " for unsupported dtype"); } -void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -240,32 +240,32 @@ void gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl:: throw unimplemented("blas", "gemm_batch", " for unsupported dtype"); } -void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, +void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("blas", "trsm_batch", ""); } -void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, +void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("blas", "trsm_batch", ""); } -void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, +void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("blas", "trsm_batch", ""); } -void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, +void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, @@ -273,7 +273,7 @@ void trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::u throw unimplemented("blas", "trsm_batch", ""); } -void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { @@ -281,7 +281,7 @@ void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64 stride_b, batch_size); } -void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { @@ -289,7 +289,7 @@ void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64 stride_b, batch_size); } -void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -297,7 +297,7 @@ void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64 throw unimplemented("blas", "omatcopy_batch", ""); } -void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -305,33 +305,33 @@ void omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64 throw unimplemented("blas", "omatcopy_batch", ""); } -void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", ""); } -void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", ""); } -void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", ""); } -void imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", ""); } -void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, @@ -340,7 +340,7 @@ void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mk beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, @@ -349,7 +349,7 @@ void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mk beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &b, @@ -358,7 +358,7 @@ void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mk throw unimplemented("blas", "omatadd_batch", ""); } -void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -370,24 +370,24 @@ void omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mk // USM APIs -sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *k, +sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *k, +sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *k, +sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -395,8 +395,8 @@ sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *k, +sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -404,24 +404,24 @@ sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo *upper_lower, throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, float alpha, +sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, double alpha, +sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, @@ -429,8 +429,8 @@ sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, +sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, @@ -438,7 +438,7 @@ sycl::event syrk_batch(sycl::queue &queue, oneapi::mkl::uplo upper_lower, throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float beta, float *y, std::int64_t incy, @@ -447,7 +447,7 @@ sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::in throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double beta, double *y, std::int64_t incy, @@ -456,7 +456,7 @@ sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::in throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, @@ -465,7 +465,7 @@ sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::in throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, @@ -474,7 +474,7 @@ sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::in throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, +sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, @@ -482,7 +482,7 @@ sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::i throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, +sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, @@ -490,7 +490,7 @@ sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::i throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, +sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, std::complex **y, std::int64_t *incy, @@ -499,7 +499,7 @@ sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::i throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, +sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, std::complex **y, std::int64_t *incy, @@ -508,7 +508,7 @@ sycl::event gemv_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::i throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, @@ -516,7 +516,7 @@ sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::in throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, @@ -524,7 +524,7 @@ sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::in throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, @@ -533,7 +533,7 @@ sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::in throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, +sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, @@ -542,21 +542,21 @@ sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side left_right, std::in throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, +sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, +sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, +sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -564,7 +564,7 @@ sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, std::i throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, +sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -680,8 +680,8 @@ sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, @@ -708,8 +708,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, @@ -718,8 +718,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, sycl::half *beta, sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -727,8 +727,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -736,8 +736,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -745,8 +745,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, - oneapi::mkl::transpose *transb, std::int64_t *m, std::int64_t *n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, + oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -754,8 +754,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose *transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, @@ -766,8 +766,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, dependencies); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, double *c, std::int64_t ldc, @@ -778,8 +778,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, dependencies); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, @@ -788,8 +788,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, @@ -798,8 +798,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, @@ -808,8 +808,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, @@ -818,8 +818,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, @@ -828,8 +828,8 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, std::int32_t *c, std::int64_t ldc, @@ -838,27 +838,27 @@ sycl::event gemm_batch(sycl::queue &queue, oneapi::mkl::transpose transa, throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, +sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, +sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -866,9 +866,9 @@ sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -876,36 +876,36 @@ sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side left_right, throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, - oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, - oneapi::mkl::diag *unit_diag, std::int64_t *m, std::int64_t *n, float *alpha, +sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, + oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, + oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, - oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, - oneapi::mkl::diag *unit_diag, std::int64_t *m, std::int64_t *n, +sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, + oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, + oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, - oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, - oneapi::mkl::diag *unit_diag, std::int64_t *m, std::int64_t *n, +sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, + oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, + oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, - oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, - oneapi::mkl::diag *unit_diag, std::int64_t *m, std::int64_t *n, +sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, + oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, + oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, @@ -913,7 +913,7 @@ sycl::event trsm_batch(sycl::queue &queue, oneapi::mkl::side *left_right, throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies) { @@ -921,7 +921,7 @@ sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -930,7 +930,7 @@ sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -938,7 +938,7 @@ sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std throw unimplemented("blas", "omatcopy_batch", " for USM"); } -sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -946,36 +946,36 @@ sycl::event omatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std throw unimplemented("blas", "omatcopy_batch", " for USM"); } -sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { throw unimplemented("blas", "imatcopy_batch", " for USM"); } -sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { throw unimplemented("blas", "imatcopy_batch", " for USM"); } -sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { throw unimplemented("blas", "imatcopy_batch", " for USM"); } -sycl::event imatcopy_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { throw unimplemented("blas", "imatcopy_batch", " for USM"); } -sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, std::int64_t ldc, std::int64_t stride_c, @@ -985,8 +985,8 @@ sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, dependencies); } -sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, std::int64_t ldc, std::int64_t stride_c, @@ -996,8 +996,8 @@ sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, dependencies); } -sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -1006,8 +1006,8 @@ sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, throw unimplemented("blas", "omatadd_batch", " for USM"); } -sycl::event omatadd_batch(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, +sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, diff --git a/src/blas/backends/portblas/portblas_common.hpp b/src/blas/backends/portblas/portblas_common.hpp index b66759d4a..27d68b3d4 100644 --- a/src/blas/backends/portblas/portblas_common.hpp +++ b/src/blas/backends/portblas/portblas_common.hpp @@ -28,7 +28,7 @@ #include namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace portblas { @@ -64,10 +64,10 @@ DEF_PORTBLAS_TYPE(int64_t, int64_t) DEF_PORTBLAS_TYPE(sycl::half, sycl::half) DEF_PORTBLAS_TYPE(float, float) DEF_PORTBLAS_TYPE(double, double) -DEF_PORTBLAS_TYPE(oneapi::mkl::transpose, char) -DEF_PORTBLAS_TYPE(oneapi::mkl::uplo, char) -DEF_PORTBLAS_TYPE(oneapi::mkl::side, char) -DEF_PORTBLAS_TYPE(oneapi::mkl::diag, char) +DEF_PORTBLAS_TYPE(oneapi::math::transpose, char) +DEF_PORTBLAS_TYPE(oneapi::math::uplo, char) +DEF_PORTBLAS_TYPE(oneapi::math::side, char) +DEF_PORTBLAS_TYPE(oneapi::math::diag, char) DEF_PORTBLAS_TYPE(std::complex, sycl_complex_t) DEF_PORTBLAS_TYPE(std::complex, sycl_complex_t) // Passthrough of portBLAS arg types for more complex wrapping. @@ -113,21 +113,21 @@ inline typename portblas_type::type convert_to_portblas_type(InputT& inp } template <> -inline char convert_to_portblas_type(oneapi::mkl::transpose& trans) { - if (trans == oneapi::mkl::transpose::nontrans) { +inline char convert_to_portblas_type(oneapi::math::transpose& trans) { + if (trans == oneapi::math::transpose::nontrans) { return 'n'; } - else if (trans == oneapi::mkl::transpose::trans) { + else if (trans == oneapi::math::transpose::trans) { return 't'; } - else { // trans == oneapi::mkl::transpose::conjtrans + else { // trans == oneapi::math::transpose::conjtrans return 'c'; } } template <> -inline char convert_to_portblas_type(oneapi::mkl::uplo& upper_lower) { - if (upper_lower == oneapi::mkl::uplo::upper) { +inline char convert_to_portblas_type(oneapi::math::uplo& upper_lower) { + if (upper_lower == oneapi::math::uplo::upper) { return 'u'; } else { @@ -136,8 +136,8 @@ inline char convert_to_portblas_type(oneapi::mkl::uplo& upper } template <> -inline char convert_to_portblas_type(oneapi::mkl::side& left_right) { - if (left_right == oneapi::mkl::side::left) { +inline char convert_to_portblas_type(oneapi::math::side& left_right) { + if (left_right == oneapi::math::side::left) { return 'l'; } else { @@ -146,8 +146,8 @@ inline char convert_to_portblas_type(oneapi::mkl::side& left_ } template <> -inline char convert_to_portblas_type(oneapi::mkl::diag& unit_diag) { - if (unit_diag == oneapi::mkl::diag::unit) { +inline char convert_to_portblas_type(oneapi::math::diag& unit_diag) { + if (unit_diag == oneapi::math::diag::unit) { return 'u'; } else { @@ -180,7 +180,7 @@ struct throw_if_unsupported_by_device { static constexpr bool checkTypeInPack = (std::is_same_v || ...); if (checkTypeInPack) { if (!q.get_info().has(AspectVal)) { - throw mkl::unsupported_device("blas", message, + throw math::unsupported_device("blas", message, q.get_info()); } } @@ -233,7 +233,7 @@ struct throw_if_unsupported_by_device { } // namespace portblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _PORTBLAS_COMMON_HPP_ diff --git a/src/blas/backends/portblas/portblas_gemm_bias.cxx b/src/blas/backends/portblas/portblas_gemm_bias.cxx index 30f638f3e..2af4d0bd2 100644 --- a/src/blas/backends/portblas/portblas_gemm_bias.cxx +++ b/src/blas/backends/portblas/portblas_gemm_bias.cxx @@ -19,32 +19,32 @@ // Buffer APIs -void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, +void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { throw unimplemented("blas", "gemm_bias", ""); } -void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, +void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { throw unimplemented("blas", "gemm_bias", ""); } -void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, +void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { throw unimplemented("blas", "gemm_bias", ""); } -void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, +void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { @@ -53,8 +53,8 @@ void gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::t // USM APIs -sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, +sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, @@ -62,8 +62,8 @@ sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, throw unimplemented("blas", "gemm_bias", " for USM"); } -sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, +sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, @@ -71,8 +71,8 @@ sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, throw unimplemented("blas", "gemm_bias", " for USM"); } -sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, +sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, @@ -80,8 +80,8 @@ sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, throw unimplemented("blas", "gemm_bias", " for USM"); } -sycl::event gemm_bias(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, std::int64_t m, +sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, diff --git a/src/blas/backends/portblas/portblas_level1_double.cpp b/src/blas/backends/portblas/portblas_level1_double.cpp index 9f7bd9428..172fcfa8d 100644 --- a/src/blas/backends/portblas/portblas_level1_double.cpp +++ b/src/blas/backends/portblas/portblas_level1_double.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace portblas { @@ -58,5 +58,5 @@ constexpr bool is_column_major() { } // namespace row_major } // namespace portblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/portblas/portblas_level1_float.cpp b/src/blas/backends/portblas/portblas_level1_float.cpp index b3f9cc5cf..b4a54375d 100644 --- a/src/blas/backends/portblas/portblas_level1_float.cpp +++ b/src/blas/backends/portblas/portblas_level1_float.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace portblas { @@ -56,5 +56,5 @@ constexpr bool is_column_major() { } // namespace row_major } // namespace portblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/portblas/portblas_level2.cxx b/src/blas/backends/portblas/portblas_level2.cxx index b3d8b6766..8d01deec2 100644 --- a/src/blas/backends/portblas/portblas_level2.cxx +++ b/src/blas/backends/portblas/portblas_level2.cxx @@ -19,20 +19,20 @@ // Buffer APIs -void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, real_t beta, sycl::buffer &y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_gemv, queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -void gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { throw unimplemented("blas", "gemv", " for complex"); } -void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, real_t beta, sycl::buffer &y, std::int64_t incy) { @@ -40,7 +40,7 @@ void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std: incy); } -void gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, @@ -68,172 +68,172 @@ void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { throw unimplemented("blas", "hbmv", ""); } -void hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +void hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { throw unimplemented("blas", "hemv", ""); } -void her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +void her(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { throw unimplemented("blas", "her", ""); } -void her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +void her2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { throw unimplemented("blas", "her2", ""); } -void hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +void hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { throw unimplemented("blas", "hpmv", ""); } -void hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +void hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { throw unimplemented("blas", "hpr", ""); } -void hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +void hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { throw unimplemented("blas", "hpr2", ""); } -void sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, +void sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, real_t beta, sycl::buffer &y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_sbmv, queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -void symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +void symv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, real_t beta, sycl::buffer &y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_symv, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +void syr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { CALL_PORTBLAS_FN(::blas::_syr, queue, upper_lower, n, alpha, x, incx, a, lda); } -void syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +void syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { CALL_PORTBLAS_FN(::blas::_syr2, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +void spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, real_t beta, sycl::buffer &y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_spmv, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +void spr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { CALL_PORTBLAS_FN(::blas::_spr, queue, upper_lower, n, alpha, x, incx, a); } -void spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +void spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { CALL_PORTBLAS_FN(::blas::_spr2, queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, +void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_tbmv, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, +void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { throw unimplemented("blas", "tbmv", ""); } -void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, +void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_tbsv, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, +void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { throw unimplemented("blas", "tbsv", ""); } -void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, +void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_tpmv, queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, +void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { throw unimplemented("blas", "tpmv", ""); } -void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, +void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_tpsv, queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, +void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { throw unimplemented("blas", "tpsv", ""); } -void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, +void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_trmv, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, +void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { throw unimplemented("blas", "trmv", " for complex"); } -void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, +void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_trsv, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, +void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { throw unimplemented("blas", "trsv", ""); } // USM APIs -sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, real_t alpha, const real_t *a, std::int64_t lda, const real_t *x, std::int64_t incx, real_t beta, real_t *y, std::int64_t incy, const std::vector &dependencies) { @@ -241,7 +241,7 @@ sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t dependencies); } -sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -249,7 +249,7 @@ sycl::event gemv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t throw unimplemented("blas", "gemv", " for USM"); } -sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, real_t alpha, const real_t *a, std::int64_t lda, const real_t *x, std::int64_t incx, real_t beta, real_t *y, std::int64_t incy, const std::vector &dependencies) { @@ -257,7 +257,7 @@ sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t incy, dependencies); } -sycl::event gbmv(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, @@ -285,7 +285,7 @@ sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::comple throw unimplemented("blas", "geru", " for USM"); } -sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, +sycl::event hbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -293,7 +293,7 @@ sycl::event hbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t throw unimplemented("blas", "hbmv", " for USM"); } -sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +sycl::event hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -301,20 +301,20 @@ sycl::event hemv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t throw unimplemented("blas", "hemv", " for USM"); } -sycl::event her(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +sycl::event her(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { throw unimplemented("blas", "her", " for USM"); } -sycl::event her2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +sycl::event her2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { throw unimplemented("blas", "her2", " for USM"); } -sycl::event hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +sycl::event hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -322,20 +322,20 @@ sycl::event hpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t throw unimplemented("blas", "hpmv", " for USM"); } -sycl::event hpr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +sycl::event hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { throw unimplemented("blas", "hpr", " for USM"); } -sycl::event hpr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, +sycl::event hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { throw unimplemented("blas", "hpr2", " for USM"); } -sycl::event sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, +sycl::event sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, real_t alpha, const real_t *a, std::int64_t lda, const real_t *x, std::int64_t incx, real_t beta, real_t *y, std::int64_t incy, const std::vector &dependencies) { @@ -343,127 +343,127 @@ sycl::event sbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t incy, dependencies); } -sycl::event symv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +sycl::event symv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, const real_t *a, std::int64_t lda, const real_t *x, std::int64_t incx, real_t beta, real_t *y, std::int64_t incy, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_symv, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event syr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +sycl::event syr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, const real_t *x, std::int64_t incx, real_t *a, std::int64_t lda, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_syr, queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event syr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +sycl::event syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, const real_t *x, std::int64_t incx, const real_t *y, std::int64_t incy, real_t *a, std::int64_t lda, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_syr2, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event spmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +sycl::event spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, const real_t *a, const real_t *x, std::int64_t incx, real_t beta, real_t *y, std::int64_t incy, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_spmv, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event spr(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +sycl::event spr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, const real_t *x, std::int64_t incx, real_t *a, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_spr, queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event spr2(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, real_t alpha, +sycl::event spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, const real_t *x, std::int64_t incx, const real_t *y, std::int64_t incy, real_t *a, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_spr2, queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, const real_t *a, +sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const real_t *a, std::int64_t lda, real_t *x, std::int64_t incx, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_tbmv, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, +sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { throw unimplemented("blas", "tbmv", " for USM"); } -sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, const real_t *a, +sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const real_t *a, std::int64_t lda, real_t *x, std::int64_t incx, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_tbsv, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, std::int64_t k, +sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { throw unimplemented("blas", "tbsv", " for USM"); } -sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, const real_t *a, real_t *x, +sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const real_t *a, real_t *x, std::int64_t incx, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_tpmv, queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, +sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { throw unimplemented("blas", "tpmv", " for USM"); } -sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, const real_t *a, real_t *x, +sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const real_t *a, real_t *x, std::int64_t incx, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_tpsv, queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, +sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { throw unimplemented("blas", "tpsv", " for USM"); } -sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, const real_t *a, std::int64_t lda, +sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const real_t *a, std::int64_t lda, real_t *x, std::int64_t incx, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_trmv, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, +sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { throw unimplemented("blas", "trmv", " for USM"); } -sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, const real_t *a, std::int64_t lda, +sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const real_t *a, std::int64_t lda, real_t *x, std::int64_t incx, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_trsv, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, const std::complex *a, +sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { throw unimplemented("blas", "trsv", " for USM"); diff --git a/src/blas/backends/portblas/portblas_level2_double.cpp b/src/blas/backends/portblas/portblas_level2_double.cpp index 67a606cb1..5b26558f0 100644 --- a/src/blas/backends/portblas/portblas_level2_double.cpp +++ b/src/blas/backends/portblas/portblas_level2_double.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace portblas { @@ -56,5 +56,5 @@ constexpr bool is_column_major() { } // namespace row_major } // namespace portblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/portblas/portblas_level2_float.cpp b/src/blas/backends/portblas/portblas_level2_float.cpp index ccb422f39..87c5b7842 100644 --- a/src/blas/backends/portblas/portblas_level2_float.cpp +++ b/src/blas/backends/portblas/portblas_level2_float.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace portblas { @@ -56,5 +56,5 @@ constexpr bool is_column_major() { } // namespace row_major } // namespace portblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/portblas/portblas_level3.cxx b/src/blas/backends/portblas/portblas_level3.cxx index 4eeb1e8f1..9887d270d 100644 --- a/src/blas/backends/portblas/portblas_level3.cxx +++ b/src/blas/backends/portblas/portblas_level3.cxx @@ -19,7 +19,7 @@ // Buffer APIs -void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, real_t beta, sycl::buffer &c, std::int64_t ldc) { @@ -27,14 +27,14 @@ void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transp ldc); } -void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { using sycl_complex_real_t = sycl::ext::oneapi::experimental::complex; - if (transa == oneapi::mkl::transpose::conjtrans || - transb == oneapi::mkl::transpose::conjtrans) { + if (transa == oneapi::math::transpose::conjtrans || + transb == oneapi::math::transpose::conjtrans) { throw unimplemented("blas", "gemm", "Conjugate Transpose unsupported yet on portBLAS"); } // Intermediate buffers for conversion purposes as portBLAS expects sycl::complex instead of std::complex @@ -63,7 +63,7 @@ void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transp queue.copy(out_pb_acc, out_acc); } -void symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, +void symm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, real_t beta, sycl::buffer &c, std::int64_t ldc) { @@ -71,7 +71,7 @@ void symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo up beta, c, ldc); } -void symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, +void symm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -79,7 +79,7 @@ void symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo up throw unimplemented("blas", "symm", ""); } -void hemm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, +void hemm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -87,34 +87,34 @@ void hemm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo up throw unimplemented("blas", "hemm", ""); } -void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer &a, std::int64_t lda, real_t beta, sycl::buffer &c, std::int64_t ldc) { throw unimplemented("blas", "syrk", ""); } -void syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { throw unimplemented("blas", "syrk", ""); } -void herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +void herk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer, 1> &a, std::int64_t lda, real_t beta, sycl::buffer, 1> &c, std::int64_t ldc) { throw unimplemented("blas", "herk", ""); } -void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, real_t beta, sycl::buffer &c, std::int64_t ldc) { throw unimplemented("blas", "syr2k", ""); } -void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -122,7 +122,7 @@ void syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::trans throw unimplemented("blas", "syr2k", ""); } -void her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +void her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, real_t beta, @@ -130,44 +130,44 @@ void her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::trans throw unimplemented("blas", "her2k", ""); } -void trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +void trmm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { throw unimplemented("blas", "trmm", ""); } -void trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +void trmm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { throw unimplemented("blas", "trmm", ""); } -void trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +void trsm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { CALL_PORTBLAS_FN(::blas::_trsm, queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, +void trsm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { throw unimplemented("blas", "trsm", " for complex"); } -void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, real_t alpha, +void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, real_t beta, sycl::buffer &c, std::int64_t ldc) { throw unimplemented("blas", "gemmt", ""); } -void gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, +void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { @@ -228,7 +228,7 @@ void omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_ // USM APIs -sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, real_t alpha, const real_t *a, std::int64_t lda, const real_t *b, std::int64_t ldb, real_t beta, real_t *c, std::int64_t ldc, const std::vector &dependencies) { @@ -236,20 +236,20 @@ sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl: c, ldc, dependencies); } -sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - if (transa == oneapi::mkl::transpose::conjtrans || - transb == oneapi::mkl::transpose::conjtrans) { + if (transa == oneapi::math::transpose::conjtrans || + transb == oneapi::math::transpose::conjtrans) { throw unimplemented("blas", "gemm", "Conjugate Transpose unsupported yet on portBLAS"); } CALL_PORTBLAS_USM_FN(::blas::_gemm, queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, +sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, real_t alpha, const real_t *a, std::int64_t lda, const real_t *b, std::int64_t ldb, real_t beta, real_t *c, std::int64_t ldc, const std::vector &dependencies) { @@ -257,7 +257,7 @@ sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl:: beta, c, ldc, dependencies); } -sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, +sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -265,7 +265,7 @@ sycl::event symm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl:: throw unimplemented("blas", "symm", " for USM"); } -sycl::event hemm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, +sycl::event hemm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -273,14 +273,14 @@ sycl::event hemm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl:: throw unimplemented("blas", "hemm", " for USM"); } -sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, real_t alpha, const real_t *a, std::int64_t lda, real_t beta, real_t *c, std::int64_t ldc, const std::vector &dependencies) { throw unimplemented("blas", "syrk", " for USM"); } -sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, @@ -288,21 +288,21 @@ sycl::event syrk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl: throw unimplemented("blas", "syrk", " for USM"); } -sycl::event herk(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +sycl::event herk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, real_t alpha, const std::complex *a, std::int64_t lda, real_t beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { throw unimplemented("blas", "herk", " for USM"); } -sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, real_t alpha, const real_t *a, std::int64_t lda, const real_t *b, std::int64_t ldb, real_t beta, real_t *c, std::int64_t ldc, const std::vector &dependencies) { throw unimplemented("blas", "syr2k", " for USM"); } -sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -310,7 +310,7 @@ sycl::event syr2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl throw unimplemented("blas", "syr2k", " for USM"); } -sycl::event her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, +sycl::event her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, real_t beta, std::complex *c, std::int64_t ldc, @@ -318,46 +318,46 @@ sycl::event her2k(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl throw unimplemented("blas", "her2k", " for USM"); } -sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, +sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, real_t alpha, const real_t *a, std::int64_t lda, real_t *b, std::int64_t ldb, const std::vector &dependencies) { throw unimplemented("blas", "trmm", " for USM"); } -sycl::event trmm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, +sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { throw unimplemented("blas", "trmm", " for USM"); } -sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, +sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, real_t alpha, const real_t *a, std::int64_t lda, real_t *b, std::int64_t ldb, const std::vector &dependencies) { CALL_PORTBLAS_USM_FN(::blas::_trsm, queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm(sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, +sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { throw unimplemented("blas", "trsm", " for USM"); } -sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, real_t alpha, +sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, real_t alpha, const real_t *a, std::int64_t lda, const real_t *b, std::int64_t ldb, real_t beta, real_t *c, std::int64_t ldc, const std::vector &dependencies) { throw unimplemented("blas", "gemmt", " for USM"); } -sycl::event gemmt(sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, +sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, diff --git a/src/blas/backends/portblas/portblas_level3_bfloat16.cpp b/src/blas/backends/portblas/portblas_level3_bfloat16.cpp index f0ff06452..d432647d6 100644 --- a/src/blas/backends/portblas/portblas_level3_bfloat16.cpp +++ b/src/blas/backends/portblas/portblas_level3_bfloat16.cpp @@ -27,24 +27,24 @@ #include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace portblas { namespace column_major { // BUFFER -void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, + sycl::buffer &a, std::int64_t lda, + sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { throw unimplemented("blas", "gemm", " for bfloat16"); } // USM -sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const oneapi::mkl::bfloat16 *a, std::int64_t lda, const oneapi::mkl::bfloat16 *b, + const oneapi::math::bfloat16 *a, std::int64_t lda, const oneapi::math::bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { throw unimplemented("blas", "gemm", " for USM"); @@ -54,18 +54,18 @@ sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl: namespace row_major { // BUFFER -void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, + sycl::buffer &a, std::int64_t lda, + sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { throw unimplemented("blas", "gemm", " for bfloat16"); } // USM -sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const oneapi::mkl::bfloat16 *a, std::int64_t lda, const oneapi::mkl::bfloat16 *b, + const oneapi::math::bfloat16 *a, std::int64_t lda, const oneapi::math::bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { throw unimplemented("blas", "gemm", " for USM"); @@ -74,5 +74,5 @@ sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl: } // namespace row_major } // namespace portblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/portblas/portblas_level3_double.cpp b/src/blas/backends/portblas/portblas_level3_double.cpp index 1ca4beb4b..80f2d0bce 100644 --- a/src/blas/backends/portblas/portblas_level3_double.cpp +++ b/src/blas/backends/portblas/portblas_level3_double.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace portblas { @@ -56,5 +56,5 @@ constexpr bool is_column_major() { } // namespace row_major } // namespace portblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/portblas/portblas_level3_float.cpp b/src/blas/backends/portblas/portblas_level3_float.cpp index d0168f7d6..dea42ed9d 100644 --- a/src/blas/backends/portblas/portblas_level3_float.cpp +++ b/src/blas/backends/portblas/portblas_level3_float.cpp @@ -28,7 +28,7 @@ #include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace portblas { @@ -58,5 +58,5 @@ constexpr bool is_column_major() { } // namespace row_major } // namespace portblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/portblas/portblas_level3_half.cpp b/src/blas/backends/portblas/portblas_level3_half.cpp index c262f9e5b..65570e5d9 100644 --- a/src/blas/backends/portblas/portblas_level3_half.cpp +++ b/src/blas/backends/portblas/portblas_level3_half.cpp @@ -27,20 +27,20 @@ #include "oneapi/math/blas/detail/portblas/onemath_blas_portblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace portblas { namespace column_major { // BUFFER -void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { throw unimplemented("blas", "gemm", " half"); } -void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -48,7 +48,7 @@ void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transp } // USM -sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, @@ -56,7 +56,7 @@ sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl: throw unimplemented("blas", "gemm", " for USM"); } -sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -66,14 +66,14 @@ sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl: namespace row_major { // BUFFER -void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { throw unimplemented("blas", "gemm", " half"); } -void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -81,7 +81,7 @@ void gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transp } // USM -sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, @@ -89,7 +89,7 @@ sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl: throw unimplemented("blas", "gemm", " for USM"); } -sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, +sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -99,5 +99,5 @@ sycl::event gemm(sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl: } // namespace row_major } // namespace portblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/rocblas/rocblas_batch.cpp b/src/blas/backends/rocblas/rocblas_batch.cpp index 5e8be4f29..64338b2a3 100644 --- a/src/blas/backends/rocblas/rocblas_batch.cpp +++ b/src/blas/backends/rocblas/rocblas_batch.cpp @@ -64,7 +64,7 @@ static inline void conj_vector(sycl::handler &cgh, T **ptr, const int64_t len, c } namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { namespace column_major { @@ -394,8 +394,8 @@ inline void omatcopy_batch(Func func, sycl::queue &queue, transpose trans, int64 overflow_check(m, n, lda, ldb, stridea, strideb, batch_size); const T beta = 0; - const int64_t new_m = trans == oneapi::mkl::transpose::nontrans ? m : n; - const int64_t new_n = trans == oneapi::mkl::transpose::nontrans ? n : m; + const int64_t new_m = trans == oneapi::math::transpose::nontrans ? m : n; + const int64_t new_n = trans == oneapi::math::transpose::nontrans ? n : m; queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); @@ -1205,8 +1205,8 @@ inline sycl::event omatcopy_batch(Func func, sycl::queue &queue, transpose trans overflow_check(m, n, lda, ldb, stridea, strideb, batch_size); const T beta = 0; - const int64_t new_m = trans == oneapi::mkl::transpose::nontrans ? m : n; - const int64_t new_n = trans == oneapi::mkl::transpose::nontrans ? n : m; + const int64_t new_m = trans == oneapi::math::transpose::nontrans ? m : n; + const int64_t new_n = trans == oneapi::math::transpose::nontrans ? n : m; auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); @@ -1335,8 +1335,8 @@ inline sycl::event omatcopy_batch(Func func, sycl::queue &queue, transpose *tran auto **b_ = reinterpret_cast(b); const T beta = 0; - const auto new_m = trans[i] == oneapi::mkl::transpose::nontrans ? m[i] : n[i]; - const auto new_n = trans[i] == oneapi::mkl::transpose::nontrans ? n[i] : m[i]; + const auto new_m = trans[i] == oneapi::math::transpose::nontrans ? m[i] : n[i]; + const auto new_n = trans[i] == oneapi::math::transpose::nontrans ? n[i] : m[i]; rocblas_native_func(func, err, handle, get_rocblas_operation(trans[i]), get_rocblas_operation(trans[i]), (int)new_m, (int)new_n, @@ -1450,10 +1450,10 @@ inline void gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m int64_t stridea, sycl::buffer, 1> &x, int64_t incx, int64_t stridex, std::complex beta, sycl::buffer, 1> &y, int64_t incy, int64_t stridey, int64_t batch_size) { - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); beta = std::conj(beta); @@ -1471,7 +1471,7 @@ inline void gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m column_major::gemv_batch(func, queue, new_trans, n, m, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit( [&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy, stridey, batch_size); }); @@ -1484,8 +1484,8 @@ inline void gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m T alpha, sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &x, int64_t incx, int64_t stridex, T beta, sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::gemv_batch(func, queue, new_trans, n, m, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); @@ -1513,7 +1513,7 @@ inline void dgmm_batch(Func func, sycl::queue &queue, side left_right, int64_t m int64_t incx, int64_t stridex, sycl::buffer &c, int64_t ldc, int64_t stridec, int64_t batch_size) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; column_major::dgmm_batch(func, queue, new_side, n, m, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); @@ -1593,9 +1593,9 @@ inline void trsm_batch(Func func, sycl::queue &queue, side left_right, uplo uppe sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &b, int64_t ldb, int64_t strideb, int64_t batch_size) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::trsm_batch(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, stridea, b, ldb, strideb, batch_size); @@ -1622,10 +1622,10 @@ inline void syrk_batch(Func func, sycl::queue &queue, uplo upper_lower, transpos int64_t k, T alpha, sycl::buffer &a, int64_t lda, int64_t stridea, T beta, sycl::buffer &c, int64_t ldc, int64_t stridec, int64_t batch_size) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::syrk_batch(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, stridea, beta, c, ldc, stridec, batch_size); @@ -1825,10 +1825,10 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose trans, in const std::vector &dependencies) { sycl::event done; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); beta = std::conj(beta); @@ -1849,7 +1849,7 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose trans, in done = column_major::gemv_batch(func, queue, new_trans, n, m, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(done); @@ -1867,8 +1867,8 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose trans, in int64_t incx, int64_t stridex, T beta, T *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::gemv_batch(func, queue, new_trans, n, m, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); @@ -1902,7 +1902,7 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, i int64_t stride = 0; for (int64_t i = 0; i < group_count; i++) { - if (trans[i] == oneapi::mkl::transpose::conjtrans) { + if (trans[i] == oneapi::math::transpose::conjtrans) { alpha[i] = std::conj(alpha[i]); beta[i] = std::conj(beta[i]); @@ -1925,9 +1925,9 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, i auto tmp_trans = std::vector{ (std::size_t)group_count }; for (int64_t i = 0; i < group_count; i++) { - const auto new_trans = trans[i] == oneapi::mkl::transpose::nontrans - ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + const auto new_trans = trans[i] == oneapi::math::transpose::nontrans + ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; tmp_trans[i] = trans[i]; trans[i] = new_trans; } @@ -1940,7 +1940,7 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, i stride = 0; for (int64_t i = 0; i < group_count; i++) { - if (trans[i] == oneapi::mkl::transpose::conjtrans) { + if (trans[i] == oneapi::math::transpose::conjtrans) { if (n[i] > 0) { done = queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n[i], incy[i], stride, group_size[i]); @@ -1961,9 +1961,9 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, i auto tmp_trans = std::vector{ static_cast(group_count) }; for (int64_t i = 0; i < group_count; i++) { - const auto new_trans = trans[i] == oneapi::mkl::transpose::nontrans - ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + const auto new_trans = trans[i] == oneapi::math::transpose::nontrans + ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; tmp_trans[i] = trans[i]; trans[i] = new_trans; } @@ -1999,7 +1999,7 @@ inline sycl::event dgmm_batch(Func func, sycl::queue &queue, side left_right, in int64_t stridex, T *c, int64_t ldc, int64_t stridec, int64_t batch_size, const std::vector &dependencies) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; return column_major::dgmm_batch(func, queue, new_side, n, m, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); @@ -2027,8 +2027,8 @@ inline sycl::event dgmm_batch(Func func, sycl::queue &queue, side *left_right, i T **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { for (int64_t i = 0; i < group_count; i++) { - const auto new_side = left_right[i] == oneapi::mkl::side::left ? oneapi::mkl::side::right - : oneapi::mkl::side::left; + const auto new_side = left_right[i] == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; left_right[i] = new_side; } @@ -2166,9 +2166,9 @@ inline sycl::event trsm_batch(Func func, sycl::queue &queue, side left_right, up int64_t strideb, int64_t batch_size, const std::vector &dependencies) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::trsm_batch(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, stridea, b, ldb, strideb, batch_size, dependencies); @@ -2196,12 +2196,12 @@ inline sycl::event trsm_batch(Func func, sycl::queue &queue, side *left_right, u const T **a, int64_t *lda, T **b, int64_t *ldb, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { for (int64_t i = 0; i < group_count; i++) { - const auto new_side = left_right[i] == oneapi::mkl::side::left ? oneapi::mkl::side::right - : oneapi::mkl::side::left; + const auto new_side = left_right[i] == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; left_right[i] = new_side; - const auto new_uplo = upper_lower[i] == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + const auto new_uplo = upper_lower[i] == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; upper_lower[i] = new_uplo; } @@ -2232,13 +2232,13 @@ inline sycl::event syrk_batch(Func func, sycl::queue &queue, uplo *upper_lower, T **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { for (int64_t i = 0; i < group_count; i++) { - const auto new_uplo = upper_lower[i] == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + const auto new_uplo = upper_lower[i] == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; upper_lower[i] = new_uplo; - const auto new_trans = trans[i] == oneapi::mkl::transpose::nontrans - ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + const auto new_trans = trans[i] == oneapi::math::transpose::nontrans + ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; trans[i] = new_trans; } @@ -2267,10 +2267,10 @@ inline sycl::event syrk_batch(Func func, sycl::queue &queue, uplo upper_lower, t int64_t n, int64_t k, const T alpha, const T *a, int64_t lda, int64_t stridea, const T beta, T *c, int64_t ldc, int64_t stridec, int64_t batch_size, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::syrk_batch(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, stridea, beta, c, ldc, stridec, batch_size, dependencies); @@ -2429,5 +2429,5 @@ sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int } // namespace row_major } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/rocblas/rocblas_extensions.cpp b/src/blas/backends/rocblas/rocblas_extensions.cpp index 7d4343f4d..a0b7d56e8 100644 --- a/src/blas/backends/rocblas/rocblas_extensions.cpp +++ b/src/blas/backends/rocblas/rocblas_extensions.cpp @@ -26,7 +26,7 @@ #include "oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { namespace column_major { @@ -97,8 +97,8 @@ inline void omatcopy(Func func, sycl::queue &queue, transpose trans, int64_t m, overflow_check(m, n, lda, ldb); const T beta = 0; - const int64_t new_m = trans == oneapi::mkl::transpose::nontrans ? m : n; - const int64_t new_n = trans == oneapi::mkl::transpose::nontrans ? n : m; + const int64_t new_m = trans == oneapi::math::transpose::nontrans ? m : n; + const int64_t new_n = trans == oneapi::math::transpose::nontrans ? n : m; queue.submit([&](sycl::handler &cgh) { auto a_acc = a.template get_access(cgh); @@ -284,8 +284,8 @@ inline sycl::event omatcopy(Func func, sycl::queue &queue, transpose trans, int6 overflow_check(m, n, lda, ldb); const T beta = 0; - const int64_t new_m = trans == oneapi::mkl::transpose::nontrans ? m : n; - const int64_t new_n = trans == oneapi::mkl::transpose::nontrans ? n : m; + const int64_t new_m = trans == oneapi::math::transpose::nontrans ? m : n; + const int64_t new_n = trans == oneapi::math::transpose::nontrans ? n : m; auto done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(dependencies); @@ -712,5 +712,5 @@ OMATADD_LAUNCHER_USM(std::complex, rocblas_zgeam) } // namespace row_major } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/rocblas/rocblas_handle.hpp b/src/blas/backends/rocblas/rocblas_handle.hpp index 7a8dfe91f..66d348619 100644 --- a/src/blas/backends/rocblas/rocblas_handle.hpp +++ b/src/blas/backends/rocblas/rocblas_handle.hpp @@ -24,7 +24,7 @@ #include "rocblas_helper.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { @@ -57,7 +57,7 @@ struct rocblas_handle_ { } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ROCBLAS_HANDLE_HPP_ diff --git a/src/blas/backends/rocblas/rocblas_helper.hpp b/src/blas/backends/rocblas/rocblas_helper.hpp index 261518951..f21d63655 100644 --- a/src/blas/backends/rocblas/rocblas_helper.hpp +++ b/src/blas/backends/rocblas/rocblas_helper.hpp @@ -34,7 +34,7 @@ #include "dtype_string.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { @@ -183,35 +183,35 @@ inline void rocblas_native_func(Func func, rocblas_status err, #endif }; -inline rocblas_operation get_rocblas_operation(oneapi::mkl::transpose trn) { +inline rocblas_operation get_rocblas_operation(oneapi::math::transpose trn) { switch (trn) { - case oneapi::mkl::transpose::nontrans: return rocblas_operation_none; - case oneapi::mkl::transpose::trans: return rocblas_operation_transpose; - case oneapi::mkl::transpose::conjtrans: return rocblas_operation_conjugate_transpose; + case oneapi::math::transpose::nontrans: return rocblas_operation_none; + case oneapi::math::transpose::trans: return rocblas_operation_transpose; + case oneapi::math::transpose::conjtrans: return rocblas_operation_conjugate_transpose; default: throw "Wrong transpose Operation."; } } -inline rocblas_fill get_rocblas_fill_mode(oneapi::mkl::uplo ul) { +inline rocblas_fill get_rocblas_fill_mode(oneapi::math::uplo ul) { switch (ul) { - case oneapi::mkl::uplo::upper: return rocblas_fill_upper; - case oneapi::mkl::uplo::lower: return rocblas_fill_lower; + case oneapi::math::uplo::upper: return rocblas_fill_upper; + case oneapi::math::uplo::lower: return rocblas_fill_lower; default: throw "Wrong fill mode."; } } -inline rocblas_diagonal get_rocblas_diag_type(oneapi::mkl::diag un) { +inline rocblas_diagonal get_rocblas_diag_type(oneapi::math::diag un) { switch (un) { - case oneapi::mkl::diag::unit: return rocblas_diagonal_unit; - case oneapi::mkl::diag::nonunit: return rocblas_diagonal_non_unit; + case oneapi::math::diag::unit: return rocblas_diagonal_unit; + case oneapi::math::diag::nonunit: return rocblas_diagonal_non_unit; default: throw "Wrong diag type."; } } -inline rocblas_side get_rocblas_side_mode(oneapi::mkl::side lr) { +inline rocblas_side get_rocblas_side_mode(oneapi::math::side lr) { switch (lr) { - case oneapi::mkl::side::left: return rocblas_side_left; - case oneapi::mkl::side::right: return rocblas_side_right; + case oneapi::math::side::left: return rocblas_side_left; + case oneapi::math::side::right: return rocblas_side_right; default: throw "Wrong side mode."; } } @@ -298,6 +298,6 @@ struct RocEquivalentType { } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ROCBLAS_HELPER_HPP_ diff --git a/src/blas/backends/rocblas/rocblas_level1.cpp b/src/blas/backends/rocblas/rocblas_level1.cpp index 69f0e2883..b5966108d 100644 --- a/src/blas/backends/rocblas/rocblas_level1.cpp +++ b/src/blas/backends/rocblas/rocblas_level1.cpp @@ -26,7 +26,7 @@ #include "oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { namespace column_major { @@ -1778,5 +1778,5 @@ NRM2_LAUNCHER_USM(std::complex, double, rocblas_dznrm2) } // namespace row_major } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/rocblas/rocblas_level2.cpp b/src/blas/backends/rocblas/rocblas_level2.cpp index c887fbd61..dd62a8558 100644 --- a/src/blas/backends/rocblas/rocblas_level2.cpp +++ b/src/blas/backends/rocblas/rocblas_level2.cpp @@ -74,7 +74,7 @@ static inline void conj_vector(sycl::handler &cgh, T *ptr_a, T *ptr_b, const int } namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { namespace column_major { @@ -1810,10 +1810,10 @@ inline void gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int6 std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); beta = std::conj(beta); @@ -1828,7 +1828,7 @@ inline void gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int6 column_major::gemv(func, queue, new_trans, n, m, alpha, a, lda, x, incx, beta, y, incy); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); } @@ -1839,8 +1839,8 @@ template inline void gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, sycl::buffer &y, int64_t incy) { - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::gemv(func, queue, new_trans, n, m, alpha, a, lda, x, incx, beta, y, incy); } @@ -1864,10 +1864,10 @@ inline void gbmv(Func func, sycl::queue &queue, transpose trans, int64_t m, int6 int64_t ku, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx, std::complex beta, sycl::buffer, 1> &y, int64_t incy) { - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); beta = std::conj(beta); @@ -1882,7 +1882,7 @@ inline void gbmv(Func func, sycl::queue &queue, transpose trans, int64_t m, int6 column_major::gbmv(func, queue, new_trans, n, m, ku, kl, alpha, a, lda, x, incx, beta, y, incy); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); } @@ -1893,8 +1893,8 @@ template inline void gbmv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, sycl::buffer &y, int64_t incy) { - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::gbmv(func, queue, new_trans, n, m, ku, kl, alpha, a, lda, x, incx, beta, y, incy); } @@ -1960,8 +1960,8 @@ template inline void hbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, sycl::buffer &y, int64_t incy) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); @@ -1992,8 +1992,8 @@ template inline void hemv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, sycl::buffer &y, int64_t incy) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); @@ -2024,8 +2024,8 @@ template inline void her(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, ScalarType alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); @@ -2050,8 +2050,8 @@ template inline void her2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, int64_t lda) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, y, n, incx, incy); }); @@ -2076,8 +2076,8 @@ template inline void hpmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, sycl::buffer &a, sycl::buffer &x, int64_t incx, T beta, sycl::buffer &y, int64_t incy) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); @@ -2107,8 +2107,8 @@ HPMV_LAUNCHER(std::complex, rocblas_zhpmv) template inline void hpr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, ScalarType alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); @@ -2132,8 +2132,8 @@ template inline void hpr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, y, n, incx, incy); }); @@ -2158,8 +2158,8 @@ template inline void sbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, sycl::buffer &y, int64_t incy) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::sbmv(func, queue, new_uplo, n, k, alpha, a, lda, x, incx, beta, y, incy); } @@ -2180,8 +2180,8 @@ template inline void symv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, sycl::buffer &y, int64_t incy) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::symv(func, queue, new_uplo, n, alpha, a, lda, x, incx, beta, y, incy); } @@ -2201,8 +2201,8 @@ SYMV_LAUNCHER(double, rocblas_dsymv) template inline void syr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::syr(func, queue, new_uplo, n, alpha, x, incx, a, lda); } @@ -2225,8 +2225,8 @@ template inline void syr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, int64_t lda) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::syr2(func, queue, new_uplo, n, alpha, x, incx, y, incy, a, lda); } @@ -2250,8 +2250,8 @@ template inline void spmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, sycl::buffer &a, sycl::buffer &x, int64_t incx, T beta, sycl::buffer &y, int64_t incy) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::spmv(func, queue, new_uplo, n, alpha, a, x, incx, beta, y, incy); } @@ -2271,8 +2271,8 @@ SPMV_LAUNCHER(double, rocblas_dspmv) template inline void spr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::spr(func, queue, new_uplo, n, alpha, x, incx, a); } @@ -2292,8 +2292,8 @@ template inline void spr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::spr2(func, queue, new_uplo, n, alpha, x, incx, y, incy, a); } @@ -2314,12 +2314,12 @@ template inline void tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2327,7 +2327,7 @@ inline void tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran column_major::tbmv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2338,10 +2338,10 @@ template inline void tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::tbmv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx); } @@ -2364,12 +2364,12 @@ template inline void tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2377,7 +2377,7 @@ inline void tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran column_major::tbsv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2388,10 +2388,10 @@ template inline void tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::tbsv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx); } @@ -2414,12 +2414,12 @@ template inline void tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2427,7 +2427,7 @@ inline void tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran column_major::tpmv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2437,10 +2437,10 @@ inline void tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran template inline void tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, sycl::buffer &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::tpmv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx); } @@ -2462,12 +2462,12 @@ template inline void tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2475,7 +2475,7 @@ inline void tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran column_major::tpsv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2485,10 +2485,10 @@ inline void tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran template inline void tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, sycl::buffer &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::tpsv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx); } @@ -2510,12 +2510,12 @@ template inline void trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2523,7 +2523,7 @@ inline void trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran column_major::trmv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2534,10 +2534,10 @@ template inline void trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::trmv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx); } @@ -2559,12 +2559,12 @@ template inline void trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2572,7 +2572,7 @@ inline void trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran column_major::trsv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); } @@ -2583,10 +2583,10 @@ template inline void trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::trsv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx); } @@ -2614,10 +2614,10 @@ inline sycl::event gemv(Func func, sycl::queue &queue, transpose trans, int64_t const std::vector &dependencies) { sycl::event done; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); beta = std::conj(beta); @@ -2636,7 +2636,7 @@ inline sycl::event gemv(Func func, sycl::queue &queue, transpose trans, int64_t done = column_major::gemv(func, queue, new_trans, n, m, alpha, a, lda, x, incx, beta, y, incy, dependencies); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(done); @@ -2652,8 +2652,8 @@ template inline sycl::event gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, T alpha, const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, int64_t incy, const std::vector &dependencies) { - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::gemv(func, queue, new_trans, n, m, alpha, a, lda, x, incx, beta, y, incy, dependencies); @@ -2682,10 +2682,10 @@ inline sycl::event gbmv(Func func, sycl::queue &queue, transpose trans, int64_t const std::vector &dependencies) { sycl::event done; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); beta = std::conj(beta); @@ -2704,7 +2704,7 @@ inline sycl::event gbmv(Func func, sycl::queue &queue, transpose trans, int64_t done = column_major::gbmv(func, queue, new_trans, n, m, ku, kl, alpha, a, lda, x, incx, beta, y, incy, dependencies); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(done); @@ -2721,8 +2721,8 @@ inline sycl::event gbmv(Func func, sycl::queue &queue, transpose trans, int64_t int64_t kl, int64_t ku, T alpha, const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, int64_t incy, const std::vector &dependencies) { - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::gbmv(func, queue, new_trans, n, m, ku, kl, alpha, a, lda, x, incx, beta, y, incy, dependencies); @@ -2795,8 +2795,8 @@ inline sycl::event hbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t int64_t incy, const std::vector &dependencies) { sycl::event done; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); @@ -2837,8 +2837,8 @@ inline sycl::event hemv(Func func, sycl::queue &queue, uplo upper_lower, int64_t int64_t incy, const std::vector &dependencies) { sycl::event done; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); @@ -2877,8 +2877,8 @@ template inline sycl::event her(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, const ScalarType alpha, const DataType *x, int64_t incx, DataType *a, int64_t lda, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (DataType *)x, n, incx); }) @@ -2904,8 +2904,8 @@ template inline sycl::event her2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, T *a, int64_t lda, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (T *)x, (T *)y, n, incx, incy); }) @@ -2935,8 +2935,8 @@ inline sycl::event hpmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t const std::vector &dependencies) { sycl::event done; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); @@ -2975,8 +2975,8 @@ template inline sycl::event hpr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, const ScalarType alpha, const DataType *x, int64_t incx, DataType *a, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (DataType *)x, n, incx); }) @@ -3002,8 +3002,8 @@ template inline sycl::event hpr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, T *a, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (T *)x, (T *)y, n, incx, incy); }) @@ -3030,8 +3030,8 @@ template inline sycl::event sbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, T alpha, const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, int64_t incy, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::sbmv(func, queue, new_uplo, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); @@ -3054,8 +3054,8 @@ template inline sycl::event symv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, int64_t incy, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::symv(func, queue, new_uplo, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); @@ -3078,8 +3078,8 @@ template inline sycl::event syr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, const T *x, int64_t incx, T *a, int64_t lda, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::syr(func, queue, new_uplo, n, alpha, x, incx, a, lda, dependencies); } @@ -3103,8 +3103,8 @@ template inline sycl::event syr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, T *a, int64_t lda, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::syr2(func, queue, new_uplo, n, alpha, x, incx, y, incy, a, lda, dependencies); @@ -3130,8 +3130,8 @@ template inline sycl::event spmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, const T *a, const T *x, int64_t incx, T beta, T *y, int64_t incy, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::spmv(func, queue, new_uplo, n, alpha, a, x, incx, beta, y, incy, dependencies); @@ -3154,8 +3154,8 @@ template inline sycl::event spr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, const T *x, int64_t incx, T *a, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::spr(func, queue, new_uplo, n, alpha, x, incx, a, dependencies); } @@ -3175,8 +3175,8 @@ template inline sycl::event spr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, T *a, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::spr2(func, queue, new_uplo, n, alpha, x, incx, y, incy, a, dependencies); } @@ -3201,12 +3201,12 @@ inline sycl::event tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpo const std::vector &dependencies) { sycl::event done; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); @@ -3216,7 +3216,7 @@ inline sycl::event tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpo done = column_major::tbmv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx, dependencies); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(done); @@ -3232,10 +3232,10 @@ template inline sycl::event tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, const T *a, int64_t lda, T *x, int64_t incx, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::tbmv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx, dependencies); @@ -3263,12 +3263,12 @@ inline sycl::event tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpo const std::vector &dependencies) { sycl::event done; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); @@ -3278,7 +3278,7 @@ inline sycl::event tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpo done = column_major::tbsv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx, dependencies); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(done); @@ -3294,10 +3294,10 @@ template inline sycl::event tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, int64_t k, const T *a, int64_t lda, T *x, int64_t incx, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::tbsv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx, dependencies); @@ -3324,12 +3324,12 @@ inline sycl::event tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpo int64_t incx, const std::vector &dependencies) { sycl::event done; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); @@ -3339,7 +3339,7 @@ inline sycl::event tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpo done = column_major::tpmv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx, dependencies); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { incx = std::abs(incx); @@ -3357,10 +3357,10 @@ template inline sycl::event tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const T *a, T *x, int64_t incx, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::tpmv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx, dependencies); @@ -3387,12 +3387,12 @@ inline sycl::event tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpo int64_t incx, const std::vector &dependencies) { sycl::event done; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); @@ -3402,7 +3402,7 @@ inline sycl::event tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpo done = column_major::tpsv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx, dependencies); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { incx = std::abs(incx); @@ -3420,10 +3420,10 @@ template inline sycl::event tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const T *a, T *x, int64_t incx, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::tpsv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx, dependencies); @@ -3451,12 +3451,12 @@ inline sycl::event trmv(Func func, sycl::queue &queue, uplo upper_lower, transpo const std::vector &dependencies) { sycl::event done; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); @@ -3466,7 +3466,7 @@ inline sycl::event trmv(Func func, sycl::queue &queue, uplo upper_lower, transpo done = column_major::trmv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx, dependencies); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(done); @@ -3482,10 +3482,10 @@ template inline sycl::event trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const T *a, int64_t lda, T *x, int64_t incx, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::trmv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx, dependencies); @@ -3513,12 +3513,12 @@ inline sycl::event trsv(Func func, sycl::queue &queue, uplo upper_lower, transpo const std::vector &dependencies) { sycl::event done; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); @@ -3528,7 +3528,7 @@ inline sycl::event trsv(Func func, sycl::queue &queue, uplo upper_lower, transpo done = column_major::trsv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx, dependencies); - if (trans == oneapi::mkl::transpose::conjtrans) { + if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { done = queue.submit([&](sycl::handler &cgh) { cgh.depends_on(done); @@ -3544,10 +3544,10 @@ template inline sycl::event trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, const T *a, int64_t lda, T *x, int64_t incx, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::trsv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx, dependencies); @@ -3571,5 +3571,5 @@ TRSV_LAUNCHER_USM(std::complex, rocblas_ztrsv) } // namespace row_major } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/rocblas/rocblas_level3.cpp b/src/blas/backends/rocblas/rocblas_level3.cpp index 085df77ed..904d83114 100644 --- a/src/blas/backends/rocblas/rocblas_level3.cpp +++ b/src/blas/backends/rocblas/rocblas_level3.cpp @@ -26,7 +26,7 @@ #include "oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { namespace column_major { @@ -960,9 +960,9 @@ inline void symm(Func func, sycl::queue &queue, side left_right, uplo upper_lowe int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, int64_t ldc) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::symm(func, queue, new_side, new_uplo, n, m, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -987,9 +987,9 @@ inline void hemm(Func func, sycl::queue &queue, side left_right, uplo upper_lowe int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, int64_t ldc) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::hemm(func, queue, new_side, new_uplo, n, m, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -1011,10 +1011,10 @@ template inline void syrk(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, T alpha, sycl::buffer &a, int64_t lda, T beta, sycl::buffer &c, int64_t ldc) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::syrk(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -1037,10 +1037,10 @@ template inline void herk(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, ScalarType alpha, sycl::buffer &a, int64_t lda, ScalarType beta, sycl::buffer &c, int64_t ldc) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::conjtrans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::conjtrans + : oneapi::math::transpose::nontrans; column_major::herk(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, beta, c, ldc); } @@ -1061,10 +1061,10 @@ template inline void syr2k(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, int64_t ldc) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; column_major::syr2k(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); @@ -1090,10 +1090,10 @@ inline void her2k(Func func, sycl::queue &queue, uplo upper_lower, transpose tra int64_t k, DataType alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, ScalarType beta, sycl::buffer &c, int64_t ldc) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::conjtrans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::conjtrans + : oneapi::math::transpose::nontrans; auto new_alpha = std::conj(alpha); column_major::her2k(func, queue, new_uplo, new_trans, n, k, new_alpha, a, lda, b, ldb, beta, c, @@ -1123,9 +1123,9 @@ inline void trmm(Func func, sycl::queue &queue, side left_right, uplo upper_lowe diag unit_diag, int64_t m, int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::trmm(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, b, ldb); @@ -1151,9 +1151,9 @@ inline void trsm(Func func, sycl::queue &queue, side left_right, uplo upper_lowe diag unit_diag, int64_t m, int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; column_major::trsm(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, b, ldb); @@ -1247,9 +1247,9 @@ inline sycl::event symm(Func func, sycl::queue &queue, side left_right, uplo upp int64_t n, T alpha, const T *a, int64_t lda, const T *b, int64_t ldb, T beta, T *c, int64_t ldc, const std::vector &dependencies) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::symm(func, queue, new_side, new_uplo, n, m, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); @@ -1276,9 +1276,9 @@ inline sycl::event hemm(Func func, sycl::queue &queue, side left_right, uplo upp int64_t n, T alpha, const T *a, int64_t lda, const T *b, int64_t ldb, T beta, T *c, int64_t ldc, const std::vector &dependencies) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::hemm(func, queue, new_side, new_uplo, n, m, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); @@ -1302,10 +1302,10 @@ template inline sycl::event syrk(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, T alpha, const T *a, int64_t lda, T beta, T *c, int64_t ldc, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::syrk(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); @@ -1331,10 +1331,10 @@ inline sycl::event herk(Func func, sycl::queue &queue, uplo upper_lower, transpo int64_t k, const ScalarType alpha, const DataType *a, int64_t lda, const ScalarType beta, DataType *c, int64_t ldc, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::conjtrans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::conjtrans + : oneapi::math::transpose::nontrans; return column_major::herk(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); @@ -1359,10 +1359,10 @@ inline sycl::event syr2k(Func func, sycl::queue &queue, uplo upper_lower, transp int64_t n, int64_t k, T alpha, const T *a, int64_t lda, const T *b, int64_t ldb, T beta, T *c, int64_t ldc, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans + : oneapi::math::transpose::nontrans; return column_major::syr2k(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); @@ -1389,10 +1389,10 @@ inline sycl::event her2k(Func func, sycl::queue &queue, uplo upper_lower, transp int64_t n, int64_t k, const DataType alpha, const DataType *a, int64_t lda, const DataType *b, int64_t ldb, const ScalarType beta, DataType *c, int64_t ldc, const std::vector &dependencies) { - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; - auto new_trans = trans == oneapi::mkl::transpose::nontrans ? oneapi::mkl::transpose::conjtrans - : oneapi::mkl::transpose::nontrans; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; + auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::conjtrans + : oneapi::math::transpose::nontrans; auto new_alpha = std::conj(alpha); return column_major::her2k(func, queue, new_uplo, new_trans, n, k, new_alpha, a, lda, b, ldb, @@ -1423,9 +1423,9 @@ inline sycl::event trmm(Func func, sycl::queue &queue, side left_right, uplo upp int64_t lda, T *b, int64_t ldb, const std::vector &dependencies) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::trmm(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, b, ldb, dependencies); @@ -1452,9 +1452,9 @@ inline sycl::event trsm(Func func, sycl::queue &queue, side left_right, uplo upp int64_t lda, T *b, int64_t ldb, const std::vector &dependencies) { auto new_side = - left_right == oneapi::mkl::side::left ? oneapi::mkl::side::right : oneapi::mkl::side::left; - auto new_uplo = upper_lower == oneapi::mkl::uplo::lower ? oneapi::mkl::uplo::upper - : oneapi::mkl::uplo::lower; + left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; return column_major::trsm(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, b, ldb, dependencies); @@ -1478,5 +1478,5 @@ TRSM_LAUNCHER_USM(std::complex, rocblas_ztrsm) } // namespace row_major } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/rocblas/rocblas_scope_handle.cpp b/src/blas/backends/rocblas/rocblas_scope_handle.cpp index ecec4dded..65187223e 100644 --- a/src/blas/backends/rocblas/rocblas_scope_handle.cpp +++ b/src/blas/backends/rocblas/rocblas_scope_handle.cpp @@ -20,7 +20,7 @@ #include "rocblas_scope_handle.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { @@ -163,5 +163,5 @@ sycl::context RocblasScopedContextHandler::get_context(const sycl::queue &queue) } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/blas/backends/rocblas/rocblas_scope_handle.hpp b/src/blas/backends/rocblas/rocblas_scope_handle.hpp index 50f650299..c08db4a78 100644 --- a/src/blas/backends/rocblas/rocblas_scope_handle.hpp +++ b/src/blas/backends/rocblas/rocblas_scope_handle.hpp @@ -39,7 +39,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { @@ -79,6 +79,6 @@ class RocblasScopedContextHandler { } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ROCBLAS_SCOPED_HANDLE_HPP_ diff --git a/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.cpp b/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.cpp index da9791411..bc23cf26a 100644 --- a/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.cpp +++ b/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.cpp @@ -22,7 +22,7 @@ #include "rocblas_scope_handle_hipsycl.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { @@ -90,5 +90,5 @@ hipStream_t RocblasScopedContextHandler::get_stream(const sycl::queue &queue) { } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi \ No newline at end of file diff --git a/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.hpp b/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.hpp index 3c156ab6c..d6fe2a7e7 100644 --- a/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.hpp +++ b/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.hpp @@ -30,7 +30,7 @@ #include #include "rocblas_helper.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { @@ -61,6 +61,6 @@ class RocblasScopedContextHandler { } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ROCBLAS_SCOPED_HANDLE_HPP_ diff --git a/src/blas/backends/rocblas/rocblas_task.hpp b/src/blas/backends/rocblas/rocblas_task.hpp index a74063ad8..bf1197897 100644 --- a/src/blas/backends/rocblas/rocblas_task.hpp +++ b/src/blas/backends/rocblas/rocblas_task.hpp @@ -47,7 +47,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace rocblas { @@ -79,6 +79,6 @@ static inline void onemath_rocblas_host_task(H &cgh, sycl::queue queue, F f) { } // namespace rocblas } // namespace blas -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ROCBLAS_TASK_HPP_ diff --git a/src/blas/backends/rocblas/rocblas_wrappers.cpp b/src/blas/backends/rocblas/rocblas_wrappers.cpp index 8d4915ea0..3c2fb8c38 100644 --- a/src/blas/backends/rocblas/rocblas_wrappers.cpp +++ b/src/blas/backends/rocblas/rocblas_wrappers.cpp @@ -25,984 +25,984 @@ extern "C" blas_function_table_t mkl_blas_table = { WRAPPER_VERSION, - oneapi::mkl::blas::rocblas::column_major::asum, - oneapi::mkl::blas::rocblas::column_major::asum, - oneapi::mkl::blas::rocblas::column_major::asum, - oneapi::mkl::blas::rocblas::column_major::asum, - oneapi::mkl::blas::rocblas::column_major::axpy, - oneapi::mkl::blas::rocblas::column_major::axpy, - oneapi::mkl::blas::rocblas::column_major::axpy, - oneapi::mkl::blas::rocblas::column_major::axpy, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpby, - oneapi::mkl::blas::rocblas::column_major::axpby, - oneapi::mkl::blas::rocblas::column_major::axpby, - oneapi::mkl::blas::rocblas::column_major::axpby, - oneapi::mkl::blas::rocblas::column_major::copy, - oneapi::mkl::blas::rocblas::column_major::copy, - oneapi::mkl::blas::rocblas::column_major::copy, - oneapi::mkl::blas::rocblas::column_major::copy, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::dot, - oneapi::mkl::blas::rocblas::column_major::dot, - oneapi::mkl::blas::rocblas::column_major::dot, - oneapi::mkl::blas::rocblas::column_major::dotc, - oneapi::mkl::blas::rocblas::column_major::dotc, - oneapi::mkl::blas::rocblas::column_major::dotu, - oneapi::mkl::blas::rocblas::column_major::dotu, - oneapi::mkl::blas::rocblas::column_major::iamin, - oneapi::mkl::blas::rocblas::column_major::iamin, - oneapi::mkl::blas::rocblas::column_major::iamin, - oneapi::mkl::blas::rocblas::column_major::iamin, - oneapi::mkl::blas::rocblas::column_major::iamax, - oneapi::mkl::blas::rocblas::column_major::iamax, - oneapi::mkl::blas::rocblas::column_major::iamax, - oneapi::mkl::blas::rocblas::column_major::iamax, - oneapi::mkl::blas::rocblas::column_major::nrm2, - oneapi::mkl::blas::rocblas::column_major::nrm2, - oneapi::mkl::blas::rocblas::column_major::nrm2, - oneapi::mkl::blas::rocblas::column_major::nrm2, - oneapi::mkl::blas::rocblas::column_major::rot, - oneapi::mkl::blas::rocblas::column_major::rot, - oneapi::mkl::blas::rocblas::column_major::rot, - oneapi::mkl::blas::rocblas::column_major::rot, - oneapi::mkl::blas::rocblas::column_major::rotg, - oneapi::mkl::blas::rocblas::column_major::rotg, - oneapi::mkl::blas::rocblas::column_major::rotg, - oneapi::mkl::blas::rocblas::column_major::rotg, - oneapi::mkl::blas::rocblas::column_major::rotm, - oneapi::mkl::blas::rocblas::column_major::rotm, - oneapi::mkl::blas::rocblas::column_major::rotmg, - oneapi::mkl::blas::rocblas::column_major::rotmg, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::sdsdot, - oneapi::mkl::blas::rocblas::column_major::swap, - oneapi::mkl::blas::rocblas::column_major::swap, - oneapi::mkl::blas::rocblas::column_major::swap, - oneapi::mkl::blas::rocblas::column_major::swap, - oneapi::mkl::blas::rocblas::column_major::gbmv, - oneapi::mkl::blas::rocblas::column_major::gbmv, - oneapi::mkl::blas::rocblas::column_major::gbmv, - oneapi::mkl::blas::rocblas::column_major::gbmv, - oneapi::mkl::blas::rocblas::column_major::gemv, - oneapi::mkl::blas::rocblas::column_major::gemv, - oneapi::mkl::blas::rocblas::column_major::gemv, - oneapi::mkl::blas::rocblas::column_major::gemv, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::ger, - oneapi::mkl::blas::rocblas::column_major::ger, - oneapi::mkl::blas::rocblas::column_major::gerc, - oneapi::mkl::blas::rocblas::column_major::gerc, - oneapi::mkl::blas::rocblas::column_major::geru, - oneapi::mkl::blas::rocblas::column_major::geru, - oneapi::mkl::blas::rocblas::column_major::hbmv, - oneapi::mkl::blas::rocblas::column_major::hbmv, - oneapi::mkl::blas::rocblas::column_major::hemv, - oneapi::mkl::blas::rocblas::column_major::hemv, - oneapi::mkl::blas::rocblas::column_major::her, - oneapi::mkl::blas::rocblas::column_major::her, - oneapi::mkl::blas::rocblas::column_major::her2, - oneapi::mkl::blas::rocblas::column_major::her2, - oneapi::mkl::blas::rocblas::column_major::hpmv, - oneapi::mkl::blas::rocblas::column_major::hpmv, - oneapi::mkl::blas::rocblas::column_major::hpr, - oneapi::mkl::blas::rocblas::column_major::hpr, - oneapi::mkl::blas::rocblas::column_major::hpr2, - oneapi::mkl::blas::rocblas::column_major::hpr2, - oneapi::mkl::blas::rocblas::column_major::sbmv, - oneapi::mkl::blas::rocblas::column_major::sbmv, - oneapi::mkl::blas::rocblas::column_major::spmv, - oneapi::mkl::blas::rocblas::column_major::spmv, - oneapi::mkl::blas::rocblas::column_major::spr, - oneapi::mkl::blas::rocblas::column_major::spr, - oneapi::mkl::blas::rocblas::column_major::spr2, - oneapi::mkl::blas::rocblas::column_major::spr2, - oneapi::mkl::blas::rocblas::column_major::symv, - oneapi::mkl::blas::rocblas::column_major::symv, - oneapi::mkl::blas::rocblas::column_major::syr, - oneapi::mkl::blas::rocblas::column_major::syr, - oneapi::mkl::blas::rocblas::column_major::syr2, - oneapi::mkl::blas::rocblas::column_major::syr2, - oneapi::mkl::blas::rocblas::column_major::tbmv, - oneapi::mkl::blas::rocblas::column_major::tbmv, - oneapi::mkl::blas::rocblas::column_major::tbmv, - oneapi::mkl::blas::rocblas::column_major::tbmv, - oneapi::mkl::blas::rocblas::column_major::tbsv, - oneapi::mkl::blas::rocblas::column_major::tbsv, - oneapi::mkl::blas::rocblas::column_major::tbsv, - oneapi::mkl::blas::rocblas::column_major::tbsv, - oneapi::mkl::blas::rocblas::column_major::tpmv, - oneapi::mkl::blas::rocblas::column_major::tpmv, - oneapi::mkl::blas::rocblas::column_major::tpmv, - oneapi::mkl::blas::rocblas::column_major::tpmv, - oneapi::mkl::blas::rocblas::column_major::tpsv, - oneapi::mkl::blas::rocblas::column_major::tpsv, - oneapi::mkl::blas::rocblas::column_major::tpsv, - oneapi::mkl::blas::rocblas::column_major::tpsv, - oneapi::mkl::blas::rocblas::column_major::trmv, - oneapi::mkl::blas::rocblas::column_major::trmv, - oneapi::mkl::blas::rocblas::column_major::trmv, - oneapi::mkl::blas::rocblas::column_major::trmv, - oneapi::mkl::blas::rocblas::column_major::trsv, - oneapi::mkl::blas::rocblas::column_major::trsv, - oneapi::mkl::blas::rocblas::column_major::trsv, - oneapi::mkl::blas::rocblas::column_major::trsv, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::hemm, - oneapi::mkl::blas::rocblas::column_major::hemm, - oneapi::mkl::blas::rocblas::column_major::herk, - oneapi::mkl::blas::rocblas::column_major::herk, - oneapi::mkl::blas::rocblas::column_major::her2k, - oneapi::mkl::blas::rocblas::column_major::her2k, - oneapi::mkl::blas::rocblas::column_major::symm, - oneapi::mkl::blas::rocblas::column_major::symm, - oneapi::mkl::blas::rocblas::column_major::symm, - oneapi::mkl::blas::rocblas::column_major::symm, - oneapi::mkl::blas::rocblas::column_major::syrk, - oneapi::mkl::blas::rocblas::column_major::syrk, - oneapi::mkl::blas::rocblas::column_major::syrk, - oneapi::mkl::blas::rocblas::column_major::syrk, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syr2k, - oneapi::mkl::blas::rocblas::column_major::syr2k, - oneapi::mkl::blas::rocblas::column_major::syr2k, - oneapi::mkl::blas::rocblas::column_major::syr2k, - oneapi::mkl::blas::rocblas::column_major::trmm, - oneapi::mkl::blas::rocblas::column_major::trmm, - oneapi::mkl::blas::rocblas::column_major::trmm, - oneapi::mkl::blas::rocblas::column_major::trmm, - oneapi::mkl::blas::rocblas::column_major::trsm, - oneapi::mkl::blas::rocblas::column_major::trsm, - oneapi::mkl::blas::rocblas::column_major::trsm, - oneapi::mkl::blas::rocblas::column_major::trsm, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::gemmt, - oneapi::mkl::blas::rocblas::column_major::gemmt, - oneapi::mkl::blas::rocblas::column_major::gemmt, - oneapi::mkl::blas::rocblas::column_major::gemmt, - oneapi::mkl::blas::rocblas::column_major::gemm_bias, - oneapi::mkl::blas::rocblas::column_major::gemm_bias, - oneapi::mkl::blas::rocblas::column_major::gemm_bias, - oneapi::mkl::blas::rocblas::column_major::gemm_bias, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::omatadd_batch, - oneapi::mkl::blas::rocblas::column_major::omatadd_batch, - oneapi::mkl::blas::rocblas::column_major::omatadd_batch, - oneapi::mkl::blas::rocblas::column_major::omatadd_batch, - oneapi::mkl::blas::rocblas::column_major::omatcopy, - oneapi::mkl::blas::rocblas::column_major::omatcopy, - oneapi::mkl::blas::rocblas::column_major::omatcopy, - oneapi::mkl::blas::rocblas::column_major::omatcopy, - oneapi::mkl::blas::rocblas::column_major::omatcopy2, - oneapi::mkl::blas::rocblas::column_major::omatcopy2, - oneapi::mkl::blas::rocblas::column_major::omatcopy2, - oneapi::mkl::blas::rocblas::column_major::omatcopy2, - oneapi::mkl::blas::rocblas::column_major::imatcopy, - oneapi::mkl::blas::rocblas::column_major::imatcopy, - oneapi::mkl::blas::rocblas::column_major::imatcopy, - oneapi::mkl::blas::rocblas::column_major::imatcopy, - oneapi::mkl::blas::rocblas::column_major::omatadd, - oneapi::mkl::blas::rocblas::column_major::omatadd, - oneapi::mkl::blas::rocblas::column_major::omatadd, - oneapi::mkl::blas::rocblas::column_major::omatadd, - oneapi::mkl::blas::rocblas::column_major::asum, - oneapi::mkl::blas::rocblas::column_major::asum, - oneapi::mkl::blas::rocblas::column_major::asum, - oneapi::mkl::blas::rocblas::column_major::asum, - oneapi::mkl::blas::rocblas::column_major::axpy, - oneapi::mkl::blas::rocblas::column_major::axpy, - oneapi::mkl::blas::rocblas::column_major::axpy, - oneapi::mkl::blas::rocblas::column_major::axpy, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpy_batch, - oneapi::mkl::blas::rocblas::column_major::axpby, - oneapi::mkl::blas::rocblas::column_major::axpby, - oneapi::mkl::blas::rocblas::column_major::axpby, - oneapi::mkl::blas::rocblas::column_major::axpby, - oneapi::mkl::blas::rocblas::column_major::copy, - oneapi::mkl::blas::rocblas::column_major::copy, - oneapi::mkl::blas::rocblas::column_major::copy, - oneapi::mkl::blas::rocblas::column_major::copy, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::copy_batch, - oneapi::mkl::blas::rocblas::column_major::dot, - oneapi::mkl::blas::rocblas::column_major::dot, - oneapi::mkl::blas::rocblas::column_major::dot, - oneapi::mkl::blas::rocblas::column_major::dotc, - oneapi::mkl::blas::rocblas::column_major::dotc, - oneapi::mkl::blas::rocblas::column_major::dotu, - oneapi::mkl::blas::rocblas::column_major::dotu, - oneapi::mkl::blas::rocblas::column_major::iamin, - oneapi::mkl::blas::rocblas::column_major::iamin, - oneapi::mkl::blas::rocblas::column_major::iamin, - oneapi::mkl::blas::rocblas::column_major::iamin, - oneapi::mkl::blas::rocblas::column_major::iamax, - oneapi::mkl::blas::rocblas::column_major::iamax, - oneapi::mkl::blas::rocblas::column_major::iamax, - oneapi::mkl::blas::rocblas::column_major::iamax, - oneapi::mkl::blas::rocblas::column_major::nrm2, - oneapi::mkl::blas::rocblas::column_major::nrm2, - oneapi::mkl::blas::rocblas::column_major::nrm2, - oneapi::mkl::blas::rocblas::column_major::nrm2, - oneapi::mkl::blas::rocblas::column_major::rot, - oneapi::mkl::blas::rocblas::column_major::rot, - oneapi::mkl::blas::rocblas::column_major::rot, - oneapi::mkl::blas::rocblas::column_major::rot, - oneapi::mkl::blas::rocblas::column_major::rotg, - oneapi::mkl::blas::rocblas::column_major::rotg, - oneapi::mkl::blas::rocblas::column_major::rotg, - oneapi::mkl::blas::rocblas::column_major::rotg, - oneapi::mkl::blas::rocblas::column_major::rotm, - oneapi::mkl::blas::rocblas::column_major::rotm, - oneapi::mkl::blas::rocblas::column_major::rotmg, - oneapi::mkl::blas::rocblas::column_major::rotmg, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::scal, - oneapi::mkl::blas::rocblas::column_major::sdsdot, - oneapi::mkl::blas::rocblas::column_major::swap, - oneapi::mkl::blas::rocblas::column_major::swap, - oneapi::mkl::blas::rocblas::column_major::swap, - oneapi::mkl::blas::rocblas::column_major::swap, - oneapi::mkl::blas::rocblas::column_major::gbmv, - oneapi::mkl::blas::rocblas::column_major::gbmv, - oneapi::mkl::blas::rocblas::column_major::gbmv, - oneapi::mkl::blas::rocblas::column_major::gbmv, - oneapi::mkl::blas::rocblas::column_major::gemv, - oneapi::mkl::blas::rocblas::column_major::gemv, - oneapi::mkl::blas::rocblas::column_major::gemv, - oneapi::mkl::blas::rocblas::column_major::gemv, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::gemv_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::dgmm_batch, - oneapi::mkl::blas::rocblas::column_major::ger, - oneapi::mkl::blas::rocblas::column_major::ger, - oneapi::mkl::blas::rocblas::column_major::gerc, - oneapi::mkl::blas::rocblas::column_major::gerc, - oneapi::mkl::blas::rocblas::column_major::geru, - oneapi::mkl::blas::rocblas::column_major::geru, - oneapi::mkl::blas::rocblas::column_major::hbmv, - oneapi::mkl::blas::rocblas::column_major::hbmv, - oneapi::mkl::blas::rocblas::column_major::hemv, - oneapi::mkl::blas::rocblas::column_major::hemv, - oneapi::mkl::blas::rocblas::column_major::her, - oneapi::mkl::blas::rocblas::column_major::her, - oneapi::mkl::blas::rocblas::column_major::her2, - oneapi::mkl::blas::rocblas::column_major::her2, - oneapi::mkl::blas::rocblas::column_major::hpmv, - oneapi::mkl::blas::rocblas::column_major::hpmv, - oneapi::mkl::blas::rocblas::column_major::hpr, - oneapi::mkl::blas::rocblas::column_major::hpr, - oneapi::mkl::blas::rocblas::column_major::hpr2, - oneapi::mkl::blas::rocblas::column_major::hpr2, - oneapi::mkl::blas::rocblas::column_major::sbmv, - oneapi::mkl::blas::rocblas::column_major::sbmv, - oneapi::mkl::blas::rocblas::column_major::spmv, - oneapi::mkl::blas::rocblas::column_major::spmv, - oneapi::mkl::blas::rocblas::column_major::spr, - oneapi::mkl::blas::rocblas::column_major::spr, - oneapi::mkl::blas::rocblas::column_major::spr2, - oneapi::mkl::blas::rocblas::column_major::spr2, - oneapi::mkl::blas::rocblas::column_major::symv, - oneapi::mkl::blas::rocblas::column_major::symv, - oneapi::mkl::blas::rocblas::column_major::syr, - oneapi::mkl::blas::rocblas::column_major::syr, - oneapi::mkl::blas::rocblas::column_major::syr2, - oneapi::mkl::blas::rocblas::column_major::syr2, - oneapi::mkl::blas::rocblas::column_major::tbmv, - oneapi::mkl::blas::rocblas::column_major::tbmv, - oneapi::mkl::blas::rocblas::column_major::tbmv, - oneapi::mkl::blas::rocblas::column_major::tbmv, - oneapi::mkl::blas::rocblas::column_major::tbsv, - oneapi::mkl::blas::rocblas::column_major::tbsv, - oneapi::mkl::blas::rocblas::column_major::tbsv, - oneapi::mkl::blas::rocblas::column_major::tbsv, - oneapi::mkl::blas::rocblas::column_major::tpmv, - oneapi::mkl::blas::rocblas::column_major::tpmv, - oneapi::mkl::blas::rocblas::column_major::tpmv, - oneapi::mkl::blas::rocblas::column_major::tpmv, - oneapi::mkl::blas::rocblas::column_major::tpsv, - oneapi::mkl::blas::rocblas::column_major::tpsv, - oneapi::mkl::blas::rocblas::column_major::tpsv, - oneapi::mkl::blas::rocblas::column_major::tpsv, - oneapi::mkl::blas::rocblas::column_major::trmv, - oneapi::mkl::blas::rocblas::column_major::trmv, - oneapi::mkl::blas::rocblas::column_major::trmv, - oneapi::mkl::blas::rocblas::column_major::trmv, - oneapi::mkl::blas::rocblas::column_major::trsv, - oneapi::mkl::blas::rocblas::column_major::trsv, - oneapi::mkl::blas::rocblas::column_major::trsv, - oneapi::mkl::blas::rocblas::column_major::trsv, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::gemm, - oneapi::mkl::blas::rocblas::column_major::hemm, - oneapi::mkl::blas::rocblas::column_major::hemm, - oneapi::mkl::blas::rocblas::column_major::herk, - oneapi::mkl::blas::rocblas::column_major::herk, - oneapi::mkl::blas::rocblas::column_major::her2k, - oneapi::mkl::blas::rocblas::column_major::her2k, - oneapi::mkl::blas::rocblas::column_major::symm, - oneapi::mkl::blas::rocblas::column_major::symm, - oneapi::mkl::blas::rocblas::column_major::symm, - oneapi::mkl::blas::rocblas::column_major::symm, - oneapi::mkl::blas::rocblas::column_major::syrk, - oneapi::mkl::blas::rocblas::column_major::syrk, - oneapi::mkl::blas::rocblas::column_major::syrk, - oneapi::mkl::blas::rocblas::column_major::syrk, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syrk_batch, - oneapi::mkl::blas::rocblas::column_major::syr2k, - oneapi::mkl::blas::rocblas::column_major::syr2k, - oneapi::mkl::blas::rocblas::column_major::syr2k, - oneapi::mkl::blas::rocblas::column_major::syr2k, - oneapi::mkl::blas::rocblas::column_major::trmm, - oneapi::mkl::blas::rocblas::column_major::trmm, - oneapi::mkl::blas::rocblas::column_major::trmm, - oneapi::mkl::blas::rocblas::column_major::trmm, - oneapi::mkl::blas::rocblas::column_major::trsm, - oneapi::mkl::blas::rocblas::column_major::trsm, - oneapi::mkl::blas::rocblas::column_major::trsm, - oneapi::mkl::blas::rocblas::column_major::trsm, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::trsm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemm_batch, - oneapi::mkl::blas::rocblas::column_major::gemmt, - oneapi::mkl::blas::rocblas::column_major::gemmt, - oneapi::mkl::blas::rocblas::column_major::gemmt, - oneapi::mkl::blas::rocblas::column_major::gemmt, - oneapi::mkl::blas::rocblas::column_major::gemm_bias, - oneapi::mkl::blas::rocblas::column_major::gemm_bias, - oneapi::mkl::blas::rocblas::column_major::gemm_bias, - oneapi::mkl::blas::rocblas::column_major::gemm_bias, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::omatadd_batch, - oneapi::mkl::blas::rocblas::column_major::omatadd_batch, - oneapi::mkl::blas::rocblas::column_major::omatadd_batch, - oneapi::mkl::blas::rocblas::column_major::omatadd_batch, - oneapi::mkl::blas::rocblas::column_major::omatcopy, - oneapi::mkl::blas::rocblas::column_major::omatcopy, - oneapi::mkl::blas::rocblas::column_major::omatcopy, - oneapi::mkl::blas::rocblas::column_major::omatcopy, - oneapi::mkl::blas::rocblas::column_major::omatcopy2, - oneapi::mkl::blas::rocblas::column_major::omatcopy2, - oneapi::mkl::blas::rocblas::column_major::omatcopy2, - oneapi::mkl::blas::rocblas::column_major::omatcopy2, - oneapi::mkl::blas::rocblas::column_major::imatcopy, - oneapi::mkl::blas::rocblas::column_major::imatcopy, - oneapi::mkl::blas::rocblas::column_major::imatcopy, - oneapi::mkl::blas::rocblas::column_major::imatcopy, - oneapi::mkl::blas::rocblas::column_major::omatadd, - oneapi::mkl::blas::rocblas::column_major::omatadd, - oneapi::mkl::blas::rocblas::column_major::omatadd, - oneapi::mkl::blas::rocblas::column_major::omatadd, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::column_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::asum, - oneapi::mkl::blas::rocblas::row_major::asum, - oneapi::mkl::blas::rocblas::row_major::asum, - oneapi::mkl::blas::rocblas::row_major::asum, - oneapi::mkl::blas::rocblas::row_major::axpy, - oneapi::mkl::blas::rocblas::row_major::axpy, - oneapi::mkl::blas::rocblas::row_major::axpy, - oneapi::mkl::blas::rocblas::row_major::axpy, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpby, - oneapi::mkl::blas::rocblas::row_major::axpby, - oneapi::mkl::blas::rocblas::row_major::axpby, - oneapi::mkl::blas::rocblas::row_major::axpby, - oneapi::mkl::blas::rocblas::row_major::copy, - oneapi::mkl::blas::rocblas::row_major::copy, - oneapi::mkl::blas::rocblas::row_major::copy, - oneapi::mkl::blas::rocblas::row_major::copy, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::dot, - oneapi::mkl::blas::rocblas::row_major::dot, - oneapi::mkl::blas::rocblas::row_major::dot, - oneapi::mkl::blas::rocblas::row_major::dotc, - oneapi::mkl::blas::rocblas::row_major::dotc, - oneapi::mkl::blas::rocblas::row_major::dotu, - oneapi::mkl::blas::rocblas::row_major::dotu, - oneapi::mkl::blas::rocblas::row_major::iamin, - oneapi::mkl::blas::rocblas::row_major::iamin, - oneapi::mkl::blas::rocblas::row_major::iamin, - oneapi::mkl::blas::rocblas::row_major::iamin, - oneapi::mkl::blas::rocblas::row_major::iamax, - oneapi::mkl::blas::rocblas::row_major::iamax, - oneapi::mkl::blas::rocblas::row_major::iamax, - oneapi::mkl::blas::rocblas::row_major::iamax, - oneapi::mkl::blas::rocblas::row_major::nrm2, - oneapi::mkl::blas::rocblas::row_major::nrm2, - oneapi::mkl::blas::rocblas::row_major::nrm2, - oneapi::mkl::blas::rocblas::row_major::nrm2, - oneapi::mkl::blas::rocblas::row_major::rot, - oneapi::mkl::blas::rocblas::row_major::rot, - oneapi::mkl::blas::rocblas::row_major::rot, - oneapi::mkl::blas::rocblas::row_major::rot, - oneapi::mkl::blas::rocblas::row_major::rotg, - oneapi::mkl::blas::rocblas::row_major::rotg, - oneapi::mkl::blas::rocblas::row_major::rotg, - oneapi::mkl::blas::rocblas::row_major::rotg, - oneapi::mkl::blas::rocblas::row_major::rotm, - oneapi::mkl::blas::rocblas::row_major::rotm, - oneapi::mkl::blas::rocblas::row_major::rotmg, - oneapi::mkl::blas::rocblas::row_major::rotmg, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::sdsdot, - oneapi::mkl::blas::rocblas::row_major::swap, - oneapi::mkl::blas::rocblas::row_major::swap, - oneapi::mkl::blas::rocblas::row_major::swap, - oneapi::mkl::blas::rocblas::row_major::swap, - oneapi::mkl::blas::rocblas::row_major::gbmv, - oneapi::mkl::blas::rocblas::row_major::gbmv, - oneapi::mkl::blas::rocblas::row_major::gbmv, - oneapi::mkl::blas::rocblas::row_major::gbmv, - oneapi::mkl::blas::rocblas::row_major::gemv, - oneapi::mkl::blas::rocblas::row_major::gemv, - oneapi::mkl::blas::rocblas::row_major::gemv, - oneapi::mkl::blas::rocblas::row_major::gemv, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::ger, - oneapi::mkl::blas::rocblas::row_major::ger, - oneapi::mkl::blas::rocblas::row_major::gerc, - oneapi::mkl::blas::rocblas::row_major::gerc, - oneapi::mkl::blas::rocblas::row_major::geru, - oneapi::mkl::blas::rocblas::row_major::geru, - oneapi::mkl::blas::rocblas::row_major::hbmv, - oneapi::mkl::blas::rocblas::row_major::hbmv, - oneapi::mkl::blas::rocblas::row_major::hemv, - oneapi::mkl::blas::rocblas::row_major::hemv, - oneapi::mkl::blas::rocblas::row_major::her, - oneapi::mkl::blas::rocblas::row_major::her, - oneapi::mkl::blas::rocblas::row_major::her2, - oneapi::mkl::blas::rocblas::row_major::her2, - oneapi::mkl::blas::rocblas::row_major::hpmv, - oneapi::mkl::blas::rocblas::row_major::hpmv, - oneapi::mkl::blas::rocblas::row_major::hpr, - oneapi::mkl::blas::rocblas::row_major::hpr, - oneapi::mkl::blas::rocblas::row_major::hpr2, - oneapi::mkl::blas::rocblas::row_major::hpr2, - oneapi::mkl::blas::rocblas::row_major::sbmv, - oneapi::mkl::blas::rocblas::row_major::sbmv, - oneapi::mkl::blas::rocblas::row_major::spmv, - oneapi::mkl::blas::rocblas::row_major::spmv, - oneapi::mkl::blas::rocblas::row_major::spr, - oneapi::mkl::blas::rocblas::row_major::spr, - oneapi::mkl::blas::rocblas::row_major::spr2, - oneapi::mkl::blas::rocblas::row_major::spr2, - oneapi::mkl::blas::rocblas::row_major::symv, - oneapi::mkl::blas::rocblas::row_major::symv, - oneapi::mkl::blas::rocblas::row_major::syr, - oneapi::mkl::blas::rocblas::row_major::syr, - oneapi::mkl::blas::rocblas::row_major::syr2, - oneapi::mkl::blas::rocblas::row_major::syr2, - oneapi::mkl::blas::rocblas::row_major::tbmv, - oneapi::mkl::blas::rocblas::row_major::tbmv, - oneapi::mkl::blas::rocblas::row_major::tbmv, - oneapi::mkl::blas::rocblas::row_major::tbmv, - oneapi::mkl::blas::rocblas::row_major::tbsv, - oneapi::mkl::blas::rocblas::row_major::tbsv, - oneapi::mkl::blas::rocblas::row_major::tbsv, - oneapi::mkl::blas::rocblas::row_major::tbsv, - oneapi::mkl::blas::rocblas::row_major::tpmv, - oneapi::mkl::blas::rocblas::row_major::tpmv, - oneapi::mkl::blas::rocblas::row_major::tpmv, - oneapi::mkl::blas::rocblas::row_major::tpmv, - oneapi::mkl::blas::rocblas::row_major::tpsv, - oneapi::mkl::blas::rocblas::row_major::tpsv, - oneapi::mkl::blas::rocblas::row_major::tpsv, - oneapi::mkl::blas::rocblas::row_major::tpsv, - oneapi::mkl::blas::rocblas::row_major::trmv, - oneapi::mkl::blas::rocblas::row_major::trmv, - oneapi::mkl::blas::rocblas::row_major::trmv, - oneapi::mkl::blas::rocblas::row_major::trmv, - oneapi::mkl::blas::rocblas::row_major::trsv, - oneapi::mkl::blas::rocblas::row_major::trsv, - oneapi::mkl::blas::rocblas::row_major::trsv, - oneapi::mkl::blas::rocblas::row_major::trsv, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::hemm, - oneapi::mkl::blas::rocblas::row_major::hemm, - oneapi::mkl::blas::rocblas::row_major::herk, - oneapi::mkl::blas::rocblas::row_major::herk, - oneapi::mkl::blas::rocblas::row_major::her2k, - oneapi::mkl::blas::rocblas::row_major::her2k, - oneapi::mkl::blas::rocblas::row_major::symm, - oneapi::mkl::blas::rocblas::row_major::symm, - oneapi::mkl::blas::rocblas::row_major::symm, - oneapi::mkl::blas::rocblas::row_major::symm, - oneapi::mkl::blas::rocblas::row_major::syrk, - oneapi::mkl::blas::rocblas::row_major::syrk, - oneapi::mkl::blas::rocblas::row_major::syrk, - oneapi::mkl::blas::rocblas::row_major::syrk, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syr2k, - oneapi::mkl::blas::rocblas::row_major::syr2k, - oneapi::mkl::blas::rocblas::row_major::syr2k, - oneapi::mkl::blas::rocblas::row_major::syr2k, - oneapi::mkl::blas::rocblas::row_major::trmm, - oneapi::mkl::blas::rocblas::row_major::trmm, - oneapi::mkl::blas::rocblas::row_major::trmm, - oneapi::mkl::blas::rocblas::row_major::trmm, - oneapi::mkl::blas::rocblas::row_major::trsm, - oneapi::mkl::blas::rocblas::row_major::trsm, - oneapi::mkl::blas::rocblas::row_major::trsm, - oneapi::mkl::blas::rocblas::row_major::trsm, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::gemmt, - oneapi::mkl::blas::rocblas::row_major::gemmt, - oneapi::mkl::blas::rocblas::row_major::gemmt, - oneapi::mkl::blas::rocblas::row_major::gemmt, - oneapi::mkl::blas::rocblas::row_major::gemm_bias, - oneapi::mkl::blas::rocblas::row_major::gemm_bias, - oneapi::mkl::blas::rocblas::row_major::gemm_bias, - oneapi::mkl::blas::rocblas::row_major::gemm_bias, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::omatadd_batch, - oneapi::mkl::blas::rocblas::row_major::omatadd_batch, - oneapi::mkl::blas::rocblas::row_major::omatadd_batch, - oneapi::mkl::blas::rocblas::row_major::omatadd_batch, - oneapi::mkl::blas::rocblas::row_major::omatcopy, - oneapi::mkl::blas::rocblas::row_major::omatcopy, - oneapi::mkl::blas::rocblas::row_major::omatcopy, - oneapi::mkl::blas::rocblas::row_major::omatcopy, - oneapi::mkl::blas::rocblas::row_major::omatcopy2, - oneapi::mkl::blas::rocblas::row_major::omatcopy2, - oneapi::mkl::blas::rocblas::row_major::omatcopy2, - oneapi::mkl::blas::rocblas::row_major::omatcopy2, - oneapi::mkl::blas::rocblas::row_major::imatcopy, - oneapi::mkl::blas::rocblas::row_major::imatcopy, - oneapi::mkl::blas::rocblas::row_major::imatcopy, - oneapi::mkl::blas::rocblas::row_major::imatcopy, - oneapi::mkl::blas::rocblas::row_major::omatadd, - oneapi::mkl::blas::rocblas::row_major::omatadd, - oneapi::mkl::blas::rocblas::row_major::omatadd, - oneapi::mkl::blas::rocblas::row_major::omatadd, - oneapi::mkl::blas::rocblas::row_major::asum, - oneapi::mkl::blas::rocblas::row_major::asum, - oneapi::mkl::blas::rocblas::row_major::asum, - oneapi::mkl::blas::rocblas::row_major::asum, - oneapi::mkl::blas::rocblas::row_major::axpy, - oneapi::mkl::blas::rocblas::row_major::axpy, - oneapi::mkl::blas::rocblas::row_major::axpy, - oneapi::mkl::blas::rocblas::row_major::axpy, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpy_batch, - oneapi::mkl::blas::rocblas::row_major::axpby, - oneapi::mkl::blas::rocblas::row_major::axpby, - oneapi::mkl::blas::rocblas::row_major::axpby, - oneapi::mkl::blas::rocblas::row_major::axpby, - oneapi::mkl::blas::rocblas::row_major::copy, - oneapi::mkl::blas::rocblas::row_major::copy, - oneapi::mkl::blas::rocblas::row_major::copy, - oneapi::mkl::blas::rocblas::row_major::copy, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::copy_batch, - oneapi::mkl::blas::rocblas::row_major::dot, - oneapi::mkl::blas::rocblas::row_major::dot, - oneapi::mkl::blas::rocblas::row_major::dot, - oneapi::mkl::blas::rocblas::row_major::dotc, - oneapi::mkl::blas::rocblas::row_major::dotc, - oneapi::mkl::blas::rocblas::row_major::dotu, - oneapi::mkl::blas::rocblas::row_major::dotu, - oneapi::mkl::blas::rocblas::row_major::iamin, - oneapi::mkl::blas::rocblas::row_major::iamin, - oneapi::mkl::blas::rocblas::row_major::iamin, - oneapi::mkl::blas::rocblas::row_major::iamin, - oneapi::mkl::blas::rocblas::row_major::iamax, - oneapi::mkl::blas::rocblas::row_major::iamax, - oneapi::mkl::blas::rocblas::row_major::iamax, - oneapi::mkl::blas::rocblas::row_major::iamax, - oneapi::mkl::blas::rocblas::row_major::nrm2, - oneapi::mkl::blas::rocblas::row_major::nrm2, - oneapi::mkl::blas::rocblas::row_major::nrm2, - oneapi::mkl::blas::rocblas::row_major::nrm2, - oneapi::mkl::blas::rocblas::row_major::rot, - oneapi::mkl::blas::rocblas::row_major::rot, - oneapi::mkl::blas::rocblas::row_major::rot, - oneapi::mkl::blas::rocblas::row_major::rot, - oneapi::mkl::blas::rocblas::row_major::rotg, - oneapi::mkl::blas::rocblas::row_major::rotg, - oneapi::mkl::blas::rocblas::row_major::rotg, - oneapi::mkl::blas::rocblas::row_major::rotg, - oneapi::mkl::blas::rocblas::row_major::rotm, - oneapi::mkl::blas::rocblas::row_major::rotm, - oneapi::mkl::blas::rocblas::row_major::rotmg, - oneapi::mkl::blas::rocblas::row_major::rotmg, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::scal, - oneapi::mkl::blas::rocblas::row_major::sdsdot, - oneapi::mkl::blas::rocblas::row_major::swap, - oneapi::mkl::blas::rocblas::row_major::swap, - oneapi::mkl::blas::rocblas::row_major::swap, - oneapi::mkl::blas::rocblas::row_major::swap, - oneapi::mkl::blas::rocblas::row_major::gbmv, - oneapi::mkl::blas::rocblas::row_major::gbmv, - oneapi::mkl::blas::rocblas::row_major::gbmv, - oneapi::mkl::blas::rocblas::row_major::gbmv, - oneapi::mkl::blas::rocblas::row_major::gemv, - oneapi::mkl::blas::rocblas::row_major::gemv, - oneapi::mkl::blas::rocblas::row_major::gemv, - oneapi::mkl::blas::rocblas::row_major::gemv, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::gemv_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::dgmm_batch, - oneapi::mkl::blas::rocblas::row_major::ger, - oneapi::mkl::blas::rocblas::row_major::ger, - oneapi::mkl::blas::rocblas::row_major::gerc, - oneapi::mkl::blas::rocblas::row_major::gerc, - oneapi::mkl::blas::rocblas::row_major::geru, - oneapi::mkl::blas::rocblas::row_major::geru, - oneapi::mkl::blas::rocblas::row_major::hbmv, - oneapi::mkl::blas::rocblas::row_major::hbmv, - oneapi::mkl::blas::rocblas::row_major::hemv, - oneapi::mkl::blas::rocblas::row_major::hemv, - oneapi::mkl::blas::rocblas::row_major::her, - oneapi::mkl::blas::rocblas::row_major::her, - oneapi::mkl::blas::rocblas::row_major::her2, - oneapi::mkl::blas::rocblas::row_major::her2, - oneapi::mkl::blas::rocblas::row_major::hpmv, - oneapi::mkl::blas::rocblas::row_major::hpmv, - oneapi::mkl::blas::rocblas::row_major::hpr, - oneapi::mkl::blas::rocblas::row_major::hpr, - oneapi::mkl::blas::rocblas::row_major::hpr2, - oneapi::mkl::blas::rocblas::row_major::hpr2, - oneapi::mkl::blas::rocblas::row_major::sbmv, - oneapi::mkl::blas::rocblas::row_major::sbmv, - oneapi::mkl::blas::rocblas::row_major::spmv, - oneapi::mkl::blas::rocblas::row_major::spmv, - oneapi::mkl::blas::rocblas::row_major::spr, - oneapi::mkl::blas::rocblas::row_major::spr, - oneapi::mkl::blas::rocblas::row_major::spr2, - oneapi::mkl::blas::rocblas::row_major::spr2, - oneapi::mkl::blas::rocblas::row_major::symv, - oneapi::mkl::blas::rocblas::row_major::symv, - oneapi::mkl::blas::rocblas::row_major::syr, - oneapi::mkl::blas::rocblas::row_major::syr, - oneapi::mkl::blas::rocblas::row_major::syr2, - oneapi::mkl::blas::rocblas::row_major::syr2, - oneapi::mkl::blas::rocblas::row_major::tbmv, - oneapi::mkl::blas::rocblas::row_major::tbmv, - oneapi::mkl::blas::rocblas::row_major::tbmv, - oneapi::mkl::blas::rocblas::row_major::tbmv, - oneapi::mkl::blas::rocblas::row_major::tbsv, - oneapi::mkl::blas::rocblas::row_major::tbsv, - oneapi::mkl::blas::rocblas::row_major::tbsv, - oneapi::mkl::blas::rocblas::row_major::tbsv, - oneapi::mkl::blas::rocblas::row_major::tpmv, - oneapi::mkl::blas::rocblas::row_major::tpmv, - oneapi::mkl::blas::rocblas::row_major::tpmv, - oneapi::mkl::blas::rocblas::row_major::tpmv, - oneapi::mkl::blas::rocblas::row_major::tpsv, - oneapi::mkl::blas::rocblas::row_major::tpsv, - oneapi::mkl::blas::rocblas::row_major::tpsv, - oneapi::mkl::blas::rocblas::row_major::tpsv, - oneapi::mkl::blas::rocblas::row_major::trmv, - oneapi::mkl::blas::rocblas::row_major::trmv, - oneapi::mkl::blas::rocblas::row_major::trmv, - oneapi::mkl::blas::rocblas::row_major::trmv, - oneapi::mkl::blas::rocblas::row_major::trsv, - oneapi::mkl::blas::rocblas::row_major::trsv, - oneapi::mkl::blas::rocblas::row_major::trsv, - oneapi::mkl::blas::rocblas::row_major::trsv, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::gemm, - oneapi::mkl::blas::rocblas::row_major::hemm, - oneapi::mkl::blas::rocblas::row_major::hemm, - oneapi::mkl::blas::rocblas::row_major::herk, - oneapi::mkl::blas::rocblas::row_major::herk, - oneapi::mkl::blas::rocblas::row_major::her2k, - oneapi::mkl::blas::rocblas::row_major::her2k, - oneapi::mkl::blas::rocblas::row_major::symm, - oneapi::mkl::blas::rocblas::row_major::symm, - oneapi::mkl::blas::rocblas::row_major::symm, - oneapi::mkl::blas::rocblas::row_major::symm, - oneapi::mkl::blas::rocblas::row_major::syrk, - oneapi::mkl::blas::rocblas::row_major::syrk, - oneapi::mkl::blas::rocblas::row_major::syrk, - oneapi::mkl::blas::rocblas::row_major::syrk, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syrk_batch, - oneapi::mkl::blas::rocblas::row_major::syr2k, - oneapi::mkl::blas::rocblas::row_major::syr2k, - oneapi::mkl::blas::rocblas::row_major::syr2k, - oneapi::mkl::blas::rocblas::row_major::syr2k, - oneapi::mkl::blas::rocblas::row_major::trmm, - oneapi::mkl::blas::rocblas::row_major::trmm, - oneapi::mkl::blas::rocblas::row_major::trmm, - oneapi::mkl::blas::rocblas::row_major::trmm, - oneapi::mkl::blas::rocblas::row_major::trsm, - oneapi::mkl::blas::rocblas::row_major::trsm, - oneapi::mkl::blas::rocblas::row_major::trsm, - oneapi::mkl::blas::rocblas::row_major::trsm, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::trsm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemm_batch, - oneapi::mkl::blas::rocblas::row_major::gemmt, - oneapi::mkl::blas::rocblas::row_major::gemmt, - oneapi::mkl::blas::rocblas::row_major::gemmt, - oneapi::mkl::blas::rocblas::row_major::gemmt, - oneapi::mkl::blas::rocblas::row_major::gemm_bias, - oneapi::mkl::blas::rocblas::row_major::gemm_bias, - oneapi::mkl::blas::rocblas::row_major::gemm_bias, - oneapi::mkl::blas::rocblas::row_major::gemm_bias, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::omatadd_batch, - oneapi::mkl::blas::rocblas::row_major::omatadd_batch, - oneapi::mkl::blas::rocblas::row_major::omatadd_batch, - oneapi::mkl::blas::rocblas::row_major::omatadd_batch, - oneapi::mkl::blas::rocblas::row_major::omatcopy, - oneapi::mkl::blas::rocblas::row_major::omatcopy, - oneapi::mkl::blas::rocblas::row_major::omatcopy, - oneapi::mkl::blas::rocblas::row_major::omatcopy, - oneapi::mkl::blas::rocblas::row_major::omatcopy2, - oneapi::mkl::blas::rocblas::row_major::omatcopy2, - oneapi::mkl::blas::rocblas::row_major::omatcopy2, - oneapi::mkl::blas::rocblas::row_major::omatcopy2, - oneapi::mkl::blas::rocblas::row_major::imatcopy, - oneapi::mkl::blas::rocblas::row_major::imatcopy, - oneapi::mkl::blas::rocblas::row_major::imatcopy, - oneapi::mkl::blas::rocblas::row_major::imatcopy, - oneapi::mkl::blas::rocblas::row_major::omatadd, - oneapi::mkl::blas::rocblas::row_major::omatadd, - oneapi::mkl::blas::rocblas::row_major::omatadd, - oneapi::mkl::blas::rocblas::row_major::omatadd, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::omatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, - oneapi::mkl::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::asum, + oneapi::math::blas::rocblas::column_major::asum, + oneapi::math::blas::rocblas::column_major::asum, + oneapi::math::blas::rocblas::column_major::asum, + oneapi::math::blas::rocblas::column_major::axpy, + oneapi::math::blas::rocblas::column_major::axpy, + oneapi::math::blas::rocblas::column_major::axpy, + oneapi::math::blas::rocblas::column_major::axpy, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpby, + oneapi::math::blas::rocblas::column_major::axpby, + oneapi::math::blas::rocblas::column_major::axpby, + oneapi::math::blas::rocblas::column_major::axpby, + oneapi::math::blas::rocblas::column_major::copy, + oneapi::math::blas::rocblas::column_major::copy, + oneapi::math::blas::rocblas::column_major::copy, + oneapi::math::blas::rocblas::column_major::copy, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::dot, + oneapi::math::blas::rocblas::column_major::dot, + oneapi::math::blas::rocblas::column_major::dot, + oneapi::math::blas::rocblas::column_major::dotc, + oneapi::math::blas::rocblas::column_major::dotc, + oneapi::math::blas::rocblas::column_major::dotu, + oneapi::math::blas::rocblas::column_major::dotu, + oneapi::math::blas::rocblas::column_major::iamin, + oneapi::math::blas::rocblas::column_major::iamin, + oneapi::math::blas::rocblas::column_major::iamin, + oneapi::math::blas::rocblas::column_major::iamin, + oneapi::math::blas::rocblas::column_major::iamax, + oneapi::math::blas::rocblas::column_major::iamax, + oneapi::math::blas::rocblas::column_major::iamax, + oneapi::math::blas::rocblas::column_major::iamax, + oneapi::math::blas::rocblas::column_major::nrm2, + oneapi::math::blas::rocblas::column_major::nrm2, + oneapi::math::blas::rocblas::column_major::nrm2, + oneapi::math::blas::rocblas::column_major::nrm2, + oneapi::math::blas::rocblas::column_major::rot, + oneapi::math::blas::rocblas::column_major::rot, + oneapi::math::blas::rocblas::column_major::rot, + oneapi::math::blas::rocblas::column_major::rot, + oneapi::math::blas::rocblas::column_major::rotg, + oneapi::math::blas::rocblas::column_major::rotg, + oneapi::math::blas::rocblas::column_major::rotg, + oneapi::math::blas::rocblas::column_major::rotg, + oneapi::math::blas::rocblas::column_major::rotm, + oneapi::math::blas::rocblas::column_major::rotm, + oneapi::math::blas::rocblas::column_major::rotmg, + oneapi::math::blas::rocblas::column_major::rotmg, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::sdsdot, + oneapi::math::blas::rocblas::column_major::swap, + oneapi::math::blas::rocblas::column_major::swap, + oneapi::math::blas::rocblas::column_major::swap, + oneapi::math::blas::rocblas::column_major::swap, + oneapi::math::blas::rocblas::column_major::gbmv, + oneapi::math::blas::rocblas::column_major::gbmv, + oneapi::math::blas::rocblas::column_major::gbmv, + oneapi::math::blas::rocblas::column_major::gbmv, + oneapi::math::blas::rocblas::column_major::gemv, + oneapi::math::blas::rocblas::column_major::gemv, + oneapi::math::blas::rocblas::column_major::gemv, + oneapi::math::blas::rocblas::column_major::gemv, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::ger, + oneapi::math::blas::rocblas::column_major::ger, + oneapi::math::blas::rocblas::column_major::gerc, + oneapi::math::blas::rocblas::column_major::gerc, + oneapi::math::blas::rocblas::column_major::geru, + oneapi::math::blas::rocblas::column_major::geru, + oneapi::math::blas::rocblas::column_major::hbmv, + oneapi::math::blas::rocblas::column_major::hbmv, + oneapi::math::blas::rocblas::column_major::hemv, + oneapi::math::blas::rocblas::column_major::hemv, + oneapi::math::blas::rocblas::column_major::her, + oneapi::math::blas::rocblas::column_major::her, + oneapi::math::blas::rocblas::column_major::her2, + oneapi::math::blas::rocblas::column_major::her2, + oneapi::math::blas::rocblas::column_major::hpmv, + oneapi::math::blas::rocblas::column_major::hpmv, + oneapi::math::blas::rocblas::column_major::hpr, + oneapi::math::blas::rocblas::column_major::hpr, + oneapi::math::blas::rocblas::column_major::hpr2, + oneapi::math::blas::rocblas::column_major::hpr2, + oneapi::math::blas::rocblas::column_major::sbmv, + oneapi::math::blas::rocblas::column_major::sbmv, + oneapi::math::blas::rocblas::column_major::spmv, + oneapi::math::blas::rocblas::column_major::spmv, + oneapi::math::blas::rocblas::column_major::spr, + oneapi::math::blas::rocblas::column_major::spr, + oneapi::math::blas::rocblas::column_major::spr2, + oneapi::math::blas::rocblas::column_major::spr2, + oneapi::math::blas::rocblas::column_major::symv, + oneapi::math::blas::rocblas::column_major::symv, + oneapi::math::blas::rocblas::column_major::syr, + oneapi::math::blas::rocblas::column_major::syr, + oneapi::math::blas::rocblas::column_major::syr2, + oneapi::math::blas::rocblas::column_major::syr2, + oneapi::math::blas::rocblas::column_major::tbmv, + oneapi::math::blas::rocblas::column_major::tbmv, + oneapi::math::blas::rocblas::column_major::tbmv, + oneapi::math::blas::rocblas::column_major::tbmv, + oneapi::math::blas::rocblas::column_major::tbsv, + oneapi::math::blas::rocblas::column_major::tbsv, + oneapi::math::blas::rocblas::column_major::tbsv, + oneapi::math::blas::rocblas::column_major::tbsv, + oneapi::math::blas::rocblas::column_major::tpmv, + oneapi::math::blas::rocblas::column_major::tpmv, + oneapi::math::blas::rocblas::column_major::tpmv, + oneapi::math::blas::rocblas::column_major::tpmv, + oneapi::math::blas::rocblas::column_major::tpsv, + oneapi::math::blas::rocblas::column_major::tpsv, + oneapi::math::blas::rocblas::column_major::tpsv, + oneapi::math::blas::rocblas::column_major::tpsv, + oneapi::math::blas::rocblas::column_major::trmv, + oneapi::math::blas::rocblas::column_major::trmv, + oneapi::math::blas::rocblas::column_major::trmv, + oneapi::math::blas::rocblas::column_major::trmv, + oneapi::math::blas::rocblas::column_major::trsv, + oneapi::math::blas::rocblas::column_major::trsv, + oneapi::math::blas::rocblas::column_major::trsv, + oneapi::math::blas::rocblas::column_major::trsv, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::hemm, + oneapi::math::blas::rocblas::column_major::hemm, + oneapi::math::blas::rocblas::column_major::herk, + oneapi::math::blas::rocblas::column_major::herk, + oneapi::math::blas::rocblas::column_major::her2k, + oneapi::math::blas::rocblas::column_major::her2k, + oneapi::math::blas::rocblas::column_major::symm, + oneapi::math::blas::rocblas::column_major::symm, + oneapi::math::blas::rocblas::column_major::symm, + oneapi::math::blas::rocblas::column_major::symm, + oneapi::math::blas::rocblas::column_major::syrk, + oneapi::math::blas::rocblas::column_major::syrk, + oneapi::math::blas::rocblas::column_major::syrk, + oneapi::math::blas::rocblas::column_major::syrk, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syr2k, + oneapi::math::blas::rocblas::column_major::syr2k, + oneapi::math::blas::rocblas::column_major::syr2k, + oneapi::math::blas::rocblas::column_major::syr2k, + oneapi::math::blas::rocblas::column_major::trmm, + oneapi::math::blas::rocblas::column_major::trmm, + oneapi::math::blas::rocblas::column_major::trmm, + oneapi::math::blas::rocblas::column_major::trmm, + oneapi::math::blas::rocblas::column_major::trsm, + oneapi::math::blas::rocblas::column_major::trsm, + oneapi::math::blas::rocblas::column_major::trsm, + oneapi::math::blas::rocblas::column_major::trsm, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::gemmt, + oneapi::math::blas::rocblas::column_major::gemmt, + oneapi::math::blas::rocblas::column_major::gemmt, + oneapi::math::blas::rocblas::column_major::gemmt, + oneapi::math::blas::rocblas::column_major::gemm_bias, + oneapi::math::blas::rocblas::column_major::gemm_bias, + oneapi::math::blas::rocblas::column_major::gemm_bias, + oneapi::math::blas::rocblas::column_major::gemm_bias, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::omatadd_batch, + oneapi::math::blas::rocblas::column_major::omatadd_batch, + oneapi::math::blas::rocblas::column_major::omatadd_batch, + oneapi::math::blas::rocblas::column_major::omatadd_batch, + oneapi::math::blas::rocblas::column_major::omatcopy, + oneapi::math::blas::rocblas::column_major::omatcopy, + oneapi::math::blas::rocblas::column_major::omatcopy, + oneapi::math::blas::rocblas::column_major::omatcopy, + oneapi::math::blas::rocblas::column_major::omatcopy2, + oneapi::math::blas::rocblas::column_major::omatcopy2, + oneapi::math::blas::rocblas::column_major::omatcopy2, + oneapi::math::blas::rocblas::column_major::omatcopy2, + oneapi::math::blas::rocblas::column_major::imatcopy, + oneapi::math::blas::rocblas::column_major::imatcopy, + oneapi::math::blas::rocblas::column_major::imatcopy, + oneapi::math::blas::rocblas::column_major::imatcopy, + oneapi::math::blas::rocblas::column_major::omatadd, + oneapi::math::blas::rocblas::column_major::omatadd, + oneapi::math::blas::rocblas::column_major::omatadd, + oneapi::math::blas::rocblas::column_major::omatadd, + oneapi::math::blas::rocblas::column_major::asum, + oneapi::math::blas::rocblas::column_major::asum, + oneapi::math::blas::rocblas::column_major::asum, + oneapi::math::blas::rocblas::column_major::asum, + oneapi::math::blas::rocblas::column_major::axpy, + oneapi::math::blas::rocblas::column_major::axpy, + oneapi::math::blas::rocblas::column_major::axpy, + oneapi::math::blas::rocblas::column_major::axpy, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpy_batch, + oneapi::math::blas::rocblas::column_major::axpby, + oneapi::math::blas::rocblas::column_major::axpby, + oneapi::math::blas::rocblas::column_major::axpby, + oneapi::math::blas::rocblas::column_major::axpby, + oneapi::math::blas::rocblas::column_major::copy, + oneapi::math::blas::rocblas::column_major::copy, + oneapi::math::blas::rocblas::column_major::copy, + oneapi::math::blas::rocblas::column_major::copy, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::copy_batch, + oneapi::math::blas::rocblas::column_major::dot, + oneapi::math::blas::rocblas::column_major::dot, + oneapi::math::blas::rocblas::column_major::dot, + oneapi::math::blas::rocblas::column_major::dotc, + oneapi::math::blas::rocblas::column_major::dotc, + oneapi::math::blas::rocblas::column_major::dotu, + oneapi::math::blas::rocblas::column_major::dotu, + oneapi::math::blas::rocblas::column_major::iamin, + oneapi::math::blas::rocblas::column_major::iamin, + oneapi::math::blas::rocblas::column_major::iamin, + oneapi::math::blas::rocblas::column_major::iamin, + oneapi::math::blas::rocblas::column_major::iamax, + oneapi::math::blas::rocblas::column_major::iamax, + oneapi::math::blas::rocblas::column_major::iamax, + oneapi::math::blas::rocblas::column_major::iamax, + oneapi::math::blas::rocblas::column_major::nrm2, + oneapi::math::blas::rocblas::column_major::nrm2, + oneapi::math::blas::rocblas::column_major::nrm2, + oneapi::math::blas::rocblas::column_major::nrm2, + oneapi::math::blas::rocblas::column_major::rot, + oneapi::math::blas::rocblas::column_major::rot, + oneapi::math::blas::rocblas::column_major::rot, + oneapi::math::blas::rocblas::column_major::rot, + oneapi::math::blas::rocblas::column_major::rotg, + oneapi::math::blas::rocblas::column_major::rotg, + oneapi::math::blas::rocblas::column_major::rotg, + oneapi::math::blas::rocblas::column_major::rotg, + oneapi::math::blas::rocblas::column_major::rotm, + oneapi::math::blas::rocblas::column_major::rotm, + oneapi::math::blas::rocblas::column_major::rotmg, + oneapi::math::blas::rocblas::column_major::rotmg, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::scal, + oneapi::math::blas::rocblas::column_major::sdsdot, + oneapi::math::blas::rocblas::column_major::swap, + oneapi::math::blas::rocblas::column_major::swap, + oneapi::math::blas::rocblas::column_major::swap, + oneapi::math::blas::rocblas::column_major::swap, + oneapi::math::blas::rocblas::column_major::gbmv, + oneapi::math::blas::rocblas::column_major::gbmv, + oneapi::math::blas::rocblas::column_major::gbmv, + oneapi::math::blas::rocblas::column_major::gbmv, + oneapi::math::blas::rocblas::column_major::gemv, + oneapi::math::blas::rocblas::column_major::gemv, + oneapi::math::blas::rocblas::column_major::gemv, + oneapi::math::blas::rocblas::column_major::gemv, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::gemv_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::dgmm_batch, + oneapi::math::blas::rocblas::column_major::ger, + oneapi::math::blas::rocblas::column_major::ger, + oneapi::math::blas::rocblas::column_major::gerc, + oneapi::math::blas::rocblas::column_major::gerc, + oneapi::math::blas::rocblas::column_major::geru, + oneapi::math::blas::rocblas::column_major::geru, + oneapi::math::blas::rocblas::column_major::hbmv, + oneapi::math::blas::rocblas::column_major::hbmv, + oneapi::math::blas::rocblas::column_major::hemv, + oneapi::math::blas::rocblas::column_major::hemv, + oneapi::math::blas::rocblas::column_major::her, + oneapi::math::blas::rocblas::column_major::her, + oneapi::math::blas::rocblas::column_major::her2, + oneapi::math::blas::rocblas::column_major::her2, + oneapi::math::blas::rocblas::column_major::hpmv, + oneapi::math::blas::rocblas::column_major::hpmv, + oneapi::math::blas::rocblas::column_major::hpr, + oneapi::math::blas::rocblas::column_major::hpr, + oneapi::math::blas::rocblas::column_major::hpr2, + oneapi::math::blas::rocblas::column_major::hpr2, + oneapi::math::blas::rocblas::column_major::sbmv, + oneapi::math::blas::rocblas::column_major::sbmv, + oneapi::math::blas::rocblas::column_major::spmv, + oneapi::math::blas::rocblas::column_major::spmv, + oneapi::math::blas::rocblas::column_major::spr, + oneapi::math::blas::rocblas::column_major::spr, + oneapi::math::blas::rocblas::column_major::spr2, + oneapi::math::blas::rocblas::column_major::spr2, + oneapi::math::blas::rocblas::column_major::symv, + oneapi::math::blas::rocblas::column_major::symv, + oneapi::math::blas::rocblas::column_major::syr, + oneapi::math::blas::rocblas::column_major::syr, + oneapi::math::blas::rocblas::column_major::syr2, + oneapi::math::blas::rocblas::column_major::syr2, + oneapi::math::blas::rocblas::column_major::tbmv, + oneapi::math::blas::rocblas::column_major::tbmv, + oneapi::math::blas::rocblas::column_major::tbmv, + oneapi::math::blas::rocblas::column_major::tbmv, + oneapi::math::blas::rocblas::column_major::tbsv, + oneapi::math::blas::rocblas::column_major::tbsv, + oneapi::math::blas::rocblas::column_major::tbsv, + oneapi::math::blas::rocblas::column_major::tbsv, + oneapi::math::blas::rocblas::column_major::tpmv, + oneapi::math::blas::rocblas::column_major::tpmv, + oneapi::math::blas::rocblas::column_major::tpmv, + oneapi::math::blas::rocblas::column_major::tpmv, + oneapi::math::blas::rocblas::column_major::tpsv, + oneapi::math::blas::rocblas::column_major::tpsv, + oneapi::math::blas::rocblas::column_major::tpsv, + oneapi::math::blas::rocblas::column_major::tpsv, + oneapi::math::blas::rocblas::column_major::trmv, + oneapi::math::blas::rocblas::column_major::trmv, + oneapi::math::blas::rocblas::column_major::trmv, + oneapi::math::blas::rocblas::column_major::trmv, + oneapi::math::blas::rocblas::column_major::trsv, + oneapi::math::blas::rocblas::column_major::trsv, + oneapi::math::blas::rocblas::column_major::trsv, + oneapi::math::blas::rocblas::column_major::trsv, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::gemm, + oneapi::math::blas::rocblas::column_major::hemm, + oneapi::math::blas::rocblas::column_major::hemm, + oneapi::math::blas::rocblas::column_major::herk, + oneapi::math::blas::rocblas::column_major::herk, + oneapi::math::blas::rocblas::column_major::her2k, + oneapi::math::blas::rocblas::column_major::her2k, + oneapi::math::blas::rocblas::column_major::symm, + oneapi::math::blas::rocblas::column_major::symm, + oneapi::math::blas::rocblas::column_major::symm, + oneapi::math::blas::rocblas::column_major::symm, + oneapi::math::blas::rocblas::column_major::syrk, + oneapi::math::blas::rocblas::column_major::syrk, + oneapi::math::blas::rocblas::column_major::syrk, + oneapi::math::blas::rocblas::column_major::syrk, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syrk_batch, + oneapi::math::blas::rocblas::column_major::syr2k, + oneapi::math::blas::rocblas::column_major::syr2k, + oneapi::math::blas::rocblas::column_major::syr2k, + oneapi::math::blas::rocblas::column_major::syr2k, + oneapi::math::blas::rocblas::column_major::trmm, + oneapi::math::blas::rocblas::column_major::trmm, + oneapi::math::blas::rocblas::column_major::trmm, + oneapi::math::blas::rocblas::column_major::trmm, + oneapi::math::blas::rocblas::column_major::trsm, + oneapi::math::blas::rocblas::column_major::trsm, + oneapi::math::blas::rocblas::column_major::trsm, + oneapi::math::blas::rocblas::column_major::trsm, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::trsm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemm_batch, + oneapi::math::blas::rocblas::column_major::gemmt, + oneapi::math::blas::rocblas::column_major::gemmt, + oneapi::math::blas::rocblas::column_major::gemmt, + oneapi::math::blas::rocblas::column_major::gemmt, + oneapi::math::blas::rocblas::column_major::gemm_bias, + oneapi::math::blas::rocblas::column_major::gemm_bias, + oneapi::math::blas::rocblas::column_major::gemm_bias, + oneapi::math::blas::rocblas::column_major::gemm_bias, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::omatadd_batch, + oneapi::math::blas::rocblas::column_major::omatadd_batch, + oneapi::math::blas::rocblas::column_major::omatadd_batch, + oneapi::math::blas::rocblas::column_major::omatadd_batch, + oneapi::math::blas::rocblas::column_major::omatcopy, + oneapi::math::blas::rocblas::column_major::omatcopy, + oneapi::math::blas::rocblas::column_major::omatcopy, + oneapi::math::blas::rocblas::column_major::omatcopy, + oneapi::math::blas::rocblas::column_major::omatcopy2, + oneapi::math::blas::rocblas::column_major::omatcopy2, + oneapi::math::blas::rocblas::column_major::omatcopy2, + oneapi::math::blas::rocblas::column_major::omatcopy2, + oneapi::math::blas::rocblas::column_major::imatcopy, + oneapi::math::blas::rocblas::column_major::imatcopy, + oneapi::math::blas::rocblas::column_major::imatcopy, + oneapi::math::blas::rocblas::column_major::imatcopy, + oneapi::math::blas::rocblas::column_major::omatadd, + oneapi::math::blas::rocblas::column_major::omatadd, + oneapi::math::blas::rocblas::column_major::omatadd, + oneapi::math::blas::rocblas::column_major::omatadd, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::omatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::column_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::asum, + oneapi::math::blas::rocblas::row_major::asum, + oneapi::math::blas::rocblas::row_major::asum, + oneapi::math::blas::rocblas::row_major::asum, + oneapi::math::blas::rocblas::row_major::axpy, + oneapi::math::blas::rocblas::row_major::axpy, + oneapi::math::blas::rocblas::row_major::axpy, + oneapi::math::blas::rocblas::row_major::axpy, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpby, + oneapi::math::blas::rocblas::row_major::axpby, + oneapi::math::blas::rocblas::row_major::axpby, + oneapi::math::blas::rocblas::row_major::axpby, + oneapi::math::blas::rocblas::row_major::copy, + oneapi::math::blas::rocblas::row_major::copy, + oneapi::math::blas::rocblas::row_major::copy, + oneapi::math::blas::rocblas::row_major::copy, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::dot, + oneapi::math::blas::rocblas::row_major::dot, + oneapi::math::blas::rocblas::row_major::dot, + oneapi::math::blas::rocblas::row_major::dotc, + oneapi::math::blas::rocblas::row_major::dotc, + oneapi::math::blas::rocblas::row_major::dotu, + oneapi::math::blas::rocblas::row_major::dotu, + oneapi::math::blas::rocblas::row_major::iamin, + oneapi::math::blas::rocblas::row_major::iamin, + oneapi::math::blas::rocblas::row_major::iamin, + oneapi::math::blas::rocblas::row_major::iamin, + oneapi::math::blas::rocblas::row_major::iamax, + oneapi::math::blas::rocblas::row_major::iamax, + oneapi::math::blas::rocblas::row_major::iamax, + oneapi::math::blas::rocblas::row_major::iamax, + oneapi::math::blas::rocblas::row_major::nrm2, + oneapi::math::blas::rocblas::row_major::nrm2, + oneapi::math::blas::rocblas::row_major::nrm2, + oneapi::math::blas::rocblas::row_major::nrm2, + oneapi::math::blas::rocblas::row_major::rot, + oneapi::math::blas::rocblas::row_major::rot, + oneapi::math::blas::rocblas::row_major::rot, + oneapi::math::blas::rocblas::row_major::rot, + oneapi::math::blas::rocblas::row_major::rotg, + oneapi::math::blas::rocblas::row_major::rotg, + oneapi::math::blas::rocblas::row_major::rotg, + oneapi::math::blas::rocblas::row_major::rotg, + oneapi::math::blas::rocblas::row_major::rotm, + oneapi::math::blas::rocblas::row_major::rotm, + oneapi::math::blas::rocblas::row_major::rotmg, + oneapi::math::blas::rocblas::row_major::rotmg, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::sdsdot, + oneapi::math::blas::rocblas::row_major::swap, + oneapi::math::blas::rocblas::row_major::swap, + oneapi::math::blas::rocblas::row_major::swap, + oneapi::math::blas::rocblas::row_major::swap, + oneapi::math::blas::rocblas::row_major::gbmv, + oneapi::math::blas::rocblas::row_major::gbmv, + oneapi::math::blas::rocblas::row_major::gbmv, + oneapi::math::blas::rocblas::row_major::gbmv, + oneapi::math::blas::rocblas::row_major::gemv, + oneapi::math::blas::rocblas::row_major::gemv, + oneapi::math::blas::rocblas::row_major::gemv, + oneapi::math::blas::rocblas::row_major::gemv, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::ger, + oneapi::math::blas::rocblas::row_major::ger, + oneapi::math::blas::rocblas::row_major::gerc, + oneapi::math::blas::rocblas::row_major::gerc, + oneapi::math::blas::rocblas::row_major::geru, + oneapi::math::blas::rocblas::row_major::geru, + oneapi::math::blas::rocblas::row_major::hbmv, + oneapi::math::blas::rocblas::row_major::hbmv, + oneapi::math::blas::rocblas::row_major::hemv, + oneapi::math::blas::rocblas::row_major::hemv, + oneapi::math::blas::rocblas::row_major::her, + oneapi::math::blas::rocblas::row_major::her, + oneapi::math::blas::rocblas::row_major::her2, + oneapi::math::blas::rocblas::row_major::her2, + oneapi::math::blas::rocblas::row_major::hpmv, + oneapi::math::blas::rocblas::row_major::hpmv, + oneapi::math::blas::rocblas::row_major::hpr, + oneapi::math::blas::rocblas::row_major::hpr, + oneapi::math::blas::rocblas::row_major::hpr2, + oneapi::math::blas::rocblas::row_major::hpr2, + oneapi::math::blas::rocblas::row_major::sbmv, + oneapi::math::blas::rocblas::row_major::sbmv, + oneapi::math::blas::rocblas::row_major::spmv, + oneapi::math::blas::rocblas::row_major::spmv, + oneapi::math::blas::rocblas::row_major::spr, + oneapi::math::blas::rocblas::row_major::spr, + oneapi::math::blas::rocblas::row_major::spr2, + oneapi::math::blas::rocblas::row_major::spr2, + oneapi::math::blas::rocblas::row_major::symv, + oneapi::math::blas::rocblas::row_major::symv, + oneapi::math::blas::rocblas::row_major::syr, + oneapi::math::blas::rocblas::row_major::syr, + oneapi::math::blas::rocblas::row_major::syr2, + oneapi::math::blas::rocblas::row_major::syr2, + oneapi::math::blas::rocblas::row_major::tbmv, + oneapi::math::blas::rocblas::row_major::tbmv, + oneapi::math::blas::rocblas::row_major::tbmv, + oneapi::math::blas::rocblas::row_major::tbmv, + oneapi::math::blas::rocblas::row_major::tbsv, + oneapi::math::blas::rocblas::row_major::tbsv, + oneapi::math::blas::rocblas::row_major::tbsv, + oneapi::math::blas::rocblas::row_major::tbsv, + oneapi::math::blas::rocblas::row_major::tpmv, + oneapi::math::blas::rocblas::row_major::tpmv, + oneapi::math::blas::rocblas::row_major::tpmv, + oneapi::math::blas::rocblas::row_major::tpmv, + oneapi::math::blas::rocblas::row_major::tpsv, + oneapi::math::blas::rocblas::row_major::tpsv, + oneapi::math::blas::rocblas::row_major::tpsv, + oneapi::math::blas::rocblas::row_major::tpsv, + oneapi::math::blas::rocblas::row_major::trmv, + oneapi::math::blas::rocblas::row_major::trmv, + oneapi::math::blas::rocblas::row_major::trmv, + oneapi::math::blas::rocblas::row_major::trmv, + oneapi::math::blas::rocblas::row_major::trsv, + oneapi::math::blas::rocblas::row_major::trsv, + oneapi::math::blas::rocblas::row_major::trsv, + oneapi::math::blas::rocblas::row_major::trsv, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::hemm, + oneapi::math::blas::rocblas::row_major::hemm, + oneapi::math::blas::rocblas::row_major::herk, + oneapi::math::blas::rocblas::row_major::herk, + oneapi::math::blas::rocblas::row_major::her2k, + oneapi::math::blas::rocblas::row_major::her2k, + oneapi::math::blas::rocblas::row_major::symm, + oneapi::math::blas::rocblas::row_major::symm, + oneapi::math::blas::rocblas::row_major::symm, + oneapi::math::blas::rocblas::row_major::symm, + oneapi::math::blas::rocblas::row_major::syrk, + oneapi::math::blas::rocblas::row_major::syrk, + oneapi::math::blas::rocblas::row_major::syrk, + oneapi::math::blas::rocblas::row_major::syrk, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syr2k, + oneapi::math::blas::rocblas::row_major::syr2k, + oneapi::math::blas::rocblas::row_major::syr2k, + oneapi::math::blas::rocblas::row_major::syr2k, + oneapi::math::blas::rocblas::row_major::trmm, + oneapi::math::blas::rocblas::row_major::trmm, + oneapi::math::blas::rocblas::row_major::trmm, + oneapi::math::blas::rocblas::row_major::trmm, + oneapi::math::blas::rocblas::row_major::trsm, + oneapi::math::blas::rocblas::row_major::trsm, + oneapi::math::blas::rocblas::row_major::trsm, + oneapi::math::blas::rocblas::row_major::trsm, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::gemmt, + oneapi::math::blas::rocblas::row_major::gemmt, + oneapi::math::blas::rocblas::row_major::gemmt, + oneapi::math::blas::rocblas::row_major::gemmt, + oneapi::math::blas::rocblas::row_major::gemm_bias, + oneapi::math::blas::rocblas::row_major::gemm_bias, + oneapi::math::blas::rocblas::row_major::gemm_bias, + oneapi::math::blas::rocblas::row_major::gemm_bias, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::omatadd_batch, + oneapi::math::blas::rocblas::row_major::omatadd_batch, + oneapi::math::blas::rocblas::row_major::omatadd_batch, + oneapi::math::blas::rocblas::row_major::omatadd_batch, + oneapi::math::blas::rocblas::row_major::omatcopy, + oneapi::math::blas::rocblas::row_major::omatcopy, + oneapi::math::blas::rocblas::row_major::omatcopy, + oneapi::math::blas::rocblas::row_major::omatcopy, + oneapi::math::blas::rocblas::row_major::omatcopy2, + oneapi::math::blas::rocblas::row_major::omatcopy2, + oneapi::math::blas::rocblas::row_major::omatcopy2, + oneapi::math::blas::rocblas::row_major::omatcopy2, + oneapi::math::blas::rocblas::row_major::imatcopy, + oneapi::math::blas::rocblas::row_major::imatcopy, + oneapi::math::blas::rocblas::row_major::imatcopy, + oneapi::math::blas::rocblas::row_major::imatcopy, + oneapi::math::blas::rocblas::row_major::omatadd, + oneapi::math::blas::rocblas::row_major::omatadd, + oneapi::math::blas::rocblas::row_major::omatadd, + oneapi::math::blas::rocblas::row_major::omatadd, + oneapi::math::blas::rocblas::row_major::asum, + oneapi::math::blas::rocblas::row_major::asum, + oneapi::math::blas::rocblas::row_major::asum, + oneapi::math::blas::rocblas::row_major::asum, + oneapi::math::blas::rocblas::row_major::axpy, + oneapi::math::blas::rocblas::row_major::axpy, + oneapi::math::blas::rocblas::row_major::axpy, + oneapi::math::blas::rocblas::row_major::axpy, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpy_batch, + oneapi::math::blas::rocblas::row_major::axpby, + oneapi::math::blas::rocblas::row_major::axpby, + oneapi::math::blas::rocblas::row_major::axpby, + oneapi::math::blas::rocblas::row_major::axpby, + oneapi::math::blas::rocblas::row_major::copy, + oneapi::math::blas::rocblas::row_major::copy, + oneapi::math::blas::rocblas::row_major::copy, + oneapi::math::blas::rocblas::row_major::copy, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::copy_batch, + oneapi::math::blas::rocblas::row_major::dot, + oneapi::math::blas::rocblas::row_major::dot, + oneapi::math::blas::rocblas::row_major::dot, + oneapi::math::blas::rocblas::row_major::dotc, + oneapi::math::blas::rocblas::row_major::dotc, + oneapi::math::blas::rocblas::row_major::dotu, + oneapi::math::blas::rocblas::row_major::dotu, + oneapi::math::blas::rocblas::row_major::iamin, + oneapi::math::blas::rocblas::row_major::iamin, + oneapi::math::blas::rocblas::row_major::iamin, + oneapi::math::blas::rocblas::row_major::iamin, + oneapi::math::blas::rocblas::row_major::iamax, + oneapi::math::blas::rocblas::row_major::iamax, + oneapi::math::blas::rocblas::row_major::iamax, + oneapi::math::blas::rocblas::row_major::iamax, + oneapi::math::blas::rocblas::row_major::nrm2, + oneapi::math::blas::rocblas::row_major::nrm2, + oneapi::math::blas::rocblas::row_major::nrm2, + oneapi::math::blas::rocblas::row_major::nrm2, + oneapi::math::blas::rocblas::row_major::rot, + oneapi::math::blas::rocblas::row_major::rot, + oneapi::math::blas::rocblas::row_major::rot, + oneapi::math::blas::rocblas::row_major::rot, + oneapi::math::blas::rocblas::row_major::rotg, + oneapi::math::blas::rocblas::row_major::rotg, + oneapi::math::blas::rocblas::row_major::rotg, + oneapi::math::blas::rocblas::row_major::rotg, + oneapi::math::blas::rocblas::row_major::rotm, + oneapi::math::blas::rocblas::row_major::rotm, + oneapi::math::blas::rocblas::row_major::rotmg, + oneapi::math::blas::rocblas::row_major::rotmg, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::scal, + oneapi::math::blas::rocblas::row_major::sdsdot, + oneapi::math::blas::rocblas::row_major::swap, + oneapi::math::blas::rocblas::row_major::swap, + oneapi::math::blas::rocblas::row_major::swap, + oneapi::math::blas::rocblas::row_major::swap, + oneapi::math::blas::rocblas::row_major::gbmv, + oneapi::math::blas::rocblas::row_major::gbmv, + oneapi::math::blas::rocblas::row_major::gbmv, + oneapi::math::blas::rocblas::row_major::gbmv, + oneapi::math::blas::rocblas::row_major::gemv, + oneapi::math::blas::rocblas::row_major::gemv, + oneapi::math::blas::rocblas::row_major::gemv, + oneapi::math::blas::rocblas::row_major::gemv, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::gemv_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::dgmm_batch, + oneapi::math::blas::rocblas::row_major::ger, + oneapi::math::blas::rocblas::row_major::ger, + oneapi::math::blas::rocblas::row_major::gerc, + oneapi::math::blas::rocblas::row_major::gerc, + oneapi::math::blas::rocblas::row_major::geru, + oneapi::math::blas::rocblas::row_major::geru, + oneapi::math::blas::rocblas::row_major::hbmv, + oneapi::math::blas::rocblas::row_major::hbmv, + oneapi::math::blas::rocblas::row_major::hemv, + oneapi::math::blas::rocblas::row_major::hemv, + oneapi::math::blas::rocblas::row_major::her, + oneapi::math::blas::rocblas::row_major::her, + oneapi::math::blas::rocblas::row_major::her2, + oneapi::math::blas::rocblas::row_major::her2, + oneapi::math::blas::rocblas::row_major::hpmv, + oneapi::math::blas::rocblas::row_major::hpmv, + oneapi::math::blas::rocblas::row_major::hpr, + oneapi::math::blas::rocblas::row_major::hpr, + oneapi::math::blas::rocblas::row_major::hpr2, + oneapi::math::blas::rocblas::row_major::hpr2, + oneapi::math::blas::rocblas::row_major::sbmv, + oneapi::math::blas::rocblas::row_major::sbmv, + oneapi::math::blas::rocblas::row_major::spmv, + oneapi::math::blas::rocblas::row_major::spmv, + oneapi::math::blas::rocblas::row_major::spr, + oneapi::math::blas::rocblas::row_major::spr, + oneapi::math::blas::rocblas::row_major::spr2, + oneapi::math::blas::rocblas::row_major::spr2, + oneapi::math::blas::rocblas::row_major::symv, + oneapi::math::blas::rocblas::row_major::symv, + oneapi::math::blas::rocblas::row_major::syr, + oneapi::math::blas::rocblas::row_major::syr, + oneapi::math::blas::rocblas::row_major::syr2, + oneapi::math::blas::rocblas::row_major::syr2, + oneapi::math::blas::rocblas::row_major::tbmv, + oneapi::math::blas::rocblas::row_major::tbmv, + oneapi::math::blas::rocblas::row_major::tbmv, + oneapi::math::blas::rocblas::row_major::tbmv, + oneapi::math::blas::rocblas::row_major::tbsv, + oneapi::math::blas::rocblas::row_major::tbsv, + oneapi::math::blas::rocblas::row_major::tbsv, + oneapi::math::blas::rocblas::row_major::tbsv, + oneapi::math::blas::rocblas::row_major::tpmv, + oneapi::math::blas::rocblas::row_major::tpmv, + oneapi::math::blas::rocblas::row_major::tpmv, + oneapi::math::blas::rocblas::row_major::tpmv, + oneapi::math::blas::rocblas::row_major::tpsv, + oneapi::math::blas::rocblas::row_major::tpsv, + oneapi::math::blas::rocblas::row_major::tpsv, + oneapi::math::blas::rocblas::row_major::tpsv, + oneapi::math::blas::rocblas::row_major::trmv, + oneapi::math::blas::rocblas::row_major::trmv, + oneapi::math::blas::rocblas::row_major::trmv, + oneapi::math::blas::rocblas::row_major::trmv, + oneapi::math::blas::rocblas::row_major::trsv, + oneapi::math::blas::rocblas::row_major::trsv, + oneapi::math::blas::rocblas::row_major::trsv, + oneapi::math::blas::rocblas::row_major::trsv, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::gemm, + oneapi::math::blas::rocblas::row_major::hemm, + oneapi::math::blas::rocblas::row_major::hemm, + oneapi::math::blas::rocblas::row_major::herk, + oneapi::math::blas::rocblas::row_major::herk, + oneapi::math::blas::rocblas::row_major::her2k, + oneapi::math::blas::rocblas::row_major::her2k, + oneapi::math::blas::rocblas::row_major::symm, + oneapi::math::blas::rocblas::row_major::symm, + oneapi::math::blas::rocblas::row_major::symm, + oneapi::math::blas::rocblas::row_major::symm, + oneapi::math::blas::rocblas::row_major::syrk, + oneapi::math::blas::rocblas::row_major::syrk, + oneapi::math::blas::rocblas::row_major::syrk, + oneapi::math::blas::rocblas::row_major::syrk, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syrk_batch, + oneapi::math::blas::rocblas::row_major::syr2k, + oneapi::math::blas::rocblas::row_major::syr2k, + oneapi::math::blas::rocblas::row_major::syr2k, + oneapi::math::blas::rocblas::row_major::syr2k, + oneapi::math::blas::rocblas::row_major::trmm, + oneapi::math::blas::rocblas::row_major::trmm, + oneapi::math::blas::rocblas::row_major::trmm, + oneapi::math::blas::rocblas::row_major::trmm, + oneapi::math::blas::rocblas::row_major::trsm, + oneapi::math::blas::rocblas::row_major::trsm, + oneapi::math::blas::rocblas::row_major::trsm, + oneapi::math::blas::rocblas::row_major::trsm, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::trsm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemm_batch, + oneapi::math::blas::rocblas::row_major::gemmt, + oneapi::math::blas::rocblas::row_major::gemmt, + oneapi::math::blas::rocblas::row_major::gemmt, + oneapi::math::blas::rocblas::row_major::gemmt, + oneapi::math::blas::rocblas::row_major::gemm_bias, + oneapi::math::blas::rocblas::row_major::gemm_bias, + oneapi::math::blas::rocblas::row_major::gemm_bias, + oneapi::math::blas::rocblas::row_major::gemm_bias, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::omatadd_batch, + oneapi::math::blas::rocblas::row_major::omatadd_batch, + oneapi::math::blas::rocblas::row_major::omatadd_batch, + oneapi::math::blas::rocblas::row_major::omatadd_batch, + oneapi::math::blas::rocblas::row_major::omatcopy, + oneapi::math::blas::rocblas::row_major::omatcopy, + oneapi::math::blas::rocblas::row_major::omatcopy, + oneapi::math::blas::rocblas::row_major::omatcopy, + oneapi::math::blas::rocblas::row_major::omatcopy2, + oneapi::math::blas::rocblas::row_major::omatcopy2, + oneapi::math::blas::rocblas::row_major::omatcopy2, + oneapi::math::blas::rocblas::row_major::omatcopy2, + oneapi::math::blas::rocblas::row_major::imatcopy, + oneapi::math::blas::rocblas::row_major::imatcopy, + oneapi::math::blas::rocblas::row_major::imatcopy, + oneapi::math::blas::rocblas::row_major::imatcopy, + oneapi::math::blas::rocblas::row_major::omatadd, + oneapi::math::blas::rocblas::row_major::omatadd, + oneapi::math::blas::rocblas::row_major::omatadd, + oneapi::math::blas::rocblas::row_major::omatadd, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::omatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, + oneapi::math::blas::rocblas::row_major::imatcopy_batch, }; diff --git a/src/blas/blas_loader.cpp b/src/blas/blas_loader.cpp index c638770ec..d92e188c6 100644 --- a/src/blas/blas_loader.cpp +++ b/src/blas/blas_loader.cpp @@ -23,62 +23,62 @@ #include "blas/function_table.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace blas { namespace column_major { namespace detail { -static oneapi::mkl::detail::table_initializer function_tables; +static oneapi::math::detail::table_initializer function_tables; // Buffer APIs -void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_scasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_dzasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_sasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_dasum_sycl(queue, n, x, incx, result); } -void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_saxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_daxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, +void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_caxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zaxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -86,7 +86,7 @@ void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -94,7 +94,7 @@ void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -102,7 +102,7 @@ void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -110,58 +110,58 @@ void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_saxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_daxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_caxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zaxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_scopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_dcopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_ccopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zcopy_sycl(queue, n, x, incx, y, incy); } -void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -169,7 +169,7 @@ void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -177,7 +177,7 @@ void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -185,7 +185,7 @@ void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -193,247 +193,247 @@ void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_sdot_sycl(queue, n, x, incx, y, incy, result); } -void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_ddot_sycl(queue, n, x, incx, y, incy, result); } -void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_dsdot_sycl(queue, n, x, incx, y, incy, result); } -void dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { function_tables[{ libkey, queue }].column_major_cdotc_sycl(queue, n, x, incx, y, incy, result); } -void dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { function_tables[{ libkey, queue }].column_major_zdotc_sycl(queue, n, x, incx, y, incy, result); } -void dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { function_tables[{ libkey, queue }].column_major_cdotu_sycl(queue, n, x, incx, y, incy, result); } -void dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { function_tables[{ libkey, queue }].column_major_zdotu_sycl(queue, n, x, incx, y, incy, result); } -void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_isamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_idamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_icamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_izamin_sycl(queue, n, x, incx, result); } -void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_isamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_idamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_icamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_izamax_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_scnrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_dznrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_snrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_dnrm2_sycl(queue, n, x, incx, result); } -void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { function_tables[{ libkey, queue }].column_major_srot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { function_tables[{ libkey, queue }].column_major_drot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { function_tables[{ libkey, queue }].column_major_csrot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { function_tables[{ libkey, queue }].column_major_zdrot_sycl(queue, n, x, incx, y, incy, c, s); } -void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &a, +void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { function_tables[{ libkey, queue }].column_major_srotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &a, +void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { function_tables[{ libkey, queue }].column_major_drotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, +void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { function_tables[{ libkey, queue }].column_major_crotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, +void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { function_tables[{ libkey, queue }].column_major_zrotg_sycl(queue, a, b, c, s); } -void rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { function_tables[{ libkey, queue }].column_major_srotm_sycl(queue, n, x, incx, y, incy, param); } -void rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { function_tables[{ libkey, queue }].column_major_drotm_sycl(queue, n, x, incx, y, incy, param); } -void rotmg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &d1, +void rotmg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m) { function_tables[{ libkey, queue }].column_major_srotmg_sycl(queue, d1, d2, x1, y1, param); } -void rotmg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &d1, +void rotmg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m) { function_tables[{ libkey, queue }].column_major_drotmg_sycl(queue, d1, d2, x1, y1, param); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_sscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_cscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_csscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_zscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_zdscal_sycl(queue, n, alpha, x, incx); } -void sdsdot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float sb, +void sdsdot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float sb, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { function_tables[{ libkey, queue }].column_major_sdsdot_sycl(queue, n, sb, x, incx, y, incy, result); } -void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_sswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_dswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_cswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zswap_sycl(queue, n, x, incx, y, incy); } -void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { @@ -441,7 +441,7 @@ void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: lda, x, incx, beta, y, incy); } -void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { @@ -449,7 +449,7 @@ void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: lda, x, incx, beta, y, incy); } -void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, @@ -458,7 +458,7 @@ void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: lda, x, incx, beta, y, incy); } -void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, @@ -467,7 +467,7 @@ void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: lda, x, incx, beta, y, incy); } -void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { @@ -475,7 +475,7 @@ void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: incx, beta, y, incy); } -void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { @@ -483,7 +483,7 @@ void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: incx, beta, y, incy); } -void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -491,7 +491,7 @@ void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: incx, beta, y, incy); } -void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -499,7 +499,7 @@ void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: incx, beta, y, incy); } -void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, @@ -509,7 +509,7 @@ void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, batch_size); } -void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, @@ -519,7 +519,7 @@ void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, batch_size); } -void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, std::complex beta, @@ -530,7 +530,7 @@ void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, batch_size); } -void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, @@ -541,7 +541,7 @@ void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, batch_size); } -void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, +void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, @@ -550,7 +550,7 @@ void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, +void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, @@ -559,7 +559,7 @@ void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, +void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, @@ -568,7 +568,7 @@ void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, +void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, @@ -577,21 +577,21 @@ void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_sger_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_dger_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -599,7 +599,7 @@ void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::i a, lda); } -void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -607,7 +607,7 @@ void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::i a, lda); } -void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -615,7 +615,7 @@ void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::i a, lda); } -void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -623,7 +623,7 @@ void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::i a, lda); } -void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -631,7 +631,7 @@ void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: lda, x, incx, beta, y, incy); } -void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -639,7 +639,7 @@ void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: lda, x, incx, beta, y, incy); } -void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -647,7 +647,7 @@ void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: x, incx, beta, y, incy); } -void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -655,21 +655,21 @@ void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: x, incx, beta, y, incy); } -void her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_cher_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_zher_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -677,7 +677,7 @@ void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: incx, y, incy, a, lda); } -void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -685,7 +685,7 @@ void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: incx, y, incy, a, lda); } -void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -693,7 +693,7 @@ void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: incx, beta, y, incy); } -void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -701,21 +701,21 @@ void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: incx, beta, y, incy); } -void hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { function_tables[{ libkey, queue }].column_major_chpr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { function_tables[{ libkey, queue }].column_major_zhpr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { @@ -723,7 +723,7 @@ void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: incx, y, incy, a); } -void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { @@ -731,7 +731,7 @@ void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: incx, y, incy, a); } -void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { @@ -739,7 +739,7 @@ void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: lda, x, incx, beta, y, incy); } -void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { @@ -747,257 +747,257 @@ void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: lda, x, incx, beta, y, incy); } -void spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_sspmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_dspmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { function_tables[{ libkey, queue }].column_major_sspr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { function_tables[{ libkey, queue }].column_major_dspr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { function_tables[{ libkey, queue }].column_major_sspr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { function_tables[{ libkey, queue }].column_major_dspr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_ssymv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_dsymv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_ssyr_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_dsyr_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_ssyr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_dsyr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_stbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_stbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_stpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_stpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_strmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_strsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -1005,7 +1005,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { @@ -1013,7 +1013,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -1022,7 +1022,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -1031,7 +1031,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { @@ -1039,7 +1039,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -1047,7 +1047,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -1055,7 +1055,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -1064,7 +1064,7 @@ void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -1073,14 +1073,14 @@ void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_cherk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { @@ -1088,7 +1088,7 @@ void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tran alpha, a, lda, beta, c, ldc); } -void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, @@ -1097,7 +1097,7 @@ void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, @@ -1106,7 +1106,7 @@ void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -1114,7 +1114,7 @@ void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { @@ -1122,7 +1122,7 @@ void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -1131,7 +1131,7 @@ void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -1140,21 +1140,21 @@ void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_ssyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_dsyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { @@ -1162,7 +1162,7 @@ void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tran alpha, a, lda, beta, c, ldc); } -void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { @@ -1170,7 +1170,7 @@ void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tran alpha, a, lda, beta, c, ldc); } -void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { @@ -1179,7 +1179,7 @@ void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower batch_size); } -void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { @@ -1188,7 +1188,7 @@ void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower batch_size); } -void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, @@ -1198,7 +1198,7 @@ void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower batch_size); } -void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, @@ -1208,7 +1208,7 @@ void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower batch_size); } -void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -1216,7 +1216,7 @@ void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { @@ -1224,7 +1224,7 @@ void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -1233,7 +1233,7 @@ void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -1242,7 +1242,7 @@ void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { @@ -1250,7 +1250,7 @@ void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { @@ -1258,7 +1258,7 @@ void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { @@ -1266,7 +1266,7 @@ void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { @@ -1274,7 +1274,7 @@ void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { @@ -1282,7 +1282,7 @@ void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { @@ -1290,7 +1290,7 @@ void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { @@ -1298,7 +1298,7 @@ void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { @@ -1306,7 +1306,7 @@ void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -1317,7 +1317,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, @@ -1328,7 +1328,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, @@ -1339,7 +1339,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, @@ -1350,7 +1350,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -1361,7 +1361,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -1372,7 +1372,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -1383,7 +1383,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -1394,7 +1394,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -1404,7 +1404,7 @@ void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, stride_b, batch_size); } -void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -1414,7 +1414,7 @@ void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, stride_b, batch_size); } -void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -1424,7 +1424,7 @@ void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, stride_b, batch_size); } -void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -1434,7 +1434,7 @@ void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, stride_b, batch_size); } -void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -1442,7 +1442,7 @@ void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { @@ -1450,7 +1450,7 @@ void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -1459,7 +1459,7 @@ void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -1468,7 +1468,7 @@ void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, @@ -1477,7 +1477,7 @@ void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, @@ -1486,7 +1486,7 @@ void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, @@ -1495,7 +1495,7 @@ void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, @@ -1504,7 +1504,7 @@ void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { @@ -1512,7 +1512,7 @@ void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { @@ -1520,7 +1520,7 @@ void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -1529,7 +1529,7 @@ void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -1538,21 +1538,21 @@ void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_simatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_dimatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { @@ -1560,7 +1560,7 @@ void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { @@ -1568,7 +1568,7 @@ void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -1579,7 +1579,7 @@ void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tra stride_c, batch_size); } -void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -1590,7 +1590,7 @@ void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tra stride_c, batch_size); } -void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &b, @@ -1601,7 +1601,7 @@ void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tra stride_c, batch_size); } -void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -1613,35 +1613,35 @@ void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tra stride_c, batch_size); } -void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_somatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_domatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_comatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_zomatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { @@ -1649,7 +1649,7 @@ void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, lda, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { @@ -1657,7 +1657,7 @@ void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, lda, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { @@ -1665,7 +1665,7 @@ void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, lda, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { @@ -1673,35 +1673,35 @@ void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, lda, stridea, b, ldb, strideb); } -void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_simatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_dimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_cimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_zimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { @@ -1709,7 +1709,7 @@ void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, t queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { @@ -1717,7 +1717,7 @@ void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, t queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, @@ -1726,7 +1726,7 @@ void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, t queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, @@ -1737,47 +1737,47 @@ void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, t // USM APIs -sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_scasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dzasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_sasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_saxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_daxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { @@ -1785,7 +1785,7 @@ sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, y, incy, dependencies); } -sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { @@ -1793,7 +1793,7 @@ sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, y, incy, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, float *alpha, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -1801,7 +1801,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, double *alpha, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -1809,7 +1809,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::complex *alpha, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, @@ -1818,7 +1818,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::complex *alpha, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, @@ -1827,7 +1827,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -1835,7 +1835,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -1843,7 +1843,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, @@ -1852,7 +1852,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, @@ -1861,21 +1861,21 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_saxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double beta, double *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_daxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { @@ -1883,7 +1883,7 @@ sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { @@ -1891,35 +1891,35 @@ sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_scopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dcopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_ccopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_zcopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -1927,7 +1927,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -1935,7 +1935,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -1943,7 +1943,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -1951,7 +1951,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -1959,7 +1959,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -1967,7 +1967,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -1975,7 +1975,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -1983,28 +1983,28 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_sdot_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_ddot_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dsdot_usm_sycl( queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { @@ -2012,7 +2012,7 @@ sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { @@ -2020,7 +2020,7 @@ sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { @@ -2028,7 +2028,7 @@ sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { @@ -2036,96 +2036,96 @@ sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, y, incy, result, dependencies); } -sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_isamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_idamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_icamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_izamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_isamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_idamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_icamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_izamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_scnrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dznrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_snrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dnrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_srot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { @@ -2133,149 +2133,149 @@ sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, c, s, dependencies); } -sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *x, +sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_csrot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *x, +sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_zdrot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, float *a, float *b, float *c, +sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, float *a, float *b, float *c, float *s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_srotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, double *a, double *b, double *c, +sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, double *a, double *b, double *c, double *s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_drotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *a, +sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, std::complex *a, std::complex *b, float *c, std::complex *s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_crotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *a, +sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, std::complex *a, std::complex *b, double *c, std::complex *s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_zrotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *x, +sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float *param, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_srotm_usm_sycl( queue, n, x, incx, y, incy, param, dependencies); } -sycl::event rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *x, +sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double *param, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_drotm_usm_sycl( queue, n, x, incx, y, incy, param, dependencies); } -sycl::event rotmg(oneapi::mkl::device libkey, sycl::queue &queue, float *d1, float *d2, float *x1, +sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_srotmg_usm_sycl(queue, d1, d2, x1, y1, param, dependencies); } -sycl::event rotmg(oneapi::mkl::device libkey, sycl::queue &queue, double *d1, double *d2, +sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_drotmg_usm_sycl(queue, d1, d2, x1, y1, param, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_sscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_cscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_csscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_zscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_zdscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event sdsdot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float sb, +sycl::event sdsdot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float sb, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_sdsdot_usm_sycl( queue, n, sb, x, incx, y, incy, result, dependencies); } -sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *x, +sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_sswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *x, +sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_cswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_zswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { @@ -2283,7 +2283,7 @@ sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { @@ -2291,7 +2291,7 @@ sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, @@ -2300,7 +2300,7 @@ sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, @@ -2309,7 +2309,7 @@ sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { @@ -2317,7 +2317,7 @@ sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { @@ -2325,7 +2325,7 @@ sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -2334,7 +2334,7 @@ sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -2343,7 +2343,7 @@ sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float beta, float *y, std::int64_t incy, @@ -2354,7 +2354,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose batch_size, dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double beta, double *y, std::int64_t incy, @@ -2365,7 +2365,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose batch_size, dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, @@ -2377,7 +2377,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose batch_size, dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, @@ -2389,7 +2389,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose batch_size, dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, float **y, std::int64_t *incy, std::int64_t group_count, @@ -2399,7 +2399,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, double **y, std::int64_t *incy, std::int64_t group_count, @@ -2409,7 +2409,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, @@ -2420,7 +2420,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, @@ -2432,7 +2432,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, @@ -2442,7 +2442,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, @@ -2452,7 +2452,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *c, @@ -2463,7 +2463,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *c, @@ -2474,7 +2474,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -2483,7 +2483,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -2492,7 +2492,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, @@ -2501,7 +2501,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, @@ -2510,14 +2510,14 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_sger_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { @@ -2525,7 +2525,7 @@ sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -2533,7 +2533,7 @@ sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -2541,7 +2541,7 @@ sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -2549,7 +2549,7 @@ sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -2557,7 +2557,7 @@ sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -2566,7 +2566,7 @@ sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -2575,7 +2575,7 @@ sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -2584,7 +2584,7 @@ sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -2593,7 +2593,7 @@ sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -2601,7 +2601,7 @@ sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -2609,7 +2609,7 @@ sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -2617,7 +2617,7 @@ sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -2625,7 +2625,7 @@ sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -2634,7 +2634,7 @@ sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -2643,21 +2643,21 @@ sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_chpr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_zhpr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { @@ -2665,7 +2665,7 @@ sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { @@ -2673,7 +2673,7 @@ sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { @@ -2681,7 +2681,7 @@ sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { @@ -2689,49 +2689,49 @@ sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_sspmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dspmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_sspr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dspr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_sspr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dspr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { @@ -2739,7 +2739,7 @@ sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { @@ -2747,28 +2747,28 @@ sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_ssyr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dsyr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_ssyr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { @@ -2776,21 +2776,21 @@ sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_stbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dtbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2798,7 +2798,7 @@ sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2806,21 +2806,21 @@ sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_stbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dtbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2828,7 +2828,7 @@ sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2836,21 +2836,21 @@ sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_stpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dtpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2858,7 +2858,7 @@ sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2866,21 +2866,21 @@ sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_stpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dtpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2888,7 +2888,7 @@ sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2896,21 +2896,21 @@ sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_strmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dtrmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2918,7 +2918,7 @@ sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2926,21 +2926,21 @@ sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_strsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dtrsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2948,7 +2948,7 @@ sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -2956,7 +2956,7 @@ sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -2964,7 +2964,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -2972,7 +2972,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -2981,7 +2981,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -2990,7 +2990,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, @@ -2999,7 +2999,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3007,7 +3007,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3015,7 +3015,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -3024,7 +3024,7 @@ sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -3033,7 +3033,7 @@ sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3041,7 +3041,7 @@ sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3049,7 +3049,7 @@ sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, @@ -3058,7 +3058,7 @@ sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, double beta, std::complex *c, std::int64_t ldc, @@ -3067,7 +3067,7 @@ sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3075,7 +3075,7 @@ sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3083,7 +3083,7 @@ sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -3092,7 +3092,7 @@ sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -3101,7 +3101,7 @@ sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3109,7 +3109,7 @@ sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3117,7 +3117,7 @@ sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, @@ -3126,7 +3126,7 @@ sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, @@ -3135,7 +3135,7 @@ sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -3145,7 +3145,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upp dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -3155,7 +3155,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upp dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, @@ -3166,7 +3166,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upp dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, @@ -3177,7 +3177,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upp dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, @@ -3187,7 +3187,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo uppe batch_size, dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, @@ -3197,7 +3197,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo uppe batch_size, dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, std::int64_t ldc, @@ -3208,7 +3208,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo uppe batch_size, dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, std::int64_t ldc, @@ -3219,7 +3219,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo uppe batch_size, dependencies); } -sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3227,7 +3227,7 @@ sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3235,7 +3235,7 @@ sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -3244,7 +3244,7 @@ sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -3253,7 +3253,7 @@ sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { @@ -3262,7 +3262,7 @@ sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { @@ -3271,7 +3271,7 @@ sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, @@ -3281,7 +3281,7 @@ sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, @@ -3291,7 +3291,7 @@ sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { @@ -3300,7 +3300,7 @@ sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { @@ -3309,7 +3309,7 @@ sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, @@ -3319,7 +3319,7 @@ sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, @@ -3329,7 +3329,7 @@ sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, @@ -3339,7 +3339,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, @@ -3349,7 +3349,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, @@ -3360,7 +3360,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, @@ -3371,7 +3371,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, @@ -3381,7 +3381,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef group_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, @@ -3391,7 +3391,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef group_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, @@ -3402,7 +3402,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef group_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, @@ -3413,7 +3413,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, const float **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, @@ -3424,7 +3424,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, const double **b, std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, @@ -3435,7 +3435,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, std::complex *beta, @@ -3446,7 +3446,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, @@ -3458,7 +3458,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, sycl::half *beta, sycl::half **c, @@ -3469,7 +3469,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, @@ -3480,7 +3480,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, @@ -3491,7 +3491,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, @@ -3502,7 +3502,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -3513,7 +3513,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, @@ -3524,7 +3524,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, @@ -3536,7 +3536,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, @@ -3548,7 +3548,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, @@ -3560,7 +3560,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -3571,7 +3571,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -3582,7 +3582,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -3593,7 +3593,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3602,7 +3602,7 @@ sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low dependencies); } -sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3611,7 +3611,7 @@ sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low dependencies); } -sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, @@ -3622,7 +3622,7 @@ sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low dependencies); } -sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, @@ -3633,7 +3633,7 @@ sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low dependencies); } -sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, @@ -3644,7 +3644,7 @@ sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, @@ -3655,7 +3655,7 @@ sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, @@ -3666,7 +3666,7 @@ sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, @@ -3677,7 +3677,7 @@ sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -3686,7 +3686,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -3695,7 +3695,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -3704,7 +3704,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -3713,7 +3713,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { @@ -3721,7 +3721,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { @@ -3729,7 +3729,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, @@ -3738,7 +3738,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, @@ -3747,7 +3747,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, @@ -3758,7 +3758,7 @@ sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transp stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, @@ -3769,7 +3769,7 @@ sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transp stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -3781,7 +3781,7 @@ sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transp stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -3793,14 +3793,14 @@ sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transp stride_c, batch_size, dependencies); } -sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_somatcopy_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { @@ -3808,7 +3808,7 @@ sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose t queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { @@ -3816,7 +3816,7 @@ sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose t queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { @@ -3824,7 +3824,7 @@ sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose t queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { @@ -3832,7 +3832,7 @@ sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { @@ -3840,7 +3840,7 @@ sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, @@ -3849,7 +3849,7 @@ sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, @@ -3858,21 +3858,21 @@ sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_simatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { return function_tables[{ libkey, queue }].column_major_dimatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { @@ -3880,7 +3880,7 @@ sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose t queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { @@ -3888,7 +3888,7 @@ sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose t queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3896,7 +3896,7 @@ sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -3904,7 +3904,7 @@ sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, @@ -3913,7 +3913,7 @@ sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, @@ -3922,7 +3922,7 @@ sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, @@ -3931,7 +3931,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, @@ -3940,7 +3940,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, @@ -3949,7 +3949,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, @@ -3958,7 +3958,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, float **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { @@ -3966,7 +3966,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, double **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { @@ -3974,7 +3974,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, std::complex **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, @@ -3983,7 +3983,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, std::complex **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, @@ -3997,57 +3997,57 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans namespace row_major { namespace detail { -static oneapi::mkl::detail::table_initializer function_tables; +static oneapi::math::detail::table_initializer function_tables; // Buffer APIs -void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_scasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_dzasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_sasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_dasum_sycl(queue, n, x, incx, result); } -void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_saxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_daxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, +void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_caxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zaxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -4055,7 +4055,7 @@ void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -4063,7 +4063,7 @@ void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -4071,7 +4071,7 @@ void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -4079,58 +4079,58 @@ void axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_saxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_daxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_caxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zaxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_scopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_dcopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_ccopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zcopy_sycl(queue, n, x, incx, y, incy); } -void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -4138,7 +4138,7 @@ void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -4146,7 +4146,7 @@ void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -4154,7 +4154,7 @@ void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { @@ -4162,247 +4162,247 @@ void copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_sdot_sycl(queue, n, x, incx, y, incy, result); } -void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_ddot_sycl(queue, n, x, incx, y, incy, result); } -void dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_dsdot_sycl(queue, n, x, incx, y, incy, result); } -void dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { function_tables[{ libkey, queue }].row_major_cdotc_sycl(queue, n, x, incx, y, incy, result); } -void dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { function_tables[{ libkey, queue }].row_major_zdotc_sycl(queue, n, x, incx, y, incy, result); } -void dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { function_tables[{ libkey, queue }].row_major_cdotu_sycl(queue, n, x, incx, y, incy, result); } -void dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { function_tables[{ libkey, queue }].row_major_zdotu_sycl(queue, n, x, incx, y, incy, result); } -void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_isamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_idamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_icamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_izamin_sycl(queue, n, x, incx, result); } -void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_isamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_idamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_icamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_izamax_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_scnrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_dznrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_snrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_dnrm2_sycl(queue, n, x, incx, result); } -void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { function_tables[{ libkey, queue }].row_major_srot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { function_tables[{ libkey, queue }].row_major_drot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { function_tables[{ libkey, queue }].row_major_csrot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { function_tables[{ libkey, queue }].row_major_zdrot_sycl(queue, n, x, incx, y, incy, c, s); } -void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &a, +void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { function_tables[{ libkey, queue }].row_major_srotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &a, +void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { function_tables[{ libkey, queue }].row_major_drotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, +void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { function_tables[{ libkey, queue }].row_major_crotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, +void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { function_tables[{ libkey, queue }].row_major_zrotg_sycl(queue, a, b, c, s); } -void rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { function_tables[{ libkey, queue }].row_major_srotm_sycl(queue, n, x, incx, y, incy, param); } -void rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { function_tables[{ libkey, queue }].row_major_drotm_sycl(queue, n, x, incx, y, incy, param); } -void rotmg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &d1, +void rotmg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m) { function_tables[{ libkey, queue }].row_major_srotmg_sycl(queue, d1, d2, x1, y1, param); } -void rotmg(oneapi::mkl::device libkey, sycl::queue &queue, sycl::buffer &d1, +void rotmg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m) { function_tables[{ libkey, queue }].row_major_drotmg_sycl(queue, d1, d2, x1, y1, param); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_sscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_cscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_csscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_zscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_zdscal_sycl(queue, n, alpha, x, incx); } -void sdsdot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float sb, +void sdsdot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float sb, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { function_tables[{ libkey, queue }].row_major_sdsdot_sycl(queue, n, sb, x, incx, y, incy, result); } -void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, +void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_sswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_dswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_cswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zswap_sycl(queue, n, x, incx, y, incy); } -void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { @@ -4410,7 +4410,7 @@ void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: lda, x, incx, beta, y, incy); } -void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { @@ -4418,7 +4418,7 @@ void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: lda, x, incx, beta, y, incy); } -void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, @@ -4427,7 +4427,7 @@ void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: lda, x, incx, beta, y, incy); } -void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, @@ -4436,7 +4436,7 @@ void gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: lda, x, incx, beta, y, incy); } -void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { @@ -4444,7 +4444,7 @@ void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: incx, beta, y, incy); } -void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { @@ -4452,7 +4452,7 @@ void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: incx, beta, y, incy); } -void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -4460,7 +4460,7 @@ void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: incx, beta, y, incy); } -void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -4468,7 +4468,7 @@ void gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std:: incx, beta, y, incy); } -void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, @@ -4478,7 +4478,7 @@ void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, batch_size); } -void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, @@ -4488,7 +4488,7 @@ void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, batch_size); } -void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, std::complex beta, @@ -4499,7 +4499,7 @@ void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, batch_size); } -void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, @@ -4510,7 +4510,7 @@ void gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, batch_size); } -void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, +void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, @@ -4519,7 +4519,7 @@ void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, +void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, @@ -4528,7 +4528,7 @@ void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, +void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, @@ -4537,7 +4537,7 @@ void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, std::int64_t m, +void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, @@ -4546,21 +4546,21 @@ void dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_sger_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_dger_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -4568,7 +4568,7 @@ void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::i lda); } -void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -4576,7 +4576,7 @@ void gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::i lda); } -void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -4584,7 +4584,7 @@ void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::i lda); } -void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -4592,7 +4592,7 @@ void geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::i lda); } -void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -4600,7 +4600,7 @@ void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: x, incx, beta, y, incy); } -void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -4608,7 +4608,7 @@ void hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: x, incx, beta, y, incy); } -void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -4616,7 +4616,7 @@ void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: incx, beta, y, incy); } -void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -4624,21 +4624,21 @@ void hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: incx, beta, y, incy); } -void her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_cher_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_zher_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -4646,7 +4646,7 @@ void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: y, incy, a, lda); } -void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { @@ -4654,7 +4654,7 @@ void her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: y, incy, a, lda); } -void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -4662,7 +4662,7 @@ void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: incx, beta, y, incy); } -void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { @@ -4670,21 +4670,21 @@ void hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: incx, beta, y, incy); } -void hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { function_tables[{ libkey, queue }].row_major_chpr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { function_tables[{ libkey, queue }].row_major_zhpr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { @@ -4692,7 +4692,7 @@ void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: y, incy, a); } -void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { @@ -4700,7 +4700,7 @@ void hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: y, incy, a); } -void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { @@ -4708,7 +4708,7 @@ void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: x, incx, beta, y, incy); } -void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { @@ -4716,257 +4716,257 @@ void sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std: x, incx, beta, y, incy); } -void spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_sspmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_dspmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { function_tables[{ libkey, queue }].row_major_sspr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { function_tables[{ libkey, queue }].row_major_dspr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { function_tables[{ libkey, queue }].row_major_sspr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { function_tables[{ libkey, queue }].row_major_dspr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_ssymv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_dsymv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_ssyr_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_dsyr_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_ssyr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_dsyr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_stbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_stbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_stpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_stpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_strmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_strsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -4974,7 +4974,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { @@ -4982,7 +4982,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -4991,7 +4991,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -5000,7 +5000,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { @@ -5008,7 +5008,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -5016,7 +5016,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -5024,7 +5024,7 @@ void gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, tran queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -5033,7 +5033,7 @@ void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -5042,14 +5042,14 @@ void hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_cherk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { @@ -5057,7 +5057,7 @@ void herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tran a, lda, beta, c, ldc); } -void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, @@ -5066,7 +5066,7 @@ void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra a, lda, b, ldb, beta, c, ldc); } -void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, @@ -5075,7 +5075,7 @@ void her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -5083,7 +5083,7 @@ void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { @@ -5091,7 +5091,7 @@ void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -5100,7 +5100,7 @@ void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -5109,21 +5109,21 @@ void symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo alpha, a, lda, b, ldb, beta, c, ldc); } -void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_ssyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_dsyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { @@ -5131,7 +5131,7 @@ void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tran a, lda, beta, c, ldc); } -void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { @@ -5139,7 +5139,7 @@ void syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tran a, lda, beta, c, ldc); } -void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { @@ -5148,7 +5148,7 @@ void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower batch_size); } -void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { @@ -5157,7 +5157,7 @@ void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower batch_size); } -void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, @@ -5167,7 +5167,7 @@ void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower batch_size); } -void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, @@ -5177,7 +5177,7 @@ void syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower batch_size); } -void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -5185,7 +5185,7 @@ void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { @@ -5193,7 +5193,7 @@ void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -5202,7 +5202,7 @@ void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -5211,7 +5211,7 @@ void syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra a, lda, b, ldb, beta, c, ldc); } -void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { @@ -5219,7 +5219,7 @@ void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { @@ -5227,7 +5227,7 @@ void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { @@ -5235,7 +5235,7 @@ void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { @@ -5243,7 +5243,7 @@ void trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { @@ -5251,7 +5251,7 @@ void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { @@ -5259,7 +5259,7 @@ void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { @@ -5267,7 +5267,7 @@ void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { @@ -5275,7 +5275,7 @@ void trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo unit_diag, m, n, alpha, a, lda, b, ldb); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -5286,7 +5286,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, @@ -5297,7 +5297,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, @@ -5308,7 +5308,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, @@ -5319,7 +5319,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -5330,7 +5330,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -5341,7 +5341,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -5352,7 +5352,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -5363,7 +5363,7 @@ void gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa stride_c, batch_size); } -void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -5373,7 +5373,7 @@ void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, stride_b, batch_size); } -void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -5383,7 +5383,7 @@ void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, stride_b, batch_size); } -void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -5393,7 +5393,7 @@ void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, stride_b, batch_size); } -void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -5403,7 +5403,7 @@ void trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, stride_b, batch_size); } -void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { @@ -5411,7 +5411,7 @@ void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { @@ -5419,7 +5419,7 @@ void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -5428,7 +5428,7 @@ void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, @@ -5437,7 +5437,7 @@ void gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, tra queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, @@ -5446,7 +5446,7 @@ void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, @@ -5455,7 +5455,7 @@ void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, @@ -5464,7 +5464,7 @@ void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, @@ -5473,7 +5473,7 @@ void gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { @@ -5481,7 +5481,7 @@ void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { @@ -5489,7 +5489,7 @@ void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -5498,7 +5498,7 @@ void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, @@ -5507,21 +5507,21 @@ void omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_simatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_dimatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { @@ -5529,7 +5529,7 @@ void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { @@ -5537,7 +5537,7 @@ void imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -5548,7 +5548,7 @@ void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tra stride_c, batch_size); } -void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -5559,7 +5559,7 @@ void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tra stride_c, batch_size); } -void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &b, @@ -5570,7 +5570,7 @@ void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tra stride_c, batch_size); } -void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -5582,35 +5582,35 @@ void omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose tra stride_c, batch_size); } -void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_somatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_domatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_comatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_zomatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { @@ -5618,7 +5618,7 @@ void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { @@ -5626,7 +5626,7 @@ void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { @@ -5634,7 +5634,7 @@ void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb) { @@ -5642,35 +5642,35 @@ void omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, stridea, b, ldb, strideb); } -void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_simatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_dimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_cimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_zimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { @@ -5678,7 +5678,7 @@ void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, t a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { @@ -5686,7 +5686,7 @@ void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, t a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, @@ -5695,7 +5695,7 @@ void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, t a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, @@ -5706,47 +5706,47 @@ void omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, t // USM APIs -sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_scasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dzasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_sasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_saxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_daxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { @@ -5754,7 +5754,7 @@ sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, incy, dependencies); } -sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { @@ -5762,7 +5762,7 @@ sycl::event axpy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, incy, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, float *alpha, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -5770,7 +5770,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, double *alpha, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -5778,7 +5778,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::complex *alpha, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, @@ -5787,7 +5787,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::complex *alpha, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, @@ -5796,7 +5796,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -5804,7 +5804,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -5812,7 +5812,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, @@ -5821,7 +5821,7 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, @@ -5830,21 +5830,21 @@ sycl::event axpy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_saxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double beta, double *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_daxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { @@ -5852,7 +5852,7 @@ sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { @@ -5860,35 +5860,35 @@ sycl::event axpby(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_scopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dcopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_ccopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_zcopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const float **x, std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -5896,7 +5896,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const double **x, std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -5904,7 +5904,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -5912,7 +5912,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies) { @@ -5920,7 +5920,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -5928,7 +5928,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -5936,7 +5936,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -5944,7 +5944,7 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { @@ -5952,28 +5952,28 @@ sycl::event copy_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int6 queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_sdot_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_ddot_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dsdot_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { @@ -5981,7 +5981,7 @@ sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, result, dependencies); } -sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { @@ -5989,7 +5989,7 @@ sycl::event dotc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, result, dependencies); } -sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { @@ -5997,7 +5997,7 @@ sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, result, dependencies); } -sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { @@ -6005,96 +6005,96 @@ sycl::event dotu(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, result, dependencies); } -sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_isamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_idamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_icamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_izamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_isamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_idamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_icamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_izamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_scnrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dznrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const float *x, +sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_snrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, const double *x, +sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dnrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_srot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { @@ -6102,149 +6102,149 @@ sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, s, dependencies); } -sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *x, +sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_csrot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *x, +sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_zdrot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, float *a, float *b, float *c, +sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, float *a, float *b, float *c, float *s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_srotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, double *a, double *b, double *c, +sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, double *a, double *b, double *c, double *s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_drotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *a, +sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, std::complex *a, std::complex *b, float *c, std::complex *s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_crotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::mkl::device libkey, sycl::queue &queue, std::complex *a, +sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, std::complex *a, std::complex *b, double *c, std::complex *s, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_zrotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *x, +sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float *param, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_srotm_usm_sycl(queue, n, x, incx, y, incy, param, dependencies); } -sycl::event rotm(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *x, +sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double *param, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_drotm_usm_sycl(queue, n, x, incx, y, incy, param, dependencies); } -sycl::event rotmg(oneapi::mkl::device libkey, sycl::queue &queue, float *d1, float *d2, float *x1, +sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_srotmg_usm_sycl(queue, d1, d2, x1, y1, param, dependencies); } -sycl::event rotmg(oneapi::mkl::device libkey, sycl::queue &queue, double *d1, double *d2, +sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_drotmg_usm_sycl(queue, d1, d2, x1, y1, param, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_sscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_cscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_csscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float alpha, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_zscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double alpha, +sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_zdscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event sdsdot(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float sb, +sycl::event sdsdot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float sb, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_sdsdot_usm_sycl(queue, n, sb, x, incx, y, incy, result, dependencies); } -sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *x, +sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_sswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *x, +sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_cswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_zswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { @@ -6252,7 +6252,7 @@ sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { @@ -6260,7 +6260,7 @@ sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, @@ -6269,7 +6269,7 @@ sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, @@ -6278,7 +6278,7 @@ sycl::event gbmv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { @@ -6286,7 +6286,7 @@ sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { @@ -6294,7 +6294,7 @@ sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -6303,7 +6303,7 @@ sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -6312,7 +6312,7 @@ sycl::event gemv(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float beta, float *y, std::int64_t incy, @@ -6323,7 +6323,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose batch_size, dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double beta, double *y, std::int64_t incy, @@ -6334,7 +6334,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose batch_size, dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, @@ -6346,7 +6346,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose batch_size, dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, @@ -6358,7 +6358,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose batch_size, dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, float **y, std::int64_t *incy, std::int64_t group_count, @@ -6368,7 +6368,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, double **y, std::int64_t *incy, std::int64_t group_count, @@ -6378,7 +6378,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, @@ -6389,7 +6389,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, @@ -6401,7 +6401,7 @@ sycl::event gemv_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, @@ -6411,7 +6411,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, @@ -6421,7 +6421,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *c, @@ -6432,7 +6432,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *c, @@ -6443,7 +6443,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -6452,7 +6452,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -6461,7 +6461,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, @@ -6470,7 +6470,7 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, @@ -6479,14 +6479,14 @@ sycl::event dgmm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_sger_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { @@ -6494,7 +6494,7 @@ sycl::event ger(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -6502,7 +6502,7 @@ sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -6510,7 +6510,7 @@ sycl::event gerc(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -6518,7 +6518,7 @@ sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -6526,7 +6526,7 @@ sycl::event geru(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -6535,7 +6535,7 @@ sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -6544,7 +6544,7 @@ sycl::event hbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -6553,7 +6553,7 @@ sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -6562,7 +6562,7 @@ sycl::event hemv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -6570,7 +6570,7 @@ sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -6578,7 +6578,7 @@ sycl::event her(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -6586,7 +6586,7 @@ sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { @@ -6594,7 +6594,7 @@ sycl::event her2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -6603,7 +6603,7 @@ sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, @@ -6612,21 +6612,21 @@ sycl::event hpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_chpr_usm_sycl(queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event hpr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_zhpr_usm_sycl(queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { @@ -6634,7 +6634,7 @@ sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { @@ -6642,7 +6642,7 @@ sycl::event hpr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { @@ -6650,7 +6650,7 @@ sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { @@ -6658,49 +6658,49 @@ sycl::event sbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_sspmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event spmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dspmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_sspr_usm_sycl(queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event spr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dspr_usm_sycl(queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_sspr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event spr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dspr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { @@ -6708,7 +6708,7 @@ sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { @@ -6716,28 +6716,28 @@ sycl::event symv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_ssyr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event syr(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dsyr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_ssyr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, +sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { @@ -6745,21 +6745,21 @@ sycl::event syr2(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_stbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dtbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6767,7 +6767,7 @@ sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6775,21 +6775,21 @@ sycl::event tbmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_stbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dtbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6797,7 +6797,7 @@ sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6805,21 +6805,21 @@ sycl::event tbsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_stpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dtpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6827,7 +6827,7 @@ sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6835,21 +6835,21 @@ sycl::event tpmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_stpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dtpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6857,7 +6857,7 @@ sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6865,21 +6865,21 @@ sycl::event tpsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_strmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dtrmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6887,7 +6887,7 @@ sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6895,21 +6895,21 @@ sycl::event trmv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_strsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dtrsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6917,7 +6917,7 @@ sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { @@ -6925,7 +6925,7 @@ sycl::event trsv(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -6933,7 +6933,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -6941,7 +6941,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -6950,7 +6950,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -6959,7 +6959,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, @@ -6968,7 +6968,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -6976,7 +6976,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, transpose transb, +sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -6984,7 +6984,7 @@ sycl::event gemm(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -6993,7 +6993,7 @@ sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -7002,7 +7002,7 @@ sycl::event hemm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7010,7 +7010,7 @@ sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7018,7 +7018,7 @@ sycl::event herk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, @@ -7027,7 +7027,7 @@ sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, double beta, std::complex *c, std::int64_t ldc, @@ -7036,7 +7036,7 @@ sycl::event her2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7044,7 +7044,7 @@ sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7052,7 +7052,7 @@ sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -7061,7 +7061,7 @@ sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -7070,7 +7070,7 @@ sycl::event symm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7078,7 +7078,7 @@ sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7086,7 +7086,7 @@ sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, @@ -7095,7 +7095,7 @@ sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, @@ -7104,7 +7104,7 @@ sycl::event syrk(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lowe queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -7114,7 +7114,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upp dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, @@ -7124,7 +7124,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upp dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, @@ -7135,7 +7135,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upp dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, @@ -7146,7 +7146,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo *upp dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, @@ -7156,7 +7156,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo uppe batch_size, dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, @@ -7166,7 +7166,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo uppe batch_size, dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, std::int64_t ldc, @@ -7177,7 +7177,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo uppe batch_size, dependencies); } -sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, std::int64_t ldc, @@ -7188,7 +7188,7 @@ sycl::event syrk_batch(oneapi::mkl::device libkey, sycl::queue &queue, uplo uppe batch_size, dependencies); } -sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7196,7 +7196,7 @@ sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7204,7 +7204,7 @@ sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -7213,7 +7213,7 @@ sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, @@ -7222,7 +7222,7 @@ sycl::event syr2k(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { @@ -7231,7 +7231,7 @@ sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { @@ -7240,7 +7240,7 @@ sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, @@ -7250,7 +7250,7 @@ sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, @@ -7260,7 +7260,7 @@ sycl::event trmm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { @@ -7269,7 +7269,7 @@ sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { @@ -7278,7 +7278,7 @@ sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, @@ -7288,7 +7288,7 @@ sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, @@ -7298,7 +7298,7 @@ sycl::event trsm(oneapi::mkl::device libkey, sycl::queue &queue, side left_right dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, @@ -7308,7 +7308,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, @@ -7318,7 +7318,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, @@ -7329,7 +7329,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, @@ -7340,7 +7340,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side left stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, @@ -7350,7 +7350,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef group_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, @@ -7360,7 +7360,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef group_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, @@ -7371,7 +7371,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef group_size, dependencies); } -sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, @@ -7382,7 +7382,7 @@ sycl::event trsm_batch(oneapi::mkl::device libkey, sycl::queue &queue, side *lef group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, const float **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, @@ -7393,7 +7393,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, const double **b, std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, @@ -7404,7 +7404,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, std::complex *beta, @@ -7415,7 +7415,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, @@ -7427,7 +7427,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, sycl::half *beta, sycl::half **c, @@ -7438,7 +7438,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, @@ -7449,7 +7449,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, @@ -7460,7 +7460,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, @@ -7471,7 +7471,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose group_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -7482,7 +7482,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, @@ -7493,7 +7493,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, @@ -7505,7 +7505,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, @@ -7517,7 +7517,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, @@ -7529,7 +7529,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -7540,7 +7540,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -7551,7 +7551,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, @@ -7562,7 +7562,7 @@ sycl::event gemm_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose stride_c, batch_size, dependencies); } -sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7571,7 +7571,7 @@ sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low dependencies); } -sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7580,7 +7580,7 @@ sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low dependencies); } -sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, @@ -7591,7 +7591,7 @@ sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low dependencies); } -sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, @@ -7602,7 +7602,7 @@ sycl::event gemmt(oneapi::mkl::device libkey, sycl::queue &queue, uplo upper_low dependencies); } -sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, @@ -7613,7 +7613,7 @@ sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, @@ -7624,7 +7624,7 @@ sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, @@ -7635,7 +7635,7 @@ sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, @@ -7646,7 +7646,7 @@ sycl::event gemm_bias(oneapi::mkl::device libkey, sycl::queue &queue, transpose dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -7655,7 +7655,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -7664,7 +7664,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -7673,7 +7673,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -7682,7 +7682,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { @@ -7690,7 +7690,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies) { @@ -7698,7 +7698,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, @@ -7707,7 +7707,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, @@ -7716,7 +7716,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, @@ -7727,7 +7727,7 @@ sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transp stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, @@ -7738,7 +7738,7 @@ sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transp stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -7750,7 +7750,7 @@ sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transp stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, @@ -7762,14 +7762,14 @@ sycl::event omatadd_batch(oneapi::mkl::device libkey, sycl::queue &queue, transp stride_c, batch_size, dependencies); } -sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_somatcopy_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { @@ -7777,7 +7777,7 @@ sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose t queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { @@ -7785,7 +7785,7 @@ sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose t queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { @@ -7793,7 +7793,7 @@ sycl::event omatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose t queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { @@ -7801,7 +7801,7 @@ sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies) { @@ -7809,7 +7809,7 @@ sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, @@ -7818,7 +7818,7 @@ sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, @@ -7827,21 +7827,21 @@ sycl::event omatcopy2(oneapi::mkl::device libkey, sycl::queue &queue, transpose queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_simatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { return function_tables[{ libkey, queue }].row_major_dimatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { @@ -7849,7 +7849,7 @@ sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose t queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies) { @@ -7857,7 +7857,7 @@ sycl::event imatcopy(oneapi::mkl::device libkey, sycl::queue &queue, transpose t queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7865,7 +7865,7 @@ sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies) { @@ -7873,7 +7873,7 @@ sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, @@ -7882,7 +7882,7 @@ sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, @@ -7891,7 +7891,7 @@ sycl::event omatadd(oneapi::mkl::device libkey, sycl::queue &queue, transpose tr queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, @@ -7900,7 +7900,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, @@ -7909,7 +7909,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, @@ -7918,7 +7918,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, @@ -7927,7 +7927,7 @@ sycl::event omatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, float **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { @@ -7935,7 +7935,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, double **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies) { @@ -7943,7 +7943,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, std::complex **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, @@ -7952,7 +7952,7 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, transpose *trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, std::complex **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, @@ -7964,5 +7964,5 @@ sycl::event imatcopy_batch(oneapi::mkl::device libkey, sycl::queue &queue, trans } //namespace detail } //namespace row_major } //namespace blas -} //namespace mkl +} //namespace math } //namespace oneapi diff --git a/src/blas/function_table.hpp b/src/blas/function_table.hpp index 68ca9c282..71321bdc7 100644 --- a/src/blas/function_table.hpp +++ b/src/blas/function_table.hpp @@ -251,53 +251,53 @@ typedef struct { void (*column_major_zswap_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_sgbmv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_sgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_dgbmv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_dgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_cgbmv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_cgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zgbmv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_zgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_sgemv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_sgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_dgemv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_dgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_cgemv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_cgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zgemv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_zgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_sgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_sgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, @@ -305,7 +305,7 @@ typedef struct { float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_dgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_dgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, @@ -314,25 +314,25 @@ typedef struct { std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); void (*column_major_cgemv_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); void (*column_major_zgemv_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_sdgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, + void (*column_major_sdgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); - void (*column_major_ddgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, + void (*column_major_ddgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, @@ -340,13 +340,13 @@ typedef struct { sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); void (*column_major_cdgmm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); void (*column_major_zdgmm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, @@ -379,489 +379,489 @@ typedef struct { sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_chbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_chbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zhbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_zhbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_chemv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_chemv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zhemv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_zhemv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_cher_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_cher_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_zher_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_zher_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_cher2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_cher2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_zher2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_zher2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_chpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_chpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zhpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_zhpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_chpr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_chpr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); - void (*column_major_zhpr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_zhpr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); - void (*column_major_chpr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_chpr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); - void (*column_major_zhpr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_zhpr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); - void (*column_major_ssbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_ssbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_dsbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_dsbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_sspmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_sspmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_dspmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_dspmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_sspr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_sspr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); - void (*column_major_dspr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_dspr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); - void (*column_major_sspr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_sspr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); - void (*column_major_dspr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_dspr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); - void (*column_major_ssymv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_ssymv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_dsymv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_dsymv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_ssyr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_ssyr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); - void (*column_major_dsyr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_dsyr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); - void (*column_major_ssyr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_ssyr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); - void (*column_major_dsyr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + void (*column_major_dsyr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); - void (*column_major_stbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_stbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*column_major_dtbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_dtbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*column_major_ctbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ctbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_ztbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ztbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_stbsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_stbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*column_major_dtbsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_dtbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*column_major_ctbsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ctbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_ztbsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ztbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_stpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_stpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); - void (*column_major_dtpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_dtpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); - void (*column_major_ctpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ctpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_ztpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ztpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_stpsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_stpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); - void (*column_major_dtpsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_dtpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); - void (*column_major_ctpsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ctpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_ztpsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ztpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_strmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_strmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*column_major_dtrmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_dtrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*column_major_ctrmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ctrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_ztrmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ztrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_strsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_strsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*column_major_dtrsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_dtrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*column_major_ctrsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ctrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_ztrsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*column_major_ztrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_sgemm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*column_major_sgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_dgemm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*column_major_dgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_cgemm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*column_major_cgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zgemm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*column_major_zgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_hgemm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*column_major_hgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_gemm_f16f16f32_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*column_major_gemm_f16f16f32_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_gemm_bf16bf16f32_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*column_major_gemm_bf16bf16f32_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, + sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, + sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_chemm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*column_major_chemm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zhemm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*column_major_zhemm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_cherk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_cherk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zherk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_zherk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_cher2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_cher2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zher2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_zher2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_ssymm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*column_major_ssymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_dsymm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*column_major_dsymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_csymm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*column_major_csymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zsymm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*column_major_zsymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_ssyrk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_ssyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_dsyrk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_dsyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_csyrk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_csyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zsyrk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_zsyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_ssyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + void (*column_major_ssyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*column_major_dsyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + void (*column_major_dsyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_csyrk_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_zsyrk_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*column_major_ssyr2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_ssyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_dsyr2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_dsyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_csyr2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_csyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zsyr2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*column_major_zsyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_strmm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*column_major_strmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*column_major_dtrmm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*column_major_dtrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*column_major_ctrmm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*column_major_ctrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); - void (*column_major_ztrmm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*column_major_ztrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); - void (*column_major_strsm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*column_major_strsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*column_major_dtrsm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*column_major_dtrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*column_major_ctrsm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*column_major_ctrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); - void (*column_major_ztrsm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*column_major_ztrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); void (*column_major_sgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_dgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_cgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_zgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*column_major_hgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*column_major_hgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, @@ -870,245 +870,245 @@ typedef struct { sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_gemm_f16f16f32_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_gemm_s8s8f32_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_gemm_s8s8s32_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_strsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_dtrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_ctrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_ztrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - void (*column_major_sgemmt_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + void (*column_major_sgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_dgemmt_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + void (*column_major_dgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_cgemmt_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + void (*column_major_cgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zgemmt_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + void (*column_major_zgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); void (*column_major_gemm_s8u8s32_bias_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); void (*column_major_gemm_s8s8s32_bias_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); void (*column_major_gemm_u8s8s32_bias_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); void (*column_major_gemm_u8u8s32_bias_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); void (*column_major_somatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_domatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_comatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_zomatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_simatcopy_batch_strided_sycl)(sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t m, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); void (*column_major_dimatcopy_batch_strided_sycl)(sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t m, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); void (*column_major_cimatcopy_batch_strided_sycl)(sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t m, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); void (*column_major_zimatcopy_batch_strided_sycl)(sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t m, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); void (*column_major_somatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_domatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_comatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_zomatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*column_major_somatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_somatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*column_major_domatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_domatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*column_major_comatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_comatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); - void (*column_major_zomatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_zomatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); - void (*column_major_somatcopy2_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_somatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); - void (*column_major_domatcopy2_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_domatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); - void (*column_major_comatcopy2_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_comatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); - void (*column_major_zomatcopy2_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_zomatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); - void (*column_major_simatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_simatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); - void (*column_major_dimatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_dimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); - void (*column_major_cimatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_cimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); - void (*column_major_zimatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*column_major_zimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); - void (*column_major_somatadd_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*column_major_somatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); - void (*column_major_domatadd_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*column_major_domatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); - void (*column_major_comatadd_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*column_major_comatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zomatadd_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*column_major_zomatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, @@ -1409,24 +1409,24 @@ typedef struct { std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_sgbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_sgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_dgbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_dgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies); sycl::event (*column_major_cgbmv_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_zgbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_zgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -1434,114 +1434,114 @@ typedef struct { std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_sgemv_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_sgemv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_dgemv_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_dgemv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies); sycl::event (*column_major_cgemv_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); sycl::event (*column_major_zgemv_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); sycl::event (*column_major_sgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float beta, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_dgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double beta, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_cgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_zgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_sgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_dgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_cgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_zgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_sdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_ddgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_cdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_zdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_sdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_ddgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_cdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_zdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); @@ -1580,891 +1580,891 @@ typedef struct { std::complex *a, std::int64_t lda, const std::vector &dependencies); sycl::event (*column_major_chbmv_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, + sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); sycl::event (*column_major_zhbmv_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, + sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_chemv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_chemv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_zhemv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_zhemv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_cher_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_cher_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*column_major_zher_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_zher_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*column_major_cher2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_cher2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*column_major_zher2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_zher2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*column_major_chpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_chpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_zhpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_zhpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_chpr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_chpr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies); - sycl::event (*column_major_zhpr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_zhpr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies); - sycl::event (*column_major_chpr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_chpr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies); - sycl::event (*column_major_zhpr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_zhpr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies); - sycl::event (*column_major_ssbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_ssbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_dsbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_dsbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_sspmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_sspmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_dspmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_dspmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_sspr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_sspr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies); - sycl::event (*column_major_dspr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_dspr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies); - sycl::event (*column_major_sspr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_sspr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies); - sycl::event (*column_major_dspr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_dspr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies); - sycl::event (*column_major_ssymv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_ssymv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_dsymv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_dsymv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_ssyr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_ssyr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*column_major_dsyr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_dsyr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*column_major_ssyr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_ssyr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*column_major_dsyr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*column_major_dsyr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*column_major_stbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_stbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_dtbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_dtbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ctbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ctbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ztbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ztbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_stbsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_stbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_dtbsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_dtbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ctbsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ctbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ztbsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ztbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_stpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_stpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_dtpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_dtpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ctpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ctpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ztpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ztpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_stpsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_stpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_dtpsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_dtpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ctpsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ctpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ztpsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ztpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_strmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_strmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_dtrmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_dtrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ctrmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ctrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ztrmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ztrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_strsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_strsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_dtrsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_dtrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ctrsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ctrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_ztrsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*column_major_ztrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*column_major_sgemm_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*column_major_sgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_dgemm_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*column_major_dgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_cgemm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_zgemm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_hgemm_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*column_major_hgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_gemm_f16f16f32_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_gemm_bf16bf16f32_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const oneapi::mkl::bfloat16 *a, - std::int64_t lda, const oneapi::mkl::bfloat16 *b, std::int64_t ldb, float beta, float *c, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const oneapi::math::bfloat16 *a, + std::int64_t lda, const oneapi::math::bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_chemm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_zhemm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, + sycl::event (*column_major_zhemm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_cherk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*column_major_cherk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_zherk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*column_major_zherk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_cher2k_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_zher2k_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_ssymm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, + sycl::event (*column_major_ssymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_dsymm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, + sycl::event (*column_major_dsymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_csymm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_zsymm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, + sycl::event (*column_major_zsymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_ssyrk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*column_major_ssyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_dsyrk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*column_major_dsyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_csyrk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*column_major_csyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_zsyrk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*column_major_zsyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_ssyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, + sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_dsyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, + sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_csyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, + sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_zsyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, + sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_ssyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_dsyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_csyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_zsyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); - sycl::event (*column_major_ssyr2k_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*column_major_ssyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_dsyr2k_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*column_major_dsyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_csyr2k_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_zsyr2k_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*column_major_zsyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_strmm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, + sycl::event (*column_major_strmm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*column_major_dtrmm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, + sycl::event (*column_major_dtrmm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies); sycl::event (*column_major_ctrmm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); sycl::event (*column_major_ztrmm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*column_major_strsm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, + sycl::event (*column_major_strsm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*column_major_dtrsm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, + sycl::event (*column_major_dtrsm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies); sycl::event (*column_major_ctrsm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); sycl::event (*column_major_ztrsm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); sycl::event (*column_major_strsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_dtrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_ctrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_ztrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_strsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, + sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_dtrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, + sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_ctrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, + sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_ztrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, + sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_sgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, const float **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_dgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, const double **b, std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_cgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_zgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_hgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, sycl::half *beta, sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_gemm_f16f16f32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_gemm_s8s8f32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_gemm_s8s8s32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*column_major_sgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_dgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_cgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_zgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_hgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_gemm_f16f16f32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_gemm_s8s8f32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_gemm_s8s8s32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); - sycl::event (*column_major_sgemmt_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, + sycl::event (*column_major_sgemmt_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_dgemmt_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, + sycl::event (*column_major_dgemmt_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_cgemmt_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_zgemmt_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_gemm_s8u8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies); sycl::event (*column_major_gemm_s8s8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies); sycl::event (*column_major_gemm_u8s8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies); sycl::event (*column_major_gemm_u8u8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies); sycl::event (*column_major_somatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_domatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_comatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_zomatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_simatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_dimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_cimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_zimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_somatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_domatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_comatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*column_major_zomatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); - sycl::event (*column_major_somatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_somatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*column_major_domatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_domatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*column_major_comatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_comatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*column_major_zomatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_zomatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); sycl::event (*column_major_somatcopy2_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies); sycl::event (*column_major_domatcopy2_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies); sycl::event (*column_major_comatcopy2_usm_sycl)(sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t m, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies); sycl::event (*column_major_zomatcopy2_usm_sycl)(sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t m, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies); - sycl::event (*column_major_simatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_simatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*column_major_dimatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_dimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*column_major_cimatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_cimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*column_major_zimatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*column_major_zimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*column_major_somatadd_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*column_major_somatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_domatadd_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*column_major_domatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_comatadd_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_zomatadd_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*column_major_zomatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, @@ -2472,40 +2472,40 @@ typedef struct { std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*column_major_somatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*column_major_domatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*column_major_comatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*column_major_zomatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*column_major_simatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, float **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*column_major_dimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, double **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*column_major_cimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, std::complex **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*column_major_zimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, std::complex **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); @@ -2724,57 +2724,57 @@ typedef struct { void (*row_major_zswap_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_sgbmv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, + void (*row_major_sgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_dgbmv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, + void (*row_major_dgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_cgbmv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, + void (*row_major_cgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_zgbmv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, + void (*row_major_zgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_sgemv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, + void (*row_major_sgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_dgemv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, + void (*row_major_dgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_cgemv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, + void (*row_major_cgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_zgemv_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, + void (*row_major_zgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_sgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_sgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_dgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_dgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, @@ -2782,25 +2782,25 @@ typedef struct { sycl::buffer &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); void (*row_major_cgemv_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); void (*row_major_zgemv_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_sdgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, + void (*row_major_sdgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, std::int64_t incx, std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); - void (*row_major_ddgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, + void (*row_major_ddgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &x, @@ -2808,13 +2808,13 @@ typedef struct { sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); void (*row_major_cdgmm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); void (*row_major_zdgmm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, @@ -2847,455 +2847,455 @@ typedef struct { sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); - void (*row_major_chbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_chbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_zhbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_zhbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_chemv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_chemv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_zhemv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_zhemv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_cher_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_cher_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); - void (*row_major_zher_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_zher_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda); - void (*row_major_cher2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_cher2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); - void (*row_major_zher2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_zher2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda); - void (*row_major_chpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_chpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_zhpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_zhpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_chpr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_chpr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); - void (*row_major_zhpr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_zhpr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a); - void (*row_major_chpr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_chpr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); - void (*row_major_zhpr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_zhpr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a); - void (*row_major_ssbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_ssbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_dsbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_dsbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_sspmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_sspmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_dspmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_dspmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_sspr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_sspr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); - void (*row_major_dspr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_dspr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a); - void (*row_major_sspr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_sspr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); - void (*row_major_dspr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_dspr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a); - void (*row_major_ssymv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_ssymv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_dsymv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_dsymv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_ssyr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_ssyr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); - void (*row_major_dsyr_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_dsyr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); - void (*row_major_ssyr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_ssyr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); - void (*row_major_dsyr2_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, + void (*row_major_dsyr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda); - void (*row_major_stbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_stbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_dtbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_dtbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_ctbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ctbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_ztbmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ztbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_stbsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_stbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_dtbsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_dtbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_ctbsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ctbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_ztbsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ztbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_stpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_stpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); - void (*row_major_dtpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_dtpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); - void (*row_major_ctpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ctpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_ztpmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ztpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_stpsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_stpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); - void (*row_major_dtpsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_dtpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx); - void (*row_major_ctpsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ctpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_ztpsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ztpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_strmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_strmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_dtrmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_dtrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_ctrmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ctrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_ztrmv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ztrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_strsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_strsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_dtrsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_dtrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_ctrsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ctrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_ztrsv_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, + void (*row_major_ztrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_sgemm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*row_major_sgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_dgemm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*row_major_dgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_cgemm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*row_major_cgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zgemm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*row_major_zgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_hgemm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*row_major_hgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_gemm_f16f16f32_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*row_major_gemm_f16f16f32_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_gemm_bf16bf16f32_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*row_major_gemm_bf16bf16f32_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, + sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, + sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_chemm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*row_major_chemm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zhemm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*row_major_zhemm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_cherk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_cherk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zherk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_zherk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_cher2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_cher2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zher2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_zher2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_ssymm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*row_major_ssymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_dsymm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*row_major_dsymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_csymm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*row_major_csymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zsymm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, std::int64_t n, + void (*row_major_zsymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_ssyrk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_ssyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_dsyrk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_dsyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_csyrk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_csyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zsyrk_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_zsyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_ssyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + void (*row_major_ssyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*row_major_dsyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + void (*row_major_dsyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_csyrk_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_zsyrk_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*row_major_ssyr2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_ssyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_dsyr2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_dsyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_csyr2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_csyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zsyr2k_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t k, + void (*row_major_zsyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_strmm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*row_major_strmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*row_major_dtrmm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*row_major_dtrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*row_major_ctrmm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*row_major_ctrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); - void (*row_major_ztrmm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*row_major_ztrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); - void (*row_major_strsm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*row_major_strsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*row_major_dtrsm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*row_major_dtrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*row_major_ctrsm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*row_major_ctrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); - void (*row_major_ztrsm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + void (*row_major_ztrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); - void (*row_major_sgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*row_major_sgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*row_major_dgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*row_major_dgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, @@ -3303,21 +3303,21 @@ typedef struct { sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_cgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_zgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*row_major_hgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + void (*row_major_hgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, @@ -3326,242 +3326,242 @@ typedef struct { sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_gemm_f16f16f32_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_gemm_s8s8f32_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_gemm_s8s8s32_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_strsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*row_major_dtrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*row_major_ctrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*row_major_ztrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - void (*row_major_sgemmt_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + void (*row_major_sgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_dgemmt_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + void (*row_major_dgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_cgemmt_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + void (*row_major_cgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zgemmt_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + void (*row_major_zgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); void (*row_major_gemm_s8u8s32_bias_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); void (*row_major_gemm_s8s8s32_bias_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); void (*row_major_gemm_u8s8s32_bias_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); void (*row_major_gemm_u8u8s32_bias_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); - void (*row_major_somatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_somatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - void (*row_major_domatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_domatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*row_major_comatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*row_major_zomatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - void (*row_major_simatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_simatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - void (*row_major_dimatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_dimatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - void (*row_major_cimatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_cimatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - void (*row_major_zimatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_zimatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); void (*row_major_somatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_domatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_comatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_zomatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*row_major_somatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_somatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*row_major_domatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_domatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - void (*row_major_comatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_comatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); - void (*row_major_zomatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_zomatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); - void (*row_major_somatcopy2_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_somatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); - void (*row_major_domatcopy2_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_domatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); - void (*row_major_comatcopy2_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_comatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); - void (*row_major_zomatcopy2_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_zomatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t strideb); - void (*row_major_simatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_simatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); - void (*row_major_dimatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_dimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); - void (*row_major_cimatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_cimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); - void (*row_major_zimatcopy_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + void (*row_major_zimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb); - void (*row_major_somatadd_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*row_major_somatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); - void (*row_major_domatadd_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*row_major_domatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, std::int64_t ldc); - void (*row_major_comatadd_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*row_major_comatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zomatadd_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, std::int64_t n, + void (*row_major_zomatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, @@ -3858,24 +3858,24 @@ typedef struct { std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_sgbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_sgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_dgbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_dgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies); sycl::event (*row_major_cgbmv_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_zgbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_zgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, std::int64_t lda, @@ -3883,19 +3883,19 @@ typedef struct { std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_sgemv_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_sgemv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_dgemv_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_dgemv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_cgemv_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_cgemv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, @@ -3903,96 +3903,96 @@ typedef struct { std::complex *y, std::int64_t incy, const std::vector &dependencies); sycl::event (*row_major_zgemv_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); sycl::event (*row_major_sgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float beta, float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_dgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double beta, double *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_cgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_zgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_sgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_dgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_cgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_zgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex *beta, std::complex **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_sdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_ddgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_cdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_zdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, const std::complex *a, std::int64_t lda, std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_sdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_ddgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_cdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_zdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, const std::complex **a, std::int64_t *lda, const std::complex **x, std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); @@ -4030,7 +4030,7 @@ typedef struct { const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*row_major_chbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_chbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, @@ -4038,309 +4038,309 @@ typedef struct { std::complex *y, std::int64_t incy, const std::vector &dependencies); sycl::event (*row_major_zhbmv_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, std::int64_t n, std::int64_t k, + sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_chemv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_chemv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_zhemv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_zhemv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_cher_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_cher_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*row_major_zher_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_zher_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*row_major_cher2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_cher2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*row_major_zher2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_zher2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*row_major_chpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_chpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_zhpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_zhpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_chpr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_chpr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies); - sycl::event (*row_major_zhpr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_zhpr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies); - sycl::event (*row_major_chpr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_chpr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies); - sycl::event (*row_major_zhpr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_zhpr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies); - sycl::event (*row_major_ssbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_ssbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_dsbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_dsbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_sspmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_sspmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_dspmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_dspmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_sspr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_sspr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies); - sycl::event (*row_major_dspr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_dspr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies); - sycl::event (*row_major_sspr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_sspr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies); - sycl::event (*row_major_dspr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_dspr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies); - sycl::event (*row_major_ssymv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_ssymv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_dsymv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_dsymv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_ssyr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_ssyr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*row_major_dsyr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_dsyr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*row_major_ssyr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_ssyr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*row_major_dsyr2_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, + sycl::event (*row_major_dsyr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies); - sycl::event (*row_major_stbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_stbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_dtbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_dtbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ctbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ctbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ztbmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ztbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_stbsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_stbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_dtbsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_dtbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ctbsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ctbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ztbsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ztbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_stpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_stpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_dtpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_dtpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ctpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ctpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ztpmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ztpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_stpsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_stpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_dtpsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_dtpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ctpsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ctpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ztpsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ztpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_strmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_strmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_dtrmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_dtrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ctrmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ctrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ztrmv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ztrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_strsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_strsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_dtrsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_dtrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ctrsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ctrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_ztrsv_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t n, + sycl::event (*row_major_ztrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies); - sycl::event (*row_major_sgemm_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*row_major_sgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_dgemm_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*row_major_dgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_cgemm_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*row_major_cgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, @@ -4348,582 +4348,582 @@ typedef struct { std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_zgemm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_hgemm_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*row_major_hgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_gemm_f16f16f32_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_gemm_bf16bf16f32_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const oneapi::mkl::bfloat16 *a, - std::int64_t lda, const oneapi::mkl::bfloat16 *b, std::int64_t ldb, float beta, float *c, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const oneapi::math::bfloat16 *a, + std::int64_t lda, const oneapi::math::bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_chemm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_zhemm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, + sycl::event (*row_major_zhemm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_cherk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*row_major_cherk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_zherk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*row_major_zherk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_cher2k_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*row_major_cher2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_zher2k_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*row_major_zher2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_ssymm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, + sycl::event (*row_major_ssymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_dsymm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, + sycl::event (*row_major_dsymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_csymm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_zsymm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, std::int64_t m, + sycl::event (*row_major_zsymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_ssyrk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*row_major_ssyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_dsyrk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*row_major_dsyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_csyrk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*row_major_csyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_zsyrk_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*row_major_zsyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_ssyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, + sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_dsyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, + sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_csyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, + sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_zsyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo *upper_lower, oneapi::mkl::transpose *trans, + sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_ssyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_dsyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_csyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_zsyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); - sycl::event (*row_major_ssyr2k_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*row_major_ssyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_dsyr2k_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*row_major_dsyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_csyr2k_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose trans, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_zsyr2k_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*row_major_zsyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_strmm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, + sycl::event (*row_major_strmm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_dtrmm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, + sycl::event (*row_major_dtrmm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies); sycl::event (*row_major_ctrmm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); sycl::event (*row_major_ztrmm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_strsm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, + sycl::event (*row_major_strsm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_dtrsm_usm_sycl)(sycl::queue &queue, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, - oneapi::mkl::diag unit_diag, std::int64_t m, + sycl::event (*row_major_dtrsm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies); sycl::event (*row_major_ctrsm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); sycl::event (*row_major_ztrsm_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); sycl::event (*row_major_strsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_dtrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_ctrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_ztrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side left_right, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, oneapi::mkl::diag unit_diag, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_strsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, + sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_dtrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, + sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_ctrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, + sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_ztrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::side *left_right, oneapi::mkl::uplo *upper_lower, - oneapi::mkl::transpose *trans, oneapi::mkl::diag *unit_diag, std::int64_t *m, + sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, + oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_sgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, const float **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_dgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, const double **b, std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_cgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_zgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, const std::complex **b, std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_hgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, sycl::half *beta, sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_gemm_f16f16f32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_gemm_s8s8f32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_gemm_s8s8s32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *transa, oneapi::mkl::transpose *transb, + sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, const std::vector &dependencies); sycl::event (*row_major_sgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_dgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_cgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_zgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_hgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_gemm_f16f16f32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_gemm_s8s8f32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_gemm_s8s8s32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); - sycl::event (*row_major_sgemmt_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, + sycl::event (*row_major_sgemmt_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_dgemmt_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, + sycl::event (*row_major_dgemmt_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_cgemmt_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_zgemmt_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_gemm_s8u8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies); sycl::event (*row_major_gemm_s8s8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies); sycl::event (*row_major_gemm_u8s8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies); sycl::event (*row_major_gemm_u8u8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, - oneapi::mkl::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, const std::vector &dependencies); sycl::event (*row_major_somatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_domatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_comatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_zomatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_simatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_dimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_cimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_zimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_somatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float beta, const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_domatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double beta, const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_comatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); sycl::event (*row_major_zomatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, const std::vector &dependencies); - sycl::event (*row_major_somatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_somatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_domatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_domatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_comatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_comatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_zomatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_zomatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_somatcopy2_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_somatcopy2_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies); - sycl::event (*row_major_domatcopy2_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_domatcopy2_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies); - sycl::event (*row_major_comatcopy2_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_comatcopy2_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies); - sycl::event (*row_major_zomatcopy2_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_zomatcopy2_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, std::complex *b, std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies); - sycl::event (*row_major_simatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_simatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_dimatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_dimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_cimatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_cimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_zimatcopy_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*row_major_zimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_somatadd_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*row_major_somatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_domatadd_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*row_major_domatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_comatadd_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, + sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_zomatadd_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, std::int64_t m, + sycl::event (*row_major_zomatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, @@ -4931,40 +4931,40 @@ typedef struct { std::complex *c, std::int64_t ldc, const std::vector &dependencies); sycl::event (*row_major_somatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*row_major_domatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*row_major_comatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*row_major_zomatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*row_major_simatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, float *alpha, float **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*row_major_dimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, double *alpha, double **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*row_major_cimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, std::complex **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); sycl::event (*row_major_zimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *m, std::int64_t *n, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, std::complex *alpha, std::complex **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, const std::vector &dependencies); diff --git a/src/dft/backends/backend_compute_signature.cxx b/src/dft/backends/backend_compute_signature.cxx index d011cb995..0a154eeb6 100644 --- a/src/dft/backends/backend_compute_signature.cxx +++ b/src/dft/backends/backend_compute_signature.cxx @@ -32,12 +32,12 @@ using descriptor_type = typename dft::detail::descriptor; void forward_ip_cc(descriptor_type& desc, sycl::buffer& inout) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_forward"); - oneapi::mkl::dft::BACKEND::compute_forward(desc, inout); + oneapi::math::dft::BACKEND::compute_forward(desc, inout); } sycl::event forward_ip_cc(descriptor_type& desc, fwd_type* inout, const std::vector& dependencies) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_forward"); - return oneapi::mkl::dft::BACKEND::compute_forward(desc, inout, dependencies); + return oneapi::math::dft::BACKEND::compute_forward(desc, inout, dependencies); } // forward inplace REAL_REAL @@ -45,12 +45,12 @@ void forward_ip_rr(descriptor_type& desc, sycl::buffer& inout_re sycl::buffer& inout_im) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_forward"); - oneapi::mkl::dft::BACKEND::compute_forward(desc, inout_re, inout_im); + oneapi::math::dft::BACKEND::compute_forward(desc, inout_re, inout_im); } sycl::event forward_ip_rr(descriptor_type& desc, scalar_type* inout_re, scalar_type* inout_im, const std::vector& dependencies) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_forward"); - return oneapi::mkl::dft::BACKEND::compute_forward(desc, inout_re, inout_im, dependencies); + return oneapi::math::dft::BACKEND::compute_forward(desc, inout_re, inout_im, dependencies); } // forward out-of-place COMPLEX_COMPLEX @@ -58,12 +58,12 @@ void forward_op_cc(descriptor_type& desc, sycl::buffer& in, sycl::buffer& out) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_forward"); - oneapi::mkl::dft::BACKEND::compute_forward(desc, in, out); + oneapi::math::dft::BACKEND::compute_forward(desc, in, out); } sycl::event forward_op_cc(descriptor_type& desc, fwd_type* in, bwd_type* out, const std::vector& dependencies) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_forward"); - return oneapi::mkl::dft::BACKEND::compute_forward(desc, in, out, dependencies); + return oneapi::math::dft::BACKEND::compute_forward(desc, in, out, dependencies); } // forward out-of-place REAL_REAL @@ -72,13 +72,13 @@ void forward_op_rr(descriptor_type& desc, sycl::buffer& in_re, sycl::buffer& out_im) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_forward"); - oneapi::mkl::dft::BACKEND::compute_forward(desc, in_re, in_im, out_re, out_im); + oneapi::math::dft::BACKEND::compute_forward(desc, in_re, in_im, out_re, out_im); } sycl::event forward_op_rr(descriptor_type& desc, scalar_type* in_re, scalar_type* in_im, scalar_type* out_re, scalar_type* out_im, const std::vector& dependencies) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_forward"); - return oneapi::mkl::dft::BACKEND::compute_forward(desc, in_re, in_im, out_re, out_im, + return oneapi::math::dft::BACKEND::compute_forward(desc, in_re, in_im, out_re, out_im, dependencies); } @@ -86,12 +86,12 @@ sycl::event forward_op_rr(descriptor_type& desc, scalar_type* in_re, scalar_type void backward_ip_cc(descriptor_type& desc, sycl::buffer& inout) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_backward"); - oneapi::mkl::dft::BACKEND::compute_backward(desc, inout); + oneapi::math::dft::BACKEND::compute_backward(desc, inout); } sycl::event backward_ip_cc(descriptor_type& desc, fwd_type* inout, const std::vector& dependencies) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_backward"); - return oneapi::mkl::dft::BACKEND::compute_backward(desc, inout, dependencies); + return oneapi::math::dft::BACKEND::compute_backward(desc, inout, dependencies); } // backward inplace REAL_REAL @@ -99,12 +99,12 @@ void backward_ip_rr(descriptor_type& desc, sycl::buffer& inout_r sycl::buffer& inout_im) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_backward"); - oneapi::mkl::dft::BACKEND::compute_backward(desc, inout_re, inout_im); + oneapi::math::dft::BACKEND::compute_backward(desc, inout_re, inout_im); } sycl::event backward_ip_rr(descriptor_type& desc, scalar_type* inout_re, scalar_type* inout_im, const std::vector& dependencies) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_backward"); - return oneapi::mkl::dft::BACKEND::compute_backward(desc, inout_re, inout_im, dependencies); + return oneapi::math::dft::BACKEND::compute_backward(desc, inout_re, inout_im, dependencies); } // backward out-of-place COMPLEX_COMPLEX @@ -112,12 +112,12 @@ void backward_op_cc(descriptor_type& desc, sycl::buffer& in, sycl::buffer& out) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_backward"); - oneapi::mkl::dft::BACKEND::compute_backward(desc, in, out); + oneapi::math::dft::BACKEND::compute_backward(desc, in, out); } sycl::event backward_op_cc(descriptor_type& desc, bwd_type* in, fwd_type* out, const std::vector& dependencies) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_backward"); - return oneapi::mkl::dft::BACKEND::compute_backward(desc, in, out, dependencies); + return oneapi::math::dft::BACKEND::compute_backward(desc, in, out, dependencies); } // backward out-of-place REAL_REAL @@ -126,12 +126,12 @@ void backward_op_rr(descriptor_type& desc, sycl::buffer& in_re, sycl::buffer& out_im) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_backward"); - oneapi::mkl::dft::BACKEND::compute_backward(desc, in_re, in_im, out_re, out_im); + oneapi::math::dft::BACKEND::compute_backward(desc, in_re, in_im, out_re, out_im); } sycl::event backward_op_rr(descriptor_type& desc, scalar_type* in_re, scalar_type* in_im, scalar_type* out_re, scalar_type* out_im, const std::vector& dependencies) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_backward"); - return oneapi::mkl::dft::BACKEND::compute_backward(desc, in_re, in_im, out_re, out_im, + return oneapi::math::dft::BACKEND::compute_backward(desc, in_re, in_im, out_re, out_im, dependencies); } diff --git a/src/dft/backends/backend_wrappers.cxx b/src/dft/backends/backend_wrappers.cxx index 5d0d2bddc..d265bc91c 100644 --- a/src/dft/backends/backend_wrappers.cxx +++ b/src/dft/backends/backend_wrappers.cxx @@ -37,10 +37,10 @@ and backend_forward_instantiations.cxx. */ // clang-format off -oneapi::mkl::dft::BACKEND::create_commit, -oneapi::mkl::dft::BACKEND::create_commit, -oneapi::mkl::dft::BACKEND::create_commit, -oneapi::mkl::dft::BACKEND::create_commit, +oneapi::math::dft::BACKEND::create_commit, +oneapi::math::dft::BACKEND::create_commit, +oneapi::math::dft::BACKEND::create_commit, +oneapi::math::dft::BACKEND::create_commit, // clang-format on #undef ONEAPI_MKL_DFT_BACKEND_SIGNATURES diff --git a/src/dft/backends/cufft/backward.cpp b/src/dft/backends/cufft/backward.cpp index d70c49611..4dfff0c8b 100644 --- a/src/dft/backends/cufft/backward.cpp +++ b/src/dft/backends/cufft/backward.cpp @@ -34,7 +34,7 @@ #include -namespace oneapi::mkl::dft::cufft { +namespace oneapi::math::dft::cufft { namespace detail { //forward declaration template @@ -62,7 +62,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, if constexpr (std::is_floating_point_v>) { offsets[0] *= 2; // offset is supplied in complex but we offset scalar pointer if (offsets[1] % 2 != 0) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "cuFFT requires offset (first value in strides) to be multiple of 2!"); } @@ -88,7 +88,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, template ONEMATH_EXPORT void compute_backward(descriptor_type &, sycl::buffer, 1> &, sycl::buffer, 1> &) { - throw oneapi::mkl::unimplemented("DFT", "compute_backward(desc, inout_re, inout_im)", + throw oneapi::math::unimplemented("DFT", "compute_backward(desc, inout_re, inout_im)", "cuFFT does not support real-real complex storage."); } @@ -107,7 +107,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, if constexpr (std::is_floating_point_v>) { if (offsets[1] % 2 != 0) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "cuFFT requires offset (first value in strides) to be multiple of 2!"); } @@ -141,7 +141,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &, sycl::buffer, 1> &, sycl::buffer, 1> &, sycl::buffer, 1> &) { - throw oneapi::mkl::unimplemented("DFT", "compute_backward(desc, in_re, in_im, out_re, out_im)", + throw oneapi::math::unimplemented("DFT", "compute_backward(desc, in_re, in_im, out_re, out_im)", "cuFFT does not support real-real complex storage."); } @@ -162,7 +162,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd>) { offsets[0] *= 2; // offset is supplied in complex but we offset scalar pointer if (offsets[1] % 2 != 0) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "cuFFT requires offset (first value in strides) to be multiple of 2!"); } @@ -188,7 +188,7 @@ template ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &, scalar *, scalar *, const std::vector &) { - throw oneapi::mkl::unimplemented("DFT", + throw oneapi::math::unimplemented("DFT", "compute_backward(desc, inout_re, inout_im, dependencies)", "cuFFT does not support real-real complex storage."); } @@ -208,7 +208,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd>) { if (offsets[1] % 2 != 0) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "cuFFT requires offset (first value in strides) to be multiple of 2!"); } @@ -235,7 +235,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &, scalar *, scalar *, scalar *, const std::vector &) { - throw oneapi::mkl::unimplemented("DFT", + throw oneapi::math::unimplemented("DFT", "compute_backward(desc, in_re, in_im, out_re, out_im, deps)", "cuFFT does not support real-real complex storage."); } @@ -243,4 +243,4 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &, scalar #include -namespace oneapi::mkl::dft::cufft { +namespace oneapi::math::dft::cufft { namespace detail { /// Commit impl class specialization for cuFFT. @@ -56,11 +56,11 @@ class cufft_commit final : public dft::detail::commit_impl { public: cufft_commit(sycl::queue& queue, const dft::detail::dft_values& config_values) - : oneapi::mkl::dft::detail::commit_impl(queue, backend::cufft, + : oneapi::math::dft::detail::commit_impl(queue, backend::cufft, config_values) { if constexpr (prec == dft::detail::precision::DOUBLE) { if (!queue.get_device().has(sycl::aspect::fp64)) { - throw mkl::exception("DFT", "commit", "Device does not support double precision."); + throw math::exception("DFT", "commit", "Device does not support double precision."); } } } @@ -69,14 +69,14 @@ class cufft_commit final : public dft::detail::commit_impl { auto fix_context = plans[0].has_value() || plans[1].has_value(); if (plans[0]) { if (cufftDestroy(plans[0].value()) != CUFFT_SUCCESS) { - throw mkl::exception("dft/backends/cufft", __FUNCTION__, + throw math::exception("dft/backends/cufft", __FUNCTION__, "Failed to destroy forward cuFFT plan."); } plans[0] = std::nullopt; } if (plans[1]) { if (cufftDestroy(plans[1].value()) != CUFFT_SUCCESS) { - throw mkl::exception("dft/backends/cufft", __FUNCTION__, + throw math::exception("dft/backends/cufft", __FUNCTION__, "Failed to destroy backward cuFFT plan."); } plans[1] = std::nullopt; @@ -87,7 +87,7 @@ class cufft_commit final : public dft::detail::commit_impl { sycl::get_native(this->get_queue().get_device()); CUcontext interopContext; if (cuDevicePrimaryCtxRetain(&interopContext, interopDevice) != CUDA_SUCCESS) { - throw mkl::exception("dft/backends/cufft", __FUNCTION__, + throw math::exception("dft/backends/cufft", __FUNCTION__, "Failed to change cuda context."); } } @@ -96,13 +96,13 @@ class cufft_commit final : public dft::detail::commit_impl { void commit(const dft::detail::dft_values& config_values) override { // this could be a recommit this->external_workspace_helper_ = - oneapi::mkl::dft::detail::external_workspace_helper( + oneapi::math::dft::detail::external_workspace_helper( config_values.workspace_placement == - oneapi::mkl::dft::detail::config_value::WORKSPACE_EXTERNAL); + oneapi::math::dft::detail::config_value::WORKSPACE_EXTERNAL); clean_plans(); if (config_values.fwd_scale != 1.0 || config_values.bwd_scale != 1.0) { - throw mkl::unimplemented( + throw math::unimplemented( "dft/backends/cufft", __FUNCTION__, "cuFFT does not support values other than 1 for FORWARD/BACKWARD_SCALE"); } @@ -164,14 +164,14 @@ class cufft_commit final : public dft::detail::commit_impl { if constexpr (dom == dft::domain::REAL) { if ((a_min != stride_vecs.vec_a.begin() + rank) || (b_min != stride_vecs.vec_b.begin() + rank)) { - throw mkl::unimplemented( + throw math::unimplemented( "dft/backends/cufft", __FUNCTION__, "cufft requires the last stride to be the the smallest one for real transforms!"); } } else { if (a_min - stride_vecs.vec_a.begin() != b_min - stride_vecs.vec_b.begin()) { - throw mkl::unimplemented( + throw math::unimplemented( "dft/backends/cufft", __FUNCTION__, "cufft requires that if ordered by stride length, the order of strides is the same for input/output or fwd/bwd strides!"); } @@ -192,7 +192,7 @@ class cufft_commit final : public dft::detail::commit_impl { } for (int i = 1; i < rank; i++) { if ((stride_vecs.vec_a[i] % a_stride != 0) || (stride_vecs.vec_b[i] % b_stride != 0)) { - throw mkl::unimplemented( + throw math::unimplemented( "dft/backends/cufft", __FUNCTION__, "cufft requires a stride to be divisible by all smaller strides!"); } @@ -202,7 +202,7 @@ class cufft_commit final : public dft::detail::commit_impl { if (rank > 2) { if (stride_vecs.vec_a[1] > stride_vecs.vec_a[2] && stride_vecs.vec_b[1] < stride_vecs.vec_b[2]) { - throw mkl::unimplemented( + throw math::unimplemented( "dft/backends/cufft", __FUNCTION__, "cufft requires that if ordered by stride length, the order of strides is the same for input and output strides!"); } @@ -215,7 +215,7 @@ class cufft_commit final : public dft::detail::commit_impl { } if ((stride_vecs.vec_a[1] % stride_vecs.vec_a[2] != 0) || (stride_vecs.vec_b[1] % stride_vecs.vec_b[2] != 0)) { - throw mkl::unimplemented( + throw math::unimplemented( "dft/backends/cufft", __FUNCTION__, "cufft requires a stride to be divisible by all smaller strides!"); } @@ -252,14 +252,14 @@ class cufft_commit final : public dft::detail::commit_impl { : check_stride_validity(stride_vecs.bwd_out, stride_vecs.bwd_in); if (!valid_forward && !valid_backward) { - throw mkl::exception("dft/backends/cufft", __FUNCTION__, "Invalid strides."); + throw math::exception("dft/backends/cufft", __FUNCTION__, "Invalid strides."); } if (valid_forward) { cufftHandle fwd_plan; auto res = cufftCreate(&fwd_plan); if (res != CUFFT_SUCCESS) { - throw mkl::exception("dft/backends/cufft", __FUNCTION__, "cufftCreate failed."); + throw math::exception("dft/backends/cufft", __FUNCTION__, "cufftCreate failed."); } apply_external_workspace_setting(fwd_plan, config_values.workspace_placement); res = cufftPlanMany(&fwd_plan, // plan @@ -276,7 +276,7 @@ class cufft_commit final : public dft::detail::commit_impl { ); if (res != CUFFT_SUCCESS) { - throw mkl::exception("dft/backends/cufft", __FUNCTION__, + throw math::exception("dft/backends/cufft", __FUNCTION__, "Failed to create forward cuFFT plan."); } @@ -287,7 +287,7 @@ class cufft_commit final : public dft::detail::commit_impl { cufftHandle bwd_plan; auto res = cufftCreate(&bwd_plan); if (res != CUFFT_SUCCESS) { - throw mkl::exception("dft/backends/cufft", __FUNCTION__, "cufftCreate failed."); + throw math::exception("dft/backends/cufft", __FUNCTION__, "cufftCreate failed."); } apply_external_workspace_setting(bwd_plan, config_values.workspace_placement); // flip fwd_distance and bwd_distance because cuFFt uses input distance and output distance. @@ -304,7 +304,7 @@ class cufft_commit final : public dft::detail::commit_impl { batch // batch ); if (res != CUFFT_SUCCESS) { - throw mkl::exception("dft/backends/cufft", __FUNCTION__, + throw math::exception("dft/backends/cufft", __FUNCTION__, "Failed to create backward cuFFT plan."); } plans[1] = bwd_plan; @@ -320,7 +320,7 @@ class cufft_commit final : public dft::detail::commit_impl { if (workspace_setting == config_value::WORKSPACE_EXTERNAL) { auto res = cufftSetAutoAllocation(handle, 0); if (res != CUFFT_SUCCESS) { - throw mkl::exception("dft/backends/cufft", "commit", + throw math::exception("dft/backends/cufft", "commit", "cufftSetAutoAllocation(plan, 0) failed."); } } @@ -357,7 +357,7 @@ class cufft_commit final : public dft::detail::commit_impl { auto stream = ih.get_native_queue(); auto result = cufftSetStream(plan, stream); if (result != CUFFT_SUCCESS) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/backends/cufft", "set_workspace", "cufftSetStream returned " + std::to_string(result)); } @@ -459,4 +459,4 @@ get_offsets_bwd( dft::detail::commit_impl*); } //namespace detail -} // namespace oneapi::mkl::dft::cufft +} // namespace oneapi::math::dft::cufft diff --git a/src/dft/backends/cufft/descriptor.cpp b/src/dft/backends/cufft/descriptor.cpp index 48e22b7e7..40d96cdee 100644 --- a/src/dft/backends/cufft/descriptor.cpp +++ b/src/dft/backends/cufft/descriptor.cpp @@ -22,7 +22,7 @@ #include "oneapi/math/dft/detail/cufft/onemath_dft_cufft.hpp" -namespace oneapi::mkl::dft::detail { +namespace oneapi::math::dft::detail { template void descriptor::commit(backend_selector selector) { @@ -42,4 +42,4 @@ template void descriptor::commit( backend_selector); template void descriptor::commit(backend_selector); -} //namespace oneapi::mkl::dft::detail +} //namespace oneapi::math::dft::detail diff --git a/src/dft/backends/cufft/execute_helper.hpp b/src/dft/backends/cufft/execute_helper.hpp index 629cf1767..877136bfe 100644 --- a/src/dft/backends/cufft/execute_helper.hpp +++ b/src/dft/backends/cufft/execute_helper.hpp @@ -34,27 +34,27 @@ #include #include -namespace oneapi::mkl::dft::cufft::detail { +namespace oneapi::math::dft::cufft::detail { template inline dft::detail::commit_impl *checked_get_commit( dft::detail::descriptor &desc) { auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::cufft) { - throw mkl::invalid_argument("dft/backends/cufft", "get_commit", + throw math::invalid_argument("dft/backends/cufft", "get_commit", "DFT descriptor has not been commited for cuFFT"); } return commit_handle; } -/// Throw an mkl::invalid_argument if the runtime param in the descriptor does not match +/// Throw an math::invalid_argument if the runtime param in the descriptor does not match /// the expected value. template inline auto expect_config(DescT &desc, const char *message) { dft::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { - throw mkl::invalid_argument("dft/backends/cufft", "expect_config", message); + throw math::invalid_argument("dft/backends/cufft", "expect_config", message); } } @@ -73,7 +73,7 @@ void cufft_execute(const std::string &func, CUstream stream, cufftHandle plan, v auto result = cufftExecR2C(plan, reinterpret_cast(input), reinterpret_cast(output)); if (result != CUFFT_SUCCESS) { - throw oneapi::mkl::exception("dft/backends/cufft", func, + throw oneapi::math::exception("dft/backends/cufft", func, "cufftExecR2C returned " + std::to_string(result)); } } @@ -81,7 +81,7 @@ void cufft_execute(const std::string &func, CUstream stream, cufftHandle plan, v auto result = cufftExecD2Z(plan, reinterpret_cast(input), reinterpret_cast(output)); if (result != CUFFT_SUCCESS) { - throw oneapi::mkl::exception("dft/backends/cufft", func, + throw oneapi::math::exception("dft/backends/cufft", func, "cufftExecD2Z returned " + std::to_string(result)); } } @@ -91,7 +91,7 @@ void cufft_execute(const std::string &func, CUstream stream, cufftHandle plan, v auto result = cufftExecC2R(plan, reinterpret_cast(input), reinterpret_cast(output)); if (result != CUFFT_SUCCESS) { - throw oneapi::mkl::exception("dft/backends/cufft", func, + throw oneapi::math::exception("dft/backends/cufft", func, "cufftExecC2R returned " + std::to_string(result)); } } @@ -99,7 +99,7 @@ void cufft_execute(const std::string &func, CUstream stream, cufftHandle plan, v auto result = cufftExecZ2D(plan, reinterpret_cast(input), reinterpret_cast(output)); if (result != CUFFT_SUCCESS) { - throw oneapi::mkl::exception("dft/backends/cufft", func, + throw oneapi::math::exception("dft/backends/cufft", func, "cufftExecZ2D returned " + std::to_string(result)); } } @@ -111,7 +111,7 @@ void cufft_execute(const std::string &func, CUstream stream, cufftHandle plan, v cufftExecC2C(plan, reinterpret_cast(input), reinterpret_cast(output), static_cast(dir)); if (result != CUFFT_SUCCESS) { - throw oneapi::mkl::exception("dft/backends/cufft", func, + throw oneapi::math::exception("dft/backends/cufft", func, "cufftExecC2C returned " + std::to_string(result)); } } @@ -120,7 +120,7 @@ void cufft_execute(const std::string &func, CUstream stream, cufftHandle plan, v cufftExecZ2Z(plan, reinterpret_cast(input), reinterpret_cast(output), static_cast(dir)); if (result != CUFFT_SUCCESS) { - throw oneapi::mkl::exception("dft/backends/cufft", func, + throw oneapi::math::exception("dft/backends/cufft", func, "cufftExecZ2Z returned " + std::to_string(result)); } } @@ -131,7 +131,7 @@ void cufft_execute(const std::string &func, CUstream stream, cufftHandle plan, v // as complete early. auto result = cuStreamSynchronize(stream); if (result != CUDA_SUCCESS) { - throw oneapi::mkl::exception("dft/backends/cufft", func, + throw oneapi::math::exception("dft/backends/cufft", func, "cuStreamSynchronize returned " + std::to_string(result)); } #endif @@ -141,12 +141,12 @@ inline CUstream setup_stream(const std::string &func, sycl::interop_handle ih, c auto stream = ih.get_native_queue(); auto result = cufftSetStream(plan, stream); if (result != CUFFT_SUCCESS) { - throw oneapi::mkl::exception("dft/backends/cufft", func, + throw oneapi::math::exception("dft/backends/cufft", func, "cufftSetStream returned " + std::to_string(result)); } return stream; } -} // namespace oneapi::mkl::dft::cufft::detail +} // namespace oneapi::math::dft::cufft::detail #endif // _ONEMATH_DFT_SRC_EXECUTE_HELPER_CUFFT_HPP_ diff --git a/src/dft/backends/cufft/forward.cpp b/src/dft/backends/cufft/forward.cpp index 757c2b97d..ca7b2bc62 100644 --- a/src/dft/backends/cufft/forward.cpp +++ b/src/dft/backends/cufft/forward.cpp @@ -35,7 +35,7 @@ #include -namespace oneapi::mkl::dft::cufft { +namespace oneapi::math::dft::cufft { namespace detail { //forward declaration @@ -64,7 +64,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, if constexpr (std::is_floating_point_v>) { if (offsets[0] % 2 != 0) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "cuFFT requires offset (first value in strides) to be multiple of 2!"); } @@ -91,7 +91,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, template ONEMATH_EXPORT void compute_forward(descriptor_type &, sycl::buffer, 1> &, sycl::buffer, 1> &) { - throw oneapi::mkl::unimplemented("DFT", "compute_forward(desc, inout_re, inout_im)", + throw oneapi::math::unimplemented("DFT", "compute_forward(desc, inout_re, inout_im)", "cuFFT does not support real-real complex storage."); } @@ -109,7 +109,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer>) { if (offsets[0] % 2 != 0) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "cuFFT requires offset (first value in strides) to be multiple of 2!"); } @@ -143,7 +143,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &, sycl::buffer, 1> &, sycl::buffer, 1> &, sycl::buffer, 1> &) { - throw oneapi::mkl::unimplemented("DFT", "compute_forward(desc, in_re, in_im, out_re, out_im)", + throw oneapi::math::unimplemented("DFT", "compute_forward(desc, in_re, in_im, out_re, out_im)", "cuFFT does not support real-real complex storage."); } @@ -163,7 +163,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd>) { if (offsets[0] % 2 != 0) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "cuFFT requires offset (first value in strides) to be multiple of 2!"); } @@ -190,7 +190,7 @@ template ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &, scalar *, scalar *, const std::vector &) { - throw oneapi::mkl::unimplemented("DFT", + throw oneapi::math::unimplemented("DFT", "compute_forward(desc, inout_re, inout_im, dependencies)", "cuFFT does not support real-real complex storage."); } @@ -210,7 +210,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd>) { if (offsets[0] % 2 != 0) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "cuFFT requires offset (first value in strides) to be multiple of 2!"); } @@ -237,7 +237,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &, scalar *, scalar *, scalar *, const std::vector &) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", "compute_forward(desc, in_re, in_im, out_re, out_im, dependencies)", "cuFFT does not support real-real complex storage."); } @@ -245,4 +245,4 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &, scalar void descriptor::commit(sycl::queue &queue) { @@ -39,4 +39,4 @@ template void descriptor::commit(sycl::queue &) template void descriptor::commit(sycl::queue &); template void descriptor::commit(sycl::queue &); -} //namespace oneapi::mkl::dft::detail +} //namespace oneapi::math::dft::detail diff --git a/src/dft/backends/mklcpu/backward.cpp b/src/dft/backends/mklcpu/backward.cpp index 0cf42948e..063648167 100644 --- a/src/dft/backends/mklcpu/backward.cpp +++ b/src/dft/backends/mklcpu/backward.cpp @@ -34,7 +34,7 @@ // MKLCPU header #include "mkl_dfti.h" -namespace oneapi::mkl::dft::mklcpu { +namespace oneapi::math::dft::mklcpu { namespace detail { // BUFFER version @@ -43,7 +43,7 @@ template inline void check_bwd_commit(dft::descriptor &desc) { auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::mklcpu) { - throw mkl::invalid_argument("DFT", "computer_backward", + throw math::invalid_argument("DFT", "computer_backward", "DFT descriptor has not been commited for MKLCPU"); } @@ -51,19 +51,19 @@ inline void check_bwd_commit(dft::descriptor &desc) { MKL_LONG commit_status{ DFTI_UNCOMMITTED }; DftiGetValue(mklcpu_desc[1], DFTI_COMMIT_STATUS, &commit_status); if (commit_status != DFTI_COMMITTED) { - throw mkl::invalid_argument("DFT", "compute_backward", + throw math::invalid_argument("DFT", "compute_backward", "MKLCPU DFT descriptor was not successfully committed."); } } -// Throw an mkl::invalid_argument if the runtime param in the descriptor does not match +// Throw an math::invalid_argument if the runtime param in the descriptor does not match // the expected value. template inline auto expect_config(DescT &desc, const char *message) { dft::detail::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { - throw mkl::invalid_argument("DFT", "compute_backward", message); + throw math::invalid_argument("DFT", "compute_backward", message); } } // convert the base commit class to derived cpu commit class @@ -94,7 +94,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, DFT_ERROR status = DftiComputeBackward(desc_acc[detail::DIR::bwd], detail::acc_to_ptr(inout_acc)); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/backends/mklcpu", "compute_backward", std::string("DftiComputeBackward failed : ") + DftiErrorMessage(status)); } @@ -126,7 +126,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, DFT_ERROR status = DftiComputeBackward( desc_acc[detail::DIR::bwd], detail::acc_to_ptr(re_acc), detail::acc_to_ptr(im_acc)); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/backends/mklcpu", "compute_backward", std::string("DftiComputeBackward failed : ") + DftiErrorMessage(status)); } @@ -159,7 +159,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, DFT_ERROR status = DftiComputeBackward(desc_acc[detail::DIR::bwd], in_ptr, detail::acc_to_ptr(out_acc)); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/backends/mklcpu", "compute_backward", std::string("DftiComputeBackward failed : ") + DftiErrorMessage(status)); } @@ -198,7 +198,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, DftiComputeBackward(desc_acc[detail::DIR::bwd], inre_ptr, inim_ptr, detail::acc_to_ptr(outre_acc), detail::acc_to_ptr(outim_acc)); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/backends/mklcpu", "compute_backward", std::string("DftiComputeBackward failed : ") + DftiErrorMessage(status)); } @@ -227,7 +227,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd(cgh, [=]() { DFT_ERROR status = DftiComputeBackward(desc_acc[detail::DIR::bwd], inout); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/backends/mklcpu", "compute_backward", std::string("DftiComputeBackward failed : ") + DftiErrorMessage(status)); } @@ -255,7 +255,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar(cgh, [=]() { DFT_ERROR status = DftiComputeBackward(desc_acc[detail::DIR::bwd], inout_re, inout_im); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/backends/mklcpu", "compute_backward", std::string("DftiComputeBackward failed : ") + DftiErrorMessage(status)); } @@ -285,7 +285,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd(cgh, [=]() { DFT_ERROR status = DftiComputeBackward(desc_acc[detail::DIR::bwd], in, out); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/backends/mklcpu", "compute_backward", std::string("DftiComputeBackward failed : ") + DftiErrorMessage(status)); } @@ -316,7 +316,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar commit_derived_impl::commit_derived_impl( sycl::queue queue, const dft::detail::dft_values& config_values) - : oneapi::mkl::dft::detail::commit_impl(queue, backend::mklcpu, config_values) { + : oneapi::math::dft::detail::commit_impl(queue, backend::mklcpu, config_values) { // create the descriptor once for the lifetime of the descriptor class DFT_ERROR status[2] = { DFTI_BAD_DESCRIPTOR, DFTI_BAD_DESCRIPTOR }; @@ -66,7 +66,7 @@ commit_derived_impl::commit_derived_impl( std::string err = std::string("DftiCreateDescriptor failed with status : ") + DftiErrorMessage(status[0]) + std::string(", ") + DftiErrorMessage(status[1]); - throw oneapi::mkl::exception("dft/backends/mklcpu", "create_descriptor", err); + throw oneapi::math::exception("dft/backends/mklcpu", "create_descriptor", err); } } @@ -81,9 +81,9 @@ template void commit_derived_impl::commit( const dft::detail::dft_values& config_values) { this->external_workspace_helper_ = - oneapi::mkl::dft::detail::external_workspace_helper( + oneapi::math::dft::detail::external_workspace_helper( config_values.workspace_placement == - oneapi::mkl::dft::detail::config_value::WORKSPACE_EXTERNAL); + oneapi::math::dft::detail::config_value::WORKSPACE_EXTERNAL); set_value(bidirection_handle.data(), config_values); this->get_queue() @@ -104,7 +104,7 @@ void commit_derived_impl::commit( std::string err = std::string("DftiCommitDescriptor failed with status : ") + DftiErrorMessage(status[0]) + std::string(", ") + DftiErrorMessage(status[1]); - throw oneapi::mkl::exception("dft/backends/mklcpu", "commit", err); + throw oneapi::math::exception("dft/backends/mklcpu", "commit", err); } }); }) @@ -122,7 +122,7 @@ void commit_derived_impl::set_value_item(mklcpu_desc_t hand, enum DFT Args... args) { DFT_ERROR value_err = DftiSetValue(hand, name, args...); if (value_err != DFTI_NO_ERROR) { - throw oneapi::mkl::exception("dft/backends/mklcpu", "set_value_item", + throw oneapi::math::exception("dft/backends/mklcpu", "set_value_item", DftiErrorMessage(value_err)); } } @@ -166,17 +166,17 @@ void commit_derived_impl::set_value(mklcpu_desc_t* descHandle, to_mklcpu(config.packed_format)); // Setting the workspace causes an FFT_INVALID_DESCRIPTOR. if (config.workspace != config_value::ALLOW) { - throw mkl::invalid_argument("dft/backends/mklcpu", "commit", + throw math::invalid_argument("dft/backends/mklcpu", "commit", "MKLCPU only supports workspace set to allow"); } // Setting the ordering causes an FFT_INVALID_DESCRIPTOR. Check that default is used: if (config.ordering != dft::detail::config_value::ORDERED) { - throw mkl::invalid_argument("dft/backends/mklcpu", "commit", + throw math::invalid_argument("dft/backends/mklcpu", "commit", "MKLCPU only supports ordered ordering."); } // Setting the transpose causes an FFT_INVALID_DESCRIPTOR. Check that default is used: if (config.transpose != false) { - throw mkl::invalid_argument("dft/backends/mklcpu", "commit", + throw math::invalid_argument("dft/backends/mklcpu", "commit", "MKLCPU only supports non-transposed."); } } @@ -208,5 +208,5 @@ create_commit( } // namespace mklcpu } // namespace dft -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/dft/backends/mklcpu/commit_derived_impl.hpp b/src/dft/backends/mklcpu/commit_derived_impl.hpp index a2e3320c5..ee4a4e53f 100644 --- a/src/dft/backends/mklcpu/commit_derived_impl.hpp +++ b/src/dft/backends/mklcpu/commit_derived_impl.hpp @@ -28,7 +28,7 @@ #include "mkl_dfti.h" namespace oneapi { -namespace mkl { +namespace math { namespace dft { namespace mklcpu { namespace detail { @@ -82,7 +82,7 @@ using commit_derived_t = detail::commit_derived_impl; } // namespace detail } // namespace mklcpu } // namespace dft -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ONEMATH_DFT_COMMIT_DERIVED_IMPL_HPP_ diff --git a/src/dft/backends/mklcpu/descriptor.cpp b/src/dft/backends/mklcpu/descriptor.cpp index 9a6acd7cf..a89203717 100644 --- a/src/dft/backends/mklcpu/descriptor.cpp +++ b/src/dft/backends/mklcpu/descriptor.cpp @@ -22,7 +22,7 @@ #include "oneapi/math/dft/detail/mklcpu/onemath_dft_mklcpu.hpp" -namespace oneapi::mkl::dft::detail { +namespace oneapi::math::dft::detail { template void descriptor::commit(backend_selector selector) { @@ -44,4 +44,4 @@ template void descriptor::commit( template void descriptor::commit( backend_selector); -} //namespace oneapi::mkl::dft::detail +} //namespace oneapi::math::dft::detail diff --git a/src/dft/backends/mklcpu/forward.cpp b/src/dft/backends/mklcpu/forward.cpp index 7152f3322..f915c9ce5 100644 --- a/src/dft/backends/mklcpu/forward.cpp +++ b/src/dft/backends/mklcpu/forward.cpp @@ -34,7 +34,7 @@ // MKLCPU header #include "mkl_dfti.h" -namespace oneapi::mkl::dft::mklcpu { +namespace oneapi::math::dft::mklcpu { namespace detail { // BUFFER version @@ -43,7 +43,7 @@ template inline void check_fwd_commit(dft::descriptor &desc) { auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::mklcpu) { - throw mkl::invalid_argument("DFT", "computer_forward", + throw math::invalid_argument("DFT", "computer_forward", "DFT descriptor has not been commited for MKLCPU"); } @@ -51,19 +51,19 @@ inline void check_fwd_commit(dft::descriptor &desc) { MKL_LONG commit_status{ DFTI_UNCOMMITTED }; DftiGetValue(mklcpu_desc[0], DFTI_COMMIT_STATUS, &commit_status); if (commit_status != DFTI_COMMITTED) { - throw mkl::invalid_argument("DFT", "compute_forward", + throw math::invalid_argument("DFT", "compute_forward", "MKLCPU DFT descriptor was not successfully committed."); } } -// Throw an mkl::invalid_argument if the runtime param in the descriptor does not match +// Throw an math::invalid_argument if the runtime param in the descriptor does not match // the expected value. template inline auto expect_config(DescT &desc, const char *message) { dft::detail::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { - throw mkl::invalid_argument("DFT", "compute_forward", message); + throw math::invalid_argument("DFT", "compute_forward", message); } } @@ -95,7 +95,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, DFT_ERROR status = DftiComputeForward(desc_acc[detail::DIR::fwd], detail::acc_to_ptr(inout_acc)); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/forward/mklcpu", "compute_forward", std::string("DftiComputeForward failed : ") + DftiErrorMessage(status)); } @@ -127,7 +127,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, DFT_ERROR status = DftiComputeForward( desc_acc[detail::DIR::fwd], detail::acc_to_ptr(re_acc), detail::acc_to_ptr(im_acc)); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/forward/mklcpu", "compute_forward", std::string("DftiComputeForward failed : ") + DftiErrorMessage(status)); } @@ -159,7 +159,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer(cgh, [=]() { DFT_ERROR status = DftiComputeForward(desc_acc[detail::DIR::fwd], inout); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/forward/mklcpu", "compute_forward", std::string("DftiComputeForward failed : ") + DftiErrorMessage(status)); } @@ -258,7 +258,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar(cgh, [=]() { DFT_ERROR status = DftiComputeForward(desc_acc[detail::DIR::fwd], inout_re, inout_im); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/forward/mklcpu", "compute_forward", std::string("DftiComputeForward failed : ") + DftiErrorMessage(status)); } @@ -289,7 +289,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd(cgh, [=]() { DFT_ERROR status = DftiComputeForward(desc_acc[detail::DIR::fwd], in, out); if (status != DFTI_NO_ERROR) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/forward/mklcpu", "compute_forward", std::string("DftiComputeForward failed : ") + DftiErrorMessage(status)); } @@ -322,7 +322,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar static inline auto host_task_internal(H& cgh, F f, int) -> decltype(cgh.host_task(f)) { @@ -82,7 +82,7 @@ inline constexpr DFTI_CONFIG_PARAM to_mklcpu(dft::detail::config_param param) { case iparam::PACKED_FORMAT: return DFTI_PACKED_FORMAT; case iparam::COMMIT_STATUS: return DFTI_COMMIT_STATUS; default: - throw mkl::invalid_argument("dft", "MKLCPU descriptor set_value()", + throw math::invalid_argument("dft", "MKLCPU descriptor set_value()", "Invalid config param."); return static_cast(0); } @@ -105,7 +105,7 @@ inline constexpr int to_mklcpu( return DFTI_REAL_REAL; } else { - throw mkl::invalid_argument("dft", "MKLCPU descriptor set_value()", + throw math::invalid_argument("dft", "MKLCPU descriptor set_value()", "Invalid config value for complex storage."); return 0; } @@ -118,7 +118,7 @@ inline constexpr int to_mklcpu( return DFTI_REAL_REAL; } else { - throw mkl::invalid_argument("dft", "MKLCPU descriptor set_value()", + throw math::invalid_argument("dft", "MKLCPU descriptor set_value()", "Invalid config value for real storage."); return 0; } @@ -130,7 +130,7 @@ inline constexpr int to_mklcpu( return DFTI_NOT_INPLACE; } else { - throw mkl::invalid_argument("dft", "MKLCPU descriptor set_value()", + throw math::invalid_argument("dft", "MKLCPU descriptor set_value()", "Invalid config value for inplace."); return 0; } @@ -159,7 +159,7 @@ inline constexpr int to_mklcpu( return DFTI_CCE_FORMAT; } else { - throw mkl::invalid_argument("dft", "MKLCPU descriptor set_value()", + throw math::invalid_argument("dft", "MKLCPU descriptor set_value()", "Invalid config value for packed format."); return 0; } @@ -173,6 +173,6 @@ typename AccType::value_type* acc_to_ptr(AccType acc) { return acc.template get_multi_ptr().get(); } -} // namespace oneapi::mkl::dft::mklcpu::detail +} // namespace oneapi::math::dft::mklcpu::detail #endif // _ONEMATH_DFT_SRC_MKLCPU_HELPERS_HPP_ diff --git a/src/dft/backends/mklgpu/backward.cpp b/src/dft/backends/mklgpu/backward.cpp index 17afddeaf..ec0e4ca08 100644 --- a/src/dft/backends/mklgpu/backward.cpp +++ b/src/dft/backends/mklgpu/backward.cpp @@ -38,7 +38,7 @@ #include #endif -namespace oneapi::mkl::dft::mklgpu { +namespace oneapi::math::dft::mklgpu { namespace detail { /// Forward a MKLGPU DFT call to the backend, checking that the commit impl is valid. @@ -50,7 +50,7 @@ inline auto compute_backward(dft::detail::descriptor &desc, ArgTs &&. using handle_t = std::pair; auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::mklgpu) { - throw mkl::invalid_argument("DFT", "compute_backward", + throw math::invalid_argument("DFT", "compute_backward", "DFT descriptor has not been commited for MKLGPU"); } auto handle = reinterpret_cast(commit_handle->get_handle()); @@ -58,7 +58,7 @@ inline auto compute_backward(dft::detail::descriptor &desc, ArgTs &&. int commit_status{ DFTI_UNCOMMITTED }; mklgpu_desc->get_value(dft::config_param::COMMIT_STATUS, &commit_status); if (commit_status != DFTI_COMMITTED) { - throw mkl::invalid_argument("DFT", "compute_backward", + throw math::invalid_argument("DFT", "compute_backward", "MKLGPU DFT descriptor was not successfully committed."); } // The MKLGPU backend's interface contains fewer function signatures than in this @@ -67,14 +67,14 @@ inline auto compute_backward(dft::detail::descriptor &desc, ArgTs &&. return dft::compute_backward(*mklgpu_desc, std::forward(args)...); } -/// Throw an mkl::invalid_argument if the runtime param in the descriptor does not match +/// Throw an math::invalid_argument if the runtime param in the descriptor does not match /// the expected value. template inline auto expect_config(DescT &desc, const char *message) { dft::detail::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { - throw mkl::invalid_argument("DFT", "compute_backward", message); + throw math::invalid_argument("DFT", "compute_backward", message); } } } // namespace detail @@ -95,7 +95,7 @@ template ONEMATH_EXPORT void compute_backward(descriptor_type & /*desc*/, sycl::buffer, 1> & /*inout_re*/, sycl::buffer, 1> & /*inout_im*/) { - throw mkl::unimplemented("DFT", "compute_backward", + throw math::unimplemented("DFT", "compute_backward", "MKLGPU does not support compute_backward(desc, inout_re, inout_im)."); } @@ -120,7 +120,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, detail::expect_config( desc, "Unexpected value for complex storage"); - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", "compute_backward(desc, in_re, in_im, out_re, out_im)", "MKLGPU does not support out-of-place FFT with real-real complex storage."); } @@ -142,7 +142,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type & /*desc*/, scalar * /*inout_re*/, scalar * /*inout_im*/, const std::vector & /*dependencies*/) { - throw mkl::unimplemented( + throw math::unimplemented( "DFT", "compute_backward", "MKLGPU does not support compute_backward(desc, inout_re, inout_im, dependencies)."); } @@ -169,7 +169,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, detail::expect_config( desc, "Unexpected value for complex storage"); - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", "compute_backward(desc, in_re, in_im, out_re, out_im, deps)", "MKLGPU does not support out-of-place FFT with real-real complex storage."); } @@ -177,4 +177,4 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, // Template function instantiations #include "dft/backends/backend_backward_instantiations.cxx" -} // namespace oneapi::mkl::dft::mklgpu +} // namespace oneapi::math::dft::mklgpu diff --git a/src/dft/backends/mklgpu/commit.cpp b/src/dft/backends/mklgpu/commit.cpp index 87efa22a2..4463eafe9 100644 --- a/src/dft/backends/mklgpu/commit.cpp +++ b/src/dft/backends/mklgpu/commit.cpp @@ -55,7 +55,7 @@ interface of this oneMath library. Consequently, the types under dft::TYPE are Intel oneMKL types, and types under dft::detail::TYPE are from this library. **/ -namespace oneapi::mkl::dft::mklgpu { +namespace oneapi::math::dft::mklgpu { namespace detail { /// Commit impl class specialization for MKLGPU. @@ -77,14 +77,14 @@ class mklgpu_commit final : public dft::detail::commit_impl { public: mklgpu_commit(sycl::queue queue, const dft::detail::dft_values& config_values) - : oneapi::mkl::dft::detail::commit_impl(queue, backend::mklgpu, + : oneapi::math::dft::detail::commit_impl(queue, backend::mklgpu, config_values), handle(std::make_shared(config_values.dimensions), nullptr) { handle.second = handle.first; // Make sure the bwd pointer is valid. // MKLGPU does not throw an informative exception for the following: if constexpr (prec == dft::detail::precision::DOUBLE) { if (!queue.get_device().has(sycl::aspect::fp64)) { - throw mkl::exception("dft/backends/mklgpu", "commit", + throw math::exception("dft/backends/mklgpu", "commit", "Device does not support double precision."); } } @@ -92,9 +92,9 @@ class mklgpu_commit final : public dft::detail::commit_impl { virtual void commit(const dft::detail::dft_values& config_values) override { this->external_workspace_helper_ = - oneapi::mkl::dft::detail::external_workspace_helper( + oneapi::math::dft::detail::external_workspace_helper( config_values.workspace_placement == - oneapi::mkl::dft::detail::config_value::WORKSPACE_EXTERNAL); + oneapi::math::dft::detail::config_value::WORKSPACE_EXTERNAL); auto stride_choice = dft::detail::get_stride_api(config_values); throw_on_invalid_stride_api("MKLGPU commit", stride_choice); @@ -115,7 +115,7 @@ class mklgpu_commit final : public dft::detail::commit_impl { // Catching the real Intel oneMKL exception causes headaches with naming forward_good = false; if (one_descriptor) { - throw mkl::exception("dft/backends/mklgpu" + throw math::exception("dft/backends/mklgpu" "commit", mkl_exception.what()); } @@ -131,7 +131,7 @@ class mklgpu_commit final : public dft::detail::commit_impl { catch (const std::exception& mkl_exception) { // Catching the real Intel oneMKL exception causes headaches with naming. if (!forward_good) { - throw mkl::exception("dft/backends/mklgpu" + throw math::exception("dft/backends/mklgpu" "commit", mkl_exception.what()); } @@ -183,7 +183,7 @@ class mklgpu_commit final : public dft::detail::commit_impl { desc.set_value(backend_param::COMPLEX_STORAGE, to_mklgpu(config.complex_storage)); if (config.real_storage != dft::detail::config_value::REAL_REAL) { - throw mkl::invalid_argument("dft/backends/mklgpu", "commit", + throw math::invalid_argument("dft/backends/mklgpu", "commit", "MKLGPU only supports real-real real storage."); } desc.set_value(backend_param::CONJUGATE_EVEN_STORAGE, @@ -193,7 +193,7 @@ class mklgpu_commit final : public dft::detail::commit_impl { if (stride_choice == dft::detail::stride_api::FB_STRIDES) { if (config.fwd_strides[0] != 0 || config.fwd_strides[0] != 0) { - throw mkl::unimplemented("dft/backends/mklgpu", "commit", + throw math::unimplemented("dft/backends/mklgpu", "commit", "MKLGPU does not support nonzero offsets."); } desc.set_value(backend_param::FWD_STRIDES, config.fwd_strides.data()); @@ -201,7 +201,7 @@ class mklgpu_commit final : public dft::detail::commit_impl { } else { if (config.input_strides[0] != 0 || config.output_strides[0] != 0) { - throw mkl::unimplemented("dft/backends/mklgpu", "commit", + throw math::unimplemented("dft/backends/mklgpu", "commit", "MKLGPU does not support nonzero offsets."); } if (assume_fwd_dft) { @@ -223,12 +223,12 @@ class mklgpu_commit final : public dft::detail::commit_impl { } // Setting the ordering causes an FFT_INVALID_DESCRIPTOR. Check that default is used: if (config.ordering != dft::detail::config_value::ORDERED) { - throw mkl::invalid_argument("dft/backends/mklgpu", "commit", + throw math::invalid_argument("dft/backends/mklgpu", "commit", "MKLGPU only supports ordered ordering."); } // Setting the transpose causes an FFT_INVALID_DESCRIPTOR. Check that default is used: if (config.transpose != false) { - throw mkl::invalid_argument("dft/backends/mklgpu", "commit", + throw math::invalid_argument("dft/backends/mklgpu", "commit", "MKLGPU only supports non-transposed."); } } @@ -266,4 +266,4 @@ create_commit( const dft::detail::descriptor&, sycl::queue&); -} // namespace oneapi::mkl::dft::mklgpu +} // namespace oneapi::math::dft::mklgpu diff --git a/src/dft/backends/mklgpu/descriptor.cpp b/src/dft/backends/mklgpu/descriptor.cpp index ca25a13ec..8a0fee21b 100644 --- a/src/dft/backends/mklgpu/descriptor.cpp +++ b/src/dft/backends/mklgpu/descriptor.cpp @@ -22,7 +22,7 @@ #include "oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp" -namespace oneapi::mkl::dft::detail { +namespace oneapi::math::dft::detail { template void descriptor::commit(backend_selector selector) { @@ -44,4 +44,4 @@ template void descriptor::commit( template void descriptor::commit( backend_selector); -} //namespace oneapi::mkl::dft::detail +} //namespace oneapi::math::dft::detail diff --git a/src/dft/backends/mklgpu/forward.cpp b/src/dft/backends/mklgpu/forward.cpp index d17183e92..56421e110 100644 --- a/src/dft/backends/mklgpu/forward.cpp +++ b/src/dft/backends/mklgpu/forward.cpp @@ -45,7 +45,7 @@ interface of this oneMath library. Consequently, the types under dft::TYPE are Intel oneMKL types, and types under dft::detail::TYPE are from this library. **/ -namespace oneapi::mkl::dft::mklgpu { +namespace oneapi::math::dft::mklgpu { namespace detail { /// Forward a MKLGPU DFT call to the backend, checking that the commit impl is valid. /// Assumes backend descriptor values match those of the frontend. @@ -56,7 +56,7 @@ inline auto compute_forward(dft::detail::descriptor &desc, ArgTs &&.. using handle_t = std::pair; auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::mklgpu) { - throw mkl::invalid_argument("DFT", "compute_forward", + throw math::invalid_argument("DFT", "compute_forward", "DFT descriptor has not been commited for MKLGPU"); } auto handle = reinterpret_cast(commit_handle->get_handle()); @@ -64,7 +64,7 @@ inline auto compute_forward(dft::detail::descriptor &desc, ArgTs &&.. int commit_status{ DFTI_UNCOMMITTED }; mklgpu_desc->get_value(dft::config_param::COMMIT_STATUS, &commit_status); if (commit_status != DFTI_COMMITTED) { - throw mkl::invalid_argument("DFT", "compute_forward", + throw math::invalid_argument("DFT", "compute_forward", "MKLGPU DFT descriptor was not successfully committed."); } // The MKLGPU backend's interface contains fewer function signatures than in this @@ -73,14 +73,14 @@ inline auto compute_forward(dft::detail::descriptor &desc, ArgTs &&.. return dft::compute_forward(*mklgpu_desc, std::forward(args)...); } -/// Throw an mkl::invalid_argument if the runtime param in the descriptor does not match +/// Throw an math::invalid_argument if the runtime param in the descriptor does not match /// the expected value. template inline auto expect_config(DescT &desc, const char *message) { dft::detail::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { - throw mkl::invalid_argument("DFT", "compute_forward", message); + throw math::invalid_argument("DFT", "compute_forward", message); } } } // namespace detail @@ -101,7 +101,7 @@ template ONEMATH_EXPORT void compute_forward(descriptor_type & /*desc*/, sycl::buffer, 1> & /*inout_re*/, sycl::buffer, 1> & /*inout_im*/) { - throw mkl::unimplemented("DFT", "compute_forward", + throw math::unimplemented("DFT", "compute_forward", "MKLGPU does not support compute_forward(desc, inout_re, inout_im)."); } @@ -125,7 +125,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, detail::expect_config( desc, "Unexpected value for complex storage"); - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", "compute_forward(desc, in_re, in_im, out_re, out_im)", "MKLGPU does not support out-of-place FFT with real-real complex storage."); } @@ -147,7 +147,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type & /*desc*/, scalar * /*inout_re*/, scalar * /*inout_im*/, const std::vector & /*dependencies*/) { - throw mkl::unimplemented( + throw math::unimplemented( "DFT", "compute_forward", "MKLGPU does not support compute_forward(desc, inout_re, inout_im, dependencies)."); } @@ -174,7 +174,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, detail::expect_config( desc, "Unexpected value for complex storage"); - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", "compute_forward(desc, in_re, in_im, out_re, out_im, dependencies)", "MKLGPU does not support out-of-place FFT with real-real complex storage."); } @@ -182,4 +182,4 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, // Template function instantiations #include "dft/backends/backend_forward_instantiations.cxx" -} // namespace oneapi::mkl::dft::mklgpu +} // namespace oneapi::math::dft::mklgpu diff --git a/src/dft/backends/mklgpu/mklgpu_helpers.hpp b/src/dft/backends/mklgpu/mklgpu_helpers.hpp index afecb984d..73c0d37f3 100644 --- a/src/dft/backends/mklgpu/mklgpu_helpers.hpp +++ b/src/dft/backends/mklgpu/mklgpu_helpers.hpp @@ -32,7 +32,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace dft { namespace mklgpu { namespace detail { @@ -78,7 +78,7 @@ inline constexpr dft::config_param to_mklgpu(dft::detail::config_param param) { case iparam::WORKSPACE_EXTERNAL_BYTES: return oparam::WORKSPACE_BYTES; case iparam::COMMIT_STATUS: return oparam::COMMIT_STATUS; default: - throw mkl::invalid_argument("dft", "MKLGPU descriptor set_value()", + throw math::invalid_argument("dft", "MKLGPU descriptor set_value()", "Invalid config param."); return static_cast(0); } @@ -98,7 +98,7 @@ inline constexpr int to_mklgpu( return DFTI_COMPLEX_COMPLEX; } else { - throw mkl::unimplemented("dft", "MKLGPU descriptor set_value()", + throw math::unimplemented("dft", "MKLGPU descriptor set_value()", "MKLGPU only supports complex-complex for complex storage."); return 0; } @@ -111,7 +111,7 @@ inline constexpr int to_mklgpu( return DFTI_NOT_INPLACE; } else { - throw mkl::invalid_argument("dft", "MKLGPU descriptor set_value()", + throw math::invalid_argument("dft", "MKLGPU descriptor set_value()", "Invalid config value for inplace."); return 0; } @@ -140,7 +140,7 @@ inline constexpr int to_mklgpu( return DFTI_CCE_FORMAT; } else { - throw mkl::invalid_argument("dft", "MKLGPU descriptor set_value()", + throw math::invalid_argument("dft", "MKLGPU descriptor set_value()", "Invalid config value for packed format."); return 0; } @@ -165,7 +165,7 @@ to_mklgpu_config_value( return dft::config_value::WORKSPACE_EXTERNAL; } else { - throw mkl::invalid_argument("dft", "MKLGPU descriptor set_value()", + throw math::invalid_argument("dft", "MKLGPU descriptor set_value()", "Invalid config value for workspace placement."); return dft::config_value::WORKSPACE_INTERNAL; } @@ -173,7 +173,7 @@ to_mklgpu_config_value( } // namespace detail } // namespace mklgpu } // namespace dft -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ONEMATH_DFT_SRC_MKLGPU_HELPERS_HPP_ diff --git a/src/dft/backends/portfft/commit.cpp b/src/dft/backends/portfft/commit.cpp index 4d20831c9..f7dbcbf83 100644 --- a/src/dft/backends/portfft/commit.cpp +++ b/src/dft/backends/portfft/commit.cpp @@ -42,7 +42,7 @@ // alias to avoid ambiguity namespace pfft = portfft; -namespace oneapi::mkl::dft::portfft { +namespace oneapi::math::dft::portfft { namespace detail { template @@ -61,11 +61,11 @@ class portfft_commit final : public dft::detail::commit_impl { public: portfft_commit(sycl::queue& queue, const dft::detail::dft_values& config_values) - : oneapi::mkl::dft::detail::commit_impl(queue, backend::portfft, + : oneapi::math::dft::detail::commit_impl(queue, backend::portfft, config_values) { if constexpr (prec == dft::detail::precision::DOUBLE) { if (!queue.get_device().has(sycl::aspect::fp64)) { - throw mkl::exception("DFT", "commit", "Device does not support double precision."); + throw math::exception("DFT", "commit", "Device does not support double precision."); } } } @@ -73,19 +73,19 @@ class portfft_commit final : public dft::detail::commit_impl { void commit(const dft::detail::dft_values& config_values) override { // not available in portFFT: this->external_workspace_helper_ = - oneapi::mkl::dft::detail::external_workspace_helper( + oneapi::math::dft::detail::external_workspace_helper( config_values.workspace_placement == - oneapi::mkl::dft::detail::config_value::WORKSPACE_EXTERNAL); + oneapi::math::dft::detail::config_value::WORKSPACE_EXTERNAL); if (config_values.workspace != config_value::ALLOW) { - throw mkl::unimplemented("dft/backends/portfft", __FUNCTION__, + throw math::unimplemented("dft/backends/portfft", __FUNCTION__, "portFFT only supports ALLOW for the WORKSPACE parameter"); } if (config_values.ordering != config_value::ORDERED) { - throw mkl::unimplemented("dft/backends/portfft", __FUNCTION__, + throw math::unimplemented("dft/backends/portfft", __FUNCTION__, "portFFT only supports ORDERED for the ORDERING parameter"); } if (config_values.transpose) { - throw mkl::unimplemented("dft/backends/portfft", __FUNCTION__, + throw math::unimplemented("dft/backends/portfft", __FUNCTION__, "portFFT does not supported transposed output"); } @@ -140,7 +140,7 @@ class portfft_commit final : public dft::detail::commit_impl { committed_descriptors[1] = bwd_desc.commit(q); } catch (const pfft::unsupported_configuration& e) { - throw oneapi::mkl::unimplemented("dft/backends/portfft", __FUNCTION__, e.what()); + throw oneapi::math::unimplemented("dft/backends/portfft", __FUNCTION__, e.what()); } } @@ -181,13 +181,13 @@ class portfft_commit final : public dft::detail::commit_impl { sycl::buffer&) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_forward"); - throw oneapi::mkl::unimplemented("DFT", "compute_forward(desc, inout_re, inout_im)", + throw oneapi::math::unimplemented("DFT", "compute_forward(desc, inout_re, inout_im)", "portFFT does not support real-real complex storage."); } sycl::event forward_ip_rr(descriptor_type& desc, scalar_type*, scalar_type*, const std::vector&) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_forward"); - throw oneapi::mkl::unimplemented("DFT", + throw oneapi::math::unimplemented("DFT", "compute_forward(desc, inout_re, inout_im, dependencies)", "portFFT does not support real-real complex storage."); } @@ -222,14 +222,14 @@ class portfft_commit final : public dft::detail::commit_impl { sycl::buffer&) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_forward"); - throw oneapi::mkl::unimplemented("DFT", + throw oneapi::math::unimplemented("DFT", "compute_forward(desc, in_re, in_im, out_re, out_im)", "portFFT does not support real-real complex storage."); } sycl::event forward_op_rr(descriptor_type& desc, scalar_type*, scalar_type*, scalar_type*, scalar_type*, const std::vector&) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_forward"); - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", "compute_forward(desc, in_re, in_im, out_re, out_im, dependencies)", "portFFT does not support real-real complex storage."); } @@ -262,14 +262,14 @@ class portfft_commit final : public dft::detail::commit_impl { sycl::buffer&) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_backward"); - throw oneapi::mkl::unimplemented("DFT", "compute_backward(desc, inout_re, inout_im)", + throw oneapi::math::unimplemented("DFT", "compute_backward(desc, inout_re, inout_im)", "portFFT does not support real-real complex storage."); } sycl::event backward_ip_rr(descriptor_type& desc, scalar_type*, scalar_type*, const std::vector&) override { dft::detail::get_commit(desc)->template compute_call_throw( "compute_backward"); - throw oneapi::mkl::unimplemented("DFT", + throw oneapi::math::unimplemented("DFT", "compute_backward(desc, inout_re, inout_im, dependencies)", "portFFT does not support real-real complex storage."); } @@ -304,7 +304,7 @@ class portfft_commit final : public dft::detail::commit_impl { sycl::buffer&) override { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_backward"); - throw oneapi::mkl::unimplemented("DFT", + throw oneapi::math::unimplemented("DFT", "compute_backward(desc, in_re, in_im, out_re, out_im)", "portFFT does not support real-real complex storage."); } @@ -312,7 +312,7 @@ class portfft_commit final : public dft::detail::commit_impl { scalar_type*, const std::vector&) override { dft::detail::get_commit(desc)->template compute_call_throw( "compute_backward"); - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", "compute_backward(desc, in_re, in_im, out_re, out_im, deps)", "portFFT does not support real-real complex storage."); } @@ -342,4 +342,4 @@ create_commit( const dft::detail::descriptor&, sycl::queue&); -} // namespace oneapi::mkl::dft::portfft +} // namespace oneapi::math::dft::portfft diff --git a/src/dft/backends/portfft/descriptor.cpp b/src/dft/backends/portfft/descriptor.cpp index 21c59e8d3..9308bca93 100644 --- a/src/dft/backends/portfft/descriptor.cpp +++ b/src/dft/backends/portfft/descriptor.cpp @@ -22,7 +22,7 @@ #include "oneapi/math/dft/detail/portfft/onemath_dft_portfft.hpp" -namespace oneapi::mkl::dft::detail { +namespace oneapi::math::dft::detail { template void descriptor::commit(backend_selector selector) { @@ -44,4 +44,4 @@ template void descriptor::commit( template void descriptor::commit( backend_selector); -} // namespace oneapi::mkl::dft::detail +} // namespace oneapi::math::dft::detail diff --git a/src/dft/backends/portfft/portfft_helper.hpp b/src/dft/backends/portfft/portfft_helper.hpp index 446dd3b3f..c4270cd3c 100644 --- a/src/dft/backends/portfft/portfft_helper.hpp +++ b/src/dft/backends/portfft/portfft_helper.hpp @@ -29,13 +29,13 @@ namespace pfft = portfft; -namespace oneapi::mkl::dft::portfft::detail { +namespace oneapi::math::dft::portfft::detail { template inline dft::detail::commit_impl *checked_get_commit( dft::detail::descriptor &desc) { auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::portfft) { - throw mkl::invalid_argument("dft/backends/portfft", "get_commit", + throw math::invalid_argument("dft/backends/portfft", "get_commit", "DFT descriptor has not been commited for portFFT"); } return commit_handle; @@ -57,6 +57,6 @@ auto get_descriptors(descriptor_type &desc) { auto commit = detail::checked_get_commit(desc); return reinterpret_cast *>(commit->get_handle()); } -} // namespace oneapi::mkl::dft::portfft::detail +} // namespace oneapi::math::dft::portfft::detail #endif diff --git a/src/dft/backends/rocfft/backward.cpp b/src/dft/backends/rocfft/backward.cpp index d9583a1c6..d0f47bd00 100644 --- a/src/dft/backends/rocfft/backward.cpp +++ b/src/dft/backends/rocfft/backward.cpp @@ -35,7 +35,7 @@ #include #include -namespace oneapi::mkl::dft::rocfft { +namespace oneapi::math::dft::rocfft { namespace detail { //forward declaration template @@ -70,7 +70,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, offsets[0] *= 2; // offset is supplied in complex but we offset scalar pointer } if (offsets[0] != offsets[1]) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "rocFFT requires input and output offsets (first value in strides) to be equal for in-place transforms!"); } @@ -103,7 +103,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, auto offsets = detail::get_offsets_bwd(commit); if (offsets[0] != offsets[1]) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "rocFFT requires input and output offsets (first value in strides) to be equal for in-place transforms!"); } @@ -226,7 +226,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd #include -namespace oneapi::mkl::dft::rocfft { +namespace oneapi::math::dft::rocfft { namespace detail { // rocfft has global setup and cleanup functions which use some global state internally. @@ -55,7 +55,7 @@ class rocfft_singleton { rocfft_singleton() { const auto result = rocfft_setup(); if (result != rocfft_status_success) { - throw mkl::exception( + throw math::exception( "DFT", "rocfft", "Failed to setup rocfft. returned status " + std::to_string(result)); } @@ -93,11 +93,11 @@ class rocfft_commit final : public dft::detail::commit_impl { public: rocfft_commit(sycl::queue& queue, const dft::detail::dft_values& config_values) - : oneapi::mkl::dft::detail::commit_impl(queue, backend::rocfft, + : oneapi::math::dft::detail::commit_impl(queue, backend::rocfft, config_values) { if constexpr (prec == dft::detail::precision::DOUBLE) { if (!queue.get_device().has(sycl::aspect::fp64)) { - throw mkl::exception("DFT", "commit", "Device does not support double precision."); + throw math::exception("DFT", "commit", "Device does not support double precision."); } } // initialise the rocFFT global state @@ -107,14 +107,14 @@ class rocfft_commit final : public dft::detail::commit_impl { void clean_plans() { if (handles[0].plan) { if (rocfft_plan_destroy(handles[0].plan.value()) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to destroy forward plan."); } handles[0].plan = std::nullopt; } if (handles[1].plan) { if (rocfft_plan_destroy(handles[1].plan.value()) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to destroy backward plan."); } handles[1].plan = std::nullopt; @@ -122,14 +122,14 @@ class rocfft_commit final : public dft::detail::commit_impl { if (handles[0].info) { if (rocfft_execution_info_destroy(handles[0].info.value()) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to destroy forward execution info ."); } handles[0].info = std::nullopt; } if (handles[1].info) { if (rocfft_execution_info_destroy(handles[1].info.value()) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to destroy backward execution info ."); } handles[1].info = std::nullopt; @@ -141,9 +141,9 @@ class rocfft_commit final : public dft::detail::commit_impl { void commit(const dft::detail::dft_values& config_values) override { // this could be a recommit this->external_workspace_helper_ = - oneapi::mkl::dft::detail::external_workspace_helper( + oneapi::math::dft::detail::external_workspace_helper( config_values.workspace_placement == - oneapi::mkl::dft::detail::config_value::WORKSPACE_EXTERNAL); + oneapi::math::dft::detail::config_value::WORKSPACE_EXTERNAL); clean_plans(); const rocfft_result_placement placement = @@ -214,7 +214,7 @@ class rocfft_commit final : public dft::detail::commit_impl { } else { if (config_values.conj_even_storage != dft::config_value::COMPLEX_COMPLEX) { - throw mkl::exception( + throw math::exception( "dft/backends/rocfft", __FUNCTION__, "only COMPLEX_COMPLEX conjugate_even_storage is supported"); } @@ -243,7 +243,7 @@ class rocfft_commit final : public dft::detail::commit_impl { if (strides[i] > strides[j] && strides[i] % cplx_dim != 0 && strides[i] % real_dim != 0) { // rocfft does not throw, it just produces wrong results - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func, "rocfft requires a stride to be divisible by all dimensions associated with smaller strides!"); } @@ -271,7 +271,7 @@ class rocfft_commit final : public dft::detail::commit_impl { if (dom == dft::domain::COMPLEX && config_values.placement == dft::config_value::NOT_INPLACE && dimensions > 2) { if (stride_vecs.vec_a != stride_vecs.vec_b) - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func, "due to a bug in rocfft version in use, it requires fwd and bwd stride to be the same for COMPLEX out_of_place computations"); } @@ -279,17 +279,17 @@ class rocfft_commit final : public dft::detail::commit_impl { rocfft_plan_description plan_desc_fwd, plan_desc_bwd; // Can't reuse with ROCm 6 due to bug. if (rocfft_plan_description_create(&plan_desc_fwd) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to create plan description."); } if (rocfft_plan_description_create(&plan_desc_bwd) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to create plan description."); } // plan_description can be destroyed afted plan_create auto description_destroy = [](rocfft_plan_description p) { if (rocfft_plan_description_destroy(p) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to destroy plan description."); } }; @@ -342,7 +342,7 @@ class rocfft_commit final : public dft::detail::commit_impl { (vec_b_valid_as_bwd_domain && vec_a_valid_as_fwd_domain); if (!valid_forward && !valid_backward) { - throw mkl::exception("dft/backends/cufft", __FUNCTION__, "Invalid strides."); + throw math::exception("dft/backends/cufft", __FUNCTION__, "Invalid strides."); } if (valid_forward) { @@ -358,13 +358,13 @@ class rocfft_commit final : public dft::detail::commit_impl { bwd_distance // out distance ); if (res != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to set forward data layout."); } if (rocfft_plan_description_set_scale_factor(plan_desc_fwd, config_values.fwd_scale) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to set forward scale factor."); } @@ -373,7 +373,7 @@ class rocfft_commit final : public dft::detail::commit_impl { lengths.data(), number_of_transforms, plan_desc_fwd); if (res != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to create forward plan."); } @@ -381,7 +381,7 @@ class rocfft_commit final : public dft::detail::commit_impl { rocfft_execution_info fwd_info; if (rocfft_execution_info_create(&fwd_info) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to create forward execution info."); } handles[0].info = fwd_info; @@ -391,7 +391,7 @@ class rocfft_commit final : public dft::detail::commit_impl { if (work_buf_size != 0) { void* work_buf; if (hipMalloc(&work_buf, work_buf_size) != hipSuccess) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to get allocate forward work buffer."); } set_workspace_impl(handles[0], reinterpret_cast(work_buf), @@ -414,13 +414,13 @@ class rocfft_commit final : public dft::detail::commit_impl { fwd_distance // out distance ); if (res != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to set backward data layout."); } if (rocfft_plan_description_set_scale_factor(plan_desc_bwd, config_values.bwd_scale) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to set backward scale factor."); } @@ -428,14 +428,14 @@ class rocfft_commit final : public dft::detail::commit_impl { res = rocfft_plan_create(&bwd_plan, placement, bwd_type, precision, dimensions, lengths.data(), number_of_transforms, plan_desc_bwd); if (res != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to create backward rocFFT plan."); } handles[1].plan = bwd_plan; rocfft_execution_info bwd_info; if (rocfft_execution_info_create(&bwd_info) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to create backward execution info."); } handles[1].info = bwd_info; @@ -445,7 +445,7 @@ class rocfft_commit final : public dft::detail::commit_impl { if (work_buf_size != 0) { void* work_buf; if (hipMalloc(&work_buf, work_buf_size) != hipSuccess) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to get allocate backward work buffer."); } set_workspace_impl(handles[1], reinterpret_cast(work_buf), @@ -484,11 +484,11 @@ class rocfft_commit final : public dft::detail::commit_impl { **/ std::int64_t get_rocfft_workspace_bytes(rocfft_handle& handle, const char* function) { if (!handle.plan) { - throw mkl::exception("dft/backends/rocfft", function, "Missing internal rocfft plan"); + throw math::exception("dft/backends/rocfft", function, "Missing internal rocfft plan"); } std::size_t size = 0; if (rocfft_plan_get_work_buffer_size(*handle.plan, &size) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", function, + throw math::exception("dft/backends/rocfft", function, "Failed to get rocfft work buffer size."); } return static_cast(size); @@ -505,31 +505,31 @@ class rocfft_commit final : public dft::detail::commit_impl { void set_workspace_impl(const rocfft_handle& handle, scalar_type* workspace, std::int64_t workspace_bytes, const char* function) { if (!handle.info) { - throw mkl::exception( + throw math::exception( "dft/backends/rocfft", function, "Could not set rocFFT workspace - handle has no associated rocfft_info."); } if (handle.buffer) { - throw mkl::exception( + throw math::exception( "dft/backends/rocfft", function, "Could not set rocFFT workspace - an internal buffer is already set."); } if (workspace_bytes && workspace == nullptr) { - throw mkl::exception("dft/backends/rocfft", function, "Trying to nullptr workspace."); + throw math::exception("dft/backends/rocfft", function, "Trying to nullptr workspace."); } auto info = *handle.info; if (workspace_bytes && rocfft_execution_info_set_work_buffer(info, static_cast(workspace), static_cast(workspace_bytes)) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", function, "Failed to set work buffer."); + throw math::exception("dft/backends/rocfft", function, "Failed to set work buffer."); } } void free_internal_workspace_if_rqd(rocfft_handle& handle, const char* function) { if (handle.buffer) { if (hipFree(*handle.buffer) != hipSuccess) { - throw mkl::exception("dft/backends/rocfft", function, + throw math::exception("dft/backends/rocfft", function, "Failed to free internal buffer."); } handle.buffer = std::nullopt; @@ -583,12 +583,12 @@ class rocfft_commit final : public dft::detail::commit_impl { std::int64_t get_plan_workspace_size_bytes(rocfft_plan_t* plan) { // plan work buffer if (plan == nullptr) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Missing internal rocFFT plan."); } std::size_t work_buf_size; if (rocfft_plan_get_work_buffer_size(plan, &work_buf_size) != rocfft_status_success) { - throw mkl::exception("dft/backends/rocfft", __FUNCTION__, + throw math::exception("dft/backends/rocfft", __FUNCTION__, "Failed to get work buffer size."); } return static_cast(work_buf_size); @@ -668,4 +668,4 @@ get_offsets_bwd( } //namespace detail -} // namespace oneapi::mkl::dft::rocfft +} // namespace oneapi::math::dft::rocfft diff --git a/src/dft/backends/rocfft/descriptor.cpp b/src/dft/backends/rocfft/descriptor.cpp index ba5fbb139..d1bd88887 100644 --- a/src/dft/backends/rocfft/descriptor.cpp +++ b/src/dft/backends/rocfft/descriptor.cpp @@ -22,7 +22,7 @@ #include "oneapi/math/dft/detail/rocfft/onemath_dft_rocfft.hpp" -namespace oneapi::mkl::dft::detail { +namespace oneapi::math::dft::detail { template void descriptor::commit(backend_selector selector) { @@ -44,4 +44,4 @@ template void descriptor::commit( template void descriptor::commit( backend_selector); -} //namespace oneapi::mkl::dft::detail +} //namespace oneapi::math::dft::detail diff --git a/src/dft/backends/rocfft/execute_helper.hpp b/src/dft/backends/rocfft/execute_helper.hpp index b5b4cc353..b73bc3a32 100644 --- a/src/dft/backends/rocfft/execute_helper.hpp +++ b/src/dft/backends/rocfft/execute_helper.hpp @@ -34,27 +34,27 @@ #include #include -namespace oneapi::mkl::dft::rocfft::detail { +namespace oneapi::math::dft::rocfft::detail { template inline dft::detail::commit_impl *checked_get_commit( dft::detail::descriptor &desc) { auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::rocfft) { - throw mkl::invalid_argument("dft/backends/rocfft", "get_commit", + throw math::invalid_argument("dft/backends/rocfft", "get_commit", "DFT descriptor has not been commited for rocFFT"); } return commit_handle; } -/// Throw an mkl::invalid_argument if the runtime param in the descriptor does not match +/// Throw an math::invalid_argument if the runtime param in the descriptor does not match /// the expected value. template inline auto expect_config(DescT &desc, const char *message) { dft::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { - throw mkl::invalid_argument("dft/backends/rocfft", "expect_config", message); + throw math::invalid_argument("dft/backends/rocfft", "expect_config", message); } } @@ -68,7 +68,7 @@ inline hipStream_t setup_stream(const std::string &func, sycl::interop_handle &i auto stream = ih.get_native_queue(); auto result = rocfft_execution_info_set_stream(info, stream); if (result != rocfft_status_success) { - throw oneapi::mkl::exception( + throw oneapi::math::exception( "dft/backends/rocfft", func, "rocfft_execution_info_set_stream returned " + std::to_string(result)); } @@ -78,8 +78,8 @@ inline hipStream_t setup_stream(const std::string &func, sycl::interop_handle &i inline void sync_checked(const std::string &func, hipStream_t stream) { auto result = hipStreamSynchronize(stream); if (result != hipSuccess) { - throw oneapi::mkl::exception("dft/backends/rocfft", func, - "hipStreamSynchronize returned " + std::to_string(result)); + throw oneapi::math::exception("dft/backends/rocfft", func, + "hipStreamSynchronize returned " + std::to_string(result)); } } @@ -87,8 +87,8 @@ inline void execute_checked(const std::string &func, hipStream_t stream, const r void *out_buffer[], rocfft_execution_info info) { auto result = rocfft_execute(plan, in_buffer, out_buffer, info); if (result != rocfft_status_success) { - throw oneapi::mkl::exception("dft/backends/rocfft", func, - "rocfft_execute returned " + std::to_string(result)); + throw oneapi::math::exception("dft/backends/rocfft", func, + "rocfft_execute returned " + std::to_string(result)); } #ifndef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND // If not using equeue native extension, the host task must wait on the @@ -98,6 +98,6 @@ inline void execute_checked(const std::string &func, hipStream_t stream, const r #endif } -} // namespace oneapi::mkl::dft::rocfft::detail +} // namespace oneapi::math::dft::rocfft::detail #endif // _ONEMATH_DFT_SRC_EXECUTE_HELPER_ROCFFT_HPP_ diff --git a/src/dft/backends/rocfft/forward.cpp b/src/dft/backends/rocfft/forward.cpp index ebafc9e2f..dfec5fbd2 100644 --- a/src/dft/backends/rocfft/forward.cpp +++ b/src/dft/backends/rocfft/forward.cpp @@ -36,7 +36,7 @@ #include #include -namespace oneapi::mkl::dft::rocfft { +namespace oneapi::math::dft::rocfft { namespace detail { //forward declaration @@ -73,7 +73,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, offsets[1] *= 2; // offset is supplied in complex but we offset scalar pointer } if (offsets[0] != offsets[1]) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "rocFFT requires input and output offsets (first value in strides) to be equal for in-place transforms!"); } @@ -106,7 +106,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, auto offsets = detail::get_offsets_fwd(commit); if (offsets[0] != offsets[1]) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "DFT", func_name, "rocFFT requires input and output offsets (first value in strides) to be equal for in-place transforms!"); } @@ -228,7 +228,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd) { if (v1[i] < 0 || v2[i] < 0) { - throw mkl::unimplemented("DFT", "commit", + throw math::unimplemented("DFT", "commit", "Backend does not support negative strides."); } } @@ -146,6 +146,6 @@ inline stride_api get_stride_api(const ConfigT& config_values) { return stride_api::INVALID; } -} // namespace oneapi::mkl::dft::detail +} // namespace oneapi::math::dft::detail #endif //_DFT_DETAIL_STRIDE_HELPER_HPP_ diff --git a/src/dft/descriptor.cxx b/src/dft/descriptor.cxx index a823817c0..df90d9f0e 100644 --- a/src/dft/descriptor.cxx +++ b/src/dft/descriptor.cxx @@ -24,7 +24,7 @@ #include "dft/descriptor_config_helper.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace dft { namespace detail { @@ -58,10 +58,10 @@ void descriptor::set_value(config_param param, ...) { va_start(vl, param); switch (param) { case config_param::FORWARD_DOMAIN: - throw mkl::invalid_argument("DFT", "set_value", "Read-only parameter."); + throw math::invalid_argument("DFT", "set_value", "Read-only parameter."); break; case config_param::DIMENSION: - throw mkl::invalid_argument("DFT", "set_value", "Read-only parameter."); + throw math::invalid_argument("DFT", "set_value", "Read-only parameter."); break; case config_param::LENGTHS: { if (values_.dimensions.size() == 1) { @@ -74,7 +74,7 @@ void descriptor::set_value(config_param param, ...) { break; } case config_param::PRECISION: - throw mkl::invalid_argument("DFT", "set_value", "Read-only parameter."); + throw math::invalid_argument("DFT", "set_value", "Read-only parameter."); break; #pragma clang diagnostic push #pragma clang diagnostic ignored "-Wdeprecated-declarations" @@ -136,15 +136,15 @@ void descriptor::set_value(config_param param, ...) { detail::set_value(values_, va_arg(vl, config_value)); break; case config_param::WORKSPACE_EXTERNAL_BYTES: - throw mkl::invalid_argument("DFT", "set_value", "Read-only parameter."); + throw math::invalid_argument("DFT", "set_value", "Read-only parameter."); break; case config_param::PACKED_FORMAT: detail::set_value(values_, va_arg(vl, config_value)); break; case config_param::COMMIT_STATUS: - throw mkl::invalid_argument("DFT", "set_value", "Read-only parameter."); + throw math::invalid_argument("DFT", "set_value", "Read-only parameter."); break; - default: throw mkl::invalid_argument("DFT", "set_value", "Invalid config_param argument."); + default: throw math::invalid_argument("DFT", "set_value", "Invalid config_param argument."); } va_end(vl); } @@ -152,11 +152,11 @@ void descriptor::set_value(config_param param, ...) { template descriptor::descriptor(std::vector dimensions) { if (dimensions.size() == 0) { - throw mkl::invalid_argument("DFT", "descriptor", "Cannot have 0 dimensional DFT."); + throw math::invalid_argument("DFT", "descriptor", "Cannot have 0 dimensional DFT."); } for (const auto& dim : dimensions) { if (dim <= 0) { - throw mkl::invalid_argument("DFT", "descriptor", + throw math::invalid_argument("DFT", "descriptor", "Invalid dimension value (negative or 0)."); } } @@ -198,7 +198,7 @@ void descriptor::get_value(config_param param, ...) const { va_list vl; va_start(vl, param); if (va_arg(vl, void*) == nullptr) { - throw mkl::invalid_argument("DFT", "get_value", "config_param is nullptr."); + throw math::invalid_argument("DFT", "get_value", "config_param is nullptr."); } va_end(vl); va_start(vl, param); @@ -256,7 +256,7 @@ void descriptor::get_value(config_param param, ...) const { break; case config_param::WORKSPACE_EXTERNAL_BYTES: if (!pimpl_) { - throw mkl::invalid_argument( + throw math::invalid_argument( "DFT", "get_value", "Cannot query WORKSPACE_EXTERNAL_BYTES on uncommitted descriptor."); } @@ -271,7 +271,7 @@ void descriptor::get_value(config_param param, ...) const { *va_arg(vl, config_value*) = pimpl_ ? config_value::COMMITTED : config_value::UNCOMMITTED; break; - default: throw mkl::invalid_argument("DFT", "get_value", "Invalid config_param argument."); + default: throw math::invalid_argument("DFT", "get_value", "Invalid config_param argument."); } va_end(vl); } @@ -282,7 +282,7 @@ void descriptor::set_workspace(scalar_type* usm_workspace) { return pimpl_->set_workspace(usm_workspace); } else { - throw mkl::uninitialized("DFT", "set_workspace", + throw math::uninitialized("DFT", "set_workspace", "Can only set workspace on committed descriptor."); } } @@ -293,7 +293,7 @@ void descriptor::set_workspace(sycl::buffer& buffer_work return pimpl_->set_workspace(buffer_workspace); } else { - throw mkl::uninitialized("DFT", "set_workspace", + throw math::uninitialized("DFT", "set_workspace", "Can only set workspace on committed descriptor."); } } @@ -305,5 +305,5 @@ template class descriptor; } //namespace detail } //namespace dft -} //namespace mkl +} //namespace math } //namespace oneapi diff --git a/src/dft/descriptor_config_helper.hpp b/src/dft/descriptor_config_helper.hpp index 6a6c4cced..c4d319096 100644 --- a/src/dft/descriptor_config_helper.hpp +++ b/src/dft/descriptor_config_helper.hpp @@ -26,7 +26,7 @@ #include "oneapi/math/dft/descriptor.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace dft { namespace detail { @@ -113,18 +113,18 @@ void set_value(dft_values& vals, param_type_helper_t, Param>&& set_val) { if constexpr (Param == config_param::LENGTHS) { if (set_val == nullptr) { - throw mkl::invalid_argument("DFT", "set_value", "Given nullptr."); + throw math::invalid_argument("DFT", "set_value", "Given nullptr."); } for (std::size_t i{ 0 }; i < vals.dimensions.size(); ++i) { if (set_val[i] <= 0) { - throw mkl::invalid_argument("DFT", "set_value", + throw math::invalid_argument("DFT", "set_value", "Invalid length value (negative or 0)."); } } std::copy(set_val, set_val + vals.dimensions.size(), vals.dimensions.begin()); } else if constexpr (Param == config_param::PRECISION) { - throw mkl::invalid_argument("DFT", "set_value", "Read-only parameter."); + throw math::invalid_argument("DFT", "set_value", "Read-only parameter."); } else if constexpr (Param == config_param::FORWARD_SCALE) { vals.fwd_scale = set_val; @@ -134,7 +134,7 @@ void set_value(dft_values& vals, } else if constexpr (Param == config_param::NUMBER_OF_TRANSFORMS) { if (set_val <= 0) { - throw mkl::invalid_argument("DFT", "set_value", + throw math::invalid_argument("DFT", "set_value", "Number of transforms must be positive."); } vals.number_of_transforms = set_val; @@ -144,7 +144,7 @@ void set_value(dft_values& vals, vals.complex_storage = set_val; } else { - throw mkl::invalid_argument("DFT", "set_value", + throw math::invalid_argument("DFT", "set_value", "Complex storage must be complex_complex or real_real."); } } @@ -153,7 +153,7 @@ void set_value(dft_values& vals, vals.real_storage = set_val; } else { - throw mkl::invalid_argument("DFT", "set_value", "Real storage must be real_real."); + throw math::invalid_argument("DFT", "set_value", "Real storage must be real_real."); } } else if constexpr (Param == config_param::CONJUGATE_EVEN_STORAGE) { @@ -161,7 +161,7 @@ void set_value(dft_values& vals, vals.conj_even_storage = set_val; } else { - throw mkl::invalid_argument("DFT", "set_value", + throw math::invalid_argument("DFT", "set_value", "Conjugate even storage must be complex_complex."); } } @@ -170,7 +170,7 @@ void set_value(dft_values& vals, vals.placement = set_val; } else { - throw mkl::invalid_argument("DFT", "set_value", + throw math::invalid_argument("DFT", "set_value", "Placement must be inplace or not inplace."); } } @@ -178,14 +178,14 @@ void set_value(dft_values& vals, #pragma clang diagnostic ignored "-Wdeprecated-declarations" else if constexpr (Param == config_param::INPUT_STRIDES) { if (set_val == nullptr) { - throw mkl::invalid_argument("DFT", "set_value", "Given nullptr."); + throw math::invalid_argument("DFT", "set_value", "Given nullptr."); } reset_strides_to_zero(vals.fwd_strides, vals.bwd_strides); std::copy(set_val, set_val + vals.dimensions.size() + 1, vals.input_strides.begin()); } else if constexpr (Param == config_param::OUTPUT_STRIDES) { if (set_val == nullptr) { - throw mkl::invalid_argument("DFT", "set_value", "Given nullptr."); + throw math::invalid_argument("DFT", "set_value", "Given nullptr."); } reset_strides_to_zero(vals.fwd_strides, vals.bwd_strides); std::copy(set_val, set_val + vals.dimensions.size() + 1, vals.output_strides.begin()); @@ -202,7 +202,7 @@ void set_value(dft_values& vals, vals.workspace = set_val; } else { - throw mkl::invalid_argument("DFT", "set_value", "Workspace must be allow or avoid."); + throw math::invalid_argument("DFT", "set_value", "Workspace must be allow or avoid."); } } else if constexpr (Param == config_param::WORKSPACE_PLACEMENT) { @@ -211,19 +211,19 @@ void set_value(dft_values& vals, vals.workspace_placement = set_val; } else { - throw mkl::invalid_argument( + throw math::invalid_argument( "DFT", "set_value", "Workspace must be WORKSPACE_AUTOMATIC or WORKSPACE_EXTERNAL."); } } else if constexpr (Param == config_param::WORKSPACE_EXTERNAL_BYTES) { - throw mkl::invalid_argument("DFT", "set_value", "Read-only parameter."); + throw math::invalid_argument("DFT", "set_value", "Read-only parameter."); } else if constexpr (Param == config_param::ORDERING) { if (set_val == config_value::ORDERED || set_val == config_value::BACKWARD_SCRAMBLED) { vals.ordering = set_val; } else { - throw mkl::invalid_argument("DFT", "set_value", + throw math::invalid_argument("DFT", "set_value", "Ordering must be ordered or backwards scrambled."); } } @@ -235,19 +235,19 @@ void set_value(dft_values& vals, vals.packed_format = set_val; } else { - throw mkl::invalid_argument("DFT", "set_value", "Packed format must be CCE."); + throw math::invalid_argument("DFT", "set_value", "Packed format must be CCE."); } } else if constexpr (Param == config_param::FWD_STRIDES) { if (set_val == nullptr) { - throw mkl::invalid_argument("DFT", "set_value", "Given nullptr."); + throw math::invalid_argument("DFT", "set_value", "Given nullptr."); } reset_strides_to_zero(vals.input_strides, vals.output_strides); std::copy(set_val, set_val + vals.dimensions.size() + 1, vals.fwd_strides.begin()); } else if constexpr (Param == config_param::BWD_STRIDES) { if (set_val == nullptr) { - throw mkl::invalid_argument("DFT", "set_value", "Given nullptr."); + throw math::invalid_argument("DFT", "set_value", "Given nullptr."); } reset_strides_to_zero(vals.input_strides, vals.output_strides); std::copy(set_val, set_val + vals.dimensions.size() + 1, vals.bwd_strides.begin()); @@ -256,7 +256,7 @@ void set_value(dft_values& vals, } // namespace detail } // namespace dft -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ONEMATH_DETAIL_DESCRIPTOR_CONFIG_HELPER_HPP_ diff --git a/src/dft/dft_loader.cpp b/src/dft/dft_loader.cpp index f5adec10b..3984208b2 100644 --- a/src/dft/dft_loader.cpp +++ b/src/dft/dft_loader.cpp @@ -25,9 +25,9 @@ #include "dft/function_table.hpp" #include "oneapi/math/detail/get_device_id.hpp" -namespace oneapi::mkl::dft::detail { +namespace oneapi::math::dft::detail { -static oneapi::mkl::detail::table_initializer +static oneapi::math::detail::table_initializer function_tables; template <> @@ -59,14 +59,14 @@ commit_impl* create_commit -inline oneapi::mkl::device get_device(descriptor& desc, const char* func_name) { +inline oneapi::math::device get_device(descriptor& desc, const char* func_name) { config_value is_committed{ config_value::UNCOMMITTED }; desc.get_value(config_param::COMMIT_STATUS, &is_committed); if (is_committed != config_value::COMMITTED) { - throw mkl::invalid_argument("DFT", func_name, "Descriptor not committed."); + throw math::invalid_argument("DFT", func_name, "Descriptor not committed."); } // Committed means that the commit pointer is not null. return get_device_id(get_commit(desc)->get_queue()); } -} // namespace oneapi::mkl::dft::detail +} // namespace oneapi::math::dft::detail diff --git a/src/dft/function_table.hpp b/src/dft/function_table.hpp index 9a491361a..244a06808 100644 --- a/src/dft/function_table.hpp +++ b/src/dft/function_table.hpp @@ -35,27 +35,27 @@ typedef struct { int version; - oneapi::mkl::dft::detail::commit_impl* ( + oneapi::math::dft::detail::commit_impl* ( *create_commit_sycl_fz)( - const oneapi::mkl::dft::descriptor& desc, + const oneapi::math::dft::descriptor& desc, sycl::queue& sycl_queue); - oneapi::mkl::dft::detail::commit_impl* ( + oneapi::math::dft::detail::commit_impl* ( *create_commit_sycl_dz)( - const oneapi::mkl::dft::descriptor& desc, + const oneapi::math::dft::descriptor& desc, sycl::queue& sycl_queue); - oneapi::mkl::dft::detail::commit_impl* (*create_commit_sycl_fr)( - const oneapi::mkl::dft::descriptor& desc, + oneapi::math::dft::detail::commit_impl* (*create_commit_sycl_fr)( + const oneapi::math::dft::descriptor& desc, sycl::queue& sycl_queue); - oneapi::mkl::dft::detail::commit_impl* (*create_commit_sycl_dr)( - const oneapi::mkl::dft::descriptor& desc, + oneapi::math::dft::detail::commit_impl* (*create_commit_sycl_dr)( + const oneapi::math::dft::descriptor& desc, sycl::queue& sycl_queue); } dft_function_table_t; diff --git a/src/include/allocator_helper.hpp b/src/include/allocator_helper.hpp index 8ea802dd1..9dfd6e36e 100644 --- a/src/include/allocator_helper.hpp +++ b/src/include/allocator_helper.hpp @@ -27,7 +27,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { static inline void *aligned_alloc(size_t align, size_t size) { #ifdef _WIN64 @@ -44,7 +44,7 @@ static inline void aligned_free(void *p) { ::free(p); #endif } -} // namespace mkl +} // namespace math } // namespace oneapi #endif // __ALLOCATOR_HELPER_HPP diff --git a/src/include/exceptions_helper.hpp b/src/include/exceptions_helper.hpp index 9db0f0ddd..cc1c46ba5 100644 --- a/src/include/exceptions_helper.hpp +++ b/src/include/exceptions_helper.hpp @@ -23,14 +23,14 @@ #include namespace oneapi { -namespace mkl { +namespace math { class backend_unsupported_exception : public std::runtime_error { public: backend_unsupported_exception() : std::runtime_error("Not yet supported for this backend") {} }; -} // namespace mkl +} // namespace math } // namespace oneapi #endif // __EXCEPTIONS_HELPER_HPP diff --git a/src/include/function_table_initializer.hpp b/src/include/function_table_initializer.hpp index 7655b5f4c..c32cb691d 100644 --- a/src/include/function_table_initializer.hpp +++ b/src/include/function_table_initializer.hpp @@ -45,10 +45,10 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace detail { -template +template class table_initializer { struct handle_deleter { using pointer = LIB_TYPE; @@ -59,7 +59,7 @@ class table_initializer { using dlhandle = std::unique_ptr; public: - function_table_t &operator[](std::pair device_queue_pair) { + function_table_t &operator[](std::pair device_queue_pair) { auto lib = tables.find(device_queue_pair.first); if (lib != tables.end()) return lib->second; @@ -96,7 +96,7 @@ class table_initializer { } #endif - function_table_t &add_table(oneapi::mkl::device key, sycl::queue &q) { + function_table_t &add_table(oneapi::math::device key, sycl::queue &q) { dlhandle handle; // check all available libraries for the key(device) for (const char *libname : libraries[domain_id][key]) { @@ -105,12 +105,12 @@ class table_initializer { break; } if (!handle) { - if (!is_generic_device_supported && key == oneapi::mkl::device::generic_device) { - throw mkl::unsupported_device("", "", q.get_device()); + if (!is_generic_device_supported && key == oneapi::math::device::generic_device) { + throw math::unsupported_device("", "", q.get_device()); } else { std::cerr << ERROR_MSG << '\n'; - throw mkl::backend_not_found(); + throw math::backend_not_found(); } } auto t = @@ -118,22 +118,22 @@ class table_initializer { if (!t) { std::cerr << ERROR_MSG << '\n'; - throw mkl::function_not_found(); + throw math::function_not_found(); } if (t->version != SPEC_VERSION) - throw mkl::specification_mismatch(); + throw math::specification_mismatch(); handles[key] = std::move(handle); tables[key] = *t; return *t; } - std::map tables; - std::map handles; + std::map tables; + std::map handles; }; } //namespace detail -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_LOADER_HPP_ diff --git a/src/lapack/backends/cusolver/cusolver_batch.cpp b/src/lapack/backends/cusolver/cusolver_batch.cpp index b193113e6..f2c990963 100644 --- a/src/lapack/backends/cusolver/cusolver_batch.cpp +++ b/src/lapack/backends/cusolver/cusolver_batch.cpp @@ -24,7 +24,7 @@ #include "oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace cusolver { @@ -185,7 +185,7 @@ GETRI_STRIDED_BATCH_LAUNCHER(std::complex, cublasZgetriBatched) template inline void getrs_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -235,7 +235,7 @@ inline void getrs_batch(const char *func_name, Func func, sycl::queue &queue, } #define GETRS_STRIDED_BATCH_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, \ + void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, \ std::int64_t stride_a, sycl::buffer &ipiv, \ std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, \ @@ -366,7 +366,7 @@ ORGQR_STRIDED_BATCH_LAUNCHER(double, cusolverDnDorgqr) template inline void potrf_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -401,7 +401,7 @@ inline void potrf_batch(const char *func_name, Func func, sycl::queue &queue, // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRF_STRIDED_BATCH_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, \ + void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, \ std::int64_t batch_size, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ @@ -418,7 +418,7 @@ POTRF_STRIDED_BATCH_LAUNCHER(std::complex, cusolverDnZpotrfBatched) template inline void potrs_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, @@ -468,7 +468,7 @@ inline void potrs_batch(const char *func_name, Func func, sycl::queue &queue, // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRS_STRIDED_BATCH_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, \ + void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, \ std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, \ std::int64_t stride_b, std::int64_t batch_size, \ @@ -936,7 +936,7 @@ sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, T *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, T *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, T *scratchpad, @@ -986,7 +986,7 @@ inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &qu } #define GETRS_STRIDED_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, \ + sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ std::int64_t nrhs, TYPE *a, std::int64_t lda, std::int64_t stride_a, \ std::int64_t *ipiv, std::int64_t stride_ipiv, TYPE *b, \ std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, \ @@ -1006,7 +1006,7 @@ GETRS_STRIDED_BATCH_LAUNCHER_USM(std::complex, cusolverDnZgetrs) template inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, T **a, std::int64_t *lda, std::int64_t **ipiv, T **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, @@ -1081,7 +1081,7 @@ inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &qu #define GETRS_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ sycl::event getrs_batch( \ - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, \ + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, \ TYPE **a, std::int64_t *lda, std::int64_t **ipiv, TYPE **b, std::int64_t *ldb, \ std::int64_t group_count, std::int64_t *group_sizes, TYPE *scratchpad, \ std::int64_t scratchpad_size, const std::vector &dependencies) { \ @@ -1200,7 +1200,7 @@ ORGQR_BATCH_LAUNCHER_USM(double, cusolverDnDorgqr) template inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1237,7 +1237,7 @@ inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &qu // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRF_STRIDED_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, \ TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1254,7 +1254,7 @@ POTRF_STRIDED_BATCH_LAUNCHER_USM(std::complex, cusolverDnZpotrfBatched) template inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, T **a, std::int64_t *lda, + oneapi::math::uplo *uplo, std::int64_t *n, T **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1300,7 +1300,7 @@ inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &qu // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRF_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ sycl::event potrf_batch( \ - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, TYPE **a, std::int64_t *lda, \ + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, TYPE **a, std::int64_t *lda, \ std::int64_t group_count, std::int64_t *group_sizes, TYPE *scratchpad, \ std::int64_t scratchpad_size, const std::vector &dependencies) { \ return potrf_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, \ @@ -1316,7 +1316,7 @@ POTRF_BATCH_LAUNCHER_USM(std::complex, cusolverDnZpotrfBatched) template inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, T *a, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, T *a, std::int64_t lda, std::int64_t stride_a, T *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, @@ -1365,7 +1365,7 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRS_STRIDED_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ sycl::event potrs_batch( \ - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, TYPE *a, \ + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, TYPE *a, \ std::int64_t lda, std::int64_t stride_a, TYPE *b, std::int64_t ldb, std::int64_t stride_b, \ std::int64_t batch_size, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1383,7 +1383,7 @@ POTRS_STRIDED_BATCH_LAUNCHER_USM(std::complex, cusolverDnZpotrsBatched) template inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, T **a, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, T **a, std::int64_t *lda, T **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, @@ -1440,7 +1440,7 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRS_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ sycl::event potrs_batch( \ - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, \ + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, \ TYPE **a, std::int64_t *lda, TYPE **b, std::int64_t *ldb, std::int64_t group_count, \ std::int64_t *group_sizes, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1614,7 +1614,7 @@ GETRI_STRIDED_BATCH_LAUNCHER_SCRATCH(std::complex) #define GETRS_STRIDED_BATCH_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t getrs_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, \ + sycl::queue & queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, \ std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, \ std::int64_t stride_b, std::int64_t batch_size) { \ return 0; \ @@ -1665,7 +1665,7 @@ GEQRF_STRIDED_BATCH_LAUNCHER_SCRATCH(std::complex, cusolverDnZgeqrf_buff #define POTRF_STRIDED_BATCH_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t potrf_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, \ + sycl::queue & queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, \ std::int64_t stride_a, std::int64_t batch_size) { \ return 0; \ } @@ -1681,7 +1681,7 @@ POTRF_STRIDED_BATCH_LAUNCHER_SCRATCH(std::complex) #define POTRS_STRIDED_BATCH_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t potrs_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, \ + sycl::queue & queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, \ std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, \ std::int64_t batch_size) { \ return 0; \ @@ -1828,7 +1828,7 @@ GETRI_GROUP_LAUNCHER_SCRATCH(std::complex) #define GETRS_GROUP_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t getrs_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::transpose * trans, std::int64_t * n, \ + sycl::queue & queue, oneapi::math::transpose * trans, std::int64_t * n, \ std::int64_t * nrhs, std::int64_t * lda, std::int64_t * ldb, std::int64_t group_count, \ std::int64_t * group_sizes) { \ return 0; \ @@ -1928,7 +1928,7 @@ ORGQR_GROUP_LAUNCHER_SCRATCH(double, cusolverDnDorgqr_bufferSize) #define POTRF_GROUP_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t potrf_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::uplo * uplo, std::int64_t * n, std::int64_t * lda, \ + sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t * n, std::int64_t * lda, \ std::int64_t group_count, std::int64_t * group_sizes) { \ return 0; \ } @@ -1944,7 +1944,7 @@ POTRF_GROUP_LAUNCHER_SCRATCH(std::complex) #define POTRS_GROUP_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t potrs_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::uplo * uplo, std::int64_t * n, std::int64_t * nrhs, \ + sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t * n, std::int64_t * nrhs, \ std::int64_t * lda, std::int64_t * ldb, std::int64_t group_count, \ std::int64_t * group_sizes) { \ return 0; \ @@ -2000,5 +2000,5 @@ UNGQR_GROUP_LAUNCHER_SCRATCH(std::complex, cusolverDnZungqr_bufferSize) } // namespace cusolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/lapack/backends/cusolver/cusolver_handle.hpp b/src/lapack/backends/cusolver/cusolver_handle.hpp index f3b587039..269295cd5 100644 --- a/src/lapack/backends/cusolver/cusolver_handle.hpp +++ b/src/lapack/backends/cusolver/cusolver_handle.hpp @@ -22,7 +22,7 @@ #include namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace cusolver { @@ -55,7 +55,7 @@ struct cusolver_handle { } // namespace cusolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif // CUSOLVER_HANDLE_HPP diff --git a/src/lapack/backends/cusolver/cusolver_helper.hpp b/src/lapack/backends/cusolver/cusolver_helper.hpp index 4163fe893..f51827f9f 100644 --- a/src/lapack/backends/cusolver/cusolver_helper.hpp +++ b/src/lapack/backends/cusolver/cusolver_helper.hpp @@ -39,7 +39,7 @@ #include "oneapi/math/lapack/exceptions.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace cusolver { @@ -220,52 +220,52 @@ inline cusolverEigType_t get_cusolver_itype(std::int64_t itype) { } } -inline cusolverEigMode_t get_cusolver_job(oneapi::mkl::job jobz) { +inline cusolverEigMode_t get_cusolver_job(oneapi::math::job jobz) { switch (jobz) { - case oneapi::mkl::job::N: return CUSOLVER_EIG_MODE_NOVECTOR; - case oneapi::mkl::job::V: return CUSOLVER_EIG_MODE_VECTOR; + case oneapi::math::job::N: return CUSOLVER_EIG_MODE_NOVECTOR; + case oneapi::math::job::V: return CUSOLVER_EIG_MODE_VECTOR; default: throw "Wrong jobz."; } } -inline signed char get_cusolver_jobsvd(oneapi::mkl::jobsvd job) { +inline signed char get_cusolver_jobsvd(oneapi::math::jobsvd job) { switch (job) { - case oneapi::mkl::jobsvd::N: return 'N'; - case oneapi::mkl::jobsvd::A: return 'A'; - case oneapi::mkl::jobsvd::O: return 'O'; - case oneapi::mkl::jobsvd::S: return 'S'; + case oneapi::math::jobsvd::N: return 'N'; + case oneapi::math::jobsvd::A: return 'A'; + case oneapi::math::jobsvd::O: return 'O'; + case oneapi::math::jobsvd::S: return 'S'; } } -inline cublasOperation_t get_cublas_operation(oneapi::mkl::transpose trn) { +inline cublasOperation_t get_cublas_operation(oneapi::math::transpose trn) { switch (trn) { - case oneapi::mkl::transpose::nontrans: return CUBLAS_OP_N; - case oneapi::mkl::transpose::trans: return CUBLAS_OP_T; - case oneapi::mkl::transpose::conjtrans: return CUBLAS_OP_C; + case oneapi::math::transpose::nontrans: return CUBLAS_OP_N; + case oneapi::math::transpose::trans: return CUBLAS_OP_T; + case oneapi::math::transpose::conjtrans: return CUBLAS_OP_C; default: throw "Wrong transpose Operation."; } } -inline cublasFillMode_t get_cublas_fill_mode(oneapi::mkl::uplo ul) { +inline cublasFillMode_t get_cublas_fill_mode(oneapi::math::uplo ul) { switch (ul) { - case oneapi::mkl::uplo::upper: return CUBLAS_FILL_MODE_UPPER; - case oneapi::mkl::uplo::lower: return CUBLAS_FILL_MODE_LOWER; + case oneapi::math::uplo::upper: return CUBLAS_FILL_MODE_UPPER; + case oneapi::math::uplo::lower: return CUBLAS_FILL_MODE_LOWER; default: throw "Wrong fill mode."; } } -inline cublasSideMode_t get_cublas_side_mode(oneapi::mkl::side lr) { +inline cublasSideMode_t get_cublas_side_mode(oneapi::math::side lr) { switch (lr) { - case oneapi::mkl::side::left: return CUBLAS_SIDE_LEFT; - case oneapi::mkl::side::right: return CUBLAS_SIDE_RIGHT; + case oneapi::math::side::left: return CUBLAS_SIDE_LEFT; + case oneapi::math::side::right: return CUBLAS_SIDE_RIGHT; default: throw "Wrong side mode."; } } -inline cublasSideMode_t get_cublas_generate(oneapi::mkl::generate qp) { +inline cublasSideMode_t get_cublas_generate(oneapi::math::generate qp) { switch (qp) { - case oneapi::mkl::generate::Q: return CUBLAS_SIDE_LEFT; - case oneapi::mkl::generate::P: return CUBLAS_SIDE_RIGHT; + case oneapi::math::generate::Q: return CUBLAS_SIDE_LEFT; + case oneapi::math::generate::P: return CUBLAS_SIDE_RIGHT; default: throw "Wrong generate."; } } @@ -311,7 +311,7 @@ inline void lapack_info_check(sycl::queue &queue, DEVINFO_T devinfo, const char get_cusolver_devinfo(queue, devinfo, dev_info_); for (const auto &val : dev_info_) { if (val > 0) - throw oneapi::mkl::lapack::computation_error( + throw oneapi::math::lapack::computation_error( func_name, std::string(cufunc_name) + " failed with info = " + std::to_string(val), val); } @@ -332,6 +332,6 @@ T **create_ptr_list_from_stride(T *ptr, int64_t ptr_stride, int64_t batch_size) } // namespace cusolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _CUSOLVER_HELPER_HPP_ diff --git a/src/lapack/backends/cusolver/cusolver_lapack.cpp b/src/lapack/backends/cusolver/cusolver_lapack.cpp index d9624cdaf..52d28edc3 100644 --- a/src/lapack/backends/cusolver/cusolver_lapack.cpp +++ b/src/lapack/backends/cusolver/cusolver_lapack.cpp @@ -23,7 +23,7 @@ #include "oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace cusolver { @@ -212,7 +212,7 @@ GETRI_LAUNCHER(std::complex) // cusolverDnXgetrs does not use scratchpad memory template inline void getrs(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -250,7 +250,7 @@ inline void getrs(const char *func_name, Func func, sycl::queue &queue, } #define GETRS_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, \ + void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, \ sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ @@ -266,8 +266,8 @@ GETRS_LAUNCHER(std::complex, cusolverDnZgetrs) #undef GETRS_LAUNCHER template -inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, +inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -301,7 +301,7 @@ inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define GESVD_LAUNCHER(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, \ + void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, \ sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, \ @@ -318,8 +318,8 @@ GESVD_LAUNCHER(std::complex, double, cusolverDnZgesvd) #undef GESVD_LAUNCHER template -inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, +inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using cuDataType_A = typename CudaEquivalentType::Type; @@ -347,7 +347,7 @@ inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define HEEVD_LAUNCHER(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, \ + void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &w, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ heevd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, scratchpad, \ @@ -361,7 +361,7 @@ HEEVD_LAUNCHER(std::complex, double, cusolverDnZheevd) template inline void hegvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -392,8 +392,8 @@ inline void hegvd(const char *func_name, Func func, sycl::queue &queue, std::int } #define HEGVD_LAUNCHER(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ + void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ hegvd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, ldb, w, \ @@ -406,7 +406,7 @@ HEGVD_LAUNCHER(std::complex, double, cusolverDnZhegvd) #undef HEGVD_LAUNCHER template -inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -438,7 +438,7 @@ inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define HETRD_LAUNCHER(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, \ + void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &d, \ sycl::buffer &e, sycl::buffer &tau, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ @@ -451,12 +451,12 @@ HETRD_LAUNCHER(std::complex, double, cusolverDnZhetrd) #undef HETRD_LAUNCHER -void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "hetrf"); } -void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { @@ -464,7 +464,7 @@ void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, } template -inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::generate vec, +inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -487,7 +487,7 @@ inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORGBR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, \ + void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ orgbr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, scratchpad, \ @@ -535,7 +535,7 @@ ORGQR_LAUNCHER(double, cusolverDnDorgqr) #undef ORGQR_LAUNCHER template -inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -557,7 +557,7 @@ inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORGTR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ orgtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ @@ -570,8 +570,8 @@ ORGTR_LAUNCHER(double, cusolverDnDorgtr) #undef ORGTR_LAUNCHER template -inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -598,8 +598,8 @@ inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORMTR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, \ - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, \ + void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ @@ -612,13 +612,13 @@ ORMTR_LAUNCHER(double, cusolverDnDormtr) #undef ORMTR_LAUNCHER -void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "ormrq"); } -void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -626,8 +626,8 @@ void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose tr } template -inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -652,7 +652,7 @@ inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORMQR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, \ + void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ @@ -666,7 +666,7 @@ ORMQR_LAUNCHER(double, cusolverDnDormqr) #undef ORMQR_LAUNCHER template -inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -690,7 +690,7 @@ inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define POTRF_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ potrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ scratchpad_size); \ @@ -704,7 +704,7 @@ POTRF_LAUNCHER(std::complex, cusolverDnZpotrf) #undef POTRF_LAUNCHER template -inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -728,7 +728,7 @@ inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define POTRI_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ potri(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ scratchpad_size); \ @@ -743,7 +743,7 @@ POTRI_LAUNCHER(std::complex, cusolverDnZpotri) // cusolverDnXpotrs does not use scratchpad memory template -inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -764,7 +764,7 @@ inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define POTRS_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, \ + void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ potrs(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, nrhs, a, lda, b, ldb, \ @@ -779,8 +779,8 @@ POTRS_LAUNCHER(std::complex, cusolverDnZpotrs) #undef POTRS_LAUNCHER template -inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, +inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); @@ -806,7 +806,7 @@ inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define SYEVD_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, \ + void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &w, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ syevd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, scratchpad, \ @@ -820,7 +820,7 @@ SYEVD_LAUNCHER(double, cusolverDnDsyevd) template inline void sygvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -849,8 +849,8 @@ inline void sygvd(const char *func_name, Func func, sycl::queue &queue, std::int } #define SYGVD_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ + void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ sygvd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, ldb, w, \ @@ -863,7 +863,7 @@ SYGVD_LAUNCHER(double, cusolverDnDsygvd) #undef SYGVD_LAUNCH template -inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -894,7 +894,7 @@ inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define SYTRD_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &d, sycl::buffer &e, \ sycl::buffer &tau, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ @@ -908,7 +908,7 @@ SYTRD_LAUNCHER(double, cusolverDnDsytrd) #undef SYTRD_LAUNCHER template -inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -951,7 +951,7 @@ inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define SYTRF_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ sytrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, ipiv, scratchpad, \ @@ -965,27 +965,27 @@ SYTRF_LAUNCHER(std::complex, cusolverDnZsytrf) #undef SYTRF_LAUNCHER -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { @@ -993,7 +993,7 @@ void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose tr } template -inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::generate vec, +inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -1016,7 +1016,7 @@ inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNGBR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, \ + void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ ungbr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, scratchpad, \ @@ -1064,7 +1064,7 @@ UNGQR_LAUNCHER(std::complex, cusolverDnZungqr) #undef UNGQR_LAUNCHER template -inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -1086,7 +1086,7 @@ inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNGTR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ ungtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ @@ -1098,14 +1098,14 @@ UNGTR_LAUNCHER(std::complex, cusolverDnZungtr) #undef UNGTR_LAUNCHER -void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "unmrq"); } -void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, @@ -1114,8 +1114,8 @@ void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose tr } template -inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -1140,7 +1140,7 @@ inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNMQR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, \ + void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ @@ -1154,8 +1154,8 @@ UNMQR_LAUNCHER(std::complex, cusolverDnZunmqr) #undef UNMQR_LAUNCHER template -inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -1182,8 +1182,8 @@ inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNMTR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, \ - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, \ + void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ @@ -1391,7 +1391,7 @@ GETRI_LAUNCHER_USM(std::complex) // cusolverDnXgetrs does not use scratchpad memory template inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, T *a, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, T *a, std::int64_t lda, std::int64_t *ipiv, T *b, std::int64_t ldb, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1435,7 +1435,7 @@ inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, } #define GETRS_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, \ + sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ std::int64_t nrhs, TYPE *a, std::int64_t lda, std::int64_t *ipiv, TYPE *b, \ std::int64_t ldb, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1452,7 +1452,7 @@ GETRS_LAUNCHER_USM(std::complex, cusolverDnZgetrs) template inline sycl::event gesvd(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, T_A *a, std::int64_t lda, T_B *s, T_A *u, std::int64_t ldu, T_A *vt, std::int64_t ldvt, T_A *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1486,7 +1486,7 @@ inline sycl::event gesvd(const char *func_name, Func func, sycl::queue &queue, } #define GESVD_LAUNCHER_USM(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, \ + sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ std::int64_t m, std::int64_t n, TYPE_A *a, std::int64_t lda, TYPE_B *s, \ TYPE_A *u, std::int64_t ldu, TYPE_A *vt, std::int64_t ldvt, \ TYPE_A *scratchpad, std::int64_t scratchpad_size, \ @@ -1504,7 +1504,7 @@ GESVD_LAUNCHER_USM(std::complex, double, cusolverDnZgesvd) template inline sycl::event heevd(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, T_A *&a, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A *&a, std::int64_t lda, T_B *&w, T_A *&scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using cuDataType_A = typename CudaEquivalentType::Type; @@ -1534,7 +1534,7 @@ inline sycl::event heevd(const char *func_name, Func func, sycl::queue &queue, } #define HEEVD_LAUNCHER_USM(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, \ + sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, TYPE_A *a, std::int64_t lda, TYPE_B *w, TYPE_A *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1549,7 +1549,7 @@ HEEVD_LAUNCHER_USM(std::complex, double, cusolverDnZheevd) template inline sycl::event hegvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, T_A *&a, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A *&a, std::int64_t lda, T_A *&b, std::int64_t ldb, T_B *&w, T_A *&scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1581,8 +1581,8 @@ inline sycl::event hegvd(const char *func_name, Func func, sycl::queue &queue, s } #define HEGVD_LAUNCHER_USM(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, TYPE_A *a, std::int64_t lda, \ + sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, TYPE_A *a, std::int64_t lda, \ TYPE_A *b, std::int64_t ldb, TYPE_B *w, TYPE_A *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1597,7 +1597,7 @@ HEGVD_LAUNCHER_USM(std::complex, double, cusolverDnZhegvd) template inline sycl::event hetrd(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T_A *a, std::int64_t lda, T_B *d, + oneapi::math::uplo uplo, std::int64_t n, T_A *a, std::int64_t lda, T_B *d, T_B *e, T_A *tau, T_A *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using cuDataType_A = typename CudaEquivalentType::Type; @@ -1628,7 +1628,7 @@ inline sycl::event hetrd(const char *func_name, Func func, sycl::queue &queue, } #define HETRD_LAUNCHER_USM(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE_A *a, \ + sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE_A *a, \ std::int64_t lda, TYPE_B *d, TYPE_B *e, TYPE_A *tau, TYPE_A *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1641,13 +1641,13 @@ HETRD_LAUNCHER_USM(std::complex, double, cusolverDnZhetrd) #undef HETRD_LAUNCHER_USM -sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "hetrf"); } -sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1656,7 +1656,7 @@ sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, template inline sycl::event orgbr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1681,7 +1681,7 @@ inline sycl::event orgbr(const char *func_name, Func func, sycl::queue &queue, } #define ORGBR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, \ + sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, \ std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, TYPE *tau, \ TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1734,7 +1734,7 @@ ORGQR_LAUNCHER_USM(double, cusolverDnDorgqr) template inline sycl::event orgtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using cuDataType = typename CudaEquivalentType::Type; @@ -1758,7 +1758,7 @@ inline sycl::event orgtr(const char *func_name, Func func, sycl::queue &queue, } #define ORGTR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ return orgtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ @@ -1772,8 +1772,8 @@ ORGTR_LAUNCHER_USM(double, cusolverDnDorgtr) template inline sycl::event ormtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, T *a, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1801,8 +1801,8 @@ inline sycl::event ormtr(const char *func_name, Func func, sycl::queue &queue, } #define ORMTR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, \ - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ + sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1815,13 +1815,13 @@ ORMTR_LAUNCHER_USM(double, cusolverDnDormtr) #undef ORMTR_LAUNCHER_USM -sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "ormrq"); } -sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1830,7 +1830,7 @@ sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::trans template inline sycl::event ormqr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1857,7 +1857,7 @@ inline sycl::event ormqr(const char *func_name, Func func, sycl::queue &queue, } #define ORMQR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, \ + sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, \ TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ @@ -1873,7 +1873,7 @@ ORMQR_LAUNCHER_USM(double, cusolverDnDormqr) template inline sycl::event potrf(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using cuDataType = typename CudaEquivalentType::Type; @@ -1900,7 +1900,7 @@ inline sycl::event potrf(const char *func_name, Func func, sycl::queue &queue, } #define POTRF_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ return potrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ @@ -1916,7 +1916,7 @@ POTRF_LAUNCHER_USM(std::complex, cusolverDnZpotrf) template inline sycl::event potri(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using cuDataType = typename CudaEquivalentType::Type; @@ -1943,7 +1943,7 @@ inline sycl::event potri(const char *func_name, Func func, sycl::queue &queue, } #define POTRI_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ return potri(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ @@ -1960,7 +1960,7 @@ POTRI_LAUNCHER_USM(std::complex, cusolverDnZpotri) // cusolverDnXpotrs does not use scratchpad memory template inline sycl::event potrs(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, T *a, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, T *a, std::int64_t lda, T *b, std::int64_t ldb, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1984,7 +1984,7 @@ inline sycl::event potrs(const char *func_name, Func func, sycl::queue &queue, } #define POTRS_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, \ + sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ std::int64_t nrhs, TYPE *a, std::int64_t lda, TYPE *b, std::int64_t ldb, \ TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -2001,7 +2001,7 @@ POTRS_LAUNCHER_USM(std::complex, cusolverDnZpotrs) template inline sycl::event syevd(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, T *a, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *w, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using cuDataType = typename CudaEquivalentType::Type; @@ -2030,7 +2030,7 @@ inline sycl::event syevd(const char *func_name, Func func, sycl::queue &queue, } #define SYEVD_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, \ + sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, TYPE *a, std::int64_t lda, TYPE *w, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -2045,7 +2045,7 @@ SYEVD_LAUNCHER_USM(double, cusolverDnDsyevd) template inline sycl::event sygvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, T *a, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *b, std::int64_t ldb, T *w, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -2076,8 +2076,8 @@ inline sycl::event sygvd(const char *func_name, Func func, sycl::queue &queue, s } #define SYGVD_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, std::int64_t lda, TYPE *b, \ + sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, TYPE *a, std::int64_t lda, TYPE *b, \ std::int64_t ldb, TYPE *w, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ return sygvd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, \ @@ -2091,7 +2091,7 @@ SYGVD_LAUNCHER_USM(double, cusolverDnDsygvd) template inline sycl::event sytrd(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *d, T *e, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *d, T *e, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using cuDataType = typename CudaEquivalentType::Type; @@ -2121,7 +2121,7 @@ inline sycl::event sytrd(const char *func_name, Func func, sycl::queue &queue, } #define SYTRD_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *d, TYPE *e, TYPE *tau, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -2136,7 +2136,7 @@ SYTRD_LAUNCHER_USM(double, cusolverDnDsytrd) template inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, std::int64_t *ipiv, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using cuDataType = typename CudaEquivalentType::Type; @@ -2184,7 +2184,7 @@ inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, } #define SYTRF_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, std::int64_t *ipiv, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -2199,27 +2199,27 @@ SYTRF_LAUNCHER_USM(std::complex, cusolverDnZsytrf) #undef SYTRF_LAUNCHER_USM -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, double *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -2228,7 +2228,7 @@ sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::trans template inline sycl::event ungbr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -2253,7 +2253,7 @@ inline sycl::event ungbr(const char *func_name, Func func, sycl::queue &queue, } #define UNGBR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, \ + sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, \ std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, TYPE *tau, \ TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -2306,7 +2306,7 @@ UNGQR_LAUNCHER_USM(std::complex, cusolverDnZungqr) template inline sycl::event ungtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using cuDataType = typename CudaEquivalentType::Type; @@ -2330,7 +2330,7 @@ inline sycl::event ungtr(const char *func_name, Func func, sycl::queue &queue, } #define UNGTR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ return ungtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ @@ -2342,14 +2342,14 @@ UNGTR_LAUNCHER_USM(std::complex, cusolverDnZungtr) #undef UNGTR_LAUNCHER_USM -sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "unmrq"); } -sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -2359,7 +2359,7 @@ sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::trans template inline sycl::event unmqr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -2386,7 +2386,7 @@ inline sycl::event unmqr(const char *func_name, Func func, sycl::queue &queue, } #define UNMQR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, \ + sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, \ TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ @@ -2402,8 +2402,8 @@ UNMQR_LAUNCHER_USM(std::complex, cusolverDnZunmqr) template inline sycl::event unmtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, T *a, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -2431,8 +2431,8 @@ inline sycl::event unmtr(const char *func_name, Func func, sycl::queue &queue, } #define UNMTR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, \ - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ + sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -2530,7 +2530,7 @@ GEQRF_LAUNCHER_SCRATCH(std::complex, cusolverDnZgeqrf_bufferSize) template inline void gesvd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { @@ -2545,7 +2545,7 @@ inline void gesvd_scratchpad_size(const char *func_name, Func func, sycl::queue #define GESVD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ std::int64_t gesvd_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, \ + sycl::queue & queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, \ std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { \ int scratch_size; \ gesvd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobu, jobvt, m, n, lda, \ @@ -2607,7 +2607,7 @@ GETRI_LAUNCHER_SCRATCH(std::complex) // cusolverDnXgetrs does not use scratchpad memory #define GETRS_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t getrs_scratchpad_size(sycl::queue & queue, oneapi::mkl::transpose trans, \ + std::int64_t getrs_scratchpad_size(sycl::queue & queue, oneapi::math::transpose trans, \ std::int64_t n, std::int64_t nrhs, std::int64_t lda, \ std::int64_t ldb) { \ return 0; \ @@ -2622,7 +2622,7 @@ GETRS_LAUNCHER_SCRATCH(std::complex) template inline void heevd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -2637,8 +2637,8 @@ inline void heevd_scratchpad_size(const char *func_name, Func func, sycl::queue #define HEEVD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t heevd_scratchpad_size(sycl::queue & queue, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, \ + std::int64_t heevd_scratchpad_size(sycl::queue & queue, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, \ std::int64_t lda) { \ int scratch_size; \ heevd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, lda, \ @@ -2653,7 +2653,7 @@ HEEVD_LAUNCHER_SCRATCH(std::complex, cusolverDnZheevd_bufferSize) template inline void hegvd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { @@ -2670,7 +2670,7 @@ inline void hegvd_scratchpad_size(const char *func_name, Func func, sycl::queue #define HEGVD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ std::int64_t hegvd_scratchpad_size(sycl::queue & queue, std::int64_t itype, \ - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, \ + oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda, std::int64_t ldb) { \ int scratch_size; \ hegvd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, \ @@ -2685,7 +2685,7 @@ HEGVD_LAUNCHER_SCRATCH(std::complex, cusolverDnZhegvd_bufferSize) template inline void hetrd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -2699,7 +2699,7 @@ inline void hetrd_scratchpad_size(const char *func_name, Func func, sycl::queue #define HETRD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t hetrd_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t hetrd_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ int scratch_size; \ hetrd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ @@ -2713,19 +2713,19 @@ HETRD_LAUNCHER_SCRATCH(std::complex, cusolverDnZhetrd_bufferSize) #undef HETRD_LAUNCHER_SCRATCH template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "hetrf_scratchpad_size"); } template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "hetrf_scratchpad_size"); } template inline void orgbr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -2739,7 +2739,7 @@ inline void orgbr_scratchpad_size(const char *func_name, Func func, sycl::queue #define ORGBR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t orgbr_scratchpad_size(sycl::queue & queue, oneapi::mkl::generate vec, \ + std::int64_t orgbr_scratchpad_size(sycl::queue & queue, oneapi::math::generate vec, \ std::int64_t m, std::int64_t n, std::int64_t k, \ std::int64_t lda) { \ int scratch_size; \ @@ -2755,7 +2755,7 @@ ORGBR_LAUNCHER_SCRATCH(double, cusolverDnDorgbr_bufferSize) template inline void orgtr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -2769,7 +2769,7 @@ inline void orgtr_scratchpad_size(const char *func_name, Func func, sycl::queue #define ORGTR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t orgtr_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t orgtr_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ int scratch_size; \ orgtr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ @@ -2812,15 +2812,15 @@ ORGQR_LAUNCHER_SCRATCH(double, cusolverDnDorgqr_bufferSize) #undef ORGQR_LAUNCHER_SCRATCH template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { throw unimplemented("lapack", "ormrq_scratchpad_size"); } template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { throw unimplemented("lapack", "ormrq_scratchpad_size"); @@ -2828,7 +2828,7 @@ std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side template inline void ormqr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { @@ -2845,7 +2845,7 @@ inline void ormqr_scratchpad_size(const char *func_name, Func func, sycl::queue #define ORMQRF_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ std::int64_t ormqr_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, \ + sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, \ std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ int scratch_size; \ ormqr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, \ @@ -2860,8 +2860,8 @@ ORMQRF_LAUNCHER_SCRATCH(double, cusolverDnDormqr_bufferSize) template inline void ormtr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -2876,8 +2876,8 @@ inline void ormtr_scratchpad_size(const char *func_name, Func func, sycl::queue #define ORMTR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t ormtr_scratchpad_size(sycl::queue & queue, oneapi::mkl::side side, \ - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, \ + std::int64_t ormtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ + oneapi::math::uplo uplo, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t lda, \ std::int64_t ldc) { \ int scratch_size; \ @@ -2893,7 +2893,7 @@ ORMTR_LAUNCHER_SCRATCH(double, cusolverDnDormtr_bufferSize) template inline void potrf_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -2907,7 +2907,7 @@ inline void potrf_scratchpad_size(const char *func_name, Func func, sycl::queue #define POTRF_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t potrf_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t potrf_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ int scratch_size; \ potrf_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ @@ -2925,7 +2925,7 @@ POTRF_LAUNCHER_SCRATCH(std::complex, cusolverDnZpotrf_bufferSize) // cusolverDnXpotrs does not use scratchpad memory #define POTRS_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t potrs_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t potrs_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t nrhs, std::int64_t lda, \ std::int64_t ldb) { \ return 0; \ @@ -2940,7 +2940,7 @@ POTRS_LAUNCHER_SCRATCH(std::complex) template inline void potri_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -2954,7 +2954,7 @@ inline void potri_scratchpad_size(const char *func_name, Func func, sycl::queue #define POTRI_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t potri_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t potri_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ int scratch_size; \ potri_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ @@ -2971,7 +2971,7 @@ POTRI_LAUNCHER_SCRATCH(std::complex, cusolverDnZpotri_bufferSize) template inline void sytrf_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -2984,7 +2984,7 @@ inline void sytrf_scratchpad_size(const char *func_name, Func func, sycl::queue #define SYTRF_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t sytrf_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t sytrf_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ int scratch_size; \ sytrf_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ @@ -3001,7 +3001,7 @@ SYTRF_LAUNCHER_SCRATCH(std::complex, cusolverDnZsytrf_bufferSize) template inline void syevd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -3016,8 +3016,8 @@ inline void syevd_scratchpad_size(const char *func_name, Func func, sycl::queue #define SYEVD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t syevd_scratchpad_size(sycl::queue & queue, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, \ + std::int64_t syevd_scratchpad_size(sycl::queue & queue, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, \ std::int64_t lda) { \ int scratch_size; \ syevd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, lda, \ @@ -3032,7 +3032,7 @@ SYEVD_LAUNCHER_SCRATCH(double, cusolverDnDsyevd_bufferSize) template inline void sygvd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { @@ -3049,7 +3049,7 @@ inline void sygvd_scratchpad_size(const char *func_name, Func func, sycl::queue #define SYGVD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ std::int64_t sygvd_scratchpad_size(sycl::queue & queue, std::int64_t itype, \ - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, \ + oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda, std::int64_t ldb) { \ int scratch_size; \ sygvd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, \ @@ -3064,7 +3064,7 @@ SYGVD_LAUNCHER_SCRATCH(double, cusolverDnDsygvd_bufferSize) template inline void sytrd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -3078,7 +3078,7 @@ inline void sytrd_scratchpad_size(const char *func_name, Func func, sycl::queue #define SYTRD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t sytrd_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t sytrd_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ int scratch_size; \ sytrd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ @@ -3092,31 +3092,31 @@ SYTRD_LAUNCHER_SCRATCH(double, cusolverDnDsytrd_bufferSize) #undef SYTRD_LAUNCHER_SCRATCH template <> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { throw unimplemented("lapack", "trtrs_scratchpad_size"); } template <> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { throw unimplemented("lapack", "trtrs_scratchpad_size"); } template <> -std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { throw unimplemented("lapack", "trtrs_scratchpad_size"); } template <> -std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { throw unimplemented("lapack", "trtrs_scratchpad_size"); @@ -3124,7 +3124,7 @@ std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, one template inline void ungbr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -3138,7 +3138,7 @@ inline void ungbr_scratchpad_size(const char *func_name, Func func, sycl::queue #define UNGBR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t ungbr_scratchpad_size(sycl::queue & queue, oneapi::mkl::generate vec, \ + std::int64_t ungbr_scratchpad_size(sycl::queue & queue, oneapi::math::generate vec, \ std::int64_t m, std::int64_t n, std::int64_t k, \ std::int64_t lda) { \ int scratch_size; \ @@ -3183,7 +3183,7 @@ UNGQR_LAUNCHER_SCRATCH(std::complex, cusolverDnZungqr_bufferSize) template inline void ungtr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -3197,7 +3197,7 @@ inline void ungtr_scratchpad_size(const char *func_name, Func func, sycl::queue #define UNGTR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t ungtr_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t ungtr_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ int scratch_size; \ ungtr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ @@ -3211,16 +3211,16 @@ UNGTR_LAUNCHER_SCRATCH(std::complex, cusolverDnZungtr_bufferSize) #undef UNGTR_LAUNCHER_SCRATCH template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { throw unimplemented("lapack", "unmrq_scratchpad_size"); } template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { @@ -3229,7 +3229,7 @@ std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, one template inline void unmqr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { @@ -3246,7 +3246,7 @@ inline void unmqr_scratchpad_size(const char *func_name, Func func, sycl::queue #define UNMQR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ std::int64_t unmqr_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, \ + sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, \ std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ int scratch_size; \ unmqr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, \ @@ -3261,8 +3261,8 @@ UNMQR_LAUNCHER_SCRATCH(std::complex, cusolverDnZunmqr_bufferSize) template inline void unmtr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc, int *scratch_size) { queue.submit([&](sycl::handler &cgh) { onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { @@ -3277,8 +3277,8 @@ inline void unmtr_scratchpad_size(const char *func_name, Func func, sycl::queue #define UNMTR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t unmtr_scratchpad_size(sycl::queue & queue, oneapi::mkl::side side, \ - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, \ + std::int64_t unmtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ + oneapi::math::uplo uplo, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t lda, \ std::int64_t ldc) { \ int scratch_size; \ @@ -3294,5 +3294,5 @@ UNMTR_LAUNCHER_SCRATCH(std::complex, cusolverDnZunmtr_bufferSize) } // namespace cusolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/lapack/backends/cusolver/cusolver_scope_handle.cpp b/src/lapack/backends/cusolver/cusolver_scope_handle.cpp index 513254851..67175ad8d 100644 --- a/src/lapack/backends/cusolver/cusolver_scope_handle.cpp +++ b/src/lapack/backends/cusolver/cusolver_scope_handle.cpp @@ -24,7 +24,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace cusolver { @@ -149,5 +149,5 @@ sycl::context CusolverScopedContextHandler::get_context(const sycl::queue &queue } // namespace cusolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/lapack/backends/cusolver/cusolver_scope_handle.hpp b/src/lapack/backends/cusolver/cusolver_scope_handle.hpp index 9ee4afa40..b552b4af2 100644 --- a/src/lapack/backends/cusolver/cusolver_scope_handle.hpp +++ b/src/lapack/backends/cusolver/cusolver_scope_handle.hpp @@ -52,7 +52,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace cusolver { @@ -127,6 +127,6 @@ class CusolverScopedContextHandler { } // namespace cusolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_CUSOLVER_SCOPED_HANDLE_HPP_ diff --git a/src/lapack/backends/cusolver/cusolver_task.hpp b/src/lapack/backends/cusolver/cusolver_task.hpp index 7059b5063..da008897e 100644 --- a/src/lapack/backends/cusolver/cusolver_task.hpp +++ b/src/lapack/backends/cusolver/cusolver_task.hpp @@ -44,7 +44,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace cusolver { @@ -67,6 +67,6 @@ static inline void onemath_cusolver_host_task(H &cgh, sycl::queue queue, F f) { } // namespace cusolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _MKL_LAPACK_CUSOLVER_TASK_HPP_ diff --git a/src/lapack/backends/cusolver/cusolver_wrappers.cpp b/src/lapack/backends/cusolver/cusolver_wrappers.cpp index 34dd68da2..52a4eec26 100644 --- a/src/lapack/backends/cusolver/cusolver_wrappers.cpp +++ b/src/lapack/backends/cusolver/cusolver_wrappers.cpp @@ -24,403 +24,403 @@ extern "C" ONEMATH_EXPORT lapack_function_table_t mkl_lapack_table = { WRAPPER_VERSION, #define LAPACK_BACKEND cusolver - oneapi::mkl::lapack::cusolver::gebrd, - oneapi::mkl::lapack::cusolver::gebrd, - oneapi::mkl::lapack::cusolver::gebrd, - oneapi::mkl::lapack::cusolver::gebrd, - oneapi::mkl::lapack::cusolver::gerqf, - oneapi::mkl::lapack::cusolver::gerqf, - oneapi::mkl::lapack::cusolver::gerqf, - oneapi::mkl::lapack::cusolver::gerqf, - oneapi::mkl::lapack::cusolver::geqrf, - oneapi::mkl::lapack::cusolver::geqrf, - oneapi::mkl::lapack::cusolver::geqrf, - oneapi::mkl::lapack::cusolver::geqrf, - oneapi::mkl::lapack::cusolver::getrf, - oneapi::mkl::lapack::cusolver::getrf, - oneapi::mkl::lapack::cusolver::getrf, - oneapi::mkl::lapack::cusolver::getrf, - oneapi::mkl::lapack::cusolver::getri, - oneapi::mkl::lapack::cusolver::getri, - oneapi::mkl::lapack::cusolver::getri, - oneapi::mkl::lapack::cusolver::getri, - oneapi::mkl::lapack::cusolver::getrs, - oneapi::mkl::lapack::cusolver::getrs, - oneapi::mkl::lapack::cusolver::getrs, - oneapi::mkl::lapack::cusolver::getrs, - oneapi::mkl::lapack::cusolver::gesvd, - oneapi::mkl::lapack::cusolver::gesvd, - oneapi::mkl::lapack::cusolver::gesvd, - oneapi::mkl::lapack::cusolver::gesvd, - oneapi::mkl::lapack::cusolver::heevd, - oneapi::mkl::lapack::cusolver::heevd, - oneapi::mkl::lapack::cusolver::hegvd, - oneapi::mkl::lapack::cusolver::hegvd, - oneapi::mkl::lapack::cusolver::hetrd, - oneapi::mkl::lapack::cusolver::hetrd, - oneapi::mkl::lapack::cusolver::hetrf, - oneapi::mkl::lapack::cusolver::hetrf, - oneapi::mkl::lapack::cusolver::orgbr, - oneapi::mkl::lapack::cusolver::orgbr, - oneapi::mkl::lapack::cusolver::orgqr, - oneapi::mkl::lapack::cusolver::orgqr, - oneapi::mkl::lapack::cusolver::orgtr, - oneapi::mkl::lapack::cusolver::orgtr, - oneapi::mkl::lapack::cusolver::ormtr, - oneapi::mkl::lapack::cusolver::ormtr, - oneapi::mkl::lapack::cusolver::ormrq, - oneapi::mkl::lapack::cusolver::ormrq, - oneapi::mkl::lapack::cusolver::ormqr, - oneapi::mkl::lapack::cusolver::ormqr, - oneapi::mkl::lapack::cusolver::potrf, - oneapi::mkl::lapack::cusolver::potrf, - oneapi::mkl::lapack::cusolver::potrf, - oneapi::mkl::lapack::cusolver::potrf, - oneapi::mkl::lapack::cusolver::potri, - oneapi::mkl::lapack::cusolver::potri, - oneapi::mkl::lapack::cusolver::potri, - oneapi::mkl::lapack::cusolver::potri, - oneapi::mkl::lapack::cusolver::potrs, - oneapi::mkl::lapack::cusolver::potrs, - oneapi::mkl::lapack::cusolver::potrs, - oneapi::mkl::lapack::cusolver::potrs, - oneapi::mkl::lapack::cusolver::syevd, - oneapi::mkl::lapack::cusolver::syevd, - oneapi::mkl::lapack::cusolver::sygvd, - oneapi::mkl::lapack::cusolver::sygvd, - oneapi::mkl::lapack::cusolver::sytrd, - oneapi::mkl::lapack::cusolver::sytrd, - oneapi::mkl::lapack::cusolver::sytrf, - oneapi::mkl::lapack::cusolver::sytrf, - oneapi::mkl::lapack::cusolver::sytrf, - oneapi::mkl::lapack::cusolver::sytrf, - oneapi::mkl::lapack::cusolver::trtrs, - oneapi::mkl::lapack::cusolver::trtrs, - oneapi::mkl::lapack::cusolver::trtrs, - oneapi::mkl::lapack::cusolver::trtrs, - oneapi::mkl::lapack::cusolver::ungbr, - oneapi::mkl::lapack::cusolver::ungbr, - oneapi::mkl::lapack::cusolver::ungqr, - oneapi::mkl::lapack::cusolver::ungqr, - oneapi::mkl::lapack::cusolver::ungtr, - oneapi::mkl::lapack::cusolver::ungtr, - oneapi::mkl::lapack::cusolver::unmrq, - oneapi::mkl::lapack::cusolver::unmrq, - oneapi::mkl::lapack::cusolver::unmqr, - oneapi::mkl::lapack::cusolver::unmqr, - oneapi::mkl::lapack::cusolver::unmtr, - oneapi::mkl::lapack::cusolver::unmtr, - oneapi::mkl::lapack::cusolver::gebrd, - oneapi::mkl::lapack::cusolver::gebrd, - oneapi::mkl::lapack::cusolver::gebrd, - oneapi::mkl::lapack::cusolver::gebrd, - oneapi::mkl::lapack::cusolver::gerqf, - oneapi::mkl::lapack::cusolver::gerqf, - oneapi::mkl::lapack::cusolver::gerqf, - oneapi::mkl::lapack::cusolver::gerqf, - oneapi::mkl::lapack::cusolver::geqrf, - oneapi::mkl::lapack::cusolver::geqrf, - oneapi::mkl::lapack::cusolver::geqrf, - oneapi::mkl::lapack::cusolver::geqrf, - oneapi::mkl::lapack::cusolver::getrf, - oneapi::mkl::lapack::cusolver::getrf, - oneapi::mkl::lapack::cusolver::getrf, - oneapi::mkl::lapack::cusolver::getrf, - oneapi::mkl::lapack::cusolver::getri, - oneapi::mkl::lapack::cusolver::getri, - oneapi::mkl::lapack::cusolver::getri, - oneapi::mkl::lapack::cusolver::getri, - oneapi::mkl::lapack::cusolver::getrs, - oneapi::mkl::lapack::cusolver::getrs, - oneapi::mkl::lapack::cusolver::getrs, - oneapi::mkl::lapack::cusolver::getrs, - oneapi::mkl::lapack::cusolver::gesvd, - oneapi::mkl::lapack::cusolver::gesvd, - oneapi::mkl::lapack::cusolver::gesvd, - oneapi::mkl::lapack::cusolver::gesvd, - oneapi::mkl::lapack::cusolver::heevd, - oneapi::mkl::lapack::cusolver::heevd, - oneapi::mkl::lapack::cusolver::hegvd, - oneapi::mkl::lapack::cusolver::hegvd, - oneapi::mkl::lapack::cusolver::hetrd, - oneapi::mkl::lapack::cusolver::hetrd, - oneapi::mkl::lapack::cusolver::hetrf, - oneapi::mkl::lapack::cusolver::hetrf, - oneapi::mkl::lapack::cusolver::orgbr, - oneapi::mkl::lapack::cusolver::orgbr, - oneapi::mkl::lapack::cusolver::orgqr, - oneapi::mkl::lapack::cusolver::orgqr, - oneapi::mkl::lapack::cusolver::orgtr, - oneapi::mkl::lapack::cusolver::orgtr, - oneapi::mkl::lapack::cusolver::ormtr, - oneapi::mkl::lapack::cusolver::ormtr, - oneapi::mkl::lapack::cusolver::ormrq, - oneapi::mkl::lapack::cusolver::ormrq, - oneapi::mkl::lapack::cusolver::ormqr, - oneapi::mkl::lapack::cusolver::ormqr, - oneapi::mkl::lapack::cusolver::potrf, - oneapi::mkl::lapack::cusolver::potrf, - oneapi::mkl::lapack::cusolver::potrf, - oneapi::mkl::lapack::cusolver::potrf, - oneapi::mkl::lapack::cusolver::potri, - oneapi::mkl::lapack::cusolver::potri, - oneapi::mkl::lapack::cusolver::potri, - oneapi::mkl::lapack::cusolver::potri, - oneapi::mkl::lapack::cusolver::potrs, - oneapi::mkl::lapack::cusolver::potrs, - oneapi::mkl::lapack::cusolver::potrs, - oneapi::mkl::lapack::cusolver::potrs, - oneapi::mkl::lapack::cusolver::syevd, - oneapi::mkl::lapack::cusolver::syevd, - oneapi::mkl::lapack::cusolver::sygvd, - oneapi::mkl::lapack::cusolver::sygvd, - oneapi::mkl::lapack::cusolver::sytrd, - oneapi::mkl::lapack::cusolver::sytrd, - oneapi::mkl::lapack::cusolver::sytrf, - oneapi::mkl::lapack::cusolver::sytrf, - oneapi::mkl::lapack::cusolver::sytrf, - oneapi::mkl::lapack::cusolver::sytrf, - oneapi::mkl::lapack::cusolver::trtrs, - oneapi::mkl::lapack::cusolver::trtrs, - oneapi::mkl::lapack::cusolver::trtrs, - oneapi::mkl::lapack::cusolver::trtrs, - oneapi::mkl::lapack::cusolver::ungbr, - oneapi::mkl::lapack::cusolver::ungbr, - oneapi::mkl::lapack::cusolver::ungqr, - oneapi::mkl::lapack::cusolver::ungqr, - oneapi::mkl::lapack::cusolver::ungtr, - oneapi::mkl::lapack::cusolver::ungtr, - oneapi::mkl::lapack::cusolver::unmrq, - oneapi::mkl::lapack::cusolver::unmrq, - oneapi::mkl::lapack::cusolver::unmqr, - oneapi::mkl::lapack::cusolver::unmqr, - oneapi::mkl::lapack::cusolver::unmtr, - oneapi::mkl::lapack::cusolver::unmtr, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::orgqr_batch, - oneapi::mkl::lapack::cusolver::orgqr_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::ungqr_batch, - oneapi::mkl::lapack::cusolver::ungqr_batch, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::orgqr_batch, - oneapi::mkl::lapack::cusolver::orgqr_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::ungqr_batch, - oneapi::mkl::lapack::cusolver::ungqr_batch, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::geqrf_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::getrf_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getri_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::getrs_batch, - oneapi::mkl::lapack::cusolver::orgqr_batch, - oneapi::mkl::lapack::cusolver::orgqr_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrf_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::potrs_batch, - oneapi::mkl::lapack::cusolver::ungqr_batch, - oneapi::mkl::lapack::cusolver::ungqr_batch, - oneapi::mkl::lapack::cusolver::gebrd_scratchpad_size, - oneapi::mkl::lapack::cusolver::gebrd_scratchpad_size, - oneapi::mkl::lapack::cusolver::gebrd_scratchpad_size>, - oneapi::mkl::lapack::cusolver::gebrd_scratchpad_size>, - oneapi::mkl::lapack::cusolver::gerqf_scratchpad_size, - oneapi::mkl::lapack::cusolver::gerqf_scratchpad_size, - oneapi::mkl::lapack::cusolver::gerqf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::gerqf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::geqrf_scratchpad_size, - oneapi::mkl::lapack::cusolver::geqrf_scratchpad_size, - oneapi::mkl::lapack::cusolver::geqrf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::geqrf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::gesvd_scratchpad_size, - oneapi::mkl::lapack::cusolver::gesvd_scratchpad_size, - oneapi::mkl::lapack::cusolver::gesvd_scratchpad_size>, - oneapi::mkl::lapack::cusolver::gesvd_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrf_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrf_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getri_scratchpad_size, - oneapi::mkl::lapack::cusolver::getri_scratchpad_size, - oneapi::mkl::lapack::cusolver::getri_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getri_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrs_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrs_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrs_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrs_scratchpad_size>, - oneapi::mkl::lapack::cusolver::heevd_scratchpad_size>, - oneapi::mkl::lapack::cusolver::heevd_scratchpad_size>, - oneapi::mkl::lapack::cusolver::hegvd_scratchpad_size>, - oneapi::mkl::lapack::cusolver::hegvd_scratchpad_size>, - oneapi::mkl::lapack::cusolver::hetrd_scratchpad_size>, - oneapi::mkl::lapack::cusolver::hetrd_scratchpad_size>, - oneapi::mkl::lapack::cusolver::hetrf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::hetrf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::orgbr_scratchpad_size, - oneapi::mkl::lapack::cusolver::orgbr_scratchpad_size, - oneapi::mkl::lapack::cusolver::orgtr_scratchpad_size, - oneapi::mkl::lapack::cusolver::orgtr_scratchpad_size, - oneapi::mkl::lapack::cusolver::orgqr_scratchpad_size, - oneapi::mkl::lapack::cusolver::orgqr_scratchpad_size, - oneapi::mkl::lapack::cusolver::ormrq_scratchpad_size, - oneapi::mkl::lapack::cusolver::ormrq_scratchpad_size, - oneapi::mkl::lapack::cusolver::ormqr_scratchpad_size, - oneapi::mkl::lapack::cusolver::ormqr_scratchpad_size, - oneapi::mkl::lapack::cusolver::ormtr_scratchpad_size, - oneapi::mkl::lapack::cusolver::ormtr_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrf_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrf_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potrf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potrs_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrs_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrs_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potrs_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potri_scratchpad_size, - oneapi::mkl::lapack::cusolver::potri_scratchpad_size, - oneapi::mkl::lapack::cusolver::potri_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potri_scratchpad_size>, - oneapi::mkl::lapack::cusolver::sytrf_scratchpad_size, - oneapi::mkl::lapack::cusolver::sytrf_scratchpad_size, - oneapi::mkl::lapack::cusolver::sytrf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::sytrf_scratchpad_size>, - oneapi::mkl::lapack::cusolver::syevd_scratchpad_size, - oneapi::mkl::lapack::cusolver::syevd_scratchpad_size, - oneapi::mkl::lapack::cusolver::sygvd_scratchpad_size, - oneapi::mkl::lapack::cusolver::sygvd_scratchpad_size, - oneapi::mkl::lapack::cusolver::sytrd_scratchpad_size, - oneapi::mkl::lapack::cusolver::sytrd_scratchpad_size, - oneapi::mkl::lapack::cusolver::trtrs_scratchpad_size, - oneapi::mkl::lapack::cusolver::trtrs_scratchpad_size, - oneapi::mkl::lapack::cusolver::trtrs_scratchpad_size>, - oneapi::mkl::lapack::cusolver::trtrs_scratchpad_size>, - oneapi::mkl::lapack::cusolver::ungbr_scratchpad_size>, - oneapi::mkl::lapack::cusolver::ungbr_scratchpad_size>, - oneapi::mkl::lapack::cusolver::ungqr_scratchpad_size>, - oneapi::mkl::lapack::cusolver::ungqr_scratchpad_size>, - oneapi::mkl::lapack::cusolver::ungtr_scratchpad_size>, - oneapi::mkl::lapack::cusolver::ungtr_scratchpad_size>, - oneapi::mkl::lapack::cusolver::unmrq_scratchpad_size>, - oneapi::mkl::lapack::cusolver::unmrq_scratchpad_size>, - oneapi::mkl::lapack::cusolver::unmqr_scratchpad_size>, - oneapi::mkl::lapack::cusolver::unmqr_scratchpad_size>, - oneapi::mkl::lapack::cusolver::unmtr_scratchpad_size>, - oneapi::mkl::lapack::cusolver::unmtr_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getri_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getri_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::ungqr_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::ungqr_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getri_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getri_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::cusolver::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::ungqr_batch_scratchpad_size>, - oneapi::mkl::lapack::cusolver::ungqr_batch_scratchpad_size> + oneapi::math::lapack::cusolver::gebrd, + oneapi::math::lapack::cusolver::gebrd, + oneapi::math::lapack::cusolver::gebrd, + oneapi::math::lapack::cusolver::gebrd, + oneapi::math::lapack::cusolver::gerqf, + oneapi::math::lapack::cusolver::gerqf, + oneapi::math::lapack::cusolver::gerqf, + oneapi::math::lapack::cusolver::gerqf, + oneapi::math::lapack::cusolver::geqrf, + oneapi::math::lapack::cusolver::geqrf, + oneapi::math::lapack::cusolver::geqrf, + oneapi::math::lapack::cusolver::geqrf, + oneapi::math::lapack::cusolver::getrf, + oneapi::math::lapack::cusolver::getrf, + oneapi::math::lapack::cusolver::getrf, + oneapi::math::lapack::cusolver::getrf, + oneapi::math::lapack::cusolver::getri, + oneapi::math::lapack::cusolver::getri, + oneapi::math::lapack::cusolver::getri, + oneapi::math::lapack::cusolver::getri, + oneapi::math::lapack::cusolver::getrs, + oneapi::math::lapack::cusolver::getrs, + oneapi::math::lapack::cusolver::getrs, + oneapi::math::lapack::cusolver::getrs, + oneapi::math::lapack::cusolver::gesvd, + oneapi::math::lapack::cusolver::gesvd, + oneapi::math::lapack::cusolver::gesvd, + oneapi::math::lapack::cusolver::gesvd, + oneapi::math::lapack::cusolver::heevd, + oneapi::math::lapack::cusolver::heevd, + oneapi::math::lapack::cusolver::hegvd, + oneapi::math::lapack::cusolver::hegvd, + oneapi::math::lapack::cusolver::hetrd, + oneapi::math::lapack::cusolver::hetrd, + oneapi::math::lapack::cusolver::hetrf, + oneapi::math::lapack::cusolver::hetrf, + oneapi::math::lapack::cusolver::orgbr, + oneapi::math::lapack::cusolver::orgbr, + oneapi::math::lapack::cusolver::orgqr, + oneapi::math::lapack::cusolver::orgqr, + oneapi::math::lapack::cusolver::orgtr, + oneapi::math::lapack::cusolver::orgtr, + oneapi::math::lapack::cusolver::ormtr, + oneapi::math::lapack::cusolver::ormtr, + oneapi::math::lapack::cusolver::ormrq, + oneapi::math::lapack::cusolver::ormrq, + oneapi::math::lapack::cusolver::ormqr, + oneapi::math::lapack::cusolver::ormqr, + oneapi::math::lapack::cusolver::potrf, + oneapi::math::lapack::cusolver::potrf, + oneapi::math::lapack::cusolver::potrf, + oneapi::math::lapack::cusolver::potrf, + oneapi::math::lapack::cusolver::potri, + oneapi::math::lapack::cusolver::potri, + oneapi::math::lapack::cusolver::potri, + oneapi::math::lapack::cusolver::potri, + oneapi::math::lapack::cusolver::potrs, + oneapi::math::lapack::cusolver::potrs, + oneapi::math::lapack::cusolver::potrs, + oneapi::math::lapack::cusolver::potrs, + oneapi::math::lapack::cusolver::syevd, + oneapi::math::lapack::cusolver::syevd, + oneapi::math::lapack::cusolver::sygvd, + oneapi::math::lapack::cusolver::sygvd, + oneapi::math::lapack::cusolver::sytrd, + oneapi::math::lapack::cusolver::sytrd, + oneapi::math::lapack::cusolver::sytrf, + oneapi::math::lapack::cusolver::sytrf, + oneapi::math::lapack::cusolver::sytrf, + oneapi::math::lapack::cusolver::sytrf, + oneapi::math::lapack::cusolver::trtrs, + oneapi::math::lapack::cusolver::trtrs, + oneapi::math::lapack::cusolver::trtrs, + oneapi::math::lapack::cusolver::trtrs, + oneapi::math::lapack::cusolver::ungbr, + oneapi::math::lapack::cusolver::ungbr, + oneapi::math::lapack::cusolver::ungqr, + oneapi::math::lapack::cusolver::ungqr, + oneapi::math::lapack::cusolver::ungtr, + oneapi::math::lapack::cusolver::ungtr, + oneapi::math::lapack::cusolver::unmrq, + oneapi::math::lapack::cusolver::unmrq, + oneapi::math::lapack::cusolver::unmqr, + oneapi::math::lapack::cusolver::unmqr, + oneapi::math::lapack::cusolver::unmtr, + oneapi::math::lapack::cusolver::unmtr, + oneapi::math::lapack::cusolver::gebrd, + oneapi::math::lapack::cusolver::gebrd, + oneapi::math::lapack::cusolver::gebrd, + oneapi::math::lapack::cusolver::gebrd, + oneapi::math::lapack::cusolver::gerqf, + oneapi::math::lapack::cusolver::gerqf, + oneapi::math::lapack::cusolver::gerqf, + oneapi::math::lapack::cusolver::gerqf, + oneapi::math::lapack::cusolver::geqrf, + oneapi::math::lapack::cusolver::geqrf, + oneapi::math::lapack::cusolver::geqrf, + oneapi::math::lapack::cusolver::geqrf, + oneapi::math::lapack::cusolver::getrf, + oneapi::math::lapack::cusolver::getrf, + oneapi::math::lapack::cusolver::getrf, + oneapi::math::lapack::cusolver::getrf, + oneapi::math::lapack::cusolver::getri, + oneapi::math::lapack::cusolver::getri, + oneapi::math::lapack::cusolver::getri, + oneapi::math::lapack::cusolver::getri, + oneapi::math::lapack::cusolver::getrs, + oneapi::math::lapack::cusolver::getrs, + oneapi::math::lapack::cusolver::getrs, + oneapi::math::lapack::cusolver::getrs, + oneapi::math::lapack::cusolver::gesvd, + oneapi::math::lapack::cusolver::gesvd, + oneapi::math::lapack::cusolver::gesvd, + oneapi::math::lapack::cusolver::gesvd, + oneapi::math::lapack::cusolver::heevd, + oneapi::math::lapack::cusolver::heevd, + oneapi::math::lapack::cusolver::hegvd, + oneapi::math::lapack::cusolver::hegvd, + oneapi::math::lapack::cusolver::hetrd, + oneapi::math::lapack::cusolver::hetrd, + oneapi::math::lapack::cusolver::hetrf, + oneapi::math::lapack::cusolver::hetrf, + oneapi::math::lapack::cusolver::orgbr, + oneapi::math::lapack::cusolver::orgbr, + oneapi::math::lapack::cusolver::orgqr, + oneapi::math::lapack::cusolver::orgqr, + oneapi::math::lapack::cusolver::orgtr, + oneapi::math::lapack::cusolver::orgtr, + oneapi::math::lapack::cusolver::ormtr, + oneapi::math::lapack::cusolver::ormtr, + oneapi::math::lapack::cusolver::ormrq, + oneapi::math::lapack::cusolver::ormrq, + oneapi::math::lapack::cusolver::ormqr, + oneapi::math::lapack::cusolver::ormqr, + oneapi::math::lapack::cusolver::potrf, + oneapi::math::lapack::cusolver::potrf, + oneapi::math::lapack::cusolver::potrf, + oneapi::math::lapack::cusolver::potrf, + oneapi::math::lapack::cusolver::potri, + oneapi::math::lapack::cusolver::potri, + oneapi::math::lapack::cusolver::potri, + oneapi::math::lapack::cusolver::potri, + oneapi::math::lapack::cusolver::potrs, + oneapi::math::lapack::cusolver::potrs, + oneapi::math::lapack::cusolver::potrs, + oneapi::math::lapack::cusolver::potrs, + oneapi::math::lapack::cusolver::syevd, + oneapi::math::lapack::cusolver::syevd, + oneapi::math::lapack::cusolver::sygvd, + oneapi::math::lapack::cusolver::sygvd, + oneapi::math::lapack::cusolver::sytrd, + oneapi::math::lapack::cusolver::sytrd, + oneapi::math::lapack::cusolver::sytrf, + oneapi::math::lapack::cusolver::sytrf, + oneapi::math::lapack::cusolver::sytrf, + oneapi::math::lapack::cusolver::sytrf, + oneapi::math::lapack::cusolver::trtrs, + oneapi::math::lapack::cusolver::trtrs, + oneapi::math::lapack::cusolver::trtrs, + oneapi::math::lapack::cusolver::trtrs, + oneapi::math::lapack::cusolver::ungbr, + oneapi::math::lapack::cusolver::ungbr, + oneapi::math::lapack::cusolver::ungqr, + oneapi::math::lapack::cusolver::ungqr, + oneapi::math::lapack::cusolver::ungtr, + oneapi::math::lapack::cusolver::ungtr, + oneapi::math::lapack::cusolver::unmrq, + oneapi::math::lapack::cusolver::unmrq, + oneapi::math::lapack::cusolver::unmqr, + oneapi::math::lapack::cusolver::unmqr, + oneapi::math::lapack::cusolver::unmtr, + oneapi::math::lapack::cusolver::unmtr, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::orgqr_batch, + oneapi::math::lapack::cusolver::orgqr_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::ungqr_batch, + oneapi::math::lapack::cusolver::ungqr_batch, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::orgqr_batch, + oneapi::math::lapack::cusolver::orgqr_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::ungqr_batch, + oneapi::math::lapack::cusolver::ungqr_batch, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::geqrf_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::getrf_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getri_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::getrs_batch, + oneapi::math::lapack::cusolver::orgqr_batch, + oneapi::math::lapack::cusolver::orgqr_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrf_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::potrs_batch, + oneapi::math::lapack::cusolver::ungqr_batch, + oneapi::math::lapack::cusolver::ungqr_batch, + oneapi::math::lapack::cusolver::gebrd_scratchpad_size, + oneapi::math::lapack::cusolver::gebrd_scratchpad_size, + oneapi::math::lapack::cusolver::gebrd_scratchpad_size>, + oneapi::math::lapack::cusolver::gebrd_scratchpad_size>, + oneapi::math::lapack::cusolver::gerqf_scratchpad_size, + oneapi::math::lapack::cusolver::gerqf_scratchpad_size, + oneapi::math::lapack::cusolver::gerqf_scratchpad_size>, + oneapi::math::lapack::cusolver::gerqf_scratchpad_size>, + oneapi::math::lapack::cusolver::geqrf_scratchpad_size, + oneapi::math::lapack::cusolver::geqrf_scratchpad_size, + oneapi::math::lapack::cusolver::geqrf_scratchpad_size>, + oneapi::math::lapack::cusolver::geqrf_scratchpad_size>, + oneapi::math::lapack::cusolver::gesvd_scratchpad_size, + oneapi::math::lapack::cusolver::gesvd_scratchpad_size, + oneapi::math::lapack::cusolver::gesvd_scratchpad_size>, + oneapi::math::lapack::cusolver::gesvd_scratchpad_size>, + oneapi::math::lapack::cusolver::getrf_scratchpad_size, + oneapi::math::lapack::cusolver::getrf_scratchpad_size, + oneapi::math::lapack::cusolver::getrf_scratchpad_size>, + oneapi::math::lapack::cusolver::getrf_scratchpad_size>, + oneapi::math::lapack::cusolver::getri_scratchpad_size, + oneapi::math::lapack::cusolver::getri_scratchpad_size, + oneapi::math::lapack::cusolver::getri_scratchpad_size>, + oneapi::math::lapack::cusolver::getri_scratchpad_size>, + oneapi::math::lapack::cusolver::getrs_scratchpad_size, + oneapi::math::lapack::cusolver::getrs_scratchpad_size, + oneapi::math::lapack::cusolver::getrs_scratchpad_size>, + oneapi::math::lapack::cusolver::getrs_scratchpad_size>, + oneapi::math::lapack::cusolver::heevd_scratchpad_size>, + oneapi::math::lapack::cusolver::heevd_scratchpad_size>, + oneapi::math::lapack::cusolver::hegvd_scratchpad_size>, + oneapi::math::lapack::cusolver::hegvd_scratchpad_size>, + oneapi::math::lapack::cusolver::hetrd_scratchpad_size>, + oneapi::math::lapack::cusolver::hetrd_scratchpad_size>, + oneapi::math::lapack::cusolver::hetrf_scratchpad_size>, + oneapi::math::lapack::cusolver::hetrf_scratchpad_size>, + oneapi::math::lapack::cusolver::orgbr_scratchpad_size, + oneapi::math::lapack::cusolver::orgbr_scratchpad_size, + oneapi::math::lapack::cusolver::orgtr_scratchpad_size, + oneapi::math::lapack::cusolver::orgtr_scratchpad_size, + oneapi::math::lapack::cusolver::orgqr_scratchpad_size, + oneapi::math::lapack::cusolver::orgqr_scratchpad_size, + oneapi::math::lapack::cusolver::ormrq_scratchpad_size, + oneapi::math::lapack::cusolver::ormrq_scratchpad_size, + oneapi::math::lapack::cusolver::ormqr_scratchpad_size, + oneapi::math::lapack::cusolver::ormqr_scratchpad_size, + oneapi::math::lapack::cusolver::ormtr_scratchpad_size, + oneapi::math::lapack::cusolver::ormtr_scratchpad_size, + oneapi::math::lapack::cusolver::potrf_scratchpad_size, + oneapi::math::lapack::cusolver::potrf_scratchpad_size, + oneapi::math::lapack::cusolver::potrf_scratchpad_size>, + oneapi::math::lapack::cusolver::potrf_scratchpad_size>, + oneapi::math::lapack::cusolver::potrs_scratchpad_size, + oneapi::math::lapack::cusolver::potrs_scratchpad_size, + oneapi::math::lapack::cusolver::potrs_scratchpad_size>, + oneapi::math::lapack::cusolver::potrs_scratchpad_size>, + oneapi::math::lapack::cusolver::potri_scratchpad_size, + oneapi::math::lapack::cusolver::potri_scratchpad_size, + oneapi::math::lapack::cusolver::potri_scratchpad_size>, + oneapi::math::lapack::cusolver::potri_scratchpad_size>, + oneapi::math::lapack::cusolver::sytrf_scratchpad_size, + oneapi::math::lapack::cusolver::sytrf_scratchpad_size, + oneapi::math::lapack::cusolver::sytrf_scratchpad_size>, + oneapi::math::lapack::cusolver::sytrf_scratchpad_size>, + oneapi::math::lapack::cusolver::syevd_scratchpad_size, + oneapi::math::lapack::cusolver::syevd_scratchpad_size, + oneapi::math::lapack::cusolver::sygvd_scratchpad_size, + oneapi::math::lapack::cusolver::sygvd_scratchpad_size, + oneapi::math::lapack::cusolver::sytrd_scratchpad_size, + oneapi::math::lapack::cusolver::sytrd_scratchpad_size, + oneapi::math::lapack::cusolver::trtrs_scratchpad_size, + oneapi::math::lapack::cusolver::trtrs_scratchpad_size, + oneapi::math::lapack::cusolver::trtrs_scratchpad_size>, + oneapi::math::lapack::cusolver::trtrs_scratchpad_size>, + oneapi::math::lapack::cusolver::ungbr_scratchpad_size>, + oneapi::math::lapack::cusolver::ungbr_scratchpad_size>, + oneapi::math::lapack::cusolver::ungqr_scratchpad_size>, + oneapi::math::lapack::cusolver::ungqr_scratchpad_size>, + oneapi::math::lapack::cusolver::ungtr_scratchpad_size>, + oneapi::math::lapack::cusolver::ungtr_scratchpad_size>, + oneapi::math::lapack::cusolver::unmrq_scratchpad_size>, + oneapi::math::lapack::cusolver::unmrq_scratchpad_size>, + oneapi::math::lapack::cusolver::unmqr_scratchpad_size>, + oneapi::math::lapack::cusolver::unmqr_scratchpad_size>, + oneapi::math::lapack::cusolver::unmtr_scratchpad_size>, + oneapi::math::lapack::cusolver::unmtr_scratchpad_size>, + oneapi::math::lapack::cusolver::getrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::getrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::getri_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getri_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getri_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::getri_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::getrs_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getrs_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getrs_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::getrs_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::geqrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::geqrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::potrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::potrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::potrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::potrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::potrs_batch_scratchpad_size, + oneapi::math::lapack::cusolver::potrs_batch_scratchpad_size, + oneapi::math::lapack::cusolver::potrs_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::potrs_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::orgqr_batch_scratchpad_size, + oneapi::math::lapack::cusolver::orgqr_batch_scratchpad_size, + oneapi::math::lapack::cusolver::ungqr_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::ungqr_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::getrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::getrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::getri_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getri_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getri_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::getri_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::getrs_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getrs_batch_scratchpad_size, + oneapi::math::lapack::cusolver::getrs_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::getrs_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::geqrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::geqrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::orgqr_batch_scratchpad_size, + oneapi::math::lapack::cusolver::orgqr_batch_scratchpad_size, + oneapi::math::lapack::cusolver::potrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::potrf_batch_scratchpad_size, + oneapi::math::lapack::cusolver::potrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::potrf_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::potrs_batch_scratchpad_size, + oneapi::math::lapack::cusolver::potrs_batch_scratchpad_size, + oneapi::math::lapack::cusolver::potrs_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::potrs_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::ungqr_batch_scratchpad_size>, + oneapi::math::lapack::cusolver::ungqr_batch_scratchpad_size> #undef LAPACK_BACKEND }; diff --git a/src/lapack/backends/mkl_common/lapack_wrappers.cxx b/src/lapack/backends/mkl_common/lapack_wrappers.cxx index a80f807b6..fa2069035 100644 --- a/src/lapack/backends/mkl_common/lapack_wrappers.cxx +++ b/src/lapack/backends/mkl_common/lapack_wrappers.cxx @@ -17,315 +17,315 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -oneapi::mkl::lapack::LAPACK_BACKEND::gebrd, oneapi::mkl::lapack::LAPACK_BACKEND::gebrd, - oneapi::mkl::lapack::LAPACK_BACKEND::gebrd, oneapi::mkl::lapack::LAPACK_BACKEND::gebrd, - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf, oneapi::mkl::lapack::LAPACK_BACKEND::gerqf, - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf, oneapi::mkl::lapack::LAPACK_BACKEND::gerqf, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf, oneapi::mkl::lapack::LAPACK_BACKEND::geqrf, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf, oneapi::mkl::lapack::LAPACK_BACKEND::geqrf, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf, oneapi::mkl::lapack::LAPACK_BACKEND::getrf, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf, oneapi::mkl::lapack::LAPACK_BACKEND::getrf, - oneapi::mkl::lapack::LAPACK_BACKEND::getri, oneapi::mkl::lapack::LAPACK_BACKEND::getri, - oneapi::mkl::lapack::LAPACK_BACKEND::getri, oneapi::mkl::lapack::LAPACK_BACKEND::getri, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs, oneapi::mkl::lapack::LAPACK_BACKEND::getrs, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs, oneapi::mkl::lapack::LAPACK_BACKEND::getrs, - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd, oneapi::mkl::lapack::LAPACK_BACKEND::gesvd, - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd, oneapi::mkl::lapack::LAPACK_BACKEND::gesvd, - oneapi::mkl::lapack::LAPACK_BACKEND::heevd, oneapi::mkl::lapack::LAPACK_BACKEND::heevd, - oneapi::mkl::lapack::LAPACK_BACKEND::hegvd, oneapi::mkl::lapack::LAPACK_BACKEND::hegvd, - oneapi::mkl::lapack::LAPACK_BACKEND::hetrd, oneapi::mkl::lapack::LAPACK_BACKEND::hetrd, - oneapi::mkl::lapack::LAPACK_BACKEND::hetrf, oneapi::mkl::lapack::LAPACK_BACKEND::hetrf, - oneapi::mkl::lapack::LAPACK_BACKEND::orgbr, oneapi::mkl::lapack::LAPACK_BACKEND::orgbr, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr, oneapi::mkl::lapack::LAPACK_BACKEND::orgqr, - oneapi::mkl::lapack::LAPACK_BACKEND::orgtr, oneapi::mkl::lapack::LAPACK_BACKEND::orgtr, - oneapi::mkl::lapack::LAPACK_BACKEND::ormtr, oneapi::mkl::lapack::LAPACK_BACKEND::ormtr, - oneapi::mkl::lapack::LAPACK_BACKEND::ormrq, oneapi::mkl::lapack::LAPACK_BACKEND::ormrq, - oneapi::mkl::lapack::LAPACK_BACKEND::ormqr, oneapi::mkl::lapack::LAPACK_BACKEND::ormqr, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf, oneapi::mkl::lapack::LAPACK_BACKEND::potrf, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf, oneapi::mkl::lapack::LAPACK_BACKEND::potrf, - oneapi::mkl::lapack::LAPACK_BACKEND::potri, oneapi::mkl::lapack::LAPACK_BACKEND::potri, - oneapi::mkl::lapack::LAPACK_BACKEND::potri, oneapi::mkl::lapack::LAPACK_BACKEND::potri, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs, oneapi::mkl::lapack::LAPACK_BACKEND::potrs, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs, oneapi::mkl::lapack::LAPACK_BACKEND::potrs, - oneapi::mkl::lapack::LAPACK_BACKEND::syevd, oneapi::mkl::lapack::LAPACK_BACKEND::syevd, - oneapi::mkl::lapack::LAPACK_BACKEND::sygvd, oneapi::mkl::lapack::LAPACK_BACKEND::sygvd, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrd, oneapi::mkl::lapack::LAPACK_BACKEND::sytrd, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf, oneapi::mkl::lapack::LAPACK_BACKEND::sytrf, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf, oneapi::mkl::lapack::LAPACK_BACKEND::sytrf, - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs, oneapi::mkl::lapack::LAPACK_BACKEND::trtrs, - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs, oneapi::mkl::lapack::LAPACK_BACKEND::trtrs, - oneapi::mkl::lapack::LAPACK_BACKEND::ungbr, oneapi::mkl::lapack::LAPACK_BACKEND::ungbr, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr, oneapi::mkl::lapack::LAPACK_BACKEND::ungqr, - oneapi::mkl::lapack::LAPACK_BACKEND::ungtr, oneapi::mkl::lapack::LAPACK_BACKEND::ungtr, - oneapi::mkl::lapack::LAPACK_BACKEND::unmrq, oneapi::mkl::lapack::LAPACK_BACKEND::unmrq, - oneapi::mkl::lapack::LAPACK_BACKEND::unmqr, oneapi::mkl::lapack::LAPACK_BACKEND::unmqr, - oneapi::mkl::lapack::LAPACK_BACKEND::unmtr, oneapi::mkl::lapack::LAPACK_BACKEND::unmtr, - oneapi::mkl::lapack::LAPACK_BACKEND::gebrd, oneapi::mkl::lapack::LAPACK_BACKEND::gebrd, - oneapi::mkl::lapack::LAPACK_BACKEND::gebrd, oneapi::mkl::lapack::LAPACK_BACKEND::gebrd, - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf, oneapi::mkl::lapack::LAPACK_BACKEND::gerqf, - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf, oneapi::mkl::lapack::LAPACK_BACKEND::gerqf, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf, oneapi::mkl::lapack::LAPACK_BACKEND::geqrf, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf, oneapi::mkl::lapack::LAPACK_BACKEND::geqrf, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf, oneapi::mkl::lapack::LAPACK_BACKEND::getrf, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf, oneapi::mkl::lapack::LAPACK_BACKEND::getrf, - oneapi::mkl::lapack::LAPACK_BACKEND::getri, oneapi::mkl::lapack::LAPACK_BACKEND::getri, - oneapi::mkl::lapack::LAPACK_BACKEND::getri, oneapi::mkl::lapack::LAPACK_BACKEND::getri, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs, oneapi::mkl::lapack::LAPACK_BACKEND::getrs, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs, oneapi::mkl::lapack::LAPACK_BACKEND::getrs, - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd, oneapi::mkl::lapack::LAPACK_BACKEND::gesvd, - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd, oneapi::mkl::lapack::LAPACK_BACKEND::gesvd, - oneapi::mkl::lapack::LAPACK_BACKEND::heevd, oneapi::mkl::lapack::LAPACK_BACKEND::heevd, - oneapi::mkl::lapack::LAPACK_BACKEND::hegvd, oneapi::mkl::lapack::LAPACK_BACKEND::hegvd, - oneapi::mkl::lapack::LAPACK_BACKEND::hetrd, oneapi::mkl::lapack::LAPACK_BACKEND::hetrd, - oneapi::mkl::lapack::LAPACK_BACKEND::hetrf, oneapi::mkl::lapack::LAPACK_BACKEND::hetrf, - oneapi::mkl::lapack::LAPACK_BACKEND::orgbr, oneapi::mkl::lapack::LAPACK_BACKEND::orgbr, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr, oneapi::mkl::lapack::LAPACK_BACKEND::orgqr, - oneapi::mkl::lapack::LAPACK_BACKEND::orgtr, oneapi::mkl::lapack::LAPACK_BACKEND::orgtr, - oneapi::mkl::lapack::LAPACK_BACKEND::ormtr, oneapi::mkl::lapack::LAPACK_BACKEND::ormtr, - oneapi::mkl::lapack::LAPACK_BACKEND::ormrq, oneapi::mkl::lapack::LAPACK_BACKEND::ormrq, - oneapi::mkl::lapack::LAPACK_BACKEND::ormqr, oneapi::mkl::lapack::LAPACK_BACKEND::ormqr, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf, oneapi::mkl::lapack::LAPACK_BACKEND::potrf, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf, oneapi::mkl::lapack::LAPACK_BACKEND::potrf, - oneapi::mkl::lapack::LAPACK_BACKEND::potri, oneapi::mkl::lapack::LAPACK_BACKEND::potri, - oneapi::mkl::lapack::LAPACK_BACKEND::potri, oneapi::mkl::lapack::LAPACK_BACKEND::potri, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs, oneapi::mkl::lapack::LAPACK_BACKEND::potrs, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs, oneapi::mkl::lapack::LAPACK_BACKEND::potrs, - oneapi::mkl::lapack::LAPACK_BACKEND::syevd, oneapi::mkl::lapack::LAPACK_BACKEND::syevd, - oneapi::mkl::lapack::LAPACK_BACKEND::sygvd, oneapi::mkl::lapack::LAPACK_BACKEND::sygvd, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrd, oneapi::mkl::lapack::LAPACK_BACKEND::sytrd, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf, oneapi::mkl::lapack::LAPACK_BACKEND::sytrf, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf, oneapi::mkl::lapack::LAPACK_BACKEND::sytrf, - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs, oneapi::mkl::lapack::LAPACK_BACKEND::trtrs, - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs, oneapi::mkl::lapack::LAPACK_BACKEND::trtrs, - oneapi::mkl::lapack::LAPACK_BACKEND::ungbr, oneapi::mkl::lapack::LAPACK_BACKEND::ungbr, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr, oneapi::mkl::lapack::LAPACK_BACKEND::ungqr, - oneapi::mkl::lapack::LAPACK_BACKEND::ungtr, oneapi::mkl::lapack::LAPACK_BACKEND::ungtr, - oneapi::mkl::lapack::LAPACK_BACKEND::unmrq, oneapi::mkl::lapack::LAPACK_BACKEND::unmrq, - oneapi::mkl::lapack::LAPACK_BACKEND::unmqr, oneapi::mkl::lapack::LAPACK_BACKEND::unmqr, - oneapi::mkl::lapack::LAPACK_BACKEND::unmtr, oneapi::mkl::lapack::LAPACK_BACKEND::unmtr, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch, - oneapi::mkl::lapack::LAPACK_BACKEND::gebrd_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::gebrd_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::gebrd_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::gebrd_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::gerqf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::gesvd_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::heevd_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::heevd_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::hegvd_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::hegvd_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::hetrd_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::hetrd_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::hetrf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::hetrf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::orgbr_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::orgbr_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::orgtr_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::orgtr_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::ormrq_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::ormrq_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::ormqr_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::ormqr_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::ormtr_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::ormtr_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potri_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potri_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potri_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potri_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrf_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::syevd_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::syevd_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::sygvd_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::sygvd_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrd_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::sytrd_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::trtrs_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::ungbr_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::ungbr_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::ungtr_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::ungtr_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::unmrq_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::unmrq_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::unmqr_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::unmqr_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::unmtr_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::unmtr_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size>, - oneapi::mkl::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size> +oneapi::math::lapack::LAPACK_BACKEND::gebrd, oneapi::math::lapack::LAPACK_BACKEND::gebrd, + oneapi::math::lapack::LAPACK_BACKEND::gebrd, oneapi::math::lapack::LAPACK_BACKEND::gebrd, + oneapi::math::lapack::LAPACK_BACKEND::gerqf, oneapi::math::lapack::LAPACK_BACKEND::gerqf, + oneapi::math::lapack::LAPACK_BACKEND::gerqf, oneapi::math::lapack::LAPACK_BACKEND::gerqf, + oneapi::math::lapack::LAPACK_BACKEND::geqrf, oneapi::math::lapack::LAPACK_BACKEND::geqrf, + oneapi::math::lapack::LAPACK_BACKEND::geqrf, oneapi::math::lapack::LAPACK_BACKEND::geqrf, + oneapi::math::lapack::LAPACK_BACKEND::getrf, oneapi::math::lapack::LAPACK_BACKEND::getrf, + oneapi::math::lapack::LAPACK_BACKEND::getrf, oneapi::math::lapack::LAPACK_BACKEND::getrf, + oneapi::math::lapack::LAPACK_BACKEND::getri, oneapi::math::lapack::LAPACK_BACKEND::getri, + oneapi::math::lapack::LAPACK_BACKEND::getri, oneapi::math::lapack::LAPACK_BACKEND::getri, + oneapi::math::lapack::LAPACK_BACKEND::getrs, oneapi::math::lapack::LAPACK_BACKEND::getrs, + oneapi::math::lapack::LAPACK_BACKEND::getrs, oneapi::math::lapack::LAPACK_BACKEND::getrs, + oneapi::math::lapack::LAPACK_BACKEND::gesvd, oneapi::math::lapack::LAPACK_BACKEND::gesvd, + oneapi::math::lapack::LAPACK_BACKEND::gesvd, oneapi::math::lapack::LAPACK_BACKEND::gesvd, + oneapi::math::lapack::LAPACK_BACKEND::heevd, oneapi::math::lapack::LAPACK_BACKEND::heevd, + oneapi::math::lapack::LAPACK_BACKEND::hegvd, oneapi::math::lapack::LAPACK_BACKEND::hegvd, + oneapi::math::lapack::LAPACK_BACKEND::hetrd, oneapi::math::lapack::LAPACK_BACKEND::hetrd, + oneapi::math::lapack::LAPACK_BACKEND::hetrf, oneapi::math::lapack::LAPACK_BACKEND::hetrf, + oneapi::math::lapack::LAPACK_BACKEND::orgbr, oneapi::math::lapack::LAPACK_BACKEND::orgbr, + oneapi::math::lapack::LAPACK_BACKEND::orgqr, oneapi::math::lapack::LAPACK_BACKEND::orgqr, + oneapi::math::lapack::LAPACK_BACKEND::orgtr, oneapi::math::lapack::LAPACK_BACKEND::orgtr, + oneapi::math::lapack::LAPACK_BACKEND::ormtr, oneapi::math::lapack::LAPACK_BACKEND::ormtr, + oneapi::math::lapack::LAPACK_BACKEND::ormrq, oneapi::math::lapack::LAPACK_BACKEND::ormrq, + oneapi::math::lapack::LAPACK_BACKEND::ormqr, oneapi::math::lapack::LAPACK_BACKEND::ormqr, + oneapi::math::lapack::LAPACK_BACKEND::potrf, oneapi::math::lapack::LAPACK_BACKEND::potrf, + oneapi::math::lapack::LAPACK_BACKEND::potrf, oneapi::math::lapack::LAPACK_BACKEND::potrf, + oneapi::math::lapack::LAPACK_BACKEND::potri, oneapi::math::lapack::LAPACK_BACKEND::potri, + oneapi::math::lapack::LAPACK_BACKEND::potri, oneapi::math::lapack::LAPACK_BACKEND::potri, + oneapi::math::lapack::LAPACK_BACKEND::potrs, oneapi::math::lapack::LAPACK_BACKEND::potrs, + oneapi::math::lapack::LAPACK_BACKEND::potrs, oneapi::math::lapack::LAPACK_BACKEND::potrs, + oneapi::math::lapack::LAPACK_BACKEND::syevd, oneapi::math::lapack::LAPACK_BACKEND::syevd, + oneapi::math::lapack::LAPACK_BACKEND::sygvd, oneapi::math::lapack::LAPACK_BACKEND::sygvd, + oneapi::math::lapack::LAPACK_BACKEND::sytrd, oneapi::math::lapack::LAPACK_BACKEND::sytrd, + oneapi::math::lapack::LAPACK_BACKEND::sytrf, oneapi::math::lapack::LAPACK_BACKEND::sytrf, + oneapi::math::lapack::LAPACK_BACKEND::sytrf, oneapi::math::lapack::LAPACK_BACKEND::sytrf, + oneapi::math::lapack::LAPACK_BACKEND::trtrs, oneapi::math::lapack::LAPACK_BACKEND::trtrs, + oneapi::math::lapack::LAPACK_BACKEND::trtrs, oneapi::math::lapack::LAPACK_BACKEND::trtrs, + oneapi::math::lapack::LAPACK_BACKEND::ungbr, oneapi::math::lapack::LAPACK_BACKEND::ungbr, + oneapi::math::lapack::LAPACK_BACKEND::ungqr, oneapi::math::lapack::LAPACK_BACKEND::ungqr, + oneapi::math::lapack::LAPACK_BACKEND::ungtr, oneapi::math::lapack::LAPACK_BACKEND::ungtr, + oneapi::math::lapack::LAPACK_BACKEND::unmrq, oneapi::math::lapack::LAPACK_BACKEND::unmrq, + oneapi::math::lapack::LAPACK_BACKEND::unmqr, oneapi::math::lapack::LAPACK_BACKEND::unmqr, + oneapi::math::lapack::LAPACK_BACKEND::unmtr, oneapi::math::lapack::LAPACK_BACKEND::unmtr, + oneapi::math::lapack::LAPACK_BACKEND::gebrd, oneapi::math::lapack::LAPACK_BACKEND::gebrd, + oneapi::math::lapack::LAPACK_BACKEND::gebrd, oneapi::math::lapack::LAPACK_BACKEND::gebrd, + oneapi::math::lapack::LAPACK_BACKEND::gerqf, oneapi::math::lapack::LAPACK_BACKEND::gerqf, + oneapi::math::lapack::LAPACK_BACKEND::gerqf, oneapi::math::lapack::LAPACK_BACKEND::gerqf, + oneapi::math::lapack::LAPACK_BACKEND::geqrf, oneapi::math::lapack::LAPACK_BACKEND::geqrf, + oneapi::math::lapack::LAPACK_BACKEND::geqrf, oneapi::math::lapack::LAPACK_BACKEND::geqrf, + oneapi::math::lapack::LAPACK_BACKEND::getrf, oneapi::math::lapack::LAPACK_BACKEND::getrf, + oneapi::math::lapack::LAPACK_BACKEND::getrf, oneapi::math::lapack::LAPACK_BACKEND::getrf, + oneapi::math::lapack::LAPACK_BACKEND::getri, oneapi::math::lapack::LAPACK_BACKEND::getri, + oneapi::math::lapack::LAPACK_BACKEND::getri, oneapi::math::lapack::LAPACK_BACKEND::getri, + oneapi::math::lapack::LAPACK_BACKEND::getrs, oneapi::math::lapack::LAPACK_BACKEND::getrs, + oneapi::math::lapack::LAPACK_BACKEND::getrs, oneapi::math::lapack::LAPACK_BACKEND::getrs, + oneapi::math::lapack::LAPACK_BACKEND::gesvd, oneapi::math::lapack::LAPACK_BACKEND::gesvd, + oneapi::math::lapack::LAPACK_BACKEND::gesvd, oneapi::math::lapack::LAPACK_BACKEND::gesvd, + oneapi::math::lapack::LAPACK_BACKEND::heevd, oneapi::math::lapack::LAPACK_BACKEND::heevd, + oneapi::math::lapack::LAPACK_BACKEND::hegvd, oneapi::math::lapack::LAPACK_BACKEND::hegvd, + oneapi::math::lapack::LAPACK_BACKEND::hetrd, oneapi::math::lapack::LAPACK_BACKEND::hetrd, + oneapi::math::lapack::LAPACK_BACKEND::hetrf, oneapi::math::lapack::LAPACK_BACKEND::hetrf, + oneapi::math::lapack::LAPACK_BACKEND::orgbr, oneapi::math::lapack::LAPACK_BACKEND::orgbr, + oneapi::math::lapack::LAPACK_BACKEND::orgqr, oneapi::math::lapack::LAPACK_BACKEND::orgqr, + oneapi::math::lapack::LAPACK_BACKEND::orgtr, oneapi::math::lapack::LAPACK_BACKEND::orgtr, + oneapi::math::lapack::LAPACK_BACKEND::ormtr, oneapi::math::lapack::LAPACK_BACKEND::ormtr, + oneapi::math::lapack::LAPACK_BACKEND::ormrq, oneapi::math::lapack::LAPACK_BACKEND::ormrq, + oneapi::math::lapack::LAPACK_BACKEND::ormqr, oneapi::math::lapack::LAPACK_BACKEND::ormqr, + oneapi::math::lapack::LAPACK_BACKEND::potrf, oneapi::math::lapack::LAPACK_BACKEND::potrf, + oneapi::math::lapack::LAPACK_BACKEND::potrf, oneapi::math::lapack::LAPACK_BACKEND::potrf, + oneapi::math::lapack::LAPACK_BACKEND::potri, oneapi::math::lapack::LAPACK_BACKEND::potri, + oneapi::math::lapack::LAPACK_BACKEND::potri, oneapi::math::lapack::LAPACK_BACKEND::potri, + oneapi::math::lapack::LAPACK_BACKEND::potrs, oneapi::math::lapack::LAPACK_BACKEND::potrs, + oneapi::math::lapack::LAPACK_BACKEND::potrs, oneapi::math::lapack::LAPACK_BACKEND::potrs, + oneapi::math::lapack::LAPACK_BACKEND::syevd, oneapi::math::lapack::LAPACK_BACKEND::syevd, + oneapi::math::lapack::LAPACK_BACKEND::sygvd, oneapi::math::lapack::LAPACK_BACKEND::sygvd, + oneapi::math::lapack::LAPACK_BACKEND::sytrd, oneapi::math::lapack::LAPACK_BACKEND::sytrd, + oneapi::math::lapack::LAPACK_BACKEND::sytrf, oneapi::math::lapack::LAPACK_BACKEND::sytrf, + oneapi::math::lapack::LAPACK_BACKEND::sytrf, oneapi::math::lapack::LAPACK_BACKEND::sytrf, + oneapi::math::lapack::LAPACK_BACKEND::trtrs, oneapi::math::lapack::LAPACK_BACKEND::trtrs, + oneapi::math::lapack::LAPACK_BACKEND::trtrs, oneapi::math::lapack::LAPACK_BACKEND::trtrs, + oneapi::math::lapack::LAPACK_BACKEND::ungbr, oneapi::math::lapack::LAPACK_BACKEND::ungbr, + oneapi::math::lapack::LAPACK_BACKEND::ungqr, oneapi::math::lapack::LAPACK_BACKEND::ungqr, + oneapi::math::lapack::LAPACK_BACKEND::ungtr, oneapi::math::lapack::LAPACK_BACKEND::ungtr, + oneapi::math::lapack::LAPACK_BACKEND::unmrq, oneapi::math::lapack::LAPACK_BACKEND::unmrq, + oneapi::math::lapack::LAPACK_BACKEND::unmqr, oneapi::math::lapack::LAPACK_BACKEND::unmqr, + oneapi::math::lapack::LAPACK_BACKEND::unmtr, oneapi::math::lapack::LAPACK_BACKEND::unmtr, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch, + oneapi::math::lapack::LAPACK_BACKEND::gebrd_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::gebrd_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::gebrd_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::gebrd_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::gerqf_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::gerqf_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::gerqf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::gerqf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::gesvd_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::gesvd_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::gesvd_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::gesvd_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrf_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrf_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getri_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getri_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getri_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getri_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrs_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrs_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrs_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrs_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::heevd_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::heevd_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::hegvd_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::hegvd_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::hetrd_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::hetrd_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::hetrf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::hetrf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::orgbr_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::orgbr_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::orgtr_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::orgtr_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::ormrq_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::ormrq_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::ormqr_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::ormqr_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::ormtr_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::ormtr_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrf_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrf_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potrf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potrs_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrs_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrs_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potrs_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potri_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potri_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potri_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potri_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::sytrf_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::sytrf_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::sytrf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::sytrf_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::syevd_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::syevd_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::sygvd_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::sygvd_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::sytrd_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::sytrd_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::trtrs_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::trtrs_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::trtrs_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::trtrs_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::ungbr_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::ungbr_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::ungtr_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::ungtr_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::unmrq_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::unmrq_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::unmqr_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::unmqr_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::unmtr_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::unmtr_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size>, + oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size> diff --git a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp b/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp index 706a76dcc..cc31eb4c1 100644 --- a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp +++ b/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp @@ -25,7 +25,7 @@ #include "oneapi/math/lapack/types.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, @@ -96,74 +96,74 @@ void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int6 void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, @@ -172,127 +172,127 @@ void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, +void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, +void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); @@ -304,42 +304,42 @@ void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, @@ -378,24 +378,24 @@ void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, @@ -427,37 +427,37 @@ void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_ sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, @@ -534,78 +534,78 @@ sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, +sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, +sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, @@ -614,131 +614,131 @@ sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64 sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float *a, +sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double *a, +sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, +sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, +sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, double *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); @@ -750,41 +750,41 @@ sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64 std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); @@ -900,49 +900,49 @@ sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, @@ -966,81 +966,81 @@ sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, st double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, float **a, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, double **a, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, @@ -1070,19 +1070,19 @@ sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, st template = nullptr> std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); -template = nullptr> +template = nullptr> std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template = nullptr> std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, @@ -1090,174 +1090,174 @@ std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int6 template = nullptr> std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); +std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, +std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, + oneapi::math::lapack::internal::is_real_floating_point = nullptr> +std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); -template = nullptr> -std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); +std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, +std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> +std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); -template = nullptr> +template = nullptr> std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); -template = nullptr> +template = nullptr> std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); -template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +template = nullptr> +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -template = nullptr> +template = nullptr> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); -template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); -template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +template = nullptr> +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); -template = nullptr> +template = nullptr> std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> +template = nullptr> std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose *trans, +template = nullptr> +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> +template = nullptr> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); template = nullptr> + oneapi::math::lapack::internal::is_real_floating_point = nullptr> std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +template = nullptr> +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); -template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +template = nullptr> +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); template = nullptr> + oneapi::math::lapack::internal::is_complex_floating_point = nullptr> std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/lapack/backends/mklcpu/mkl_lapack.cpp b/src/lapack/backends/mklcpu/mkl_lapack.cpp index f64b538a8..246722d15 100644 --- a/src/lapack/backends/mklcpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklcpu/mkl_lapack.cpp @@ -29,7 +29,7 @@ #include "../mkl_common/mkl_lapack_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace mklcpu { @@ -37,5 +37,5 @@ namespace mklcpu { } // namespace mklcpu } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/lapack/backends/mklgpu/mkl_lapack.cpp b/src/lapack/backends/mklgpu/mkl_lapack.cpp index 8d24c0eb9..4650ccdd5 100644 --- a/src/lapack/backends/mklgpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklgpu/mkl_lapack.cpp @@ -29,7 +29,7 @@ #include "../mkl_common/mkl_lapack_backend.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace mklgpu { @@ -37,5 +37,5 @@ namespace mklgpu { } // namespace mklgpu } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/lapack/backends/rocsolver/rocsolver_batch.cpp b/src/lapack/backends/rocsolver/rocsolver_batch.cpp index 03217b85b..ae21f70af 100644 --- a/src/lapack/backends/rocsolver/rocsolver_batch.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_batch.cpp @@ -25,7 +25,7 @@ #include "oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace rocsolver { @@ -81,14 +81,14 @@ void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getri_batch"); } -void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getrs_batch"); } -void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -96,7 +96,7 @@ void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t std::int64_t scratchpad_size) { throw unimplemented("lapack", "getrs_batch"); } -void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, @@ -104,7 +104,7 @@ void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t std::int64_t scratchpad_size) { throw unimplemented("lapack", "getrs_batch"); } -void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, @@ -150,51 +150,51 @@ void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "orgqr_batch"); } -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrf_batch"); } -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrf_batch"); } -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrf_batch"); } -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrf_batch"); } -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrs_batch"); } -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrs_batch"); } -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrs_batch"); } -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, @@ -378,7 +378,7 @@ sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex &dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, @@ -386,7 +386,7 @@ sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::i const std::vector &dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, @@ -394,7 +394,7 @@ sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::i const std::vector &dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -403,7 +403,7 @@ sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::i const std::vector &dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, @@ -412,21 +412,21 @@ sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::i const std::vector &dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, @@ -434,7 +434,7 @@ sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std:: const std::vector &dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, @@ -468,26 +468,26 @@ sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, st const std::vector &dependencies) { throw unimplemented("lapack", "orgqr_batch"); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "potrf_batch"); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "potrf_batch"); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "potrf_batch"); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -497,7 +497,7 @@ sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t template inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, T **a, std::int64_t *lda, + oneapi::math::uplo *uplo, std::int64_t *n, T **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -540,7 +540,7 @@ inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &qu // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRF_BATCH_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ sycl::event potrf_batch( \ - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, TYPE **a, std::int64_t *lda, \ + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, TYPE **a, std::int64_t *lda, \ std::int64_t group_count, std::int64_t *group_sizes, TYPE *scratchpad, \ std::int64_t scratchpad_size, const std::vector &dependencies) { \ return potrf_batch(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, \ @@ -554,21 +554,21 @@ POTRF_BATCH_LAUNCHER_USM(std::complex, rocsolver_zpotrf_batched) #undef POTRF_BATCH_LAUNCHER_USM -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "potrs_batch"); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "potrs_batch"); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -576,7 +576,7 @@ sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t const std::vector &dependencies) { throw unimplemented("lapack", "potrs_batch"); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -587,7 +587,7 @@ sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t template inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, T **a, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, T **a, std::int64_t *lda, T **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, std::int64_t scratchpad_size, @@ -640,7 +640,7 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRS_BATCH_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ sycl::event potrs_batch( \ - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, \ + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, \ TYPE **a, std::int64_t *lda, TYPE **b, std::int64_t *ldb, std::int64_t group_count, \ std::int64_t *group_sizes, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -746,7 +746,7 @@ std::int64_t getri_batch_scratchpad_size>(sycl::queue &queu throw unimplemented("lapack", "getri_batch_scratchpad_size"); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, @@ -754,7 +754,7 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl: throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, @@ -763,14 +763,14 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("lapack", "getrs_batch_scratchpad_size"); @@ -805,20 +805,20 @@ std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queu } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { throw unimplemented("lapack", "potrf_batch_scratchpad_size"); } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { throw unimplemented("lapack", "potrf_batch_scratchpad_size"); } template <> std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { @@ -826,21 +826,21 @@ std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue } template <> std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { throw unimplemented("lapack", "potrf_batch_scratchpad_size"); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("lapack", "potrs_batch_scratchpad_size"); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, @@ -849,13 +849,13 @@ std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("lapack", "potrs_batch_scratchpad_size"); } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("lapack", "potrs_batch_scratchpad_size"); } @@ -940,7 +940,7 @@ std::int64_t getri_batch_scratchpad_size>(sycl::queue &queu throw unimplemented("lapack", "getri_batch_scratchpad_size"); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose *trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, @@ -948,7 +948,7 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl: throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose *trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, @@ -957,13 +957,13 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } @@ -1014,7 +1014,7 @@ std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_ #define POTRF_GROUP_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t potrf_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::uplo * uplo, std::int64_t * n, std::int64_t * lda, \ + sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t * n, std::int64_t * lda, \ std::int64_t group_count, std::int64_t * group_sizes) { \ return 0; \ } @@ -1030,7 +1030,7 @@ POTRF_GROUP_LAUNCHER_SCRATCH(std::complex) #define POTRS_GROUP_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t potrs_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::uplo * uplo, std::int64_t * n, std::int64_t * nrhs, \ + sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t * n, std::int64_t * nrhs, \ std::int64_t * lda, std::int64_t * ldb, std::int64_t group_count, \ std::int64_t * group_sizes) { \ return 0; \ @@ -1062,5 +1062,5 @@ std::int64_t ungqr_batch_scratchpad_size>(sycl::queue &queu } // namespace rocsolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/lapack/backends/rocsolver/rocsolver_handle.hpp b/src/lapack/backends/rocsolver/rocsolver_handle.hpp index c44463ef4..3e1a14b34 100644 --- a/src/lapack/backends/rocsolver/rocsolver_handle.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_handle.hpp @@ -24,7 +24,7 @@ #include namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace rocsolver { @@ -57,7 +57,7 @@ struct rocsolver_handle { } // namespace rocsolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif // ROCSOLVER_HANDLE_HPP diff --git a/src/lapack/backends/rocsolver/rocsolver_helper.hpp b/src/lapack/backends/rocsolver/rocsolver_helper.hpp index 36f78d00d..36464f2e6 100644 --- a/src/lapack/backends/rocsolver/rocsolver_helper.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_helper.hpp @@ -38,7 +38,7 @@ #include "oneapi/math/lapack/exceptions.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace rocsolver { @@ -186,53 +186,53 @@ inline rocblas_eform get_rocsolver_itype(std::int64_t itype) { } } -inline rocblas_evect get_rocsolver_job(oneapi::mkl::job jobz) { +inline rocblas_evect get_rocsolver_job(oneapi::math::job jobz) { switch (jobz) { - case oneapi::mkl::job::V: return rocblas_evect_original; - case oneapi::mkl::job::N: return rocblas_evect_none; + case oneapi::math::job::V: return rocblas_evect_original; + case oneapi::math::job::N: return rocblas_evect_none; default: throw "Wrong jobz."; } } -inline rocblas_svect get_rocsolver_jobsvd(oneapi::mkl::jobsvd job) { +inline rocblas_svect get_rocsolver_jobsvd(oneapi::math::jobsvd job) { switch (job) { - case oneapi::mkl::jobsvd::N: return rocblas_svect_none; - case oneapi::mkl::jobsvd::A: return rocblas_svect_all; - case oneapi::mkl::jobsvd::O: return rocblas_svect_overwrite; - case oneapi::mkl::jobsvd::S: return rocblas_svect_singular; + case oneapi::math::jobsvd::N: return rocblas_svect_none; + case oneapi::math::jobsvd::A: return rocblas_svect_all; + case oneapi::math::jobsvd::O: return rocblas_svect_overwrite; + case oneapi::math::jobsvd::S: return rocblas_svect_singular; default: throw "Wrong jobsvd."; } } -inline rocblas_operation get_rocblas_operation(oneapi::mkl::transpose trn) { +inline rocblas_operation get_rocblas_operation(oneapi::math::transpose trn) { switch (trn) { - case oneapi::mkl::transpose::nontrans: return rocblas_operation_none; - case oneapi::mkl::transpose::trans: return rocblas_operation_transpose; - case oneapi::mkl::transpose::conjtrans: return rocblas_operation_conjugate_transpose; + case oneapi::math::transpose::nontrans: return rocblas_operation_none; + case oneapi::math::transpose::trans: return rocblas_operation_transpose; + case oneapi::math::transpose::conjtrans: return rocblas_operation_conjugate_transpose; default: throw "Wrong transpose Operation."; } } -inline rocblas_fill get_rocblas_fill_mode(oneapi::mkl::uplo ul) { +inline rocblas_fill get_rocblas_fill_mode(oneapi::math::uplo ul) { switch (ul) { - case oneapi::mkl::uplo::upper: return rocblas_fill_upper; - case oneapi::mkl::uplo::lower: return rocblas_fill_lower; + case oneapi::math::uplo::upper: return rocblas_fill_upper; + case oneapi::math::uplo::lower: return rocblas_fill_lower; default: throw "Wrong fill mode."; } } -inline rocblas_side get_rocblas_side_mode(oneapi::mkl::side lr) { +inline rocblas_side get_rocblas_side_mode(oneapi::math::side lr) { switch (lr) { - case oneapi::mkl::side::left: return rocblas_side_left; - case oneapi::mkl::side::right: return rocblas_side_right; + case oneapi::math::side::left: return rocblas_side_left; + case oneapi::math::side::right: return rocblas_side_right; default: throw "Wrong side mode."; } } -inline rocblas_storev get_rocblas_generate(oneapi::mkl::generate qp) { +inline rocblas_storev get_rocblas_generate(oneapi::math::generate qp) { switch (qp) { - case oneapi::mkl::generate::Q: return rocblas_column_wise; - case oneapi::mkl::generate::P: return rocblas_row_wise; + case oneapi::math::generate::Q: return rocblas_column_wise; + case oneapi::math::generate::P: return rocblas_row_wise; default: throw "Wrong generate."; } } @@ -276,13 +276,13 @@ inline void lapack_info_check(sycl::queue &queue, DEVINFO_T devinfo, const char queue.wait(); const int devinfo_ = get_rocsolver_devinfo(queue, devinfo); if (devinfo_ > 0) - throw oneapi::mkl::lapack::computation_error( + throw oneapi::math::lapack::computation_error( func_name, std::string(cufunc_name) + " failed with info = " + std::to_string(devinfo_), devinfo_); } } // namespace rocsolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _ROCSOLVER_HELPER_HPP_ diff --git a/src/lapack/backends/rocsolver/rocsolver_lapack.cpp b/src/lapack/backends/rocsolver/rocsolver_lapack.cpp index 311c9985a..ea7a61937 100644 --- a/src/lapack/backends/rocsolver/rocsolver_lapack.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_lapack.cpp @@ -25,7 +25,7 @@ #include "oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace rocsolver { @@ -211,7 +211,7 @@ void getri(sycl::queue &queue, std::int64_t n, sycl::buffer template inline void getrs(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -249,7 +249,7 @@ inline void getrs(const char *func_name, Func func, sycl::queue &queue, } #define GETRS_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, \ + void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, \ sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ @@ -265,8 +265,8 @@ GETRS_LAUNCHER(std::complex, rocsolver_zgetrs) #undef GETRS_LAUNCHER template -inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, +inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -300,7 +300,7 @@ inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define GESVD_LAUNCHER(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, \ + void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, \ sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, \ @@ -317,8 +317,8 @@ GESVD_LAUNCHER(std::complex, double, rocsolver_zgesvd) #undef GESVD_LAUNCHER template -inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, +inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using rocmDataType_A = typename RocmEquivalentType::Type; @@ -346,7 +346,7 @@ inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define HEEVD_LAUNCHER(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, \ + void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &w, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ heevd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, scratchpad, \ @@ -360,7 +360,7 @@ HEEVD_LAUNCHER(std::complex, double, rocsolver_zheevd) template inline void hegvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -391,8 +391,8 @@ inline void hegvd(const char *func_name, Func func, sycl::queue &queue, std::int } #define HEGVD_LAUNCHER(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ + void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ hegvd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, ldb, \ @@ -405,7 +405,7 @@ HEGVD_LAUNCHER(std::complex, double, rocsolver_zhegvd) #undef HEGVD_LAUNCHER template -inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -431,7 +431,7 @@ inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define HETRD_LAUNCHER(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, \ + void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &d, \ sycl::buffer &e, sycl::buffer &tau, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ @@ -444,12 +444,12 @@ HETRD_LAUNCHER(std::complex, double, rocsolver_zhetrd) #undef HETRD_LAUNCHER -void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "hetrf"); } -void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { @@ -457,7 +457,7 @@ void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, } template -inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::generate vec, +inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -478,7 +478,7 @@ inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORGBR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, \ + void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ orgbr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, scratchpad, \ @@ -523,7 +523,7 @@ ORGQR_LAUNCHER(double, rocsolver_dorgqr) #undef ORGQR_LAUNCHER template -inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; @@ -543,7 +543,7 @@ inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORGTR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ orgtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ @@ -556,8 +556,8 @@ ORGTR_LAUNCHER(double, rocsolver_dorgtr) #undef ORGTR_LAUNCHER template -inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -581,8 +581,8 @@ inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORMTR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, \ - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, \ + void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ @@ -595,13 +595,13 @@ ORMTR_LAUNCHER(double, rocsolver_dormtr) #undef ORMTR_LAUNCHER -void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "ormrq"); } -void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -609,8 +609,8 @@ void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose tr } template -inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; @@ -633,7 +633,7 @@ inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORMQR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, \ + void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ @@ -647,7 +647,7 @@ ORMQR_LAUNCHER(double, rocsolver_dormqr) #undef ORMQR_LAUNCHER template -inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; @@ -669,7 +669,7 @@ inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define POTRF_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ potrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ scratchpad_size); \ @@ -683,7 +683,7 @@ POTRF_LAUNCHER(std::complex, rocsolver_zpotrf) #undef POTRF_LAUNCHER template -inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; @@ -705,7 +705,7 @@ inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define POTRI_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ potri(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ scratchpad_size); \ @@ -719,7 +719,7 @@ POTRI_LAUNCHER(std::complex, rocsolver_zpotri) #undef POTRI_LAUNCHER template -inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -740,7 +740,7 @@ inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define POTRS_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, \ + void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ potrs(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, nrhs, a, lda, b, ldb, \ @@ -755,8 +755,8 @@ POTRS_LAUNCHER(std::complex, rocsolver_zpotrs) #undef POTRS_LAUNCHER template -inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, +inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); @@ -782,7 +782,7 @@ inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define SYEVD_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, \ + void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &w, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ syevd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, scratchpad, \ @@ -796,7 +796,7 @@ SYEVD_LAUNCHER(double, rocsolver_dsyevd) template inline void sygvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; @@ -825,8 +825,8 @@ inline void sygvd(const char *func_name, Func func, sycl::queue &queue, std::int } #define SYGVD_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ + void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ sygvd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, ldb, \ @@ -839,7 +839,7 @@ SYGVD_LAUNCHER(double, rocsolver_dsygvd) #undef SYGVD_LAUNCH template -inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -864,7 +864,7 @@ inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define SYTRD_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &d, sycl::buffer &e, \ sycl::buffer &tau, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ @@ -878,7 +878,7 @@ SYTRD_LAUNCHER(double, rocsolver_dsytrd) #undef SYTRD_LAUNCHER template -inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -920,7 +920,7 @@ inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define SYTRF_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ sytrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, ipiv, scratchpad, \ @@ -934,27 +934,27 @@ SYTRF_LAUNCHER(std::complex, rocsolver_zsytrf) #undef SYTRF_LAUNCHER -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { @@ -962,7 +962,7 @@ void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose tr } template -inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::generate vec, +inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -983,7 +983,7 @@ inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNGBR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, \ + void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ ungbr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, scratchpad, \ @@ -1028,7 +1028,7 @@ UNGQR_LAUNCHER(std::complex, rocsolver_zungqr) #undef UNGQR_LAUNCHER template -inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::uplo uplo, +inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; @@ -1048,7 +1048,7 @@ inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNGTR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, \ + void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ ungtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ @@ -1060,14 +1060,14 @@ UNGTR_LAUNCHER(std::complex, rocsolver_zungtr) #undef UNGTR_LAUNCHER -void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "unmrq"); } -void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, @@ -1076,8 +1076,8 @@ void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose tr } template -inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; @@ -1100,7 +1100,7 @@ inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNMQR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, \ + void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, \ std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, \ sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ @@ -1114,8 +1114,8 @@ UNMQR_LAUNCHER(std::complex, rocsolver_zunmqr) #undef UNMQR_LAUNCHER template -inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { @@ -1139,8 +1139,8 @@ inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNMTR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, \ - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, \ + void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, \ std::int64_t scratchpad_size) { \ @@ -1343,7 +1343,7 @@ sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, s template inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, T *a, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, T *a, std::int64_t lda, std::int64_t *ipiv, T *b, std::int64_t ldb, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1387,7 +1387,7 @@ inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, } #define GETRS_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, \ + sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ std::int64_t nrhs, TYPE *a, std::int64_t lda, std::int64_t *ipiv, TYPE *b, \ std::int64_t ldb, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1404,7 +1404,7 @@ GETRS_LAUNCHER_USM(std::complex, rocsolver_zgetrs) template inline sycl::event gesvd(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, T_A *a, std::int64_t lda, T_B *s, T_A *u, std::int64_t ldu, T_A *vt, std::int64_t ldvt, T_A *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1438,7 +1438,7 @@ inline sycl::event gesvd(const char *func_name, Func func, sycl::queue &queue, } #define GESVD_LAUNCHER_USM(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, \ + sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ std::int64_t m, std::int64_t n, TYPE_A *a, std::int64_t lda, TYPE_B *s, \ TYPE_A *u, std::int64_t ldu, TYPE_A *vt, std::int64_t ldvt, \ TYPE_A *scratchpad, std::int64_t scratchpad_size, \ @@ -1456,7 +1456,7 @@ GESVD_LAUNCHER_USM(std::complex, double, rocsolver_zgesvd) template inline sycl::event heevd(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, T_A *&a, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A *&a, std::int64_t lda, T_B *&w, T_A *&scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using rocmDataType_A = typename RocmEquivalentType::Type; @@ -1486,7 +1486,7 @@ inline sycl::event heevd(const char *func_name, Func func, sycl::queue &queue, } #define HEEVD_LAUNCHER_USM(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, \ + sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, TYPE_A *a, std::int64_t lda, TYPE_B *w, TYPE_A *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1501,7 +1501,7 @@ HEEVD_LAUNCHER_USM(std::complex, double, rocsolver_zheevd) template inline sycl::event hegvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, T_A *&a, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A *&a, std::int64_t lda, T_A *&b, std::int64_t ldb, T_B *&w, T_A *&scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1533,8 +1533,8 @@ inline sycl::event hegvd(const char *func_name, Func func, sycl::queue &queue, s } #define HEGVD_LAUNCHER_USM(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, TYPE_A *a, std::int64_t lda, \ + sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, TYPE_A *a, std::int64_t lda, \ TYPE_A *b, std::int64_t ldb, TYPE_B *w, TYPE_A *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1549,7 +1549,7 @@ HEGVD_LAUNCHER_USM(std::complex, double, rocsolver_zhegvd) template inline sycl::event hetrd(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T_A *a, std::int64_t lda, T_B *d, + oneapi::math::uplo uplo, std::int64_t n, T_A *a, std::int64_t lda, T_B *d, T_B *e, T_A *tau, T_A *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using rocmDataType_A = typename RocmEquivalentType::Type; @@ -1575,7 +1575,7 @@ inline sycl::event hetrd(const char *func_name, Func func, sycl::queue &queue, } #define HETRD_LAUNCHER_USM(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE_A *a, \ + sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE_A *a, \ std::int64_t lda, TYPE_B *d, TYPE_B *e, TYPE_A *tau, TYPE_A *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1588,13 +1588,13 @@ HETRD_LAUNCHER_USM(std::complex, double, rocsolver_zhetrd) #undef HETRD_LAUNCHER_USM -sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "hetrf"); } -sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1603,7 +1603,7 @@ sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, template inline sycl::event orgbr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1627,7 +1627,7 @@ inline sycl::event orgbr(const char *func_name, Func func, sycl::queue &queue, } #define ORGBR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, \ + sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, \ std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, TYPE *tau, \ TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1678,7 +1678,7 @@ ORGQR_LAUNCHER_USM(double, rocsolver_dorgqr) template inline sycl::event orgtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using rocmDataType = typename RocmEquivalentType::Type; @@ -1701,7 +1701,7 @@ inline sycl::event orgtr(const char *func_name, Func func, sycl::queue &queue, } #define ORGTR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ return orgtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, \ @@ -1715,8 +1715,8 @@ ORGTR_LAUNCHER_USM(double, rocsolver_dorgtr) template inline sycl::event ormtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, T *a, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1742,8 +1742,8 @@ inline sycl::event ormtr(const char *func_name, Func func, sycl::queue &queue, } #define ORMTR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, \ - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ + sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1756,13 +1756,13 @@ ORMTR_LAUNCHER_USM(double, rocsolver_dormtr) #undef ORMTR_LAUNCHER_USM -sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "ormrq"); } -sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1771,7 +1771,7 @@ sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::trans template inline sycl::event ormqr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1797,7 +1797,7 @@ inline sycl::event ormqr(const char *func_name, Func func, sycl::queue &queue, } #define ORMQR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, \ + sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, \ TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ @@ -1813,7 +1813,7 @@ ORMQR_LAUNCHER_USM(double, rocsolver_dormqr) template inline sycl::event potrf(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using rocmDataType = typename RocmEquivalentType::Type; @@ -1839,7 +1839,7 @@ inline sycl::event potrf(const char *func_name, Func func, sycl::queue &queue, } #define POTRF_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ return potrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ @@ -1855,7 +1855,7 @@ POTRF_LAUNCHER_USM(std::complex, rocsolver_zpotrf) template inline sycl::event potri(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using rocmDataType = typename RocmEquivalentType::Type; @@ -1882,7 +1882,7 @@ inline sycl::event potri(const char *func_name, Func func, sycl::queue &queue, } #define POTRI_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ return potri(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ @@ -1898,7 +1898,7 @@ POTRI_LAUNCHER_USM(std::complex, rocsolver_zpotri) template inline sycl::event potrs(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, T *a, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, T *a, std::int64_t lda, T *b, std::int64_t ldb, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1922,7 +1922,7 @@ inline sycl::event potrs(const char *func_name, Func func, sycl::queue &queue, } #define POTRS_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, \ + sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ std::int64_t nrhs, TYPE *a, std::int64_t lda, TYPE *b, std::int64_t ldb, \ TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1939,7 +1939,7 @@ POTRS_LAUNCHER_USM(std::complex, rocsolver_zpotrs) template inline sycl::event syevd(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, T *a, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *w, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using rocmDataType = typename RocmEquivalentType::Type; @@ -1968,7 +1968,7 @@ inline sycl::event syevd(const char *func_name, Func func, sycl::queue &queue, } #define SYEVD_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, \ + sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, TYPE *a, std::int64_t lda, TYPE *w, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -1983,7 +1983,7 @@ SYEVD_LAUNCHER_USM(double, rocsolver_dsyevd) template inline sycl::event sygvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, T *a, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *b, std::int64_t ldb, T *w, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -2014,8 +2014,8 @@ inline sycl::event sygvd(const char *func_name, Func func, sycl::queue &queue, s } #define SYGVD_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, std::int64_t lda, TYPE *b, \ + sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, TYPE *a, std::int64_t lda, TYPE *b, \ std::int64_t ldb, TYPE *w, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ return sygvd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, \ @@ -2029,7 +2029,7 @@ SYGVD_LAUNCHER_USM(double, rocsolver_dsygvd) template inline sycl::event sytrd(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *d, T *e, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *d, T *e, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using rocmDataType = typename RocmEquivalentType::Type; @@ -2054,7 +2054,7 @@ inline sycl::event sytrd(const char *func_name, Func func, sycl::queue &queue, } #define SYTRD_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *d, TYPE *e, TYPE *tau, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -2069,7 +2069,7 @@ SYTRD_LAUNCHER_USM(double, rocsolver_dsytrd) template inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, std::int64_t *ipiv, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using rocmDataType = typename RocmEquivalentType::Type; @@ -2113,7 +2113,7 @@ inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, } #define SYTRF_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, std::int64_t *ipiv, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -2128,27 +2128,27 @@ SYTRF_LAUNCHER_USM(std::complex, rocsolver_zsytrf) #undef SYTRF_LAUNCHER_USM -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, double *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -2157,7 +2157,7 @@ sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::trans template inline sycl::event ungbr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -2181,7 +2181,7 @@ inline sycl::event ungbr(const char *func_name, Func func, sycl::queue &queue, } #define UNGBR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, \ + sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, \ std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, TYPE *tau, \ TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -2232,7 +2232,7 @@ UNGQR_LAUNCHER_USM(std::complex, rocsolver_zungqr) template inline sycl::event ungtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, + oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { using rocmDataType = typename RocmEquivalentType::Type; @@ -2255,7 +2255,7 @@ inline sycl::event ungtr(const char *func_name, Func func, sycl::queue &queue, } #define UNGTR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ return ungtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, \ @@ -2267,14 +2267,14 @@ UNGTR_LAUNCHER_USM(std::complex, rocsolver_zungtr) #undef UNGTR_LAUNCHER_USM -sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { throw unimplemented("lapack", "unmrq"); } -sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -2284,7 +2284,7 @@ sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::trans template inline sycl::event unmqr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -2310,7 +2310,7 @@ inline sycl::event unmqr(const char *func_name, Func func, sycl::queue &queue, } #define UNMQR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, \ + sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, \ TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ @@ -2326,8 +2326,8 @@ UNMQR_LAUNCHER_USM(std::complex, rocsolver_zunmqr) template inline sycl::event unmtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, T *a, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T *a, std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -2353,8 +2353,8 @@ inline sycl::event unmtr(const char *func_name, Func func, sycl::queue &queue, } #define UNMTR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, \ - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ + sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ std::int64_t lda, TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ std::int64_t scratchpad_size, \ const std::vector &dependencies) { \ @@ -2421,7 +2421,7 @@ GEQRF_LAUNCHER_SCRATCH(std::complex) #define GESVD_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t gesvd_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, \ + sycl::queue & queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, \ std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { \ return std::min(m, n) - 1; \ } @@ -2468,7 +2468,7 @@ std::int64_t getri_scratchpad_size>(sycl::queue &queue, std #define GETRS_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t getrs_scratchpad_size(sycl::queue & queue, oneapi::mkl::transpose trans, \ + std::int64_t getrs_scratchpad_size(sycl::queue & queue, oneapi::math::transpose trans, \ std::int64_t n, std::int64_t nrhs, std::int64_t lda, \ std::int64_t ldb) { \ return 0; \ @@ -2483,8 +2483,8 @@ GETRS_LAUNCHER_SCRATCH(std::complex) #define HEEVD_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t heevd_scratchpad_size(sycl::queue & queue, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, \ + std::int64_t heevd_scratchpad_size(sycl::queue & queue, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, \ std::int64_t lda) { \ return n; \ } @@ -2497,7 +2497,7 @@ HEEVD_LAUNCHER_SCRATCH(std::complex) #define HEGVD_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t hegvd_scratchpad_size(sycl::queue & queue, std::int64_t itype, \ - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, \ + oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda, std::int64_t ldb) { \ return n; \ } @@ -2509,7 +2509,7 @@ HEGVD_LAUNCHER_SCRATCH(std::complex) #define HETRD_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t hetrd_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t hetrd_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ return 0; \ } @@ -2520,19 +2520,19 @@ HETRD_LAUNCHER_SCRATCH(std::complex) #undef HETRD_LAUNCHER_SCRATCH template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "hetrf_scratchpad_size"); } template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "hetrf_scratchpad_size"); } #define ORGBR_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t orgbr_scratchpad_size(sycl::queue & queue, oneapi::mkl::generate vec, \ + std::int64_t orgbr_scratchpad_size(sycl::queue & queue, oneapi::math::generate vec, \ std::int64_t m, std::int64_t n, std::int64_t k, \ std::int64_t lda) { \ return 0; \ @@ -2545,7 +2545,7 @@ ORGBR_LAUNCHER_SCRATCH(double) #define ORGTR_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t orgtr_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t orgtr_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ return 0; \ } @@ -2568,15 +2568,15 @@ ORGQR_LAUNCHER_SCRATCH(double) #undef ORGQR_LAUNCHER_SCRATCH template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { throw unimplemented("lapack", "ormrq_scratchpad_size"); } template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { throw unimplemented("lapack", "ormrq_scratchpad_size"); @@ -2585,7 +2585,7 @@ std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side #define ORMQRF_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t ormqr_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, \ + sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, \ std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ return 0; \ } @@ -2597,8 +2597,8 @@ ORMQRF_LAUNCHER_SCRATCH(double) #define ORMTR_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t ormtr_scratchpad_size(sycl::queue & queue, oneapi::mkl::side side, \ - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, \ + std::int64_t ormtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ + oneapi::math::uplo uplo, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t lda, \ std::int64_t ldc) { \ return 0; \ @@ -2611,7 +2611,7 @@ ORMTR_LAUNCHER_SCRATCH(double) #define POTRF_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t potrf_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t potrf_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ return 0; \ } @@ -2625,7 +2625,7 @@ POTRF_LAUNCHER_SCRATCH(std::complex) #define POTRS_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t potrs_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t potrs_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t nrhs, std::int64_t lda, \ std::int64_t ldb) { \ return 0; \ @@ -2640,7 +2640,7 @@ POTRS_LAUNCHER_SCRATCH(std::complex) #define POTRI_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t potri_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t potri_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ return 0; \ } @@ -2654,7 +2654,7 @@ POTRI_LAUNCHER_SCRATCH(std::complex) #define SYTRF_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t sytrf_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t sytrf_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ return 0; \ } @@ -2668,8 +2668,8 @@ SYTRF_LAUNCHER_SCRATCH(std::complex) #define SYEVD_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t syevd_scratchpad_size(sycl::queue & queue, oneapi::mkl::job jobz, \ - oneapi::mkl::uplo uplo, std::int64_t n, \ + std::int64_t syevd_scratchpad_size(sycl::queue & queue, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, \ std::int64_t lda) { \ return n; \ } @@ -2682,7 +2682,7 @@ SYEVD_LAUNCHER_SCRATCH(double) #define SYGVD_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t sygvd_scratchpad_size(sycl::queue & queue, std::int64_t itype, \ - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, \ + oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda, std::int64_t ldb) { \ return n; \ } @@ -2694,7 +2694,7 @@ SYGVD_LAUNCHER_SCRATCH(double) #define SYTRD_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t sytrd_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t sytrd_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ return 0; \ } @@ -2706,8 +2706,8 @@ SYTRD_LAUNCHER_SCRATCH(double) #define TRTRS_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t trtrs_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, \ + std::int64_t trtrs_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, oneapi::math::diag diag, \ std::int64_t n, std::int64_t nrhs, std::int64_t lda, \ std::int64_t ldb) { \ return 0; \ @@ -2722,7 +2722,7 @@ TRTRS_LAUNCHER_SCRATCH(std::complex) #define UNGBR_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t ungbr_scratchpad_size(sycl::queue & queue, oneapi::mkl::generate vec, \ + std::int64_t ungbr_scratchpad_size(sycl::queue & queue, oneapi::math::generate vec, \ std::int64_t m, std::int64_t n, std::int64_t k, \ std::int64_t lda) { \ return 0; \ @@ -2747,7 +2747,7 @@ UNGQR_LAUNCHER_SCRATCH(std::complex) #define UNGTR_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t ungtr_scratchpad_size(sycl::queue & queue, oneapi::mkl::uplo uplo, \ + std::int64_t ungtr_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda) { \ return 0; \ } @@ -2758,16 +2758,16 @@ UNGTR_LAUNCHER_SCRATCH(std::complex) #undef UNGTR_LAUNCHER_SCRATCH template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { throw unimplemented("lapack", "unmrq_scratchpad_size"); } template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { @@ -2777,7 +2777,7 @@ std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, one #define UNMQR_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t unmqr_scratchpad_size( \ - sycl::queue & queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, \ + sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, \ std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ return 0; \ } @@ -2789,8 +2789,8 @@ UNMQR_LAUNCHER_SCRATCH(std::complex) #define UNMTR_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t unmtr_scratchpad_size(sycl::queue & queue, oneapi::mkl::side side, \ - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, \ + std::int64_t unmtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ + oneapi::math::uplo uplo, oneapi::math::transpose trans, \ std::int64_t m, std::int64_t n, std::int64_t lda, \ std::int64_t ldc) { \ return 0; \ @@ -2803,5 +2803,5 @@ UNMTR_LAUNCHER_SCRATCH(std::complex) } // namespace rocsolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp b/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp index ee3de2367..306c43f38 100644 --- a/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp @@ -26,7 +26,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace rocsolver { @@ -151,5 +151,5 @@ sycl::context RocsolverScopedContextHandler::get_context(const sycl::queue &queu } // namespace rocsolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp b/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp index 98368f0db..e1036e034 100644 --- a/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp @@ -46,7 +46,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace rocsolver { @@ -80,6 +80,6 @@ class RocsolverScopedContextHandler { } // namespace rocsolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_ROCSOLVER_SCOPED_HANDLE_HPP_ diff --git a/src/lapack/backends/rocsolver/rocsolver_task.hpp b/src/lapack/backends/rocsolver/rocsolver_task.hpp index 965484690..75014c130 100644 --- a/src/lapack/backends/rocsolver/rocsolver_task.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_task.hpp @@ -46,7 +46,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace rocsolver { @@ -69,6 +69,6 @@ static inline void onemath_rocsolver_host_task(H &cgh, sycl::queue queue, F f) { } // namespace rocsolver } // namespace lapack -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _MKL_LAPACK_ROCSOLVER_TASK_HPP_ diff --git a/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp b/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp index 539f46922..506963182 100644 --- a/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp @@ -26,403 +26,403 @@ extern "C" ONEMATH_EXPORT lapack_function_table_t mkl_lapack_table = { WRAPPER_VERSION, #define LAPACK_BACKEND rocsolver - oneapi::mkl::lapack::rocsolver::gebrd, - oneapi::mkl::lapack::rocsolver::gebrd, - oneapi::mkl::lapack::rocsolver::gebrd, - oneapi::mkl::lapack::rocsolver::gebrd, - oneapi::mkl::lapack::rocsolver::gerqf, - oneapi::mkl::lapack::rocsolver::gerqf, - oneapi::mkl::lapack::rocsolver::gerqf, - oneapi::mkl::lapack::rocsolver::gerqf, - oneapi::mkl::lapack::rocsolver::geqrf, - oneapi::mkl::lapack::rocsolver::geqrf, - oneapi::mkl::lapack::rocsolver::geqrf, - oneapi::mkl::lapack::rocsolver::geqrf, - oneapi::mkl::lapack::rocsolver::getrf, - oneapi::mkl::lapack::rocsolver::getrf, - oneapi::mkl::lapack::rocsolver::getrf, - oneapi::mkl::lapack::rocsolver::getrf, - oneapi::mkl::lapack::rocsolver::getri, - oneapi::mkl::lapack::rocsolver::getri, - oneapi::mkl::lapack::rocsolver::getri, - oneapi::mkl::lapack::rocsolver::getri, - oneapi::mkl::lapack::rocsolver::getrs, - oneapi::mkl::lapack::rocsolver::getrs, - oneapi::mkl::lapack::rocsolver::getrs, - oneapi::mkl::lapack::rocsolver::getrs, - oneapi::mkl::lapack::rocsolver::gesvd, - oneapi::mkl::lapack::rocsolver::gesvd, - oneapi::mkl::lapack::rocsolver::gesvd, - oneapi::mkl::lapack::rocsolver::gesvd, - oneapi::mkl::lapack::rocsolver::heevd, - oneapi::mkl::lapack::rocsolver::heevd, - oneapi::mkl::lapack::rocsolver::hegvd, - oneapi::mkl::lapack::rocsolver::hegvd, - oneapi::mkl::lapack::rocsolver::hetrd, - oneapi::mkl::lapack::rocsolver::hetrd, - oneapi::mkl::lapack::rocsolver::hetrf, - oneapi::mkl::lapack::rocsolver::hetrf, - oneapi::mkl::lapack::rocsolver::orgbr, - oneapi::mkl::lapack::rocsolver::orgbr, - oneapi::mkl::lapack::rocsolver::orgqr, - oneapi::mkl::lapack::rocsolver::orgqr, - oneapi::mkl::lapack::rocsolver::orgtr, - oneapi::mkl::lapack::rocsolver::orgtr, - oneapi::mkl::lapack::rocsolver::ormtr, - oneapi::mkl::lapack::rocsolver::ormtr, - oneapi::mkl::lapack::rocsolver::ormrq, - oneapi::mkl::lapack::rocsolver::ormrq, - oneapi::mkl::lapack::rocsolver::ormqr, - oneapi::mkl::lapack::rocsolver::ormqr, - oneapi::mkl::lapack::rocsolver::potrf, - oneapi::mkl::lapack::rocsolver::potrf, - oneapi::mkl::lapack::rocsolver::potrf, - oneapi::mkl::lapack::rocsolver::potrf, - oneapi::mkl::lapack::rocsolver::potri, - oneapi::mkl::lapack::rocsolver::potri, - oneapi::mkl::lapack::rocsolver::potri, - oneapi::mkl::lapack::rocsolver::potri, - oneapi::mkl::lapack::rocsolver::potrs, - oneapi::mkl::lapack::rocsolver::potrs, - oneapi::mkl::lapack::rocsolver::potrs, - oneapi::mkl::lapack::rocsolver::potrs, - oneapi::mkl::lapack::rocsolver::syevd, - oneapi::mkl::lapack::rocsolver::syevd, - oneapi::mkl::lapack::rocsolver::sygvd, - oneapi::mkl::lapack::rocsolver::sygvd, - oneapi::mkl::lapack::rocsolver::sytrd, - oneapi::mkl::lapack::rocsolver::sytrd, - oneapi::mkl::lapack::rocsolver::sytrf, - oneapi::mkl::lapack::rocsolver::sytrf, - oneapi::mkl::lapack::rocsolver::sytrf, - oneapi::mkl::lapack::rocsolver::sytrf, - oneapi::mkl::lapack::rocsolver::trtrs, - oneapi::mkl::lapack::rocsolver::trtrs, - oneapi::mkl::lapack::rocsolver::trtrs, - oneapi::mkl::lapack::rocsolver::trtrs, - oneapi::mkl::lapack::rocsolver::ungbr, - oneapi::mkl::lapack::rocsolver::ungbr, - oneapi::mkl::lapack::rocsolver::ungqr, - oneapi::mkl::lapack::rocsolver::ungqr, - oneapi::mkl::lapack::rocsolver::ungtr, - oneapi::mkl::lapack::rocsolver::ungtr, - oneapi::mkl::lapack::rocsolver::unmrq, - oneapi::mkl::lapack::rocsolver::unmrq, - oneapi::mkl::lapack::rocsolver::unmqr, - oneapi::mkl::lapack::rocsolver::unmqr, - oneapi::mkl::lapack::rocsolver::unmtr, - oneapi::mkl::lapack::rocsolver::unmtr, - oneapi::mkl::lapack::rocsolver::gebrd, - oneapi::mkl::lapack::rocsolver::gebrd, - oneapi::mkl::lapack::rocsolver::gebrd, - oneapi::mkl::lapack::rocsolver::gebrd, - oneapi::mkl::lapack::rocsolver::gerqf, - oneapi::mkl::lapack::rocsolver::gerqf, - oneapi::mkl::lapack::rocsolver::gerqf, - oneapi::mkl::lapack::rocsolver::gerqf, - oneapi::mkl::lapack::rocsolver::geqrf, - oneapi::mkl::lapack::rocsolver::geqrf, - oneapi::mkl::lapack::rocsolver::geqrf, - oneapi::mkl::lapack::rocsolver::geqrf, - oneapi::mkl::lapack::rocsolver::getrf, - oneapi::mkl::lapack::rocsolver::getrf, - oneapi::mkl::lapack::rocsolver::getrf, - oneapi::mkl::lapack::rocsolver::getrf, - oneapi::mkl::lapack::rocsolver::getri, - oneapi::mkl::lapack::rocsolver::getri, - oneapi::mkl::lapack::rocsolver::getri, - oneapi::mkl::lapack::rocsolver::getri, - oneapi::mkl::lapack::rocsolver::getrs, - oneapi::mkl::lapack::rocsolver::getrs, - oneapi::mkl::lapack::rocsolver::getrs, - oneapi::mkl::lapack::rocsolver::getrs, - oneapi::mkl::lapack::rocsolver::gesvd, - oneapi::mkl::lapack::rocsolver::gesvd, - oneapi::mkl::lapack::rocsolver::gesvd, - oneapi::mkl::lapack::rocsolver::gesvd, - oneapi::mkl::lapack::rocsolver::heevd, - oneapi::mkl::lapack::rocsolver::heevd, - oneapi::mkl::lapack::rocsolver::hegvd, - oneapi::mkl::lapack::rocsolver::hegvd, - oneapi::mkl::lapack::rocsolver::hetrd, - oneapi::mkl::lapack::rocsolver::hetrd, - oneapi::mkl::lapack::rocsolver::hetrf, - oneapi::mkl::lapack::rocsolver::hetrf, - oneapi::mkl::lapack::rocsolver::orgbr, - oneapi::mkl::lapack::rocsolver::orgbr, - oneapi::mkl::lapack::rocsolver::orgqr, - oneapi::mkl::lapack::rocsolver::orgqr, - oneapi::mkl::lapack::rocsolver::orgtr, - oneapi::mkl::lapack::rocsolver::orgtr, - oneapi::mkl::lapack::rocsolver::ormtr, - oneapi::mkl::lapack::rocsolver::ormtr, - oneapi::mkl::lapack::rocsolver::ormrq, - oneapi::mkl::lapack::rocsolver::ormrq, - oneapi::mkl::lapack::rocsolver::ormqr, - oneapi::mkl::lapack::rocsolver::ormqr, - oneapi::mkl::lapack::rocsolver::potrf, - oneapi::mkl::lapack::rocsolver::potrf, - oneapi::mkl::lapack::rocsolver::potrf, - oneapi::mkl::lapack::rocsolver::potrf, - oneapi::mkl::lapack::rocsolver::potri, - oneapi::mkl::lapack::rocsolver::potri, - oneapi::mkl::lapack::rocsolver::potri, - oneapi::mkl::lapack::rocsolver::potri, - oneapi::mkl::lapack::rocsolver::potrs, - oneapi::mkl::lapack::rocsolver::potrs, - oneapi::mkl::lapack::rocsolver::potrs, - oneapi::mkl::lapack::rocsolver::potrs, - oneapi::mkl::lapack::rocsolver::syevd, - oneapi::mkl::lapack::rocsolver::syevd, - oneapi::mkl::lapack::rocsolver::sygvd, - oneapi::mkl::lapack::rocsolver::sygvd, - oneapi::mkl::lapack::rocsolver::sytrd, - oneapi::mkl::lapack::rocsolver::sytrd, - oneapi::mkl::lapack::rocsolver::sytrf, - oneapi::mkl::lapack::rocsolver::sytrf, - oneapi::mkl::lapack::rocsolver::sytrf, - oneapi::mkl::lapack::rocsolver::sytrf, - oneapi::mkl::lapack::rocsolver::trtrs, - oneapi::mkl::lapack::rocsolver::trtrs, - oneapi::mkl::lapack::rocsolver::trtrs, - oneapi::mkl::lapack::rocsolver::trtrs, - oneapi::mkl::lapack::rocsolver::ungbr, - oneapi::mkl::lapack::rocsolver::ungbr, - oneapi::mkl::lapack::rocsolver::ungqr, - oneapi::mkl::lapack::rocsolver::ungqr, - oneapi::mkl::lapack::rocsolver::ungtr, - oneapi::mkl::lapack::rocsolver::ungtr, - oneapi::mkl::lapack::rocsolver::unmrq, - oneapi::mkl::lapack::rocsolver::unmrq, - oneapi::mkl::lapack::rocsolver::unmqr, - oneapi::mkl::lapack::rocsolver::unmqr, - oneapi::mkl::lapack::rocsolver::unmtr, - oneapi::mkl::lapack::rocsolver::unmtr, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::orgqr_batch, - oneapi::mkl::lapack::rocsolver::orgqr_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::ungqr_batch, - oneapi::mkl::lapack::rocsolver::ungqr_batch, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::orgqr_batch, - oneapi::mkl::lapack::rocsolver::orgqr_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::ungqr_batch, - oneapi::mkl::lapack::rocsolver::ungqr_batch, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::geqrf_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::getrf_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getri_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::getrs_batch, - oneapi::mkl::lapack::rocsolver::orgqr_batch, - oneapi::mkl::lapack::rocsolver::orgqr_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrf_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::potrs_batch, - oneapi::mkl::lapack::rocsolver::ungqr_batch, - oneapi::mkl::lapack::rocsolver::ungqr_batch, - oneapi::mkl::lapack::rocsolver::gebrd_scratchpad_size, - oneapi::mkl::lapack::rocsolver::gebrd_scratchpad_size, - oneapi::mkl::lapack::rocsolver::gebrd_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::gebrd_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::gerqf_scratchpad_size, - oneapi::mkl::lapack::rocsolver::gerqf_scratchpad_size, - oneapi::mkl::lapack::rocsolver::gerqf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::gerqf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::geqrf_scratchpad_size, - oneapi::mkl::lapack::rocsolver::geqrf_scratchpad_size, - oneapi::mkl::lapack::rocsolver::geqrf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::geqrf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::gesvd_scratchpad_size, - oneapi::mkl::lapack::rocsolver::gesvd_scratchpad_size, - oneapi::mkl::lapack::rocsolver::gesvd_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::gesvd_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrf_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrf_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getri_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getri_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getri_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getri_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrs_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrs_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrs_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrs_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::heevd_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::heevd_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::hegvd_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::hegvd_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::hetrd_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::hetrd_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::hetrf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::hetrf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::orgbr_scratchpad_size, - oneapi::mkl::lapack::rocsolver::orgbr_scratchpad_size, - oneapi::mkl::lapack::rocsolver::orgtr_scratchpad_size, - oneapi::mkl::lapack::rocsolver::orgtr_scratchpad_size, - oneapi::mkl::lapack::rocsolver::orgqr_scratchpad_size, - oneapi::mkl::lapack::rocsolver::orgqr_scratchpad_size, - oneapi::mkl::lapack::rocsolver::ormrq_scratchpad_size, - oneapi::mkl::lapack::rocsolver::ormrq_scratchpad_size, - oneapi::mkl::lapack::rocsolver::ormqr_scratchpad_size, - oneapi::mkl::lapack::rocsolver::ormqr_scratchpad_size, - oneapi::mkl::lapack::rocsolver::ormtr_scratchpad_size, - oneapi::mkl::lapack::rocsolver::ormtr_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrf_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrf_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potrf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potrs_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrs_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrs_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potrs_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potri_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potri_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potri_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potri_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::sytrf_scratchpad_size, - oneapi::mkl::lapack::rocsolver::sytrf_scratchpad_size, - oneapi::mkl::lapack::rocsolver::sytrf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::sytrf_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::syevd_scratchpad_size, - oneapi::mkl::lapack::rocsolver::syevd_scratchpad_size, - oneapi::mkl::lapack::rocsolver::sygvd_scratchpad_size, - oneapi::mkl::lapack::rocsolver::sygvd_scratchpad_size, - oneapi::mkl::lapack::rocsolver::sytrd_scratchpad_size, - oneapi::mkl::lapack::rocsolver::sytrd_scratchpad_size, - oneapi::mkl::lapack::rocsolver::trtrs_scratchpad_size, - oneapi::mkl::lapack::rocsolver::trtrs_scratchpad_size, - oneapi::mkl::lapack::rocsolver::trtrs_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::trtrs_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::ungbr_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::ungbr_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::ungqr_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::ungqr_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::ungtr_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::ungtr_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::unmrq_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::unmrq_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::unmqr_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::unmqr_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::unmtr_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::unmtr_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getri_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getri_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::ungqr_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::ungqr_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getri_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getri_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getri_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrs_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::getrs_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::geqrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::geqrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::orgqr_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrf_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potrf_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrs_batch_scratchpad_size, - oneapi::mkl::lapack::rocsolver::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::potrs_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::ungqr_batch_scratchpad_size>, - oneapi::mkl::lapack::rocsolver::ungqr_batch_scratchpad_size> + oneapi::math::lapack::rocsolver::gebrd, + oneapi::math::lapack::rocsolver::gebrd, + oneapi::math::lapack::rocsolver::gebrd, + oneapi::math::lapack::rocsolver::gebrd, + oneapi::math::lapack::rocsolver::gerqf, + oneapi::math::lapack::rocsolver::gerqf, + oneapi::math::lapack::rocsolver::gerqf, + oneapi::math::lapack::rocsolver::gerqf, + oneapi::math::lapack::rocsolver::geqrf, + oneapi::math::lapack::rocsolver::geqrf, + oneapi::math::lapack::rocsolver::geqrf, + oneapi::math::lapack::rocsolver::geqrf, + oneapi::math::lapack::rocsolver::getrf, + oneapi::math::lapack::rocsolver::getrf, + oneapi::math::lapack::rocsolver::getrf, + oneapi::math::lapack::rocsolver::getrf, + oneapi::math::lapack::rocsolver::getri, + oneapi::math::lapack::rocsolver::getri, + oneapi::math::lapack::rocsolver::getri, + oneapi::math::lapack::rocsolver::getri, + oneapi::math::lapack::rocsolver::getrs, + oneapi::math::lapack::rocsolver::getrs, + oneapi::math::lapack::rocsolver::getrs, + oneapi::math::lapack::rocsolver::getrs, + oneapi::math::lapack::rocsolver::gesvd, + oneapi::math::lapack::rocsolver::gesvd, + oneapi::math::lapack::rocsolver::gesvd, + oneapi::math::lapack::rocsolver::gesvd, + oneapi::math::lapack::rocsolver::heevd, + oneapi::math::lapack::rocsolver::heevd, + oneapi::math::lapack::rocsolver::hegvd, + oneapi::math::lapack::rocsolver::hegvd, + oneapi::math::lapack::rocsolver::hetrd, + oneapi::math::lapack::rocsolver::hetrd, + oneapi::math::lapack::rocsolver::hetrf, + oneapi::math::lapack::rocsolver::hetrf, + oneapi::math::lapack::rocsolver::orgbr, + oneapi::math::lapack::rocsolver::orgbr, + oneapi::math::lapack::rocsolver::orgqr, + oneapi::math::lapack::rocsolver::orgqr, + oneapi::math::lapack::rocsolver::orgtr, + oneapi::math::lapack::rocsolver::orgtr, + oneapi::math::lapack::rocsolver::ormtr, + oneapi::math::lapack::rocsolver::ormtr, + oneapi::math::lapack::rocsolver::ormrq, + oneapi::math::lapack::rocsolver::ormrq, + oneapi::math::lapack::rocsolver::ormqr, + oneapi::math::lapack::rocsolver::ormqr, + oneapi::math::lapack::rocsolver::potrf, + oneapi::math::lapack::rocsolver::potrf, + oneapi::math::lapack::rocsolver::potrf, + oneapi::math::lapack::rocsolver::potrf, + oneapi::math::lapack::rocsolver::potri, + oneapi::math::lapack::rocsolver::potri, + oneapi::math::lapack::rocsolver::potri, + oneapi::math::lapack::rocsolver::potri, + oneapi::math::lapack::rocsolver::potrs, + oneapi::math::lapack::rocsolver::potrs, + oneapi::math::lapack::rocsolver::potrs, + oneapi::math::lapack::rocsolver::potrs, + oneapi::math::lapack::rocsolver::syevd, + oneapi::math::lapack::rocsolver::syevd, + oneapi::math::lapack::rocsolver::sygvd, + oneapi::math::lapack::rocsolver::sygvd, + oneapi::math::lapack::rocsolver::sytrd, + oneapi::math::lapack::rocsolver::sytrd, + oneapi::math::lapack::rocsolver::sytrf, + oneapi::math::lapack::rocsolver::sytrf, + oneapi::math::lapack::rocsolver::sytrf, + oneapi::math::lapack::rocsolver::sytrf, + oneapi::math::lapack::rocsolver::trtrs, + oneapi::math::lapack::rocsolver::trtrs, + oneapi::math::lapack::rocsolver::trtrs, + oneapi::math::lapack::rocsolver::trtrs, + oneapi::math::lapack::rocsolver::ungbr, + oneapi::math::lapack::rocsolver::ungbr, + oneapi::math::lapack::rocsolver::ungqr, + oneapi::math::lapack::rocsolver::ungqr, + oneapi::math::lapack::rocsolver::ungtr, + oneapi::math::lapack::rocsolver::ungtr, + oneapi::math::lapack::rocsolver::unmrq, + oneapi::math::lapack::rocsolver::unmrq, + oneapi::math::lapack::rocsolver::unmqr, + oneapi::math::lapack::rocsolver::unmqr, + oneapi::math::lapack::rocsolver::unmtr, + oneapi::math::lapack::rocsolver::unmtr, + oneapi::math::lapack::rocsolver::gebrd, + oneapi::math::lapack::rocsolver::gebrd, + oneapi::math::lapack::rocsolver::gebrd, + oneapi::math::lapack::rocsolver::gebrd, + oneapi::math::lapack::rocsolver::gerqf, + oneapi::math::lapack::rocsolver::gerqf, + oneapi::math::lapack::rocsolver::gerqf, + oneapi::math::lapack::rocsolver::gerqf, + oneapi::math::lapack::rocsolver::geqrf, + oneapi::math::lapack::rocsolver::geqrf, + oneapi::math::lapack::rocsolver::geqrf, + oneapi::math::lapack::rocsolver::geqrf, + oneapi::math::lapack::rocsolver::getrf, + oneapi::math::lapack::rocsolver::getrf, + oneapi::math::lapack::rocsolver::getrf, + oneapi::math::lapack::rocsolver::getrf, + oneapi::math::lapack::rocsolver::getri, + oneapi::math::lapack::rocsolver::getri, + oneapi::math::lapack::rocsolver::getri, + oneapi::math::lapack::rocsolver::getri, + oneapi::math::lapack::rocsolver::getrs, + oneapi::math::lapack::rocsolver::getrs, + oneapi::math::lapack::rocsolver::getrs, + oneapi::math::lapack::rocsolver::getrs, + oneapi::math::lapack::rocsolver::gesvd, + oneapi::math::lapack::rocsolver::gesvd, + oneapi::math::lapack::rocsolver::gesvd, + oneapi::math::lapack::rocsolver::gesvd, + oneapi::math::lapack::rocsolver::heevd, + oneapi::math::lapack::rocsolver::heevd, + oneapi::math::lapack::rocsolver::hegvd, + oneapi::math::lapack::rocsolver::hegvd, + oneapi::math::lapack::rocsolver::hetrd, + oneapi::math::lapack::rocsolver::hetrd, + oneapi::math::lapack::rocsolver::hetrf, + oneapi::math::lapack::rocsolver::hetrf, + oneapi::math::lapack::rocsolver::orgbr, + oneapi::math::lapack::rocsolver::orgbr, + oneapi::math::lapack::rocsolver::orgqr, + oneapi::math::lapack::rocsolver::orgqr, + oneapi::math::lapack::rocsolver::orgtr, + oneapi::math::lapack::rocsolver::orgtr, + oneapi::math::lapack::rocsolver::ormtr, + oneapi::math::lapack::rocsolver::ormtr, + oneapi::math::lapack::rocsolver::ormrq, + oneapi::math::lapack::rocsolver::ormrq, + oneapi::math::lapack::rocsolver::ormqr, + oneapi::math::lapack::rocsolver::ormqr, + oneapi::math::lapack::rocsolver::potrf, + oneapi::math::lapack::rocsolver::potrf, + oneapi::math::lapack::rocsolver::potrf, + oneapi::math::lapack::rocsolver::potrf, + oneapi::math::lapack::rocsolver::potri, + oneapi::math::lapack::rocsolver::potri, + oneapi::math::lapack::rocsolver::potri, + oneapi::math::lapack::rocsolver::potri, + oneapi::math::lapack::rocsolver::potrs, + oneapi::math::lapack::rocsolver::potrs, + oneapi::math::lapack::rocsolver::potrs, + oneapi::math::lapack::rocsolver::potrs, + oneapi::math::lapack::rocsolver::syevd, + oneapi::math::lapack::rocsolver::syevd, + oneapi::math::lapack::rocsolver::sygvd, + oneapi::math::lapack::rocsolver::sygvd, + oneapi::math::lapack::rocsolver::sytrd, + oneapi::math::lapack::rocsolver::sytrd, + oneapi::math::lapack::rocsolver::sytrf, + oneapi::math::lapack::rocsolver::sytrf, + oneapi::math::lapack::rocsolver::sytrf, + oneapi::math::lapack::rocsolver::sytrf, + oneapi::math::lapack::rocsolver::trtrs, + oneapi::math::lapack::rocsolver::trtrs, + oneapi::math::lapack::rocsolver::trtrs, + oneapi::math::lapack::rocsolver::trtrs, + oneapi::math::lapack::rocsolver::ungbr, + oneapi::math::lapack::rocsolver::ungbr, + oneapi::math::lapack::rocsolver::ungqr, + oneapi::math::lapack::rocsolver::ungqr, + oneapi::math::lapack::rocsolver::ungtr, + oneapi::math::lapack::rocsolver::ungtr, + oneapi::math::lapack::rocsolver::unmrq, + oneapi::math::lapack::rocsolver::unmrq, + oneapi::math::lapack::rocsolver::unmqr, + oneapi::math::lapack::rocsolver::unmqr, + oneapi::math::lapack::rocsolver::unmtr, + oneapi::math::lapack::rocsolver::unmtr, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::orgqr_batch, + oneapi::math::lapack::rocsolver::orgqr_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::ungqr_batch, + oneapi::math::lapack::rocsolver::ungqr_batch, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::orgqr_batch, + oneapi::math::lapack::rocsolver::orgqr_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::ungqr_batch, + oneapi::math::lapack::rocsolver::ungqr_batch, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::geqrf_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::getrf_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getri_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::getrs_batch, + oneapi::math::lapack::rocsolver::orgqr_batch, + oneapi::math::lapack::rocsolver::orgqr_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrf_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::potrs_batch, + oneapi::math::lapack::rocsolver::ungqr_batch, + oneapi::math::lapack::rocsolver::ungqr_batch, + oneapi::math::lapack::rocsolver::gebrd_scratchpad_size, + oneapi::math::lapack::rocsolver::gebrd_scratchpad_size, + oneapi::math::lapack::rocsolver::gebrd_scratchpad_size>, + oneapi::math::lapack::rocsolver::gebrd_scratchpad_size>, + oneapi::math::lapack::rocsolver::gerqf_scratchpad_size, + oneapi::math::lapack::rocsolver::gerqf_scratchpad_size, + oneapi::math::lapack::rocsolver::gerqf_scratchpad_size>, + oneapi::math::lapack::rocsolver::gerqf_scratchpad_size>, + oneapi::math::lapack::rocsolver::geqrf_scratchpad_size, + oneapi::math::lapack::rocsolver::geqrf_scratchpad_size, + oneapi::math::lapack::rocsolver::geqrf_scratchpad_size>, + oneapi::math::lapack::rocsolver::geqrf_scratchpad_size>, + oneapi::math::lapack::rocsolver::gesvd_scratchpad_size, + oneapi::math::lapack::rocsolver::gesvd_scratchpad_size, + oneapi::math::lapack::rocsolver::gesvd_scratchpad_size>, + oneapi::math::lapack::rocsolver::gesvd_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrf_scratchpad_size, + oneapi::math::lapack::rocsolver::getrf_scratchpad_size, + oneapi::math::lapack::rocsolver::getrf_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrf_scratchpad_size>, + oneapi::math::lapack::rocsolver::getri_scratchpad_size, + oneapi::math::lapack::rocsolver::getri_scratchpad_size, + oneapi::math::lapack::rocsolver::getri_scratchpad_size>, + oneapi::math::lapack::rocsolver::getri_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrs_scratchpad_size, + oneapi::math::lapack::rocsolver::getrs_scratchpad_size, + oneapi::math::lapack::rocsolver::getrs_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrs_scratchpad_size>, + oneapi::math::lapack::rocsolver::heevd_scratchpad_size>, + oneapi::math::lapack::rocsolver::heevd_scratchpad_size>, + oneapi::math::lapack::rocsolver::hegvd_scratchpad_size>, + oneapi::math::lapack::rocsolver::hegvd_scratchpad_size>, + oneapi::math::lapack::rocsolver::hetrd_scratchpad_size>, + oneapi::math::lapack::rocsolver::hetrd_scratchpad_size>, + oneapi::math::lapack::rocsolver::hetrf_scratchpad_size>, + oneapi::math::lapack::rocsolver::hetrf_scratchpad_size>, + oneapi::math::lapack::rocsolver::orgbr_scratchpad_size, + oneapi::math::lapack::rocsolver::orgbr_scratchpad_size, + oneapi::math::lapack::rocsolver::orgtr_scratchpad_size, + oneapi::math::lapack::rocsolver::orgtr_scratchpad_size, + oneapi::math::lapack::rocsolver::orgqr_scratchpad_size, + oneapi::math::lapack::rocsolver::orgqr_scratchpad_size, + oneapi::math::lapack::rocsolver::ormrq_scratchpad_size, + oneapi::math::lapack::rocsolver::ormrq_scratchpad_size, + oneapi::math::lapack::rocsolver::ormqr_scratchpad_size, + oneapi::math::lapack::rocsolver::ormqr_scratchpad_size, + oneapi::math::lapack::rocsolver::ormtr_scratchpad_size, + oneapi::math::lapack::rocsolver::ormtr_scratchpad_size, + oneapi::math::lapack::rocsolver::potrf_scratchpad_size, + oneapi::math::lapack::rocsolver::potrf_scratchpad_size, + oneapi::math::lapack::rocsolver::potrf_scratchpad_size>, + oneapi::math::lapack::rocsolver::potrf_scratchpad_size>, + oneapi::math::lapack::rocsolver::potrs_scratchpad_size, + oneapi::math::lapack::rocsolver::potrs_scratchpad_size, + oneapi::math::lapack::rocsolver::potrs_scratchpad_size>, + oneapi::math::lapack::rocsolver::potrs_scratchpad_size>, + oneapi::math::lapack::rocsolver::potri_scratchpad_size, + oneapi::math::lapack::rocsolver::potri_scratchpad_size, + oneapi::math::lapack::rocsolver::potri_scratchpad_size>, + oneapi::math::lapack::rocsolver::potri_scratchpad_size>, + oneapi::math::lapack::rocsolver::sytrf_scratchpad_size, + oneapi::math::lapack::rocsolver::sytrf_scratchpad_size, + oneapi::math::lapack::rocsolver::sytrf_scratchpad_size>, + oneapi::math::lapack::rocsolver::sytrf_scratchpad_size>, + oneapi::math::lapack::rocsolver::syevd_scratchpad_size, + oneapi::math::lapack::rocsolver::syevd_scratchpad_size, + oneapi::math::lapack::rocsolver::sygvd_scratchpad_size, + oneapi::math::lapack::rocsolver::sygvd_scratchpad_size, + oneapi::math::lapack::rocsolver::sytrd_scratchpad_size, + oneapi::math::lapack::rocsolver::sytrd_scratchpad_size, + oneapi::math::lapack::rocsolver::trtrs_scratchpad_size, + oneapi::math::lapack::rocsolver::trtrs_scratchpad_size, + oneapi::math::lapack::rocsolver::trtrs_scratchpad_size>, + oneapi::math::lapack::rocsolver::trtrs_scratchpad_size>, + oneapi::math::lapack::rocsolver::ungbr_scratchpad_size>, + oneapi::math::lapack::rocsolver::ungbr_scratchpad_size>, + oneapi::math::lapack::rocsolver::ungqr_scratchpad_size>, + oneapi::math::lapack::rocsolver::ungqr_scratchpad_size>, + oneapi::math::lapack::rocsolver::ungtr_scratchpad_size>, + oneapi::math::lapack::rocsolver::ungtr_scratchpad_size>, + oneapi::math::lapack::rocsolver::unmrq_scratchpad_size>, + oneapi::math::lapack::rocsolver::unmrq_scratchpad_size>, + oneapi::math::lapack::rocsolver::unmqr_scratchpad_size>, + oneapi::math::lapack::rocsolver::unmqr_scratchpad_size>, + oneapi::math::lapack::rocsolver::unmtr_scratchpad_size>, + oneapi::math::lapack::rocsolver::unmtr_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::getri_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getri_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getri_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::getri_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrs_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getrs_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getrs_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrs_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::geqrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::geqrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::potrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::potrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::potrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::potrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::potrs_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::potrs_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::potrs_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::potrs_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::orgqr_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::orgqr_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::ungqr_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::ungqr_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::getri_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getri_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getri_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::getri_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrs_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getrs_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::getrs_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::getrs_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::geqrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::geqrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::geqrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::orgqr_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::orgqr_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::potrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::potrf_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::potrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::potrf_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::potrs_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::potrs_batch_scratchpad_size, + oneapi::math::lapack::rocsolver::potrs_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::potrs_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::ungqr_batch_scratchpad_size>, + oneapi::math::lapack::rocsolver::ungqr_batch_scratchpad_size> #undef LAPACK_BACKEND }; diff --git a/src/lapack/function_table.hpp b/src/lapack/function_table.hpp index 4a0b9293e..9f72752bd 100644 --- a/src/lapack/function_table.hpp +++ b/src/lapack/function_table.hpp @@ -116,95 +116,95 @@ typedef struct { std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*cgetrs_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + void (*cgetrs_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*dgetrs_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + void (*dgetrs_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sgetrs_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + void (*sgetrs_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*zgetrs_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + void (*zgetrs_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*dgesvd_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, + void (*dgesvd_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sgesvd_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, + void (*sgesvd_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*cgesvd_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, + void (*cgesvd_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zgesvd_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, + void (*zgesvd_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*cheevd_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + void (*cheevd_sycl)(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zheevd_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + void (*zheevd_sycl)(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*chegvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + void (*chegvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zhegvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + void (*zhegvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*chetrd_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*chetrd_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zhetrd_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*zhetrd_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*chetrf_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*chetrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zhetrf_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*zhetrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*sorgbr_sycl)(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, + void (*sorgbr_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dorgbr_sycl)(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, + void (*dorgbr_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); @@ -214,158 +214,158 @@ typedef struct { void (*sorgqr_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sorgtr_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*sorgtr_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dorgtr_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*dorgtr_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sormtr_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + void (*sormtr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dormtr_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + void (*dormtr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sormrq_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, + void (*sormrq_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dormrq_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, + void (*dormrq_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dormqr_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, + void (*dormqr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sormqr_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, + void (*sormqr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*spotrf_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*spotrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dpotrf_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*dpotrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*cpotrf_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*cpotrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zpotrf_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*zpotrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*spotri_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*spotri_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dpotri_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*dpotri_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*cpotri_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*cpotri_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zpotri_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*zpotri_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*spotrs_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*spotrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dpotrs_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*dpotrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*cpotrs_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*cpotrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zpotrs_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*zpotrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*dsyevd_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + void (*dsyevd_sycl)(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*ssyevd_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + void (*ssyevd_sycl)(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dsygvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, + void (*dsygvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*ssygvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, + void (*ssygvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dsytrd_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*dsytrd_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*ssytrd_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*ssytrd_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*ssytrf_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*ssytrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dsytrf_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*dsytrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*csytrf_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*csytrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zsytrf_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*zsytrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*ctrtrs_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + void (*ctrtrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*dtrtrs_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + void (*dtrtrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*strtrs_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + void (*strtrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*ztrtrs_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, + void (*ztrtrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*cungbr_sycl)(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, + void (*cungbr_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zungbr_sycl)(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, + void (*zungbr_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, @@ -380,53 +380,53 @@ typedef struct { sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*cungtr_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*cungtr_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zungtr_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*zungtr_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*cunmrq_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, + void (*cunmrq_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zunmrq_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, + void (*zunmrq_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*cunmqr_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, + void (*cunmqr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zunmqr_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, + void (*zunmqr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*cunmtr_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + void (*cunmtr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zunmtr_sycl)(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + void (*zunmtr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, @@ -518,97 +518,97 @@ typedef struct { std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cgetrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*cgetrs_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dgetrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*dgetrs_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*sgetrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*sgetrs_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zgetrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + sycl::event (*zgetrs_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dgesvd_usm_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, + sycl::event (*dgesvd_usm_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*sgesvd_usm_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, + sycl::event (*sgesvd_usm_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cgesvd_usm_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, + sycl::event (*cgesvd_usm_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zgesvd_usm_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, + sycl::event (*zgesvd_usm_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cheevd_usm_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + sycl::event (*cheevd_usm_sycl)(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zheevd_usm_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + sycl::event (*zheevd_usm_sycl)(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*chegvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + sycl::event (*chegvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zhegvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, + sycl::event (*zhegvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*chetrd_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*chetrd_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zhetrd_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*zhetrd_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*chetrf_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*chetrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zhetrf_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*zhetrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*sorgbr_usm_sycl)(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, + sycl::event (*sorgbr_usm_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dorgbr_usm_sycl)(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, + sycl::event (*dorgbr_usm_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); @@ -620,176 +620,176 @@ typedef struct { std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*sorgtr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*sorgtr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dorgtr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*dorgtr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*sormtr_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, + sycl::event (*sormtr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dormtr_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, + sycl::event (*dormtr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*sormrq_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::event (*sormrq_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dormrq_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::event (*dormrq_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dormqr_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::event (*dormqr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*sormqr_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::event (*sormqr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*spotrf_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*spotrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dpotrf_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*dpotrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cpotrf_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*cpotrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zpotrf_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*zpotrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*spotri_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*spotri_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dpotri_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*dpotri_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cpotri_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*cpotri_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zpotri_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*zpotri_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*spotrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*spotrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dpotrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*dpotrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cpotrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*cpotrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zpotrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*zpotrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dsyevd_usm_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, + sycl::event (*dsyevd_usm_sycl)(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*ssyevd_usm_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, + sycl::event (*ssyevd_usm_sycl)(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dsygvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, + sycl::event (*dsygvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*ssygvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, + sycl::event (*ssygvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dsytrd_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*dsytrd_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*ssytrd_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*ssytrd_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*ssytrf_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*ssytrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dsytrf_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*dsytrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*csytrf_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*csytrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zsytrf_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*zsytrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*ctrtrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, + sycl::event (*ctrtrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dtrtrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, + sycl::event (*dtrtrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*strtrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, + sycl::event (*strtrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*ztrtrs_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, + sycl::event (*ztrtrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cungbr_usm_sycl)(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, + sycl::event (*cungbr_usm_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zungbr_usm_sycl)(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, + sycl::event (*zungbr_usm_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -804,53 +804,53 @@ typedef struct { std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cungtr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*cungtr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zungtr_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*zungtr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cunmrq_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::event (*cunmrq_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zunmrq_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::event (*zunmrq_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cunmqr_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::event (*cunmqr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zunmqr_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, + sycl::event (*zunmqr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cunmtr_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, + sycl::event (*cunmtr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zunmtr_usm_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, + sycl::event (*zunmtr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, @@ -900,19 +900,19 @@ typedef struct { std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*sgetrs_batch_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + void (*sgetrs_batch_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dgetrs_batch_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + void (*dgetrs_batch_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*cgetrs_batch_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + void (*cgetrs_batch_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -920,7 +920,7 @@ typedef struct { std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zgetrs_batch_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + void (*zgetrs_batch_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, @@ -960,42 +960,42 @@ typedef struct { sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*spotrf_batch_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*spotrf_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dpotrf_batch_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*dpotrf_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*cpotrf_batch_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*cpotrf_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zpotrf_batch_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*zpotrf_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*spotrs_batch_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*spotrs_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dpotrs_batch_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*dpotrs_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*cpotrs_batch_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*cpotrs_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size); - void (*zpotrs_batch_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + void (*zpotrs_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, @@ -1091,7 +1091,7 @@ typedef struct { std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*sgetrs_batch_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*sgetrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, @@ -1099,7 +1099,7 @@ typedef struct { std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dgetrs_batch_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + sycl::event (*dgetrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, @@ -1108,13 +1108,13 @@ typedef struct { std::int64_t scratchpad_size, const std::vector &dependencies); sycl::event (*cgetrs_batch_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); sycl::event (*zgetrs_batch_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1131,41 +1131,41 @@ typedef struct { std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*spotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*spotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dpotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*dpotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cpotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*cpotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zpotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*zpotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*spotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*spotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dpotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*dpotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cpotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*cpotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, @@ -1173,7 +1173,7 @@ typedef struct { std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zpotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + sycl::event (*zpotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, @@ -1264,21 +1264,21 @@ typedef struct { std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*sgetrs_group_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose *trans, + sycl::event (*sgetrs_group_usm_sycl)(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dgetrs_group_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose *trans, + sycl::event (*dgetrs_group_usm_sycl)(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cgetrs_group_usm_sycl)(sycl::queue &queue, oneapi::mkl::transpose *trans, + sycl::event (*cgetrs_group_usm_sycl)(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, @@ -1287,7 +1287,7 @@ typedef struct { std::int64_t scratchpad_size, const std::vector &dependencies); sycl::event (*zgetrs_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1303,48 +1303,48 @@ typedef struct { std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*spotrf_group_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + sycl::event (*spotrf_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dpotrf_group_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + sycl::event (*dpotrf_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cpotrf_group_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + sycl::event (*cpotrf_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*zpotrf_group_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + sycl::event (*zpotrf_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*spotrs_group_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + sycl::event (*spotrs_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*dpotrs_group_usm_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + sycl::event (*dpotrs_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); sycl::event (*cpotrs_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); sycl::event (*zpotrs_group_usm_sycl)( - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies); @@ -1387,20 +1387,20 @@ typedef struct { std::int64_t lda); std::int64_t (*zgeqrf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*sgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + std::int64_t (*sgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); - std::int64_t (*dgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + std::int64_t (*dgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); - std::int64_t (*cgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + std::int64_t (*cgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); - std::int64_t (*zgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + std::int64_t (*zgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); std::int64_t (*sgetrf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, @@ -1419,185 +1419,185 @@ typedef struct { std::int64_t lda); std::int64_t (*zgetri_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t n, std::int64_t lda); - std::int64_t (*sgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + std::int64_t (*sgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*dgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + std::int64_t (*dgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*cgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + std::int64_t (*cgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*zgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::transpose trans, + std::int64_t (*zgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*cheevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + std::int64_t (*cheevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zheevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + std::int64_t (*zheevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); std::int64_t (*chegvd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); std::int64_t (*zhegvd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); - std::int64_t (*chetrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*chetrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zhetrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*zhetrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*chetrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*chetrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zhetrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*zhetrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*sorgbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::generate vect, + std::int64_t (*sorgbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*dorgbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::generate vect, + std::int64_t (*dorgbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*sorgtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*sorgtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dorgtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*dorgtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); std::int64_t (*sorgqr_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); std::int64_t (*dorgqr_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*sormrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*sormrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*dormrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*dormrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*sormqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*sormqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*dormqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*dormqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*sormtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*sormtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); - std::int64_t (*dormtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*dormtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); - std::int64_t (*spotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*spotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dpotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*dpotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*cpotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*cpotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zpotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*zpotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*spotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*spotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*dpotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*dpotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*cpotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*cpotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*zpotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*zpotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*spotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*spotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dpotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*dpotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*cpotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*cpotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zpotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*zpotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*ssytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*ssytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dsytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*dsytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*csytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*csytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zsytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*zsytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*ssyevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + std::int64_t (*ssyevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dsyevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + std::int64_t (*dsyevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); std::int64_t (*ssygvd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); std::int64_t (*dsygvd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); - std::int64_t (*ssytrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*ssytrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dsytrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*dsytrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*strtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, + std::int64_t (*strtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*dtrtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, + std::int64_t (*dtrtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*ctrtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, + std::int64_t (*ctrtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*ztrtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, + std::int64_t (*ztrtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*cungbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::generate vect, + std::int64_t (*cungbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*zungbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::generate vect, + std::int64_t (*zungbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); std::int64_t (*cungqr_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); std::int64_t (*zungqr_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*cungtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*cungtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zungtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*zungtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*cunmrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*cunmrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*zunmrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*zunmrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*cunmqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*cunmqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*zunmqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*zunmqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*cunmtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*cunmtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); - std::int64_t (*zunmtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, + std::int64_t (*zunmtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); std::int64_t (*sgetrf_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, @@ -1636,19 +1636,19 @@ typedef struct { std::int64_t stride_ipiv, std::int64_t batch_size); std::int64_t (*sgetrs_batch_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); std::int64_t (*dgetrs_batch_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); std::int64_t (*cgetrs_batch_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); std::int64_t (*zgetrs_batch_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); std::int64_t (*sgeqrf_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, @@ -1671,38 +1671,38 @@ typedef struct { std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); - std::int64_t (*spotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*spotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); - std::int64_t (*dpotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*dpotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); - std::int64_t (*cpotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*cpotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); - std::int64_t (*zpotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*zpotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); - std::int64_t (*spotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*spotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - std::int64_t (*dpotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*dpotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - std::int64_t (*cpotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*cpotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - std::int64_t (*zpotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo uplo, + std::int64_t (*zpotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, @@ -1756,16 +1756,16 @@ typedef struct { std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); std::int64_t (*sgetrs_group_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); std::int64_t (*dgetrs_group_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); std::int64_t (*cgetrs_group_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); std::int64_t (*zgetrs_group_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); std::int64_t (*sgeqrf_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, @@ -1791,38 +1791,38 @@ typedef struct { std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); - std::int64_t (*spotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + std::int64_t (*spotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); - std::int64_t (*dpotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + std::int64_t (*dpotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); - std::int64_t (*cpotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + std::int64_t (*cpotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); - std::int64_t (*zpotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + std::int64_t (*zpotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); - std::int64_t (*spotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + std::int64_t (*spotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); - std::int64_t (*dpotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + std::int64_t (*dpotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); - std::int64_t (*cpotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + std::int64_t (*cpotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); - std::int64_t (*zpotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::mkl::uplo *uplo, + std::int64_t (*zpotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, diff --git a/src/lapack/lapack_loader.cpp b/src/lapack/lapack_loader.cpp index a27c40626..3283d458c 100644 --- a/src/lapack/lapack_loader.cpp +++ b/src/lapack/lapack_loader.cpp @@ -23,14 +23,14 @@ #include "lapack/function_table.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace lapack { namespace detail { -static oneapi::mkl::detail::table_initializer +static oneapi::math::detail::table_initializer function_tables; -void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, sycl::buffer> &taup, sycl::buffer> &scratchpad, @@ -38,21 +38,21 @@ void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std:: function_tables[{ libkey, queue }].cgebrd_sycl(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgebrd_sycl(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgebrd_sycl(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, sycl::buffer> &taup, sycl::buffer> &scratchpad, @@ -60,109 +60,109 @@ void gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std:: function_tables[{ libkey, queue }].zgebrd_sycl(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgerqf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgerqf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgerqf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgerqf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgeqrf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgeqrf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgeqrf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgeqrf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgetrf_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgetrf_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgetrf_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgetrf_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgetri_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, +void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgetri_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, +void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgetri_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgetri_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +void getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, @@ -170,21 +170,21 @@ void getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpos function_tables[{ libkey, queue }].cgetrs_sycl(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +void getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgetrs_sycl(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +void getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgetrs_sycl(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +void getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, @@ -192,24 +192,24 @@ void getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpos function_tables[{ libkey, queue }].zgetrs_sycl(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, +void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgesvd_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, +void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgesvd_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, @@ -217,8 +217,8 @@ void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd j function_tables[{ libkey, queue }].cgesvd_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, @@ -226,363 +226,363 @@ void gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd j function_tables[{ libkey, queue }].zgesvd_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, +void heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cheevd_sycl(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -void heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer> &a, +void heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zheevd_sycl(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -void hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +void hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].chegvd_sycl(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +void hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zhegvd_sycl(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].chetrd_sycl(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zhetrd_sycl(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].chetrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zhetrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +void orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sorgbr_sycl(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +void orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dorgbr_sycl(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dorgqr_sycl(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sorgqr_sycl(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sorgtr_sycl(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -void orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dorgtr_sycl(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -void ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sormtr_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dormtr_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +void ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sormrq_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +void ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dormrq_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +void ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dormqr_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +void ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sormqr_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].spotrf_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dpotrf_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cpotrf_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zpotrf_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].spotri_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dpotri_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cpotri_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zpotri_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].spotrs_sycl(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dpotrs_sycl(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cpotrs_sycl(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zpotrs_sycl(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, +void syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dsyevd_sycl(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -void syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, +void syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ssyevd_sycl(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -void sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dsygvd_sycl(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ssygvd_sycl(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dsytrd_sycl(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ssytrd_sycl(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ssytrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dsytrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].csytrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zsytrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ctrtrs_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dtrtrs_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].strtrs_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ztrtrs_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +void ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cungbr_sycl(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +void ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zungbr_sycl(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cungqr_sycl(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zungqr_sycl(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cungtr_sycl(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -void ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zungtr_sycl(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -void unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +void unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, @@ -590,8 +590,8 @@ void unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side sid function_tables[{ libkey, queue }].cunmrq_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +void unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, @@ -599,8 +599,8 @@ void unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side sid function_tables[{ libkey, queue }].zunmrq_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +void unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, @@ -608,8 +608,8 @@ void unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side sid function_tables[{ libkey, queue }].cunmqr_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +void unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, @@ -617,8 +617,8 @@ void unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side sid function_tables[{ libkey, queue }].zunmqr_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, @@ -626,8 +626,8 @@ void unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side sid function_tables[{ libkey, queue }].cunmtr_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, @@ -635,7 +635,7 @@ void unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side sid function_tables[{ libkey, queue }].zunmtr_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -643,21 +643,21 @@ sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m return function_tables[{ libkey, queue }].cgebrd_usm_sycl( queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tauq, double *taup, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dgebrd_usm_sycl( queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tauq, float *taup, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].sgebrd_usm_sycl( queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -665,111 +665,111 @@ sycl::event gebrd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m return function_tables[{ libkey, queue }].zgebrd_usm_sycl( queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].sgerqf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dgerqf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].cgerqf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event gerqf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zgerqf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].cgeqrf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dgeqrf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].sgeqrf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zgeqrf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].cgetrf_usm_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dgetrf_usm_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].sgetrf_usm_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zgetrf_usm_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].cgetri_usm_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *a, +sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dgetri_usm_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *a, +sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].sgetri_usm_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getri(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zgetri_usm_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -777,21 +777,21 @@ sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::t return function_tables[{ libkey, queue }].cgetrs_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dgetrs_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].sgetrs_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -799,8 +799,8 @@ sycl::event getrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::t return function_tables[{ libkey, queue }].zgetrs_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, +sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -808,8 +808,8 @@ sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::j u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, +sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -817,8 +817,8 @@ sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::j u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, +sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -826,8 +826,8 @@ sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::j u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -836,104 +836,104 @@ sycl::event gesvd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::j u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, +sycl::event heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].cheevd_usm_sycl( queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event heevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, +sycl::event heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zheevd_usm_sycl( queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].chegvd_usm_sycl( queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event hegvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zhegvd_usm_sycl( queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].chetrd_usm_sycl( queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zhetrd_usm_sycl( queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].chetrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zhetrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +sycl::event orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].sorgbr_usm_sycl( queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +sycl::event orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dorgbr_usm_sycl( queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dorgqr_usm_sycl( queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].sorgqr_usm_sycl( queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].sorgtr_usm_sycl( queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dorgtr_usm_sycl( queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -941,8 +941,8 @@ sycl::event ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -950,8 +950,8 @@ sycl::event ormtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +sycl::event ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -959,8 +959,8 @@ sycl::event ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +sycl::event ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -968,8 +968,8 @@ sycl::event ormrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +sycl::event ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -977,8 +977,8 @@ sycl::event ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +sycl::event ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -986,157 +986,157 @@ sycl::event ormqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].spotrf_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dpotrf_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].cpotrf_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zpotrf_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].spotri_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dpotri_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].cpotri_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zpotri_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].spotrs_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dpotrs_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].cpotrs_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zpotrs_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, +sycl::event syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dsyevd_usm_sycl( queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event syevd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, +sycl::event syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].ssyevd_usm_sycl( queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dsygvd_usm_sycl( queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event sygvd(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].ssygvd_usm_sycl( queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dsytrd_usm_sycl( queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrd(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].ssytrd_usm_sycl( queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].ssytrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dsytrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].csytrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zsytrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1144,8 +1144,8 @@ sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::u lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1153,8 +1153,8 @@ sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::u lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1162,8 +1162,8 @@ sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::u lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, std::int64_t n, +sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1171,50 +1171,50 @@ sycl::event trtrs(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::u lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +sycl::event ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].cungbr_usm_sycl( queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungbr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::generate vec, +sycl::event ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zungbr_usm_sycl( queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].cungqr_usm_sycl( queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +sycl::event ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zungqr_usm_sycl( queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].cungtr_usm_sycl( queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].zungtr_usm_sycl( queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +sycl::event unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1222,8 +1222,8 @@ sycl::event unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +sycl::event unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1231,8 +1231,8 @@ sycl::event unmrq(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +sycl::event unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1240,8 +1240,8 @@ sycl::event unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, +sycl::event unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1249,8 +1249,8 @@ sycl::event unmqr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1259,8 +1259,8 @@ sycl::event unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::int64_t m, +sycl::event unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1269,21 +1269,21 @@ sycl::event unmtr(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::s lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgeqrf_batch_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgeqrf_batch_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, @@ -1291,7 +1291,7 @@ void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, function_tables[{ libkey, queue }].cgeqrf_batch_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, @@ -1299,7 +1299,7 @@ void geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, function_tables[{ libkey, queue }].zgeqrf_batch_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, @@ -1307,7 +1307,7 @@ void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, function_tables[{ libkey, queue }].sgetri_batch_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, @@ -1315,7 +1315,7 @@ void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, function_tables[{ libkey, queue }].dgetri_batch_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, @@ -1323,7 +1323,7 @@ void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, function_tables[{ libkey, queue }].cgetri_batch_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, @@ -1331,7 +1331,7 @@ void getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, function_tables[{ libkey, queue }].zgetri_batch_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -1341,7 +1341,7 @@ void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::tr ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, @@ -1351,7 +1351,7 @@ void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::tr ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, @@ -1361,7 +1361,7 @@ void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::tr ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, +void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, @@ -1371,7 +1371,7 @@ void getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::tr ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, @@ -1379,7 +1379,7 @@ void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, function_tables[{ libkey, queue }].sgetrf_batch_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, @@ -1387,7 +1387,7 @@ void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, function_tables[{ libkey, queue }].dgetrf_batch_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, @@ -1395,7 +1395,7 @@ void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, function_tables[{ libkey, queue }].cgetrf_batch_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, @@ -1403,49 +1403,49 @@ void getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, function_tables[{ libkey, queue }].zgetrf_batch_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sorgqr_batch_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dorgqr_batch_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].spotrf_batch_sycl(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dpotrf_batch_sycl(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cpotrf_batch_sycl(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zpotrf_batch_sycl(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, @@ -1454,7 +1454,7 @@ void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::up ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, @@ -1463,7 +1463,7 @@ void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::up ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -1472,7 +1472,7 @@ void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::up ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -1481,7 +1481,7 @@ void potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::up ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, @@ -1489,7 +1489,7 @@ void ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, function_tables[{ libkey, queue }].cungqr_batch_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, +void ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, @@ -1497,7 +1497,7 @@ void ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, function_tables[{ libkey, queue }].zungqr_batch_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, @@ -1506,7 +1506,7 @@ sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, @@ -1515,7 +1515,7 @@ sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, @@ -1525,7 +1525,7 @@ sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, @@ -1535,7 +1535,7 @@ sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, @@ -1544,7 +1544,7 @@ sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, @@ -1553,7 +1553,7 @@ sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, @@ -1563,7 +1563,7 @@ sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, @@ -1573,7 +1573,7 @@ sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, float *a, +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, @@ -1582,7 +1582,7 @@ sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, double *a, +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, @@ -1591,7 +1591,7 @@ sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1600,7 +1600,7 @@ sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1609,8 +1609,8 @@ sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, @@ -1619,8 +1619,8 @@ sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, @@ -1630,8 +1630,8 @@ sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -1641,8 +1641,8 @@ sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -1652,7 +1652,7 @@ sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, @@ -1661,7 +1661,7 @@ sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, @@ -1670,21 +1670,21 @@ sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].spotrf_batch_usm_sycl( queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { return function_tables[{ libkey, queue }].dpotrf_batch_usm_sycl( queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1692,7 +1692,7 @@ sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: return function_tables[{ libkey, queue }].cpotrf_batch_usm_sycl( queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1700,7 +1700,7 @@ sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: return function_tables[{ libkey, queue }].zpotrf_batch_usm_sycl( queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, @@ -1709,7 +1709,7 @@ sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, @@ -1718,7 +1718,7 @@ sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -1728,7 +1728,7 @@ sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, @@ -1738,7 +1738,7 @@ sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, @@ -1748,7 +1748,7 @@ sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, +sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, @@ -1758,7 +1758,7 @@ sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, @@ -1767,7 +1767,7 @@ sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, @@ -1776,7 +1776,7 @@ sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, @@ -1786,7 +1786,7 @@ sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, @@ -1796,7 +1796,7 @@ sycl::event geqrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, @@ -1805,7 +1805,7 @@ sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, @@ -1814,7 +1814,7 @@ sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1823,7 +1823,7 @@ sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1832,7 +1832,7 @@ sycl::event getrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, float **a, +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1840,7 +1840,7 @@ sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, double **a, +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1848,7 +1848,7 @@ sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1857,7 +1857,7 @@ sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *n, +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1866,8 +1866,8 @@ sycl::event getri_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, @@ -1876,8 +1876,8 @@ sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, @@ -1886,8 +1886,8 @@ sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, @@ -1897,8 +1897,8 @@ sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, @@ -1908,7 +1908,7 @@ sycl::event getrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, @@ -1917,7 +1917,7 @@ sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, @@ -1926,7 +1926,7 @@ sycl::event orgqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1934,7 +1934,7 @@ sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { @@ -1942,7 +1942,7 @@ sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1951,7 +1951,7 @@ sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, @@ -1960,7 +1960,7 @@ sycl::event potrf_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, @@ -1969,7 +1969,7 @@ sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, @@ -1978,7 +1978,7 @@ sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, @@ -1988,7 +1988,7 @@ sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, @@ -1998,7 +1998,7 @@ sycl::event potrs_batch(oneapi::mkl::device libkey, sycl::queue &queue, oneapi:: queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, @@ -2008,7 +2008,7 @@ sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, +sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, @@ -2020,555 +2020,555 @@ sycl::event ungqr_batch(oneapi::mkl::device libkey, sycl::queue &queue, std::int } template <> -std::int64_t gebrd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sgebrd_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gebrd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dgebrd_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gebrd_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t gebrd_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cgebrd_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gebrd_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t gebrd_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zgebrd_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gerqf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sgerqf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gerqf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dgerqf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gerqf_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t gerqf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cgerqf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gerqf_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t gerqf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zgerqf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t geqrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sgeqrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t geqrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dgeqrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t geqrf_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t geqrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cgeqrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t geqrf_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t geqrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zgeqrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gesvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { return function_tables[{ libkey, queue }].sgesvd_scratchpad_size_sycl(queue, jobu, jobvt, m, n, lda, ldu, ldvt); } template <> -std::int64_t gesvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { return function_tables[{ libkey, queue }].dgesvd_scratchpad_size_sycl(queue, jobu, jobvt, m, n, lda, ldu, ldvt); } template <> -std::int64_t gesvd_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t gesvd_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { return function_tables[{ libkey, queue }].cgesvd_scratchpad_size_sycl(queue, jobu, jobvt, m, n, lda, ldu, ldvt); } template <> -std::int64_t gesvd_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t gesvd_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { return function_tables[{ libkey, queue }].zgesvd_scratchpad_size_sycl(queue, jobu, jobvt, m, n, lda, ldu, ldvt); } template <> -std::int64_t getrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sgetrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t getrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dgetrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t getrf_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t getrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cgetrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t getrf_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t getrf_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zgetrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t getri_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sgetri_scratchpad_size_sycl(queue, n, lda); } template <> -std::int64_t getri_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dgetri_scratchpad_size_sycl(queue, n, lda); } template <> -std::int64_t getri_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t getri_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cgetri_scratchpad_size_sycl(queue, n, lda); } template <> -std::int64_t getri_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t getri_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zgetri_scratchpad_size_sycl(queue, n, lda); } template <> -std::int64_t getrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, +std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].sgetrs_scratchpad_size_sycl(queue, trans, n, nrhs, lda, ldb); } template <> -std::int64_t getrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, +std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].dgetrs_scratchpad_size_sycl(queue, trans, n, nrhs, lda, ldb); } template <> -std::int64_t getrs_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t getrs_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].cgetrs_scratchpad_size_sycl(queue, trans, n, nrhs, lda, ldb); } template <> -std::int64_t getrs_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t getrs_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].zgetrs_scratchpad_size_sycl(queue, trans, n, nrhs, lda, ldb); } template <> -std::int64_t heevd_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t heevd_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cheevd_scratchpad_size_sycl(queue, jobz, uplo, n, lda); } template <> -std::int64_t heevd_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t heevd_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zheevd_scratchpad_size_sycl(queue, jobz, uplo, n, lda); } template <> -std::int64_t hegvd_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t hegvd_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].chegvd_scratchpad_size_sycl(queue, itype, jobz, uplo, n, lda, ldb); } template <> -std::int64_t hegvd_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t hegvd_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].zhegvd_scratchpad_size_sycl(queue, itype, jobz, uplo, n, lda, ldb); } template <> -std::int64_t hetrd_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrd_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].chetrd_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t hetrd_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrd_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zhetrd_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t hetrf_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].chetrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t hetrf_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zhetrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t orgbr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, +std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].sorgbr_scratchpad_size_sycl(queue, vect, m, n, k, lda); } template <> -std::int64_t orgbr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, +std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].dorgbr_scratchpad_size_sycl(queue, vect, m, n, k, lda); } template <> -std::int64_t orgtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sorgtr_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t orgtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dorgtr_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t orgqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].sorgqr_scratchpad_size_sycl(queue, m, n, k, lda); } template <> -std::int64_t orgqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].dorgqr_scratchpad_size_sycl(queue, m, n, k, lda); } template <> -std::int64_t ormrq_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return function_tables[{ libkey, queue }].sormrq_scratchpad_size_sycl(queue, side, trans, m, n, k, lda, ldc); } template <> -std::int64_t ormrq_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return function_tables[{ libkey, queue }].dormrq_scratchpad_size_sycl(queue, side, trans, m, n, k, lda, ldc); } template <> -std::int64_t ormqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return function_tables[{ libkey, queue }].sormqr_scratchpad_size_sycl(queue, side, trans, m, n, k, lda, ldc); } template <> -std::int64_t ormqr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::transpose trans, +std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return function_tables[{ libkey, queue }].dormqr_scratchpad_size_sycl(queue, side, trans, m, n, k, lda, ldc); } template <> -std::int64_t ormtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, +std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { return function_tables[{ libkey, queue }].sormtr_scratchpad_size_sycl(queue, side, uplo, trans, m, n, lda, ldc); } template <> -std::int64_t ormtr_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, +std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { return function_tables[{ libkey, queue }].dormtr_scratchpad_size_sycl(queue, side, uplo, trans, m, n, lda, ldc); } template <> -std::int64_t potrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].spotrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t potrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dpotrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t potrf_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cpotrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t potrf_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zpotrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t potrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].spotrs_scratchpad_size_sycl(queue, uplo, n, nrhs, lda, ldb); } template <> -std::int64_t potrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].dpotrs_scratchpad_size_sycl(queue, uplo, n, nrhs, lda, ldb); } template <> -std::int64_t potrs_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrs_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].cpotrs_scratchpad_size_sycl(queue, uplo, n, nrhs, lda, ldb); } template <> -std::int64_t potrs_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrs_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].zpotrs_scratchpad_size_sycl(queue, uplo, n, nrhs, lda, ldb); } template <> -std::int64_t potri_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].spotri_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t potri_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dpotri_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t potri_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potri_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cpotri_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t potri_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potri_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zpotri_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t sytrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].ssytrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t sytrf_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dsytrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t sytrf_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t sytrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].csytrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t sytrf_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t sytrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zsytrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t syevd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].ssyevd_scratchpad_size_sycl(queue, jobz, uplo, n, lda); } template <> -std::int64_t syevd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dsyevd_scratchpad_size_sycl(queue, jobz, uplo, n, lda); } template <> -std::int64_t sygvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, +std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].ssygvd_scratchpad_size_sycl(queue, itype, jobz, uplo, n, lda, ldb); } template <> -std::int64_t sygvd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, +std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].dsygvd_scratchpad_size_sycl(queue, itype, jobz, uplo, n, lda, ldb); } template <> -std::int64_t sytrd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].ssytrd_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t sytrd_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dsytrd_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t trtrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].strtrs_scratchpad_size_sycl(queue, uplo, trans, diag, n, nrhs, lda, ldb); } template <> -std::int64_t trtrs_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].dtrtrs_scratchpad_size_sycl(queue, uplo, trans, diag, n, nrhs, lda, ldb); } template <> -std::int64_t trtrs_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +std::int64_t trtrs_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].ctrtrs_scratchpad_size_sycl(queue, uplo, trans, diag, n, nrhs, lda, ldb); } template <> -std::int64_t trtrs_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +std::int64_t trtrs_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].ztrtrs_scratchpad_size_sycl(queue, uplo, trans, diag, n, nrhs, lda, ldb); } template <> -std::int64_t ungbr_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t ungbr_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].cungbr_scratchpad_size_sycl(queue, vect, m, n, k, lda); } template <> -std::int64_t ungbr_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t ungbr_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].zungbr_scratchpad_size_sycl(queue, vect, m, n, k, lda); } template <> -std::int64_t ungqr_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t ungqr_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].cungqr_scratchpad_size_sycl(queue, m, n, k, lda); } template <> -std::int64_t ungqr_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t ungqr_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].zungqr_scratchpad_size_sycl(queue, m, n, k, lda); } template <> -std::int64_t ungtr_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t ungtr_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cungtr_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t ungtr_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t ungtr_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zungtr_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t unmrq_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmrq_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { @@ -2576,9 +2576,9 @@ std::int64_t unmrq_scratchpad_size>(oneapi::mkl::device libk k, lda, ldc); } template <> -std::int64_t unmrq_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmrq_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { @@ -2586,9 +2586,9 @@ std::int64_t unmrq_scratchpad_size>(oneapi::mkl::device lib k, lda, ldc); } template <> -std::int64_t unmqr_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmqr_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { @@ -2596,9 +2596,9 @@ std::int64_t unmqr_scratchpad_size>(oneapi::mkl::device libk k, lda, ldc); } template <> -std::int64_t unmqr_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmqr_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { @@ -2606,27 +2606,27 @@ std::int64_t unmqr_scratchpad_size>(oneapi::mkl::device lib k, lda, ldc); } template <> -std::int64_t unmtr_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, +std::int64_t unmtr_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { return function_tables[{ libkey, queue }].cunmtr_scratchpad_size_sycl(queue, side, uplo, trans, m, n, lda, ldc); } template <> -std::int64_t unmtr_scratchpad_size>(oneapi::mkl::device libkey, - sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, +std::int64_t unmtr_scratchpad_size>(oneapi::math::device libkey, + sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { return function_tables[{ libkey, queue }].zunmtr_scratchpad_size_sycl(queue, side, uplo, trans, m, n, lda, ldc); } template <> -std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { @@ -2634,7 +2634,7 @@ std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, m, n, lda, stride_a, stride_ipiv, batch_size); } template <> -std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { @@ -2643,20 +2643,20 @@ std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, syc } template <> std::int64_t getrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cgetrf_batch_scratchpad_size_sycl( queue, m, n, lda, stride_a, stride_ipiv, batch_size); } template <> std::int64_t getrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zgetrf_batch_scratchpad_size_sycl( queue, m, n, lda, stride_a, stride_ipiv, batch_size); } template <> -std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { @@ -2664,7 +2664,7 @@ std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, n, lda, stride_a, stride_ipiv, batch_size); } template <> -std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { @@ -2673,21 +2673,21 @@ std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, syc } template <> std::int64_t getri_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cgetri_batch_scratchpad_size_sycl( queue, n, lda, stride_a, stride_ipiv, batch_size); } template <> std::int64_t getri_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zgetri_batch_scratchpad_size_sycl( queue, n, lda, stride_a, stride_ipiv, batch_size); } template <> -std::int64_t getrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, @@ -2696,8 +2696,8 @@ std::int64_t getrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } template <> -std::int64_t getrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose trans, std::int64_t n, +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, @@ -2707,7 +2707,7 @@ std::int64_t getrs_batch_scratchpad_size(oneapi::mkl::device libkey, syc } template <> std::int64_t getrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cgetrs_batch_scratchpad_size_sycl( @@ -2715,14 +2715,14 @@ std::int64_t getrs_batch_scratchpad_size>( } template <> std::int64_t getrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zgetrs_batch_scratchpad_size_sycl( queue, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } template <> -std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { @@ -2730,7 +2730,7 @@ std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, m, n, lda, stride_a, stride_tau, batch_size); } template <> -std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { @@ -2739,29 +2739,29 @@ std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, syc } template <> std::int64_t geqrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cgeqrf_batch_scratchpad_size_sycl( queue, m, n, lda, stride_a, stride_tau, batch_size); } template <> std::int64_t geqrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zgeqrf_batch_scratchpad_size_sycl( queue, m, n, lda, stride_a, stride_tau, batch_size); } template <> -std::int64_t potrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { return function_tables[{ libkey, queue }].spotrf_batch_scratchpad_size_sycl( queue, uplo, n, lda, stride_a, batch_size); } template <> -std::int64_t potrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { return function_tables[{ libkey, queue }].dpotrf_batch_scratchpad_size_sycl( @@ -2769,21 +2769,21 @@ std::int64_t potrf_batch_scratchpad_size(oneapi::mkl::device libkey, syc } template <> std::int64_t potrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cpotrf_batch_scratchpad_size_sycl( queue, uplo, n, lda, stride_a, batch_size); } template <> std::int64_t potrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zpotrf_batch_scratchpad_size_sycl( queue, uplo, n, lda, stride_a, batch_size); } template <> -std::int64_t potrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { @@ -2791,8 +2791,8 @@ std::int64_t potrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template <> -std::int64_t potrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { @@ -2801,7 +2801,7 @@ std::int64_t potrs_batch_scratchpad_size(oneapi::mkl::device libkey, syc } template <> std::int64_t potrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cpotrs_batch_scratchpad_size_sycl( @@ -2809,14 +2809,14 @@ std::int64_t potrs_batch_scratchpad_size>( } template <> std::int64_t potrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zpotrs_batch_scratchpad_size_sycl( queue, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template <> -std::int64_t orgqr_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { @@ -2824,7 +2824,7 @@ std::int64_t orgqr_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, m, n, k, lda, stride_a, stride_tau, batch_size); } template <> -std::int64_t orgqr_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { @@ -2833,20 +2833,20 @@ std::int64_t orgqr_batch_scratchpad_size(oneapi::mkl::device libkey, syc } template <> std::int64_t ungqr_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cungqr_batch_scratchpad_size_sycl( queue, m, n, k, lda, stride_a, stride_tau, batch_size); } template <> std::int64_t ungqr_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zungqr_batch_scratchpad_size_sycl( queue, m, n, k, lda, stride_a, stride_tau, batch_size); } template <> -std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2854,7 +2854,7 @@ std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2862,7 +2862,7 @@ std::int64_t getrf_batch_scratchpad_size(oneapi::mkl::device libkey, syc queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t getrf_batch_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t getrf_batch_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, @@ -2871,7 +2871,7 @@ std::int64_t getrf_batch_scratchpad_size>(oneapi::mkl::devic queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t getrf_batch_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t getrf_batch_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, @@ -2880,7 +2880,7 @@ std::int64_t getrf_batch_scratchpad_size>(oneapi::mkl::devi queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2888,7 +2888,7 @@ std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, n, lda, group_count, group_sizes); } template <> -std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2896,7 +2896,7 @@ std::int64_t getri_batch_scratchpad_size(oneapi::mkl::device libkey, syc queue, n, lda, group_count, group_sizes); } template <> -std::int64_t getri_batch_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t getri_batch_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, @@ -2905,7 +2905,7 @@ std::int64_t getri_batch_scratchpad_size>(oneapi::mkl::devic queue, n, lda, group_count, group_sizes); } template <> -std::int64_t getri_batch_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t getri_batch_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, @@ -2914,8 +2914,8 @@ std::int64_t getri_batch_scratchpad_size>(oneapi::mkl::devi queue, n, lda, group_count, group_sizes); } template <> -std::int64_t getrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose *trans, std::int64_t *n, +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2923,8 +2923,8 @@ std::int64_t getrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, trans, n, nrhs, lda, ldb, group_count, group_sizes); } template <> -std::int64_t getrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::transpose *trans, std::int64_t *n, +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2933,7 +2933,7 @@ std::int64_t getrs_batch_scratchpad_size(oneapi::mkl::device libkey, syc } template <> std::int64_t getrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { return function_tables[{ libkey, queue }].cgetrs_group_scratchpad_size_sycl( @@ -2941,14 +2941,14 @@ std::int64_t getrs_batch_scratchpad_size>( } template <> std::int64_t getrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { return function_tables[{ libkey, queue }].zgetrs_group_scratchpad_size_sycl( queue, trans, n, nrhs, lda, ldb, group_count, group_sizes); } template <> -std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2956,7 +2956,7 @@ std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2964,7 +2964,7 @@ std::int64_t geqrf_batch_scratchpad_size(oneapi::mkl::device libkey, syc queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t geqrf_batch_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t geqrf_batch_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, @@ -2973,7 +2973,7 @@ std::int64_t geqrf_batch_scratchpad_size>(oneapi::mkl::devic queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t geqrf_batch_scratchpad_size>(oneapi::mkl::device libkey, +std::int64_t geqrf_batch_scratchpad_size>(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, @@ -2982,7 +2982,7 @@ std::int64_t geqrf_batch_scratchpad_size>(oneapi::mkl::devi queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t orgqr_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2990,7 +2990,7 @@ std::int64_t orgqr_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, m, n, k, lda, group_count, group_sizes); } template <> -std::int64_t orgqr_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { @@ -2998,16 +2998,16 @@ std::int64_t orgqr_batch_scratchpad_size(oneapi::mkl::device libkey, syc queue, m, n, k, lda, group_count, group_sizes); } template <> -std::int64_t potrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return function_tables[{ libkey, queue }].spotrf_group_scratchpad_size_sycl( queue, uplo, n, lda, group_count, group_sizes); } template <> -std::int64_t potrf_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return function_tables[{ libkey, queue }].dpotrf_group_scratchpad_size_sycl( @@ -3015,21 +3015,21 @@ std::int64_t potrf_batch_scratchpad_size(oneapi::mkl::device libkey, syc } template <> std::int64_t potrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return function_tables[{ libkey, queue }].cpotrf_group_scratchpad_size_sycl( queue, uplo, n, lda, group_count, group_sizes); } template <> std::int64_t potrf_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return function_tables[{ libkey, queue }].zpotrf_group_scratchpad_size_sycl( queue, uplo, n, lda, group_count, group_sizes); } template <> -std::int64_t potrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { @@ -3037,8 +3037,8 @@ std::int64_t potrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl queue, uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template <> -std::int64_t potrs_batch_scratchpad_size(oneapi::mkl::device libkey, sycl::queue &queue, - oneapi::mkl::uplo *uplo, std::int64_t *n, +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { @@ -3047,7 +3047,7 @@ std::int64_t potrs_batch_scratchpad_size(oneapi::mkl::device libkey, syc } template <> std::int64_t potrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { return function_tables[{ libkey, queue }].cpotrs_group_scratchpad_size_sycl( @@ -3055,7 +3055,7 @@ std::int64_t potrs_batch_scratchpad_size>( } template <> std::int64_t potrs_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { return function_tables[{ libkey, queue }].zpotrs_group_scratchpad_size_sycl( @@ -3063,14 +3063,14 @@ std::int64_t potrs_batch_scratchpad_size>( } template <> std::int64_t ungqr_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return function_tables[{ libkey, queue }].cungqr_group_scratchpad_size_sycl( queue, m, n, k, lda, group_count, group_sizes); } template <> std::int64_t ungqr_batch_scratchpad_size>( - oneapi::mkl::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, + oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return function_tables[{ libkey, queue }].zungqr_group_scratchpad_size_sycl( queue, m, n, k, lda, group_count, group_sizes); @@ -3078,5 +3078,5 @@ std::int64_t ungqr_batch_scratchpad_size>( } //namespace detail } //namespace lapack -} //namespace mkl +} //namespace math } //namespace oneapi diff --git a/src/rng/backends/curand/curand_helper.hpp b/src/rng/backends/curand/curand_helper.hpp index b38946260..203f527c7 100644 --- a/src/rng/backends/curand/curand_helper.hpp +++ b/src/rng/backends/curand/curand_helper.hpp @@ -70,7 +70,7 @@ #include "oneapi/math/types.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace curand { @@ -190,7 +190,7 @@ class cuda_error : virtual public std::runtime_error { throw curand_error(std::string(#func) + std::string(" : "), status); \ } -// Static template functions oneapi::mkl::rng::curand::range_transform_fp for +// Static template functions oneapi::math::rng::curand::range_transform_fp for // Buffer and USM APIs // // cuRAND has no built-in functionality to specify a custom range for sampling @@ -252,7 +252,7 @@ static inline sycl::event range_transform_fp_accurate(sycl::queue& queue, T a, T }); } -// Static template functions oneapi::mkl::rng::curand::range_transform_int for +// Static template functions oneapi::math::rng::curand::range_transform_int for // Buffer and USM APIs // // cuRAND has no built-in functionality to specify a custom range for sampling @@ -285,7 +285,7 @@ inline sycl::event range_transform_int(sycl::queue& queue, T a, T b, std::int64_ [=](sycl::id<1> id) { out[id] = a + in[id] % (b - a); }); } -// Static template functions oneapi::mkl::rng::curand::sample_bernoulli for +// Static template functions oneapi::math::rng::curand::sample_bernoulli for // Buffer and USM APIs // // cuRAND has no built-in functionality to sample from a Bernoulli distribution. @@ -320,7 +320,7 @@ static inline sycl::event sample_bernoulli_from_uniform(sycl::queue& queue, floa } // namespace curand } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _MKL_RNG_CURAND_HELPER_HPP_ diff --git a/src/rng/backends/curand/curand_task.hpp b/src/rng/backends/curand/curand_task.hpp index 49455cfbc..58ce982de 100644 --- a/src/rng/backends/curand/curand_task.hpp +++ b/src/rng/backends/curand/curand_task.hpp @@ -10,7 +10,7 @@ #include "curand_helper.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace curand { #ifdef __HIPSYCL__ @@ -84,7 +84,7 @@ static inline void onemath_curand_host_task(H &cgh, Engine e, F f) { } // namespace curand } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif diff --git a/src/rng/backends/curand/curand_wrappers.cpp b/src/rng/backends/curand/curand_wrappers.cpp index 4b30b9a76..c3e25a63c 100644 --- a/src/rng/backends/curand/curand_wrappers.cpp +++ b/src/rng/backends/curand/curand_wrappers.cpp @@ -62,7 +62,7 @@ #define WRAPPER_VERSION 1 extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { - WRAPPER_VERSION, oneapi::mkl::rng::curand::create_philox4x32x10, - oneapi::mkl::rng::curand::create_philox4x32x10, oneapi::mkl::rng::curand::create_mrg32k3a, - oneapi::mkl::rng::curand::create_mrg32k3a + WRAPPER_VERSION, oneapi::math::rng::curand::create_philox4x32x10, + oneapi::math::rng::curand::create_philox4x32x10, oneapi::math::rng::curand::create_mrg32k3a, + oneapi::math::rng::curand::create_mrg32k3a }; diff --git a/src/rng/backends/curand/mrg32k3a.cpp b/src/rng/backends/curand/mrg32k3a.cpp index d3e77a810..4e2b6d263 100644 --- a/src/rng/backends/curand/mrg32k3a.cpp +++ b/src/rng/backends/curand/mrg32k3a.cpp @@ -80,35 +80,35 @@ #include "oneapi/math/rng/engines.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace curand { #if !defined(_WIN64) -class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { +class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { public: mrg32k3a_impl(sycl::queue queue, std::uint32_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { + : oneapi::math::rng::detail::engine_impl(queue) { curandStatus_t status; CURAND_CALL(curandCreateGenerator, status, &engine_, CURAND_RNG_PSEUDO_MRG32K3A); CURAND_CALL(curandSetPseudoRandomGeneratorSeed, status, engine_, (unsigned long long)seed); } mrg32k3a_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine", + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "mrg32ka engine", "multi-seed unsupported by cuRAND backend"); } - mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::mkl::rng::detail::engine_impl(*other) { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine", + mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::math::rng::detail::engine_impl(*other) { + throw oneapi::math::unimplemented("rng", "mrg32ka engine", "copy construction unsupported by cuRAND backend"); } // Buffers API virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -121,7 +121,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -133,8 +133,8 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { range_transform_fp(queue_, distr.a(), distr.b(), n, r); } - virtual void generate(const oneapi::mkl::rng::uniform< - std::int32_t, oneapi::mkl::rng::uniform_method::standard>& distr, + virtual void generate(const oneapi::math::rng::uniform< + std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { sycl::buffer ib(n); queue_.submit([&](sycl::handler& cgh) { @@ -148,7 +148,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -161,7 +161,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -173,8 +173,8 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { range_transform_fp_accurate(queue_, distr.a(), distr.b(), n, r); } - virtual void generate(const oneapi::mkl::rng::gaussian< - float, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -186,8 +186,8 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { }); } - virtual void generate(const oneapi::mkl::rng::gaussian< - double, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -200,23 +200,23 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } - virtual void generate(const oneapi::mkl::rng::lognormal< - float, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -228,8 +228,8 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { }); } - virtual void generate(const oneapi::mkl::rng::lognormal< - double, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -242,45 +242,45 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } @@ -299,7 +299,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { // USM APIs virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -312,7 +312,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -325,7 +325,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& + const oneapi::math::rng::uniform& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { auto usm_deleter = [this](std::uint32_t* ptr) { @@ -348,7 +348,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -362,7 +362,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -375,7 +375,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -389,7 +389,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -403,25 +403,25 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -434,7 +434,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -448,18 +448,18 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; @@ -468,7 +468,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; @@ -477,7 +477,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; @@ -486,7 +486,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; @@ -495,7 +495,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; @@ -512,7 +512,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { }); } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { return new mrg32k3a_impl(this); } @@ -522,12 +522,12 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void skip_ahead(std::initializer_list num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "skip_ahead", + throw oneapi::math::unimplemented("rng", "skip_ahead", "initializer list unsupported by cuRAND backend"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "leapfrog", "unsupported by cuRAND backend"); + throw oneapi::math::unimplemented("rng", "leapfrog", "unsupported by cuRAND backend"); } virtual ~mrg32k3a_impl() override { @@ -539,290 +539,290 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::uint32_t seed_; }; #else // cuRAND backend is currently not supported on Windows -class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { +class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { public: mrg32k3a_impl(sycl::queue queue, std::uint32_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } mrg32k3a_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::mkl::rng::detail::engine_impl(*other) { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::math::rng::detail::engine_impl(*other) { + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } // Buffers API virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate(const oneapi::mkl::rng::uniform< - std::int32_t, oneapi::mkl::rng::uniform_method::standard>& distr, + virtual void generate(const oneapi::math::rng::uniform< + std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate(const oneapi::mkl::rng::gaussian< - float, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate(const oneapi::mkl::rng::gaussian< - double, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate(const oneapi::mkl::rng::lognormal< - float, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate(const oneapi::mkl::rng::lognormal< - double, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate(const bits& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } // USM APIs virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& + const oneapi::math::rng::uniform& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate(const bits& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return nullptr; } virtual void skip_ahead(std::uint64_t num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void skip_ahead(std::initializer_list num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual ~mrg32k3a_impl() override {} }; #endif -oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed) { +oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed) { return new mrg32k3a_impl(queue, seed); } -oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::initializer_list seed) { return new mrg32k3a_impl(queue, seed); } } // namespace curand } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/rng/backends/curand/philox4x32x10.cpp b/src/rng/backends/curand/philox4x32x10.cpp index 89ca45853..375be1d6d 100644 --- a/src/rng/backends/curand/philox4x32x10.cpp +++ b/src/rng/backends/curand/philox4x32x10.cpp @@ -80,7 +80,7 @@ #include "oneapi/math/rng/detail/curand/onemath_rng_curand.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace curand { @@ -107,31 +107,31 @@ namespace curand { * generates random numbers. * */ -class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { +class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { public: philox4x32x10_impl(sycl::queue queue, std::uint64_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { + : oneapi::math::rng::detail::engine_impl(queue) { curandStatus_t status; CURAND_CALL(curandCreateGenerator, status, &engine_, CURAND_RNG_PSEUDO_PHILOX4_32_10); CURAND_CALL(curandSetPseudoRandomGeneratorSeed, status, engine_, (unsigned long long)seed); } philox4x32x10_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine", + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine", "multi-seed unsupported by cuRAND backend"); } philox4x32x10_impl(const philox4x32x10_impl* other) - : oneapi::mkl::rng::detail::engine_impl(*other) { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine", + : oneapi::math::rng::detail::engine_impl(*other) { + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine", "copy construction unsupported by cuRAND backend"); } // Buffers API virtual inline void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -144,7 +144,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -156,8 +156,8 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { range_transform_fp(queue_, distr.a(), distr.b(), n, r); } - virtual void generate(const oneapi::mkl::rng::uniform< - std::int32_t, oneapi::mkl::rng::uniform_method::standard>& distr, + virtual void generate(const oneapi::math::rng::uniform< + std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { sycl::buffer ib(n); queue_.submit([&](sycl::handler& cgh) { @@ -171,7 +171,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -184,7 +184,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -196,8 +196,8 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { range_transform_fp_accurate(queue_, distr.a(), distr.b(), n, r); } - virtual void generate(const oneapi::mkl::rng::gaussian< - float, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -209,8 +209,8 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { }); } - virtual void generate(const oneapi::mkl::rng::gaussian< - double, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -223,23 +223,23 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } - virtual void generate(const oneapi::mkl::rng::lognormal< - float, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -251,8 +251,8 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { }); } - virtual void generate(const oneapi::mkl::rng::lognormal< - double, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -265,45 +265,45 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); } @@ -322,7 +322,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { // USM APIs virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -335,7 +335,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -348,7 +348,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& + const oneapi::math::rng::uniform& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { auto usm_deleter = [this](std::uint32_t* ptr) { @@ -371,7 +371,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -384,7 +384,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -397,7 +397,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -411,7 +411,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -425,25 +425,25 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -456,7 +456,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -470,18 +470,18 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; @@ -490,7 +490,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; @@ -499,7 +499,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; @@ -508,7 +508,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; @@ -517,7 +517,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); return sycl::event{}; @@ -534,7 +534,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { }); } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { return new philox4x32x10_impl(this); } @@ -544,12 +544,12 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void skip_ahead(std::initializer_list num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "skip_ahead", + throw oneapi::math::unimplemented("rng", "skip_ahead", "initializer list unsupported by cuRAND backend"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "leapfrog", "unsupported by cuRAND backend"); + throw oneapi::math::unimplemented("rng", "leapfrog", "unsupported by cuRAND backend"); } virtual ~philox4x32x10_impl() override { @@ -560,291 +560,291 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { curandGenerator_t engine_; }; #else // cuRAND backend is currently not supported on Windows -class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { +class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { public: philox4x32x10_impl(sycl::queue queue, std::uint64_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } philox4x32x10_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } philox4x32x10_impl(const philox4x32x10_impl* other) - : oneapi::mkl::rng::detail::engine_impl(*other) { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + : oneapi::math::rng::detail::engine_impl(*other) { + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } // Buffers API virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate(const oneapi::mkl::rng::uniform< - std::int32_t, oneapi::mkl::rng::uniform_method::standard>& distr, + virtual void generate(const oneapi::math::rng::uniform< + std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate(const oneapi::mkl::rng::gaussian< - float, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate(const oneapi::mkl::rng::gaussian< - double, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate(const oneapi::mkl::rng::lognormal< - float, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate(const oneapi::mkl::rng::lognormal< - double, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate(const bits& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } // USM APIs virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& + const oneapi::math::rng::uniform& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate(const bits& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return nullptr; } virtual void skip_ahead(std::uint64_t num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void skip_ahead(std::initializer_list num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual ~philox4x32x10_impl() override {} }; #endif -oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { +oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { return new philox4x32x10_impl(queue, seed); } -oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed) { return new philox4x32x10_impl(queue, seed); } } // namespace curand } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/rng/backends/mklcpu/cpu_common.hpp b/src/rng/backends/mklcpu/cpu_common.hpp index cbd6cae59..b83dfef6b 100644 --- a/src/rng/backends/mklcpu/cpu_common.hpp +++ b/src/rng/backends/mklcpu/cpu_common.hpp @@ -27,7 +27,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace mklcpu { @@ -68,7 +68,7 @@ typename Acc::value_type *get_raw_ptr(Acc acc) { } // namespace mklcpu } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_RNG_CPU_COMMON_HPP_ diff --git a/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp b/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp index 04d4a7c7b..36b8d8169 100644 --- a/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp +++ b/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp @@ -23,7 +23,7 @@ #define WRAPPER_VERSION 1 extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { - WRAPPER_VERSION, oneapi::mkl::rng::mklcpu::create_philox4x32x10, - oneapi::mkl::rng::mklcpu::create_philox4x32x10, oneapi::mkl::rng::mklcpu::create_mrg32k3a, - oneapi::mkl::rng::mklcpu::create_mrg32k3a + WRAPPER_VERSION, oneapi::math::rng::mklcpu::create_philox4x32x10, + oneapi::math::rng::mklcpu::create_philox4x32x10, oneapi::math::rng::mklcpu::create_mrg32k3a, + oneapi::math::rng::mklcpu::create_mrg32k3a }; diff --git a/src/rng/backends/mklcpu/mrg32k3a.cpp b/src/rng/backends/mklcpu/mrg32k3a.cpp index 27bb86203..4e9958211 100644 --- a/src/rng/backends/mklcpu/mrg32k3a.cpp +++ b/src/rng/backends/mklcpu/mrg32k3a.cpp @@ -33,26 +33,26 @@ #include "cpu_common.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace mklcpu { -class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { +class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { public: mrg32k3a_impl(sycl::queue queue, std::uint32_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { + : oneapi::math::rng::detail::engine_impl(queue) { vslNewStream(&stream_, VSL_BRNG_MRG32K3A, seed); state_size_ = vslGetStreamSize(stream_); } mrg32k3a_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { + : oneapi::math::rng::detail::engine_impl(queue) { vslNewStreamEx(&stream_, VSL_BRNG_MRG32K3A, 2 * seed.size(), reinterpret_cast(seed.begin())); state_size_ = vslGetStreamSize(stream_); } - mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::mkl::rng::detail::engine_impl(*other) { + mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::math::rng::detail::engine_impl(*other) { vslCopyStream(&stream_, other->stream_); state_size_ = vslGetStreamSize(stream_); } @@ -545,7 +545,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { }); } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { return new mrg32k3a_impl(this); } @@ -558,7 +558,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "leapfrog"); + throw oneapi::math::unimplemented("rng", "leapfrog"); } virtual ~mrg32k3a_impl() override { @@ -570,16 +570,16 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { std::int32_t state_size_; }; -oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed) { +oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed) { return new mrg32k3a_impl(queue, seed); } -oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::initializer_list seed) { return new mrg32k3a_impl(queue, seed); } } // namespace mklcpu } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/rng/backends/mklcpu/philox4x32x10.cpp b/src/rng/backends/mklcpu/philox4x32x10.cpp index 2d3e660fd..ecea512ce 100644 --- a/src/rng/backends/mklcpu/philox4x32x10.cpp +++ b/src/rng/backends/mklcpu/philox4x32x10.cpp @@ -33,28 +33,28 @@ #include "cpu_common.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace mklcpu { -class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { +class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { public: philox4x32x10_impl(sycl::queue queue, std::uint64_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { + : oneapi::math::rng::detail::engine_impl(queue) { vslNewStreamEx(&stream_, VSL_BRNG_PHILOX4X32X10, 2, reinterpret_cast(&seed)); state_size_ = vslGetStreamSize(stream_); } philox4x32x10_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { + : oneapi::math::rng::detail::engine_impl(queue) { vslNewStreamEx(&stream_, VSL_BRNG_PHILOX4X32X10, 2 * seed.size(), reinterpret_cast(seed.begin())); state_size_ = vslGetStreamSize(stream_); } philox4x32x10_impl(const philox4x32x10_impl* other) - : oneapi::mkl::rng::detail::engine_impl(*other) { + : oneapi::math::rng::detail::engine_impl(*other) { vslCopyStream(&stream_, other->stream_); state_size_ = vslGetStreamSize(stream_); } @@ -547,7 +547,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { }); } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { return new philox4x32x10_impl(this); } @@ -560,7 +560,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "leapfrog"); + throw oneapi::math::unimplemented("rng", "leapfrog"); } virtual ~philox4x32x10_impl() override { @@ -572,16 +572,16 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { std::int32_t state_size_; }; -oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { +oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { return new philox4x32x10_impl(queue, seed); } -oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed) { return new philox4x32x10_impl(queue, seed); } } // namespace mklcpu } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp b/src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp index 5ca480ba5..d18b275fa 100755 --- a/src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp +++ b/src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp @@ -27,7 +27,7 @@ #endif namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace detail { @@ -75,7 +75,7 @@ sycl::event generate(sycl::queue& queue, const DistrType& distr, } // namespace gpu } // namespace detail } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif //_MKL_INTERNAL_RNG_GPU_HPP_ diff --git a/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp b/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp index 216aab951..47b36245e 100644 --- a/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp +++ b/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp @@ -23,7 +23,7 @@ #define WRAPPER_VERSION 1 extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { - WRAPPER_VERSION, oneapi::mkl::rng::mklgpu::create_philox4x32x10, - oneapi::mkl::rng::mklgpu::create_philox4x32x10, oneapi::mkl::rng::mklgpu::create_mrg32k3a, - oneapi::mkl::rng::mklgpu::create_mrg32k3a + WRAPPER_VERSION, oneapi::math::rng::mklgpu::create_philox4x32x10, + oneapi::math::rng::mklgpu::create_philox4x32x10, oneapi::math::rng::mklgpu::create_mrg32k3a, + oneapi::math::rng::mklgpu::create_mrg32k3a }; diff --git a/src/rng/backends/mklgpu/mrg32k3a.cpp b/src/rng/backends/mklgpu/mrg32k3a.cpp index 1dd44775d..886598ca3 100644 --- a/src/rng/backends/mklgpu/mrg32k3a.cpp +++ b/src/rng/backends/mklgpu/mrg32k3a.cpp @@ -34,284 +34,284 @@ #include "mkl_internal_rng_gpu.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace mklgpu { -class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { +class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { public: mrg32k3a_impl(sycl::queue queue, std::uint32_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { + : oneapi::math::rng::detail::engine_impl(queue) { engine_ = - oneapi::mkl::rng::detail::gpu::create_engine(queue, seed); + oneapi::math::rng::detail::gpu::create_engine(queue, seed); } mrg32k3a_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - engine_ = oneapi::mkl::rng::detail::gpu::create_engine( + : oneapi::math::rng::detail::engine_impl(queue) { + engine_ = oneapi::math::rng::detail::gpu::create_engine( queue, (std::int64_t)(seed.size()), (const unsigned int*)seed.begin()); } - mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::mkl::rng::detail::engine_impl(*other) { + mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::math::rng::detail::engine_impl(*other) { sycl::queue queue(other->queue_); - engine_ = oneapi::mkl::rng::detail::gpu::create_engine( + engine_ = oneapi::math::rng::detail::gpu::create_engine( queue, other->engine_); } // Buffers API virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } - virtual void generate(const oneapi::mkl::rng::uniform< - std::int32_t, oneapi::mkl::rng::uniform_method::standard>& distr, + virtual void generate(const oneapi::math::rng::uniform< + std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } - virtual void generate(const oneapi::mkl::rng::gaussian< - float, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } - virtual void generate(const oneapi::mkl::rng::gaussian< - double, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } - virtual void generate(const oneapi::mkl::rng::lognormal< - float, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } - virtual void generate(const oneapi::mkl::rng::lognormal< - double, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate(const bits& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } // USM APIs virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& + const oneapi::math::rng::uniform& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { ; - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate(const bits& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { return new mrg32k3a_impl(this); } virtual void skip_ahead(std::uint64_t num_to_skip) override { - oneapi::mkl::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); + oneapi::math::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); } virtual void skip_ahead(std::initializer_list num_to_skip) override { - oneapi::mkl::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); + oneapi::math::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "leapfrog"); + throw oneapi::math::unimplemented("rng", "leapfrog"); } virtual ~mrg32k3a_impl() override { - oneapi::mkl::rng::detail::gpu::delete_engine(queue_, engine_); + oneapi::math::rng::detail::gpu::delete_engine(queue_, engine_); } private: - oneapi::mkl::rng::detail::engine_base_impl* engine_; + oneapi::math::rng::detail::engine_base_impl* engine_; }; -oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed) { +oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed) { return new mrg32k3a_impl(queue, seed); } -oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::initializer_list seed) { return new mrg32k3a_impl(queue, seed); } } // namespace mklgpu } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/rng/backends/mklgpu/philox4x32x10.cpp b/src/rng/backends/mklgpu/philox4x32x10.cpp index 6410876fb..2bf91145a 100644 --- a/src/rng/backends/mklgpu/philox4x32x10.cpp +++ b/src/rng/backends/mklgpu/philox4x32x10.cpp @@ -34,285 +34,285 @@ #include "mkl_internal_rng_gpu.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace mklgpu { -class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { +class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { public: philox4x32x10_impl(sycl::queue queue, std::uint64_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - engine_ = oneapi::mkl::rng::detail::gpu::create_engine( + : oneapi::math::rng::detail::engine_impl(queue) { + engine_ = oneapi::math::rng::detail::gpu::create_engine( queue, seed); } philox4x32x10_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - engine_ = oneapi::mkl::rng::detail::gpu::create_engine( + : oneapi::math::rng::detail::engine_impl(queue) { + engine_ = oneapi::math::rng::detail::gpu::create_engine( queue, (std::int64_t)(seed.size() * 2), (const unsigned int*)seed.begin()); } philox4x32x10_impl(const philox4x32x10_impl* other) - : oneapi::mkl::rng::detail::engine_impl(*other) { + : oneapi::math::rng::detail::engine_impl(*other) { sycl::queue queue(other->queue_); - engine_ = oneapi::mkl::rng::detail::gpu::create_engine( + engine_ = oneapi::math::rng::detail::gpu::create_engine( queue, other->engine_); } // Buffers API virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } - virtual void generate(const oneapi::mkl::rng::uniform< - std::int32_t, oneapi::mkl::rng::uniform_method::standard>& distr, + virtual void generate(const oneapi::math::rng::uniform< + std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } - virtual void generate(const oneapi::mkl::rng::gaussian< - float, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } - virtual void generate(const oneapi::mkl::rng::gaussian< - double, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } - virtual void generate(const oneapi::mkl::rng::lognormal< - float, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } - virtual void generate(const oneapi::mkl::rng::lognormal< - double, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } virtual void generate(const bits& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); } // USM APIs virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& + const oneapi::math::rng::uniform& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { ; - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } virtual sycl::event generate(const bits& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { return new philox4x32x10_impl(this); } virtual void skip_ahead(std::uint64_t num_to_skip) override { - oneapi::mkl::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); + oneapi::math::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); } virtual void skip_ahead(std::initializer_list num_to_skip) override { - oneapi::mkl::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); + oneapi::math::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "leapfrog"); + throw oneapi::math::unimplemented("rng", "leapfrog"); } virtual ~philox4x32x10_impl() override { - oneapi::mkl::rng::detail::gpu::delete_engine(queue_, engine_); + oneapi::math::rng::detail::gpu::delete_engine(queue_, engine_); } private: - oneapi::mkl::rng::detail::engine_base_impl* engine_; + oneapi::math::rng::detail::engine_base_impl* engine_; }; -oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { +oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { return new philox4x32x10_impl(queue, seed); } -oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed) { return new philox4x32x10_impl(queue, seed); } } // namespace mklgpu } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/rng/backends/rocrand/mrg32k3a.cpp b/src/rng/backends/rocrand/mrg32k3a.cpp index 419434159..20d62b33f 100644 --- a/src/rng/backends/rocrand/mrg32k3a.cpp +++ b/src/rng/backends/rocrand/mrg32k3a.cpp @@ -80,15 +80,15 @@ #include "oneapi/math/rng/engines.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace rocrand { #if !defined(_WIN64) -class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { +class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { public: mrg32k3a_impl(sycl::queue queue, std::uint32_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue), + : oneapi::math::rng::detail::engine_impl(queue), seed_(seed), offset_(0) { rocrand_status status; @@ -97,13 +97,13 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } mrg32k3a_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine", + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "mrg32ka engine", "multi-seed unsupported by rocRAND backend"); } mrg32k3a_impl(const mrg32k3a_impl* other) - : oneapi::mkl::rng::detail::engine_impl(*other), + : oneapi::math::rng::detail::engine_impl(*other), seed_(other->seed_), offset_(other->offset_) { rocrand_status status; @@ -117,7 +117,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { // Buffers API virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -135,7 +135,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -152,8 +152,8 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { range_transform_fp(queue_, distr.a(), distr.b(), n, r); } - virtual void generate(const oneapi::mkl::rng::uniform< - std::int32_t, oneapi::mkl::rng::uniform_method::standard>& distr, + virtual void generate(const oneapi::math::rng::uniform< + std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { sycl::buffer ib(n); queue_ @@ -172,7 +172,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -190,7 +190,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -207,8 +207,8 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { range_transform_fp_accurate(queue_, distr.a(), distr.b(), n, r); } - virtual void generate(const oneapi::mkl::rng::gaussian< - float, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -224,8 +224,8 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { increment_internal_offset(n); } - virtual void generate(const oneapi::mkl::rng::gaussian< - double, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -242,7 +242,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -259,7 +259,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -275,8 +275,8 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { increment_internal_offset(n); } - virtual void generate(const oneapi::mkl::rng::lognormal< - float, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -292,8 +292,8 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { increment_internal_offset(n); } - virtual void generate(const oneapi::mkl::rng::lognormal< - double, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -310,7 +310,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -327,7 +327,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -345,14 +345,14 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "Bernoulli distribution method unsupported by rocRAND backend"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "Bernoulli distribution method unsupported by rocRAND backend"); } @@ -407,7 +407,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { // USM APIs virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -425,7 +425,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -443,7 +443,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& + const oneapi::math::rng::uniform& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { std::uint32_t* ib = (std::uint32_t*)malloc_device( @@ -463,7 +463,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -481,7 +481,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -499,7 +499,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -517,7 +517,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -535,7 +535,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -552,7 +552,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -569,7 +569,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -587,7 +587,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -605,7 +605,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -622,7 +622,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -641,7 +641,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "Bernoulli distribution method unsupported by rocRAND backend"); return sycl::event{}; @@ -650,7 +650,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "Bernoulli distribution method unsupported by rocRAND backend"); return sycl::event{}; @@ -705,7 +705,7 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { return event; } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { return new mrg32k3a_impl(this); } @@ -715,12 +715,12 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void skip_ahead(std::initializer_list num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "skip_ahead", + throw oneapi::math::unimplemented("rng", "skip_ahead", "initializer list unsupported by rocRAND backend"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "leapfrog", "unsupported by rocRAND backend"); + throw oneapi::math::unimplemented("rng", "leapfrog", "unsupported by rocRAND backend"); } virtual ~mrg32k3a_impl() override { @@ -737,290 +737,290 @@ class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { } }; #else // rocRAND backend is currently not supported on Windows -class mrg32k3a_impl : public oneapi::mkl::rng::detail::engine_impl { +class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { public: mrg32k3a_impl(sycl::queue queue, std::uint32_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } mrg32k3a_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::mkl::rng::detail::engine_impl(*other) { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::math::rng::detail::engine_impl(*other) { + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } // Buffers API virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate(const oneapi::mkl::rng::uniform< - std::int32_t, oneapi::mkl::rng::uniform_method::standard>& distr, + virtual void generate(const oneapi::math::rng::uniform< + std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate(const oneapi::mkl::rng::gaussian< - float, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate(const oneapi::mkl::rng::gaussian< - double, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate(const oneapi::mkl::rng::lognormal< - float, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate(const oneapi::mkl::rng::lognormal< - double, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void generate(const bits& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } // USM APIs virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& + const oneapi::math::rng::uniform& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } virtual sycl::event generate(const bits& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return nullptr; } virtual void skip_ahead(std::uint64_t num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void skip_ahead(std::initializer_list num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "mrg32ka engine"); + throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } virtual ~mrg32k3a_impl() override {} }; #endif -oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed) { +oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed) { return new mrg32k3a_impl(queue, seed); } -oneapi::mkl::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, +oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::initializer_list seed) { return new mrg32k3a_impl(queue, seed); } } // namespace rocrand } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/rng/backends/rocrand/philox4x32x10.cpp b/src/rng/backends/rocrand/philox4x32x10.cpp index e3656a43d..6753b9c1b 100644 --- a/src/rng/backends/rocrand/philox4x32x10.cpp +++ b/src/rng/backends/rocrand/philox4x32x10.cpp @@ -80,7 +80,7 @@ // #include "oneapi/math/rng/engines.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace rocrand { @@ -107,10 +107,10 @@ namespace rocrand { * generates random numbers. * */ -class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { +class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { public: philox4x32x10_impl(sycl::queue queue, std::uint64_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue), + : oneapi::math::rng::detail::engine_impl(queue), seed_(seed), offset_(0) { rocrand_status status; @@ -119,13 +119,13 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } philox4x32x10_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine", + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine", "multi-seed unsupported by rocRAND backend"); } philox4x32x10_impl(const philox4x32x10_impl* other) - : oneapi::mkl::rng::detail::engine_impl(*other), + : oneapi::math::rng::detail::engine_impl(*other), seed_(other->seed_), offset_(other->offset_) { rocrand_status status; @@ -139,7 +139,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { // Buffers API virtual inline void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -157,7 +157,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -174,8 +174,8 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { range_transform_fp(queue_, distr.a(), distr.b(), n, r); } - virtual void generate(const oneapi::mkl::rng::uniform< - std::int32_t, oneapi::mkl::rng::uniform_method::standard>& distr, + virtual void generate(const oneapi::math::rng::uniform< + std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { sycl::buffer ib(n); queue_ @@ -194,7 +194,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -212,7 +212,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -229,8 +229,8 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { range_transform_fp_accurate(queue_, distr.a(), distr.b(), n, r); } - virtual void generate(const oneapi::mkl::rng::gaussian< - float, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -246,8 +246,8 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { increment_internal_offset(n); } - virtual void generate(const oneapi::mkl::rng::gaussian< - double, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -264,7 +264,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -281,7 +281,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -297,8 +297,8 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { increment_internal_offset(n); } - virtual void generate(const oneapi::mkl::rng::lognormal< - float, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -314,8 +314,8 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { increment_internal_offset(n); } - virtual void generate(const oneapi::mkl::rng::lognormal< - double, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -332,7 +332,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -349,7 +349,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { @@ -367,14 +367,14 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "Bernoulli distribution method unsupported by rocRAND backend"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "Bernoulli distribution method unsupported by rocRAND backend"); } @@ -429,7 +429,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { // USM APIs virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -447,7 +447,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -465,7 +465,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& + const oneapi::math::rng::uniform& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { std::uint32_t* ib = (std::uint32_t*)malloc_device( @@ -485,7 +485,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -503,7 +503,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -521,7 +521,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -539,7 +539,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -557,7 +557,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -574,7 +574,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -591,7 +591,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -609,7 +609,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); @@ -627,7 +627,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -644,7 +644,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -663,7 +663,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "Bernoulli distribution method unsupported by rocRAND backend"); return sycl::event{}; @@ -672,7 +672,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "Bernoulli distribution method unsupported by rocRAND backend"); return sycl::event{}; @@ -726,7 +726,7 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { return event; } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { return new philox4x32x10_impl(this); } @@ -736,12 +736,12 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } virtual void skip_ahead(std::initializer_list num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "skip_ahead", + throw oneapi::math::unimplemented("rng", "skip_ahead", "initializer list unsupported by rocRAND backend"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "leapfrog", "unsupported by rocRAND backend"); + throw oneapi::math::unimplemented("rng", "leapfrog", "unsupported by rocRAND backend"); } virtual ~philox4x32x10_impl() override { @@ -758,291 +758,291 @@ class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { } }; #else // rocRAND backend is currently not supported on Windows -class philox4x32x10_impl : public oneapi::mkl::rng::detail::engine_impl { +class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { public: philox4x32x10_impl(sycl::queue queue, std::uint64_t seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } philox4x32x10_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::mkl::rng::detail::engine_impl(queue) { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + : oneapi::math::rng::detail::engine_impl(queue) { + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } philox4x32x10_impl(const philox4x32x10_impl* other) - : oneapi::mkl::rng::detail::engine_impl(*other) { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + : oneapi::math::rng::detail::engine_impl(*other) { + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } // Buffers API virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate(const oneapi::mkl::rng::uniform< - std::int32_t, oneapi::mkl::rng::uniform_method::standard>& distr, + virtual void generate(const oneapi::math::rng::uniform< + std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate(const oneapi::mkl::rng::gaussian< - float, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate(const oneapi::mkl::rng::gaussian< - double, oneapi::mkl::rng::gaussian_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::gaussian< + double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate(const oneapi::mkl::rng::lognormal< - float, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate(const oneapi::mkl::rng::lognormal< - double, oneapi::mkl::rng::lognormal_method::box_muller2>& distr, + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void generate(const bits& distr, std::int64_t n, sycl::buffer& r) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } // USM APIs virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& + const oneapi::math::rng::uniform& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::uniform& distr, + const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::gaussian& distr, + const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( - const oneapi::mkl::rng::lognormal& distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } virtual sycl::event generate(const bits& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; } - virtual oneapi::mkl::rng::detail::engine_impl* copy_state() override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + virtual oneapi::math::rng::detail::engine_impl* copy_state() override { + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return nullptr; } virtual void skip_ahead(std::uint64_t num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void skip_ahead(std::initializer_list num_to_skip) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { - throw oneapi::mkl::unimplemented("rng", "philox4x32x10 engine"); + throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } virtual ~philox4x32x10_impl() override {} }; #endif -oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { +oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { return new philox4x32x10_impl(queue, seed); } -oneapi::mkl::rng::detail::engine_impl* create_philox4x32x10( +oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed) { return new philox4x32x10_impl(queue, seed); } } // namespace rocrand } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/rng/backends/rocrand/rocrand_helper.hpp b/src/rng/backends/rocrand/rocrand_helper.hpp index e1d51dee3..a64da3236 100644 --- a/src/rng/backends/rocrand/rocrand_helper.hpp +++ b/src/rng/backends/rocrand/rocrand_helper.hpp @@ -70,11 +70,11 @@ #include "oneapi/math/types.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace rocrand { -// Static template functions oneapi::mkl::rng::rocrand::range_transform_fp for +// Static template functions oneapi::math::rng::rocrand::range_transform_fp for // Buffer and USM APIs // // rocRAND has no built-in functionality to specify a custom range for sampling @@ -138,7 +138,7 @@ static inline sycl::event range_transform_fp_accurate(sycl::queue& queue, T a, T }); } -// Static template functions oneapi::mkl::rng::rocrand::range_transform_int for +// Static template functions oneapi::math::rng::rocrand::range_transform_int for // Buffer and USM APIs // // rocRAND has no built-in functionality to specify a custom range for sampling @@ -173,7 +173,7 @@ inline sycl::event range_transform_int(sycl::queue& queue, T a, T b, std::int64_ }); } -// Static template functions oneapi::mkl::rng::rocrand::sample_bernoulli for +// Static template functions oneapi::math::rng::rocrand::sample_bernoulli for // Buffer and USM APIs // // rocRAND has no built-in functionality to sample from a Bernoulli distribution. @@ -329,7 +329,7 @@ class rocm_error : virtual public std::runtime_error { } // namespace rocrand } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _MKL_RNG_ROCRAND_HELPER_HPP_ diff --git a/src/rng/backends/rocrand/rocrand_task.hpp b/src/rng/backends/rocrand/rocrand_task.hpp index 3256a6e96..723721d7e 100644 --- a/src/rng/backends/rocrand/rocrand_task.hpp +++ b/src/rng/backends/rocrand/rocrand_task.hpp @@ -10,7 +10,7 @@ #include "rocrand_helper.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace rocrand { #ifdef __HIPSYCL__ @@ -84,7 +84,7 @@ static inline void onemath_rocrand_host_task(H &cgh, Engine e, F f) { } // namespace rocrand } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi #endif diff --git a/src/rng/backends/rocrand/rocrand_wrappers.cpp b/src/rng/backends/rocrand/rocrand_wrappers.cpp index 5a399a6af..821909121 100644 --- a/src/rng/backends/rocrand/rocrand_wrappers.cpp +++ b/src/rng/backends/rocrand/rocrand_wrappers.cpp @@ -64,7 +64,7 @@ #define WRAPPER_VERSION 1 extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { - WRAPPER_VERSION, oneapi::mkl::rng::rocrand::create_philox4x32x10, - oneapi::mkl::rng::rocrand::create_philox4x32x10, oneapi::mkl::rng::rocrand::create_mrg32k3a, - oneapi::mkl::rng::rocrand::create_mrg32k3a + WRAPPER_VERSION, oneapi::math::rng::rocrand::create_philox4x32x10, + oneapi::math::rng::rocrand::create_philox4x32x10, oneapi::math::rng::rocrand::create_mrg32k3a, + oneapi::math::rng::rocrand::create_mrg32k3a }; diff --git a/src/rng/function_table.hpp b/src/rng/function_table.hpp index 0708d5fbb..da3ca1b6f 100644 --- a/src/rng/function_table.hpp +++ b/src/rng/function_table.hpp @@ -32,14 +32,14 @@ typedef struct { int version; - oneapi::mkl::rng::detail::engine_impl* (*create_philox4x32x10_sycl)(sycl::queue queue, + oneapi::math::rng::detail::engine_impl* (*create_philox4x32x10_sycl)(sycl::queue queue, std::uint64_t seed); - oneapi::mkl::rng::detail::engine_impl* (*create_philox4x32x10_ex_sycl)( + oneapi::math::rng::detail::engine_impl* (*create_philox4x32x10_ex_sycl)( sycl::queue queue, std::initializer_list seed); - oneapi::mkl::rng::detail::engine_impl* (*create_mrg32k3a_sycl)(sycl::queue queue, + oneapi::math::rng::detail::engine_impl* (*create_mrg32k3a_sycl)(sycl::queue queue, std::uint32_t seed); - oneapi::mkl::rng::detail::engine_impl* (*create_mrg32k3a_ex_sycl)( + oneapi::math::rng::detail::engine_impl* (*create_mrg32k3a_ex_sycl)( sycl::queue queue, std::initializer_list seed); } rng_function_table_t; diff --git a/src/rng/rng_loader.cpp b/src/rng/rng_loader.cpp index fa969bb3c..3bd78f522 100644 --- a/src/rng/rng_loader.cpp +++ b/src/rng/rng_loader.cpp @@ -23,32 +23,32 @@ #include "rng/function_table.hpp" namespace oneapi { -namespace mkl { +namespace math { namespace rng { namespace detail { -static oneapi::mkl::detail::table_initializer function_tables; +static oneapi::math::detail::table_initializer function_tables; -engine_impl* create_philox4x32x10(oneapi::mkl::device libkey, sycl::queue queue, +engine_impl* create_philox4x32x10(oneapi::math::device libkey, sycl::queue queue, std::uint64_t seed) { return function_tables[{ libkey, queue }].create_philox4x32x10_sycl(queue, seed); } -engine_impl* create_philox4x32x10(oneapi::mkl::device libkey, sycl::queue queue, +engine_impl* create_philox4x32x10(oneapi::math::device libkey, sycl::queue queue, std::initializer_list seed) { return function_tables[{ libkey, queue }].create_philox4x32x10_ex_sycl(queue, seed); } -engine_impl* create_mrg32k3a(oneapi::mkl::device libkey, sycl::queue queue, std::uint32_t seed) { +engine_impl* create_mrg32k3a(oneapi::math::device libkey, sycl::queue queue, std::uint32_t seed) { return function_tables[{ libkey, queue }].create_mrg32k3a_sycl(queue, seed); } -engine_impl* create_mrg32k3a(oneapi::mkl::device libkey, sycl::queue queue, +engine_impl* create_mrg32k3a(oneapi::math::device libkey, sycl::queue queue, std::initializer_list seed) { return function_tables[{ libkey, queue }].create_mrg32k3a_ex_sycl(queue, seed); } } // namespace detail } // namespace rng -} // namespace mkl +} // namespace math } // namespace oneapi diff --git a/src/sparse_blas/backends/backend_wrappers.cxx b/src/sparse_blas/backends/backend_wrappers.cxx index fff743e82..99ef75a82 100644 --- a/src/sparse_blas/backends/backend_wrappers.cxx +++ b/src/sparse_blas/backends/backend_wrappers.cxx @@ -48,68 +48,68 @@ function template instantiations must be added to backend_sparse_blas_instantiat // clang-format off // Dense vector #define LIST_DENSE_VECTOR_FUNCS() \ -oneapi::mkl::sparse::BACKEND::init_dense_vector, \ -oneapi::mkl::sparse::BACKEND::init_dense_vector, \ -oneapi::mkl::sparse::BACKEND::set_dense_vector_data, \ -oneapi::mkl::sparse::BACKEND::set_dense_vector_data, +oneapi::math::sparse::BACKEND::init_dense_vector, \ +oneapi::math::sparse::BACKEND::init_dense_vector, \ +oneapi::math::sparse::BACKEND::set_dense_vector_data, \ +oneapi::math::sparse::BACKEND::set_dense_vector_data, REPEAT_FOR_EACH_FP_TYPE(LIST_DENSE_VECTOR_FUNCS) #undef LIST_DENSE_VECTOR_FUNCS -oneapi::mkl::sparse::BACKEND::release_dense_vector, +oneapi::math::sparse::BACKEND::release_dense_vector, // Dense matrix #define LIST_DENSE_MATRIX_FUNCS() \ -oneapi::mkl::sparse::BACKEND::init_dense_matrix, \ -oneapi::mkl::sparse::BACKEND::init_dense_matrix, \ -oneapi::mkl::sparse::BACKEND::set_dense_matrix_data, \ -oneapi::mkl::sparse::BACKEND::set_dense_matrix_data, +oneapi::math::sparse::BACKEND::init_dense_matrix, \ +oneapi::math::sparse::BACKEND::init_dense_matrix, \ +oneapi::math::sparse::BACKEND::set_dense_matrix_data, \ +oneapi::math::sparse::BACKEND::set_dense_matrix_data, REPEAT_FOR_EACH_FP_TYPE(LIST_DENSE_MATRIX_FUNCS) #undef LIST_DENSE_MATRIX_FUNCS -oneapi::mkl::sparse::BACKEND::release_dense_matrix, +oneapi::math::sparse::BACKEND::release_dense_matrix, // COO matrix #define LIST_COO_MATRIX_FUNCS() \ -oneapi::mkl::sparse::BACKEND::init_coo_matrix, \ -oneapi::mkl::sparse::BACKEND::init_coo_matrix, \ -oneapi::mkl::sparse::BACKEND::set_coo_matrix_data, \ -oneapi::mkl::sparse::BACKEND::set_coo_matrix_data, +oneapi::math::sparse::BACKEND::init_coo_matrix, \ +oneapi::math::sparse::BACKEND::init_coo_matrix, \ +oneapi::math::sparse::BACKEND::set_coo_matrix_data, \ +oneapi::math::sparse::BACKEND::set_coo_matrix_data, REPEAT_FOR_EACH_FP_AND_INT_TYPE(LIST_COO_MATRIX_FUNCS) #undef LIST_COO_MATRIX_FUNCS // CSR matrix #define LIST_CSR_MATRIX_FUNCS() \ -oneapi::mkl::sparse::BACKEND::init_csr_matrix, \ -oneapi::mkl::sparse::BACKEND::init_csr_matrix, \ -oneapi::mkl::sparse::BACKEND::set_csr_matrix_data, \ -oneapi::mkl::sparse::BACKEND::set_csr_matrix_data, +oneapi::math::sparse::BACKEND::init_csr_matrix, \ +oneapi::math::sparse::BACKEND::init_csr_matrix, \ +oneapi::math::sparse::BACKEND::set_csr_matrix_data, \ +oneapi::math::sparse::BACKEND::set_csr_matrix_data, REPEAT_FOR_EACH_FP_AND_INT_TYPE(LIST_CSR_MATRIX_FUNCS) #undef LIST_CSR_MATRIX_FUNCS // Common sparse matrix functions -oneapi::mkl::sparse::BACKEND::release_sparse_matrix, -oneapi::mkl::sparse::BACKEND::set_matrix_property, +oneapi::math::sparse::BACKEND::release_sparse_matrix, +oneapi::math::sparse::BACKEND::set_matrix_property, // SPMM -oneapi::mkl::sparse::BACKEND::init_spmm_descr, -oneapi::mkl::sparse::BACKEND::release_spmm_descr, -oneapi::mkl::sparse::BACKEND::spmm_buffer_size, -oneapi::mkl::sparse::BACKEND::spmm_optimize, -oneapi::mkl::sparse::BACKEND::spmm_optimize, -oneapi::mkl::sparse::BACKEND::spmm, +oneapi::math::sparse::BACKEND::init_spmm_descr, +oneapi::math::sparse::BACKEND::release_spmm_descr, +oneapi::math::sparse::BACKEND::spmm_buffer_size, +oneapi::math::sparse::BACKEND::spmm_optimize, +oneapi::math::sparse::BACKEND::spmm_optimize, +oneapi::math::sparse::BACKEND::spmm, // SPMV -oneapi::mkl::sparse::BACKEND::init_spmv_descr, -oneapi::mkl::sparse::BACKEND::release_spmv_descr, -oneapi::mkl::sparse::BACKEND::spmv_buffer_size, -oneapi::mkl::sparse::BACKEND::spmv_optimize, -oneapi::mkl::sparse::BACKEND::spmv_optimize, -oneapi::mkl::sparse::BACKEND::spmv, +oneapi::math::sparse::BACKEND::init_spmv_descr, +oneapi::math::sparse::BACKEND::release_spmv_descr, +oneapi::math::sparse::BACKEND::spmv_buffer_size, +oneapi::math::sparse::BACKEND::spmv_optimize, +oneapi::math::sparse::BACKEND::spmv_optimize, +oneapi::math::sparse::BACKEND::spmv, // SPSV -oneapi::mkl::sparse::BACKEND::init_spsv_descr, -oneapi::mkl::sparse::BACKEND::release_spsv_descr, -oneapi::mkl::sparse::BACKEND::spsv_buffer_size, -oneapi::mkl::sparse::BACKEND::spsv_optimize, -oneapi::mkl::sparse::BACKEND::spsv_optimize, -oneapi::mkl::sparse::BACKEND::spsv, +oneapi::math::sparse::BACKEND::init_spsv_descr, +oneapi::math::sparse::BACKEND::release_spsv_descr, +oneapi::math::sparse::BACKEND::spsv_buffer_size, +oneapi::math::sparse::BACKEND::spsv_optimize, +oneapi::math::sparse::BACKEND::spsv_optimize, +oneapi::math::sparse::BACKEND::spsv, // clang-format on diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.cxx b/src/sparse_blas/backends/mkl_common/mkl_handles.cxx index 85e973fe6..ea3f3abb3 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.cxx @@ -20,37 +20,37 @@ // Dense vector template void init_dense_vector(sycl::queue & /*queue*/, - oneapi::mkl::sparse::dense_vector_handle_t *p_dvhandle, std::int64_t size, + oneapi::math::sparse::dense_vector_handle_t *p_dvhandle, std::int64_t size, sycl::buffer val) { - *p_dvhandle = new oneapi::mkl::sparse::dense_vector_handle(val, size); + *p_dvhandle = new oneapi::math::sparse::dense_vector_handle(val, size); } template void init_dense_vector(sycl::queue & /*queue*/, - oneapi::mkl::sparse::dense_vector_handle_t *p_dvhandle, std::int64_t size, + oneapi::math::sparse::dense_vector_handle_t *p_dvhandle, std::int64_t size, fpType *val) { - *p_dvhandle = new oneapi::mkl::sparse::dense_vector_handle(val, size); + *p_dvhandle = new oneapi::math::sparse::dense_vector_handle(val, size); } template void check_can_reset_value_handle(const std::string &function_name, InternalHandleT *internal_handle, bool expect_buffer) { if (internal_handle->get_value_type() != detail::get_data_type()) { - throw oneapi::mkl::invalid_argument( + throw oneapi::math::invalid_argument( "sparse_blas", function_name, "Incompatible data types expected " + data_type_to_str(internal_handle->get_value_type()) + " but got " + data_type_to_str(detail::get_data_type())); } if (internal_handle->all_use_buffer() != expect_buffer) { - throw oneapi::mkl::invalid_argument( + throw oneapi::math::invalid_argument( "sparse_blas", function_name, "Cannot change the container type between buffer or USM"); } } template void set_dense_vector_data(sycl::queue & /*queue*/, - oneapi::mkl::sparse::dense_vector_handle_t dvhandle, std::int64_t size, + oneapi::math::sparse::dense_vector_handle_t dvhandle, std::int64_t size, sycl::buffer val) { check_can_reset_value_handle(__func__, dvhandle, true); dvhandle->size = size; @@ -59,7 +59,7 @@ void set_dense_vector_data(sycl::queue & /*queue*/, template void set_dense_vector_data(sycl::queue & /*queue*/, - oneapi::mkl::sparse::dense_vector_handle_t dvhandle, std::int64_t size, + oneapi::math::sparse::dense_vector_handle_t dvhandle, std::int64_t size, fpType *val) { check_can_reset_value_handle(__func__, dvhandle, false); dvhandle->size = size; @@ -68,22 +68,22 @@ void set_dense_vector_data(sycl::queue & /*queue*/, #define INSTANTIATE_DENSE_VECTOR_FUNCS(FP_TYPE, FP_SUFFIX) \ template void init_dense_vector( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_vector_handle_t * p_dvhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t * p_dvhandle, \ std::int64_t size, sycl::buffer val); \ template void init_dense_vector( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_vector_handle_t * p_dvhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t * p_dvhandle, \ std::int64_t size, FP_TYPE * val); \ template void set_dense_vector_data( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_vector_handle_t dvhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t dvhandle, \ std::int64_t size, sycl::buffer val); \ template void set_dense_vector_data( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_vector_handle_t dvhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t dvhandle, \ std::int64_t size, FP_TYPE * val) FOR_EACH_FP_TYPE(INSTANTIATE_DENSE_VECTOR_FUNCS); #undef INSTANTIATE_DENSE_VECTOR_FUNCS sycl::event release_dense_vector(sycl::queue &queue, - oneapi::mkl::sparse::dense_vector_handle_t dvhandle, + oneapi::math::sparse::dense_vector_handle_t dvhandle, const std::vector &dependencies) { return detail::submit_release(queue, dvhandle, dependencies); } @@ -91,27 +91,27 @@ sycl::event release_dense_vector(sycl::queue &queue, // Dense matrix template void init_dense_matrix(sycl::queue & /*queue*/, - oneapi::mkl::sparse::dense_matrix_handle_t *p_dmhandle, + oneapi::math::sparse::dense_matrix_handle_t *p_dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, - oneapi::mkl::layout dense_layout, sycl::buffer val) { + oneapi::math::layout dense_layout, sycl::buffer val) { *p_dmhandle = - new oneapi::mkl::sparse::dense_matrix_handle(val, num_rows, num_cols, ld, dense_layout); + new oneapi::math::sparse::dense_matrix_handle(val, num_rows, num_cols, ld, dense_layout); } template void init_dense_matrix(sycl::queue & /*queue*/, - oneapi::mkl::sparse::dense_matrix_handle_t *p_dmhandle, + oneapi::math::sparse::dense_matrix_handle_t *p_dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, - oneapi::mkl::layout dense_layout, fpType *val) { + oneapi::math::layout dense_layout, fpType *val) { *p_dmhandle = - new oneapi::mkl::sparse::dense_matrix_handle(val, num_rows, num_cols, ld, dense_layout); + new oneapi::math::sparse::dense_matrix_handle(val, num_rows, num_cols, ld, dense_layout); } template void set_dense_matrix_data(sycl::queue & /*queue*/, - oneapi::mkl::sparse::dense_matrix_handle_t dmhandle, + oneapi::math::sparse::dense_matrix_handle_t dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, - oneapi::mkl::layout dense_layout, sycl::buffer val) { + oneapi::math::layout dense_layout, sycl::buffer val) { check_can_reset_value_handle(__func__, dmhandle, true); dmhandle->num_rows = num_rows; dmhandle->num_cols = num_cols; @@ -122,9 +122,9 @@ void set_dense_matrix_data(sycl::queue & /*queue*/, template void set_dense_matrix_data(sycl::queue & /*queue*/, - oneapi::mkl::sparse::dense_matrix_handle_t dmhandle, + oneapi::math::sparse::dense_matrix_handle_t dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, - oneapi::mkl::layout dense_layout, fpType *val) { + oneapi::math::layout dense_layout, fpType *val) { check_can_reset_value_handle(__func__, dmhandle, false); dmhandle->num_rows = num_rows; dmhandle->num_cols = num_cols; @@ -135,35 +135,35 @@ void set_dense_matrix_data(sycl::queue & /*queue*/, #define INSTANTIATE_DENSE_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX) \ template void init_dense_matrix( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_matrix_handle_t * p_dmhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t * p_dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::mkl::layout dense_layout, sycl::buffer val); \ + oneapi::math::layout dense_layout, sycl::buffer val); \ template void init_dense_matrix( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_matrix_handle_t * p_dmhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t * p_dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::mkl::layout dense_layout, FP_TYPE * val); \ + oneapi::math::layout dense_layout, FP_TYPE * val); \ template void set_dense_matrix_data( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_matrix_handle_t dmhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::mkl::layout dense_layout, sycl::buffer val); \ + oneapi::math::layout dense_layout, sycl::buffer val); \ template void set_dense_matrix_data( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_matrix_handle_t dmhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::mkl::layout dense_layout, FP_TYPE * val) + oneapi::math::layout dense_layout, FP_TYPE * val) FOR_EACH_FP_TYPE(INSTANTIATE_DENSE_MATRIX_FUNCS); #undef INSTANTIATE_DENSE_MATRIX_FUNCS sycl::event release_dense_matrix(sycl::queue &queue, - oneapi::mkl::sparse::dense_matrix_handle_t dmhandle, + oneapi::math::sparse::dense_matrix_handle_t dmhandle, const std::vector &dependencies) { return detail::submit_release(queue, dmhandle, dependencies); } // COO matrix template -void init_coo_matrix(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t *p_smhandle, +void init_coo_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, - oneapi::mkl::index_base index, sycl::buffer row_ind, + oneapi::math::index_base index, sycl::buffer row_ind, sycl::buffer col_ind, sycl::buffer val) { oneapi::mkl::sparse::matrix_handle_t mkl_handle; oneapi::mkl::sparse::init_matrix_handle(&mkl_handle); @@ -174,13 +174,13 @@ void init_coo_matrix(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t *p index, internal_smhandle->row_container.get_buffer(), internal_smhandle->col_container.get_buffer(), internal_smhandle->value_container.get_buffer()); - *p_smhandle = reinterpret_cast(internal_smhandle); + *p_smhandle = reinterpret_cast(internal_smhandle); } template -void init_coo_matrix(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t *p_smhandle, +void init_coo_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, - oneapi::mkl::index_base index, intType *row_ind, intType *col_ind, + oneapi::math::index_base index, intType *row_ind, intType *col_ind, fpType *val) { oneapi::mkl::sparse::matrix_handle_t mkl_handle; oneapi::mkl::sparse::init_matrix_handle(&mkl_handle); @@ -189,7 +189,7 @@ void init_coo_matrix(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t *p queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), static_cast(nnz), index, row_ind, col_ind, val); event.wait_and_throw(); - *p_smhandle = reinterpret_cast(internal_smhandle); + *p_smhandle = reinterpret_cast(internal_smhandle); } template @@ -198,23 +198,23 @@ void check_can_reset_sparse_handle(const std::string &function_name, bool expect_buffer) { check_can_reset_value_handle(function_name, internal_smhandle, expect_buffer); if (internal_smhandle->get_int_type() != detail::get_data_type()) { - throw oneapi::mkl::invalid_argument( + throw oneapi::math::invalid_argument( "sparse_blas", function_name, "Incompatible data types expected " + data_type_to_str(internal_smhandle->get_int_type()) + " but got " + data_type_to_str(detail::get_data_type())); } if (!internal_smhandle->can_be_reset) { - throw mkl::unimplemented( + throw math::unimplemented( "sparse_blas/mkl", function_name, "Reseting the matrix handle's data after it was used in a computation is not supported."); } } template -void set_coo_matrix_data(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t smhandle, +void set_coo_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, - oneapi::mkl::index_base index, sycl::buffer row_ind, + oneapi::math::index_base index, sycl::buffer row_ind, sycl::buffer col_ind, sycl::buffer val) { auto internal_smhandle = detail::get_internal_handle(smhandle); check_can_reset_sparse_handle(__func__, internal_smhandle, true); @@ -231,9 +231,9 @@ void set_coo_matrix_data(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_ } template -void set_coo_matrix_data(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t smhandle, +void set_coo_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, - oneapi::mkl::index_base index, intType *row_ind, intType *col_ind, + oneapi::math::index_base index, intType *row_ind, intType *col_ind, fpType *val) { auto internal_smhandle = detail::get_internal_handle(smhandle); check_can_reset_sparse_handle(__func__, internal_smhandle, false); @@ -248,31 +248,31 @@ void set_coo_matrix_data(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_ #define INSTANTIATE_COO_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ template void init_coo_matrix( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t * p_smhandle, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::mkl::index_base index, sycl::buffer row_ind, \ + oneapi::math::index_base index, sycl::buffer row_ind, \ sycl::buffer col_ind, sycl::buffer val); \ template void init_coo_matrix( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t * p_smhandle, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::mkl::index_base index, INT_TYPE * row_ind, INT_TYPE * col_ind, FP_TYPE * val); \ + oneapi::math::index_base index, INT_TYPE * row_ind, INT_TYPE * col_ind, FP_TYPE * val); \ template void set_coo_matrix_data( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::mkl::index_base index, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ + std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ sycl::buffer row_ind, sycl::buffer col_ind, \ sycl::buffer val); \ template void set_coo_matrix_data( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::mkl::index_base index, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ + std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ INT_TYPE * row_ind, INT_TYPE * col_ind, FP_TYPE * val) FOR_EACH_FP_AND_INT_TYPE(INSTANTIATE_COO_MATRIX_FUNCS); #undef INSTANTIATE_COO_MATRIX_FUNCS // CSR matrix template -void init_csr_matrix(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t *p_smhandle, +void init_csr_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t /*nnz*/, - oneapi::mkl::index_base index, sycl::buffer row_ptr, + oneapi::math::index_base index, sycl::buffer row_ptr, sycl::buffer col_ind, sycl::buffer val) { oneapi::mkl::sparse::matrix_handle_t mkl_handle; oneapi::mkl::sparse::init_matrix_handle(&mkl_handle); @@ -284,13 +284,13 @@ void init_csr_matrix(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t *p internal_smhandle->row_container.get_buffer(), internal_smhandle->col_container.get_buffer(), internal_smhandle->value_container.get_buffer()); - *p_smhandle = reinterpret_cast(internal_smhandle); + *p_smhandle = reinterpret_cast(internal_smhandle); } template -void init_csr_matrix(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t *p_smhandle, +void init_csr_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t /*nnz*/, - oneapi::mkl::index_base index, intType *row_ptr, intType *col_ind, + oneapi::math::index_base index, intType *row_ptr, intType *col_ind, fpType *val) { oneapi::mkl::sparse::matrix_handle_t mkl_handle; oneapi::mkl::sparse::init_matrix_handle(&mkl_handle); @@ -300,13 +300,13 @@ void init_csr_matrix(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t *p queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), index, row_ptr, col_ind, val); event.wait_and_throw(); - *p_smhandle = reinterpret_cast(internal_smhandle); + *p_smhandle = reinterpret_cast(internal_smhandle); } template -void set_csr_matrix_data(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t smhandle, +void set_csr_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t /*nnz*/, - oneapi::mkl::index_base index, sycl::buffer row_ptr, + oneapi::math::index_base index, sycl::buffer row_ptr, sycl::buffer col_ind, sycl::buffer val) { auto internal_smhandle = detail::get_internal_handle(smhandle); check_can_reset_sparse_handle(__func__, internal_smhandle, true); @@ -324,9 +324,9 @@ void set_csr_matrix_data(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_ } template -void set_csr_matrix_data(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t smhandle, +void set_csr_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t /*nnz*/, - oneapi::mkl::index_base index, intType *row_ptr, intType *col_ind, + oneapi::math::index_base index, intType *row_ptr, intType *col_ind, fpType *val) { auto internal_smhandle = detail::get_internal_handle(smhandle); check_can_reset_sparse_handle(__func__, internal_smhandle, false); @@ -342,28 +342,28 @@ void set_csr_matrix_data(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_ #define INSTANTIATE_CSR_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ template void init_csr_matrix( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t * p_smhandle, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::mkl::index_base index, sycl::buffer row_ptr, \ + oneapi::math::index_base index, sycl::buffer row_ptr, \ sycl::buffer col_ind, sycl::buffer val); \ template void init_csr_matrix( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t * p_smhandle, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::mkl::index_base index, INT_TYPE * row_ptr, INT_TYPE * col_ind, FP_TYPE * val); \ + oneapi::math::index_base index, INT_TYPE * row_ptr, INT_TYPE * col_ind, FP_TYPE * val); \ template void set_csr_matrix_data( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::mkl::index_base index, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ + std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ sycl::buffer row_ptr, sycl::buffer col_ind, \ sycl::buffer val); \ template void set_csr_matrix_data( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::mkl::index_base index, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ + std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ INT_TYPE * row_ptr, INT_TYPE * col_ind, FP_TYPE * val) FOR_EACH_FP_AND_INT_TYPE(INSTANTIATE_CSR_MATRIX_FUNCS); #undef INSTANTIATE_CSR_MATRIX_FUNCS // Common sparse matrix functions -sycl::event release_sparse_matrix(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t smhandle, +sycl::event release_sparse_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, const std::vector &dependencies) { auto internal_smhandle = detail::get_internal_handle(smhandle); // Asynchronously release the backend's handle followed by the internal handle. @@ -372,19 +372,19 @@ sycl::event release_sparse_matrix(sycl::queue &queue, oneapi::mkl::sparse::matri return detail::submit_release(queue, internal_smhandle, event); } -bool set_matrix_property(sycl::queue & /*queue*/, oneapi::mkl::sparse::matrix_handle_t smhandle, - oneapi::mkl::sparse::matrix_property property) { +bool set_matrix_property(sycl::queue & /*queue*/, oneapi::math::sparse::matrix_handle_t smhandle, + oneapi::math::sparse::matrix_property property) { auto internal_smhandle = detail::get_internal_handle(smhandle); // Store the matrix property internally for better error checking internal_smhandle->set_matrix_property(property); // Set the matrix property on the backend handle // Backend and oneMath types for the property don't match switch (property) { - case oneapi::mkl::sparse::matrix_property::symmetric: + case oneapi::math::sparse::matrix_property::symmetric: oneapi::mkl::sparse::set_matrix_property(internal_smhandle->backend_handle, oneapi::mkl::sparse::property::symmetric); return true; - case oneapi::mkl::sparse::matrix_property::sorted: + case oneapi::math::sparse::matrix_property::sorted: oneapi::mkl::sparse::set_matrix_property(internal_smhandle->backend_handle, oneapi::mkl::sparse::property::sorted); return true; diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp index e459c9b8c..e7b2a44e0 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp @@ -21,13 +21,13 @@ #define _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HANDLES_HPP_ // MKLCPU and MKLGPU backends include -// This include defines its own oneapi::mkl::sparse namespace with some of the +// This include defines its own oneapi::math::sparse namespace with some of the // types that are used here: matrix_handle_t, index_base, transpose, uplo, diag. #include #include "sparse_blas/generic_container.hpp" -namespace oneapi::mkl::sparse { +namespace oneapi::math::sparse { // Complete the definition of incomplete types dense_vector_handle and // dense_matrix_handle as they don't exist in oneMath backends yet. @@ -56,16 +56,16 @@ struct dense_matrix_handle : public detail::generic_dense_matrix_handle { ld, dense_layout) {} }; -} // namespace oneapi::mkl::sparse +} // namespace oneapi::math::sparse -namespace oneapi::mkl::sparse::detail { +namespace oneapi::math::sparse::detail { /** * Internal sparse_matrix_handle type for MKLCPU and MKLGPU backends. * Here \p matrix_handle_t is the type of the backend's handle. * The user-facing incomplete type matrix_handle_t must be kept incomplete. * Internally matrix_handle_t is reinterpret_cast as - * oneapi::mkl::sparse::detail::sparse_matrix_handle which holds another + * oneapi::math::sparse::detail::sparse_matrix_handle which holds another * matrix_handle_t for the backend handle. */ using sparse_matrix_handle = detail::generic_sparse_handle; @@ -75,6 +75,6 @@ inline auto get_internal_handle(matrix_handle_t handle) { return reinterpret_cast(handle); } -} // namespace oneapi::mkl::sparse::detail +} // namespace oneapi::math::sparse::detail #endif // _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HANDLES_HPP_ diff --git a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp index f55c42a1d..6ab54d560 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp @@ -32,7 +32,7 @@ #include "sparse_blas/enum_data_types.hpp" #include "sparse_blas/macros.hpp" -namespace oneapi::mkl::sparse::detail { +namespace oneapi::math::sparse::detail { /// Return whether a pointer is accessible on the host template @@ -47,7 +47,7 @@ inline void check_ptr_is_host_accessible(const std::string &function_name, const std::string &scalar_name, bool is_ptr_accessible_on_host) { if (!is_ptr_accessible_on_host) { - throw mkl::invalid_argument( + throw math::invalid_argument( "sparse_blas", function_name, "Scalar " + scalar_name + " must be accessible on the host for buffer functions."); } @@ -92,7 +92,7 @@ inline sycl::event collapse_dependencies(sycl::queue &queue, case detail::data_type::complex_fp64: \ return op_functor>(__VA_ARGS__); \ default: \ - throw oneapi::mkl::exception( \ + throw oneapi::math::exception( \ "sparse_blas", function_name, \ "Internal error: unsupported type " + data_type_to_str(value_type)); \ } @@ -100,12 +100,12 @@ inline sycl::event collapse_dependencies(sycl::queue &queue, #define CHECK_DESCR_MATCH(descr, argument, optimize_func_name) \ do { \ if (descr->last_optimized_##argument != argument) { \ - throw mkl::invalid_argument( \ + throw math::invalid_argument( \ "sparse_blas", __func__, \ #argument " argument must match with the previous call to " #optimize_func_name); \ } \ } while (0) -} // namespace oneapi::mkl::sparse::detail +} // namespace oneapi::math::sparse::detail #endif // _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HELPER_HPP_ diff --git a/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx b/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx index dce464228..0510aee31 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx @@ -17,38 +17,38 @@ * **************************************************************************/ -namespace oneapi::mkl::sparse { +namespace oneapi::math::sparse { struct spmm_descr { bool buffer_size_called = false; bool optimized_called = false; - oneapi::mkl::transpose last_optimized_opA; - oneapi::mkl::transpose last_optimized_opB; - oneapi::mkl::sparse::matrix_view last_optimized_A_view; - oneapi::mkl::sparse::matrix_handle_t last_optimized_A_handle; - oneapi::mkl::sparse::dense_matrix_handle_t last_optimized_B_handle; - oneapi::mkl::sparse::dense_matrix_handle_t last_optimized_C_handle; - oneapi::mkl::sparse::spmm_alg last_optimized_alg; + oneapi::math::transpose last_optimized_opA; + oneapi::math::transpose last_optimized_opB; + oneapi::math::sparse::matrix_view last_optimized_A_view; + oneapi::math::sparse::matrix_handle_t last_optimized_A_handle; + oneapi::math::sparse::dense_matrix_handle_t last_optimized_B_handle; + oneapi::math::sparse::dense_matrix_handle_t last_optimized_C_handle; + oneapi::math::sparse::spmm_alg last_optimized_alg; }; -} // namespace oneapi::mkl::sparse +} // namespace oneapi::math::sparse -namespace oneapi::mkl::sparse::BACKEND { +namespace oneapi::math::sparse::BACKEND { -void init_spmm_descr(sycl::queue & /*queue*/, oneapi::mkl::sparse::spmm_descr_t *p_spmm_descr) { +void init_spmm_descr(sycl::queue & /*queue*/, oneapi::math::sparse::spmm_descr_t *p_spmm_descr) { *p_spmm_descr = new spmm_descr(); } -sycl::event release_spmm_descr(sycl::queue &queue, oneapi::mkl::sparse::spmm_descr_t spmm_descr, +sycl::event release_spmm_descr(sycl::queue &queue, oneapi::math::sparse::spmm_descr_t spmm_descr, const std::vector &dependencies) { return detail::submit_release(queue, spmm_descr, dependencies); } -void check_valid_spmm(const std::string &function_name, oneapi::mkl::transpose opA, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_matrix_handle_t B_handle, - oneapi::mkl::sparse::dense_matrix_handle_t C_handle, +void check_valid_spmm(const std::string &function_name, oneapi::math::transpose opA, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_matrix_handle_t B_handle, + oneapi::math::sparse::dense_matrix_handle_t C_handle, bool is_alpha_host_accessible, bool is_beta_host_accessible) { THROW_IF_NULLPTR(function_name, A_handle); THROW_IF_NULLPTR(function_name, B_handle); @@ -61,22 +61,22 @@ void check_valid_spmm(const std::string &function_name, oneapi::mkl::transpose o detail::check_ptr_is_host_accessible("spmm", "beta", is_beta_host_accessible); } if (is_alpha_host_accessible != is_beta_host_accessible) { - throw mkl::invalid_argument( + throw math::invalid_argument( "sparse_blas", function_name, "Alpha and beta must both be placed on host memory or device memory."); } if (B_handle->dense_layout != C_handle->dense_layout) { - throw mkl::invalid_argument("sparse_blas", function_name, + throw math::invalid_argument("sparse_blas", function_name, "B and C matrices must used the same layout."); } - if (A_view.type_view != oneapi::mkl::sparse::matrix_descr::general) { - throw mkl::invalid_argument("sparse_blas", function_name, + if (A_view.type_view != oneapi::math::sparse::matrix_descr::general) { + throw math::invalid_argument("sparse_blas", function_name, "Matrix view's type must be `matrix_descr::general`."); } - if (A_view.diag_view != oneapi::mkl::diag::nonunit) { - throw mkl::invalid_argument("sparse_blas", function_name, + if (A_view.diag_view != oneapi::math::diag::nonunit) { + throw math::invalid_argument("sparse_blas", function_name, "Matrix's diag_view must be `nonunit`."); } @@ -84,9 +84,9 @@ void check_valid_spmm(const std::string &function_name, oneapi::mkl::transpose o detail::data_type data_type = internal_A_handle->get_value_type(); if ((data_type == detail::data_type::complex_fp32 || data_type == detail::data_type::complex_fp64) && - opA == oneapi::mkl::transpose::conjtrans && - internal_A_handle->has_matrix_property(oneapi::mkl::sparse::matrix_property::symmetric)) { - throw mkl::unimplemented( + opA == oneapi::math::transpose::conjtrans && + internal_A_handle->has_matrix_property(oneapi::math::sparse::matrix_property::symmetric)) { + throw math::unimplemented( "sparse_blas", function_name, "The backend does not support spmm using conjtrans and the symmetric property."); } @@ -95,14 +95,14 @@ void check_valid_spmm(const std::string &function_name, oneapi::mkl::transpose o #endif // BACKEND } -void spmm_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, - oneapi::mkl::transpose /*opB*/, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_matrix_handle_t B_handle, const void *beta, - oneapi::mkl::sparse::dense_matrix_handle_t C_handle, - oneapi::mkl::sparse::spmm_alg /*alg*/, - oneapi::mkl::sparse::spmm_descr_t spmm_descr, std::size_t &temp_buffer_size) { +void spmm_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, + oneapi::math::transpose /*opB*/, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t C_handle, + oneapi::math::sparse::spmm_alg /*alg*/, + oneapi::math::sparse::spmm_descr_t spmm_descr, std::size_t &temp_buffer_size) { // TODO: Add support for external workspace once the close-source oneMath backend supports it. bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); @@ -113,17 +113,17 @@ void spmm_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, } inline void common_spmm_optimize( - sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_matrix_handle_t B_handle, const void *beta, - oneapi::mkl::sparse::dense_matrix_handle_t C_handle, oneapi::mkl::sparse::spmm_alg alg, - oneapi::mkl::sparse::spmm_descr_t spmm_descr) { + sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, + oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg alg, + oneapi::math::sparse::spmm_descr_t spmm_descr) { bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); check_valid_spmm("spmm_optimize", opA, A_view, A_handle, B_handle, C_handle, is_alpha_host_accessible, is_beta_host_accessible); if (!spmm_descr->buffer_size_called) { - throw mkl::uninitialized("sparse_blas", "spmm_optimize", + throw math::uninitialized("sparse_blas", "spmm_optimize", "spmm_buffer_size must be called before spmm_optimize."); } spmm_descr->optimized_called = true; @@ -136,12 +136,12 @@ inline void common_spmm_optimize( spmm_descr->last_optimized_alg = alg; } -void spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, - const void *alpha, oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_matrix_handle_t B_handle, const void *beta, - oneapi::mkl::sparse::dense_matrix_handle_t C_handle, - oneapi::mkl::sparse::spmm_alg alg, oneapi::mkl::sparse::spmm_descr_t spmm_descr, +void spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void *alpha, oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t C_handle, + oneapi::math::sparse::spmm_alg alg, oneapi::math::sparse::spmm_descr_t spmm_descr, sycl::buffer /*workspace*/) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (!internal_A_handle->all_use_buffer()) { @@ -149,21 +149,21 @@ void spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl:: } common_spmm_optimize(queue, opA, opB, alpha, A_view, A_handle, B_handle, beta, C_handle, alg, spmm_descr); - if (alg == oneapi::mkl::sparse::spmm_alg::no_optimize_alg) { + if (alg == oneapi::math::sparse::spmm_alg::no_optimize_alg) { return; } internal_A_handle->can_be_reset = false; // TODO: Add support for spmm_optimize once the close-source oneMath backend supports it. } -sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, - oneapi::mkl::transpose opB, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_matrix_handle_t B_handle, const void *beta, - oneapi::mkl::sparse::dense_matrix_handle_t C_handle, - oneapi::mkl::sparse::spmm_alg alg, - oneapi::mkl::sparse::spmm_descr_t spmm_descr, void * /*workspace*/, +sycl::event spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t C_handle, + oneapi::math::sparse::spmm_alg alg, + oneapi::math::sparse::spmm_descr_t spmm_descr, void * /*workspace*/, const std::vector &dependencies) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (internal_A_handle->all_use_buffer()) { @@ -171,7 +171,7 @@ sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, } common_spmm_optimize(queue, opA, opB, alpha, A_view, A_handle, B_handle, beta, C_handle, alg, spmm_descr); - if (alg == oneapi::mkl::sparse::spmm_alg::no_optimize_alg) { + if (alg == oneapi::math::sparse::spmm_alg::no_optimize_alg) { return detail::collapse_dependencies(queue, dependencies); } internal_A_handle->can_be_reset = false; @@ -181,11 +181,11 @@ sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, template sycl::event internal_spmm( - sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, const void *alpha, - oneapi::mkl::sparse::matrix_view /*A_view*/, oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_matrix_handle_t B_handle, const void *beta, - oneapi::mkl::sparse::dense_matrix_handle_t C_handle, oneapi::mkl::sparse::spmm_alg /*alg*/, - oneapi::mkl::sparse::spmm_descr_t /*spmm_descr*/, const std::vector &dependencies, + sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, + oneapi::math::sparse::matrix_view /*A_view*/, oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg /*alg*/, + oneapi::math::sparse::spmm_descr_t /*spmm_descr*/, const std::vector &dependencies, bool is_alpha_host_accessible, bool is_beta_host_accessible) { T host_alpha = detail::get_scalar_on_host(queue, static_cast(alpha), is_alpha_host_accessible); @@ -212,12 +212,12 @@ sycl::event internal_spmm( } } -sycl::event spmm(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, - const void *alpha, oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_matrix_handle_t B_handle, const void *beta, - oneapi::mkl::sparse::dense_matrix_handle_t C_handle, - oneapi::mkl::sparse::spmm_alg alg, oneapi::mkl::sparse::spmm_descr_t spmm_descr, +sycl::event spmm(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void *alpha, oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t C_handle, + oneapi::math::sparse::spmm_alg alg, oneapi::math::sparse::spmm_descr_t spmm_descr, const std::vector &dependencies) { bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); @@ -225,7 +225,7 @@ sycl::event spmm(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::tr is_beta_host_accessible); if (!spmm_descr->optimized_called) { - throw mkl::uninitialized("sparse_blas", __func__, + throw math::uninitialized("sparse_blas", __func__, "spmm_optimize must be called before spmm."); } CHECK_DESCR_MATCH(spmm_descr, opA, "spmm_optimize"); @@ -242,4 +242,4 @@ sycl::event spmm(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::tr is_alpha_host_accessible, is_beta_host_accessible); } -} // namespace oneapi::mkl::sparse::BACKEND +} // namespace oneapi::math::sparse::BACKEND diff --git a/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx b/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx index a959f1f8c..097c79c27 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx @@ -17,37 +17,37 @@ * **************************************************************************/ -namespace oneapi::mkl::sparse { +namespace oneapi::math::sparse { struct spmv_descr { bool buffer_size_called = false; bool optimized_called = false; - oneapi::mkl::transpose last_optimized_opA; - oneapi::mkl::sparse::matrix_view last_optimized_A_view; - oneapi::mkl::sparse::matrix_handle_t last_optimized_A_handle; - oneapi::mkl::sparse::dense_vector_handle_t last_optimized_x_handle; - oneapi::mkl::sparse::dense_vector_handle_t last_optimized_y_handle; - oneapi::mkl::sparse::spmv_alg last_optimized_alg; + oneapi::math::transpose last_optimized_opA; + oneapi::math::sparse::matrix_view last_optimized_A_view; + oneapi::math::sparse::matrix_handle_t last_optimized_A_handle; + oneapi::math::sparse::dense_vector_handle_t last_optimized_x_handle; + oneapi::math::sparse::dense_vector_handle_t last_optimized_y_handle; + oneapi::math::sparse::spmv_alg last_optimized_alg; }; -} // namespace oneapi::mkl::sparse +} // namespace oneapi::math::sparse -namespace oneapi::mkl::sparse::BACKEND { +namespace oneapi::math::sparse::BACKEND { -void init_spmv_descr(sycl::queue & /*queue*/, oneapi::mkl::sparse::spmv_descr_t *p_spmv_descr) { +void init_spmv_descr(sycl::queue & /*queue*/, oneapi::math::sparse::spmv_descr_t *p_spmv_descr) { *p_spmv_descr = new spmv_descr(); } -sycl::event release_spmv_descr(sycl::queue &queue, oneapi::mkl::sparse::spmv_descr_t spmv_descr, +sycl::event release_spmv_descr(sycl::queue &queue, oneapi::math::sparse::spmv_descr_t spmv_descr, const std::vector &dependencies) { return detail::submit_release(queue, spmv_descr, dependencies); } -void check_valid_spmv(const std::string &function_name, oneapi::mkl::transpose opA, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, +void check_valid_spmv(const std::string &function_name, oneapi::math::transpose opA, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, bool is_alpha_host_accessible, bool is_beta_host_accessible) { THROW_IF_NULLPTR(function_name, A_handle); THROW_IF_NULLPTR(function_name, x_handle); @@ -60,38 +60,38 @@ void check_valid_spmv(const std::string &function_name, oneapi::mkl::transpose o detail::check_ptr_is_host_accessible("spmv", "beta", is_beta_host_accessible); } if (is_alpha_host_accessible != is_beta_host_accessible) { - throw mkl::invalid_argument( + throw math::invalid_argument( "sparse_blas", function_name, "Alpha and beta must both be placed on host memory or device memory."); } - if (A_view.type_view == oneapi::mkl::sparse::matrix_descr::diagonal) { - throw mkl::invalid_argument("sparse_blas", function_name, + if (A_view.type_view == oneapi::math::sparse::matrix_descr::diagonal) { + throw math::invalid_argument("sparse_blas", function_name, "Matrix view's type cannot be diagonal."); } - if (A_view.type_view != oneapi::mkl::sparse::matrix_descr::triangular && - A_view.diag_view == oneapi::mkl::diag::unit) { - throw mkl::invalid_argument( + if (A_view.type_view != oneapi::math::sparse::matrix_descr::triangular && + A_view.diag_view == oneapi::math::diag::unit) { + throw math::invalid_argument( "sparse_blas", function_name, "`unit` diag_view can only be used with a triangular type_view."); } - if ((A_view.type_view == oneapi::mkl::sparse::matrix_descr::symmetric || - A_view.type_view == oneapi::mkl::sparse::matrix_descr::hermitian) && - opA == oneapi::mkl::transpose::conjtrans) { - throw mkl::unimplemented( + if ((A_view.type_view == oneapi::math::sparse::matrix_descr::symmetric || + A_view.type_view == oneapi::math::sparse::matrix_descr::hermitian) && + opA == oneapi::math::transpose::conjtrans) { + throw math::unimplemented( "sparse_blas", function_name, "The backend does not support Symmetric or Hermitian matrix with `conjtrans`."); } } -void spmv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, const void *beta, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spmv_alg /*alg*/, - oneapi::mkl::sparse::spmv_descr_t spmv_descr, std::size_t &temp_buffer_size) { +void spmv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spmv_alg /*alg*/, + oneapi::math::sparse::spmv_descr_t spmv_descr, std::size_t &temp_buffer_size) { // TODO: Add support for external workspace once the close-source oneMath backend supports it. bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); @@ -101,20 +101,20 @@ void spmv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void spmv_descr->buffer_size_called = true; } -inline void common_spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, +inline void common_spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spmv_alg alg, - oneapi::mkl::sparse::spmv_descr_t spmv_descr) { + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spmv_alg alg, + oneapi::math::sparse::spmv_descr_t spmv_descr) { bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); check_valid_spmv("spmv_optimize", opA, A_view, A_handle, x_handle, y_handle, is_alpha_host_accessible, is_beta_host_accessible); if (!spmv_descr->buffer_size_called) { - throw mkl::uninitialized("sparse_blas", "spmv_optimize", + throw math::uninitialized("sparse_blas", "spmv_optimize", "spmv_buffer_size must be called before spmv_optimize."); } spmv_descr->optimized_called = true; @@ -126,12 +126,12 @@ inline void common_spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, spmv_descr->last_optimized_alg = alg; } -void spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, const void *beta, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spmv_alg alg, oneapi::mkl::sparse::spmv_descr_t spmv_descr, +void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::spmv_descr_t spmv_descr, sycl::buffer /*workspace*/) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (!internal_A_handle->all_use_buffer()) { @@ -139,7 +139,7 @@ void spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *a } common_spmv_optimize(queue, opA, alpha, A_view, A_handle, x_handle, beta, y_handle, alg, spmv_descr); - if (alg == oneapi::mkl::sparse::spmv_alg::no_optimize_alg) { + if (alg == oneapi::math::sparse::spmv_alg::no_optimize_alg) { return; } internal_A_handle->can_be_reset = false; @@ -157,13 +157,13 @@ void spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *a } } -sycl::event spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, const void *beta, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spmv_alg alg, - oneapi::mkl::sparse::spmv_descr_t spmv_descr, void * /*workspace*/, +sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spmv_alg alg, + oneapi::math::sparse::spmv_descr_t spmv_descr, void * /*workspace*/, const std::vector &dependencies) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (internal_A_handle->all_use_buffer()) { @@ -171,7 +171,7 @@ sycl::event spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const } common_spmv_optimize(queue, opA, alpha, A_view, A_handle, x_handle, beta, y_handle, alg, spmv_descr); - if (alg == oneapi::mkl::sparse::spmv_alg::no_optimize_alg) { + if (alg == oneapi::math::sparse::spmv_alg::no_optimize_alg) { return detail::collapse_dependencies(queue, dependencies); } internal_A_handle->can_be_reset = false; @@ -190,13 +190,13 @@ sycl::event spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const } template -sycl::event internal_spmv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, const void *beta, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spmv_alg /*alg*/, - oneapi::mkl::sparse::spmv_descr_t /*spmv_descr*/, +sycl::event internal_spmv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spmv_alg /*alg*/, + oneapi::math::sparse::spmv_descr_t /*spmv_descr*/, const std::vector &dependencies, bool is_alpha_host_accessible, bool is_beta_host_accessible) { T host_alpha = @@ -245,12 +245,12 @@ sycl::event internal_spmv(sycl::queue &queue, oneapi::mkl::transpose opA, const } } -sycl::event spmv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, const void *beta, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spmv_alg alg, oneapi::mkl::sparse::spmv_descr_t spmv_descr, +sycl::event spmv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::spmv_descr_t spmv_descr, const std::vector &dependencies) { bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); @@ -258,7 +258,7 @@ sycl::event spmv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alp is_beta_host_accessible); if (!spmv_descr->optimized_called) { - throw mkl::uninitialized("sparse_blas", __func__, + throw math::uninitialized("sparse_blas", __func__, "spmv_optimize must be called before spmv."); } CHECK_DESCR_MATCH(spmv_descr, opA, "spmv_optimize"); @@ -274,4 +274,4 @@ sycl::event spmv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alp is_alpha_host_accessible, is_beta_host_accessible); } -} // namespace oneapi::mkl::sparse::BACKEND +} // namespace oneapi::math::sparse::BACKEND diff --git a/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx b/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx index 6a7f50eb1..554281080 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx @@ -17,46 +17,46 @@ * **************************************************************************/ -namespace oneapi::mkl::sparse { +namespace oneapi::math::sparse { struct spsv_descr { bool buffer_size_called = false; bool optimized_called = false; - oneapi::mkl::transpose last_optimized_opA; - oneapi::mkl::sparse::matrix_view last_optimized_A_view; - oneapi::mkl::sparse::matrix_handle_t last_optimized_A_handle; - oneapi::mkl::sparse::dense_vector_handle_t last_optimized_x_handle; - oneapi::mkl::sparse::dense_vector_handle_t last_optimized_y_handle; - oneapi::mkl::sparse::spsv_alg last_optimized_alg; + oneapi::math::transpose last_optimized_opA; + oneapi::math::sparse::matrix_view last_optimized_A_view; + oneapi::math::sparse::matrix_handle_t last_optimized_A_handle; + oneapi::math::sparse::dense_vector_handle_t last_optimized_x_handle; + oneapi::math::sparse::dense_vector_handle_t last_optimized_y_handle; + oneapi::math::sparse::spsv_alg last_optimized_alg; }; -} // namespace oneapi::mkl::sparse +} // namespace oneapi::math::sparse -namespace oneapi::mkl::sparse::BACKEND { +namespace oneapi::math::sparse::BACKEND { -void init_spsv_descr(sycl::queue & /*queue*/, oneapi::mkl::sparse::spsv_descr_t *p_spsv_descr) { +void init_spsv_descr(sycl::queue & /*queue*/, oneapi::math::sparse::spsv_descr_t *p_spsv_descr) { *p_spsv_descr = new spsv_descr(); } -sycl::event release_spsv_descr(sycl::queue &queue, oneapi::mkl::sparse::spsv_descr_t spsv_descr, +sycl::event release_spsv_descr(sycl::queue &queue, oneapi::math::sparse::spsv_descr_t spsv_descr, const std::vector &dependencies) { return detail::submit_release(queue, spsv_descr, dependencies); } -void check_valid_spsv(const std::string &function_name, oneapi::mkl::transpose opA, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - bool is_alpha_host_accessible, oneapi::mkl::sparse::spsv_alg alg) { +void check_valid_spsv(const std::string &function_name, oneapi::math::transpose opA, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, + bool is_alpha_host_accessible, oneapi::math::sparse::spsv_alg alg) { THROW_IF_NULLPTR(function_name, A_handle); THROW_IF_NULLPTR(function_name, x_handle); THROW_IF_NULLPTR(function_name, y_handle); auto internal_A_handle = detail::get_internal_handle(A_handle); - if (alg == oneapi::mkl::sparse::spsv_alg::no_optimize_alg && - !internal_A_handle->has_matrix_property(oneapi::mkl::sparse::matrix_property::sorted)) { - throw mkl::unimplemented( + if (alg == oneapi::math::sparse::spsv_alg::no_optimize_alg && + !internal_A_handle->has_matrix_property(oneapi::math::sparse::matrix_property::sorted)) { + throw math::unimplemented( "sparse_blas", function_name, "The backend does not support `no_optimize_alg` unless A_handle has the property `matrix_property::sorted`."); } @@ -65,8 +65,8 @@ void check_valid_spsv(const std::string &function_name, oneapi::mkl::transpose o detail::data_type data_type = internal_A_handle->get_value_type(); if ((data_type == detail::data_type::complex_fp32 || data_type == detail::data_type::complex_fp64) && - opA == oneapi::mkl::transpose::conjtrans) { - throw mkl::unimplemented("sparse_blas", function_name, + opA == oneapi::math::transpose::conjtrans) { + throw math::unimplemented("sparse_blas", function_name, "The backend does not support spsv using conjtrans."); } #else @@ -75,7 +75,7 @@ void check_valid_spsv(const std::string &function_name, oneapi::mkl::transpose o detail::check_all_containers_compatible(function_name, internal_A_handle, x_handle, y_handle); if (A_view.type_view != matrix_descr::triangular) { - throw mkl::invalid_argument("sparse_blas", function_name, + throw math::invalid_argument("sparse_blas", function_name, "Matrix view's type must be `matrix_descr::triangular`."); } @@ -84,13 +84,13 @@ void check_valid_spsv(const std::string &function_name, oneapi::mkl::transpose o } } -void spsv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spsv_alg alg, - oneapi::mkl::sparse::spsv_descr_t spsv_descr, std::size_t &temp_buffer_size) { +void spsv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spsv_alg alg, + oneapi::math::sparse::spsv_descr_t spsv_descr, std::size_t &temp_buffer_size) { // TODO: Add support for external workspace once the close-source oneMath backend supports it. bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); check_valid_spsv(__func__, opA, A_view, A_handle, x_handle, y_handle, is_alpha_host_accessible, @@ -99,18 +99,18 @@ void spsv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void spsv_descr->buffer_size_called = true; } -inline void common_spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spsv_alg alg, - oneapi::mkl::sparse::spsv_descr_t spsv_descr) { +inline void common_spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spsv_alg alg, + oneapi::math::sparse::spsv_descr_t spsv_descr) { bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); check_valid_spsv("spsv_optimize", opA, A_view, A_handle, x_handle, y_handle, is_alpha_host_accessible, alg); if (!spsv_descr->buffer_size_called) { - throw mkl::uninitialized("sparse_blas", "spsv_optimize", + throw math::uninitialized("sparse_blas", "spsv_optimize", "spsv_buffer_size must be called before spsv_optimize."); } spsv_descr->optimized_called = true; @@ -122,19 +122,19 @@ inline void common_spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, spsv_descr->last_optimized_alg = alg; } -void spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spsv_alg alg, oneapi::mkl::sparse::spsv_descr_t spsv_descr, +void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spsv_alg alg, oneapi::math::sparse::spsv_descr_t spsv_descr, sycl::buffer /*workspace*/) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (!internal_A_handle->all_use_buffer()) { detail::throw_incompatible_container(__func__); } common_spsv_optimize(queue, opA, alpha, A_view, A_handle, x_handle, y_handle, alg, spsv_descr); - if (alg == oneapi::mkl::sparse::spsv_alg::no_optimize_alg) { + if (alg == oneapi::math::sparse::spsv_alg::no_optimize_alg) { return; } internal_A_handle->can_be_reset = false; @@ -142,20 +142,20 @@ void spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *a internal_A_handle->backend_handle); } -sycl::event spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spsv_alg alg, - oneapi::mkl::sparse::spsv_descr_t spsv_descr, void * /*workspace*/, +sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spsv_alg alg, + oneapi::math::sparse::spsv_descr_t spsv_descr, void * /*workspace*/, const std::vector &dependencies) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (internal_A_handle->all_use_buffer()) { detail::throw_incompatible_container(__func__); } common_spsv_optimize(queue, opA, alpha, A_view, A_handle, x_handle, y_handle, alg, spsv_descr); - if (alg == oneapi::mkl::sparse::spsv_alg::no_optimize_alg) { + if (alg == oneapi::math::sparse::spsv_alg::no_optimize_alg) { return detail::collapse_dependencies(queue, dependencies); } internal_A_handle->can_be_reset = false; @@ -164,13 +164,13 @@ sycl::event spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const } template -sycl::event internal_spsv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spsv_alg /*alg*/, - oneapi::mkl::sparse::spsv_descr_t /*spsv_descr*/, +sycl::event internal_spsv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spsv_alg /*alg*/, + oneapi::math::sparse::spsv_descr_t /*spsv_descr*/, const std::vector &dependencies, bool is_alpha_host_accessible) { T host_alpha = @@ -192,19 +192,19 @@ sycl::event internal_spsv(sycl::queue &queue, oneapi::mkl::transpose opA, const } } -sycl::event spsv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spsv_alg alg, oneapi::mkl::sparse::spsv_descr_t spsv_descr, +sycl::event spsv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spsv_alg alg, oneapi::math::sparse::spsv_descr_t spsv_descr, const std::vector &dependencies) { bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); check_valid_spsv(__func__, opA, A_view, A_handle, x_handle, y_handle, is_alpha_host_accessible, alg); if (!spsv_descr->optimized_called) { - throw mkl::uninitialized("sparse_blas", __func__, + throw math::uninitialized("sparse_blas", __func__, "spsv_optimize must be called before spsv."); } CHECK_DESCR_MATCH(spsv_descr, opA, "spsv_optimize"); @@ -220,4 +220,4 @@ sycl::event spsv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alp is_alpha_host_accessible); } -} // namespace oneapi::mkl::sparse::BACKEND +} // namespace oneapi::math::sparse::BACKEND diff --git a/src/sparse_blas/backends/mklcpu/mklcpu_handles.cpp b/src/sparse_blas/backends/mklcpu/mklcpu_handles.cpp index 526fd45a9..24715d844 100644 --- a/src/sparse_blas/backends/mklcpu/mklcpu_handles.cpp +++ b/src/sparse_blas/backends/mklcpu/mklcpu_handles.cpp @@ -22,8 +22,8 @@ #include "sparse_blas/backends/mkl_common/mkl_helper.hpp" #include "sparse_blas/backends/mkl_common/mkl_handles.hpp" -namespace oneapi::mkl::sparse::mklcpu { +namespace oneapi::math::sparse::mklcpu { #include "sparse_blas/backends/mkl_common/mkl_handles.cxx" -} // namespace oneapi::mkl::sparse::mklcpu +} // namespace oneapi::math::sparse::mklcpu diff --git a/src/sparse_blas/backends/mklgpu/mklgpu_handles.cpp b/src/sparse_blas/backends/mklgpu/mklgpu_handles.cpp index c5b33a7da..5177373dd 100644 --- a/src/sparse_blas/backends/mklgpu/mklgpu_handles.cpp +++ b/src/sparse_blas/backends/mklgpu/mklgpu_handles.cpp @@ -22,8 +22,8 @@ #include "sparse_blas/backends/mkl_common/mkl_handles.hpp" #include "sparse_blas/backends/mkl_common/mkl_helper.hpp" -namespace oneapi::mkl::sparse::mklgpu { +namespace oneapi::math::sparse::mklgpu { #include "sparse_blas/backends/mkl_common/mkl_handles.cxx" -} // namespace oneapi::mkl::sparse::mklgpu +} // namespace oneapi::math::sparse::mklgpu diff --git a/src/sparse_blas/enum_data_types.hpp b/src/sparse_blas/enum_data_types.hpp index 17f8b35b6..7c7a2c5e3 100644 --- a/src/sparse_blas/enum_data_types.hpp +++ b/src/sparse_blas/enum_data_types.hpp @@ -22,7 +22,7 @@ #include -namespace oneapi::mkl::sparse::detail { +namespace oneapi::math::sparse::detail { enum data_type { none, int32, int64, real_fp32, real_fp64, complex_fp32, complex_fp64 }; @@ -64,6 +64,6 @@ constexpr data_type get_data_type() { } } -} // namespace oneapi::mkl::sparse::detail +} // namespace oneapi::math::sparse::detail #endif // _ONEMATH_SRC_SPARSE_BLAS_ENUM_DATA_TYPES_HPP_ diff --git a/src/sparse_blas/function_table.hpp b/src/sparse_blas/function_table.hpp index 591f89f9d..3667ff868 100644 --- a/src/sparse_blas/function_table.hpp +++ b/src/sparse_blas/function_table.hpp @@ -26,77 +26,77 @@ // Dense vector #define DEFINE_DENSE_VECTOR_FUNCS(FP_TYPE, FP_SUFFIX) \ void (*init_dense_vector_buffer##FP_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_vector_handle_t * p_dvhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t * p_dvhandle, \ std::int64_t size, sycl::buffer val); \ void (*init_dense_vector_usm##FP_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_vector_handle_t * p_dvhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t * p_dvhandle, \ std::int64_t size, FP_TYPE * val); \ void (*set_dense_vector_data_buffer##FP_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_vector_handle_t dvhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t dvhandle, \ std::int64_t size, sycl::buffer val); \ void (*set_dense_vector_data_usm##FP_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_vector_handle_t dvhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t dvhandle, \ std::int64_t size, FP_TYPE * val) // Dense matrix #define DEFINE_DENSE_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX) \ void (*init_dense_matrix_buffer##FP_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_matrix_handle_t * p_dmhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t * p_dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::mkl::layout dense_layout, sycl::buffer val); \ + oneapi::math::layout dense_layout, sycl::buffer val); \ void (*init_dense_matrix_usm##FP_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_matrix_handle_t * p_dmhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t * p_dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::mkl::layout dense_layout, FP_TYPE * val); \ + oneapi::math::layout dense_layout, FP_TYPE * val); \ void (*set_dense_matrix_data_buffer##FP_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_matrix_handle_t dmhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::mkl::layout dense_layout, sycl::buffer val); \ + oneapi::math::layout dense_layout, sycl::buffer val); \ void (*set_dense_matrix_data_usm##FP_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::dense_matrix_handle_t dmhandle, \ + sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::mkl::layout dense_layout, FP_TYPE * val) + oneapi::math::layout dense_layout, FP_TYPE * val) // COO matrix #define DEFINE_COO_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ void (*init_coo_matrix_buffer##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t * p_smhandle, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::mkl::index_base index, sycl::buffer row_ind, \ + oneapi::math::index_base index, sycl::buffer row_ind, \ sycl::buffer col_ind, sycl::buffer val); \ void (*init_coo_matrix_usm##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t * p_smhandle, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::mkl::index_base index, INT_TYPE * row_ind, INT_TYPE * col_ind, FP_TYPE * val); \ + oneapi::math::index_base index, INT_TYPE * row_ind, INT_TYPE * col_ind, FP_TYPE * val); \ void (*set_coo_matrix_data_buffer##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::mkl::index_base index, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ + std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ sycl::buffer row_ind, sycl::buffer col_ind, \ sycl::buffer val); \ void (*set_coo_matrix_data_usm##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::mkl::index_base index, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ + std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ INT_TYPE * row_ind, INT_TYPE * col_ind, FP_TYPE * val) // CSR matrix #define DEFINE_CSR_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ void (*init_csr_matrix_buffer##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t * p_smhandle, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::mkl::index_base index, sycl::buffer row_ptr, \ + oneapi::math::index_base index, sycl::buffer row_ptr, \ sycl::buffer col_ind, sycl::buffer val); \ void (*init_csr_matrix_usm##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t * p_smhandle, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::mkl::index_base index, INT_TYPE * row_ptr, INT_TYPE * col_ind, FP_TYPE * val); \ + oneapi::math::index_base index, INT_TYPE * row_ptr, INT_TYPE * col_ind, FP_TYPE * val); \ void (*set_csr_matrix_data_buffer##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::mkl::index_base index, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ + std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ sycl::buffer row_ptr, sycl::buffer col_ind, \ sycl::buffer val); \ void (*set_csr_matrix_data_usm##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::mkl::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::mkl::index_base index, \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ + std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ INT_TYPE * row_ptr, INT_TYPE * col_ind, FP_TYPE * val) typedef struct { @@ -105,13 +105,13 @@ typedef struct { // Dense vector FOR_EACH_FP_TYPE(DEFINE_DENSE_VECTOR_FUNCS); sycl::event (*release_dense_vector)(sycl::queue &queue, - oneapi::mkl::sparse::dense_vector_handle_t dvhandle, + oneapi::math::sparse::dense_vector_handle_t dvhandle, const std::vector &dependencies); // Dense matrix FOR_EACH_FP_TYPE(DEFINE_DENSE_MATRIX_FUNCS); sycl::event (*release_dense_matrix)(sycl::queue &queue, - oneapi::mkl::sparse::dense_matrix_handle_t dmhandle, + oneapi::math::sparse::dense_matrix_handle_t dmhandle, const std::vector &dependencies); // COO matrix @@ -122,140 +122,140 @@ typedef struct { // Common sparse matrix functions sycl::event (*release_sparse_matrix)(sycl::queue &queue, - oneapi::mkl::sparse::matrix_handle_t smhandle, + oneapi::math::sparse::matrix_handle_t smhandle, const std::vector &dependencies); - bool (*set_matrix_property)(sycl::queue &queue, oneapi::mkl::sparse::matrix_handle_t smhandle, - oneapi::mkl::sparse::matrix_property property); + bool (*set_matrix_property)(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, + oneapi::math::sparse::matrix_property property); // SPMM - void (*init_spmm_descr)(sycl::queue &queue, oneapi::mkl::sparse::spmm_descr_t *p_spmm_descr); + void (*init_spmm_descr)(sycl::queue &queue, oneapi::math::sparse::spmm_descr_t *p_spmm_descr); sycl::event (*release_spmm_descr)(sycl::queue &queue, - oneapi::mkl::sparse::spmm_descr_t spmm_descr, + oneapi::math::sparse::spmm_descr_t spmm_descr, const std::vector &dependencies); - void (*spmm_buffer_size)(sycl::queue &queue, oneapi::mkl::transpose opA, - oneapi::mkl::transpose opB, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_matrix_handle_t B_handle, const void *beta, - oneapi::mkl::sparse::dense_matrix_handle_t C_handle, - oneapi::mkl::sparse::spmm_alg alg, - oneapi::mkl::sparse::spmm_descr_t spmm_descr, + void (*spmm_buffer_size)(sycl::queue &queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t C_handle, + oneapi::math::sparse::spmm_alg alg, + oneapi::math::sparse::spmm_descr_t spmm_descr, std::size_t &temp_buffer_size); void (*spmm_optimize_buffer)( - sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, - const void *alpha, oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_matrix_handle_t B_handle, const void *beta, - oneapi::mkl::sparse::dense_matrix_handle_t C_handle, oneapi::mkl::sparse::spmm_alg alg, - oneapi::mkl::sparse::spmm_descr_t spmm_descr, sycl::buffer workspace); - - sycl::event (*spmm_optimize_usm)(sycl::queue &queue, oneapi::mkl::transpose opA, - oneapi::mkl::transpose opB, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_matrix_handle_t B_handle, + sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void *alpha, oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg alg, + oneapi::math::sparse::spmm_descr_t spmm_descr, sycl::buffer workspace); + + sycl::event (*spmm_optimize_usm)(sycl::queue &queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, - oneapi::mkl::sparse::dense_matrix_handle_t C_handle, - oneapi::mkl::sparse::spmm_alg alg, - oneapi::mkl::sparse::spmm_descr_t spmm_descr, void *workspace, + oneapi::math::sparse::dense_matrix_handle_t C_handle, + oneapi::math::sparse::spmm_alg alg, + oneapi::math::sparse::spmm_descr_t spmm_descr, void *workspace, const std::vector &dependencies); - sycl::event (*spmm)(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, - const void *alpha, oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_matrix_handle_t B_handle, const void *beta, - oneapi::mkl::sparse::dense_matrix_handle_t C_handle, - oneapi::mkl::sparse::spmm_alg alg, - oneapi::mkl::sparse::spmm_descr_t spmm_descr, + sycl::event (*spmm)(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void *alpha, oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t C_handle, + oneapi::math::sparse::spmm_alg alg, + oneapi::math::sparse::spmm_descr_t spmm_descr, const std::vector &dependencies); // SPMV - void (*init_spmv_descr)(sycl::queue &queue, oneapi::mkl::sparse::spmv_descr_t *p_spmv_descr); + void (*init_spmv_descr)(sycl::queue &queue, oneapi::math::sparse::spmv_descr_t *p_spmv_descr); sycl::event (*release_spmv_descr)(sycl::queue &queue, - oneapi::mkl::sparse::spmv_descr_t spmv_descr, + oneapi::math::sparse::spmv_descr_t spmv_descr, const std::vector &dependencies); - void (*spmv_buffer_size)(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, const void *beta, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spmv_alg alg, - oneapi::mkl::sparse::spmv_descr_t spmv_descr, + void (*spmv_buffer_size)(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spmv_alg alg, + oneapi::math::sparse::spmv_descr_t spmv_descr, std::size_t &temp_buffer_size); void (*spmv_optimize_buffer)( - sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, const void *beta, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, oneapi::mkl::sparse::spmv_alg alg, - oneapi::mkl::sparse::spmv_descr_t spmv_descr, sycl::buffer workspace); - - sycl::event (*spmv_optimize_usm)(sycl::queue &queue, oneapi::mkl::transpose opA, - const void *alpha, oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, + sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spmv_alg alg, + oneapi::math::sparse::spmv_descr_t spmv_descr, sycl::buffer workspace); + + sycl::event (*spmv_optimize_usm)(sycl::queue &queue, oneapi::math::transpose opA, + const void *alpha, oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spmv_alg alg, - oneapi::mkl::sparse::spmv_descr_t spmv_descr, void *workspace, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spmv_alg alg, + oneapi::math::sparse::spmv_descr_t spmv_descr, void *workspace, const std::vector &dependencies); - sycl::event (*spmv)(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, const void *beta, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spmv_alg alg, - oneapi::mkl::sparse::spmv_descr_t spmv_descr, + sycl::event (*spmv)(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spmv_alg alg, + oneapi::math::sparse::spmv_descr_t spmv_descr, const std::vector &dependencies); // SPSV - void (*init_spsv_descr)(sycl::queue &queue, oneapi::mkl::sparse::spsv_descr_t *p_spsv_descr); + void (*init_spsv_descr)(sycl::queue &queue, oneapi::math::sparse::spsv_descr_t *p_spsv_descr); sycl::event (*release_spsv_descr)(sycl::queue &queue, - oneapi::mkl::sparse::spsv_descr_t spsv_descr, + oneapi::math::sparse::spsv_descr_t spsv_descr, const std::vector &dependencies); - void (*spsv_buffer_size)(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spsv_alg alg, - oneapi::mkl::sparse::spsv_descr_t spsv_descr, + void (*spsv_buffer_size)(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spsv_alg alg, + oneapi::math::sparse::spsv_descr_t spsv_descr, std::size_t &temp_buffer_size); - void (*spsv_optimize_buffer)(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spsv_alg alg, - oneapi::mkl::sparse::spsv_descr_t spsv_descr, + void (*spsv_optimize_buffer)(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spsv_alg alg, + oneapi::math::sparse::spsv_descr_t spsv_descr, sycl::buffer workspace); - sycl::event (*spsv_optimize_usm)(sycl::queue &queue, oneapi::mkl::transpose opA, - const void *alpha, oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spsv_alg alg, - oneapi::mkl::sparse::spsv_descr_t spsv_descr, void *workspace, + sycl::event (*spsv_optimize_usm)(sycl::queue &queue, oneapi::math::transpose opA, + const void *alpha, oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spsv_alg alg, + oneapi::math::sparse::spsv_descr_t spsv_descr, void *workspace, const std::vector &dependencies); - sycl::event (*spsv)(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, - oneapi::mkl::sparse::matrix_view A_view, - oneapi::mkl::sparse::matrix_handle_t A_handle, - oneapi::mkl::sparse::dense_vector_handle_t x_handle, - oneapi::mkl::sparse::dense_vector_handle_t y_handle, - oneapi::mkl::sparse::spsv_alg alg, - oneapi::mkl::sparse::spsv_descr_t spsv_descr, + sycl::event (*spsv)(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + oneapi::math::sparse::matrix_view A_view, + oneapi::math::sparse::matrix_handle_t A_handle, + oneapi::math::sparse::dense_vector_handle_t x_handle, + oneapi::math::sparse::dense_vector_handle_t y_handle, + oneapi::math::sparse::spsv_alg alg, + oneapi::math::sparse::spsv_descr_t spsv_descr, const std::vector &dependencies); } sparse_blas_function_table_t; diff --git a/src/sparse_blas/generic_container.hpp b/src/sparse_blas/generic_container.hpp index 0b0f87c22..e2377b951 100644 --- a/src/sparse_blas/generic_container.hpp +++ b/src/sparse_blas/generic_container.hpp @@ -32,7 +32,7 @@ #include "oneapi/math/sparse_blas/types.hpp" #include "enum_data_types.hpp" -namespace oneapi::mkl::sparse::detail { +namespace oneapi::math::sparse::detail { /// Represent a non-templated container for USM or buffer. struct generic_container { @@ -156,7 +156,7 @@ struct generic_dense_vector_handle : public detail::generic_dense_handle(backend_handle, value_buffer), size(size) { if (value_buffer.size() < static_cast(size)) { - throw oneapi::mkl::invalid_argument( + throw oneapi::math::invalid_argument( "sparse_blas", "init_dense_vector", "Buffer size too small, expected at least " + std::to_string(size) + " but got " + std::to_string(value_buffer.size()) + " elements."); @@ -170,7 +170,7 @@ struct generic_dense_matrix_handle : public detail::generic_dense_handle generic_dense_matrix_handle(BackendHandleT backend_handle, T* value_ptr, std::int64_t num_rows, @@ -191,9 +191,9 @@ struct generic_dense_matrix_handle : public detail::generic_dense_handle( - (dense_layout == oneapi::mkl::layout::row_major ? num_rows : num_cols) * ld); + (dense_layout == oneapi::math::layout::row_major ? num_rows : num_cols) * ld); if (value_buffer.size() < minimum_size) { - throw oneapi::mkl::invalid_argument( + throw oneapi::math::invalid_argument( "sparse_blas", "init_dense_matrix", "Buffer size too small, expected at least " + std::to_string(minimum_size) + " but got " + std::to_string(value_buffer.size()) + " elements."); @@ -248,21 +248,21 @@ struct generic_sparse_handle { return row_container.data_type; } - void set_matrix_property(oneapi::mkl::sparse::matrix_property property) { + void set_matrix_property(oneapi::math::sparse::matrix_property property) { properties_mask |= matrix_property_to_mask(property); } - bool has_matrix_property(oneapi::mkl::sparse::matrix_property property) { + bool has_matrix_property(oneapi::math::sparse::matrix_property property) { return properties_mask & matrix_property_to_mask(property); } private: - std::int32_t matrix_property_to_mask(oneapi::mkl::sparse::matrix_property property) { + std::int32_t matrix_property_to_mask(oneapi::math::sparse::matrix_property property) { switch (property) { - case oneapi::mkl::sparse::matrix_property::symmetric: return 1 << 0; - case oneapi::mkl::sparse::matrix_property::sorted: return 1 << 1; + case oneapi::math::sparse::matrix_property::symmetric: return 1 << 0; + case oneapi::math::sparse::matrix_property::sorted: return 1 << 1; default: - throw oneapi::mkl::invalid_argument( + throw oneapi::math::invalid_argument( "sparse_blas", "set_matrix_property", "Unsupported matrix property " + std::to_string(static_cast(property))); } @@ -270,7 +270,7 @@ struct generic_sparse_handle { }; inline void throw_incompatible_container(const std::string& function_name) { - throw oneapi::mkl::invalid_argument( + throw oneapi::math::invalid_argument( "sparse_blas", function_name, "Incompatible container types. All inputs and outputs must use the same container: buffer or USM"); } @@ -306,14 +306,14 @@ void check_all_containers_compatible(const std::string& function_name, for (const auto internal_container : { internal_containers... }) { const data_type other_value_type = internal_container->get_value_type(); if (other_value_type != first_value_type) { - throw oneapi::mkl::invalid_argument( + throw oneapi::math::invalid_argument( "sparse_blas", function_name, "Incompatible data types expected " + data_type_to_str(first_value_type) + " but got " + data_type_to_str(other_value_type)); } const data_type other_int_type = internal_container->get_int_type(); if (other_int_type != data_type::none && other_int_type != first_int_type) { - throw oneapi::mkl::invalid_argument("sparse_blas", function_name, + throw oneapi::math::invalid_argument("sparse_blas", function_name, "Incompatible integer types expected " + data_type_to_str(first_int_type) + " but got " + data_type_to_str(other_int_type)); @@ -329,6 +329,6 @@ sycl::event submit_release(sycl::queue& queue, T* ptr, const DependenciesT& depe }); } -} // namespace oneapi::mkl::sparse::detail +} // namespace oneapi::math::sparse::detail #endif // _ONEMATH_SRC_SPARSE_BLAS_GENERIC_CONTAINER_HPP_ diff --git a/src/sparse_blas/macros.hpp b/src/sparse_blas/macros.hpp index 9aa08503f..34db8efc5 100644 --- a/src/sparse_blas/macros.hpp +++ b/src/sparse_blas/macros.hpp @@ -38,7 +38,7 @@ #define THROW_IF_NULLPTR(FUNC_NAME, PTR) \ if (!(PTR)) { \ - throw mkl::uninitialized("sparse_blas", FUNC_NAME, \ + throw math::uninitialized("sparse_blas", FUNC_NAME, \ std::string(#PTR) + " must not be nullptr."); \ } diff --git a/src/sparse_blas/matrix_view_comparison.hpp b/src/sparse_blas/matrix_view_comparison.hpp index 69ea94b7e..08c2f8c54 100644 --- a/src/sparse_blas/matrix_view_comparison.hpp +++ b/src/sparse_blas/matrix_view_comparison.hpp @@ -22,14 +22,14 @@ #include "oneapi/math/sparse_blas/matrix_view.hpp" -inline bool operator==(const oneapi::mkl::sparse::matrix_view& lhs, - const oneapi::mkl::sparse::matrix_view& rhs) { +inline bool operator==(const oneapi::math::sparse::matrix_view& lhs, + const oneapi::math::sparse::matrix_view& rhs) { return lhs.type_view == rhs.type_view && lhs.uplo_view == rhs.uplo_view && lhs.diag_view == rhs.diag_view; } -inline bool operator!=(const oneapi::mkl::sparse::matrix_view& lhs, - const oneapi::mkl::sparse::matrix_view& rhs) { +inline bool operator!=(const oneapi::math::sparse::matrix_view& lhs, + const oneapi::math::sparse::matrix_view& rhs) { return !(lhs == rhs); } diff --git a/src/sparse_blas/sparse_blas_loader.cpp b/src/sparse_blas/sparse_blas_loader.cpp index 2cf6fba74..bc17b490d 100644 --- a/src/sparse_blas/sparse_blas_loader.cpp +++ b/src/sparse_blas/sparse_blas_loader.cpp @@ -24,9 +24,9 @@ #include "sparse_blas/macros.hpp" #include "oneapi/math/detail/get_device_id.hpp" -namespace oneapi::mkl::sparse { +namespace oneapi::math::sparse { -static oneapi::mkl::detail::table_initializer function_tables; @@ -214,7 +214,7 @@ sycl::event release_spmm_descr(sycl::queue &queue, spmm_descr_t spmm_descr, return function_tables[{ libkey, queue }].release_spmm_descr(queue, spmm_descr, dependencies); } -void spmm_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, +void spmm_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, @@ -225,7 +225,7 @@ void spmm_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mk temp_buffer_size); } -void spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, +void spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, sycl::buffer workspace) { @@ -235,8 +235,8 @@ void spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl:: spmm_descr, workspace); } -sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, - oneapi::mkl::transpose opB, const void *alpha, matrix_view A_view, +sycl::event spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, void *workspace, @@ -247,7 +247,7 @@ sycl::event spmm_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, workspace, dependencies); } -sycl::event spmm(sycl::queue &queue, oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, +sycl::event spmm(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, @@ -270,7 +270,7 @@ sycl::event release_spmv_descr(sycl::queue &queue, spmv_descr_t spmv_descr, return function_tables[{ libkey, queue }].release_spmv_descr(queue, spmv_descr, dependencies); } -void spmv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +void spmv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, std::size_t &temp_buffer_size) { @@ -280,7 +280,7 @@ void spmv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void temp_buffer_size); } -void spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, sycl::buffer workspace) { @@ -289,7 +289,7 @@ void spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *a queue, opA, alpha, A_view, A_handle, x_handle, beta, y_handle, alg, spmv_descr, workspace); } -sycl::event spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, @@ -300,7 +300,7 @@ sycl::event spmv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const dependencies); } -sycl::event spmv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +sycl::event spmv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, const std::vector &dependencies) { @@ -321,7 +321,7 @@ sycl::event release_spsv_descr(sycl::queue &queue, spsv_descr_t spsv_descr, return function_tables[{ libkey, queue }].release_spsv_descr(queue, spsv_descr, dependencies); } -void spsv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +void spsv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, std::size_t &temp_buffer_size) { @@ -330,7 +330,7 @@ void spsv_buffer_size(sycl::queue &queue, oneapi::mkl::transpose opA, const void queue, opA, alpha, A_view, A_handle, x_handle, y_handle, alg, spsv_descr, temp_buffer_size); } -void spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, sycl::buffer workspace) { @@ -339,7 +339,7 @@ void spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *a queue, opA, alpha, A_view, A_handle, x_handle, y_handle, alg, spsv_descr, workspace); } -sycl::event spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, void *workspace, @@ -350,7 +350,7 @@ sycl::event spsv_optimize(sycl::queue &queue, oneapi::mkl::transpose opA, const workspace, dependencies); } -sycl::event spsv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alpha, +sycl::event spsv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, const std::vector &dependencies) { @@ -359,4 +359,4 @@ sycl::event spsv(sycl::queue &queue, oneapi::mkl::transpose opA, const void *alp y_handle, alg, spsv_descr, dependencies); } -} // namespace oneapi::mkl::sparse +} // namespace oneapi::math::sparse diff --git a/tests/unit_tests/blas/batch/axpy_batch_stride.cpp b/tests/unit_tests/blas/batch/axpy_batch_stride.cpp index b49305662..c83dedd5e 100644 --- a/tests/unit_tests/blas/batch/axpy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/axpy_batch_stride.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, fp alpha, +int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, fp alpha, int64_t batch_size) { // Prepare data. int64_t n, i; @@ -105,13 +105,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, fp try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::axpy_batch(main_queue, n, alpha, x_buffer, incx, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::axpy_batch(main_queue, n, alpha, x_buffer, incx, stride_x, y_buffer, incy, stride_y, batch_size); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::axpy_batch(main_queue, n, alpha, x_buffer, incx, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::axpy_batch(main_queue, n, alpha, x_buffer, incx, stride_x, y_buffer, incy, stride_y, batch_size); break; @@ -119,13 +119,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, fp } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::axpy_batch, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpy_batch, n, alpha, x_buffer, incx, stride_x, y_buffer, incy, stride_y, batch_size); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::axpy_batch, n, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::axpy_batch, n, alpha, x_buffer, incx, stride_x, y_buffer, incy, stride_y, batch_size); break; @@ -139,7 +139,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, fp print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -160,7 +160,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, fp } class AxpyBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpyBatchStrideTests, RealSinglePrecision) { float alpha = 2.0; @@ -208,8 +208,8 @@ TEST_P(AxpyBatchStrideTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(AxpyBatchStrideTestSuite, AxpyBatchStrideTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp index 7fbde684d..5951bc0f5 100644 --- a/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, fp alpha, +int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, fp alpha, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -110,13 +110,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, fp try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::axpy_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::axpy_batch( main_queue, n, alpha, &x[0], incx, stride_x, &y[0], incy, stride_y, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::axpy_batch(main_queue, n, alpha, &x[0], incx, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::axpy_batch(main_queue, n, alpha, &x[0], incx, stride_x, &y[0], incy, stride_y, batch_size, dependencies); break; @@ -125,13 +125,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, fp done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::axpy_batch, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpy_batch, n, alpha, &x[0], incx, stride_x, &y[0], incy, stride_y, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::axpy_batch, n, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::axpy_batch, n, alpha, &x[0], incx, stride_x, &y[0], incy, stride_y, batch_size, dependencies); break; @@ -146,7 +146,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, fp print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -166,7 +166,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, fp } class AxpyBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpyBatchStrideUsmTests, RealSinglePrecision) { float alpha = 2.0; @@ -214,8 +214,8 @@ TEST_P(AxpyBatchStrideUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(AxpyBatchStrideUsmTestSuite, AxpyBatchStrideUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/axpy_batch_usm.cpp b/tests/unit_tests/blas/batch/axpy_batch_usm.cpp index c511789fa..bbb62a19c 100644 --- a/tests/unit_tests/blas/batch/axpy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/axpy_batch_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { +int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -70,23 +70,23 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { // Prepare data. int64_t *n = - (int64_t *)oneapi::mkl::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); int64_t *incx = - (int64_t *)oneapi::mkl::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); int64_t *incy = - (int64_t *)oneapi::mkl::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); - fp *alpha = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * group_count, *dev, cxt); + (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + fp *alpha = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * group_count, *dev, cxt); int64_t *group_size = - (int64_t *)oneapi::mkl::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); if ((n == NULL) || (incx == NULL) || (incy == NULL) || (alpha == NULL) || (group_size == NULL)) { std::cout << "Error cannot allocate input arrays\n"; - oneapi::mkl::free_shared(n, cxt); - oneapi::mkl::free_shared(incx, cxt); - oneapi::mkl::free_shared(incy, cxt); - oneapi::mkl::free_shared(alpha, cxt); - oneapi::mkl::free_shared(group_size, cxt); + oneapi::math::free_shared(n, cxt); + oneapi::math::free_shared(incx, cxt); + oneapi::math::free_shared(incy, cxt); + oneapi::math::free_shared(alpha, cxt); + oneapi::math::free_shared(group_size, cxt); return false; } @@ -105,17 +105,17 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { } fp **x_array = - (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); + (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); fp **y_array = - (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); + (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); fp **y_ref_array = - (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); + (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); if ((x_array == NULL) || (y_array == NULL) || (y_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; - oneapi::mkl::free_shared(x_array, cxt); - oneapi::mkl::free_shared(y_array, cxt); - oneapi::mkl::free_shared(y_ref_array, cxt); + oneapi::math::free_shared(x_array, cxt); + oneapi::math::free_shared(y_array, cxt); + oneapi::math::free_shared(y_ref_array, cxt); return false; } idx = 0; @@ -124,11 +124,11 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { total_size_x = (1 + (n[i] - 1) * std::abs(incx[i])); total_size_y = (1 + (n[i] - 1) * std::abs(incy[i])); x_array[idx] = - (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * total_size_x, *dev, cxt); + (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_x, *dev, cxt); y_array[idx] = - (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); + (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); y_ref_array[idx] = - (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); + (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); rand_vector(x_array[idx], n[i], incx[i]); rand_vector(y_array[idx], n[i], incy[i]); copy_vector(y_array[idx], n[i], incy[i], y_ref_array[idx]); @@ -157,13 +157,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::axpy_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::axpy_batch( main_queue, n, alpha, (const fp **)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::axpy_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::axpy_batch( main_queue, n, alpha, (const fp **)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; @@ -172,13 +172,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::axpy_batch, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpy_batch, n, alpha, (const fp **)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::axpy_batch, n, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::axpy_batch, n, alpha, (const fp **)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; @@ -193,24 +193,24 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(x_array[idx], cxt); - oneapi::mkl::free_shared(y_array[idx], cxt); - oneapi::mkl::free_shared(y_ref_array[idx], cxt); + oneapi::math::free_shared(x_array[idx], cxt); + oneapi::math::free_shared(y_array[idx], cxt); + oneapi::math::free_shared(y_ref_array[idx], cxt); idx++; } } - oneapi::mkl::free_shared(n, cxt); - oneapi::mkl::free_shared(incx, cxt); - oneapi::mkl::free_shared(incy, cxt); - oneapi::mkl::free_shared(alpha, cxt); - oneapi::mkl::free_shared(group_size, cxt); - oneapi::mkl::free_shared(x_array, cxt); - oneapi::mkl::free_shared(y_array, cxt); - oneapi::mkl::free_shared(y_ref_array, cxt); + oneapi::math::free_shared(n, cxt); + oneapi::math::free_shared(incx, cxt); + oneapi::math::free_shared(incy, cxt); + oneapi::math::free_shared(alpha, cxt); + oneapi::math::free_shared(group_size, cxt); + oneapi::math::free_shared(x_array, cxt); + oneapi::math::free_shared(y_array, cxt); + oneapi::math::free_shared(y_ref_array, cxt); return test_skipped; } @@ -233,26 +233,26 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(x_array[idx], cxt); - oneapi::mkl::free_shared(y_array[idx], cxt); - oneapi::mkl::free_shared(y_ref_array[idx], cxt); + oneapi::math::free_shared(x_array[idx], cxt); + oneapi::math::free_shared(y_array[idx], cxt); + oneapi::math::free_shared(y_ref_array[idx], cxt); idx++; } } - oneapi::mkl::free_shared(n, cxt); - oneapi::mkl::free_shared(incx, cxt); - oneapi::mkl::free_shared(incy, cxt); - oneapi::mkl::free_shared(alpha, cxt); - oneapi::mkl::free_shared(group_size, cxt); - oneapi::mkl::free_shared(x_array, cxt); - oneapi::mkl::free_shared(y_array, cxt); - oneapi::mkl::free_shared(y_ref_array, cxt); + oneapi::math::free_shared(n, cxt); + oneapi::math::free_shared(incx, cxt); + oneapi::math::free_shared(incy, cxt); + oneapi::math::free_shared(alpha, cxt); + oneapi::math::free_shared(group_size, cxt); + oneapi::math::free_shared(x_array, cxt); + oneapi::math::free_shared(y_array, cxt); + oneapi::math::free_shared(y_ref_array, cxt); return (int)good; } class AxpyBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpyBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -278,8 +278,8 @@ TEST_P(AxpyBatchUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(AxpyBatchUsmTestSuite, AxpyBatchUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/copy_batch_stride.cpp b/tests/unit_tests/blas/batch/copy_batch_stride.cpp index a78663d8a..a0d933d2d 100644 --- a/tests/unit_tests/blas/batch/copy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/copy_batch_stride.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { // Prepare data. int64_t n, i; @@ -104,25 +104,25 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::copy_batch(main_queue, n, x_buffer, incx, stride_x, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::copy_batch(main_queue, n, x_buffer, incx, stride_x, y_buffer, incy, stride_y, batch_size); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::copy_batch(main_queue, n, x_buffer, incx, stride_x, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::copy_batch(main_queue, n, x_buffer, incx, stride_x, y_buffer, incy, stride_y, batch_size); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::copy_batch, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::copy_batch, n, x_buffer, incx, stride_x, y_buffer, incy, stride_y, batch_size); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::copy_batch, n, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::copy_batch, n, x_buffer, incx, stride_x, y_buffer, incy, stride_y, batch_size); break; @@ -136,7 +136,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -157,7 +157,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in } class CopyBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(CopyBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 2, 3, 15)); @@ -195,8 +195,8 @@ TEST_P(CopyBatchStrideTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(CopyBatchStrideTestSuite, CopyBatchStrideTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp index dd1a7a9f3..e1502799f 100644 --- a/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -109,13 +109,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::copy_batch(main_queue, n, &x[0], incx, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::copy_batch(main_queue, n, &x[0], incx, stride_x, &y[0], incy, stride_y, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::copy_batch(main_queue, n, &x[0], incx, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::copy_batch(main_queue, n, &x[0], incx, stride_x, &y[0], incy, stride_y, batch_size, dependencies); break; @@ -124,13 +124,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::copy_batch, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::copy_batch, n, &x[0], incx, stride_x, &y[0], incy, stride_y, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::copy_batch, n, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::copy_batch, n, &x[0], incx, stride_x, &y[0], incy, stride_y, batch_size, dependencies); break; @@ -145,7 +145,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -165,7 +165,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in } class CopyBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(CopyBatchStrideUsmTests, RealSinglePrecision) { float alpha = 2.0; @@ -207,8 +207,8 @@ TEST_P(CopyBatchStrideUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(CopyBatchStrideUsmTestSuite, CopyBatchStrideUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/copy_batch_usm.cpp b/tests/unit_tests/blas/batch/copy_batch_usm.cpp index ef1a30d27..ad0b5003d 100644 --- a/tests/unit_tests/blas/batch/copy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/copy_batch_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { +int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -70,20 +70,20 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { // Prepare data. int64_t *n = - (int64_t *)oneapi::mkl::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); int64_t *incx = - (int64_t *)oneapi::mkl::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); int64_t *incy = - (int64_t *)oneapi::mkl::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); int64_t *group_size = - (int64_t *)oneapi::mkl::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); if ((n == NULL) || (incx == NULL) || (incy == NULL) || (group_size == NULL)) { std::cout << "Error cannot allocate input arrays\n"; - oneapi::mkl::free_shared(n, cxt); - oneapi::mkl::free_shared(incx, cxt); - oneapi::mkl::free_shared(incy, cxt); - oneapi::mkl::free_shared(group_size, cxt); + oneapi::math::free_shared(n, cxt); + oneapi::math::free_shared(incx, cxt); + oneapi::math::free_shared(incy, cxt); + oneapi::math::free_shared(group_size, cxt); return false; } @@ -101,17 +101,17 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { } fp **x_array = - (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); + (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); fp **y_array = - (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); + (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); fp **y_ref_array = - (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); + (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); if ((x_array == NULL) || (y_array == NULL) || (y_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; - oneapi::mkl::free_shared(x_array, cxt); - oneapi::mkl::free_shared(y_array, cxt); - oneapi::mkl::free_shared(y_ref_array, cxt); + oneapi::math::free_shared(x_array, cxt); + oneapi::math::free_shared(y_array, cxt); + oneapi::math::free_shared(y_ref_array, cxt); return false; } idx = 0; @@ -120,11 +120,11 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { total_size_x = (1 + (n[i] - 1) * std::abs(incx[i])); total_size_y = (1 + (n[i] - 1) * std::abs(incy[i])); x_array[idx] = - (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * total_size_x, *dev, cxt); + (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_x, *dev, cxt); y_array[idx] = - (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); + (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); y_ref_array[idx] = - (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); + (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); rand_vector(x_array[idx], n[i], incx[i]); rand_vector(y_array[idx], n[i], incy[i]); copy_vector(y_array[idx], n[i], incy[i], y_ref_array[idx]); @@ -153,13 +153,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::copy_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::copy_batch( main_queue, n, (const fp **)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::copy_batch(main_queue, n, (const fp **)x_array, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::copy_batch(main_queue, n, (const fp **)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; @@ -168,13 +168,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::copy_batch, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::copy_batch, n, (const fp **)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::copy_batch, n, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::copy_batch, n, (const fp **)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; @@ -189,23 +189,23 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(x_array[idx], cxt); - oneapi::mkl::free_shared(y_array[idx], cxt); - oneapi::mkl::free_shared(y_ref_array[idx], cxt); + oneapi::math::free_shared(x_array[idx], cxt); + oneapi::math::free_shared(y_array[idx], cxt); + oneapi::math::free_shared(y_ref_array[idx], cxt); idx++; } } - oneapi::mkl::free_shared(n, cxt); - oneapi::mkl::free_shared(incx, cxt); - oneapi::mkl::free_shared(incy, cxt); - oneapi::mkl::free_shared(group_size, cxt); - oneapi::mkl::free_shared(x_array, cxt); - oneapi::mkl::free_shared(y_array, cxt); - oneapi::mkl::free_shared(y_ref_array, cxt); + oneapi::math::free_shared(n, cxt); + oneapi::math::free_shared(incx, cxt); + oneapi::math::free_shared(incy, cxt); + oneapi::math::free_shared(group_size, cxt); + oneapi::math::free_shared(x_array, cxt); + oneapi::math::free_shared(y_array, cxt); + oneapi::math::free_shared(y_ref_array, cxt); return test_skipped; } @@ -228,25 +228,25 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(x_array[idx], cxt); - oneapi::mkl::free_shared(y_array[idx], cxt); - oneapi::mkl::free_shared(y_ref_array[idx], cxt); + oneapi::math::free_shared(x_array[idx], cxt); + oneapi::math::free_shared(y_array[idx], cxt); + oneapi::math::free_shared(y_ref_array[idx], cxt); idx++; } } - oneapi::mkl::free_shared(n, cxt); - oneapi::mkl::free_shared(incx, cxt); - oneapi::mkl::free_shared(incy, cxt); - oneapi::mkl::free_shared(group_size, cxt); - oneapi::mkl::free_shared(x_array, cxt); - oneapi::mkl::free_shared(y_array, cxt); - oneapi::mkl::free_shared(y_ref_array, cxt); + oneapi::math::free_shared(n, cxt); + oneapi::math::free_shared(incx, cxt); + oneapi::math::free_shared(incy, cxt); + oneapi::math::free_shared(group_size, cxt); + oneapi::math::free_shared(x_array, cxt); + oneapi::math::free_shared(y_array, cxt); + oneapi::math::free_shared(y_ref_array, cxt); return (int)good; } class CopyBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(CopyBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -272,8 +272,8 @@ TEST_P(CopyBatchUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(CopyBatchUsmTestSuite, CopyBatchUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp b/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp index 16d335fcb..658c57aa6 100644 --- a/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, int64_t incx, +int test(device *dev, oneapi::math::layout layout, oneapi::math::side left_right, int64_t incx, int64_t batch_size) { // Prepare data. int64_t m, n; @@ -61,7 +61,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, lda = 38; ldc = 42; - int x_len = (left_right == oneapi::mkl::side::right) ? n : m; + int x_len = (left_right == oneapi::math::side::right) ? n : m; int64_t stride_a, stride_x, stride_c; stride_x = x_len * std::abs(incx); @@ -73,8 +73,8 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, for (i = 0; i < batch_size; i++) { rand_vector(x.data() + stride_x * i, x_len, incx); - rand_matrix(A.data() + stride_a * i, layout, oneapi::mkl::transpose::nontrans, m, n, lda); - rand_matrix(C.data() + stride_c * i, layout, oneapi::mkl::transpose::nontrans, m, n, ldc); + rand_matrix(A.data() + stride_a * i, layout, oneapi::math::transpose::nontrans, m, n, lda); + rand_matrix(C.data() + stride_c * i, layout, oneapi::math::transpose::nontrans, m, n, ldc); } C_ref = C; @@ -121,13 +121,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::dgmm_batch(main_queue, left_right, m, n, A_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::dgmm_batch(main_queue, left_right, m, n, A_buffer, lda, stride_a, x_buffer, incx, stride_x, C_buffer, ldc, stride_c, batch_size); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::dgmm_batch(main_queue, left_right, m, n, A_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::dgmm_batch(main_queue, left_right, m, n, A_buffer, lda, stride_a, x_buffer, incx, stride_x, C_buffer, ldc, stride_c, batch_size); break; @@ -135,13 +135,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::dgmm_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::dgmm_batch, left_right, m, n, A_buffer, lda, stride_a, x_buffer, incx, stride_x, C_buffer, ldc, stride_c, batch_size); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::dgmm_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dgmm_batch, left_right, m, n, A_buffer, lda, stride_a, x_buffer, incx, stride_x, C_buffer, ldc, stride_c, batch_size); break; @@ -155,7 +155,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -177,76 +177,76 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, } class DgmmBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(DgmmBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 2, 5)); + oneapi::math::side::right, 2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, -2, 5)); + oneapi::math::side::right, -2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 1, 5)); + oneapi::math::side::right, 1, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 2, 5)); + oneapi::math::side::left, 2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, -2, 5)); + oneapi::math::side::left, -2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 1, 5)); + oneapi::math::side::left, 1, 5)); } TEST_P(DgmmBatchStrideTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 2, 5)); + oneapi::math::side::right, 2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, -2, 5)); + oneapi::math::side::right, -2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 1, 5)); + oneapi::math::side::right, 1, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 2, 5)); + oneapi::math::side::left, 2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, -2, 5)); + oneapi::math::side::left, -2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 1, 5)); + oneapi::math::side::left, 1, 5)); } TEST_P(DgmmBatchStrideTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 2, 5)); + oneapi::math::side::right, 2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, -2, 5)); + oneapi::math::side::right, -2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 1, 5)); + oneapi::math::side::right, 1, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 2, 5)); + oneapi::math::side::left, 2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, -2, 5)); + oneapi::math::side::left, -2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 1, 5)); + oneapi::math::side::left, 1, 5)); } TEST_P(DgmmBatchStrideTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 2, 5)); + oneapi::math::side::right, 2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, -2, 5)); + oneapi::math::side::right, -2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 1, 5)); + oneapi::math::side::right, 1, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 2, 5)); + oneapi::math::side::left, 2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, -2, 5)); + oneapi::math::side::left, -2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 1, 5)); + oneapi::math::side::left, 1, 5)); } INSTANTIATE_TEST_SUITE_P(DgmmBatchStrideTestSuite, DgmmBatchStrideTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp index 64843b7d9..eca3eb6f1 100644 --- a/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, int64_t incx, +int test(device *dev, oneapi::math::layout layout, oneapi::math::side left_right, int64_t incx, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -80,7 +80,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, lda = 38; ldc = 42; - int x_len = (left_right == oneapi::mkl::side::right) ? n : m; + int x_len = (left_right == oneapi::math::side::right) ? n : m; int64_t stride_a, stride_x, stride_c; stride_x = x_len * std::abs(incx); @@ -96,8 +96,8 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, for (i = 0; i < batch_size; i++) { rand_vector(&x[stride_x * i], x_len, incx); - rand_matrix(&A[stride_a * i], layout, oneapi::mkl::transpose::nontrans, m, n, lda); - rand_matrix(&C[stride_c * i], layout, oneapi::mkl::transpose::nontrans, m, n, ldc); + rand_matrix(&A[stride_a * i], layout, oneapi::math::transpose::nontrans, m, n, lda); + rand_matrix(&C[stride_c * i], layout, oneapi::math::transpose::nontrans, m, n, ldc); } C_ref.resize(C.size()); @@ -126,13 +126,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::dgmm_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::dgmm_batch( main_queue, left_right, m, n, &A[0], lda, stride_a, &x[0], incx, stride_x, &C[0], ldc, stride_c, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::dgmm_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::dgmm_batch( main_queue, left_right, m, n, &A[0], lda, stride_a, &x[0], incx, stride_x, &C[0], ldc, stride_c, batch_size, dependencies); break; @@ -141,13 +141,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::dgmm_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::dgmm_batch, left_right, m, n, &A[0], lda, stride_a, &x[0], incx, stride_x, &C[0], ldc, stride_c, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::dgmm_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dgmm_batch, left_right, m, n, &A[0], lda, stride_a, &x[0], incx, stride_x, &C[0], ldc, stride_c, batch_size, dependencies); break; @@ -162,7 +162,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -182,76 +182,76 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, } class DgmmBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(DgmmBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 2, 5)); + oneapi::math::side::right, 2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, -2, 5)); + oneapi::math::side::right, -2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 1, 5)); + oneapi::math::side::right, 1, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 2, 5)); + oneapi::math::side::left, 2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, -2, 5)); + oneapi::math::side::left, -2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 1, 5)); + oneapi::math::side::left, 1, 5)); } TEST_P(DgmmBatchStrideUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 2, 5)); + oneapi::math::side::right, 2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, -2, 5)); + oneapi::math::side::right, -2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 1, 5)); + oneapi::math::side::right, 1, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 2, 5)); + oneapi::math::side::left, 2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, -2, 5)); + oneapi::math::side::left, -2, 5)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 1, 5)); + oneapi::math::side::left, 1, 5)); } TEST_P(DgmmBatchStrideUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 2, 5)); + oneapi::math::side::right, 2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, -2, 5)); + oneapi::math::side::right, -2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 1, 5)); + oneapi::math::side::right, 1, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 2, 5)); + oneapi::math::side::left, 2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, -2, 5)); + oneapi::math::side::left, -2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 1, 5)); + oneapi::math::side::left, 1, 5)); } TEST_P(DgmmBatchStrideUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 2, 5)); + oneapi::math::side::right, 2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, -2, 5)); + oneapi::math::side::right, -2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, 1, 5)); + oneapi::math::side::right, 1, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 2, 5)); + oneapi::math::side::left, 2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, -2, 5)); + oneapi::math::side::left, -2, 5)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, 1, 5)); + oneapi::math::side::left, 1, 5)); } INSTANTIATE_TEST_SUITE_P(DgmmBatchStrideUsmTestSuite, DgmmBatchStrideUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp b/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp index 18234d879..01c39cad6 100644 --- a/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { +int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -73,8 +73,8 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { vector m(uaint), n(uaint), lda(uaint), incx(uaint), ldc(uaint), group_size(uaint); - auto uaside = usm_allocator(cxt, *dev); - vector left_right(uaside); + auto uaside = usm_allocator(cxt, *dev); + vector left_right(uaside); m.resize(group_count); n.resize(group_count); @@ -98,7 +98,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { incx[i] = -3 + std::rand() % 6; incx[i] = (incx[i] == 0) ? 3 : incx[i]; ldc[i] = std::max(m[i], n[i]); - left_right[i] = (oneapi::mkl::side)(std::rand() % 2); + left_right[i] = (oneapi::math::side)(std::rand() % 2); total_batch_count += group_size[i]; } @@ -112,19 +112,19 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { - size_a = (layout == oneapi::mkl::layout::col_major) ? lda[i] * n[i] : lda[i] * m[i]; - x_len = (left_right[i] == oneapi::mkl::side::R) ? n[i] : m[i]; + size_a = (layout == oneapi::math::layout::col_major) ? lda[i] * n[i] : lda[i] * m[i]; + x_len = (left_right[i] == oneapi::math::side::R) ? n[i] : m[i]; size_x = 1 + (x_len - 1) * std::abs(incx[i]); - size_c = (layout == oneapi::mkl::layout::col_major) ? ldc[i] * n[i] : ldc[i] * m[i]; + size_c = (layout == oneapi::math::layout::col_major) ? ldc[i] * n[i] : ldc[i] * m[i]; for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); - x_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_x, *dev, cxt); - c_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); - c_ref_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); - rand_matrix(a_array[idx], layout, oneapi::mkl::transpose::nontrans, m[i], n[i], lda[i]); + a_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); + x_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_x, *dev, cxt); + c_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); + c_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); + rand_matrix(a_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], lda[i]); rand_vector(x_array[idx], x_len, incx[i]); - rand_matrix(c_array[idx], layout, oneapi::mkl::transpose::nontrans, m[i], n[i], ldc[i]); - copy_matrix(c_array[idx], layout, oneapi::mkl::transpose::nontrans, m[i], n[i], ldc[i], + rand_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldc[i]); + copy_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldc[i], c_ref_array[idx]); idx++; } @@ -132,33 +132,33 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { // Call reference DGMM_BATCH. using fp_ref = typename ref_type_info::type; - int *m_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *n_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *lda_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *incx_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *ldc_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *group_size_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); + int *m_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *n_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *lda_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *incx_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *ldc_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *group_size_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); CBLAS_SIDE *left_right_ref = - (CBLAS_SIDE *)oneapi::mkl::aligned_alloc(64, sizeof(CBLAS_SIDE) * group_count); + (CBLAS_SIDE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_SIDE) * group_count); if ((m_ref == NULL) || (n_ref == NULL) || (lda_ref == NULL) || (incx_ref == NULL) || (ldc_ref == NULL) || (left_right_ref == NULL) || (group_size_ref == NULL)) { std::cout << "Error cannot allocate input arrays\n"; - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(incx_ref); - oneapi::mkl::aligned_free(ldc_ref); - oneapi::mkl::aligned_free(left_right_ref); - oneapi::mkl::aligned_free(group_size_ref); + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(incx_ref); + oneapi::math::aligned_free(ldc_ref); + oneapi::math::aligned_free(left_right_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(x_array[idx], cxt); - oneapi::mkl::free_shared(c_array[idx], cxt); - oneapi::mkl::free_shared(c_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(x_array[idx], cxt); + oneapi::math::free_shared(c_array[idx], cxt); + oneapi::math::free_shared(c_ref_array[idx], cxt); idx++; } } @@ -187,14 +187,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::dgmm_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::dgmm_batch( main_queue, &left_right[0], &m[0], &n[0], (const fp **)&a_array[0], &lda[0], (const fp **)&x_array[0], &incx[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::dgmm_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::dgmm_batch( main_queue, &left_right[0], &m[0], &n[0], (const fp **)&a_array[0], &lda[0], (const fp **)&x_array[0], &incx[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); @@ -204,14 +204,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::dgmm_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::dgmm_batch, &left_right[0], &m[0], &n[0], (const fp **)&a_array[0], &lda[0], (const fp **)&x_array[0], &incx[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::dgmm_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dgmm_batch, &left_right[0], &m[0], &n[0], (const fp **)&a_array[0], &lda[0], (const fp **)&x_array[0], &incx[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); @@ -227,21 +227,21 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(incx_ref); - oneapi::mkl::aligned_free(ldc_ref); - oneapi::mkl::aligned_free(left_right_ref); - oneapi::mkl::aligned_free(group_size_ref); + catch (const oneapi::math::unimplemented &e) { + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(incx_ref); + oneapi::math::aligned_free(ldc_ref); + oneapi::math::aligned_free(left_right_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(x_array[idx], cxt); - oneapi::mkl::free_shared(c_array[idx], cxt); - oneapi::mkl::free_shared(c_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(x_array[idx], cxt); + oneapi::math::free_shared(c_array[idx], cxt); + oneapi::math::free_shared(c_ref_array[idx], cxt); idx++; } } @@ -263,20 +263,20 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { } } - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(incx_ref); - oneapi::mkl::aligned_free(ldc_ref); - oneapi::mkl::aligned_free(left_right_ref); - oneapi::mkl::aligned_free(group_size_ref); + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(incx_ref); + oneapi::math::aligned_free(ldc_ref); + oneapi::math::aligned_free(left_right_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(x_array[idx], cxt); - oneapi::mkl::free_shared(c_array[idx], cxt); - oneapi::mkl::free_shared(c_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(x_array[idx], cxt); + oneapi::math::free_shared(c_array[idx], cxt); + oneapi::math::free_shared(c_ref_array[idx], cxt); idx++; } } @@ -285,7 +285,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { } class DgmmBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(DgmmBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -311,8 +311,8 @@ TEST_P(DgmmBatchUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(DgmmBatchUsmTestSuite, DgmmBatchUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/gemm_batch_stride.cpp b/tests/unit_tests/blas/batch/gemm_batch_stride.cpp index 537d39ee1..999910829 100644 --- a/tests/unit_tests/blas/batch/gemm_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/gemm_batch_stride.cpp @@ -48,11 +48,11 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n, k; int64_t lda, ldb, ldc; - oneapi::mkl::transpose transa, transb; + oneapi::math::transpose transa, transb; Ts alpha, beta; int64_t i, tmp; @@ -70,31 +70,31 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { (std::is_same>::value)) { tmp = std::rand() % 3; if (tmp == 2) - transa = oneapi::mkl::transpose::conjtrans; + transa = oneapi::math::transpose::conjtrans; else - transa = (oneapi::mkl::transpose)tmp; + transa = (oneapi::math::transpose)tmp; tmp = std::rand() % 3; if (tmp == 2) - transb = oneapi::mkl::transpose::conjtrans; + transb = oneapi::math::transpose::conjtrans; else - transb = (oneapi::mkl::transpose)tmp; + transb = (oneapi::math::transpose)tmp; } else { - transa = (oneapi::mkl::transpose)(std::rand() % 2); - transb = (oneapi::mkl::transpose)(std::rand() % 2); + transa = (oneapi::math::transpose)(std::rand() % 2); + transb = (oneapi::math::transpose)(std::rand() % 2); } int64_t stride_a, stride_b, stride_c; switch (layout) { - case oneapi::mkl::layout::col_major: - stride_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * k : lda * m; - stride_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * k; + case oneapi::math::layout::col_major: + stride_a = (transa == oneapi::math::transpose::nontrans) ? lda * k : lda * m; + stride_b = (transb == oneapi::math::transpose::nontrans) ? ldb * n : ldb * k; stride_c = ldc * n; break; - case oneapi::mkl::layout::row_major: - stride_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * m : lda * k; - stride_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * k : ldb * n; + case oneapi::math::layout::row_major: + stride_a = (transa == oneapi::math::transpose::nontrans) ? lda * m : lda * k; + stride_b = (transb == oneapi::math::transpose::nontrans) ? ldb * k : ldb * n; stride_c = ldc * m; break; default: break; @@ -110,7 +110,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { for (i = 0; i < batch_size; i++) { rand_matrix(A.data() + stride_a * i, layout, transa, m, k, lda); rand_matrix(B.data() + stride_b * i, layout, transb, k, n, ldb); - rand_matrix(C.data() + stride_c * i, layout, oneapi::mkl::transpose::nontrans, m, n, ldc); + rand_matrix(C.data() + stride_c * i, layout, oneapi::math::transpose::nontrans, m, n, ldc); } for (size_t i = 0; i < A.size(); ++i) { @@ -168,13 +168,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::gemm_batch( + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::gemm_batch( main_queue, transa, transb, m, n, k, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, beta, C_buffer, ldc, stride_c, batch_size); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::gemm_batch( + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::gemm_batch( main_queue, transa, transb, m, n, k, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, beta, C_buffer, ldc, stride_c, batch_size); break; @@ -182,14 +182,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemm_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemm_batch, transa, transb, m, n, k, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, beta, C_buffer, ldc, stride_c, batch_size); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemm_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm_batch, transa, transb, m, n, k, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, beta, C_buffer, ldc, stride_c, batch_size); @@ -205,7 +205,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -224,7 +224,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { C_cast_ref[i] = C_ref[i]; } auto C_accessor = C_buffer.get_host_access(read_only); - bool good = check_almost_equal_matrix(C_accessor, C_cast_ref, oneapi::mkl::layout::col_major, + bool good = check_almost_equal_matrix(C_accessor, C_cast_ref, oneapi::math::layout::col_major, stride_c * batch_size, 1, stride_c * batch_size, error_mag, std::cout); @@ -232,7 +232,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { } class GemmBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemmBatchStrideTests, RealHalfPrecision) { EXPECT_TRUEORSKIP((test( @@ -282,8 +282,8 @@ TEST_P(GemmBatchStrideTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GemmBatchStrideTestSuite, GemmBatchStrideTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp index b4f10dc7d..d556b81e1 100644 --- a/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -71,7 +71,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n, k; int64_t lda, ldb, ldc; - oneapi::mkl::transpose transa, transb; + oneapi::math::transpose transa, transb; Ts alpha, beta; int64_t i, tmp; @@ -89,31 +89,31 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { (std::is_same>::value)) { tmp = std::rand() % 3; if (tmp == 2) - transa = oneapi::mkl::transpose::conjtrans; + transa = oneapi::math::transpose::conjtrans; else - transa = (oneapi::mkl::transpose)tmp; + transa = (oneapi::math::transpose)tmp; tmp = std::rand() % 3; if (tmp == 2) - transb = oneapi::mkl::transpose::conjtrans; + transb = oneapi::math::transpose::conjtrans; else - transb = (oneapi::mkl::transpose)tmp; + transb = (oneapi::math::transpose)tmp; } else { - transa = (oneapi::mkl::transpose)(std::rand() % 2); - transb = (oneapi::mkl::transpose)(std::rand() % 2); + transa = (oneapi::math::transpose)(std::rand() % 2); + transb = (oneapi::math::transpose)(std::rand() % 2); } int64_t stride_a, stride_b, stride_c; switch (layout) { - case oneapi::mkl::layout::col_major: - stride_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * k : lda * m; - stride_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * k; + case oneapi::math::layout::col_major: + stride_a = (transa == oneapi::math::transpose::nontrans) ? lda * k : lda * m; + stride_b = (transb == oneapi::math::transpose::nontrans) ? ldb * n : ldb * k; stride_c = ldc * n; break; - case oneapi::mkl::layout::row_major: - stride_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * m : lda * k; - stride_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * k : ldb * n; + case oneapi::math::layout::row_major: + stride_a = (transa == oneapi::math::transpose::nontrans) ? lda * m : lda * k; + stride_b = (transb == oneapi::math::transpose::nontrans) ? ldb * k : ldb * n; stride_c = ldc * m; break; default: break; @@ -136,17 +136,17 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { C_ref.resize(stride_c * batch_size); C_cast_ref.resize(stride_c * batch_size); - Ta **a_array = (Ta **)oneapi::mkl::malloc_shared(64, sizeof(Ta *) * batch_size, *dev, cxt); - Tb **b_array = (Tb **)oneapi::mkl::malloc_shared(64, sizeof(Tb *) * batch_size, *dev, cxt); - Tc **c_array = (Tc **)oneapi::mkl::malloc_shared(64, sizeof(Tc *) * batch_size, *dev, cxt); - Ts **c_ref_array = (Ts **)oneapi::mkl::malloc_shared(64, sizeof(Ts *) * batch_size, *dev, cxt); + Ta **a_array = (Ta **)oneapi::math::malloc_shared(64, sizeof(Ta *) * batch_size, *dev, cxt); + Tb **b_array = (Tb **)oneapi::math::malloc_shared(64, sizeof(Tb *) * batch_size, *dev, cxt); + Tc **c_array = (Tc **)oneapi::math::malloc_shared(64, sizeof(Tc *) * batch_size, *dev, cxt); + Ts **c_ref_array = (Ts **)oneapi::math::malloc_shared(64, sizeof(Ts *) * batch_size, *dev, cxt); if ((a_array == NULL) || (b_array == NULL) || (c_array == NULL) || (c_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(b_array, cxt); - oneapi::mkl::free_shared(c_array, cxt); - oneapi::mkl::free_shared(c_ref_array, cxt); + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(b_array, cxt); + oneapi::math::free_shared(c_array, cxt); + oneapi::math::free_shared(c_ref_array, cxt); return false; } @@ -157,17 +157,17 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { c_ref_array[i] = &C_ref[i * stride_c]; } - rand_matrix(A, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(A, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_a * batch_size, 1, stride_a * batch_size); - rand_matrix(B, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(B, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_b * batch_size, 1, stride_b * batch_size); - rand_matrix(C, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(C, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_c * batch_size, 1, stride_c * batch_size); - copy_matrix(A, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + copy_matrix(A, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_a * batch_size, 1, stride_a * batch_size, A_ref); - copy_matrix(B, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + copy_matrix(B, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_b * batch_size, 1, stride_b * batch_size, B_ref); - copy_matrix(C, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + copy_matrix(C, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_c * batch_size, 1, stride_c * batch_size, C_ref); // Call reference GEMM_BATCH_STRIDE. @@ -194,13 +194,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::gemm_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::gemm_batch( main_queue, transa, transb, m, n, k, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, beta, &C[0], ldc, stride_c, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::gemm_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::gemm_batch( main_queue, transa, transb, m, n, k, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, beta, &C[0], ldc, stride_c, batch_size, dependencies); break; @@ -209,14 +209,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { done.wait_and_throw(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemm_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemm_batch, transa, transb, m, n, k, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, beta, &C[0], ldc, stride_c, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemm_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm_batch, transa, transb, m, n, k, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, beta, &C[0], ldc, stride_c, batch_size, dependencies); @@ -232,11 +232,11 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(b_array, cxt); - oneapi::mkl::free_shared(c_array, cxt); - oneapi::mkl::free_shared(c_ref_array, cxt); + catch (const oneapi::math::unimplemented &e) { + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(b_array, cxt); + oneapi::math::free_shared(c_array, cxt); + oneapi::math::free_shared(c_ref_array, cxt); return test_skipped; } @@ -254,20 +254,20 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { for (size_t i = 0; i < C_ref.size(); ++i) { C_cast_ref[i] = C_ref[i]; } - bool good = check_almost_equal_matrix(C, C_cast_ref, oneapi::mkl::layout::col_major, + bool good = check_almost_equal_matrix(C, C_cast_ref, oneapi::math::layout::col_major, stride_c * batch_size, 1, stride_c * batch_size, error_mag, std::cout); - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(b_array, cxt); - oneapi::mkl::free_shared(c_array, cxt); - oneapi::mkl::free_shared(c_ref_array, cxt); + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(b_array, cxt); + oneapi::math::free_shared(c_array, cxt); + oneapi::math::free_shared(c_ref_array, cxt); return (int)good; } class GemmBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemmBatchStrideUsmTests, RealHalfPrecision) { EXPECT_TRUEORSKIP((test( @@ -317,8 +317,8 @@ TEST_P(GemmBatchStrideUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GemmBatchStrideUsmTestSuite, GemmBatchStrideUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/gemm_batch_usm.cpp b/tests/unit_tests/blas/batch/gemm_batch_usm.cpp index fa7ee938d..c02279b53 100644 --- a/tests/unit_tests/blas/batch/gemm_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/gemm_batch_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { +int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -73,8 +73,8 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { vector m(uaint), n(uaint), k(uaint), lda(uaint), ldb(uaint), ldc(uaint), group_size(uaint); - auto uatranspose = usm_allocator(cxt, *dev); - vector transa(uatranspose), transb(uatranspose); + auto uatranspose = usm_allocator(cxt, *dev); + vector transa(uatranspose), transb(uatranspose); auto uaTs = usm_allocator(cxt, *dev); vector alpha(uaTs), beta(uaTs); @@ -110,18 +110,18 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { (std::is_same>::value)) { tmp = std::rand() % 3; if (tmp == 2) - transa[i] = oneapi::mkl::transpose::conjtrans; + transa[i] = oneapi::math::transpose::conjtrans; else - transa[i] = (oneapi::mkl::transpose)tmp; + transa[i] = (oneapi::math::transpose)tmp; tmp = std::rand() % 3; if (tmp == 2) - transb[i] = oneapi::mkl::transpose::conjtrans; + transb[i] = oneapi::math::transpose::conjtrans; else - transb[i] = (oneapi::mkl::transpose)tmp; + transb[i] = (oneapi::math::transpose)tmp; } else { - transa[i] = (oneapi::mkl::transpose)(std::rand() % 2); - transb[i] = (oneapi::mkl::transpose)(std::rand() % 2); + transa[i] = (oneapi::math::transpose)(std::rand() % 2); + transb[i] = (oneapi::math::transpose)(std::rand() % 2); } total_batch_count += group_size[i]; } @@ -145,33 +145,33 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { switch (layout) { - case oneapi::mkl::layout::col_major: - size_a = lda[i] * ((transa[i] == oneapi::mkl::transpose::nontrans) ? k[i] : m[i]); - size_b = ldb[i] * ((transb[i] == oneapi::mkl::transpose::nontrans) ? n[i] : k[i]); + case oneapi::math::layout::col_major: + size_a = lda[i] * ((transa[i] == oneapi::math::transpose::nontrans) ? k[i] : m[i]); + size_b = ldb[i] * ((transb[i] == oneapi::math::transpose::nontrans) ? n[i] : k[i]); size_c = ldc[i] * n[i]; break; - case oneapi::mkl::layout::row_major: - size_a = lda[i] * ((transa[i] == oneapi::mkl::transpose::nontrans) ? m[i] : k[i]); - size_b = ldb[i] * ((transb[i] == oneapi::mkl::transpose::nontrans) ? k[i] : n[i]); + case oneapi::math::layout::row_major: + size_a = lda[i] * ((transa[i] == oneapi::math::transpose::nontrans) ? m[i] : k[i]); + size_b = ldb[i] * ((transb[i] == oneapi::math::transpose::nontrans) ? k[i] : n[i]); size_c = ldc[i] * m[i]; break; default: break; } for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (Ta *)oneapi::mkl::malloc_shared(64, sizeof(Ta) * size_a, *dev, cxt); - b_array[idx] = (Tb *)oneapi::mkl::malloc_shared(64, sizeof(Tb) * size_b, *dev, cxt); - c_array[idx] = (Tc *)oneapi::mkl::malloc_shared(64, sizeof(Tc) * size_c, *dev, cxt); - a_ref_array[idx] = (Ts *)oneapi::mkl::malloc_shared(64, sizeof(Ts) * size_a, *dev, cxt); - b_ref_array[idx] = (Ts *)oneapi::mkl::malloc_shared(64, sizeof(Ts) * size_b, *dev, cxt); + a_array[idx] = (Ta *)oneapi::math::malloc_shared(64, sizeof(Ta) * size_a, *dev, cxt); + b_array[idx] = (Tb *)oneapi::math::malloc_shared(64, sizeof(Tb) * size_b, *dev, cxt); + c_array[idx] = (Tc *)oneapi::math::malloc_shared(64, sizeof(Tc) * size_c, *dev, cxt); + a_ref_array[idx] = (Ts *)oneapi::math::malloc_shared(64, sizeof(Ts) * size_a, *dev, cxt); + b_ref_array[idx] = (Ts *)oneapi::math::malloc_shared(64, sizeof(Ts) * size_b, *dev, cxt); c_cast_ref_array[idx] = - (Tc *)oneapi::mkl::malloc_shared(64, sizeof(Tc) * size_c, *dev, cxt); - c_ref_array[idx] = (Ts *)oneapi::mkl::malloc_shared(64, sizeof(Ts) * size_c, *dev, cxt); + (Tc *)oneapi::math::malloc_shared(64, sizeof(Tc) * size_c, *dev, cxt); + c_ref_array[idx] = (Ts *)oneapi::math::malloc_shared(64, sizeof(Ts) * size_c, *dev, cxt); rand_matrix(a_array[idx], layout, transa[i], m[i], k[i], lda[i]); rand_matrix(b_array[idx], layout, transb[i], k[i], n[i], ldb[i]); - rand_matrix(c_array[idx], layout, oneapi::mkl::transpose::nontrans, m[i], n[i], ldc[i]); + rand_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldc[i]); copy_matrix(a_array[idx], layout, transa[i], m[i], k[i], lda[i], a_ref_array[idx]); copy_matrix(b_array[idx], layout, transb[i], k[i], n[i], ldb[i], b_ref_array[idx]); - copy_matrix(c_array[idx], layout, oneapi::mkl::transpose::nontrans, m[i], n[i], ldc[i], + copy_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldc[i], c_ref_array[idx]); idx++; } @@ -179,42 +179,42 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { // Call reference GEMM_BATCH. using fp_ref = typename ref_type_info::type; - int *m_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *n_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *k_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *lda_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *ldb_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *ldc_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *group_size_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); + int *m_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *n_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *k_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *lda_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *ldb_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *ldc_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *group_size_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); CBLAS_TRANSPOSE *transa_ref = - (CBLAS_TRANSPOSE *)oneapi::mkl::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); + (CBLAS_TRANSPOSE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); CBLAS_TRANSPOSE *transb_ref = - (CBLAS_TRANSPOSE *)oneapi::mkl::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); + (CBLAS_TRANSPOSE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); if ((m_ref == NULL) || (n_ref == NULL) || (k_ref == NULL) || (lda_ref == NULL) || (ldb_ref == NULL) || (ldc_ref == NULL) || (transa_ref == NULL) || (transb_ref == NULL) || (group_size_ref == NULL)) { std::cout << "Error cannot allocate input arrays\n"; - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(k_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(ldb_ref); - oneapi::mkl::aligned_free(ldc_ref); - oneapi::mkl::aligned_free(transa_ref); - oneapi::mkl::aligned_free(transb_ref); - oneapi::mkl::aligned_free(group_size_ref); + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(k_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(ldb_ref); + oneapi::math::aligned_free(ldc_ref); + oneapi::math::aligned_free(transa_ref); + oneapi::math::aligned_free(transb_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(b_array[idx], cxt); - oneapi::mkl::free_shared(c_array[idx], cxt); - oneapi::mkl::free_shared(a_ref_array[idx], cxt); - oneapi::mkl::free_shared(b_ref_array[idx], cxt); - oneapi::mkl::free_shared(c_cast_ref_array[idx], cxt); - oneapi::mkl::free_shared(c_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(b_array[idx], cxt); + oneapi::math::free_shared(c_array[idx], cxt); + oneapi::math::free_shared(a_ref_array[idx], cxt); + oneapi::math::free_shared(b_ref_array[idx], cxt); + oneapi::math::free_shared(c_cast_ref_array[idx], cxt); + oneapi::math::free_shared(c_ref_array[idx], cxt); idx++; } } @@ -247,14 +247,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::gemm_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::gemm_batch( main_queue, &transa[0], &transb[0], &m[0], &n[0], &k[0], &alpha[0], (const Ta **)&a_array[0], &lda[0], (const Tb **)&b_array[0], &ldb[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::gemm_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::gemm_batch( main_queue, &transa[0], &transb[0], &m[0], &n[0], &k[0], &alpha[0], (const Ta **)&a_array[0], &lda[0], (const Tb **)&b_array[0], &ldb[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); @@ -264,15 +264,15 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { done.wait_and_throw(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemm_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemm_batch, &transa[0], &transb[0], &m[0], &n[0], &k[0], &alpha[0], (const Ta **)&a_array[0], &lda[0], (const Tb **)&b_array[0], &ldb[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemm_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm_batch, &transa[0], &transb[0], &m[0], &n[0], &k[0], &alpha[0], (const Ta **)&a_array[0], &lda[0], (const Ta **)&b_array[0], &ldb[0], &beta[0], &c_array[0], &ldc[0], group_count, @@ -289,26 +289,26 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(k_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(ldb_ref); - oneapi::mkl::aligned_free(ldc_ref); - oneapi::mkl::aligned_free(transa_ref); - oneapi::mkl::aligned_free(transb_ref); - oneapi::mkl::aligned_free(group_size_ref); + catch (const oneapi::math::unimplemented &e) { + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(k_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(ldb_ref); + oneapi::math::aligned_free(ldc_ref); + oneapi::math::aligned_free(transa_ref); + oneapi::math::aligned_free(transb_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(b_array[idx], cxt); - oneapi::mkl::free_shared(c_array[idx], cxt); - oneapi::mkl::free_shared(a_ref_array[idx], cxt); - oneapi::mkl::free_shared(b_ref_array[idx], cxt); - oneapi::mkl::free_shared(c_cast_ref_array[idx], cxt); - oneapi::mkl::free_shared(c_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(b_array[idx], cxt); + oneapi::math::free_shared(c_array[idx], cxt); + oneapi::math::free_shared(a_ref_array[idx], cxt); + oneapi::math::free_shared(b_ref_array[idx], cxt); + oneapi::math::free_shared(c_cast_ref_array[idx], cxt); + oneapi::math::free_shared(c_ref_array[idx], cxt); idx++; } } @@ -330,32 +330,32 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { if (std::is_same_v) error_mag = 1; - copy_matrix(c_ref_array[idx], layout, oneapi::mkl::transpose::nontrans, m[i], n[i], + copy_matrix(c_ref_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldc[i], c_cast_ref_array[idx]); good = good && check_almost_equal_matrix(c_array[idx], c_cast_ref_array[idx], layout, m[i], n[i], ldc[i], error_mag, std::cout); idx++; } } - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(k_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(ldb_ref); - oneapi::mkl::aligned_free(ldc_ref); - oneapi::mkl::aligned_free(transa_ref); - oneapi::mkl::aligned_free(transb_ref); - oneapi::mkl::aligned_free(group_size_ref); + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(k_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(ldb_ref); + oneapi::math::aligned_free(ldc_ref); + oneapi::math::aligned_free(transa_ref); + oneapi::math::aligned_free(transb_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(b_array[idx], cxt); - oneapi::mkl::free_shared(c_array[idx], cxt); - oneapi::mkl::free_shared(a_ref_array[idx], cxt); - oneapi::mkl::free_shared(b_ref_array[idx], cxt); - oneapi::mkl::free_shared(c_cast_ref_array[idx], cxt); - oneapi::mkl::free_shared(c_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(b_array[idx], cxt); + oneapi::math::free_shared(c_array[idx], cxt); + oneapi::math::free_shared(a_ref_array[idx], cxt); + oneapi::math::free_shared(b_ref_array[idx], cxt); + oneapi::math::free_shared(c_cast_ref_array[idx], cxt); + oneapi::math::free_shared(c_ref_array[idx], cxt); idx++; } } @@ -364,7 +364,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { } class GemmBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemmBatchUsmTests, RealHalfPrecision) { EXPECT_TRUEORSKIP((test( @@ -414,8 +414,8 @@ TEST_P(GemmBatchUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GemmBatchUsmTestSuite, GemmBatchUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/gemv_batch_stride.cpp b/tests/unit_tests/blas/batch/gemv_batch_stride.cpp index 98677a0bd..8d4780a00 100644 --- a/tests/unit_tests/blas/batch/gemv_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/gemv_batch_stride.cpp @@ -48,11 +48,11 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { // Prepare data. int64_t m, n; int64_t lda; - oneapi::mkl::transpose transa; + oneapi::math::transpose transa; fp alpha, beta; int64_t i, tmp; @@ -64,14 +64,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in beta = rand_scalar(); if ((std::is_same::value) || (std::is_same::value)) { - transa = (oneapi::mkl::transpose)(std::rand() % 2); + transa = (oneapi::math::transpose)(std::rand() % 2); } else { tmp = std::rand() % 3; if (tmp == 2) - transa = oneapi::mkl::transpose::conjtrans; + transa = oneapi::math::transpose::conjtrans; else - transa = (oneapi::mkl::transpose)tmp; + transa = (oneapi::math::transpose)tmp; } int x_len = outer_dimension(transa, m, n); @@ -88,7 +88,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in for (i = 0; i < batch_size; i++) { rand_vector(x.data() + stride_x * i, x_len, incx); rand_vector(y.data() + stride_y * i, y_len, incy); - rand_matrix(A.data() + stride_a * i, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A.data() + stride_a * i, layout, oneapi::math::transpose::nontrans, m, n, lda); } y_ref = y; @@ -136,13 +136,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::gemv_batch( + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::gemv_batch( main_queue, transa, m, n, alpha, A_buffer, lda, stride_a, x_buffer, incx, stride_x, beta, y_buffer, incy, stride_y, batch_size); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::gemv_batch( + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::gemv_batch( main_queue, transa, m, n, alpha, A_buffer, lda, stride_a, x_buffer, incx, stride_x, beta, y_buffer, incy, stride_y, batch_size); break; @@ -150,13 +150,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemv_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemv_batch, transa, m, n, alpha, A_buffer, lda, stride_a, x_buffer, incx, stride_x, beta, y_buffer, incy, stride_y, batch_size); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemv_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemv_batch, transa, m, n, alpha, A_buffer, lda, stride_a, x_buffer, incx, stride_x, beta, y_buffer, incy, stride_y, batch_size); break; @@ -170,7 +170,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -192,7 +192,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in } class GemvBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemvBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 2, 3, 5)); @@ -226,8 +226,8 @@ TEST_P(GemvBatchStrideTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GemvBatchStrideTestSuite, GemvBatchStrideTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp index 0539f7eac..39eae0399 100644 --- a/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -71,7 +71,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in // Prepare data. int64_t m, n; int64_t lda; - oneapi::mkl::transpose transa; + oneapi::math::transpose transa; fp alpha, beta; int64_t i, tmp; @@ -83,14 +83,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in beta = rand_scalar(); if ((std::is_same::value) || (std::is_same::value)) { - transa = (oneapi::mkl::transpose)(std::rand() % 2); + transa = (oneapi::math::transpose)(std::rand() % 2); } else { tmp = std::rand() % 3; if (tmp == 2) - transa = oneapi::mkl::transpose::conjtrans; + transa = oneapi::math::transpose::conjtrans; else - transa = (oneapi::mkl::transpose)tmp; + transa = (oneapi::math::transpose)tmp; } int x_len = outer_dimension(transa, m, n); @@ -111,7 +111,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in for (i = 0; i < batch_size; i++) { rand_vector(&x[stride_x * i], x_len, incx); rand_vector(&y[stride_y * i], y_len, incy); - rand_matrix(&A[stride_a * i], layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(&A[stride_a * i], layout, oneapi::math::transpose::nontrans, m, n, lda); } y_ref.resize(y.size()); @@ -139,13 +139,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::gemv_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::gemv_batch( main_queue, transa, m, n, alpha, &A[0], lda, stride_a, &x[0], incx, stride_x, beta, &y[0], incy, stride_y, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::gemv_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::gemv_batch( main_queue, transa, m, n, alpha, &A[0], lda, stride_a, &x[0], incx, stride_x, beta, &y[0], incy, stride_y, batch_size, dependencies); break; @@ -154,14 +154,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemv_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemv_batch, transa, m, n, alpha, &A[0], lda, stride_a, &x[0], incx, stride_x, beta, &y[0], incy, stride_y, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemv_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemv_batch, transa, m, n, alpha, &A[0], lda, stride_a, &x[0], incx, stride_x, beta, &y[0], incy, stride_y, batch_size, dependencies); @@ -177,7 +177,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -197,7 +197,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t incx, int64_t incy, in } class GemvBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemvBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 2, 3, 5)); @@ -231,8 +231,8 @@ TEST_P(GemvBatchStrideUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GemvBatchStrideUsmTestSuite, GemvBatchStrideUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/gemv_batch_usm.cpp b/tests/unit_tests/blas/batch/gemv_batch_usm.cpp index f239c17cf..40fffe810 100644 --- a/tests/unit_tests/blas/batch/gemv_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/gemv_batch_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { +int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -73,8 +73,8 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { vector m(uaint), n(uaint), lda(uaint), incx(uaint), incy(uaint), group_size(uaint); - auto uatranspose = usm_allocator(cxt, *dev); - vector transa(uatranspose); + auto uatranspose = usm_allocator(cxt, *dev); + vector transa(uatranspose); auto uafp = usm_allocator(cxt, *dev); vector alpha(uafp), beta(uafp); @@ -107,14 +107,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { alpha[i] = rand_scalar(); beta[i] = rand_scalar(); if ((std::is_same::value) || (std::is_same::value)) { - transa[i] = (oneapi::mkl::transpose)(std::rand() % 2); + transa[i] = (oneapi::math::transpose)(std::rand() % 2); } else { tmp = std::rand() % 3; if (tmp == 2) - transa[i] = oneapi::mkl::transpose::conjtrans; + transa[i] = oneapi::math::transpose::conjtrans; else - transa[i] = (oneapi::mkl::transpose)tmp; + transa[i] = (oneapi::math::transpose)tmp; } total_batch_count += group_size[i]; } @@ -129,17 +129,17 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { - size_a = (layout == oneapi::mkl::layout::col_major) ? lda[i] * n[i] : lda[i] * m[i]; - x_len = (transa[i] == oneapi::mkl::transpose::nontrans) ? n[i] : m[i]; - y_len = (transa[i] == oneapi::mkl::transpose::nontrans) ? m[i] : n[i]; + size_a = (layout == oneapi::math::layout::col_major) ? lda[i] * n[i] : lda[i] * m[i]; + x_len = (transa[i] == oneapi::math::transpose::nontrans) ? n[i] : m[i]; + y_len = (transa[i] == oneapi::math::transpose::nontrans) ? m[i] : n[i]; size_x = 1 + (x_len - 1) * std::abs(incx[i]); size_y = 1 + (y_len - 1) * std::abs(incy[i]); for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); - x_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_x, *dev, cxt); - y_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_y, *dev, cxt); - y_ref_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_y, *dev, cxt); - rand_matrix(a_array[idx], layout, oneapi::mkl::transpose::nontrans, m[i], n[i], lda[i]); + a_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); + x_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_x, *dev, cxt); + y_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_y, *dev, cxt); + y_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_y, *dev, cxt); + rand_matrix(a_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], lda[i]); rand_vector(x_array[idx], x_len, incx[i]); rand_vector(y_array[idx], y_len, incy[i]); copy_vector(y_array[idx], y_len, incy[i], y_ref_array[idx]); @@ -149,33 +149,33 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { // Call reference GEMV_BATCH. using fp_ref = typename ref_type_info::type; - int *m_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *n_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *lda_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *incx_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *incy_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *group_size_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); + int *m_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *n_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *lda_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *incx_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *incy_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *group_size_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); CBLAS_TRANSPOSE *transa_ref = - (CBLAS_TRANSPOSE *)oneapi::mkl::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); + (CBLAS_TRANSPOSE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); if ((m_ref == NULL) || (n_ref == NULL) || (lda_ref == NULL) || (incx_ref == NULL) || (incy_ref == NULL) || (transa_ref == NULL) || (group_size_ref == NULL)) { std::cout << "Error cannot allocate input arrays\n"; - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(incx_ref); - oneapi::mkl::aligned_free(incy_ref); - oneapi::mkl::aligned_free(transa_ref); - oneapi::mkl::aligned_free(group_size_ref); + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(incx_ref); + oneapi::math::aligned_free(incy_ref); + oneapi::math::aligned_free(transa_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(x_array[idx], cxt); - oneapi::mkl::free_shared(y_array[idx], cxt); - oneapi::mkl::free_shared(y_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(x_array[idx], cxt); + oneapi::math::free_shared(y_array[idx], cxt); + oneapi::math::free_shared(y_ref_array[idx], cxt); idx++; } } @@ -205,14 +205,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::gemv_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::gemv_batch( main_queue, &transa[0], &m[0], &n[0], &alpha[0], (const fp **)&a_array[0], &lda[0], (const fp **)&x_array[0], &incx[0], &beta[0], &y_array[0], &incy[0], group_count, &group_size[0], dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::gemv_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::gemv_batch( main_queue, &transa[0], &m[0], &n[0], &alpha[0], (const fp **)&a_array[0], &lda[0], (const fp **)&x_array[0], &incx[0], &beta[0], &y_array[0], &incy[0], group_count, &group_size[0], dependencies); @@ -222,15 +222,15 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemv_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemv_batch, &transa[0], &m[0], &n[0], &alpha[0], (const fp **)&a_array[0], &lda[0], (const fp **)&x_array[0], &incx[0], &beta[0], &y_array[0], &incy[0], group_count, &group_size[0], dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemv_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemv_batch, &transa[0], &m[0], &n[0], &alpha[0], (const fp **)&a_array[0], &lda[0], (const fp **)&x_array[0], &incx[0], &beta[0], &y_array[0], &incy[0], group_count, @@ -247,21 +247,21 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(incx_ref); - oneapi::mkl::aligned_free(incy_ref); - oneapi::mkl::aligned_free(transa_ref); - oneapi::mkl::aligned_free(group_size_ref); + catch (const oneapi::math::unimplemented &e) { + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(incx_ref); + oneapi::math::aligned_free(incy_ref); + oneapi::math::aligned_free(transa_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(x_array[idx], cxt); - oneapi::mkl::free_shared(y_array[idx], cxt); - oneapi::mkl::free_shared(y_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(x_array[idx], cxt); + oneapi::math::free_shared(y_array[idx], cxt); + oneapi::math::free_shared(y_ref_array[idx], cxt); idx++; } } @@ -276,7 +276,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { // Compare the results of reference implementation and DPC++ implementation. idx = 0; for (i = 0; i < group_count; i++) { - y_len = (transa[i] == oneapi::mkl::transpose::nontrans) ? m[i] : n[i]; + y_len = (transa[i] == oneapi::math::transpose::nontrans) ? m[i] : n[i]; for (j = 0; j < group_size[i]; j++) { good = good && check_equal_vector(y_array[idx], y_ref_array[idx], y_len, incy[i], std::max(m[i], n[i]), std::cout); @@ -284,20 +284,20 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { } } - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(incx_ref); - oneapi::mkl::aligned_free(incy_ref); - oneapi::mkl::aligned_free(transa_ref); - oneapi::mkl::aligned_free(group_size_ref); + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(incx_ref); + oneapi::math::aligned_free(incy_ref); + oneapi::math::aligned_free(transa_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(x_array[idx], cxt); - oneapi::mkl::free_shared(y_array[idx], cxt); - oneapi::mkl::free_shared(y_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(x_array[idx], cxt); + oneapi::math::free_shared(y_array[idx], cxt); + oneapi::math::free_shared(y_ref_array[idx], cxt); idx++; } } @@ -306,7 +306,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { } class GemvBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemvBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -332,8 +332,8 @@ TEST_P(GemvBatchUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GemvBatchUsmTestSuite, GemvBatchUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp b/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp index fa287d187..c28a4764e 100644 --- a/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp @@ -48,11 +48,11 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n; int64_t lda, ldb; - oneapi::mkl::transpose trans; + oneapi::math::transpose trans; fp alpha; int64_t i, tmp; @@ -66,14 +66,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { int64_t stride_a, stride_b, stride; switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: stride_a = lda * n; - stride_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + stride_b = (trans == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; stride = std::max(stride_a, stride_b); break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: stride_a = lda * m; - stride_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + stride_b = (trans == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; stride = std::max(stride_a, stride_b); break; default: break; @@ -81,9 +81,9 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { vector> AB(stride * batch_size), AB_ref(stride * batch_size); - rand_matrix(AB.data(), oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(AB.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride * batch_size, 1, stride * batch_size); - copy_matrix(AB.data(), oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + copy_matrix(AB.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride * batch_size, 1, stride * batch_size, AB_ref.data()); // Call reference IMATCOPY_BATCH_STRIDE. @@ -120,25 +120,25 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::imatcopy_batch( + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::imatcopy_batch( main_queue, trans, m, n, alpha, AB_buffer, lda, ldb, stride, batch_size); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::imatcopy_batch( + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::imatcopy_batch( main_queue, trans, m, n, alpha, AB_buffer, lda, ldb, stride, batch_size); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::imatcopy_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::imatcopy_batch, trans, m, n, alpha, AB_buffer, lda, ldb, stride, batch_size); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::imatcopy_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::imatcopy_batch, trans, m, n, alpha, AB_buffer, lda, ldb, stride, batch_size); break; @@ -152,7 +152,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -164,14 +164,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { // Compare the results of reference implementation and DPC++ implementation. auto AB_accessor = AB_buffer.get_host_access(read_only); - bool good = check_equal_matrix(AB_accessor, AB_ref, oneapi::mkl::layout::col_major, + bool good = check_equal_matrix(AB_accessor, AB_ref, oneapi::math::layout::col_major, stride * batch_size, 1, stride * batch_size, 10, std::cout); return (int)good; } class ImatcopyBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(ImatcopyBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -197,8 +197,8 @@ TEST_P(ImatcopyBatchStrideTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(ImatcopyBatchStrideTestSuite, ImatcopyBatchStrideTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp index ab9fb95f4..5abb649fb 100644 --- a/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -71,7 +71,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n; int64_t lda, ldb; - oneapi::mkl::transpose trans; + oneapi::math::transpose trans; fp alpha; int64_t i; @@ -85,14 +85,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { int64_t stride_a, stride_b, stride; switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: stride_a = lda * n; - stride_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + stride_b = (trans == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; stride = std::max(stride_a, stride_b); break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: stride_a = lda * m; - stride_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + stride_b = (trans == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; stride = std::max(stride_a, stride_b); break; default: break; @@ -103,12 +103,12 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { AB.resize(stride * batch_size); AB_ref.resize(stride * batch_size); - fp **ab_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **ab_ref_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **ab_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **ab_ref_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); if ((ab_array == NULL) || (ab_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; - oneapi::mkl::free_shared(ab_array, cxt); - oneapi::mkl::free_shared(ab_ref_array, cxt); + oneapi::math::free_shared(ab_array, cxt); + oneapi::math::free_shared(ab_ref_array, cxt); return false; } @@ -117,9 +117,9 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { ab_ref_array[i] = &AB_ref[i * stride]; } - rand_matrix(AB, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride * batch_size, 1, stride * batch_size); - copy_matrix(AB, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + copy_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride * batch_size, 1, stride * batch_size, AB_ref); // Call reference IMATCOPY_BATCH_STRIDE. @@ -136,13 +136,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::imatcopy_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::imatcopy_batch( main_queue, trans, m, n, alpha, &AB[0], lda, ldb, stride, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::imatcopy_batch(main_queue, trans, m, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::imatcopy_batch(main_queue, trans, m, n, alpha, &AB[0], lda, ldb, stride, batch_size, dependencies); break; @@ -151,13 +151,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::imatcopy_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::imatcopy_batch, trans, m, n, alpha, &AB[0], lda, ldb, stride, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::imatcopy_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::imatcopy_batch, trans, m, n, alpha, &AB[0], lda, ldb, stride, batch_size, dependencies); break; @@ -172,9 +172,9 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { - oneapi::mkl::free_shared(ab_array, cxt); - oneapi::mkl::free_shared(ab_ref_array, cxt); + catch (const oneapi::math::unimplemented &e) { + oneapi::math::free_shared(ab_array, cxt); + oneapi::math::free_shared(ab_ref_array, cxt); return test_skipped; } @@ -184,17 +184,17 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { } // Compare the results of reference implementation and DPC++ implementation. - bool good = check_equal_matrix(AB, AB_ref, oneapi::mkl::layout::col_major, stride * batch_size, + bool good = check_equal_matrix(AB, AB_ref, oneapi::math::layout::col_major, stride * batch_size, 1, stride * batch_size, 10, std::cout); - oneapi::mkl::free_shared(ab_array, cxt); - oneapi::mkl::free_shared(ab_ref_array, cxt); + oneapi::math::free_shared(ab_array, cxt); + oneapi::math::free_shared(ab_ref_array, cxt); return (int)good; } class ImatcopyBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(ImatcopyBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -220,8 +220,8 @@ TEST_P(ImatcopyBatchStrideUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(ImatcopyBatchStrideUsmTestSuite, ImatcopyBatchStrideUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp b/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp index a4b8093c9..25dd4cd3f 100644 --- a/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp @@ -49,7 +49,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { +int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -73,8 +73,8 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { auto uaint = usm_allocator(cxt, *dev); vector m(uaint), n(uaint), lda(uaint), ldb(uaint), group_size(uaint); - auto uatranspose = usm_allocator(cxt, *dev); - vector trans(uatranspose); + auto uatranspose = usm_allocator(cxt, *dev); + vector trans(uatranspose); auto uafp = usm_allocator(cxt, *dev); vector alpha(uafp); @@ -112,26 +112,26 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: size_a = lda[i] * n[i]; size_b = - (trans[i] == oneapi::mkl::transpose::nontrans) ? ldb[i] * n[i] : ldb[i] * m[i]; + (trans[i] == oneapi::math::transpose::nontrans) ? ldb[i] * n[i] : ldb[i] * m[i]; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: size_a = lda[i] * m[i]; size_b = - (trans[i] == oneapi::mkl::transpose::nontrans) ? ldb[i] * m[i] : ldb[i] * n[i]; + (trans[i] == oneapi::math::transpose::nontrans) ? ldb[i] * m[i] : ldb[i] * n[i]; break; default: break; } size = std::max(size_a, size_b); for (j = 0; j < group_size[i]; j++) { - ab_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size, *dev, cxt); - ab_ref_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size, *dev, cxt); - rand_matrix(ab_array[idx], oneapi::mkl::layout::col_major, - oneapi::mkl::transpose::nontrans, size, 1, size); - copy_matrix(ab_array[idx], oneapi::mkl::layout::col_major, - oneapi::mkl::transpose::nontrans, size, 1, size, ab_ref_array[idx]); + ab_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size, *dev, cxt); + ab_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size, *dev, cxt); + rand_matrix(ab_array[idx], oneapi::math::layout::col_major, + oneapi::math::transpose::nontrans, size, 1, size); + copy_matrix(ab_array[idx], oneapi::math::layout::col_major, + oneapi::math::transpose::nontrans, size, 1, size, ab_ref_array[idx]); idx++; } } @@ -155,13 +155,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::imatcopy_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::imatcopy_batch( main_queue, trans.data(), m.data(), n.data(), alpha.data(), ab_array.data(), lda.data(), ldb.data(), group_count, group_size.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::imatcopy_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::imatcopy_batch( main_queue, trans.data(), m.data(), n.data(), alpha.data(), ab_array.data(), lda.data(), ldb.data(), group_count, group_size.data(), dependencies); break; @@ -170,14 +170,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::imatcopy_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::imatcopy_batch, trans.data(), m.data(), n.data(), alpha.data(), ab_array.data(), lda.data(), ldb.data(), group_count, group_size.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::imatcopy_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::imatcopy_batch, trans.data(), m.data(), n.data(), alpha.data(), ab_array.data(), lda.data(), ldb.data(), group_count, group_size.data(), dependencies); @@ -193,12 +193,12 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(ab_array[idx], cxt); - oneapi::mkl::free_shared(ab_ref_array[idx], cxt); + oneapi::math::free_shared(ab_array[idx], cxt); + oneapi::math::free_shared(ab_ref_array[idx], cxt); idx++; } } @@ -215,15 +215,15 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: size_a = lda[i] * n[i]; size_b = - (trans[i] == oneapi::mkl::transpose::nontrans) ? ldb[i] * n[i] : ldb[i] * m[i]; + (trans[i] == oneapi::math::transpose::nontrans) ? ldb[i] * n[i] : ldb[i] * m[i]; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: size_a = lda[i] * m[i]; size_b = - (trans[i] == oneapi::mkl::transpose::nontrans) ? ldb[i] * m[i] : ldb[i] * n[i]; + (trans[i] == oneapi::math::transpose::nontrans) ? ldb[i] * m[i] : ldb[i] * n[i]; break; default: break; } @@ -231,7 +231,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { for (j = 0; j < group_size[i]; j++) { good = good && check_equal_matrix(ab_array[idx], ab_ref_array[idx], - oneapi::mkl::layout::col_major, size, 1, size, 10, std::cout); + oneapi::math::layout::col_major, size, 1, size, 10, std::cout); idx++; } } @@ -239,8 +239,8 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(ab_array[idx], cxt); - oneapi::mkl::free_shared(ab_ref_array[idx], cxt); + oneapi::math::free_shared(ab_array[idx], cxt); + oneapi::math::free_shared(ab_ref_array[idx], cxt); idx++; } } @@ -249,7 +249,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { } class ImatcopyBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(ImatcopyBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -275,8 +275,8 @@ TEST_P(ImatcopyBatchUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(ImatcopyBatchUsmTestSuite, ImatcopyBatchUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp b/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp index 4bd5fcce3..ddb37e883 100644 --- a/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp @@ -48,11 +48,11 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n; int64_t lda, ldb, ldc; - oneapi::mkl::transpose transa, transb; + oneapi::math::transpose transa, transb; fp alpha, beta; int64_t i, tmp; @@ -70,14 +70,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { int64_t stride_a, stride_b, stride_c; switch (layout) { - case oneapi::mkl::layout::col_major: - stride_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * n : lda * m; - stride_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + case oneapi::math::layout::col_major: + stride_a = (transa == oneapi::math::transpose::nontrans) ? lda * n : lda * m; + stride_b = (transb == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; stride_c = ldc * n; break; - case oneapi::mkl::layout::row_major: - stride_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * m : lda * n; - stride_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + case oneapi::math::layout::row_major: + stride_a = (transa == oneapi::math::transpose::nontrans) ? lda * m : lda * n; + stride_b = (transb == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; stride_c = ldc * m; break; default: break; @@ -86,13 +86,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { vector> A(stride_a * batch_size), B(stride_b * batch_size), C(stride_c * batch_size), C_ref(stride_c * batch_size); - rand_matrix(A.data(), oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(A.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_a * batch_size, 1, stride_a * batch_size); - rand_matrix(B.data(), oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(B.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_b * batch_size, 1, stride_b * batch_size); - rand_matrix(C.data(), oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(C.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_c * batch_size, 1, stride_c * batch_size); - copy_matrix(C.data(), oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + copy_matrix(C.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_c * batch_size, 1, stride_c * batch_size, C_ref.data()); // Call reference OMATADD_BATCH_STRIDE. @@ -132,13 +132,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::omatadd_batch( + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::omatadd_batch( main_queue, transa, transb, m, n, alpha, A_buffer, lda, stride_a, beta, B_buffer, ldb, stride_b, C_buffer, ldc, stride_c, batch_size); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::omatadd_batch( + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::omatadd_batch( main_queue, transa, transb, m, n, alpha, A_buffer, lda, stride_a, beta, B_buffer, ldb, stride_b, C_buffer, ldc, stride_c, batch_size); break; @@ -146,14 +146,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::omatadd_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatadd_batch, transa, transb, m, n, alpha, A_buffer, lda, stride_a, beta, B_buffer, ldb, stride_b, C_buffer, ldc, stride_c, batch_size); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::omatadd_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatadd_batch, transa, transb, m, n, alpha, A_buffer, lda, stride_a, beta, B_buffer, ldb, stride_b, C_buffer, ldc, stride_c, batch_size); @@ -168,7 +168,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -180,14 +180,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { // Compare the results of reference implementation and DPC++ implementation. auto C_accessor = C_buffer.get_host_access(read_only); - bool good = check_equal_matrix(C_accessor, C_ref, oneapi::mkl::layout::col_major, + bool good = check_equal_matrix(C_accessor, C_ref, oneapi::math::layout::col_major, stride_c * batch_size, 1, stride_c * batch_size, 10, std::cout); return (int)good; } class OmataddBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmataddBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -213,8 +213,8 @@ TEST_P(OmataddBatchStrideTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(OmataddBatchStrideTestSuite, OmataddBatchStrideTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp index de6267978..b01c2ec10 100644 --- a/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -71,7 +71,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n; int64_t lda, ldb, ldc; - oneapi::mkl::transpose transa, transb; + oneapi::math::transpose transa, transb; fp alpha, beta; int64_t i, tmp; @@ -89,14 +89,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { int64_t stride_a, stride_b, stride_c; switch (layout) { - case oneapi::mkl::layout::col_major: - stride_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * n : lda * m; - stride_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + case oneapi::math::layout::col_major: + stride_a = (transa == oneapi::math::transpose::nontrans) ? lda * n : lda * m; + stride_b = (transb == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; stride_c = ldc * n; break; - case oneapi::mkl::layout::row_major: - stride_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * m : lda * n; - stride_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + case oneapi::math::layout::row_major: + stride_a = (transa == oneapi::math::transpose::nontrans) ? lda * m : lda * n; + stride_b = (transb == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; stride_c = ldc * m; break; default: break; @@ -110,17 +110,17 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { C.resize(stride_c * batch_size); C_ref.resize(stride_c * batch_size); - fp **a_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **b_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **c_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **c_ref_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **a_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **b_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **c_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **c_ref_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); if ((a_array == NULL) || (b_array == NULL) || (c_array == NULL) || (c_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(b_array, cxt); - oneapi::mkl::free_shared(c_array, cxt); - oneapi::mkl::free_shared(c_ref_array, cxt); + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(b_array, cxt); + oneapi::math::free_shared(c_array, cxt); + oneapi::math::free_shared(c_ref_array, cxt); return false; } @@ -131,13 +131,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { c_ref_array[i] = &C_ref[i * stride_c]; } - rand_matrix(A, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(A, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_a * batch_size, 1, stride_a * batch_size); - rand_matrix(B, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(B, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_b * batch_size, 1, stride_b * batch_size); - rand_matrix(C, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(C, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_c * batch_size, 1, stride_c * batch_size); - copy_matrix(C, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + copy_matrix(C, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_c * batch_size, 1, stride_c * batch_size, C_ref); // Call reference OMATADD_BATCH_STRIDE. @@ -156,13 +156,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::omatadd_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::omatadd_batch( main_queue, transa, transb, m, n, alpha, &A[0], lda, stride_a, beta, &B[0], ldb, stride_b, &C[0], ldc, stride_c, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::omatadd_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::omatadd_batch( main_queue, transa, transb, m, n, alpha, &A[0], lda, stride_a, beta, &B[0], ldb, stride_b, &C[0], ldc, stride_c, batch_size, dependencies); break; @@ -171,14 +171,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::omatadd_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatadd_batch, transa, transb, m, n, alpha, &A[0], lda, stride_a, beta, &B[0], ldb, stride_b, &C[0], ldc, stride_c, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::omatadd_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatadd_batch, transa, transb, m, n, alpha, &A[0], lda, stride_a, beta, &B[0], ldb, stride_b, &C[0], ldc, stride_c, batch_size, dependencies); @@ -194,11 +194,11 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(b_array, cxt); - oneapi::mkl::free_shared(c_array, cxt); - oneapi::mkl::free_shared(c_ref_array, cxt); + catch (const oneapi::math::unimplemented &e) { + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(b_array, cxt); + oneapi::math::free_shared(c_array, cxt); + oneapi::math::free_shared(c_ref_array, cxt); return test_skipped; } @@ -208,19 +208,19 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { } // Compare the results of reference implementation and DPC++ implementation. - bool good = check_equal_matrix(C, C_ref, oneapi::mkl::layout::col_major, stride_c * batch_size, + bool good = check_equal_matrix(C, C_ref, oneapi::math::layout::col_major, stride_c * batch_size, 1, stride_c * batch_size, 10, std::cout); - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(b_array, cxt); - oneapi::mkl::free_shared(c_array, cxt); - oneapi::mkl::free_shared(c_ref_array, cxt); + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(b_array, cxt); + oneapi::math::free_shared(c_array, cxt); + oneapi::math::free_shared(c_ref_array, cxt); return (int)good; } class OmataddBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmataddBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -246,8 +246,8 @@ TEST_P(OmataddBatchStrideUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(OmataddBatchStrideUsmTestSuite, OmataddBatchStrideUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp b/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp index bd024bbc1..20320d51d 100644 --- a/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp @@ -48,11 +48,11 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n; int64_t lda, ldb; - oneapi::mkl::transpose trans; + oneapi::math::transpose trans; fp alpha; int64_t i, tmp; @@ -67,13 +67,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { int64_t stride_a, stride_b; switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: stride_a = lda * n; - stride_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + stride_b = (trans == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: stride_a = lda * m; - stride_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + stride_b = (trans == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; break; default: break; } @@ -82,7 +82,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { B_ref(stride_b * batch_size); for (i = 0; i < batch_size; i++) { - rand_matrix(A.data() + stride_a * i, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A.data() + stride_a * i, layout, oneapi::math::transpose::nontrans, m, n, lda); rand_matrix(B.data() + stride_b * i, layout, trans, m, n, ldb); } @@ -121,13 +121,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::omatcopy_batch(main_queue, trans, m, n, alpha, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::omatcopy_batch(main_queue, trans, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, batch_size); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::omatcopy_batch(main_queue, trans, m, n, alpha, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::omatcopy_batch(main_queue, trans, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, batch_size); break; @@ -135,13 +135,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::omatcopy_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatcopy_batch, trans, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, batch_size); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::omatcopy_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatcopy_batch, trans, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, batch_size); break; @@ -155,7 +155,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -167,14 +167,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { // Compare the results of reference implementation and DPC++ implementation. auto B_accessor = B_buffer.get_host_access(read_only); - bool good = check_equal_matrix(B_accessor, B_ref, oneapi::mkl::layout::col_major, + bool good = check_equal_matrix(B_accessor, B_ref, oneapi::math::layout::col_major, stride_b * batch_size, 1, stride_b * batch_size, 10, std::cout); return (int)good; } class OmatcopyBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmatcopyBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -200,8 +200,8 @@ TEST_P(OmatcopyBatchStrideTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(OmatcopyBatchStrideTestSuite, OmatcopyBatchStrideTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp index 2fc0a32a0..72b7c8f1b 100644 --- a/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp @@ -49,7 +49,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -72,7 +72,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n; int64_t lda, ldb; - oneapi::mkl::transpose trans; + oneapi::math::transpose trans; fp alpha; int64_t i, tmp; @@ -87,13 +87,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { int64_t stride_a, stride_b; switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: stride_a = lda * n; - stride_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + stride_b = (trans == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: stride_a = lda * m; - stride_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + stride_b = (trans == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; break; default: break; } @@ -105,15 +105,15 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { B.resize(stride_b * batch_size); B_ref.resize(stride_b * batch_size); - fp **a_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **b_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **b_ref_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **a_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **b_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **b_ref_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); if ((a_array == NULL) || (b_array == NULL) || (b_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(b_array, cxt); - oneapi::mkl::free_shared(b_ref_array, cxt); + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(b_array, cxt); + oneapi::math::free_shared(b_ref_array, cxt); return false; } @@ -123,11 +123,11 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { b_ref_array[i] = &B_ref[i * stride_b]; } - rand_matrix(A, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(A, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_a * batch_size, 1, stride_a * batch_size); - rand_matrix(B, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(B, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_b * batch_size, 1, stride_b * batch_size); - copy_matrix(B, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + copy_matrix(B, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_b * batch_size, 1, stride_b * batch_size, B_ref); // Call reference OMATCOPY_BATCH_STRIDE. @@ -145,13 +145,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::omatcopy_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::omatcopy_batch( main_queue, trans, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::omatcopy_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::omatcopy_batch( main_queue, trans, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, batch_size, dependencies); break; @@ -160,13 +160,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::omatcopy_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatcopy_batch, trans, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::omatcopy_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatcopy_batch, trans, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, batch_size, dependencies); break; @@ -181,10 +181,10 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(b_array, cxt); - oneapi::mkl::free_shared(b_ref_array, cxt); + catch (const oneapi::math::unimplemented &e) { + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(b_array, cxt); + oneapi::math::free_shared(b_ref_array, cxt); return test_skipped; } @@ -194,18 +194,18 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { } // Compare the results of reference implementation and DPC++ implementation. - bool good = check_equal_matrix(B, B_ref, oneapi::mkl::layout::col_major, stride_b * batch_size, + bool good = check_equal_matrix(B, B_ref, oneapi::math::layout::col_major, stride_b * batch_size, 1, stride_b * batch_size, 10, std::cout); - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(b_array, cxt); - oneapi::mkl::free_shared(b_ref_array, cxt); + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(b_array, cxt); + oneapi::math::free_shared(b_ref_array, cxt); return (int)good; } class OmatcopyBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmatcopyBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -231,8 +231,8 @@ TEST_P(OmatcopyBatchStrideUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(OmatcopyBatchStrideUsmTestSuite, OmatcopyBatchStrideUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp b/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp index e1c0b8a3c..190c38afe 100644 --- a/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp @@ -49,7 +49,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { +int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -73,8 +73,8 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { auto uaint = usm_allocator(cxt, *dev); vector m(uaint), n(uaint), lda(uaint), ldb(uaint), group_size(uaint); - auto uatranspose = usm_allocator(cxt, *dev); - vector trans(uatranspose); + auto uatranspose = usm_allocator(cxt, *dev); + vector trans(uatranspose); auto uafp = usm_allocator(cxt, *dev); vector alpha(uafp); @@ -113,28 +113,28 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: size_a = lda[i] * n[i]; size_b = - (trans[i] == oneapi::mkl::transpose::nontrans) ? ldb[i] * n[i] : ldb[i] * m[i]; + (trans[i] == oneapi::math::transpose::nontrans) ? ldb[i] * n[i] : ldb[i] * m[i]; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: size_a = lda[i] * m[i]; size_b = - (trans[i] == oneapi::mkl::transpose::nontrans) ? ldb[i] * m[i] : ldb[i] * n[i]; + (trans[i] == oneapi::math::transpose::nontrans) ? ldb[i] * m[i] : ldb[i] * n[i]; break; default: break; } for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); - b_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); - b_ref_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); - rand_matrix(a_array[idx], oneapi::mkl::layout::col_major, - oneapi::mkl::transpose::nontrans, size_a, 1, size_a); - rand_matrix(b_array[idx], oneapi::mkl::layout::col_major, - oneapi::mkl::transpose::nontrans, size_b, 1, size_b); - copy_matrix(b_array[idx], oneapi::mkl::layout::col_major, - oneapi::mkl::transpose::nontrans, size_b, 1, size_b, b_ref_array[idx]); + a_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); + b_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); + b_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); + rand_matrix(a_array[idx], oneapi::math::layout::col_major, + oneapi::math::transpose::nontrans, size_a, 1, size_a); + rand_matrix(b_array[idx], oneapi::math::layout::col_major, + oneapi::math::transpose::nontrans, size_b, 1, size_b); + copy_matrix(b_array[idx], oneapi::math::layout::col_major, + oneapi::math::transpose::nontrans, size_b, 1, size_b, b_ref_array[idx]); idx++; } } @@ -158,14 +158,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::omatcopy_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::omatcopy_batch( main_queue, trans.data(), m.data(), n.data(), alpha.data(), (const fp **)a_array.data(), lda.data(), b_array.data(), ldb.data(), group_count, group_size.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::omatcopy_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::omatcopy_batch( main_queue, trans.data(), m.data(), n.data(), alpha.data(), (const fp **)a_array.data(), lda.data(), b_array.data(), ldb.data(), group_count, group_size.data(), dependencies); @@ -175,14 +175,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::omatcopy_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatcopy_batch, trans.data(), m.data(), n.data(), alpha.data(), (const fp **)a_array.data(), lda.data(), b_array.data(), ldb.data(), group_count, group_size.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::omatcopy_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatcopy_batch, trans.data(), m.data(), n.data(), alpha.data(), (const fp **)a_array.data(), lda.data(), b_array.data(), ldb.data(), group_count, group_size.data(), dependencies); @@ -198,13 +198,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(b_array[idx], cxt); - oneapi::mkl::free_shared(b_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(b_array[idx], cxt); + oneapi::math::free_shared(b_ref_array[idx], cxt); idx++; } } @@ -221,21 +221,21 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: size_a = lda[i] * n[i]; size_b = - (trans[i] == oneapi::mkl::transpose::nontrans) ? ldb[i] * n[i] : ldb[i] * m[i]; + (trans[i] == oneapi::math::transpose::nontrans) ? ldb[i] * n[i] : ldb[i] * m[i]; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: size_a = lda[i] * m[i]; size_b = - (trans[i] == oneapi::mkl::transpose::nontrans) ? ldb[i] * m[i] : ldb[i] * n[i]; + (trans[i] == oneapi::math::transpose::nontrans) ? ldb[i] * m[i] : ldb[i] * n[i]; break; default: break; } for (j = 0; j < group_size[i]; j++) { good = good && check_equal_matrix(b_array[idx], b_ref_array[idx], - oneapi::mkl::layout::col_major, size_b, 1, size_b, 10, + oneapi::math::layout::col_major, size_b, 1, size_b, 10, std::cout); idx++; } @@ -244,9 +244,9 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(b_array[idx], cxt); - oneapi::mkl::free_shared(b_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(b_array[idx], cxt); + oneapi::math::free_shared(b_ref_array[idx], cxt); idx++; } } @@ -255,7 +255,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { } class OmatcopyBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmatcopyBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -281,8 +281,8 @@ TEST_P(OmatcopyBatchUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(OmatcopyBatchUsmTestSuite, OmatcopyBatchUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/syrk_batch_stride.cpp b/tests/unit_tests/blas/batch/syrk_batch_stride.cpp index 4b7b8d56d..3599d36f8 100644 --- a/tests/unit_tests/blas/batch/syrk_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/syrk_batch_stride.cpp @@ -48,12 +48,12 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Prepare data. int64_t n, k; int64_t lda, ldc; - oneapi::mkl::uplo upper_lower; - oneapi::mkl::transpose trans; + oneapi::math::uplo upper_lower; + oneapi::math::transpose trans; fp alpha, beta; int64_t i, tmp; @@ -65,26 +65,26 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { alpha = rand_scalar(); beta = rand_scalar(); - upper_lower = (oneapi::mkl::uplo)(std::rand() % 2); + upper_lower = (oneapi::math::uplo)(std::rand() % 2); if ((std::is_same::value) || (std::is_same::value)) { - trans = (std::rand() % 2) == 0 ? oneapi::mkl::transpose::nontrans - : (std::rand() % 2) == 0 ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::conjtrans; + trans = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans + : (std::rand() % 2) == 0 ? oneapi::math::transpose::trans + : oneapi::math::transpose::conjtrans; } else { - trans = (std::rand() % 2) == 0 ? oneapi::mkl::transpose::nontrans - : oneapi::mkl::transpose::trans; + trans = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans + : oneapi::math::transpose::trans; } int64_t stride_a, stride_c; switch (layout) { - case oneapi::mkl::layout::col_major: - stride_a = (trans == oneapi::mkl::transpose::nontrans) ? lda * k : lda * n; + case oneapi::math::layout::col_major: + stride_a = (trans == oneapi::math::transpose::nontrans) ? lda * k : lda * n; stride_c = ldc * n; break; - case oneapi::mkl::layout::row_major: - stride_a = (trans == oneapi::mkl::transpose::nontrans) ? lda * n : lda * k; + case oneapi::math::layout::row_major: + stride_a = (trans == oneapi::math::transpose::nontrans) ? lda * n : lda * k; stride_c = ldc * n; break; default: break; @@ -95,7 +95,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { for (i = 0; i < batch_size; i++) { rand_matrix(A.data() + stride_a * i, layout, trans, n, k, lda); - rand_matrix(C.data() + stride_c * i, layout, oneapi::mkl::transpose::nontrans, n, n, ldc); + rand_matrix(C.data() + stride_c * i, layout, oneapi::math::transpose::nontrans, n, n, ldc); } C_ref = C; @@ -140,13 +140,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::syrk_batch(main_queue, upper_lower, trans, n, k, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::syrk_batch(main_queue, upper_lower, trans, n, k, alpha, A_buffer, lda, stride_a, beta, C_buffer, ldc, stride_c, batch_size); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::syrk_batch(main_queue, upper_lower, trans, n, k, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::syrk_batch(main_queue, upper_lower, trans, n, k, alpha, A_buffer, lda, stride_a, beta, C_buffer, ldc, stride_c, batch_size); break; @@ -154,13 +154,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::syrk_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syrk_batch, upper_lower, trans, n, k, alpha, A_buffer, lda, stride_a, beta, C_buffer, ldc, stride_c, batch_size); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::syrk_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syrk_batch, upper_lower, trans, n, k, alpha, A_buffer, lda, stride_a, beta, C_buffer, ldc, stride_c, batch_size); break; @@ -174,7 +174,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -187,14 +187,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { auto C_accessor = C_buffer.get_host_access(read_only); bool good = - check_equal_matrix(C_accessor, C_ref, oneapi::mkl::layout::col_major, stride_c * batch_size, + check_equal_matrix(C_accessor, C_ref, oneapi::math::layout::col_major, stride_c * batch_size, 1, stride_c * batch_size, 10 * k, std::cout); return (int)good; } class SyrkBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SyrkBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -220,8 +220,8 @@ TEST_P(SyrkBatchStrideTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(SyrkBatchStrideTestSuite, SyrkBatchStrideTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp index bd11b1765..06aea5c4f 100644 --- a/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { +int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -71,8 +71,8 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { // Prepare data. int64_t n, k; int64_t lda, ldc; - oneapi::mkl::uplo upper_lower; - oneapi::mkl::transpose trans; + oneapi::math::uplo upper_lower; + oneapi::math::transpose trans; fp alpha, beta; int64_t i, tmp; @@ -84,26 +84,26 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { ldc = std::max(n, n); alpha = rand_scalar(); beta = rand_scalar(); - upper_lower = (oneapi::mkl::uplo)(std::rand() % 2); + upper_lower = (oneapi::math::uplo)(std::rand() % 2); if ((std::is_same::value) || (std::is_same::value)) { - trans = (std::rand() % 2) == 0 ? oneapi::mkl::transpose::nontrans - : (std::rand() % 2) == 0 ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::conjtrans; + trans = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans + : (std::rand() % 2) == 0 ? oneapi::math::transpose::trans + : oneapi::math::transpose::conjtrans; } else { - trans = (std::rand() % 2) == 0 ? oneapi::mkl::transpose::nontrans - : oneapi::mkl::transpose::trans; + trans = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans + : oneapi::math::transpose::trans; } int64_t stride_a, stride_c; switch (layout) { - case oneapi::mkl::layout::col_major: - stride_a = (trans == oneapi::mkl::transpose::nontrans) ? lda * k : lda * n; + case oneapi::math::layout::col_major: + stride_a = (trans == oneapi::math::transpose::nontrans) ? lda * k : lda * n; stride_c = ldc * n; break; - case oneapi::mkl::layout::row_major: - stride_a = (trans == oneapi::mkl::transpose::nontrans) ? lda * n : lda * k; + case oneapi::math::layout::row_major: + stride_a = (trans == oneapi::math::transpose::nontrans) ? lda * n : lda * k; stride_c = ldc * n; break; default: break; @@ -116,15 +116,15 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { C.resize(stride_c * batch_size); C_ref.resize(stride_c * batch_size); - fp **a_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **c_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **c_ref_array = (fp **)oneapi::mkl::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **a_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **c_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp **c_ref_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); if ((a_array == NULL) || (c_array == NULL) || (c_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(c_array, cxt); - oneapi::mkl::free_shared(c_ref_array, cxt); + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(c_array, cxt); + oneapi::math::free_shared(c_ref_array, cxt); return false; } @@ -134,11 +134,11 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { c_ref_array[i] = &C_ref[i * stride_c]; } - rand_matrix(A, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(A, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_a * batch_size, 1, stride_a * batch_size); - rand_matrix(C, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + rand_matrix(C, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_c * batch_size, 1, stride_c * batch_size); - copy_matrix(C, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, + copy_matrix(C, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, stride_c * batch_size, 1, stride_c * batch_size, C_ref); // Call reference SYRK_BATCH_STRIDE. @@ -161,13 +161,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::syrk_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::syrk_batch( main_queue, upper_lower, trans, n, k, alpha, &A[0], lda, stride_a, beta, &C[0], ldc, stride_c, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::syrk_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::syrk_batch( main_queue, upper_lower, trans, n, k, alpha, &A[0], lda, stride_a, beta, &C[0], ldc, stride_c, batch_size, dependencies); break; @@ -176,13 +176,13 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::syrk_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syrk_batch, upper_lower, trans, n, k, alpha, &A[0], lda, stride_a, beta, &C[0], ldc, stride_c, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::syrk_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syrk_batch, upper_lower, trans, n, k, alpha, &A[0], lda, stride_a, beta, &C[0], ldc, stride_c, batch_size, dependencies); break; @@ -197,10 +197,10 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(c_array, cxt); - oneapi::mkl::free_shared(c_ref_array, cxt); + catch (const oneapi::math::unimplemented &e) { + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(c_array, cxt); + oneapi::math::free_shared(c_ref_array, cxt); return test_skipped; } @@ -210,18 +210,18 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t batch_size) { } // Compare the results of reference implementation and DPC++ implementation. - bool good = check_equal_matrix(C, C_ref, oneapi::mkl::layout::col_major, stride_c * batch_size, + bool good = check_equal_matrix(C, C_ref, oneapi::math::layout::col_major, stride_c * batch_size, 1, stride_c * batch_size, 10 * k, std::cout); - oneapi::mkl::free_shared(a_array, cxt); - oneapi::mkl::free_shared(c_array, cxt); - oneapi::mkl::free_shared(c_ref_array, cxt); + oneapi::math::free_shared(a_array, cxt); + oneapi::math::free_shared(c_array, cxt); + oneapi::math::free_shared(c_ref_array, cxt); return (int)good; } class SyrkBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SyrkBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -247,8 +247,8 @@ TEST_P(SyrkBatchStrideUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(SyrkBatchStrideUsmTestSuite, SyrkBatchStrideUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/syrk_batch_usm.cpp b/tests/unit_tests/blas/batch/syrk_batch_usm.cpp index d308f98ca..bf7358ca9 100644 --- a/tests/unit_tests/blas/batch/syrk_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/syrk_batch_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { +int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -72,11 +72,11 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { auto uaint = usm_allocator(cxt, *dev); vector n(uaint), k(uaint), lda(uaint), ldc(uaint), group_size(uaint); - auto uauplo = usm_allocator(cxt, *dev); - vector upper_lower(uauplo); + auto uauplo = usm_allocator(cxt, *dev); + vector upper_lower(uauplo); - auto uatranspose = usm_allocator(cxt, *dev); - vector trans(uatranspose); + auto uatranspose = usm_allocator(cxt, *dev); + vector trans(uatranspose); auto uafp = usm_allocator(cxt, *dev); vector alpha(uafp), beta(uafp); @@ -104,16 +104,16 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { ldc[i] = std::max(n[i], n[i]); alpha[i] = rand_scalar(); beta[i] = rand_scalar(); - upper_lower[i] = (oneapi::mkl::uplo)(std::rand() % 2); + upper_lower[i] = (oneapi::math::uplo)(std::rand() % 2); if ((std::is_same::value) || (std::is_same::value)) { trans[i] = (std::rand() % 2) == 0 - ? oneapi::mkl::transpose::nontrans - : (std::rand() % 2) == 0 ? oneapi::mkl::transpose::trans - : oneapi::mkl::transpose::conjtrans; + ? oneapi::math::transpose::nontrans + : (std::rand() % 2) == 0 ? oneapi::math::transpose::trans + : oneapi::math::transpose::conjtrans; } else { - trans[i] = (std::rand() % 2) == 0 ? oneapi::mkl::transpose::nontrans - : oneapi::mkl::transpose::trans; + trans[i] = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans + : oneapi::math::transpose::trans; } total_batch_count += group_size[i]; } @@ -127,23 +127,23 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { switch (layout) { - case oneapi::mkl::layout::col_major: - size_a = lda[i] * ((trans[i] == oneapi::mkl::transpose::nontrans) ? k[i] : n[i]); + case oneapi::math::layout::col_major: + size_a = lda[i] * ((trans[i] == oneapi::math::transpose::nontrans) ? k[i] : n[i]); size_c = ldc[i] * n[i]; break; - case oneapi::mkl::layout::row_major: - size_a = lda[i] * ((trans[i] == oneapi::mkl::transpose::nontrans) ? n[i] : k[i]); + case oneapi::math::layout::row_major: + size_a = lda[i] * ((trans[i] == oneapi::math::transpose::nontrans) ? n[i] : k[i]); size_c = ldc[i] * n[i]; break; default: break; } for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); - c_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); - c_ref_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); + a_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); + c_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); + c_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); rand_matrix(a_array[idx], layout, trans[i], n[i], k[i], lda[i]); - rand_matrix(c_array[idx], layout, oneapi::mkl::transpose::nontrans, n[i], n[i], ldc[i]); - copy_matrix(c_array[idx], layout, oneapi::mkl::transpose::nontrans, n[i], n[i], ldc[i], + rand_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, n[i], n[i], ldc[i]); + copy_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, n[i], n[i], ldc[i], c_ref_array[idx]); idx++; } @@ -151,33 +151,33 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { // Call reference SYRK_BATCH. using fp_ref = typename ref_type_info::type; - int *n_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *k_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *lda_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *ldc_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *group_size_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); + int *n_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *k_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *lda_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *ldc_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *group_size_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); CBLAS_UPLO *upper_lower_ref = - (CBLAS_UPLO *)oneapi::mkl::aligned_alloc(64, sizeof(CBLAS_UPLO) * group_count); + (CBLAS_UPLO *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_UPLO) * group_count); CBLAS_TRANSPOSE *trans_ref = - (CBLAS_TRANSPOSE *)oneapi::mkl::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); + (CBLAS_TRANSPOSE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); if ((n_ref == NULL) || (k_ref == NULL) || (lda_ref == NULL) || (ldc_ref == NULL) || (trans_ref == NULL) || (upper_lower_ref == NULL) || (group_size_ref == NULL)) { std::cout << "Error cannot allocate input arrays\n"; - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(k_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(ldc_ref); - oneapi::mkl::aligned_free(trans_ref); - oneapi::mkl::aligned_free(upper_lower_ref); - oneapi::mkl::aligned_free(group_size_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(k_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(ldc_ref); + oneapi::math::aligned_free(trans_ref); + oneapi::math::aligned_free(upper_lower_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(c_array[idx], cxt); - oneapi::mkl::free_shared(c_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(c_array[idx], cxt); + oneapi::math::free_shared(c_ref_array[idx], cxt); idx++; } } @@ -206,14 +206,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::syrk_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::syrk_batch( main_queue, &upper_lower[0], &trans[0], &n[0], &k[0], &alpha[0], (const fp **)&a_array[0], &lda[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::syrk_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::syrk_batch( main_queue, &upper_lower[0], &trans[0], &n[0], &k[0], &alpha[0], (const fp **)&a_array[0], &lda[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); @@ -223,14 +223,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::syrk_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syrk_batch, &upper_lower[0], &trans[0], &n[0], &k[0], &alpha[0], (const fp **)&a_array[0], &lda[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::syrk_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syrk_batch, &upper_lower[0], &trans[0], &n[0], &k[0], &alpha[0], (const fp **)&a_array[0], &lda[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); @@ -246,20 +246,20 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(k_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(ldc_ref); - oneapi::mkl::aligned_free(upper_lower_ref); - oneapi::mkl::aligned_free(trans_ref); - oneapi::mkl::aligned_free(group_size_ref); + catch (const oneapi::math::unimplemented &e) { + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(k_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(ldc_ref); + oneapi::math::aligned_free(upper_lower_ref); + oneapi::math::aligned_free(trans_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(c_array[idx], cxt); - oneapi::mkl::free_shared(c_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(c_array[idx], cxt); + oneapi::math::free_shared(c_ref_array[idx], cxt); idx++; } } @@ -280,19 +280,19 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx++; } } - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(k_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(ldc_ref); - oneapi::mkl::aligned_free(upper_lower_ref); - oneapi::mkl::aligned_free(trans_ref); - oneapi::mkl::aligned_free(group_size_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(k_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(ldc_ref); + oneapi::math::aligned_free(upper_lower_ref); + oneapi::math::aligned_free(trans_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(c_array[idx], cxt); - oneapi::mkl::free_shared(c_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(c_array[idx], cxt); + oneapi::math::free_shared(c_ref_array[idx], cxt); idx++; } } @@ -301,7 +301,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { } class SyrkBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SyrkBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -327,8 +327,8 @@ TEST_P(SyrkBatchUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(SyrkBatchUsmTestSuite, SyrkBatchUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/trsm_batch_stride.cpp b/tests/unit_tests/blas/batch/trsm_batch_stride.cpp index fcafbf4d6..5ae26282d 100644 --- a/tests/unit_tests/blas/batch/trsm_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/trsm_batch_stride.cpp @@ -48,14 +48,14 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb; - oneapi::mkl::transpose trans; - oneapi::mkl::side left_right; - oneapi::mkl::uplo upper_lower; - oneapi::mkl::diag unit_nonunit; + oneapi::math::transpose trans; + oneapi::math::side left_right; + oneapi::math::uplo upper_lower; + oneapi::math::diag unit_nonunit; fp alpha; int64_t batch_size; int64_t i, tmp; @@ -68,26 +68,26 @@ int test(device *dev, oneapi::mkl::layout layout) { alpha = rand_scalar(); if ((std::is_same::value) || (std::is_same::value)) { - trans = (oneapi::mkl::transpose)(std::rand() % 2); + trans = (oneapi::math::transpose)(std::rand() % 2); } else { tmp = std::rand() % 3; if (tmp == 2) - trans = oneapi::mkl::transpose::conjtrans; + trans = oneapi::math::transpose::conjtrans; else - trans = (oneapi::mkl::transpose)tmp; + trans = (oneapi::math::transpose)tmp; } - left_right = (oneapi::mkl::side)(std::rand() % 2); - upper_lower = (oneapi::mkl::uplo)(std::rand() % 2); - unit_nonunit = (oneapi::mkl::diag)(std::rand() % 2); + left_right = (oneapi::math::side)(std::rand() % 2); + upper_lower = (oneapi::math::uplo)(std::rand() % 2); + unit_nonunit = (oneapi::math::diag)(std::rand() % 2); int64_t stride_a, stride_b; int64_t total_size_b; - stride_a = (left_right == oneapi::mkl::side::left) ? lda * m : lda * n; + stride_a = (left_right == oneapi::math::side::left) ? lda * m : lda * n; switch (layout) { - case oneapi::mkl::layout::col_major: stride_b = ldb * n; break; - case oneapi::mkl::layout::row_major: stride_b = ldb * m; break; + case oneapi::math::layout::col_major: stride_b = ldb * n; break; + case oneapi::math::layout::row_major: stride_b = ldb * m; break; default: break; } total_size_b = batch_size * stride_b; @@ -96,11 +96,11 @@ int test(device *dev, oneapi::mkl::layout layout) { B_ref(total_size_b); for (i = 0; i < batch_size; i++) { - if (left_right == oneapi::mkl::side::left) + if (left_right == oneapi::math::side::left) rand_trsm_matrix(A.data() + stride_a * i, layout, trans, m, m, lda); else rand_trsm_matrix(A.data() + stride_a * i, layout, trans, n, n, lda); - rand_matrix(B.data() + stride_b * i, layout, oneapi::mkl::transpose::nontrans, m, n, ldb); + rand_matrix(B.data() + stride_b * i, layout, oneapi::math::transpose::nontrans, m, n, ldb); } B_ref = B; @@ -146,13 +146,13 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::trsm_batch( + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::trsm_batch( main_queue, left_right, upper_lower, trans, unit_nonunit, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, batch_size); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::trsm_batch( + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::trsm_batch( main_queue, left_right, upper_lower, trans, unit_nonunit, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, batch_size); break; @@ -160,14 +160,14 @@ int test(device *dev, oneapi::mkl::layout layout) { } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::trsm_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trsm_batch, left_right, upper_lower, trans, unit_nonunit, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, batch_size); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::trsm_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsm_batch, left_right, upper_lower, trans, unit_nonunit, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, batch_size); @@ -182,7 +182,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -194,14 +194,14 @@ int test(device *dev, oneapi::mkl::layout layout) { // Compare the results of reference implementation and DPC++ implementation. auto B_accessor = B_buffer.get_host_access(read_only); bool good = - check_equal_trsm_matrix(B_accessor, B_ref, oneapi::mkl::layout::col_major, total_size_b, 1, + check_equal_trsm_matrix(B_accessor, B_ref, oneapi::math::layout::col_major, total_size_b, 1, total_size_b, 10 * std::max(m, n), std::cout); return (int)good; } class TrsmBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TrsmBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); @@ -225,8 +225,8 @@ TEST_P(TrsmBatchStrideTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(TrsmBatchStrideTestSuite, TrsmBatchStrideTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp index 4dde42ef7..ec0f17572 100644 --- a/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -71,10 +71,10 @@ int test(device *dev, oneapi::mkl::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb; - oneapi::mkl::transpose trans; - oneapi::mkl::side left_right; - oneapi::mkl::uplo upper_lower; - oneapi::mkl::diag unit_nonunit; + oneapi::math::transpose trans; + oneapi::math::side left_right; + oneapi::math::uplo upper_lower; + oneapi::math::diag unit_nonunit; fp alpha; int64_t batch_size; int64_t i, tmp; @@ -87,24 +87,24 @@ int test(device *dev, oneapi::mkl::layout layout) { alpha = rand_scalar(); if ((std::is_same::value) || (std::is_same::value)) { - trans = (oneapi::mkl::transpose)(std::rand() % 2); + trans = (oneapi::math::transpose)(std::rand() % 2); } else { tmp = std::rand() % 3; if (tmp == 2) - trans = oneapi::mkl::transpose::conjtrans; + trans = oneapi::math::transpose::conjtrans; else - trans = (oneapi::mkl::transpose)tmp; + trans = (oneapi::math::transpose)tmp; } - left_right = (oneapi::mkl::side)(std::rand() % 2); - upper_lower = (oneapi::mkl::uplo)(std::rand() % 2); - unit_nonunit = (oneapi::mkl::diag)(std::rand() % 2); + left_right = (oneapi::math::side)(std::rand() % 2); + upper_lower = (oneapi::math::uplo)(std::rand() % 2); + unit_nonunit = (oneapi::math::diag)(std::rand() % 2); int64_t stride_a, stride_b; int64_t total_size_b; - stride_a = (left_right == oneapi::mkl::side::left) ? lda * m : lda * n; - stride_b = (layout == oneapi::mkl::layout::col_major) ? ldb * n : ldb * m; + stride_a = (left_right == oneapi::math::side::left) ? lda * m : lda * n; + stride_b = (layout == oneapi::math::layout::col_major) ? ldb * n : ldb * m; total_size_b = batch_size * stride_b; @@ -116,14 +116,14 @@ int test(device *dev, oneapi::mkl::layout layout) { B_ref.resize(total_size_b); for (i = 0; i < batch_size; i++) { - if (left_right == oneapi::mkl::side::left) + if (left_right == oneapi::math::side::left) rand_trsm_matrix(&A[stride_a * i], layout, trans, m, m, lda); else rand_trsm_matrix(&A[stride_a * i], layout, trans, n, n, lda); - rand_matrix(&B[stride_b * i], layout, oneapi::mkl::transpose::nontrans, m, n, ldb); + rand_matrix(&B[stride_b * i], layout, oneapi::math::transpose::nontrans, m, n, ldb); } - copy_matrix(B, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, total_size_b, + copy_matrix(B, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, total_size_b, 1, total_size_b, B_ref); // Call reference TRSM_BATCH_STRIDE. @@ -148,13 +148,13 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::trsm_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::trsm_batch( main_queue, left_right, upper_lower, trans, unit_nonunit, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::trsm_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::trsm_batch( main_queue, left_right, upper_lower, trans, unit_nonunit, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, batch_size, dependencies); break; @@ -163,14 +163,14 @@ int test(device *dev, oneapi::mkl::layout layout) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::trsm_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trsm_batch, left_right, upper_lower, trans, unit_nonunit, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, batch_size, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::trsm_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsm_batch, left_right, upper_lower, trans, unit_nonunit, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, batch_size, dependencies); @@ -186,7 +186,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -196,14 +196,14 @@ int test(device *dev, oneapi::mkl::layout layout) { } // Compare the results of reference implementation and DPC++ implementation. - bool good = check_equal_trsm_matrix(B, B_ref, oneapi::mkl::layout::col_major, total_size_b, 1, + bool good = check_equal_trsm_matrix(B, B_ref, oneapi::math::layout::col_major, total_size_b, 1, total_size_b, 10 * std::max(m, n), std::cout); return (int)good; } class TrsmBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TrsmBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); @@ -227,8 +227,8 @@ TEST_P(TrsmBatchStrideUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(TrsmBatchStrideUsmTestSuite, TrsmBatchStrideUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/batch/trsm_batch_usm.cpp b/tests/unit_tests/blas/batch/trsm_batch_usm.cpp index 1e3d2980d..8caeb9f14 100644 --- a/tests/unit_tests/blas/batch/trsm_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/trsm_batch_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { +int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -72,17 +72,17 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { auto uaint = usm_allocator(cxt, *dev); vector m(uaint), n(uaint), lda(uaint), ldb(uaint), group_size(uaint); - auto uatranspose = usm_allocator(cxt, *dev); - vector trans(uatranspose); + auto uatranspose = usm_allocator(cxt, *dev); + vector trans(uatranspose); - auto uaside = usm_allocator(cxt, *dev); - vector left_right(uaside); + auto uaside = usm_allocator(cxt, *dev); + vector left_right(uaside); - auto uauplo = usm_allocator(cxt, *dev); - vector upper_lower(uauplo); + auto uauplo = usm_allocator(cxt, *dev); + vector upper_lower(uauplo); - auto uadiag = usm_allocator(cxt, *dev); - vector unit_nonunit(uadiag); + auto uadiag = usm_allocator(cxt, *dev); + vector unit_nonunit(uadiag); auto uafp = usm_allocator(cxt, *dev); vector alpha(uafp); @@ -112,18 +112,18 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { ldb[i] = std::max(n[i], m[i]); alpha[i] = rand_scalar(); if ((std::is_same::value) || (std::is_same::value)) { - trans[i] = (oneapi::mkl::transpose)(std::rand() % 2); + trans[i] = (oneapi::math::transpose)(std::rand() % 2); } else { tmp = std::rand() % 3; if (tmp == 2) - trans[i] = oneapi::mkl::transpose::conjtrans; + trans[i] = oneapi::math::transpose::conjtrans; else - trans[i] = (oneapi::mkl::transpose)tmp; + trans[i] = (oneapi::math::transpose)tmp; } - left_right[i] = (oneapi::mkl::side)(std::rand() % 2); - upper_lower[i] = (oneapi::mkl::uplo)(std::rand() % 2); - unit_nonunit[i] = (oneapi::mkl::diag)(std::rand() % 2); + left_right[i] = (oneapi::math::side)(std::rand() % 2); + upper_lower[i] = (oneapi::math::uplo)(std::rand() % 2); + unit_nonunit[i] = (oneapi::math::diag)(std::rand() % 2); total_batch_count += group_size[i]; } @@ -137,16 +137,16 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx = 0; for (i = 0; i < group_count; i++) { - size_a = lda[i] * (left_right[i] == oneapi::mkl::side::left ? m[i] : n[i]); - Arank = left_right[i] == oneapi::mkl::side::left ? m[i] : n[i]; - size_b = ldb[i] * ((layout == oneapi::mkl::layout::col_major) ? n[i] : m[i]); + size_a = lda[i] * (left_right[i] == oneapi::math::side::left ? m[i] : n[i]); + Arank = left_right[i] == oneapi::math::side::left ? m[i] : n[i]; + size_b = ldb[i] * ((layout == oneapi::math::layout::col_major) ? n[i] : m[i]); for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); - b_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); - b_ref_array[idx] = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); + a_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); + b_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); + b_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); rand_trsm_matrix(a_array[idx], layout, trans[i], Arank, Arank, lda[i]); - rand_matrix(b_array[idx], layout, oneapi::mkl::transpose::nontrans, m[i], n[i], ldb[i]); - copy_matrix(b_array[idx], layout, oneapi::mkl::transpose::nontrans, m[i], n[i], ldb[i], + rand_matrix(b_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldb[i]); + copy_matrix(b_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldb[i], b_ref_array[idx]); idx++; } @@ -154,40 +154,40 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { // Call reference TRSM_BATCH. using fp_ref = typename ref_type_info::type; - int *m_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *n_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *lda_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *ldb_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); - int *group_size_ref = (int *)oneapi::mkl::aligned_alloc(64, sizeof(int) * group_count); + int *m_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *n_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *lda_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *ldb_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int *group_size_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); CBLAS_TRANSPOSE *trans_ref = - (CBLAS_TRANSPOSE *)oneapi::mkl::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); + (CBLAS_TRANSPOSE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); CBLAS_SIDE *left_right_ref = - (CBLAS_SIDE *)oneapi::mkl::aligned_alloc(64, sizeof(CBLAS_SIDE) * group_count); + (CBLAS_SIDE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_SIDE) * group_count); CBLAS_UPLO *upper_lower_ref = - (CBLAS_UPLO *)oneapi::mkl::aligned_alloc(64, sizeof(CBLAS_UPLO) * group_count); + (CBLAS_UPLO *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_UPLO) * group_count); CBLAS_DIAG *unit_nonunit_ref = - (CBLAS_DIAG *)oneapi::mkl::aligned_alloc(64, sizeof(CBLAS_DIAG) * group_count); + (CBLAS_DIAG *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_DIAG) * group_count); if ((m_ref == NULL) || (n_ref == NULL) || (lda_ref == NULL) || (ldb_ref == NULL) || (trans_ref == NULL) || (left_right_ref == NULL) || (upper_lower_ref == NULL) || (unit_nonunit_ref == NULL) || (group_size_ref == NULL)) { std::cout << "Error cannot allocate input arrays\n"; - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(ldb_ref); - oneapi::mkl::aligned_free(trans_ref); - oneapi::mkl::aligned_free(left_right_ref); - oneapi::mkl::aligned_free(upper_lower_ref); - oneapi::mkl::aligned_free(unit_nonunit_ref); - oneapi::mkl::aligned_free(group_size_ref); + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(ldb_ref); + oneapi::math::aligned_free(trans_ref); + oneapi::math::aligned_free(left_right_ref); + oneapi::math::aligned_free(upper_lower_ref); + oneapi::math::aligned_free(unit_nonunit_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(b_array[idx], cxt); - oneapi::mkl::free_shared(b_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(b_array[idx], cxt); + oneapi::math::free_shared(b_ref_array[idx], cxt); idx++; } } @@ -218,14 +218,14 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::trsm_batch( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::trsm_batch( main_queue, &left_right[0], &upper_lower[0], &trans[0], &unit_nonunit[0], &m[0], &n[0], &alpha[0], (const fp **)&a_array[0], &lda[0], &b_array[0], &ldb[0], group_count, &group_size[0], dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::trsm_batch( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::trsm_batch( main_queue, &left_right[0], &upper_lower[0], &trans[0], &unit_nonunit[0], &m[0], &n[0], &alpha[0], (const fp **)&a_array[0], &lda[0], &b_array[0], &ldb[0], group_count, &group_size[0], dependencies); @@ -235,15 +235,15 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::trsm_batch, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trsm_batch, &left_right[0], &upper_lower[0], &trans[0], &unit_nonunit[0], &m[0], &n[0], &alpha[0], (const fp **)&a_array[0], &lda[0], &b_array[0], &ldb[0], group_count, &group_size[0], dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::trsm_batch, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsm_batch, &left_right[0], &upper_lower[0], &trans[0], &unit_nonunit[0], &m[0], &n[0], &alpha[0], (const fp **)&a_array[0], &lda[0], &b_array[0], &ldb[0], @@ -260,22 +260,22 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(ldb_ref); - oneapi::mkl::aligned_free(trans_ref); - oneapi::mkl::aligned_free(left_right_ref); - oneapi::mkl::aligned_free(upper_lower_ref); - oneapi::mkl::aligned_free(unit_nonunit_ref); - oneapi::mkl::aligned_free(group_size_ref); + catch (const oneapi::math::unimplemented &e) { + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(ldb_ref); + oneapi::math::aligned_free(trans_ref); + oneapi::math::aligned_free(left_right_ref); + oneapi::math::aligned_free(upper_lower_ref); + oneapi::math::aligned_free(unit_nonunit_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(b_array[idx], cxt); - oneapi::mkl::free_shared(b_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(b_array[idx], cxt); + oneapi::math::free_shared(b_ref_array[idx], cxt); idx++; } } @@ -296,21 +296,21 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { idx++; } } - oneapi::mkl::aligned_free(m_ref); - oneapi::mkl::aligned_free(n_ref); - oneapi::mkl::aligned_free(lda_ref); - oneapi::mkl::aligned_free(ldb_ref); - oneapi::mkl::aligned_free(trans_ref); - oneapi::mkl::aligned_free(left_right_ref); - oneapi::mkl::aligned_free(upper_lower_ref); - oneapi::mkl::aligned_free(unit_nonunit_ref); - oneapi::mkl::aligned_free(group_size_ref); + oneapi::math::aligned_free(m_ref); + oneapi::math::aligned_free(n_ref); + oneapi::math::aligned_free(lda_ref); + oneapi::math::aligned_free(ldb_ref); + oneapi::math::aligned_free(trans_ref); + oneapi::math::aligned_free(left_right_ref); + oneapi::math::aligned_free(upper_lower_ref); + oneapi::math::aligned_free(unit_nonunit_ref); + oneapi::math::aligned_free(group_size_ref); idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { - oneapi::mkl::free_shared(a_array[idx], cxt); - oneapi::mkl::free_shared(b_array[idx], cxt); - oneapi::mkl::free_shared(b_ref_array[idx], cxt); + oneapi::math::free_shared(a_array[idx], cxt); + oneapi::math::free_shared(b_array[idx], cxt); + oneapi::math::free_shared(b_ref_array[idx], cxt); idx++; } } @@ -319,7 +319,7 @@ int test(device *dev, oneapi::mkl::layout layout, int64_t group_count) { } class TrsmBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TrsmBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); @@ -345,8 +345,8 @@ TEST_P(TrsmBatchUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(TrsmBatchUsmTestSuite, TrsmBatchUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/gemm_bias.cpp b/tests/unit_tests/blas/extensions/gemm_bias.cpp index f08015827..8ef864af7 100644 --- a/tests/unit_tests/blas/extensions/gemm_bias.cpp +++ b/tests/unit_tests/blas/extensions/gemm_bias.cpp @@ -48,8 +48,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, int m, int n, int k, int lda, +int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, int m, int n, int k, int lda, int ldb, int ldc, Ts alpha, Ts beta) { // Prepare data. vector> A; @@ -61,13 +61,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, rand_matrix(A, layout, transa, m, k, lda); rand_matrix(B, layout, transb, k, n, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, m, n, ldc); - if (offsetc == oneapi::mkl::offset::fix) - rand_matrix(co, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, 1, 1, 1); - if (offsetc == oneapi::mkl::offset::column) - rand_matrix(co, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, m, 1, m); - if (offsetc == oneapi::mkl::offset::row) - rand_matrix(co, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, n, 1, n); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, m, n, ldc); + if (offsetc == oneapi::math::offset::fix) + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, 1, 1, 1); + if (offsetc == oneapi::math::offset::column) + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, m, 1, m); + if (offsetc == oneapi::math::offset::row) + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, n, 1, n); C_ref = C; @@ -112,13 +112,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::gemm_bias(main_queue, transa, transb, offsetc, m, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::gemm_bias(main_queue, transa, transb, offsetc, m, n, k, alpha, A_buffer, lda, ao, B_buffer, ldb, bo, beta, C_buffer, ldc, CO_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::gemm_bias(main_queue, transa, transb, offsetc, m, n, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::gemm_bias(main_queue, transa, transb, offsetc, m, n, k, alpha, A_buffer, lda, ao, B_buffer, ldb, bo, beta, C_buffer, ldc, CO_buffer); break; @@ -126,13 +126,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemm_bias, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemm_bias, transa, transb, offsetc, m, n, k, alpha, A_buffer, lda, ao, B_buffer, ldb, bo, beta, C_buffer, ldc, CO_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemm_bias, transa, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm_bias, transa, transb, offsetc, m, n, k, alpha, A_buffer, lda, ao, B_buffer, ldb, bo, beta, C_buffer, ldc, CO_buffer); break; @@ -146,7 +146,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -162,58 +162,58 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, } class GemmBiasTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemmBiasTests, Int8Int8Int32Precision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); } @@ -221,52 +221,52 @@ TEST_P(GemmBiasTests, Int8Uint8Int32Precision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); } @@ -274,52 +274,52 @@ TEST_P(GemmBiasTests, Uint8Int8Int32Precision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); } @@ -327,59 +327,59 @@ TEST_P(GemmBiasTests, Uint8Uint8Int32Precision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); } INSTANTIATE_TEST_SUITE_P(GemmBiasTestSuite, GemmBiasTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp b/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp index b223c5700..19f68dc66 100644 --- a/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp +++ b/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp @@ -48,8 +48,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, oneapi::mkl::offset offsetc, int m, int n, int k, int lda, +int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, int m, int n, int k, int lda, int ldb, int ldc, Ts alpha, Ts beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -83,13 +83,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, rand_matrix(A, layout, transa, m, k, lda); rand_matrix(B, layout, transb, k, n, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, m, n, ldc); - if (offsetc == oneapi::mkl::offset::fix) - rand_matrix(co, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, 1, 1, 1); - if (offsetc == oneapi::mkl::offset::column) - rand_matrix(co, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, m, 1, m); - if (offsetc == oneapi::mkl::offset::row) - rand_matrix(co, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, n, 1, n); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, m, n, ldc); + if (offsetc == oneapi::math::offset::fix) + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, 1, 1, 1); + if (offsetc == oneapi::math::offset::column) + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, m, 1, m); + if (offsetc == oneapi::math::offset::row) + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, n, 1, n); C_ref.resize(C.size()); for (int i = 0; i < C.size(); i++) @@ -115,13 +115,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::gemm_bias( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::gemm_bias( main_queue, transa, transb, offsetc, m, n, k, alpha, A.data(), lda, ao, B.data(), ldb, bo, beta, C.data(), ldc, co.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::gemm_bias( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::gemm_bias( main_queue, transa, transb, offsetc, m, n, k, alpha, A.data(), lda, ao, B.data(), ldb, bo, beta, C.data(), ldc, co.data(), dependencies); break; @@ -130,14 +130,14 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemm_bias, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemm_bias, transa, transb, offsetc, m, n, k, alpha, A.data(), lda, ao, B.data(), ldb, bo, beta, C.data(), ldc, co.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemm_bias, transa, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm_bias, transa, transb, offsetc, m, n, k, alpha, A.data(), lda, ao, B.data(), ldb, bo, beta, C.data(), ldc, co.data(), dependencies); @@ -153,7 +153,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -168,58 +168,58 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, } class GemmBiasUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemmBiasUsmTests, Int8Int8Int32Precision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); } @@ -227,52 +227,52 @@ TEST_P(GemmBiasUsmTests, Int8Uint8Int32Precision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); } @@ -280,52 +280,52 @@ TEST_P(GemmBiasUsmTests, Uint8Int8Int32Precision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); } @@ -333,59 +333,59 @@ TEST_P(GemmBiasUsmTests, Uint8Uint8Int32Precision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::fix, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::fix, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::column, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::column, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, oneapi::mkl::offset::row, 79, 83, 91, 103, 105, 106, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, oneapi::math::offset::row, 79, 83, 91, 103, 105, 106, alpha, beta))); } INSTANTIATE_TEST_SUITE_P(GemmBiasUsmTestSuite, GemmBiasUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/gemmt.cpp b/tests/unit_tests/blas/extensions/gemmt.cpp index f83e2d1b0..fdcec77e6 100644 --- a/tests/unit_tests/blas/extensions/gemmt.cpp +++ b/tests/unit_tests/blas/extensions/gemmt.cpp @@ -48,14 +48,14 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, int n, int k, int lda, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, int n, int k, int lda, int ldb, int ldc, fp alpha, fp beta) { // Prepare data. vector> A, B, C, C_ref; rand_matrix(A, layout, transa, n, k, lda); rand_matrix(B, layout, transb, k, n, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, n, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, n, n, ldc); C_ref = C; // Call Reference GEMMT. @@ -94,13 +94,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::gemmt(main_queue, upper_lower, transa, transb, n, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::gemmt(main_queue, upper_lower, transa, transb, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::gemmt(main_queue, upper_lower, transa, transb, n, k, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::gemmt(main_queue, upper_lower, transa, transb, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -108,13 +108,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemmt, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemmt, upper_lower, transa, transb, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemmt, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemmt, upper_lower, transa, transb, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -127,7 +127,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -143,43 +143,43 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class GemmtTests : public ::testing::TestWithParam> { +class GemmtTests : public ::testing::TestWithParam> { }; TEST_P(GemmtTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); } @@ -189,36 +189,36 @@ TEST_P(GemmtTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); } @@ -226,76 +226,76 @@ TEST_P(GemmtTests, ComplexSinglePrecision) { std::complex alpha(2.0); std::complex beta(3.0); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); } @@ -305,83 +305,83 @@ TEST_P(GemmtTests, ComplexDoublePrecision) { std::complex alpha(2.0); std::complex beta(3.0); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(GemmtTestSuite, GemmtTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/gemmt_usm.cpp b/tests/unit_tests/blas/extensions/gemmt_usm.cpp index 1a3b3e450..8a5db1e15 100644 --- a/tests/unit_tests/blas/extensions/gemmt_usm.cpp +++ b/tests/unit_tests/blas/extensions/gemmt_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, int n, int k, int lda, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, int n, int k, int lda, int ldb, int ldc, fp alpha, fp beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -74,7 +74,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, vector A(ua), B(ua), C(ua); rand_matrix(A, layout, transa, n, k, lda); rand_matrix(B, layout, transb, k, n, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, n, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, n, n, ldc); auto C_ref = C; @@ -94,13 +94,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::gemmt( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::gemmt( main_queue, upper_lower, transa, transb, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::gemmt(main_queue, upper_lower, transa, transb, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::gemmt(main_queue, upper_lower, transa, transb, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -109,13 +109,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemmt, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemmt, upper_lower, transa, transb, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemmt, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemmt, upper_lower, transa, transb, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -129,7 +129,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -144,42 +144,42 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class GemmtUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemmtUsmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); } @@ -189,36 +189,36 @@ TEST_P(GemmtUsmTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); } @@ -226,76 +226,76 @@ TEST_P(GemmtUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0); std::complex beta(3.0); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); } @@ -305,83 +305,83 @@ TEST_P(GemmtUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0); std::complex beta(3.0); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, alpha, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::trans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::conjtrans, 27, 98, 101, 102, 103, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(GemmtUsmTestSuite, GemmtUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/imatcopy.cpp b/tests/unit_tests/blas/extensions/imatcopy.cpp index 739612791..46335f012 100644 --- a/tests/unit_tests/blas/extensions/imatcopy.cpp +++ b/tests/unit_tests/blas/extensions/imatcopy.cpp @@ -48,11 +48,11 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb; - oneapi::mkl::transpose trans; + oneapi::math::transpose trans; fp alpha; int64_t i, tmp; @@ -65,13 +65,13 @@ int test(device *dev, oneapi::mkl::layout layout) { int64_t size_a, size_b, size; switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: size_a = lda * n; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: size_a = lda * m; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; break; default: break; } @@ -79,9 +79,9 @@ int test(device *dev, oneapi::mkl::layout layout) { vector> AB(size), AB_ref(size); - rand_matrix(AB, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size, 1, + rand_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size, 1, size); - copy_matrix(AB, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size, 1, size, + copy_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size, 1, size, AB_ref); // Call reference IMATCOPY. @@ -114,24 +114,24 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::imatcopy(main_queue, trans, m, n, alpha, AB_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::imatcopy(main_queue, trans, m, n, alpha, AB_buffer, lda, ldb); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::imatcopy(main_queue, trans, m, n, alpha, AB_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::imatcopy(main_queue, trans, m, n, alpha, AB_buffer, lda, ldb); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::imatcopy, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::imatcopy, trans, m, n, alpha, AB_buffer, lda, ldb); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::imatcopy, trans, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::imatcopy, trans, m, n, alpha, AB_buffer, lda, ldb); break; default: break; @@ -144,7 +144,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -155,14 +155,14 @@ int test(device *dev, oneapi::mkl::layout layout) { // Compare the results of reference implementation and DPC++ implementation. auto AB_accessor = AB_buffer.get_host_access(read_only); - bool good = check_equal_matrix(AB_accessor, AB_ref, oneapi::mkl::layout::col_major, size, 1, + bool good = check_equal_matrix(AB_accessor, AB_ref, oneapi::math::layout::col_major, size, 1, size, 10, std::cout); return (int)good; } class ImatcopyTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(ImatcopyTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); @@ -186,8 +186,8 @@ TEST_P(ImatcopyTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(ImatcopyTestSuite, ImatcopyTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/imatcopy_usm.cpp b/tests/unit_tests/blas/extensions/imatcopy_usm.cpp index a8f3745bf..c247cd699 100644 --- a/tests/unit_tests/blas/extensions/imatcopy_usm.cpp +++ b/tests/unit_tests/blas/extensions/imatcopy_usm.cpp @@ -49,7 +49,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -72,7 +72,7 @@ int test(device *dev, oneapi::mkl::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb; - oneapi::mkl::transpose trans; + oneapi::math::transpose trans; fp alpha; int64_t i; @@ -85,13 +85,13 @@ int test(device *dev, oneapi::mkl::layout layout) { int64_t size_a, size_b, size; switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: size_a = lda * n; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: size_a = lda * m; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; break; default: break; } @@ -103,9 +103,9 @@ int test(device *dev, oneapi::mkl::layout layout) { AB.resize(size); AB_ref.resize(size); - rand_matrix(AB, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size, 1, + rand_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size, 1, size); - copy_matrix(AB, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size, 1, size, + copy_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size, 1, size, AB_ref); // Call reference IMATCOPY. @@ -119,12 +119,12 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::imatcopy(main_queue, trans, m, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::imatcopy(main_queue, trans, m, n, alpha, &AB[0], lda, ldb, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::imatcopy(main_queue, trans, m, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::imatcopy(main_queue, trans, m, n, alpha, &AB[0], lda, ldb, dependencies); break; default: break; @@ -132,12 +132,12 @@ int test(device *dev, oneapi::mkl::layout layout) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::imatcopy, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::imatcopy, trans, m, n, alpha, &AB[0], lda, ldb, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::imatcopy, trans, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::imatcopy, trans, m, n, alpha, &AB[0], lda, ldb, dependencies); break; default: break; @@ -151,7 +151,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -160,14 +160,14 @@ int test(device *dev, oneapi::mkl::layout layout) { } // Compare the results of reference implementation and DPC++ implementation. - bool good = check_equal_matrix(AB, AB_ref, oneapi::mkl::layout::col_major, size, 1, size, 10, + bool good = check_equal_matrix(AB, AB_ref, oneapi::math::layout::col_major, size, 1, size, 10, std::cout); return (int)good; } class ImatcopyUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(ImatcopyUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); @@ -191,8 +191,8 @@ TEST_P(ImatcopyUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(ImatcopyUsmTestSuite, ImatcopyUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/omatadd.cpp b/tests/unit_tests/blas/extensions/omatadd.cpp index 9cebed350..bf0a9ac23 100644 --- a/tests/unit_tests/blas/extensions/omatadd.cpp +++ b/tests/unit_tests/blas/extensions/omatadd.cpp @@ -48,11 +48,11 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb, ldc; - oneapi::mkl::transpose transa, transb; + oneapi::math::transpose transa, transb; fp alpha, beta; int64_t i, tmp; @@ -69,14 +69,14 @@ int test(device *dev, oneapi::mkl::layout layout) { int64_t size_a, size_b, size_c; switch (layout) { - case oneapi::mkl::layout::col_major: - size_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * n : lda * m; - size_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + case oneapi::math::layout::col_major: + size_a = (transa == oneapi::math::transpose::nontrans) ? lda * n : lda * m; + size_b = (transb == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; size_c = ldc * n; break; - case oneapi::mkl::layout::row_major: - size_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * m : lda * n; - size_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + case oneapi::math::layout::row_major: + size_a = (transa == oneapi::math::transpose::nontrans) ? lda * m : lda * n; + size_b = (transb == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; size_c = ldc * m; break; default: break; @@ -84,13 +84,13 @@ int test(device *dev, oneapi::mkl::layout layout) { vector> A(size_a), B(size_b), C(size_c), C_ref(size_c); - rand_matrix(A.data(), oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_a, + rand_matrix(A.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_a, 1, size_a); - rand_matrix(B.data(), oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_b, + rand_matrix(B.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_b, 1, size_b); - rand_matrix(C.data(), oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_c, + rand_matrix(C.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_c, 1, size_c); - copy_matrix(C.data(), oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_c, + copy_matrix(C.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_c, 1, size_c, C_ref.data()); // Call reference OMATADD. @@ -127,13 +127,13 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::omatadd(main_queue, transa, transb, m, n, alpha, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::omatadd(main_queue, transa, transb, m, n, alpha, A_buffer, lda, beta, B_buffer, ldb, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::omatadd(main_queue, transa, transb, m, n, alpha, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::omatadd(main_queue, transa, transb, m, n, alpha, A_buffer, lda, beta, B_buffer, ldb, C_buffer, ldc); break; @@ -141,13 +141,13 @@ int test(device *dev, oneapi::mkl::layout layout) { } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::omatadd, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatadd, transa, transb, m, n, alpha, A_buffer, lda, beta, B_buffer, ldb, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::omatadd, transa, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatadd, transa, transb, m, n, alpha, A_buffer, lda, beta, B_buffer, ldb, C_buffer, ldc); break; @@ -160,7 +160,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -171,14 +171,14 @@ int test(device *dev, oneapi::mkl::layout layout) { // Compare the results of reference implementation and DPC++ implementation. auto C_accessor = C_buffer.get_host_access(read_only); - bool good = check_equal_matrix(C_accessor, C_ref, oneapi::mkl::layout::col_major, size_c, 1, + bool good = check_equal_matrix(C_accessor, C_ref, oneapi::math::layout::col_major, size_c, 1, size_c, 10, std::cout); return (int)good; } class OmataddTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmataddTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); @@ -202,8 +202,8 @@ TEST_P(OmataddTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(OmataddTestSuite, OmataddTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/omatadd_usm.cpp b/tests/unit_tests/blas/extensions/omatadd_usm.cpp index 1cd838c17..cb9b45d4a 100644 --- a/tests/unit_tests/blas/extensions/omatadd_usm.cpp +++ b/tests/unit_tests/blas/extensions/omatadd_usm.cpp @@ -48,7 +48,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -71,7 +71,7 @@ int test(device *dev, oneapi::mkl::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb, ldc; - oneapi::mkl::transpose transa, transb; + oneapi::math::transpose transa, transb; fp alpha, beta; int64_t i, tmp; @@ -88,14 +88,14 @@ int test(device *dev, oneapi::mkl::layout layout) { int64_t size_a, size_b, size_c; switch (layout) { - case oneapi::mkl::layout::col_major: - size_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * n : lda * m; - size_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + case oneapi::math::layout::col_major: + size_a = (transa == oneapi::math::transpose::nontrans) ? lda * n : lda * m; + size_b = (transb == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; size_c = ldc * n; break; - case oneapi::mkl::layout::row_major: - size_a = (transa == oneapi::mkl::transpose::nontrans) ? lda * m : lda * n; - size_b = (transb == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + case oneapi::math::layout::row_major: + size_a = (transa == oneapi::math::transpose::nontrans) ? lda * m : lda * n; + size_b = (transb == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; size_c = ldc * m; break; default: break; @@ -109,13 +109,13 @@ int test(device *dev, oneapi::mkl::layout layout) { C.resize(size_c); C_ref.resize(size_c); - rand_matrix(A, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_a, 1, + rand_matrix(A, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_a, 1, size_a); - rand_matrix(B, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_b, 1, + rand_matrix(B, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_b, 1, size_b); - rand_matrix(C, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_c, 1, + rand_matrix(C, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_c, 1, size_c); - copy_matrix(C, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_c, 1, + copy_matrix(C, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_c, 1, size_c, C_ref); // Call reference OMATADD. @@ -131,13 +131,13 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::omatadd(main_queue, transa, transb, m, n, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::omatadd(main_queue, transa, transb, m, n, alpha, &A[0], lda, beta, &B[0], ldb, &C[0], ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::omatadd(main_queue, transa, transb, m, n, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::omatadd(main_queue, transa, transb, m, n, alpha, &A[0], lda, beta, &B[0], ldb, &C[0], ldc, dependencies); break; @@ -146,13 +146,13 @@ int test(device *dev, oneapi::mkl::layout layout) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::omatadd, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatadd, transa, transb, m, n, alpha, &A[0], lda, beta, &B[0], ldb, &C[0], ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::omatadd, transa, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatadd, transa, transb, m, n, alpha, &A[0], lda, beta, &B[0], ldb, &C[0], ldc, dependencies); break; @@ -166,7 +166,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -175,14 +175,14 @@ int test(device *dev, oneapi::mkl::layout layout) { } // Compare the results of reference implementation and DPC++ implementation. - bool good = check_equal_matrix(C, C_ref, oneapi::mkl::layout::col_major, size_c, 1, size_c, 10, + bool good = check_equal_matrix(C, C_ref, oneapi::math::layout::col_major, size_c, 1, size_c, 10, std::cout); return (int)good; } class OmataddUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmataddUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); @@ -206,8 +206,8 @@ TEST_P(OmataddUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(OmataddUsmTestSuite, OmataddUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/omatcopy.cpp b/tests/unit_tests/blas/extensions/omatcopy.cpp index 022cf2e63..f77de83f1 100644 --- a/tests/unit_tests/blas/extensions/omatcopy.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy.cpp @@ -48,11 +48,11 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb; - oneapi::mkl::transpose trans; + oneapi::math::transpose trans; fp alpha; int64_t i, tmp; @@ -63,33 +63,33 @@ int test(device *dev, oneapi::mkl::layout layout) { alpha = rand_scalar(); if ((std::is_same::value) || (std::is_same::value)) { - trans = (oneapi::mkl::transpose)(std::rand() % 2); + trans = (oneapi::math::transpose)(std::rand() % 2); } else { tmp = std::rand() % 3; if (tmp == 2) - trans = oneapi::mkl::transpose::conjtrans; + trans = oneapi::math::transpose::conjtrans; else - trans = (oneapi::mkl::transpose)tmp; + trans = (oneapi::math::transpose)tmp; } int64_t size_a, size_b; switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: size_a = lda * n; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: size_a = lda * m; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; break; default: break; } vector> A(size_a), B(size_b), B_ref(size_b); - rand_matrix(A.data(), layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A.data(), layout, oneapi::math::transpose::nontrans, m, n, lda); rand_matrix(B.data(), layout, trans, m, n, ldb); // Call reference OMATCOPY. @@ -123,24 +123,24 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::omatcopy(main_queue, trans, m, n, alpha, A_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::omatcopy(main_queue, trans, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::omatcopy(main_queue, trans, m, n, alpha, A_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::omatcopy(main_queue, trans, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::omatcopy, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatcopy, trans, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::omatcopy, trans, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatcopy, trans, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; default: break; @@ -153,7 +153,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -164,14 +164,14 @@ int test(device *dev, oneapi::mkl::layout layout) { // Compare the results of reference implementation and DPC++ implementation. auto B_accessor = B_buffer.get_host_access(read_only); - bool good = check_equal_matrix(B_accessor, B_ref, oneapi::mkl::layout::col_major, size_b, 1, + bool good = check_equal_matrix(B_accessor, B_ref, oneapi::math::layout::col_major, size_b, 1, size_b, 10, std::cout); return (int)good; } class OmatcopyTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmatcopyTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); @@ -195,8 +195,8 @@ TEST_P(OmatcopyTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(OmatcopyTestSuite, OmatcopyTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/omatcopy2.cpp b/tests/unit_tests/blas/extensions/omatcopy2.cpp index 93038c63b..41b4b6e7c 100644 --- a/tests/unit_tests/blas/extensions/omatcopy2.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy2.cpp @@ -48,12 +48,12 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb; int64_t stride_a, stride_b; - oneapi::mkl::transpose trans; + oneapi::math::transpose trans; fp alpha; stride_a = 1 + std::rand() % 50; @@ -68,22 +68,22 @@ int test(device *dev, oneapi::mkl::layout layout) { int64_t size_a, size_b; switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: size_a = lda * n; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: size_a = lda * m; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; break; default: break; } vector> A(size_a), B(size_b), B_ref(size_b); - rand_matrix(A.data(), layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A.data(), layout, oneapi::math::transpose::nontrans, m, n, lda); rand_matrix(B.data(), layout, trans, m, n, ldb); - copy_matrix(B.data(), oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_b, + copy_matrix(B.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_b, 1, size_b, B_ref.data()); // Call reference OMATCOPY2. @@ -120,25 +120,25 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::omatcopy2(main_queue, trans, m, n, alpha, A_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::omatcopy2(main_queue, trans, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::omatcopy2(main_queue, trans, m, n, alpha, A_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::omatcopy2(main_queue, trans, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::omatcopy2, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatcopy2, trans, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::omatcopy2, trans, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatcopy2, trans, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b); break; @@ -152,7 +152,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -163,14 +163,14 @@ int test(device *dev, oneapi::mkl::layout layout) { // Compare the results of reference implementation and DPC++ implementation. auto B_accessor = B_buffer.get_host_access(read_only); - bool good = check_equal_matrix(B_accessor, B_ref, oneapi::mkl::layout::col_major, size_b, 1, + bool good = check_equal_matrix(B_accessor, B_ref, oneapi::math::layout::col_major, size_b, 1, size_b, 10, std::cout); return (int)good; } class Omatcopy2Tests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Omatcopy2Tests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); @@ -194,8 +194,8 @@ TEST_P(Omatcopy2Tests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(Omatcopy2TestSuite, Omatcopy2Tests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp b/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp index 1e880cf5b..cdcc21020 100644 --- a/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp @@ -49,7 +49,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -73,7 +73,7 @@ int test(device *dev, oneapi::mkl::layout layout) { int64_t m, n; int64_t lda, ldb; int64_t stride_a, stride_b; - oneapi::mkl::transpose trans; + oneapi::math::transpose trans; fp alpha; stride_a = 1 + std::rand() % 50; @@ -88,13 +88,13 @@ int test(device *dev, oneapi::mkl::layout layout) { int64_t size_a, size_b; switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: size_a = lda * n; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: size_a = lda * m; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; break; default: break; } @@ -106,11 +106,11 @@ int test(device *dev, oneapi::mkl::layout layout) { B.resize(size_b); B_ref.resize(size_b); - rand_matrix(A, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_a, 1, + rand_matrix(A, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_a, 1, size_a); - rand_matrix(B, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_b, 1, + rand_matrix(B, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_b, 1, size_b); - copy_matrix(B, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_b, 1, + copy_matrix(B, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_b, 1, size_b, B_ref); // Call reference OMATCOPY2. @@ -127,13 +127,13 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::omatcopy2(main_queue, trans, m, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::omatcopy2(main_queue, trans, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::omatcopy2(main_queue, trans, m, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::omatcopy2(main_queue, trans, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, dependencies); break; @@ -142,13 +142,13 @@ int test(device *dev, oneapi::mkl::layout layout) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::omatcopy2, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatcopy2, trans, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::omatcopy2, trans, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatcopy2, trans, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, dependencies); break; @@ -163,7 +163,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -172,14 +172,14 @@ int test(device *dev, oneapi::mkl::layout layout) { } // Compare the results of reference implementation and DPC++ implementation. - bool good = check_equal_matrix(B, B_ref, oneapi::mkl::layout::col_major, size_b, 1, size_b, 10, + bool good = check_equal_matrix(B, B_ref, oneapi::math::layout::col_major, size_b, 1, size_b, 10, std::cout); return (int)good; } class Omatcopy2UsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Omatcopy2UsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); @@ -203,8 +203,8 @@ TEST_P(Omatcopy2UsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(Omatcopy2UsmTestSuite, Omatcopy2UsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/extensions/omatcopy_usm.cpp b/tests/unit_tests/blas/extensions/omatcopy_usm.cpp index beddf9c5d..7ba026ab3 100644 --- a/tests/unit_tests/blas/extensions/omatcopy_usm.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy_usm.cpp @@ -49,7 +49,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -72,7 +72,7 @@ int test(device *dev, oneapi::mkl::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb; - oneapi::mkl::transpose trans; + oneapi::math::transpose trans; fp alpha; int64_t i, tmp; @@ -85,13 +85,13 @@ int test(device *dev, oneapi::mkl::layout layout) { int64_t size_a, size_b; switch (layout) { - case oneapi::mkl::layout::col_major: + case oneapi::math::layout::col_major: size_a = lda * n; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * n : ldb * m; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * n : ldb * m; break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: size_a = lda * m; - size_b = (trans == oneapi::mkl::transpose::nontrans) ? ldb * m : ldb * n; + size_b = (trans == oneapi::math::transpose::nontrans) ? ldb * m : ldb * n; break; default: break; } @@ -103,11 +103,11 @@ int test(device *dev, oneapi::mkl::layout layout) { B.resize(size_b); B_ref.resize(size_b); - rand_matrix(A, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_a, 1, + rand_matrix(A, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_a, 1, size_a); - rand_matrix(B, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_b, 1, + rand_matrix(B, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_b, 1, size_b); - copy_matrix(B, oneapi::mkl::layout::col_major, oneapi::mkl::transpose::nontrans, size_b, 1, + copy_matrix(B, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_b, 1, size_b, B_ref); // Call reference OMATCOPY. @@ -121,12 +121,12 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::omatcopy( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::omatcopy( main_queue, trans, m, n, alpha, &A[0], lda, &B[0], ldb, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::omatcopy(main_queue, trans, m, n, alpha, &A[0], + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::omatcopy(main_queue, trans, m, n, alpha, &A[0], lda, &B[0], ldb, dependencies); break; default: break; @@ -134,12 +134,12 @@ int test(device *dev, oneapi::mkl::layout layout) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::omatcopy, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatcopy, trans, m, n, alpha, &A[0], lda, &B[0], ldb, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::omatcopy, trans, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatcopy, trans, m, n, alpha, &A[0], lda, &B[0], ldb, dependencies); break; default: break; @@ -153,7 +153,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -162,14 +162,14 @@ int test(device *dev, oneapi::mkl::layout layout) { } // Compare the results of reference implementation and DPC++ implementation. - bool good = check_equal_matrix(B, B_ref, oneapi::mkl::layout::col_major, size_b, 1, size_b, 10, + bool good = check_equal_matrix(B, B_ref, oneapi::math::layout::col_major, size_b, 1, size_b, 10, std::cout); return (int)good; } class OmatcopyUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmatcopyUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); @@ -193,8 +193,8 @@ TEST_P(OmatcopyUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(OmatcopyUsmTestSuite, OmatcopyUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/include/allocator_helper.hpp b/tests/unit_tests/blas/include/allocator_helper.hpp index 79fd22254..6254ad26c 100644 --- a/tests/unit_tests/blas/include/allocator_helper.hpp +++ b/tests/unit_tests/blas/include/allocator_helper.hpp @@ -48,7 +48,7 @@ struct allocator_helper { allocator_helper(allocator_helper&& other) noexcept {} T* allocate(size_t n) { - void* mem = oneapi::mkl::aligned_alloc(align, n * sizeof(T)); + void* mem = oneapi::math::aligned_alloc(align, n * sizeof(T)); if (!mem) throw std::bad_alloc(); @@ -56,7 +56,7 @@ struct allocator_helper { } void deallocate(T* p, size_t n) noexcept { - oneapi::mkl::aligned_free(p); + oneapi::math::aligned_free(p); } constexpr size_t max_size() const noexcept { diff --git a/tests/unit_tests/blas/include/onemath_blas_helper.hpp b/tests/unit_tests/blas/include/onemath_blas_helper.hpp index 3358c1c9c..f5944fe92 100644 --- a/tests/unit_tests/blas/include/onemath_blas_helper.hpp +++ b/tests/unit_tests/blas/include/onemath_blas_helper.hpp @@ -31,38 +31,38 @@ typedef enum { CblasRowOffset = 101, CblasColOffset = 102, CblasFixOffset = 103 * equivalents. */ -inline CBLAS_TRANSPOSE convert_to_cblas_trans(oneapi::mkl::transpose trans) { - if (trans == oneapi::mkl::transpose::trans) +inline CBLAS_TRANSPOSE convert_to_cblas_trans(oneapi::math::transpose trans) { + if (trans == oneapi::math::transpose::trans) return CBLAS_TRANSPOSE::CblasTrans; - else if (trans == oneapi::mkl::transpose::conjtrans) + else if (trans == oneapi::math::transpose::conjtrans) return CBLAS_TRANSPOSE::CblasConjTrans; else return CBLAS_TRANSPOSE::CblasNoTrans; } -inline CBLAS_UPLO convert_to_cblas_uplo(oneapi::mkl::uplo is_upper) { - return is_upper == oneapi::mkl::uplo::upper ? CBLAS_UPLO::CblasUpper : CBLAS_UPLO::CblasLower; +inline CBLAS_UPLO convert_to_cblas_uplo(oneapi::math::uplo is_upper) { + return is_upper == oneapi::math::uplo::upper ? CBLAS_UPLO::CblasUpper : CBLAS_UPLO::CblasLower; } -inline CBLAS_DIAG convert_to_cblas_diag(oneapi::mkl::diag is_unit) { - return is_unit == oneapi::mkl::diag::unit ? CBLAS_DIAG::CblasUnit : CBLAS_DIAG::CblasNonUnit; +inline CBLAS_DIAG convert_to_cblas_diag(oneapi::math::diag is_unit) { + return is_unit == oneapi::math::diag::unit ? CBLAS_DIAG::CblasUnit : CBLAS_DIAG::CblasNonUnit; } -inline CBLAS_SIDE convert_to_cblas_side(oneapi::mkl::side is_left) { - return is_left == oneapi::mkl::side::left ? CBLAS_SIDE::CblasLeft : CBLAS_SIDE::CblasRight; +inline CBLAS_SIDE convert_to_cblas_side(oneapi::math::side is_left) { + return is_left == oneapi::math::side::left ? CBLAS_SIDE::CblasLeft : CBLAS_SIDE::CblasRight; } -inline CBLAS_OFFSET convert_to_cblas_offset(oneapi::mkl::offset offsetc) { - if (offsetc == oneapi::mkl::offset::fix) +inline CBLAS_OFFSET convert_to_cblas_offset(oneapi::math::offset offsetc) { + if (offsetc == oneapi::math::offset::fix) return CBLAS_OFFSET::CblasFixOffset; - else if (offsetc == oneapi::mkl::offset::column) + else if (offsetc == oneapi::math::offset::column) return CBLAS_OFFSET::CblasColOffset; else return CBLAS_OFFSET::CblasRowOffset; } -inline CBLAS_LAYOUT convert_to_cblas_layout(oneapi::mkl::layout is_column) { - return is_column == oneapi::mkl::layout::col_major ? CBLAS_LAYOUT::CblasColMajor +inline CBLAS_LAYOUT convert_to_cblas_layout(oneapi::math::layout is_column) { + return is_column == oneapi::math::layout::col_major ? CBLAS_LAYOUT::CblasColMajor : CBLAS_LAYOUT::CblasRowMajor; } diff --git a/tests/unit_tests/blas/include/reference_blas_templates.hpp b/tests/unit_tests/blas/include/reference_blas_templates.hpp index 182fcfad6..b84e71c1b 100644 --- a/tests/unit_tests/blas/include/reference_blas_templates.hpp +++ b/tests/unit_tests/blas/include/reference_blas_templates.hpp @@ -162,18 +162,18 @@ void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, c sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; sizec = *ldc * *m; } - float *af = (float *)oneapi::mkl::aligned_alloc(64, sizeof(float) * sizea); - float *bf = (float *)oneapi::mkl::aligned_alloc(64, sizeof(float) * sizeb); - float *cf = (float *)oneapi::mkl::aligned_alloc(64, sizeof(float) * sizec); + float *af = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizea); + float *bf = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizeb); + float *cf = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizec); copy_mat(a, layout, transa, *m, *k, *lda, af); copy_mat(b, layout, transb, *k, *n, *ldb, bf); copy_mat(c, layout, CblasNoTrans, *m, *n, *ldc, cf); cblas_sgemm_wrapper(layout, transa, transb, *m, *n, *k, alphaf, af, *lda, bf, *ldb, betaf, cf, *ldc); copy_mat(cf, layout, CblasNoTrans, *m, *n, *ldc, c); - oneapi::mkl::aligned_free(af); - oneapi::mkl::aligned_free(bf); - oneapi::mkl::aligned_free(cf); + oneapi::math::aligned_free(af); + oneapi::math::aligned_free(bf); + oneapi::math::aligned_free(cf); } template <> @@ -230,20 +230,20 @@ void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, c sizea = (transa == CblasNoTrans) ? *lda * *m : *lda * *k; sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; } - float *af = (float *)oneapi::mkl::aligned_alloc(64, sizeof(float) * sizea); - float *bf = (float *)oneapi::mkl::aligned_alloc(64, sizeof(float) * sizeb); + float *af = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizea); + float *bf = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizeb); copy_mat(a, layout, transa, *m, *k, *lda, af); copy_mat(b, layout, transb, *k, *n, *ldb, bf); cblas_sgemm_wrapper(layout, transa, transb, *m, *n, *k, *alpha, af, *lda, bf, *ldb, *beta, c, *ldc); - oneapi::mkl::aligned_free(af); - oneapi::mkl::aligned_free(bf); + oneapi::math::aligned_free(af); + oneapi::math::aligned_free(bf); } template <> void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int *m, - const int *n, const int *k, const float *alpha, const oneapi::mkl::bfloat16 *a, - const int *lda, const oneapi::mkl::bfloat16 *b, const int *ldb, const float *beta, + const int *n, const int *k, const float *alpha, const oneapi::math::bfloat16 *a, + const int *lda, const oneapi::math::bfloat16 *b, const int *ldb, const float *beta, float *c, const int *ldc) { // Not supported in NETLIB. SGEMM is used as reference. int sizea, sizeb; @@ -255,14 +255,14 @@ void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, c sizea = (transa == CblasNoTrans) ? *lda * *m : *lda * *k; sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; } - float *af = (float *)oneapi::mkl::aligned_alloc(64, sizeof(float) * sizea); - float *bf = (float *)oneapi::mkl::aligned_alloc(64, sizeof(float) * sizeb); + float *af = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizea); + float *bf = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizeb); copy_mat(a, layout, transa, *m, *k, *lda, af); copy_mat(b, layout, transb, *k, *n, *ldb, bf); cblas_sgemm_wrapper(layout, transa, transb, *m, *n, *k, *alpha, af, *lda, bf, *ldb, *beta, c, *ldc); - oneapi::mkl::aligned_free(af); - oneapi::mkl::aligned_free(bf); + oneapi::math::aligned_free(af); + oneapi::math::aligned_free(bf); } template @@ -1591,9 +1591,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; sizec = *ldc * *m; } - double *ad = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizea); - double *bd = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizeb); - double *cd = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizec); + double *ad = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); + double *bd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); + double *cd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); double alphad = *alpha; double betad = *beta; double aod = *ao; @@ -1604,9 +1604,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran cblas_dgemm_wrapper(layout, transa, transb, *m, *n, *k, alphad, ad, *lda, bd, *ldb, betad, cd, *ldc); copy_mat(cd, layout, *m, *n, *ldc, offsetc, co, c); - oneapi::mkl::aligned_free(ad); - oneapi::mkl::aligned_free(bd); - oneapi::mkl::aligned_free(cd); + oneapi::math::aligned_free(ad); + oneapi::math::aligned_free(bd); + oneapi::math::aligned_free(cd); } template <> @@ -1627,9 +1627,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; sizec = *ldc * *m; } - double *ad = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizea); - double *bd = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizeb); - double *cd = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizec); + double *ad = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); + double *bd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); + double *cd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); double alphad = *alpha; double betad = *beta; double aod = *ao; @@ -1640,9 +1640,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran cblas_dgemm_wrapper(layout, transa, transb, *m, *n, *k, alphad, ad, *lda, bd, *ldb, betad, cd, *ldc); copy_mat(cd, layout, *m, *n, *ldc, offsetc, co, c); - oneapi::mkl::aligned_free(ad); - oneapi::mkl::aligned_free(bd); - oneapi::mkl::aligned_free(cd); + oneapi::math::aligned_free(ad); + oneapi::math::aligned_free(bd); + oneapi::math::aligned_free(cd); } template <> @@ -1662,9 +1662,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; sizec = *ldc * *m; } - double *ad = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizea); - double *bd = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizeb); - double *cd = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizec); + double *ad = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); + double *bd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); + double *cd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); double alphad = *alpha; double betad = *beta; double aod = *ao; @@ -1675,9 +1675,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran cblas_dgemm_wrapper(layout, transa, transb, *m, *n, *k, alphad, ad, *lda, bd, *ldb, betad, cd, *ldc); copy_mat(cd, layout, *m, *n, *ldc, offsetc, co, c); - oneapi::mkl::aligned_free(ad); - oneapi::mkl::aligned_free(bd); - oneapi::mkl::aligned_free(cd); + oneapi::math::aligned_free(ad); + oneapi::math::aligned_free(bd); + oneapi::math::aligned_free(cd); } template <> @@ -1698,9 +1698,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; sizec = *ldc * *m; } - double *ad = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizea); - double *bd = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizeb); - double *cd = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizec); + double *ad = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); + double *bd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); + double *cd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); double alphad = *alpha; double betad = *beta; double aod = *ao; @@ -1711,9 +1711,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran cblas_dgemm_wrapper(layout, transa, transb, *m, *n, *k, alphad, ad, *lda, bd, *ldb, betad, cd, *ldc); copy_mat(cd, layout, *m, *n, *ldc, offsetc, co, c); - oneapi::mkl::aligned_free(ad); - oneapi::mkl::aligned_free(bd); - oneapi::mkl::aligned_free(cd); + oneapi::math::aligned_free(ad); + oneapi::math::aligned_free(bd); + oneapi::math::aligned_free(cd); } template @@ -1730,12 +1730,12 @@ void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, // Not supported in NETLIB. SGEMM is used as reference. int sizec; sizec = *ldc * *n; - float *cf = (float *)oneapi::mkl::aligned_alloc(64, sizeof(float) * sizec); + float *cf = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizec); update_c(c, layout, upper_lower, *n, *n, *ldc, cf); cblas_sgemm_wrapper(layout, transa, transb, *n, *n, *k, *alpha, a, *lda, b, *ldb, *beta, cf, *ldc); update_c(cf, layout, upper_lower, *n, *n, *ldc, c); - oneapi::mkl::aligned_free(cf); + oneapi::math::aligned_free(cf); } template <> @@ -1746,12 +1746,12 @@ void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, // Not supported in NETLIB. DGEMM is used as reference. int sizec; sizec = *ldc * *n; - double *cf = (double *)oneapi::mkl::aligned_alloc(64, sizeof(double) * sizec); + double *cf = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); update_c(c, layout, upper_lower, *n, *n, *ldc, cf); cblas_dgemm_wrapper(layout, transa, transb, *n, *n, *k, *alpha, a, *lda, b, *ldb, *beta, cf, *ldc); update_c(cf, layout, upper_lower, *n, *n, *ldc, c); - oneapi::mkl::aligned_free(cf); + oneapi::math::aligned_free(cf); } template <> @@ -1764,12 +1764,12 @@ void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, int sizec; sizec = *ldc * *n; std::complex *cf = - (std::complex *)oneapi::mkl::aligned_alloc(64, sizeof(std::complex) * sizec); + (std::complex *)oneapi::math::aligned_alloc(64, sizeof(std::complex) * sizec); update_c(c, layout, upper_lower, *n, *n, *ldc, cf); cblas_cgemm_wrapper(layout, transa, transb, *n, *n, *k, alpha, a, *lda, b, *ldb, beta, cf, *ldc); update_c(cf, layout, upper_lower, *n, *n, *ldc, c); - oneapi::mkl::aligned_free(cf); + oneapi::math::aligned_free(cf); } template <> @@ -1781,13 +1781,13 @@ void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, // Not supported in NETLIB. ZGEMM is used as reference. int sizec; sizec = *ldc * *n; - std::complex *cf = (std::complex *)oneapi::mkl::aligned_alloc( + std::complex *cf = (std::complex *)oneapi::math::aligned_alloc( 64, sizeof(std::complex) * sizec); update_c(c, layout, upper_lower, *n, *n, *ldc, cf); cblas_zgemm_wrapper(layout, transa, transb, *n, *n, *k, alpha, a, *lda, b, *ldb, beta, cf, *ldc); update_c(cf, layout, upper_lower, *n, *n, *ldc, c); - oneapi::mkl::aligned_free(cf); + oneapi::math::aligned_free(cf); } template @@ -1978,10 +1978,10 @@ fp sametype_conj(fp x) { } template -void omatcopy_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int64_t m, int64_t n, +void omatcopy_ref(oneapi::math::layout layout, oneapi::math::transpose trans, int64_t m, int64_t n, fp alpha, fp *A, int64_t lda, fp *B, int64_t ldb) { int64_t logical_m, logical_n; - if (layout == oneapi::mkl::layout::col_major) { + if (layout == oneapi::math::layout::col_major) { logical_m = m; logical_n = n; } @@ -1989,14 +1989,14 @@ void omatcopy_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int6 logical_m = n; logical_n = m; } - if (trans == oneapi::mkl::transpose::nontrans) { + if (trans == oneapi::math::transpose::nontrans) { for (int64_t j = 0; j < logical_n; j++) { for (int64_t i = 0; i < logical_m; i++) { B[j * ldb + i] = alpha * A[j * lda + i]; } } } - else if (trans == oneapi::mkl::transpose::trans) { + else if (trans == oneapi::math::transpose::trans) { for (int64_t j = 0; j < logical_n; j++) { for (int64_t i = 0; i < logical_m; i++) { B[i * ldb + j] = alpha * A[j * lda + i]; @@ -2014,12 +2014,12 @@ void omatcopy_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int6 } template -void omatcopy2_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, const int64_t &m, +void omatcopy2_ref(oneapi::math::layout layout, oneapi::math::transpose trans, const int64_t &m, const int64_t &n, const fp &alpha, const fp *in_matrix, const int64_t &ld_in, const int64_t &inc_in, fp *out_matrix, const int64_t &ld_out, const int64_t inc_out) { int64_t logical_m, logical_n; - if (layout == oneapi::mkl::layout::col_major) { + if (layout == oneapi::math::layout::col_major) { logical_m = m; logical_n = n; } @@ -2027,7 +2027,7 @@ void omatcopy2_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, con logical_m = n; logical_n = m; } - if (trans == oneapi::mkl::transpose::trans) { + if (trans == oneapi::math::transpose::trans) { for (int64_t i = 0; i < logical_m; ++i) { for (int64_t j = 0; j < logical_n; ++j) { { @@ -2037,7 +2037,7 @@ void omatcopy2_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, con } } } - else if (trans == oneapi::mkl::transpose::nontrans) { + else if (trans == oneapi::math::transpose::nontrans) { for (int i = 0; i < logical_n; ++i) { for (int j = 0; j < logical_m; ++j) { { @@ -2060,10 +2060,10 @@ void omatcopy2_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, con } template -void imatcopy_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int64_t m, int64_t n, +void imatcopy_ref(oneapi::math::layout layout, oneapi::math::transpose trans, int64_t m, int64_t n, fp alpha, fp *A, int64_t lda, int64_t ldb) { int64_t logical_m, logical_n; - if (layout == oneapi::mkl::layout::col_major) { + if (layout == oneapi::math::layout::col_major) { logical_m = m; logical_n = n; } @@ -2072,16 +2072,16 @@ void imatcopy_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int6 logical_n = m; } std::vector temp(m * n); - int64_t ld_temp = (trans == oneapi::mkl::transpose::nontrans ? logical_m : logical_n); + int64_t ld_temp = (trans == oneapi::math::transpose::nontrans ? logical_m : logical_n); - if (trans == oneapi::mkl::transpose::nontrans) { + if (trans == oneapi::math::transpose::nontrans) { for (int64_t j = 0; j < logical_n; j++) { for (int64_t i = 0; i < logical_m; i++) { temp[j * ld_temp + i] = alpha * A[j * lda + i]; } } } - else if (trans == oneapi::mkl::transpose::trans) { + else if (trans == oneapi::math::transpose::trans) { for (int64_t j = 0; j < logical_n; j++) { for (int64_t i = 0; i < logical_m; i++) { temp[i * ld_temp + j] = alpha * A[j * lda + i]; @@ -2097,7 +2097,7 @@ void imatcopy_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int6 } } - if (trans == oneapi::mkl::transpose::nontrans) { + if (trans == oneapi::math::transpose::nontrans) { for (int64_t j = 0; j < logical_n; j++) { for (int64_t i = 0; i < logical_m; i++) { A[j * ldb + i] = temp[j * ld_temp + i]; @@ -2114,11 +2114,11 @@ void imatcopy_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int6 } template -void omatadd_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, int64_t m, int64_t n, fp alpha, fp *A, int64_t lda, +void omatadd_ref(oneapi::math::layout layout, oneapi::math::transpose transa, + oneapi::math::transpose transb, int64_t m, int64_t n, fp alpha, fp *A, int64_t lda, fp beta, fp *B, int64_t ldb, fp *C, int64_t ldc) { int64_t logical_m, logical_n; - if (layout == oneapi::mkl::layout::col_major) { + if (layout == oneapi::math::layout::col_major) { logical_m = m; logical_n = n; } @@ -2133,14 +2133,14 @@ void omatadd_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose transa, } } - if (transa == oneapi::mkl::transpose::nontrans) { + if (transa == oneapi::math::transpose::nontrans) { for (int64_t j = 0; j < logical_n; j++) { for (int64_t i = 0; i < logical_m; i++) { C[j * ldc + i] += alpha * A[j * lda + i]; } } } - else if (transa == oneapi::mkl::transpose::trans) { + else if (transa == oneapi::math::transpose::trans) { for (int64_t j = 0; j < logical_n; j++) { for (int64_t i = 0; i < logical_m; i++) { C[j * ldc + i] += alpha * A[i * lda + j]; @@ -2156,14 +2156,14 @@ void omatadd_ref(oneapi::mkl::layout layout, oneapi::mkl::transpose transa, } } - if (transb == oneapi::mkl::transpose::nontrans) { + if (transb == oneapi::math::transpose::nontrans) { for (int64_t j = 0; j < logical_n; j++) { for (int64_t i = 0; i < logical_m; i++) { C[j * ldc + i] += beta * B[j * ldb + i]; } } } - else if (transb == oneapi::mkl::transpose::trans) { + else if (transb == oneapi::math::transpose::trans) { for (int64_t j = 0; j < logical_n; j++) { for (int64_t i = 0; i < logical_m; i++) { C[j * ldc + i] += beta * B[i * ldb + j]; diff --git a/tests/unit_tests/blas/include/reference_blas_wrappers.hpp b/tests/unit_tests/blas/include/reference_blas_wrappers.hpp index 90186d8d8..3b86e1d33 100644 --- a/tests/unit_tests/blas/include/reference_blas_wrappers.hpp +++ b/tests/unit_tests/blas/include/reference_blas_wrappers.hpp @@ -43,7 +43,7 @@ static LIB_TYPE blas_library() { if (h_libblas == NULL) { h_libblas = GET_LIB_HANDLE(ONEMATH_REF_BLAS_LIBNAME); if (h_libblas == NULL) { - throw oneapi::mkl::library_not_found( + throw oneapi::math::library_not_found( "BLAS", "blas_library()", std::string("failed to load BLAS library ") + ONEMATH_REF_BLAS_LIBNAME); } @@ -56,7 +56,7 @@ static LIB_TYPE cblas_library() { if (h_libcblas == NULL) { h_libcblas = GET_LIB_HANDLE(ONEMATH_REF_CBLAS_LIBNAME); if (h_libcblas == NULL) { - throw oneapi::mkl::library_not_found( + throw oneapi::math::library_not_found( "BLAS", "cblas_library()", std::string("failed to load CBLAS library ") + ONEMATH_REF_CBLAS_LIBNAME); } diff --git a/tests/unit_tests/blas/include/test_common.hpp b/tests/unit_tests/blas/include/test_common.hpp index 5d607991e..25620f4fb 100644 --- a/tests/unit_tests/blas/include/test_common.hpp +++ b/tests/unit_tests/blas/include/test_common.hpp @@ -67,20 +67,20 @@ constexpr int num_components() { // Matrix helpers. template -constexpr T inner_dimension(oneapi::mkl::transpose trans, T m, T n) { - return (trans == oneapi::mkl::transpose::nontrans) ? m : n; +constexpr T inner_dimension(oneapi::math::transpose trans, T m, T n) { + return (trans == oneapi::math::transpose::nontrans) ? m : n; } template -constexpr T outer_dimension(oneapi::mkl::transpose trans, T m, T n) { - return (trans == oneapi::mkl::transpose::nontrans) ? n : m; +constexpr T outer_dimension(oneapi::math::transpose trans, T m, T n) { + return (trans == oneapi::math::transpose::nontrans) ? n : m; } template -constexpr T matrix_size(oneapi::mkl::transpose trans, T m, T n, T ldm) { +constexpr T matrix_size(oneapi::math::transpose trans, T m, T n, T ldm) { return outer_dimension(trans, m, n) * ldm; } template -constexpr T matrix_size(oneapi::mkl::layout layout, oneapi::mkl::transpose trans, T m, T n, T ldm) { - return (layout == oneapi::mkl::layout::col_major) ? outer_dimension(trans, m, n) * ldm +constexpr T matrix_size(oneapi::math::layout layout, oneapi::math::transpose trans, T m, T n, T ldm) { + return (layout == oneapi::math::layout::col_major) ? outer_dimension(trans, m, n) * ldm : inner_dimension(trans, m, n) * ldm; } @@ -192,28 +192,28 @@ void rand_vector(vec &v, int n, int inc) { } template -oneapi::mkl::transpose rand_trans() { +oneapi::math::transpose rand_trans() { std::int64_t tmp; - oneapi::mkl::transpose trans; + oneapi::math::transpose trans; if ((std::is_same::value) || (std::is_same::value)) { - trans = (oneapi::mkl::transpose)(std::rand() % 2); + trans = (oneapi::math::transpose)(std::rand() % 2); } else { tmp = std::rand() % 3; if (tmp == 2) - trans = oneapi::mkl::transpose::conjtrans; + trans = oneapi::math::transpose::conjtrans; else - trans = (oneapi::mkl::transpose)tmp; + trans = (oneapi::math::transpose)tmp; } return trans; } template -void print_matrix(vec &M, oneapi::mkl::transpose trans, int m, int n, int ld, char *name) { +void print_matrix(vec &M, oneapi::math::transpose trans, int m, int n, int ld, char *name) { std::cout << "Matrix " << name << ":\n"; for (int i = 0; i < m; i++) { for (int j = 0; j < n; j++) { - if (trans == oneapi::mkl::transpose::nontrans) + if (trans == oneapi::math::transpose::nontrans) std::cout << (double)M[i + j * ld] << " "; else std::cout << (double)M[j + i * ld] << " "; @@ -230,14 +230,14 @@ void copy_vector(fp *src, int n, int inc, fp *dest) { } template -void copy_matrix(vec_src &src, oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int m, +void copy_matrix(vec_src &src, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, int ld, vec_dest &dest) { using T_data = typename vec_dest::value_type; dest.resize(matrix_size(layout, trans, m, n, ld)); - if (((trans == oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::col_major)) || - ((trans != oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::row_major))) { + if (((trans == oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::col_major)) || + ((trans != oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::row_major))) { for (int j = 0; j < n; j++) for (int i = 0; i < m; i++) dest[i + j * ld] = (T_data)src[i + j * ld]; @@ -250,12 +250,12 @@ void copy_matrix(vec_src &src, oneapi::mkl::layout layout, oneapi::mkl::transpos } template -void copy_matrix(fp_src *src, oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int m, +void copy_matrix(fp_src *src, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, int ld, fp_dst *dest) { - if (((trans == oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::col_major)) || - ((trans != oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::row_major))) { + if (((trans == oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::col_major)) || + ((trans != oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::row_major))) { for (int j = 0; j < n; j++) for (int i = 0; i < m; i++) dest[i + j * ld] = (fp_dst)src[i + j * ld]; @@ -268,12 +268,12 @@ void copy_matrix(fp_src *src, oneapi::mkl::layout layout, oneapi::mkl::transpose } template -void rand_matrix(vec &M, oneapi::mkl::transpose trans, int m, int n, int ld) { +void rand_matrix(vec &M, oneapi::math::transpose trans, int m, int n, int ld) { using fp = typename vec::value_type; M.resize(matrix_size(trans, m, n, ld)); - if (trans == oneapi::mkl::transpose::nontrans) { + if (trans == oneapi::math::transpose::nontrans) { for (int j = 0; j < n; j++) for (int i = 0; i < m; i++) M[i + j * ld] = rand_scalar(); @@ -286,16 +286,16 @@ void rand_matrix(vec &M, oneapi::mkl::transpose trans, int m, int n, int ld) { } template -void rand_matrix(vec &M, oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int m, int n, +void rand_matrix(vec &M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, int ld) { using fp = typename vec::value_type; M.resize(matrix_size(layout, trans, m, n, ld)); - if (((trans == oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::col_major)) || - ((trans != oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::row_major))) { + if (((trans == oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::col_major)) || + ((trans != oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::row_major))) { for (int j = 0; j < n; j++) for (int i = 0; i < m; i++) M[i + j * ld] = rand_scalar(); @@ -308,12 +308,12 @@ void rand_matrix(vec &M, oneapi::mkl::layout layout, oneapi::mkl::transpose tran } template -void rand_matrix(fp *M, oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int m, int n, +void rand_matrix(fp *M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, int ld) { - if (((trans == oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::col_major)) || - ((trans != oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::row_major))) { + if (((trans == oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::col_major)) || + ((trans != oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::row_major))) { for (int j = 0; j < n; j++) for (int i = 0; i < m; i++) M[i + j * ld] = rand_scalar(); @@ -326,16 +326,16 @@ void rand_matrix(fp *M, oneapi::mkl::layout layout, oneapi::mkl::transpose trans } template -void rand_trsm_matrix(vec &M, oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int m, +void rand_trsm_matrix(vec &M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, int ld) { using fp = typename vec::value_type; M.resize(matrix_size(layout, trans, m, n, ld)); - if (((trans == oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::col_major)) || - ((trans != oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::row_major))) { + if (((trans == oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::col_major)) || + ((trans != oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::row_major))) { for (int j = 0; j < n; j++) for (int i = 0; i < m; i++) { if (i == j) @@ -356,12 +356,12 @@ void rand_trsm_matrix(vec &M, oneapi::mkl::layout layout, oneapi::mkl::transpose } template -void rand_trsm_matrix(fp *M, oneapi::mkl::layout layout, oneapi::mkl::transpose trans, int m, int n, +void rand_trsm_matrix(fp *M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, int ld) { - if (((trans == oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::col_major)) || - ((trans != oneapi::mkl::transpose::nontrans) && - (layout == oneapi::mkl::layout::row_major))) { + if (((trans == oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::col_major)) || + ((trans != oneapi::math::transpose::nontrans) && + (layout == oneapi::math::layout::row_major))) { for (int j = 0; j < n; j++) for (int i = 0; i < m; i++) { if (i == j) @@ -382,8 +382,8 @@ void rand_trsm_matrix(fp *M, oneapi::mkl::layout layout, oneapi::mkl::transpose } template -void rand_tpsv_matrix(vec &M, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, int m) { +void rand_tpsv_matrix(vec &M, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, int m) { using fp = typename vec::value_type; std::vector tmp; int start, end, i, j, k = 0; @@ -392,13 +392,13 @@ void rand_tpsv_matrix(vec &M, oneapi::mkl::layout layout, oneapi::mkl::uplo uppe M.resize((m * (m + 1)) / 2); for (j = 0; j < m; j++) { - if (layout == oneapi::mkl::layout::col_major) { - start = (upper_lower == oneapi::mkl::uplo::U) ? 0 : j; - end = (upper_lower == oneapi::mkl::uplo::U) ? j : m - 1; + if (layout == oneapi::math::layout::col_major) { + start = (upper_lower == oneapi::math::uplo::U) ? 0 : j; + end = (upper_lower == oneapi::math::uplo::U) ? j : m - 1; } else { - start = (upper_lower == oneapi::mkl::uplo::U) ? j : 0; - end = (upper_lower == oneapi::mkl::uplo::U) ? m - 1 : j; + start = (upper_lower == oneapi::math::uplo::U) ? j : 0; + end = (upper_lower == oneapi::math::uplo::U) ? m - 1 : j; } for (i = start; i <= end; i++) { M[k] = tmp[i + j * m]; @@ -408,8 +408,8 @@ void rand_tpsv_matrix(vec &M, oneapi::mkl::layout layout, oneapi::mkl::uplo uppe } template -void rand_tbsv_matrix(vec &M, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, int m, int k, int ld) { +void rand_tbsv_matrix(vec &M, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, int m, int k, int ld) { using fp = typename vec::value_type; std::vector tmp; int i, j, n; @@ -417,8 +417,8 @@ void rand_tbsv_matrix(vec &M, oneapi::mkl::layout layout, oneapi::mkl::uplo uppe rand_trsm_matrix(tmp, layout, trans, m, m, ld); M.resize(matrix_size(layout, trans, m, m, ld)); - if (((layout == oneapi::mkl::layout::col_major) && (upper_lower == oneapi::mkl::uplo::U)) || - ((layout == oneapi::mkl::layout::row_major) && (upper_lower == oneapi::mkl::uplo::L))) { + if (((layout == oneapi::math::layout::col_major) && (upper_lower == oneapi::math::uplo::U)) || + ((layout == oneapi::math::layout::row_major) && (upper_lower == oneapi::math::uplo::L))) { for (j = 0; j < m; j++) { n = k - j; for (i = std::max(0, j - k); i <= j; i++) { @@ -564,13 +564,13 @@ bool check_equal_trsv_vector(vec1 &v, vec2 &v_ref, int n, int inc, int error_mag } template -bool check_equal_matrix(acc1 &M, acc2 &M_ref, oneapi::mkl::layout layout, int m, int n, int ld, +bool check_equal_matrix(acc1 &M, acc2 &M_ref, oneapi::math::layout layout, int m, int n, int ld, int error_mag, std::ostream &out) { bool good = true; int idx, count = 0; for (int j = 0; j < n; j++) { for (int i = 0; i < m; i++) { - idx = (layout == oneapi::mkl::layout::col_major) ? i + j * ld : j + i * ld; + idx = (layout == oneapi::math::layout::col_major) ? i + j * ld : j + i * ld; if (!check_equal(M[idx], M_ref[idx], error_mag)) { out << "Difference in entry (" << i << ',' << j << "): DPC++ " << M[idx] << " vs. Reference " << M_ref[idx] << std::endl; @@ -586,13 +586,13 @@ bool check_equal_matrix(acc1 &M, acc2 &M_ref, oneapi::mkl::layout layout, int m, } template -bool check_equal_matrix(const fp *M, const fp *M_ref, oneapi::mkl::layout layout, int m, int n, +bool check_equal_matrix(const fp *M, const fp *M_ref, oneapi::math::layout layout, int m, int n, int ld, int error_mag, std::ostream &out) { bool good = true; int idx, count = 0; for (int j = 0; j < n; j++) { for (int i = 0; i < m; i++) { - idx = (layout == oneapi::mkl::layout::col_major) ? i + j * ld : j + i * ld; + idx = (layout == oneapi::math::layout::col_major) ? i + j * ld : j + i * ld; if (!check_equal(M[idx], M_ref[idx], error_mag)) { out << "Difference in entry (" << i << ',' << j << "): DPC++ " << M[idx] << " vs. Reference " << M_ref[idx] << std::endl; @@ -608,16 +608,16 @@ bool check_equal_matrix(const fp *M, const fp *M_ref, oneapi::mkl::layout layout } template -bool check_equal_matrix(acc1 &M, acc2 &M_ref, oneapi::mkl::layout layout, - oneapi::mkl::uplo upper_lower, int m, int n, int ld, int error_mag, +bool check_equal_matrix(acc1 &M, acc2 &M_ref, oneapi::math::layout layout, + oneapi::math::uplo upper_lower, int m, int n, int ld, int error_mag, std::ostream &out) { bool good = true; int idx, count = 0; for (int j = 0; j < n; j++) { for (int i = 0; i < m; i++) { - idx = (layout == oneapi::mkl::layout::col_major) ? i + j * ld : j + i * ld; - if (((upper_lower == oneapi::mkl::uplo::upper) && (j >= i)) || - ((upper_lower == oneapi::mkl::uplo::lower) && (j <= i))) { + idx = (layout == oneapi::math::layout::col_major) ? i + j * ld : j + i * ld; + if (((upper_lower == oneapi::math::uplo::upper) && (j >= i)) || + ((upper_lower == oneapi::math::uplo::lower) && (j <= i))) { if (!check_equal(M[idx], M_ref[idx], error_mag)) { out << "Difference in entry (" << i << ',' << j << "): DPC++ " << M[idx] << " vs. Reference " << M_ref[idx] << std::endl; @@ -634,13 +634,13 @@ bool check_equal_matrix(acc1 &M, acc2 &M_ref, oneapi::mkl::layout layout, } template -bool check_equal_trsm_matrix(acc1 &M, acc2 &M_ref, oneapi::mkl::layout layout, int m, int n, int ld, +bool check_equal_trsm_matrix(acc1 &M, acc2 &M_ref, oneapi::math::layout layout, int m, int n, int ld, int error_mag, std::ostream &out) { bool good = true; int idx, count = 0; for (int j = 0; j < n; j++) { for (int i = 0; i < m; i++) { - idx = (layout == oneapi::mkl::layout::col_major) ? i + j * ld : j + i * ld; + idx = (layout == oneapi::math::layout::col_major) ? i + j * ld : j + i * ld; if (!check_equal_trsm(M[idx], M_ref[idx], error_mag)) { out << "Difference in entry (" << i << ',' << j << "): DPC++ " << M[idx] << " vs. Reference " << M_ref[idx] << std::endl; @@ -677,14 +677,14 @@ typename std::enable_if::value, bool>::type check_almost_eq } template -bool check_almost_equal_matrix_int(Ta &M, Tb &M_ref, oneapi::mkl::layout layout, int m, int n, +bool check_almost_equal_matrix_int(Ta &M, Tb &M_ref, oneapi::math::layout layout, int m, int n, int ld, int error_mag, std::ostream &out) { static_assert(is_matrix_type_integral() && is_matrix_type_integral()); bool good = true; int idx, count = 0; for (int j = 0; j < n; j++) { for (int i = 0; i < m; i++) { - idx = (layout == oneapi::mkl::layout::col_major) ? i + j * ld : j + i * ld; + idx = (layout == oneapi::math::layout::col_major) ? i + j * ld : j + i * ld; if (!check_almost_equal_int(M[idx], M_ref[idx], error_mag)) { out << "Difference in entry (" << i << ',' << j << "): DPC++ " << M[idx] << " vs. Reference " << M_ref[idx] << std::endl; @@ -700,7 +700,7 @@ bool check_almost_equal_matrix_int(Ta &M, Tb &M_ref, oneapi::mkl::layout layout, } template -bool check_almost_equal_matrix(Ta &M, Tb &M_ref, oneapi::mkl::layout layout, int m, int n, int ld, +bool check_almost_equal_matrix(Ta &M, Tb &M_ref, oneapi::math::layout layout, int m, int n, int ld, int error_mag, std::ostream &out) { // Only call if returned dtype is integral if constexpr (is_matrix_type_integral() && is_matrix_type_integral()) diff --git a/tests/unit_tests/blas/level1/asum.cpp b/tests/unit_tests/blas/level1/asum.cpp index d0a9869d6..60263989b 100644 --- a/tests/unit_tests/blas/level1/asum.cpp +++ b/tests/unit_tests/blas/level1/asum.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int64_t N, int64_t incx) { +int test(device* dev, oneapi::math::layout layout, int64_t N, int64_t incx) { // Prepare data. vector x; fp_res result = fp_res(-1), result_ref = fp_res(-1); @@ -82,22 +82,22 @@ int test(device* dev, oneapi::mkl::layout layout, int64_t N, int64_t incx) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::asum(main_queue, N, x_buffer, incx, result_buffer); + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::asum(main_queue, N, x_buffer, incx, result_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::asum(main_queue, N, x_buffer, incx, result_buffer); + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::asum(main_queue, N, x_buffer, incx, result_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::asum, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::asum, N, x_buffer, incx, result_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::asum, N, x_buffer, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::asum, N, x_buffer, incx, result_buffer); break; default: break; @@ -109,7 +109,7 @@ int test(device* dev, oneapi::mkl::layout layout, int64_t N, int64_t incx) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -125,7 +125,7 @@ int test(device* dev, oneapi::mkl::layout layout, int64_t N, int64_t incx) { return (int)good; } -class AsumTests : public ::testing::TestWithParam> { +class AsumTests : public ::testing::TestWithParam> { }; TEST_P(AsumTests, RealSinglePrecision) { @@ -170,8 +170,8 @@ TEST_P(AsumTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(AsumTestSuite, AsumTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/asum_usm.cpp b/tests/unit_tests/blas/level1/asum_usm.cpp index 076de8f0b..8d15c1407 100644 --- a/tests/unit_tests/blas/level1/asum_usm.cpp +++ b/tests/unit_tests/blas/level1/asum_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int64_t N, int64_t incx) { +int test(device* dev, oneapi::math::layout layout, int64_t N, int64_t incx) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -83,10 +83,10 @@ int test(device* dev, oneapi::mkl::layout layout, int64_t N, int64_t incx) { fp_res* result_p; if constexpr (alloc_type == usm::alloc::shared) { - result_p = (fp_res*)oneapi::mkl::malloc_shared(64, sizeof(fp_res), *dev, cxt); + result_p = (fp_res*)oneapi::math::malloc_shared(64, sizeof(fp_res), *dev, cxt); } else if constexpr (alloc_type == usm::alloc::device) { - result_p = (fp_res*)oneapi::mkl::malloc_device(64, sizeof(fp_res), *dev, cxt); + result_p = (fp_res*)oneapi::math::malloc_device(64, sizeof(fp_res), *dev, cxt); } else { throw std::runtime_error("Bad alloc_type"); @@ -95,12 +95,12 @@ int test(device* dev, oneapi::mkl::layout layout, int64_t N, int64_t incx) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::asum(main_queue, N, x.data(), incx, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::asum(main_queue, N, x.data(), incx, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::asum(main_queue, N, x.data(), incx, result_p, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::asum(main_queue, N, x.data(), incx, result_p, dependencies); break; default: break; @@ -108,12 +108,12 @@ int test(device* dev, oneapi::mkl::layout layout, int64_t N, int64_t incx) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::asum, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::asum, N, x.data(), incx, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::asum, N, x.data(), + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::asum, N, x.data(), incx, result_p, dependencies); break; default: break; @@ -126,7 +126,7 @@ int test(device* dev, oneapi::mkl::layout layout, int64_t N, int64_t incx) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -138,13 +138,13 @@ int test(device* dev, oneapi::mkl::layout layout, int64_t N, int64_t incx) { bool good = check_equal_ptr(main_queue, result_p, result_ref, N, std::cout); - oneapi::mkl::free_usm(result_p, cxt); + oneapi::math::free_usm(result_p, cxt); return (int)good; } class AsumUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AsumUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP( @@ -196,8 +196,8 @@ TEST_P(AsumUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(AsumUsmTestSuite, AsumUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/axpby.cpp b/tests/unit_tests/blas/level1/axpby.cpp index df3550efe..6b417e913 100644 --- a/tests/unit_tests/blas/level1/axpby.cpp +++ b/tests/unit_tests/blas/level1/axpby.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp alpha, fp beta) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha, fp beta) { // Prepare data. vector x, y, y_ref; @@ -85,24 +85,24 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::axpby(main_queue, N, alpha, x_buffer, incx, beta, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::axpby(main_queue, N, alpha, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::axpby(main_queue, N, alpha, x_buffer, incx, beta, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::axpby(main_queue, N, alpha, x_buffer, incx, beta, y_buffer, incy); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::axpby, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpby, N, alpha, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::axpby, N, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::axpby, N, alpha, x_buffer, incx, beta, y_buffer, incy); break; default: break; @@ -114,7 +114,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -131,7 +131,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp } class AxpbyTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpbyTests, RealSinglePrecision) { float alpha(2.0); @@ -180,8 +180,8 @@ TEST_P(AxpbyTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(AxpbyTestSuite, AxpbyTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/axpby_usm.cpp b/tests/unit_tests/blas/level1/axpby_usm.cpp index 348731324..152879329 100644 --- a/tests/unit_tests/blas/level1/axpby_usm.cpp +++ b/tests/unit_tests/blas/level1/axpby_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp alpha, fp beta) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha, fp beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -87,12 +87,12 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::axpby(main_queue, N, alpha, x.data(), incx, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::axpby(main_queue, N, alpha, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::axpby(main_queue, N, alpha, x.data(), incx, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::axpby(main_queue, N, alpha, x.data(), incx, beta, y.data(), incy, dependencies); break; default: break; @@ -100,12 +100,12 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::axpby, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpby, N, alpha, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::axpby, N, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::axpby, N, alpha, x.data(), incx, beta, y.data(), incy, dependencies); break; default: break; @@ -118,7 +118,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -134,7 +134,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp } class AxpbyUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpbyUsmTests, RealSinglePrecision) { float alpha(2.0); @@ -183,8 +183,8 @@ TEST_P(AxpbyUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(AxpbyUsmTestSuite, AxpbyUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/axpy.cpp b/tests/unit_tests/blas/level1/axpy.cpp index ebcb6476c..dc38cb718 100644 --- a/tests/unit_tests/blas/level1/axpy.cpp +++ b/tests/unit_tests/blas/level1/axpy.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp alpha) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha) { // Prepare data. vector x, y, y_ref; @@ -85,24 +85,24 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::axpy(main_queue, N, alpha, x_buffer, incx, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::axpy(main_queue, N, alpha, x_buffer, incx, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::axpy(main_queue, N, alpha, x_buffer, incx, y_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::axpy(main_queue, N, alpha, x_buffer, incx, y_buffer, incy); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::axpy, N, alpha, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpy, N, alpha, x_buffer, incx, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::axpy, N, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::axpy, N, alpha, x_buffer, incx, y_buffer, incy); break; default: break; @@ -114,7 +114,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -130,7 +130,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp return (int)good; } -class AxpyTests : public ::testing::TestWithParam> { +class AxpyTests : public ::testing::TestWithParam> { }; TEST_P(AxpyTests, RealSinglePrecision) { @@ -176,8 +176,8 @@ TEST_P(AxpyTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(AxpyTestSuite, AxpyTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/axpy_usm.cpp b/tests/unit_tests/blas/level1/axpy_usm.cpp index 5bcf9874d..d24c500a6 100644 --- a/tests/unit_tests/blas/level1/axpy_usm.cpp +++ b/tests/unit_tests/blas/level1/axpy_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp alpha) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -87,12 +87,12 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::axpy(main_queue, N, alpha, x.data(), incx, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::axpy(main_queue, N, alpha, x.data(), incx, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::axpy(main_queue, N, alpha, x.data(), incx, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::axpy(main_queue, N, alpha, x.data(), incx, y.data(), incy, dependencies); break; default: break; @@ -100,12 +100,12 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::axpy, N, alpha, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpy, N, alpha, x.data(), incx, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::axpy, N, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::axpy, N, alpha, x.data(), incx, y.data(), incy, dependencies); break; default: break; @@ -118,7 +118,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -134,7 +134,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp } class AxpyUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpyUsmTests, RealSinglePrecision) { float alpha(2.0); @@ -179,8 +179,8 @@ TEST_P(AxpyUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(AxpyUsmTestSuite, AxpyUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/copy.cpp b/tests/unit_tests/blas/level1/copy.cpp index f5e2c81a6..9c2a2d54e 100644 --- a/tests/unit_tests/blas/level1/copy.cpp +++ b/tests/unit_tests/blas/level1/copy.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { // Prepare data. vector x, y, y_ref; @@ -84,23 +84,23 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::copy(main_queue, N, x_buffer, incx, y_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::copy(main_queue, N, x_buffer, incx, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::copy(main_queue, N, x_buffer, incx, y_buffer, incy); + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::copy(main_queue, N, x_buffer, incx, y_buffer, incy); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::copy, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::copy, N, x_buffer, incx, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::copy, N, x_buffer, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::copy, N, x_buffer, incx, y_buffer, incy); break; default: break; @@ -112,7 +112,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -128,7 +128,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { return (int)good; } -class CopyTests : public ::testing::TestWithParam> { +class CopyTests : public ::testing::TestWithParam> { }; TEST_P(CopyTests, RealSinglePrecision) { @@ -164,8 +164,8 @@ TEST_P(CopyTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(CopyTestSuite, CopyTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/copy_usm.cpp b/tests/unit_tests/blas/level1/copy_usm.cpp index ac457acf9..1243b493f 100644 --- a/tests/unit_tests/blas/level1/copy_usm.cpp +++ b/tests/unit_tests/blas/level1/copy_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -86,12 +86,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::copy(main_queue, N, x.data(), incx, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::copy(main_queue, N, x.data(), incx, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::copy(main_queue, N, x.data(), incx, y.data(), + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::copy(main_queue, N, x.data(), incx, y.data(), incy, dependencies); break; default: break; @@ -99,12 +99,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::copy, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::copy, N, x.data(), incx, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::copy, N, x.data(), + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::copy, N, x.data(), incx, y.data(), incy, dependencies); break; default: break; @@ -117,7 +117,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -133,7 +133,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { } class CopyUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(CopyUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 1357, 2, 3)); @@ -168,8 +168,8 @@ TEST_P(CopyUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(CopyUsmTestSuite, CopyUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/dot.cpp b/tests/unit_tests/blas/level1/dot.cpp index 0ac335641..b00e83a69 100644 --- a/tests/unit_tests/blas/level1/dot.cpp +++ b/tests/unit_tests/blas/level1/dot.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { // Prepare data. vector x, y; fp_res result = fp_res(-1), result_ref = fp_res(-1); @@ -84,24 +84,24 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::dot(main_queue, N, x_buffer, incx, y_buffer, incy, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::dot(main_queue, N, x_buffer, incx, y_buffer, incy, result_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::dot(main_queue, N, x_buffer, incx, y_buffer, incy, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::dot(main_queue, N, x_buffer, incx, y_buffer, incy, result_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::dot, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::dot, N, x_buffer, incx, y_buffer, incy, result_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::dot, N, x_buffer, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dot, N, x_buffer, incx, y_buffer, incy, result_buffer); break; default: break; @@ -113,7 +113,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -129,7 +129,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { return (int)good; } -class DotTests : public ::testing::TestWithParam> {}; +class DotTests : public ::testing::TestWithParam> {}; TEST_P(DotTests, RealSinglePrecision) { EXPECT_TRUEORSKIP( @@ -161,8 +161,8 @@ TEST_P(DotTests, RealDoubleSinglePrecision) { INSTANTIATE_TEST_SUITE_P(DotTestSuite, DotTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/dot_usm.cpp b/tests/unit_tests/blas/level1/dot_usm.cpp index 94ad1c008..8d17272c1 100644 --- a/tests/unit_tests/blas/level1/dot_usm.cpp +++ b/tests/unit_tests/blas/level1/dot_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -83,10 +83,10 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { fp_res* result_p; if constexpr (alloc_type == usm::alloc::shared) { - result_p = (fp_res*)oneapi::mkl::malloc_shared(64, sizeof(fp_res), *dev, cxt); + result_p = (fp_res*)oneapi::math::malloc_shared(64, sizeof(fp_res), *dev, cxt); } else if constexpr (alloc_type == usm::alloc::device) { - result_p = (fp_res*)oneapi::mkl::malloc_device(64, sizeof(fp_res), *dev, cxt); + result_p = (fp_res*)oneapi::math::malloc_device(64, sizeof(fp_res), *dev, cxt); } else { throw std::runtime_error("Bad alloc_type"); @@ -95,12 +95,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::dot(main_queue, N, x.data(), incx, y.data(), + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::dot(main_queue, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::dot(main_queue, N, x.data(), incx, y.data(), + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::dot(main_queue, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; default: break; @@ -108,12 +108,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::dot, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::dot, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::dot, N, x.data(), + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dot, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; default: break; @@ -126,7 +126,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -137,13 +137,13 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { // Compare the results of reference implementation and DPC++ implementation. bool good = check_equal_ptr(main_queue, result_p, result_ref, N, std::cout); - oneapi::mkl::free_usm(result_p, cxt); + oneapi::math::free_usm(result_p, cxt); return (int)good; } class DotUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(DotUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP( @@ -181,8 +181,8 @@ TEST_P(DotUsmTests, RealDoubleSinglePrecision) { INSTANTIATE_TEST_SUITE_P(DotUsmTestSuite, DotUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/dotc.cpp b/tests/unit_tests/blas/level1/dotc.cpp index 544e007ea..96ae89f73 100644 --- a/tests/unit_tests/blas/level1/dotc.cpp +++ b/tests/unit_tests/blas/level1/dotc.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { // Prepare data. vector x, y; fp result = 0.0, result_reference = 0.0; @@ -86,24 +86,24 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::dotc(main_queue, N, x_buffer, incx, y_buffer, incy, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::dotc(main_queue, N, x_buffer, incx, y_buffer, incy, result_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::dotc(main_queue, N, x_buffer, incx, y_buffer, incy, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::dotc(main_queue, N, x_buffer, incx, y_buffer, incy, result_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::dotc, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::dotc, N, x_buffer, incx, y_buffer, incy, result_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::dotc, N, x_buffer, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotc, N, x_buffer, incx, y_buffer, incy, result_buffer); break; default: break; @@ -115,7 +115,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -131,7 +131,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { return (int)good; } -class DotcTests : public ::testing::TestWithParam> { +class DotcTests : public ::testing::TestWithParam> { }; TEST_P(DotcTests, ComplexSinglePrecision) { @@ -155,8 +155,8 @@ TEST_P(DotcTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(DotcTestSuite, DotcTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/dotc_usm.cpp b/tests/unit_tests/blas/level1/dotc_usm.cpp index f24b5e222..3148bd3f2 100644 --- a/tests/unit_tests/blas/level1/dotc_usm.cpp +++ b/tests/unit_tests/blas/level1/dotc_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -83,17 +83,17 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { // Call DPC++ DOTC. - auto result_p = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp), *dev, cxt); + auto result_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::dotc( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::dotc( main_queue, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::dotc(main_queue, N, x.data(), incx, y.data(), + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::dotc(main_queue, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; default: break; @@ -101,12 +101,12 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::dotc, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::dotc, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::dotc, N, x.data(), + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotc, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; default: break; @@ -119,7 +119,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -131,13 +131,13 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { bool good = check_equal(*result_p, result_reference, N, std::cout); - oneapi::mkl::free_shared(result_p, cxt); + oneapi::math::free_shared(result_p, cxt); return (int)good; } class DotcUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(DotcUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP( @@ -160,8 +160,8 @@ TEST_P(DotcUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(DotcUsmTestSuite, DotcUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/dotu.cpp b/tests/unit_tests/blas/level1/dotu.cpp index 6f3163da5..ded168de0 100644 --- a/tests/unit_tests/blas/level1/dotu.cpp +++ b/tests/unit_tests/blas/level1/dotu.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { // Prepare data. vector x, y; fp result = 0.0, result_reference = 0.0; @@ -86,24 +86,24 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::dotu(main_queue, N, x_buffer, incx, y_buffer, incy, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::dotu(main_queue, N, x_buffer, incx, y_buffer, incy, result_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::dotu(main_queue, N, x_buffer, incx, y_buffer, incy, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::dotu(main_queue, N, x_buffer, incx, y_buffer, incy, result_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::dotu, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::dotu, N, x_buffer, incx, y_buffer, incy, result_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::dotu, N, x_buffer, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotu, N, x_buffer, incx, y_buffer, incy, result_buffer); break; default: break; @@ -115,7 +115,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -131,7 +131,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { return (int)good; } -class DotuTests : public ::testing::TestWithParam> { +class DotuTests : public ::testing::TestWithParam> { }; TEST_P(DotuTests, ComplexSinglePrecision) { @@ -155,8 +155,8 @@ TEST_P(DotuTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(DotuTestSuite, DotuTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/dotu_usm.cpp b/tests/unit_tests/blas/level1/dotu_usm.cpp index c67871c2c..889e01705 100644 --- a/tests/unit_tests/blas/level1/dotu_usm.cpp +++ b/tests/unit_tests/blas/level1/dotu_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -83,17 +83,17 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { // Call DPC++ DOTU. - auto result_p = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp), *dev, cxt); + auto result_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::dotu( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::dotu( main_queue, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::dotu(main_queue, N, x.data(), incx, y.data(), + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::dotu(main_queue, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; default: break; @@ -101,12 +101,12 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::dotu, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::dotu, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::dotu, N, x.data(), + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotu, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; default: break; @@ -119,7 +119,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -130,13 +130,13 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy) { // Compare the results of reference implementation and DPC++ implementation. bool good = check_equal(*result_p, result_reference, N, std::cout); - oneapi::mkl::free_shared(result_p, cxt); + oneapi::math::free_shared(result_p, cxt); return (int)good; } class DotuUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(DotuUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP( @@ -159,8 +159,8 @@ TEST_P(DotuUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(DotuUsmTestSuite, DotuUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/iamax.cpp b/tests/unit_tests/blas/level1/iamax.cpp index 41d79748c..19672e341 100644 --- a/tests/unit_tests/blas/level1/iamax.cpp +++ b/tests/unit_tests/blas/level1/iamax.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { +int test(device* dev, oneapi::math::layout layout, int N, int incx) { // Prepare data. vector x; int64_t result = -1, result_ref = -1; @@ -82,23 +82,23 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::iamax(main_queue, N, x_buffer, incx, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::iamax(main_queue, N, x_buffer, incx, result_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::iamax(main_queue, N, x_buffer, incx, result_buffer); + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::iamax(main_queue, N, x_buffer, incx, result_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::iamax, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::iamax, N, x_buffer, incx, result_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::iamax, N, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::iamax, N, x_buffer, incx, result_buffer); break; default: break; @@ -110,7 +110,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -126,7 +126,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { return (int)good; } -class IamaxTests : public ::testing::TestWithParam> { +class IamaxTests : public ::testing::TestWithParam> { }; TEST_P(IamaxTests, RealSinglePrecision) { @@ -162,8 +162,8 @@ TEST_P(IamaxTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(IamaxTestSuite, IamaxTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/iamax_usm.cpp b/tests/unit_tests/blas/level1/iamax_usm.cpp index 86e8ab9fa..88029f248 100644 --- a/tests/unit_tests/blas/level1/iamax_usm.cpp +++ b/tests/unit_tests/blas/level1/iamax_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { +int test(device* dev, oneapi::math::layout layout, int N, int incx) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -82,10 +82,10 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { int64_t* result_p; if constexpr (alloc_type == usm::alloc::shared) { - result_p = (int64_t*)oneapi::mkl::malloc_shared(64, sizeof(int64_t), *dev, cxt); + result_p = (int64_t*)oneapi::math::malloc_shared(64, sizeof(int64_t), *dev, cxt); } else if constexpr (alloc_type == usm::alloc::device) { - result_p = (int64_t*)oneapi::mkl::malloc_device(64, sizeof(int64_t), *dev, cxt); + result_p = (int64_t*)oneapi::math::malloc_device(64, sizeof(int64_t), *dev, cxt); } else { throw std::runtime_error("Bad alloc_type"); @@ -94,12 +94,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::iamax(main_queue, N, x.data(), incx, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::iamax(main_queue, N, x.data(), incx, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::iamax(main_queue, N, x.data(), incx, result_p, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::iamax(main_queue, N, x.data(), incx, result_p, dependencies); break; default: break; @@ -107,12 +107,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::iamax, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::iamax, N, x.data(), incx, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::iamax, N, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::iamax, N, x.data(), incx, result_p, dependencies); break; default: break; @@ -125,7 +125,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -136,13 +136,13 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { // Compare the results of reference implementation and DPC++ implementation. bool good = check_equal_ptr(main_queue, result_p, result_ref, 0, std::cout); - oneapi::mkl::free_usm(result_p, cxt); + oneapi::math::free_usm(result_p, cxt); return (int)good; } class IamaxUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(IamaxUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 1357, 2)); @@ -185,8 +185,8 @@ TEST_P(IamaxUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(IamaxUsmTestSuite, IamaxUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/iamin.cpp b/tests/unit_tests/blas/level1/iamin.cpp index 147ab4f51..197d7f3d1 100644 --- a/tests/unit_tests/blas/level1/iamin.cpp +++ b/tests/unit_tests/blas/level1/iamin.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { +int test(device* dev, oneapi::math::layout layout, int N, int incx) { // Prepare data. vector x; int64_t result = -1, result_ref = -1; @@ -82,23 +82,23 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::iamin(main_queue, N, x_buffer, incx, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::iamin(main_queue, N, x_buffer, incx, result_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::iamin(main_queue, N, x_buffer, incx, result_buffer); + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::iamin(main_queue, N, x_buffer, incx, result_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::iamin, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::iamin, N, x_buffer, incx, result_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::iamin, N, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::iamin, N, x_buffer, incx, result_buffer); break; default: break; @@ -110,7 +110,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -126,7 +126,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { return (int)good; } -class IaminTests : public ::testing::TestWithParam> { +class IaminTests : public ::testing::TestWithParam> { }; TEST_P(IaminTests, RealSinglePrecision) { @@ -162,8 +162,8 @@ TEST_P(IaminTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(IaminTestSuite, IaminTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/iamin_usm.cpp b/tests/unit_tests/blas/level1/iamin_usm.cpp index 437d60aed..9baea05ef 100644 --- a/tests/unit_tests/blas/level1/iamin_usm.cpp +++ b/tests/unit_tests/blas/level1/iamin_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { +int test(device* dev, oneapi::math::layout layout, int N, int incx) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -82,10 +82,10 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { int64_t* result_p; if constexpr (alloc_type == usm::alloc::shared) { - result_p = (int64_t*)oneapi::mkl::malloc_shared(64, sizeof(int64_t), *dev, cxt); + result_p = (int64_t*)oneapi::math::malloc_shared(64, sizeof(int64_t), *dev, cxt); } else if constexpr (alloc_type == usm::alloc::device) { - result_p = (int64_t*)oneapi::mkl::malloc_device(64, sizeof(int64_t), *dev, cxt); + result_p = (int64_t*)oneapi::math::malloc_device(64, sizeof(int64_t), *dev, cxt); } else { throw std::runtime_error("Bad alloc_type"); @@ -94,12 +94,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::iamin(main_queue, N, x.data(), incx, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::iamin(main_queue, N, x.data(), incx, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::iamin(main_queue, N, x.data(), incx, result_p, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::iamin(main_queue, N, x.data(), incx, result_p, dependencies); break; default: break; @@ -107,12 +107,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::iamin, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::iamin, N, x.data(), incx, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::iamin, N, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::iamin, N, x.data(), incx, result_p, dependencies); break; default: break; @@ -125,7 +125,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -136,13 +136,13 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { // Compare the results of reference implementation and DPC++ implementation. bool good = check_equal_ptr(main_queue, result_p, result_ref, 0, std::cout); - oneapi::mkl::free_usm(result_p, cxt); + oneapi::math::free_usm(result_p, cxt); return (int)good; } class IaminUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(IaminUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 1357, 2)); @@ -185,8 +185,8 @@ TEST_P(IaminUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(IaminUsmTestSuite, IaminUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/nrm2.cpp b/tests/unit_tests/blas/level1/nrm2.cpp index f9b905421..edc0edd93 100644 --- a/tests/unit_tests/blas/level1/nrm2.cpp +++ b/tests/unit_tests/blas/level1/nrm2.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { +int test(device* dev, oneapi::math::layout layout, int N, int incx) { // Prepare data. vector x; fp_res result = fp_res(-1), result_ref = fp_res(-1); @@ -83,22 +83,22 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::nrm2(main_queue, N, x_buffer, incx, result_buffer); + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::nrm2(main_queue, N, x_buffer, incx, result_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::nrm2(main_queue, N, x_buffer, incx, result_buffer); + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::nrm2(main_queue, N, x_buffer, incx, result_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::nrm2, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::nrm2, N, x_buffer, incx, result_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::nrm2, N, x_buffer, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::nrm2, N, x_buffer, incx, result_buffer); break; default: break; @@ -110,7 +110,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -126,7 +126,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { return (int)good; } -class Nrm2Tests : public ::testing::TestWithParam> { +class Nrm2Tests : public ::testing::TestWithParam> { }; TEST_P(Nrm2Tests, RealSinglePrecision) { @@ -168,8 +168,8 @@ TEST_P(Nrm2Tests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(Nrm2TestSuite, Nrm2Tests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/nrm2_usm.cpp b/tests/unit_tests/blas/level1/nrm2_usm.cpp index 065b784e1..b22701401 100644 --- a/tests/unit_tests/blas/level1/nrm2_usm.cpp +++ b/tests/unit_tests/blas/level1/nrm2_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { +int test(device* dev, oneapi::math::layout layout, int N, int incx) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -83,10 +83,10 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { fp_res* result_p; if constexpr (alloc_type == usm::alloc::shared) { - result_p = (fp_res*)oneapi::mkl::malloc_shared(64, sizeof(fp_res), *dev, cxt); + result_p = (fp_res*)oneapi::math::malloc_shared(64, sizeof(fp_res), *dev, cxt); } else if constexpr (alloc_type == usm::alloc::device) { - result_p = (fp_res*)oneapi::mkl::malloc_device(64, sizeof(fp_res), *dev, cxt); + result_p = (fp_res*)oneapi::math::malloc_device(64, sizeof(fp_res), *dev, cxt); } else { throw std::runtime_error("Bad alloc_type"); @@ -95,12 +95,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::nrm2(main_queue, N, x.data(), incx, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::nrm2(main_queue, N, x.data(), incx, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::nrm2(main_queue, N, x.data(), incx, result_p, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::nrm2(main_queue, N, x.data(), incx, result_p, dependencies); break; default: break; @@ -108,12 +108,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::nrm2, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::nrm2, N, x.data(), incx, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::nrm2, N, x.data(), + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::nrm2, N, x.data(), incx, result_p, dependencies); break; default: break; @@ -126,7 +126,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -137,13 +137,13 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx) { // Compare the results of reference implementation and DPC++ implementation. bool good = check_equal_ptr(main_queue, result_p, result_ref, N, std::cout); - oneapi::mkl::free_usm(result_p, cxt); + oneapi::math::free_usm(result_p, cxt); return (int)good; } class Nrm2UsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Nrm2UsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP( @@ -192,8 +192,8 @@ TEST_P(Nrm2UsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(Nrm2UsmTestSuite, Nrm2UsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/rot.cpp b/tests/unit_tests/blas/level1/rot.cpp index e8cdbe993..30506b4bf 100644 --- a/tests/unit_tests/blas/level1/rot.cpp +++ b/tests/unit_tests/blas/level1/rot.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp_scalar c, +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp_scalar c, fp_scalar s) { // Prepare data. vector x, x_ref, y, y_ref; @@ -86,24 +86,24 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp_ try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::rot(main_queue, N, x_buffer, incx, y_buffer, incy, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::rot(main_queue, N, x_buffer, incx, y_buffer, incy, c, s); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::rot(main_queue, N, x_buffer, incx, y_buffer, incy, c, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::rot(main_queue, N, x_buffer, incx, y_buffer, incy, c, s); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::rot, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rot, N, x_buffer, incx, y_buffer, incy, c, s); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::rot, N, x_buffer, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rot, N, x_buffer, incx, y_buffer, incy, c, s); break; default: break; @@ -115,7 +115,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp_ print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -135,7 +135,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp_ return (int)good; } -class RotTests : public ::testing::TestWithParam> { +class RotTests : public ::testing::TestWithParam> { }; TEST_P(RotTests, RealSinglePrecision) { @@ -185,8 +185,8 @@ TEST_P(RotTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(RotTestSuite, RotTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/rot_usm.cpp b/tests/unit_tests/blas/level1/rot_usm.cpp index e54bb483b..880228cc7 100644 --- a/tests/unit_tests/blas/level1/rot_usm.cpp +++ b/tests/unit_tests/blas/level1/rot_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp_scalar c, +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp_scalar c, fp_scalar s) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -88,12 +88,12 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp_ try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::rot(main_queue, N, x.data(), incx, y.data(), + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::rot(main_queue, N, x.data(), incx, y.data(), incy, c, s, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::rot(main_queue, N, x.data(), incx, y.data(), + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::rot(main_queue, N, x.data(), incx, y.data(), incy, c, s, dependencies); break; default: break; @@ -101,12 +101,12 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp_ done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::rot, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rot, N, x.data(), incx, y.data(), incy, c, s, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::rot, N, x.data(), + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rot, N, x.data(), incx, y.data(), incy, c, s, dependencies); break; default: break; @@ -119,7 +119,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp_ print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -137,7 +137,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp_ } class RotUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(RotUsmTests, RealSinglePrecision) { float c(2.0); @@ -186,8 +186,8 @@ TEST_P(RotUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(RotUsmTestSuite, RotUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/rotg.cpp b/tests/unit_tests/blas/level1/rotg.cpp index 289287d2d..23b2b9969 100644 --- a/tests/unit_tests/blas/level1/rotg.cpp +++ b/tests/unit_tests/blas/level1/rotg.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Prepare data. fp a, b, s, a_ref, b_ref, s_ref; fp_scalar c, c_ref; @@ -92,24 +92,24 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::rotg(main_queue, a_buffer, b_buffer, c_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::rotg(main_queue, a_buffer, b_buffer, c_buffer, s_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::rotg(main_queue, a_buffer, b_buffer, c_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::rotg(main_queue, a_buffer, b_buffer, c_buffer, s_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::rotg, a_buffer, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rotg, a_buffer, b_buffer, c_buffer, s_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::rotg, a_buffer, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotg, a_buffer, b_buffer, c_buffer, s_buffer); break; default: break; @@ -121,7 +121,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -144,7 +144,7 @@ int test(device *dev, oneapi::mkl::layout layout) { return (int)good; } -class RotgTests : public ::testing::TestWithParam> { +class RotgTests : public ::testing::TestWithParam> { }; TEST_P(RotgTests, RealSinglePrecision) { @@ -180,8 +180,8 @@ TEST_P(RotgTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(RotgTestSuite, RotgTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/rotg_usm.cpp b/tests/unit_tests/blas/level1/rotg_usm.cpp index e4c38ffe5..a90023a8b 100644 --- a/tests/unit_tests/blas/level1/rotg_usm.cpp +++ b/tests/unit_tests/blas/level1/rotg_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -89,16 +89,16 @@ int test(device *dev, oneapi::mkl::layout layout) { fp *a_p, *b_p, *s_p; fp_scalar *c_p; if constexpr (alloc_type == usm::alloc::shared) { - a_p = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp), *dev, cxt); - b_p = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp), *dev, cxt); - s_p = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp), *dev, cxt); - c_p = (fp_scalar *)oneapi::mkl::malloc_shared(64, sizeof(fp_scalar), *dev, cxt); + a_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + b_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + s_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + c_p = (fp_scalar *)oneapi::math::malloc_shared(64, sizeof(fp_scalar), *dev, cxt); } else if constexpr (alloc_type == usm::alloc::device) { - a_p = (fp *)oneapi::mkl::malloc_device(64, sizeof(fp), *dev, cxt); - b_p = (fp *)oneapi::mkl::malloc_device(64, sizeof(fp), *dev, cxt); - s_p = (fp *)oneapi::mkl::malloc_device(64, sizeof(fp), *dev, cxt); - c_p = (fp_scalar *)oneapi::mkl::malloc_device(64, sizeof(fp_scalar), *dev, cxt); + a_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); + b_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); + s_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); + c_p = (fp_scalar *)oneapi::math::malloc_device(64, sizeof(fp_scalar), *dev, cxt); } else { throw std::runtime_error("Bad alloc_type"); @@ -113,12 +113,12 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::rotg(main_queue, a_p, b_p, c_p, s_p, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::rotg(main_queue, a_p, b_p, c_p, s_p, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::rotg(main_queue, a_p, b_p, c_p, s_p, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::rotg(main_queue, a_p, b_p, c_p, s_p, dependencies); break; default: break; @@ -126,12 +126,12 @@ int test(device *dev, oneapi::mkl::layout layout) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::rotg, a_p, b_p, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rotg, a_p, b_p, c_p, s_p, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::rotg, a_p, b_p, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotg, a_p, b_p, c_p, s_p, dependencies); break; default: break; @@ -144,7 +144,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -161,16 +161,16 @@ int test(device *dev, oneapi::mkl::layout layout) { bool good = good_a && good_b && good_c && good_s; - oneapi::mkl::free_usm(a_p, cxt); - oneapi::mkl::free_usm(b_p, cxt); - oneapi::mkl::free_usm(s_p, cxt); - oneapi::mkl::free_usm(c_p, cxt); + oneapi::math::free_usm(a_p, cxt); + oneapi::math::free_usm(b_p, cxt); + oneapi::math::free_usm(s_p, cxt); + oneapi::math::free_usm(c_p, cxt); return (int)good; } class RotgUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(RotgUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP((test(std::get<0>(GetParam()), std::get<1>(GetParam())))); @@ -201,8 +201,8 @@ TEST_P(RotgUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(RotgUsmTestSuite, RotgUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/rotm.cpp b/tests/unit_tests/blas/level1/rotm.cpp index 31cbd54d2..81f9f137e 100644 --- a/tests/unit_tests/blas/level1/rotm.cpp +++ b/tests/unit_tests/blas/level1/rotm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp flag) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp flag) { // Prepare data. vector x, x_ref, y, y_ref; vector param; @@ -89,24 +89,24 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::rotm(main_queue, N, x_buffer, incx, y_buffer, incy, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::rotm(main_queue, N, x_buffer, incx, y_buffer, incy, param_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::rotm(main_queue, N, x_buffer, incx, y_buffer, incy, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::rotm(main_queue, N, x_buffer, incx, y_buffer, incy, param_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::rotm, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rotm, N, x_buffer, incx, y_buffer, incy, param_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::rotm, N, x_buffer, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotm, N, x_buffer, incx, y_buffer, incy, param_buffer); break; default: break; @@ -118,7 +118,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -136,7 +136,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp return (int)good; } -class RotmTests : public ::testing::TestWithParam> { +class RotmTests : public ::testing::TestWithParam> { }; TEST_P(RotmTests, RealSinglePrecision) { @@ -204,8 +204,8 @@ TEST_P(RotmTests, RealDoublePrecision) { INSTANTIATE_TEST_SUITE_P(RotmTestSuite, RotmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/rotm_usm.cpp b/tests/unit_tests/blas/level1/rotm_usm.cpp index 02472bce1..1625215a9 100644 --- a/tests/unit_tests/blas/level1/rotm_usm.cpp +++ b/tests/unit_tests/blas/level1/rotm_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp flag) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp flag) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -89,12 +89,12 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::rotm( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::rotm( main_queue, N, x.data(), incx, y.data(), incy, param.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::rotm(main_queue, N, x.data(), incx, y.data(), + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::rotm(main_queue, N, x.data(), incx, y.data(), incy, param.data(), dependencies); break; default: break; @@ -102,12 +102,12 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::rotm, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rotm, N, x.data(), incx, y.data(), incy, param.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::rotm, N, x.data(), + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotm, N, x.data(), incx, y.data(), incy, param.data(), dependencies); break; default: break; @@ -120,7 +120,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -138,7 +138,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, fp } class RotmUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(RotmUsmTests, RealSinglePrecision) { float flag(-1.0); @@ -205,8 +205,8 @@ TEST_P(RotmUsmTests, RealDoublePrecision) { INSTANTIATE_TEST_SUITE_P(RotmUsmTestSuite, RotmUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/rotmg.cpp b/tests/unit_tests/blas/level1/rotmg.cpp index bb30a7a32..efc75a4b5 100644 --- a/tests/unit_tests/blas/level1/rotmg.cpp +++ b/tests/unit_tests/blas/level1/rotmg.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Prepare data. fp d1, d2, x1, y1, d1_ref, d2_ref, x1_ref; vector param(5, fp(0)), param_ref(5, fp(0)); @@ -89,24 +89,24 @@ int test(device* dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::rotmg(main_queue, d1_buffer, d2_buffer, x1_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::rotmg(main_queue, d1_buffer, d2_buffer, x1_buffer, y1, param_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::rotmg(main_queue, d1_buffer, d2_buffer, x1_buffer, y1, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::rotmg(main_queue, d1_buffer, d2_buffer, x1_buffer, y1, param_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::rotmg, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rotmg, d1_buffer, d2_buffer, x1_buffer, y1, param_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::rotmg, d1_buffer, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotmg, d1_buffer, d2_buffer, x1_buffer, y1, param_buffer); break; default: break; @@ -118,7 +118,7 @@ int test(device* dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -187,7 +187,7 @@ int test(device* dev, oneapi::mkl::layout layout) { return (int)good; } -class RotmgTests : public ::testing::TestWithParam> { +class RotmgTests : public ::testing::TestWithParam> { }; TEST_P(RotmgTests, RealSinglePrecision) { @@ -201,8 +201,8 @@ TEST_P(RotmgTests, RealDoublePrecision) { INSTANTIATE_TEST_SUITE_P(RotmgTestSuite, RotmgTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/rotmg_usm.cpp b/tests/unit_tests/blas/level1/rotmg_usm.cpp index 3da6cdbf8..e992ad560 100644 --- a/tests/unit_tests/blas/level1/rotmg_usm.cpp +++ b/tests/unit_tests/blas/level1/rotmg_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout) { +int test(device *dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -82,14 +82,14 @@ int test(device *dev, oneapi::mkl::layout layout) { fp *d1_p, *d2_p, *x1_p; if constexpr (alloc_type == usm::alloc::device) { - d1_p = (fp *)oneapi::mkl::malloc_device(64, sizeof(fp), *dev, cxt); - d2_p = (fp *)oneapi::mkl::malloc_device(64, sizeof(fp), *dev, cxt); - x1_p = (fp *)oneapi::mkl::malloc_device(64, sizeof(fp), *dev, cxt); + d1_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); + d2_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); + x1_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); } else if constexpr (alloc_type == usm::alloc::shared) { - d1_p = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp), *dev, cxt); - d2_p = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp), *dev, cxt); - x1_p = (fp *)oneapi::mkl::malloc_shared(64, sizeof(fp), *dev, cxt); + d1_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + d2_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + x1_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); } else { throw std::runtime_error("Bad alloc_type"); @@ -108,12 +108,12 @@ int test(device *dev, oneapi::mkl::layout layout) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::rotmg(main_queue, d1_p, d2_p, x1_p, y1, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::rotmg(main_queue, d1_p, d2_p, x1_p, y1, param.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::rotmg(main_queue, d1_p, d2_p, x1_p, y1, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::rotmg(main_queue, d1_p, d2_p, x1_p, y1, param.data(), dependencies); break; default: break; @@ -121,12 +121,12 @@ int test(device *dev, oneapi::mkl::layout layout) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::rotmg, d1_p, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rotmg, d1_p, d2_p, x1_p, y1, param.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::rotmg, d1_p, d2_p, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotmg, d1_p, d2_p, x1_p, y1, param.data(), dependencies); break; default: break; @@ -139,7 +139,7 @@ int test(device *dev, oneapi::mkl::layout layout) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -205,15 +205,15 @@ int test(device *dev, oneapi::mkl::layout layout) { bool good = good_d1 && good_d2 && good_x1 && flag_good && h11_good && h12_good && h21_good && h22_good; - oneapi::mkl::free_usm(d1_p, cxt); - oneapi::mkl::free_usm(d2_p, cxt); - oneapi::mkl::free_usm(x1_p, cxt); + oneapi::math::free_usm(d1_p, cxt); + oneapi::math::free_usm(d2_p, cxt); + oneapi::math::free_usm(x1_p, cxt); return (int)good; } class RotmgUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(RotmgUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); @@ -230,8 +230,8 @@ TEST_P(RotmgUsmTests, RealDoublePrecision) { INSTANTIATE_TEST_SUITE_P(RotmgUsmTestSuite, RotmgUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/scal.cpp b/tests/unit_tests/blas/level1/scal.cpp index 5c4f8ee35..1d7fad6d8 100644 --- a/tests/unit_tests/blas/level1/scal.cpp +++ b/tests/unit_tests/blas/level1/scal.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx, fp_scalar alpha) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, fp_scalar alpha) { // Prepare data. vector x, x_ref; @@ -84,22 +84,22 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, fp_scalar alp try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::scal(main_queue, N, alpha, x_buffer, incx); + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::scal(main_queue, N, alpha, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::scal(main_queue, N, alpha, x_buffer, incx); + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::scal(main_queue, N, alpha, x_buffer, incx); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::scal, N, alpha, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::scal, N, alpha, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::scal, N, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::scal, N, alpha, x_buffer, incx); break; default: break; @@ -111,7 +111,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, fp_scalar alp print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -126,7 +126,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, fp_scalar alp return (int)good; } -class ScalTests : public ::testing::TestWithParam> { +class ScalTests : public ::testing::TestWithParam> { }; TEST_P(ScalTests, RealSinglePrecision) { @@ -180,8 +180,8 @@ TEST_P(ScalTests, ComplexRealDoublePrecision) { INSTANTIATE_TEST_SUITE_P(ScalTestSuite, ScalTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/scal_usm.cpp b/tests/unit_tests/blas/level1/scal_usm.cpp index 83a936717..a62372331 100644 --- a/tests/unit_tests/blas/level1/scal_usm.cpp +++ b/tests/unit_tests/blas/level1/scal_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx, fp_scalar alpha) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, fp_scalar alpha) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -87,12 +87,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, fp_scalar alp try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::scal(main_queue, N, alpha, x.data(), incx, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::scal(main_queue, N, alpha, x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::scal(main_queue, N, alpha, x.data(), incx, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::scal(main_queue, N, alpha, x.data(), incx, dependencies); break; default: break; @@ -100,12 +100,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, fp_scalar alp done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::scal, N, alpha, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::scal, N, alpha, x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::scal, N, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::scal, N, alpha, x.data(), incx, dependencies); break; default: break; @@ -118,7 +118,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, fp_scalar alp print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -134,7 +134,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, fp_scalar alp } class ScalUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(ScalUsmTests, RealSinglePrecision) { float alpha(2.0); @@ -187,8 +187,8 @@ TEST_P(ScalUsmTests, ComplexRealDoublePrecision) { INSTANTIATE_TEST_SUITE_P(ScalUsmTestSuite, ScalUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/sdsdot.cpp b/tests/unit_tests/blas/level1/sdsdot.cpp index 90bc2d767..67505b5f3 100644 --- a/tests/unit_tests/blas/level1/sdsdot.cpp +++ b/tests/unit_tests/blas/level1/sdsdot.cpp @@ -45,7 +45,7 @@ extern std::vector devices; namespace { -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, float alpha) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, float alpha) { // Prepare data. vector x, y; float result = float(-1), result_ref = float(-1); @@ -84,24 +84,24 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, flo try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::sdsdot(main_queue, N, alpha, x_buffer, incx, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::sdsdot(main_queue, N, alpha, x_buffer, incx, y_buffer, incy, result_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::sdsdot(main_queue, N, alpha, x_buffer, incx, y_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::sdsdot(main_queue, N, alpha, x_buffer, incx, y_buffer, incy, result_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::sdsdot, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::sdsdot, N, alpha, x_buffer, incx, y_buffer, incy, result_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::sdsdot, N, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::sdsdot, N, alpha, x_buffer, incx, y_buffer, incy, result_buffer); break; default: break; @@ -113,7 +113,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, flo print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -130,7 +130,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, flo } class SdsdotTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SdsdotTests, RealSinglePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -141,8 +141,8 @@ TEST_P(SdsdotTests, RealSinglePrecision) { INSTANTIATE_TEST_SUITE_P(SdsdotTestSuite, SdsdotTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/sdsdot_usm.cpp b/tests/unit_tests/blas/level1/sdsdot_usm.cpp index fc4f9189d..09d9e08a3 100644 --- a/tests/unit_tests/blas/level1/sdsdot_usm.cpp +++ b/tests/unit_tests/blas/level1/sdsdot_usm.cpp @@ -45,7 +45,7 @@ extern std::vector devices; namespace { -int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, float alpha) { +int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, float alpha) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const &e : exceptions) { @@ -81,17 +81,17 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, flo // Call DPC++ SDSDOT. - auto result_p = (float *)oneapi::mkl::malloc_shared(64, sizeof(float), *dev, cxt); + auto result_p = (float *)oneapi::math::malloc_shared(64, sizeof(float), *dev, cxt); try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::sdsdot( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::sdsdot( main_queue, N, alpha, x.data(), incx, y.data(), incy, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::sdsdot(main_queue, N, alpha, x.data(), incx, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::sdsdot(main_queue, N, alpha, x.data(), incx, y.data(), incy, result_p, dependencies); break; default: break; @@ -99,13 +99,13 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, flo done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::sdsdot, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::sdsdot, N, alpha, x.data(), incx, y.data(), incy, result_p, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::sdsdot, N, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::sdsdot, N, alpha, x.data(), incx, y.data(), incy, result_p, dependencies); break; default: break; @@ -118,7 +118,7 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, flo print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -129,13 +129,13 @@ int test(device *dev, oneapi::mkl::layout layout, int N, int incx, int incy, flo // Compare the results of reference implementation and DPC++ implementation. bool good = check_equal(*result_p, result_ref, N, std::cout); - oneapi::mkl::free_shared(result_p, cxt); + oneapi::math::free_shared(result_p, cxt); return (int)good; } class SdsdotUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SdsdotUsmTests, RealSinglePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -146,8 +146,8 @@ TEST_P(SdsdotUsmTests, RealSinglePrecision) { INSTANTIATE_TEST_SUITE_P(SdsdotUsmTestSuite, SdsdotUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/swap.cpp b/tests/unit_tests/blas/level1/swap.cpp index a83400233..696943c35 100644 --- a/tests/unit_tests/blas/level1/swap.cpp +++ b/tests/unit_tests/blas/level1/swap.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { // Prepare data. vector x, x_ref, y, y_ref; rand_vector(x, N, incx); @@ -84,23 +84,23 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::swap(main_queue, N, x_buffer, incx, y_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::swap(main_queue, N, x_buffer, incx, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::swap(main_queue, N, x_buffer, incx, y_buffer, incy); + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::swap(main_queue, N, x_buffer, incx, y_buffer, incy); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::swap, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::swap, N, x_buffer, incx, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::swap, N, x_buffer, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::swap, N, x_buffer, incx, y_buffer, incy); break; default: break; @@ -112,7 +112,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -131,7 +131,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { return (int)good; } -class SwapTests : public ::testing::TestWithParam> { +class SwapTests : public ::testing::TestWithParam> { }; TEST_P(SwapTests, RealSinglePrecision) { @@ -167,8 +167,8 @@ TEST_P(SwapTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(SwapTestSuite, SwapTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level1/swap_usm.cpp b/tests/unit_tests/blas/level1/swap_usm.cpp index 4a5c4f375..3fd742c36 100644 --- a/tests/unit_tests/blas/level1/swap_usm.cpp +++ b/tests/unit_tests/blas/level1/swap_usm.cpp @@ -46,7 +46,7 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -86,12 +86,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::swap(main_queue, N, x.data(), incx, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::swap(main_queue, N, x.data(), incx, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::swap(main_queue, N, x.data(), incx, y.data(), + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::swap(main_queue, N, x.data(), incx, y.data(), incy, dependencies); break; default: break; @@ -99,12 +99,12 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::swap, N, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::swap, N, x.data(), incx, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::swap, N, x.data(), + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::swap, N, x.data(), incx, y.data(), incy, dependencies); break; default: break; @@ -117,7 +117,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -135,7 +135,7 @@ int test(device* dev, oneapi::mkl::layout layout, int N, int incx, int incy) { } class SwapUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SwapUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 1357, 2, 3)); @@ -170,8 +170,8 @@ TEST_P(SwapUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(SwapUsmTestSuite, SwapUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/gbmv.cpp b/tests/unit_tests/blas/level2/gbmv.cpp index 8a532ef39..f1e8fc3ec 100644 --- a/tests/unit_tests/blas/level2/gbmv.cpp +++ b/tests/unit_tests/blas/level2/gbmv.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, int m, int n, +int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, int kl, int ku, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. int x_len = outer_dimension(transa, m, n); @@ -58,7 +58,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, rand_vector(x, x_len, incx); rand_vector(y, y_len, incy); y_ref = y; - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, n, lda); // Call Reference GBMV. const int m_ref = m, n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; @@ -94,13 +94,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::gbmv(main_queue, transa, m, n, kl, ku, alpha, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::gbmv(main_queue, transa, m, n, kl, ku, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::gbmv(main_queue, transa, m, n, kl, ku, alpha, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::gbmv(main_queue, transa, m, n, kl, ku, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; @@ -108,13 +108,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gbmv, transa, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gbmv, transa, m, n, kl, ku, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gbmv, transa, m, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gbmv, transa, m, n, kl, ku, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; @@ -127,7 +127,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -142,29 +142,29 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, return (int)good; } -class GbmvTests : public ::testing::TestWithParam> { +class GbmvTests : public ::testing::TestWithParam> { }; TEST_P(GbmvTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); } TEST_P(GbmvTests, RealDoublePrecision) { @@ -173,53 +173,53 @@ TEST_P(GbmvTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); } TEST_P(GbmvTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); } TEST_P(GbmvTests, ComplexDoublePrecision) { @@ -228,38 +228,38 @@ TEST_P(GbmvTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(GbmvTestSuite, GbmvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/gbmv_usm.cpp b/tests/unit_tests/blas/level2/gbmv_usm.cpp index 36f412340..c017d0fae 100644 --- a/tests/unit_tests/blas/level2/gbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/gbmv_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, int m, int n, +int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, int kl, int ku, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -76,7 +76,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, rand_vector(x, x_len, incx); rand_vector(y, y_len, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, n, lda); auto y_ref = y; @@ -94,13 +94,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::gbmv(main_queue, transa, m, n, kl, ku, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::gbmv(main_queue, transa, m, n, kl, ku, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::gbmv(main_queue, transa, m, n, kl, ku, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::gbmv(main_queue, transa, m, n, kl, ku, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -109,13 +109,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gbmv, transa, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gbmv, transa, m, n, kl, ku, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gbmv, transa, m, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gbmv, transa, m, n, kl, ku, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -129,7 +129,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -145,28 +145,28 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, } class GbmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GbmvUsmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); } TEST_P(GbmvUsmTests, RealDoublePrecision) { @@ -175,53 +175,53 @@ TEST_P(GbmvUsmTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); } TEST_P(GbmvUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); } TEST_P(GbmvUsmTests, ComplexDoublePrecision) { @@ -230,38 +230,38 @@ TEST_P(GbmvUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, 5, 7, + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, 5, 7, alpha, + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, 5, 7, + oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(GbmvUsmTestSuite, GbmvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/gemv.cpp b/tests/unit_tests/blas/level2/gemv.cpp index 33681d68f..d75720278 100644 --- a/tests/unit_tests/blas/level2/gemv.cpp +++ b/tests/unit_tests/blas/level2/gemv.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, int m, int n, +int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. int x_len = outer_dimension(transa, m, n); @@ -58,7 +58,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, rand_vector(x, x_len, incx); rand_vector(y, y_len, incy); y_ref = y; - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, n, lda); // Call Reference GEMV. const int m_ref = m, n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; @@ -93,25 +93,25 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::gemv(main_queue, transa, m, n, alpha, A_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::gemv(main_queue, transa, m, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::gemv(main_queue, transa, m, n, alpha, A_buffer, lda, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::gemv(main_queue, transa, m, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemv, transa, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemv, transa, m, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemv, transa, m, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemv, transa, m, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; @@ -124,7 +124,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -139,25 +139,25 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, return (int)good; } -class GemvTests : public ::testing::TestWithParam> { +class GemvTests : public ::testing::TestWithParam> { }; TEST_P(GemvTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, 2, 3, 42)); + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, -2, -3, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); } TEST_P(GemvTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -165,51 +165,51 @@ TEST_P(GemvTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, 2, 3, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, -2, -3, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, 1, 1, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); } TEST_P(GemvTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 2, 3, alpha, beta, + oneapi::math::transpose::nontrans, 2, 3, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 2, 3, alpha, beta, + oneapi::math::transpose::nontrans, 2, 3, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 2, 3, alpha, beta, + oneapi::math::transpose::nontrans, 2, 3, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 2, 3, alpha, beta, 2, + oneapi::math::transpose::trans, 2, 3, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 2, 3, alpha, beta, + oneapi::math::transpose::trans, 2, 3, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 2, 3, alpha, beta, 1, + oneapi::math::transpose::trans, 2, 3, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 2, 3, alpha, + oneapi::math::transpose::conjtrans, 2, 3, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 2, 3, alpha, + oneapi::math::transpose::conjtrans, 2, 3, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 2, 3, alpha, + oneapi::math::transpose::conjtrans, 2, 3, alpha, beta, 1, 1, 42)); } @@ -219,38 +219,38 @@ TEST_P(GemvTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, + oneapi::math::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, + oneapi::math::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, + oneapi::math::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, alpha, + oneapi::math::transpose::conjtrans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, alpha, + oneapi::math::transpose::conjtrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, alpha, + oneapi::math::transpose::conjtrans, 25, 30, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(GemvTestSuite, GemvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/gemv_usm.cpp b/tests/unit_tests/blas/level2/gemv_usm.cpp index f11815bb8..5729914d2 100644 --- a/tests/unit_tests/blas/level2/gemv_usm.cpp +++ b/tests/unit_tests/blas/level2/gemv_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, int m, int n, +int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -76,7 +76,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, rand_vector(x, x_len, incx); rand_vector(y, y_len, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, n, lda); auto y_ref = y; @@ -93,13 +93,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::gemv(main_queue, transa, m, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::gemv(main_queue, transa, m, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::gemv(main_queue, transa, m, n, alpha, A.data(), + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::gemv(main_queue, transa, m, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -108,13 +108,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemv, transa, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemv, transa, m, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemv, transa, m, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemv, transa, m, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -128,7 +128,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -144,24 +144,24 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, } class GemvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemvUsmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, 2, 3, 42)); + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, -2, -3, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); } TEST_P(GemvUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -169,50 +169,50 @@ TEST_P(GemvUsmTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, 2, 3, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, -2, -3, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, beta, 1, 1, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); } TEST_P(GemvUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, + oneapi::math::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, + oneapi::math::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, + oneapi::math::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, alpha, + oneapi::math::transpose::conjtrans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, alpha, + oneapi::math::transpose::conjtrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, alpha, + oneapi::math::transpose::conjtrans, 25, 30, alpha, beta, 1, 1, 42)); } TEST_P(GemvUsmTests, ComplexDoublePrecision) { @@ -221,38 +221,38 @@ TEST_P(GemvUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::nontrans, 25, 30, alpha, + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, + oneapi::math::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, + oneapi::math::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::trans, 25, 30, alpha, beta, + oneapi::math::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, alpha, + oneapi::math::transpose::conjtrans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, alpha, + oneapi::math::transpose::conjtrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::transpose::conjtrans, 25, 30, alpha, + oneapi::math::transpose::conjtrans, 25, 30, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(GemvUsmTestSuite, GemvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/ger.cpp b/tests/unit_tests/blas/level2/ger.cpp index 62341bf85..46b92ff17 100644 --- a/tests/unit_tests/blas/level2/ger.cpp +++ b/tests/unit_tests/blas/level2/ger.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Prepare data. @@ -55,7 +55,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in rand_vector(x, m, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, n, lda); A_ref = A; // Call Reference GER. @@ -90,24 +90,24 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::ger(main_queue, m, n, alpha, x_buffer, incx, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::ger(main_queue, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::ger(main_queue, m, n, alpha, x_buffer, incx, y_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::ger(main_queue, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::ger, m, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::ger, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::ger, m, n, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::ger, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; @@ -119,7 +119,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -135,7 +135,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in return (int)good; } -class GerTests : public ::testing::TestWithParam> { +class GerTests : public ::testing::TestWithParam> { }; TEST_P(GerTests, RealSinglePrecision) { @@ -161,8 +161,8 @@ TEST_P(GerTests, RealDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GerTestSuite, GerTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/ger_usm.cpp b/tests/unit_tests/blas/level2/ger_usm.cpp index 26f3b7c53..9d963e68a 100644 --- a/tests/unit_tests/blas/level2/ger_usm.cpp +++ b/tests/unit_tests/blas/level2/ger_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -74,7 +74,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in rand_vector(x, m, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, n, lda); auto A_ref = A; @@ -90,14 +90,14 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::ger(main_queue, m, n, alpha, x.data(), incx, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::ger(main_queue, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: done = - oneapi::mkl::blas::row_major::ger(main_queue, m, n, alpha, x.data(), incx, + oneapi::math::blas::row_major::ger(main_queue, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; default: break; @@ -105,13 +105,13 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::ger, m, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::ger, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::ger, m, n, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::ger, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; @@ -125,7 +125,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in } class GerUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GerUsmTests, RealSinglePrecision) { float alpha(2.0); @@ -166,8 +166,8 @@ TEST_P(GerUsmTests, RealDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GerUsmTestSuite, GerUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/gerc.cpp b/tests/unit_tests/blas/level2/gerc.cpp index 45b42adcf..316c496a4 100644 --- a/tests/unit_tests/blas/level2/gerc.cpp +++ b/tests/unit_tests/blas/level2/gerc.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Prepare data. @@ -55,7 +55,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in rand_vector(x, m, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, n, lda); A_ref = A; // Call Reference GERC. @@ -90,24 +90,24 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::gerc(main_queue, m, n, alpha, x_buffer, incx, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::gerc(main_queue, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::gerc(main_queue, m, n, alpha, x_buffer, incx, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::gerc(main_queue, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gerc, m, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gerc, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gerc, m, n, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gerc, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; @@ -119,7 +119,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -135,7 +135,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in return (int)good; } -class GercTests : public ::testing::TestWithParam> { +class GercTests : public ::testing::TestWithParam> { }; TEST_P(GercTests, ComplexSinglePrecision) { @@ -161,8 +161,8 @@ TEST_P(GercTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GercTestSuite, GercTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/gerc_usm.cpp b/tests/unit_tests/blas/level2/gerc_usm.cpp index 5691e74cb..cbca89a4e 100644 --- a/tests/unit_tests/blas/level2/gerc_usm.cpp +++ b/tests/unit_tests/blas/level2/gerc_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -74,7 +74,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in rand_vector(x, m, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, n, lda); auto A_ref = A; @@ -90,14 +90,14 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::gerc(main_queue, m, n, alpha, x.data(), + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::gerc(main_queue, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: done = - oneapi::mkl::blas::row_major::gerc(main_queue, m, n, alpha, x.data(), incx, + oneapi::math::blas::row_major::gerc(main_queue, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; default: break; @@ -105,13 +105,13 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gerc, m, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gerc, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gerc, m, n, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gerc, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; @@ -125,7 +125,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in } class GercUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GercUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); @@ -166,8 +166,8 @@ TEST_P(GercUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GercUsmTestSuite, GercUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/geru.cpp b/tests/unit_tests/blas/level2/geru.cpp index 333826b5a..70435af74 100644 --- a/tests/unit_tests/blas/level2/geru.cpp +++ b/tests/unit_tests/blas/level2/geru.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Prepare data. @@ -55,7 +55,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in rand_vector(x, m, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, n, lda); A_ref = A; // Call Reference GERU. @@ -90,24 +90,24 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::geru(main_queue, m, n, alpha, x_buffer, incx, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::geru(main_queue, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::geru(main_queue, m, n, alpha, x_buffer, incx, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::geru(main_queue, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::geru, m, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::geru, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::geru, m, n, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::geru, m, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; @@ -119,7 +119,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -135,7 +135,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in return (int)good; } -class GeruTests : public ::testing::TestWithParam> { +class GeruTests : public ::testing::TestWithParam> { }; TEST_P(GeruTests, ComplexSinglePrecision) { @@ -161,8 +161,8 @@ TEST_P(GeruTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GeruTestSuite, GeruTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/geru_usm.cpp b/tests/unit_tests/blas/level2/geru_usm.cpp index 4748cc995..b470ffbdc 100644 --- a/tests/unit_tests/blas/level2/geru_usm.cpp +++ b/tests/unit_tests/blas/level2/geru_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -74,7 +74,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in rand_vector(x, m, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, n, lda); auto A_ref = A; @@ -90,14 +90,14 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::geru(main_queue, m, n, alpha, x.data(), + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::geru(main_queue, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: + case oneapi::math::layout::row_major: done = - oneapi::mkl::blas::row_major::geru(main_queue, m, n, alpha, x.data(), incx, + oneapi::math::blas::row_major::geru(main_queue, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; default: break; @@ -105,13 +105,13 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::geru, m, n, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::geru, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::geru, m, n, alpha, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::geru, m, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; @@ -125,7 +125,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::mkl::layout layout, int m, int n, fp alpha, int in } class GeruUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GeruUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); @@ -166,8 +166,8 @@ TEST_P(GeruUsmTests, ComplexDoublePrecision) { INSTANTIATE_TEST_SUITE_P(GeruUsmTestSuite, GeruUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/hbmv.cpp b/tests/unit_tests/blas/level2/hbmv.cpp index efc9b8c17..02c1e0f15 100644 --- a/tests/unit_tests/blas/level2/hbmv.cpp +++ b/tests/unit_tests/blas/level2/hbmv.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, int k, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. vector x, y, y_ref, A; @@ -55,7 +55,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, rand_vector(x, n, incx); rand_vector(y, n, incy); y_ref = y; - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); // Call Reference HBMV. const int n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; @@ -91,26 +91,26 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::hbmv(main_queue, upper_lower, n, k, alpha, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::hbmv(main_queue, upper_lower, n, k, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::hbmv(main_queue, upper_lower, n, k, alpha, A_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::hbmv(main_queue, upper_lower, n, k, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hbmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hbmv, upper_lower, n, k, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hbmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hbmv, upper_lower, n, k, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; @@ -123,7 +123,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -138,29 +138,29 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class HbmvTests : public ::testing::TestWithParam> { +class HbmvTests : public ::testing::TestWithParam> { }; TEST_P(HbmvTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 2, 3, + oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 2, 3, + oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, -2, + oneapi::math::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, -2, + oneapi::math::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 1, 1, + oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 1, 1, + oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); } TEST_P(HbmvTests, ComplexDoublePrecision) { @@ -169,29 +169,29 @@ TEST_P(HbmvTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 2, 3, + oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 2, 3, + oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, -2, + oneapi::math::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, -2, + oneapi::math::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 1, 1, + oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 1, 1, + oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(HbmvTestSuite, HbmvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/hbmv_usm.cpp b/tests/unit_tests/blas/level2/hbmv_usm.cpp index ef60731c4..6adbba106 100644 --- a/tests/unit_tests/blas/level2/hbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/hbmv_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, int k, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -74,7 +74,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); auto y_ref = y; @@ -92,13 +92,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::hbmv(main_queue, upper_lower, n, k, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::hbmv(main_queue, upper_lower, n, k, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::hbmv(main_queue, upper_lower, n, k, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::hbmv(main_queue, upper_lower, n, k, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -107,13 +107,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hbmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hbmv, upper_lower, n, k, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hbmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hbmv, upper_lower, n, k, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -127,7 +127,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -143,28 +143,28 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class HbmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HbmvUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 2, 3, + oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 2, 3, + oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, -2, + oneapi::math::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, -2, + oneapi::math::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 1, 1, + oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 1, 1, + oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); } TEST_P(HbmvUsmTests, ComplexDoublePrecision) { @@ -173,29 +173,29 @@ TEST_P(HbmvUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 2, 3, + oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 2, 3, + oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, -2, + oneapi::math::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, -2, + oneapi::math::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 1, 1, + oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 1, 1, + oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(HbmvUsmTestSuite, HbmvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/hemv.cpp b/tests/unit_tests/blas/level2/hemv.cpp index b0dab2e95..c45ebd3a4 100644 --- a/tests/unit_tests/blas/level2/hemv.cpp +++ b/tests/unit_tests/blas/level2/hemv.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. vector x, y, y_ref, A; @@ -55,7 +55,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, rand_vector(x, n, incx); rand_vector(y, n, incy); y_ref = y; - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); // Call Reference HEMV. const int n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; @@ -90,25 +90,25 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::hemv(main_queue, upper_lower, n, alpha, A_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::hemv(main_queue, upper_lower, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::hemv(main_queue, upper_lower, n, alpha, A_buffer, lda, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::hemv(main_queue, upper_lower, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hemv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hemv, upper_lower, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hemv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hemv, upper_lower, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; @@ -121,7 +121,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -136,29 +136,29 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class HemvTests : public ::testing::TestWithParam> { +class HemvTests : public ::testing::TestWithParam> { }; TEST_P(HemvTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3, + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3, + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3, + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3, + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1, + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1, + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1, 42)); } TEST_P(HemvTests, ComplexDoublePrecision) { @@ -167,29 +167,29 @@ TEST_P(HemvTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3, + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3, + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3, + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3, + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1, + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1, + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(HemvTestSuite, HemvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/hemv_usm.cpp b/tests/unit_tests/blas/level2/hemv_usm.cpp index d78935e71..07f0cc386 100644 --- a/tests/unit_tests/blas/level2/hemv_usm.cpp +++ b/tests/unit_tests/blas/level2/hemv_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -74,7 +74,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); auto y_ref = y; @@ -91,13 +91,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::hemv(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::hemv(main_queue, upper_lower, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::hemv(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::hemv(main_queue, upper_lower, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -106,13 +106,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hemv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hemv, upper_lower, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hemv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hemv, upper_lower, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -126,7 +126,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -142,28 +142,28 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class HemvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HemvUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3, + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3, + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3, + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3, + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1, + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1, + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1, 42)); } TEST_P(HemvUsmTests, ComplexDoublePrecision) { @@ -172,29 +172,29 @@ TEST_P(HemvUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3, + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3, + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3, + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3, + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1, + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1, + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(HemvUsmTestSuite, HemvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/her.cpp b/tests/unit_tests/blas/level2/her.cpp index 951f37a6d..d7817ba9d 100644 --- a/tests/unit_tests/blas/level2/her.cpp +++ b/tests/unit_tests/blas/level2/her.cpp @@ -47,12 +47,12 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp_scalar alpha, int incx, int lda) { // Prepare data. vector x, A_ref, A; rand_vector(x, n, incx); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); A_ref = A; // Call Reference HER. @@ -87,24 +87,24 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::her(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::her(main_queue, upper_lower, n, alpha, x_buffer, incx, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::her(main_queue, upper_lower, n, alpha, x_buffer, incx, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::her(main_queue, upper_lower, n, alpha, x_buffer, incx, A_buffer, lda); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::her, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::her, upper_lower, n, alpha, x_buffer, incx, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::her, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::her, upper_lower, n, alpha, x_buffer, incx, A_buffer, lda); break; default: break; @@ -116,7 +116,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -131,29 +131,29 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class HerTests : public ::testing::TestWithParam> { +class HerTests : public ::testing::TestWithParam> { }; TEST_P(HerTests, ComplexSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 42))); + oneapi::math::uplo::lower, 30, alpha, 2, 42))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 42))); + oneapi::math::uplo::upper, 30, alpha, 2, 42))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, 42))); + oneapi::math::uplo::lower, 30, alpha, -2, 42))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, 42))); + oneapi::math::uplo::upper, 30, alpha, -2, 42))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 42))); + oneapi::math::uplo::lower, 30, alpha, 1, 42))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 42))); + oneapi::math::uplo::upper, 30, alpha, 1, 42))); } TEST_P(HerTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -161,28 +161,28 @@ TEST_P(HerTests, ComplexDoublePrecision) { double alpha(2.0); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 42))); + oneapi::math::uplo::lower, 30, alpha, 2, 42))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 42))); + oneapi::math::uplo::upper, 30, alpha, 2, 42))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, 42))); + oneapi::math::uplo::lower, 30, alpha, -2, 42))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, 42))); + oneapi::math::uplo::upper, 30, alpha, -2, 42))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 42))); + oneapi::math::uplo::lower, 30, alpha, 1, 42))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 42))); + oneapi::math::uplo::upper, 30, alpha, 1, 42))); } INSTANTIATE_TEST_SUITE_P(HerTestSuite, HerTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/her2.cpp b/tests/unit_tests/blas/level2/her2.cpp index 2a42bef17..9ddb0ee3f 100644 --- a/tests/unit_tests/blas/level2/her2.cpp +++ b/tests/unit_tests/blas/level2/her2.cpp @@ -47,14 +47,14 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy, int lda) { // Prepare data. vector x, y, A_ref, A; rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); A_ref = A; // Call Reference HER2. @@ -90,25 +90,25 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::her2(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::her2(main_queue, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::her2(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::her2(main_queue, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::her2, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::her2, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::her2, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::her2, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; @@ -120,7 +120,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -135,46 +135,46 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class Her2Tests : public ::testing::TestWithParam> { +class Her2Tests : public ::testing::TestWithParam> { }; TEST_P(Her2Tests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 1, 42)); } TEST_P(Her2Tests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(Her2TestSuite, Her2Tests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/her2_usm.cpp b/tests/unit_tests/blas/level2/her2_usm.cpp index 2940e3004..93c1d1f04 100644 --- a/tests/unit_tests/blas/level2/her2_usm.cpp +++ b/tests/unit_tests/blas/level2/her2_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -74,7 +74,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); auto A_ref = A; @@ -91,13 +91,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::her2(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::her2(main_queue, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::her2(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::her2(main_queue, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; @@ -106,13 +106,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::her2, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::her2, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::her2, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::her2, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; @@ -126,7 +126,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -142,45 +142,45 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class Her2UsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Her2UsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 1, 42)); } TEST_P(Her2UsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(Her2UsmTestSuite, Her2UsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/her_usm.cpp b/tests/unit_tests/blas/level2/her_usm.cpp index 589fed553..cc61fb93b 100644 --- a/tests/unit_tests/blas/level2/her_usm.cpp +++ b/tests/unit_tests/blas/level2/her_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp_scalar alpha, int incx, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -72,7 +72,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, auto ua = usm_allocator(cxt, *dev); vector x(ua), A(ua); rand_vector(x, n, incx); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); auto A_ref = A; @@ -89,12 +89,12 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::her( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::her( main_queue, upper_lower, n, alpha, x.data(), incx, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::her( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::her( main_queue, upper_lower, n, alpha, x.data(), incx, A.data(), lda, dependencies); break; default: break; @@ -102,13 +102,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::her, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::her, upper_lower, n, alpha, x.data(), incx, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::her, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::her, upper_lower, n, alpha, x.data(), incx, A.data(), lda, dependencies); break; default: break; @@ -121,7 +121,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -137,28 +137,28 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class HerUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HerUsmTests, ComplexSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 42))); + oneapi::math::uplo::lower, 30, alpha, 2, 42))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 42))); + oneapi::math::uplo::upper, 30, alpha, 2, 42))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, 42))); + oneapi::math::uplo::lower, 30, alpha, -2, 42))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, 42))); + oneapi::math::uplo::upper, 30, alpha, -2, 42))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 42))); + oneapi::math::uplo::lower, 30, alpha, 1, 42))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 42))); + oneapi::math::uplo::upper, 30, alpha, 1, 42))); } TEST_P(HerUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -166,28 +166,28 @@ TEST_P(HerUsmTests, ComplexDoublePrecision) { double alpha(2.0); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 42))); + oneapi::math::uplo::lower, 30, alpha, 2, 42))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 42))); + oneapi::math::uplo::upper, 30, alpha, 2, 42))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, 42))); + oneapi::math::uplo::lower, 30, alpha, -2, 42))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, 42))); + oneapi::math::uplo::upper, 30, alpha, -2, 42))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 42))); + oneapi::math::uplo::lower, 30, alpha, 1, 42))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 42))); + oneapi::math::uplo::upper, 30, alpha, 1, 42))); } INSTANTIATE_TEST_SUITE_P(HerUsmTestSuite, HerUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/hpmv.cpp b/tests/unit_tests/blas/level2/hpmv.cpp index 7f43d82a0..8894d1e0b 100644 --- a/tests/unit_tests/blas/level2/hpmv.cpp +++ b/tests/unit_tests/blas/level2/hpmv.cpp @@ -47,14 +47,14 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy) { // Prepare data. vector x, y, y_ref, A; rand_vector(x, n, incx); rand_vector(y, n, incy); y_ref = y; - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); // Call Reference HPMV. const int n_ref = n, incx_ref = incx, incy_ref = incy; @@ -89,25 +89,25 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::hpmv(main_queue, upper_lower, n, alpha, A_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::hpmv(main_queue, upper_lower, n, alpha, A_buffer, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::hpmv(main_queue, upper_lower, n, alpha, A_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::hpmv(main_queue, upper_lower, n, alpha, A_buffer, x_buffer, incx, beta, y_buffer, incy); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hpmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hpmv, upper_lower, n, alpha, A_buffer, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hpmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpmv, upper_lower, n, alpha, A_buffer, x_buffer, incx, beta, y_buffer, incy); break; default: break; @@ -119,7 +119,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -134,24 +134,24 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class HpmvTests : public ::testing::TestWithParam> { +class HpmvTests : public ::testing::TestWithParam> { }; TEST_P(HpmvTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1)); } TEST_P(HpmvTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -159,25 +159,25 @@ TEST_P(HpmvTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1)); } INSTANTIATE_TEST_SUITE_P(HpmvTestSuite, HpmvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/hpmv_usm.cpp b/tests/unit_tests/blas/level2/hpmv_usm.cpp index 6be30eadf..d2302f573 100644 --- a/tests/unit_tests/blas/level2/hpmv_usm.cpp +++ b/tests/unit_tests/blas/level2/hpmv_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -73,7 +73,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, vector x(ua), y(ua), A(ua); rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); auto y_ref = y; @@ -90,13 +90,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::hpmv(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::hpmv(main_queue, upper_lower, n, alpha, A.data(), x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::hpmv(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::hpmv(main_queue, upper_lower, n, alpha, A.data(), x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -105,13 +105,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hpmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hpmv, upper_lower, n, alpha, A.data(), x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hpmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpmv, upper_lower, n, alpha, A.data(), x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -125,7 +125,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -141,23 +141,23 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class HpmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HpmvUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1)); } TEST_P(HpmvUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -165,25 +165,25 @@ TEST_P(HpmvUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1)); } INSTANTIATE_TEST_SUITE_P(HpmvUsmTestSuite, HpmvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/hpr.cpp b/tests/unit_tests/blas/level2/hpr.cpp index 714da1780..157efa989 100644 --- a/tests/unit_tests/blas/level2/hpr.cpp +++ b/tests/unit_tests/blas/level2/hpr.cpp @@ -47,12 +47,12 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp_scalar alpha, int incx) { // Prepare data. vector x, A_ref, A; rand_vector(x, n, incx); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); A_ref = A; // Call Reference HPR. @@ -87,24 +87,24 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::hpr(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::hpr(main_queue, upper_lower, n, alpha, x_buffer, incx, A_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::hpr(main_queue, upper_lower, n, alpha, x_buffer, incx, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::hpr(main_queue, upper_lower, n, alpha, x_buffer, incx, A_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hpr, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hpr, upper_lower, n, alpha, x_buffer, incx, A_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hpr, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpr, upper_lower, n, alpha, x_buffer, incx, A_buffer); break; default: break; @@ -116,7 +116,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -131,25 +131,25 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class HprTests : public ::testing::TestWithParam> { +class HprTests : public ::testing::TestWithParam> { }; TEST_P(HprTests, ComplexSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, 30, alpha, 2))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 2))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, 30, alpha, 2))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 2))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2))); + oneapi::math::uplo::lower, 30, alpha, -2))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2))); + oneapi::math::uplo::upper, 30, alpha, -2))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, 30, alpha, 1))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 1))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, 30, alpha, 1))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 1))); } TEST_P(HprTests, ComplexDoublePrecision) { @@ -157,25 +157,25 @@ TEST_P(HprTests, ComplexDoublePrecision) { double alpha(2.0); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, 30, alpha, 2))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 2))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, 30, alpha, 2))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 2))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2))); + oneapi::math::uplo::lower, 30, alpha, -2))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2))); + oneapi::math::uplo::upper, 30, alpha, -2))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, 30, alpha, 1))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 1))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, 30, alpha, 1))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 1))); } INSTANTIATE_TEST_SUITE_P(HprTestSuite, HprTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/hpr2.cpp b/tests/unit_tests/blas/level2/hpr2.cpp index fe03b6d48..5ab9b7d28 100644 --- a/tests/unit_tests/blas/level2/hpr2.cpp +++ b/tests/unit_tests/blas/level2/hpr2.cpp @@ -47,13 +47,13 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy) { // Prepare data. vector x, y, A_ref, A; rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); A_ref = A; // Call Reference HPR2. @@ -89,25 +89,25 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::hpr2(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::hpr2(main_queue, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::hpr2(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::hpr2(main_queue, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hpr2, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hpr2, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hpr2, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpr2, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer); break; default: break; @@ -119,7 +119,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -134,46 +134,46 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class Hpr2Tests : public ::testing::TestWithParam> { +class Hpr2Tests : public ::testing::TestWithParam> { }; TEST_P(Hpr2Tests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, 1, 1)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, 1, 1)); } TEST_P(Hpr2Tests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, 1, 1)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, 1, 1)); } INSTANTIATE_TEST_SUITE_P(Hpr2TestSuite, Hpr2Tests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/hpr2_usm.cpp b/tests/unit_tests/blas/level2/hpr2_usm.cpp index d58d4163d..4ac809338 100644 --- a/tests/unit_tests/blas/level2/hpr2_usm.cpp +++ b/tests/unit_tests/blas/level2/hpr2_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -73,7 +73,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, vector x(ua), y(ua), A(ua); rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); auto A_ref = A; @@ -90,13 +90,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::hpr2(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::hpr2(main_queue, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::hpr2(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::hpr2(main_queue, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), dependencies); break; @@ -105,13 +105,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hpr2, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hpr2, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hpr2, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpr2, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), dependencies); break; @@ -125,7 +125,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -141,45 +141,45 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class Hpr2UsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Hpr2UsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, 1, 1)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, 1, 1)); } TEST_P(Hpr2UsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, 1, 1)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, 1, 1)); } INSTANTIATE_TEST_SUITE_P(Hpr2UsmTestSuite, Hpr2UsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/hpr_usm.cpp b/tests/unit_tests/blas/level2/hpr_usm.cpp index 15dcbd4ac..b7b38a80a 100644 --- a/tests/unit_tests/blas/level2/hpr_usm.cpp +++ b/tests/unit_tests/blas/level2/hpr_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp_scalar alpha, int incx) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -72,7 +72,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, auto ua = usm_allocator(cxt, *dev); vector x(ua), A(ua); rand_vector(x, n, incx); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); auto A_ref = A; @@ -89,12 +89,12 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::hpr(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::hpr(main_queue, upper_lower, n, alpha, x.data(), incx, A.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::hpr(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::hpr(main_queue, upper_lower, n, alpha, x.data(), incx, A.data(), dependencies); break; default: break; @@ -102,13 +102,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hpr, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hpr, upper_lower, n, alpha, x.data(), incx, A.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hpr, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpr, upper_lower, n, alpha, x.data(), incx, A.data(), dependencies); break; default: break; @@ -121,7 +121,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -137,24 +137,24 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class HprUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HprUsmTests, ComplexSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, 30, alpha, 2))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 2))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, 30, alpha, 2))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 2))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2))); + oneapi::math::uplo::lower, 30, alpha, -2))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2))); + oneapi::math::uplo::upper, 30, alpha, -2))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, 30, alpha, 1))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 1))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, 30, alpha, 1))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 1))); } TEST_P(HprUsmTests, ComplexDoublePrecision) { @@ -162,25 +162,25 @@ TEST_P(HprUsmTests, ComplexDoublePrecision) { double alpha(2.0); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, 30, alpha, 2))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 2))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, 30, alpha, 2))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 2))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2))); + oneapi::math::uplo::lower, 30, alpha, -2))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2))); + oneapi::math::uplo::upper, 30, alpha, -2))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, 30, alpha, 1))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 1))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, 30, alpha, 1))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 1))); } INSTANTIATE_TEST_SUITE_P(HprUsmTestSuite, HprUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/sbmv.cpp b/tests/unit_tests/blas/level2/sbmv.cpp index b0291945b..9291e14de 100644 --- a/tests/unit_tests/blas/level2/sbmv.cpp +++ b/tests/unit_tests/blas/level2/sbmv.cpp @@ -47,14 +47,14 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, int k, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. vector x, y, y_ref, A; rand_vector(x, n, incx); rand_vector(y, n, incy); y_ref = y; - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); // Call Reference SBMV. const int n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; @@ -89,26 +89,26 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::sbmv(main_queue, upper_lower, n, k, alpha, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::sbmv(main_queue, upper_lower, n, k, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::sbmv(main_queue, upper_lower, n, k, alpha, A_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::sbmv(main_queue, upper_lower, n, k, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::sbmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::sbmv, upper_lower, n, k, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::sbmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::sbmv, upper_lower, n, k, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; @@ -121,7 +121,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -136,24 +136,24 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class SbmvTests : public ::testing::TestWithParam> { +class SbmvTests : public ::testing::TestWithParam> { }; TEST_P(SbmvTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); } TEST_P(SbmvTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -161,23 +161,23 @@ TEST_P(SbmvTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(SbmvTestSuite, SbmvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/sbmv_usm.cpp b/tests/unit_tests/blas/level2/sbmv_usm.cpp index b14464788..ad0a8b9b0 100644 --- a/tests/unit_tests/blas/level2/sbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/sbmv_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, int k, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -73,7 +73,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, vector x(ua), y(ua), A(ua); rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); auto y_ref = y; @@ -91,13 +91,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::sbmv(main_queue, upper_lower, n, k, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::sbmv(main_queue, upper_lower, n, k, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::sbmv(main_queue, upper_lower, n, k, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::sbmv(main_queue, upper_lower, n, k, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -106,13 +106,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::sbmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::sbmv, upper_lower, n, k, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::sbmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::sbmv, upper_lower, n, k, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -126,7 +126,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -142,23 +142,23 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class SbmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SbmvUsmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); } TEST_P(SbmvUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -166,23 +166,23 @@ TEST_P(SbmvUsmTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(SbmvUsmTestSuite, SbmvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/spmv.cpp b/tests/unit_tests/blas/level2/spmv.cpp index 78d413853..0fb16c5c5 100644 --- a/tests/unit_tests/blas/level2/spmv.cpp +++ b/tests/unit_tests/blas/level2/spmv.cpp @@ -47,14 +47,14 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy) { // Prepare data. vector x, y, y_ref, A; rand_vector(x, n, incx); rand_vector(y, n, incy); y_ref = y; - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); // Call Reference SPMV. const int n_ref = n, incx_ref = incx, incy_ref = incy; @@ -89,25 +89,25 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::spmv(main_queue, upper_lower, n, alpha, A_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::spmv(main_queue, upper_lower, n, alpha, A_buffer, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::spmv(main_queue, upper_lower, n, alpha, A_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::spmv(main_queue, upper_lower, n, alpha, A_buffer, x_buffer, incx, beta, y_buffer, incy); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::spmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::spmv, upper_lower, n, alpha, A_buffer, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::spmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spmv, upper_lower, n, alpha, A_buffer, x_buffer, incx, beta, y_buffer, incy); break; default: break; @@ -119,7 +119,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -134,24 +134,24 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class SpmvTests : public ::testing::TestWithParam> { +class SpmvTests : public ::testing::TestWithParam> { }; TEST_P(SpmvTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1)); } TEST_P(SpmvTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -159,23 +159,23 @@ TEST_P(SpmvTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1)); } INSTANTIATE_TEST_SUITE_P(SpmvTestSuite, SpmvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/spmv_usm.cpp b/tests/unit_tests/blas/level2/spmv_usm.cpp index 12e1eae39..7998fc02b 100644 --- a/tests/unit_tests/blas/level2/spmv_usm.cpp +++ b/tests/unit_tests/blas/level2/spmv_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -73,7 +73,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, vector x(ua), y(ua), A(ua); rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); auto y_ref = y; @@ -90,13 +90,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::spmv(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::spmv(main_queue, upper_lower, n, alpha, A.data(), x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::spmv(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::spmv(main_queue, upper_lower, n, alpha, A.data(), x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -105,13 +105,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::spmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::spmv, upper_lower, n, alpha, A.data(), x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::spmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spmv, upper_lower, n, alpha, A.data(), x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -125,7 +125,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -141,23 +141,23 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class SpmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SpmvUsmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1)); } TEST_P(SpmvUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -165,23 +165,23 @@ TEST_P(SpmvUsmTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1)); } INSTANTIATE_TEST_SUITE_P(SpmvUsmTestSuite, SpmvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/spr.cpp b/tests/unit_tests/blas/level2/spr.cpp index 1d954af7a..7f6706cc3 100644 --- a/tests/unit_tests/blas/level2/spr.cpp +++ b/tests/unit_tests/blas/level2/spr.cpp @@ -47,12 +47,12 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx) { // Prepare data. vector x, A_ref, A; rand_vector(x, n, incx); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); A_ref = A; // Call Reference SPR. @@ -86,24 +86,24 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::spr(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::spr(main_queue, upper_lower, n, alpha, x_buffer, incx, A_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::spr(main_queue, upper_lower, n, alpha, x_buffer, incx, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::spr(main_queue, upper_lower, n, alpha, x_buffer, incx, A_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::spr, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::spr, upper_lower, n, alpha, x_buffer, incx, A_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::spr, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spr, upper_lower, n, alpha, x_buffer, incx, A_buffer); break; default: break; @@ -115,7 +115,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -130,46 +130,46 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class SprTests : public ::testing::TestWithParam> { +class SprTests : public ::testing::TestWithParam> { }; TEST_P(SprTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2)); + oneapi::math::uplo::lower, 30, alpha, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2)); + oneapi::math::uplo::upper, 30, alpha, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2)); + oneapi::math::uplo::lower, 30, alpha, -2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2)); + oneapi::math::uplo::upper, 30, alpha, -2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1)); + oneapi::math::uplo::lower, 30, alpha, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1)); + oneapi::math::uplo::upper, 30, alpha, 1)); } TEST_P(SprTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2)); + oneapi::math::uplo::lower, 30, alpha, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2)); + oneapi::math::uplo::upper, 30, alpha, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2)); + oneapi::math::uplo::lower, 30, alpha, -2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2)); + oneapi::math::uplo::upper, 30, alpha, -2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1)); + oneapi::math::uplo::lower, 30, alpha, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1)); + oneapi::math::uplo::upper, 30, alpha, 1)); } INSTANTIATE_TEST_SUITE_P(SprTestSuite, SprTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/spr2.cpp b/tests/unit_tests/blas/level2/spr2.cpp index e1f1735ce..6cec654c0 100644 --- a/tests/unit_tests/blas/level2/spr2.cpp +++ b/tests/unit_tests/blas/level2/spr2.cpp @@ -47,13 +47,13 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy) { // Prepare data. vector x, y, A_ref, A; rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); A_ref = A; // Call Reference SPR2. @@ -89,25 +89,25 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::spr2(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::spr2(main_queue, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::spr2(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::spr2(main_queue, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::spr2, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::spr2, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::spr2, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spr2, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer); break; default: break; @@ -119,7 +119,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -134,46 +134,46 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class Spr2Tests : public ::testing::TestWithParam> { +class Spr2Tests : public ::testing::TestWithParam> { }; TEST_P(Spr2Tests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, 1, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, 1, 1)); } TEST_P(Spr2Tests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, 1, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, 1, 1)); } INSTANTIATE_TEST_SUITE_P(Spr2TestSuite, Spr2Tests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/spr2_usm.cpp b/tests/unit_tests/blas/level2/spr2_usm.cpp index dd5e0fd84..27e4ce07e 100644 --- a/tests/unit_tests/blas/level2/spr2_usm.cpp +++ b/tests/unit_tests/blas/level2/spr2_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -73,7 +73,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, vector x(ua), y(ua), A(ua); rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); auto A_ref = A; @@ -90,13 +90,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::spr2(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::spr2(main_queue, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::spr2(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::spr2(main_queue, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), dependencies); break; @@ -105,13 +105,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::spr2, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::spr2, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::spr2, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spr2, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), dependencies); break; @@ -125,7 +125,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -141,45 +141,45 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class Spr2UsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Spr2UsmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, 1, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, 1, 1)); } TEST_P(Spr2UsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3)); + oneapi::math::uplo::lower, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3)); + oneapi::math::uplo::upper, 30, alpha, 2, 3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, -2, -3)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1)); + oneapi::math::uplo::lower, 30, alpha, 1, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1)); + oneapi::math::uplo::upper, 30, alpha, 1, 1)); } INSTANTIATE_TEST_SUITE_P(Spr2UsmTestSuite, Spr2UsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/spr_usm.cpp b/tests/unit_tests/blas/level2/spr_usm.cpp index 38f695b21..20e74159b 100644 --- a/tests/unit_tests/blas/level2/spr_usm.cpp +++ b/tests/unit_tests/blas/level2/spr_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -72,7 +72,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, auto ua = usm_allocator(cxt, *dev); vector x(ua), A(ua); rand_vector(x, n, incx); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, n); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, n); auto A_ref = A; @@ -88,12 +88,12 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::spr(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::spr(main_queue, upper_lower, n, alpha, x.data(), incx, A.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::spr(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::spr(main_queue, upper_lower, n, alpha, x.data(), incx, A.data(), dependencies); break; default: break; @@ -101,13 +101,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::spr, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::spr, upper_lower, n, alpha, x.data(), incx, A.data(), dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::spr, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spr, upper_lower, n, alpha, x.data(), incx, A.data(), dependencies); break; default: break; @@ -120,7 +120,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -136,45 +136,45 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class SprUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SprUsmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2)); + oneapi::math::uplo::lower, 30, alpha, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2)); + oneapi::math::uplo::upper, 30, alpha, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2)); + oneapi::math::uplo::lower, 30, alpha, -2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2)); + oneapi::math::uplo::upper, 30, alpha, -2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1)); + oneapi::math::uplo::lower, 30, alpha, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1)); + oneapi::math::uplo::upper, 30, alpha, 1)); } TEST_P(SprUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2)); + oneapi::math::uplo::lower, 30, alpha, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2)); + oneapi::math::uplo::upper, 30, alpha, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2)); + oneapi::math::uplo::lower, 30, alpha, -2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2)); + oneapi::math::uplo::upper, 30, alpha, -2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1)); + oneapi::math::uplo::lower, 30, alpha, 1)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1)); + oneapi::math::uplo::upper, 30, alpha, 1)); } INSTANTIATE_TEST_SUITE_P(SprUsmTestSuite, SprUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/symv.cpp b/tests/unit_tests/blas/level2/symv.cpp index 226fe7981..4e62799e8 100644 --- a/tests/unit_tests/blas/level2/symv.cpp +++ b/tests/unit_tests/blas/level2/symv.cpp @@ -47,14 +47,14 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. vector x, y, y_ref, A; rand_vector(x, n, incx); rand_vector(y, n, incy); y_ref = y; - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); // Call Reference SYMV. const int n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; @@ -89,25 +89,25 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::symv(main_queue, upper_lower, n, alpha, A_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::symv(main_queue, upper_lower, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::symv(main_queue, upper_lower, n, alpha, A_buffer, lda, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::symv(main_queue, upper_lower, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::symv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::symv, upper_lower, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::symv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::symv, upper_lower, n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, incy); break; @@ -120,7 +120,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -135,24 +135,24 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class SymvTests : public ::testing::TestWithParam> { +class SymvTests : public ::testing::TestWithParam> { }; TEST_P(SymvTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1, 42)); } TEST_P(SymvTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -160,23 +160,23 @@ TEST_P(SymvTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(SymvTestSuite, SymvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/symv_usm.cpp b/tests/unit_tests/blas/level2/symv_usm.cpp index aee51965c..30b82190c 100644 --- a/tests/unit_tests/blas/level2/symv_usm.cpp +++ b/tests/unit_tests/blas/level2/symv_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -73,7 +73,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, vector x(ua), y(ua), A(ua); rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); auto y_ref = y; @@ -90,13 +90,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::symv(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::symv(main_queue, upper_lower, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::symv(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::symv(main_queue, upper_lower, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -105,13 +105,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::symv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::symv, upper_lower, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::symv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::symv, upper_lower, n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), incy, dependencies); break; @@ -125,7 +125,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -141,23 +141,23 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class SymvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SymvUsmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1, 42)); } TEST_P(SymvUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -165,23 +165,23 @@ TEST_P(SymvUsmTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, beta, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, beta, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(SymvUsmTestSuite, SymvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/syr.cpp b/tests/unit_tests/blas/level2/syr.cpp index b7cf0aad7..1ce3ecf81 100644 --- a/tests/unit_tests/blas/level2/syr.cpp +++ b/tests/unit_tests/blas/level2/syr.cpp @@ -47,12 +47,12 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int lda) { // Prepare data. vector x, A_ref, A; rand_vector(x, n, incx); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); A_ref = A; // Call Reference SYR. @@ -86,24 +86,24 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::syr(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::syr(main_queue, upper_lower, n, alpha, x_buffer, incx, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::syr(main_queue, upper_lower, n, alpha, x_buffer, incx, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::syr(main_queue, upper_lower, n, alpha, x_buffer, incx, A_buffer, lda); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::syr, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syr, upper_lower, n, alpha, x_buffer, incx, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::syr, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syr, upper_lower, n, alpha, x_buffer, incx, A_buffer, lda); break; default: break; @@ -115,7 +115,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -130,46 +130,46 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class SyrTests : public ::testing::TestWithParam> { +class SyrTests : public ::testing::TestWithParam> { }; TEST_P(SyrTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 42)); } TEST_P(SyrTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 42)); } INSTANTIATE_TEST_SUITE_P(SyrTestSuite, SyrTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/syr2.cpp b/tests/unit_tests/blas/level2/syr2.cpp index b58af622d..3c3b3ceac 100644 --- a/tests/unit_tests/blas/level2/syr2.cpp +++ b/tests/unit_tests/blas/level2/syr2.cpp @@ -47,13 +47,13 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy, int lda) { // Prepare data. vector x, y, A_ref, A; rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); A_ref = A; // Call Reference SYR2. @@ -89,25 +89,25 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::syr2(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::syr2(main_queue, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::syr2(main_queue, upper_lower, n, alpha, x_buffer, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::syr2(main_queue, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::syr2, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syr2, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::syr2, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syr2, upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; @@ -119,7 +119,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -134,46 +134,46 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class Syr2Tests : public ::testing::TestWithParam> { +class Syr2Tests : public ::testing::TestWithParam> { }; TEST_P(Syr2Tests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 1, 42)); } TEST_P(Syr2Tests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(Syr2TestSuite, Syr2Tests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/syr2_usm.cpp b/tests/unit_tests/blas/level2/syr2_usm.cpp index 357200f79..94d3ce524 100644 --- a/tests/unit_tests/blas/level2/syr2_usm.cpp +++ b/tests/unit_tests/blas/level2/syr2_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -73,7 +73,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, vector x(ua), y(ua), A(ua); rand_vector(x, n, incx); rand_vector(y, n, incy); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); auto A_ref = A; @@ -90,13 +90,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::syr2(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::syr2(main_queue, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::syr2(main_queue, upper_lower, n, alpha, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::syr2(main_queue, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; @@ -105,13 +105,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::syr2, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syr2, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::syr2, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syr2, upper_lower, n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; @@ -125,7 +125,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -141,45 +141,45 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class Syr2UsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Syr2UsmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 1, 42)); } TEST_P(Syr2UsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 3, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, -3, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 1, 42)); } INSTANTIATE_TEST_SUITE_P(Syr2UsmTestSuite, Syr2UsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/syr_usm.cpp b/tests/unit_tests/blas/level2/syr_usm.cpp index 1c10bd493..9a693ae1e 100644 --- a/tests/unit_tests/blas/level2/syr_usm.cpp +++ b/tests/unit_tests/blas/level2/syr_usm.cpp @@ -47,7 +47,7 @@ extern std::vector devices; namespace { template -int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, int n, fp alpha, +int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -72,7 +72,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, auto ua = usm_allocator(cxt, *dev); vector x(ua), A(ua); rand_vector(x, n, incx); - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); auto A_ref = A; @@ -88,12 +88,12 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::syr( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::syr( main_queue, upper_lower, n, alpha, x.data(), incx, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::syr( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::syr( main_queue, upper_lower, n, alpha, x.data(), incx, A.data(), lda, dependencies); break; default: break; @@ -101,13 +101,13 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::syr, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syr, upper_lower, n, alpha, x.data(), incx, A.data(), lda, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::syr, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syr, upper_lower, n, alpha, x.data(), incx, A.data(), lda, dependencies); break; default: break; @@ -120,7 +120,7 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { return test_skipped; } @@ -136,45 +136,45 @@ int test(device *dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class SyrUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SyrUsmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 42)); } TEST_P(SyrUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 2, 42)); + oneapi::math::uplo::lower, 30, alpha, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 2, 42)); + oneapi::math::uplo::upper, 30, alpha, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, -2, 42)); + oneapi::math::uplo::lower, 30, alpha, -2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, -2, 42)); + oneapi::math::uplo::upper, 30, alpha, -2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, 30, alpha, 1, 42)); + oneapi::math::uplo::lower, 30, alpha, 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, 30, alpha, 1, 42)); + oneapi::math::uplo::upper, 30, alpha, 1, 42)); } INSTANTIATE_TEST_SUITE_P(SyrUsmTestSuite, SyrUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/tbmv.cpp b/tests/unit_tests/blas/level2/tbmv.cpp index c22f03c65..0fb975581 100644 --- a/tests/unit_tests/blas/level2/tbmv.cpp +++ b/tests/unit_tests/blas/level2/tbmv.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int k, int incx, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int k, int incx, int lda) { // Prepare data. vector x, x_ref, A; @@ -89,25 +89,25 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::tbmv(main_queue, upper_lower, transa, unit_nonunit, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::tbmv(main_queue, upper_lower, transa, unit_nonunit, n, k, A_buffer, lda, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::tbmv(main_queue, upper_lower, transa, unit_nonunit, n, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::tbmv(main_queue, upper_lower, transa, unit_nonunit, n, k, A_buffer, lda, x_buffer, incx); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::tbmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::tbmv, upper_lower, transa, unit_nonunit, n, k, A_buffer, lda, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::tbmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbmv, upper_lower, transa, unit_nonunit, n, k, A_buffer, lda, x_buffer, incx); break; default: break; @@ -119,7 +119,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -134,146 +134,146 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class TbmvTests : public ::testing::TestWithParam> { +class TbmvTests : public ::testing::TestWithParam> { }; TEST_P(TbmvTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbmvTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbmvTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbmvTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); } INSTANTIATE_TEST_SUITE_P(TbmvTestSuite, TbmvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/tbmv_usm.cpp b/tests/unit_tests/blas/level2/tbmv_usm.cpp index d3de28f21..56efc0742 100644 --- a/tests/unit_tests/blas/level2/tbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/tbmv_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int k, int incx, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int k, int incx, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -91,13 +91,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::tbmv(main_queue, upper_lower, transa, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::tbmv(main_queue, upper_lower, transa, unit_nonunit, n, k, A.data(), lda, x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::tbmv(main_queue, upper_lower, transa, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::tbmv(main_queue, upper_lower, transa, unit_nonunit, n, k, A.data(), lda, x.data(), incx, dependencies); break; @@ -106,13 +106,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::tbmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::tbmv, upper_lower, transa, unit_nonunit, n, k, A.data(), lda, x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::tbmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbmv, upper_lower, transa, unit_nonunit, n, k, A.data(), lda, x.data(), incx, dependencies); break; @@ -126,7 +126,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -142,145 +142,145 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class TbmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TbmvUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbmvUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbmvUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbmvUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); } INSTANTIATE_TEST_SUITE_P(TbmvUsmTestSuite, TbmvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/tbsv.cpp b/tests/unit_tests/blas/level2/tbsv.cpp index 305254936..7127a98d1 100644 --- a/tests/unit_tests/blas/level2/tbsv.cpp +++ b/tests/unit_tests/blas/level2/tbsv.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int k, int incx, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int k, int incx, int lda) { // Prepare data. vector x, x_ref, A; @@ -89,25 +89,25 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::tbsv(main_queue, upper_lower, transa, unit_nonunit, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::tbsv(main_queue, upper_lower, transa, unit_nonunit, n, k, A_buffer, lda, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::tbsv(main_queue, upper_lower, transa, unit_nonunit, n, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::tbsv(main_queue, upper_lower, transa, unit_nonunit, n, k, A_buffer, lda, x_buffer, incx); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::tbsv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::tbsv, upper_lower, transa, unit_nonunit, n, k, A_buffer, lda, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::tbsv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbsv, upper_lower, transa, unit_nonunit, n, k, A_buffer, lda, x_buffer, incx); break; default: break; @@ -119,7 +119,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -134,146 +134,146 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class TbsvTests : public ::testing::TestWithParam> { +class TbsvTests : public ::testing::TestWithParam> { }; TEST_P(TbsvTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbsvTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbsvTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbsvTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); } INSTANTIATE_TEST_SUITE_P(TbsvTestSuite, TbsvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/tbsv_usm.cpp b/tests/unit_tests/blas/level2/tbsv_usm.cpp index dbbadad7e..769f28669 100644 --- a/tests/unit_tests/blas/level2/tbsv_usm.cpp +++ b/tests/unit_tests/blas/level2/tbsv_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int k, int incx, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int k, int incx, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -91,13 +91,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::tbsv(main_queue, upper_lower, transa, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::tbsv(main_queue, upper_lower, transa, unit_nonunit, n, k, A.data(), lda, x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::tbsv(main_queue, upper_lower, transa, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::tbsv(main_queue, upper_lower, transa, unit_nonunit, n, k, A.data(), lda, x.data(), incx, dependencies); break; @@ -106,13 +106,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::tbsv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::tbsv, upper_lower, transa, unit_nonunit, n, k, A.data(), lda, x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::tbsv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbsv, upper_lower, transa, unit_nonunit, n, k, A.data(), lda, x.data(), incx, dependencies); break; @@ -126,7 +126,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -142,145 +142,145 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class TbsvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TbsvUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbsvUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbsvUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); } TEST_P(TbsvUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 5, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 5, 2, 42)); } INSTANTIATE_TEST_SUITE_P(TbsvUsmTestSuite, TbsvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/tpmv.cpp b/tests/unit_tests/blas/level2/tpmv.cpp index 04fbf1986..558fd901a 100644 --- a/tests/unit_tests/blas/level2/tpmv.cpp +++ b/tests/unit_tests/blas/level2/tpmv.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int incx) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx) { // Prepare data. vector x, x_ref, A; rand_vector(x, n, incx); @@ -87,25 +87,25 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::tpmv(main_queue, upper_lower, transa, unit_nonunit, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::tpmv(main_queue, upper_lower, transa, unit_nonunit, n, A_buffer, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::tpmv(main_queue, upper_lower, transa, unit_nonunit, n, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::tpmv(main_queue, upper_lower, transa, unit_nonunit, n, A_buffer, x_buffer, incx); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::tpmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::tpmv, upper_lower, transa, unit_nonunit, n, A_buffer, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::tpmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpmv, upper_lower, transa, unit_nonunit, n, A_buffer, x_buffer, incx); break; default: break; @@ -117,7 +117,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -132,146 +132,146 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class TpmvTests : public ::testing::TestWithParam> { +class TpmvTests : public ::testing::TestWithParam> { }; TEST_P(TpmvTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpmvTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpmvTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpmvTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); } INSTANTIATE_TEST_SUITE_P(TpmvTestSuite, TpmvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/tpmv_usm.cpp b/tests/unit_tests/blas/level2/tpmv_usm.cpp index 9760da308..04e0bae47 100644 --- a/tests/unit_tests/blas/level2/tpmv_usm.cpp +++ b/tests/unit_tests/blas/level2/tpmv_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int incx) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -89,13 +89,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::tpmv(main_queue, upper_lower, transa, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::tpmv(main_queue, upper_lower, transa, unit_nonunit, n, A.data(), x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::tpmv(main_queue, upper_lower, transa, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::tpmv(main_queue, upper_lower, transa, unit_nonunit, n, A.data(), x.data(), incx, dependencies); break; @@ -104,13 +104,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::tpmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::tpmv, upper_lower, transa, unit_nonunit, n, A.data(), x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::tpmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpmv, upper_lower, transa, unit_nonunit, n, A.data(), x.data(), incx, dependencies); break; @@ -124,7 +124,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -140,145 +140,145 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class TpmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TpmvUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpmvUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpmvUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpmvUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); } INSTANTIATE_TEST_SUITE_P(TpmvUsmTestSuite, TpmvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/tpsv.cpp b/tests/unit_tests/blas/level2/tpsv.cpp index 348ce132b..2f7e0423c 100644 --- a/tests/unit_tests/blas/level2/tpsv.cpp +++ b/tests/unit_tests/blas/level2/tpsv.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int incx) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx) { // Prepare data. vector x, x_ref, A; rand_vector(x, n, incx); @@ -87,25 +87,25 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::tpsv(main_queue, upper_lower, transa, unit_nonunit, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::tpsv(main_queue, upper_lower, transa, unit_nonunit, n, A_buffer, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::tpsv(main_queue, upper_lower, transa, unit_nonunit, n, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::tpsv(main_queue, upper_lower, transa, unit_nonunit, n, A_buffer, x_buffer, incx); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::tpsv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::tpsv, upper_lower, transa, unit_nonunit, n, A_buffer, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::tpsv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpsv, upper_lower, transa, unit_nonunit, n, A_buffer, x_buffer, incx); break; default: break; @@ -117,7 +117,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -132,146 +132,146 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class TpsvTests : public ::testing::TestWithParam> { +class TpsvTests : public ::testing::TestWithParam> { }; TEST_P(TpsvTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpsvTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpsvTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpsvTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); } INSTANTIATE_TEST_SUITE_P(TpsvTestSuite, TpsvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/tpsv_usm.cpp b/tests/unit_tests/blas/level2/tpsv_usm.cpp index 9cdeb57aa..f3980ed7f 100644 --- a/tests/unit_tests/blas/level2/tpsv_usm.cpp +++ b/tests/unit_tests/blas/level2/tpsv_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int incx) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -89,13 +89,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::tpsv(main_queue, upper_lower, transa, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::tpsv(main_queue, upper_lower, transa, unit_nonunit, n, A.data(), x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::tpsv(main_queue, upper_lower, transa, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::tpsv(main_queue, upper_lower, transa, unit_nonunit, n, A.data(), x.data(), incx, dependencies); break; @@ -104,13 +104,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::tpsv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::tpsv, upper_lower, transa, unit_nonunit, n, A.data(), x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::tpsv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpsv, upper_lower, transa, unit_nonunit, n, A.data(), x.data(), incx, dependencies); break; @@ -124,7 +124,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -140,145 +140,145 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class TpsvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TpsvUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpsvUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpsvUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); } TEST_P(TpsvUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2)); } INSTANTIATE_TEST_SUITE_P(TpsvUsmTestSuite, TpsvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/trmv.cpp b/tests/unit_tests/blas/level2/trmv.cpp index ef6fc4845..0ea3371bd 100644 --- a/tests/unit_tests/blas/level2/trmv.cpp +++ b/tests/unit_tests/blas/level2/trmv.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int incx, int lda) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, int lda) { // Prepare data. vector x, x_ref, A; rand_vector(x, n, incx); @@ -87,25 +87,25 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::trmv(main_queue, upper_lower, transa, unit_nonunit, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::trmv(main_queue, upper_lower, transa, unit_nonunit, n, A_buffer, lda, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::trmv(main_queue, upper_lower, transa, unit_nonunit, n, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::trmv(main_queue, upper_lower, transa, unit_nonunit, n, A_buffer, lda, x_buffer, incx); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::trmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trmv, upper_lower, transa, unit_nonunit, n, A_buffer, lda, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::trmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trmv, upper_lower, transa, unit_nonunit, n, A_buffer, lda, x_buffer, incx); break; default: break; @@ -117,7 +117,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -132,146 +132,146 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class TrmvTests : public ::testing::TestWithParam> { +class TrmvTests : public ::testing::TestWithParam> { }; TEST_P(TrmvTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrmvTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrmvTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrmvTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); } INSTANTIATE_TEST_SUITE_P(TrmvTestSuite, TrmvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/trmv_usm.cpp b/tests/unit_tests/blas/level2/trmv_usm.cpp index c444fe786..1b7fd27c6 100644 --- a/tests/unit_tests/blas/level2/trmv_usm.cpp +++ b/tests/unit_tests/blas/level2/trmv_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int incx, int lda) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -89,13 +89,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::trmv(main_queue, upper_lower, transa, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::trmv(main_queue, upper_lower, transa, unit_nonunit, n, A.data(), lda, x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::trmv(main_queue, upper_lower, transa, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::trmv(main_queue, upper_lower, transa, unit_nonunit, n, A.data(), lda, x.data(), incx, dependencies); break; @@ -104,13 +104,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::trmv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trmv, upper_lower, transa, unit_nonunit, n, A.data(), lda, x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::trmv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trmv, upper_lower, transa, unit_nonunit, n, A.data(), lda, x.data(), incx, dependencies); break; @@ -124,7 +124,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -140,145 +140,145 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class TrmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TrmvUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrmvUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrmvUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrmvUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); } INSTANTIATE_TEST_SUITE_P(TrmvUsmTestSuite, TrmvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/trsv.cpp b/tests/unit_tests/blas/level2/trsv.cpp index 137f4379e..f9279f597 100644 --- a/tests/unit_tests/blas/level2/trsv.cpp +++ b/tests/unit_tests/blas/level2/trsv.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int incx, int lda) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, int lda) { // Prepare data. vector x, x_ref, A; rand_vector(x, n, incx); @@ -87,25 +87,25 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::trsv(main_queue, upper_lower, transa, unit_nonunit, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::trsv(main_queue, upper_lower, transa, unit_nonunit, n, A_buffer, lda, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::trsv(main_queue, upper_lower, transa, unit_nonunit, n, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::trsv(main_queue, upper_lower, transa, unit_nonunit, n, A_buffer, lda, x_buffer, incx); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::trsv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trsv, upper_lower, transa, unit_nonunit, n, A_buffer, lda, x_buffer, incx); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::trsv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsv, upper_lower, transa, unit_nonunit, n, A_buffer, lda, x_buffer, incx); break; default: break; @@ -117,7 +117,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -132,146 +132,146 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class TrsvTests : public ::testing::TestWithParam> { +class TrsvTests : public ::testing::TestWithParam> { }; TEST_P(TrsvTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrsvTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrsvTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrsvTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); } INSTANTIATE_TEST_SUITE_P(TrsvTestSuite, TrsvTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level2/trsv_usm.cpp b/tests/unit_tests/blas/level2/trsv_usm.cpp index 908b36dbd..53b8c1df7 100644 --- a/tests/unit_tests/blas/level2/trsv_usm.cpp +++ b/tests/unit_tests/blas/level2/trsv_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose transa, oneapi::mkl::diag unit_nonunit, int n, int incx, int lda) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -89,13 +89,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::trsv(main_queue, upper_lower, transa, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::trsv(main_queue, upper_lower, transa, unit_nonunit, n, A.data(), lda, x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::trsv(main_queue, upper_lower, transa, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::trsv(main_queue, upper_lower, transa, unit_nonunit, n, A.data(), lda, x.data(), incx, dependencies); break; @@ -104,13 +104,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::trsv, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trsv, upper_lower, transa, unit_nonunit, n, A.data(), lda, x.data(), incx, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::trsv, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsv, upper_lower, transa, unit_nonunit, n, A.data(), lda, x.data(), incx, dependencies); break; @@ -124,7 +124,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -140,145 +140,145 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class TrsvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TrsvUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrsvUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, - oneapi::mkl::diag::nonunit, 30, 2, 42)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, + oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrsvUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); } TEST_P(TrsvUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::unit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 30, 2, 42)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 30, 2, 42)); } INSTANTIATE_TEST_SUITE_P(TrsvUsmTestSuite, TrsvUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/gemm.cpp b/tests/unit_tests/blas/level3/gemm.cpp index f892ce19d..c8f44a433 100644 --- a/tests/unit_tests/blas/level3/gemm.cpp +++ b/tests/unit_tests/blas/level3/gemm.cpp @@ -48,8 +48,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, int m, int n, int k, int lda, int ldb, int ldc, Tc alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose transa, + oneapi::math::transpose transb, int m, int n, int k, int lda, int ldb, int ldc, Tc alpha, Tc beta) { // Prepare data. vector> A, B; @@ -57,7 +57,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, rand_matrix(A, layout, transa, m, k, lda); rand_matrix(B, layout, transb, k, n, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, m, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, m, n, ldc); C_ref = C; // Call Reference GEMM. @@ -97,13 +97,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::gemm(main_queue, transa, transb, m, n, k, alpha, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::gemm(main_queue, transa, transb, m, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::gemm(main_queue, transa, transb, m, n, k, alpha, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::gemm(main_queue, transa, transb, m, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -111,13 +111,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemm, transa, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemm, transa, transb, m, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemm, transa, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm, transa, transb, m, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -130,7 +130,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -145,78 +145,78 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, return (int)good; } -class GemmTests : public ::testing::TestWithParam> { +class GemmTests : public ::testing::TestWithParam> { }; TEST_P(GemmTests, Bfloat16Bfloat16FloatPrecision) { float alpha(2.0); float beta(3.0); - EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); - EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); - EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); - EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + EXPECT_TRUEORSKIP((test( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + EXPECT_TRUEORSKIP((test( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + EXPECT_TRUEORSKIP((test( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + EXPECT_TRUEORSKIP((test( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmTests, HalfHalfFloatPrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmTests, RealHalfPrecision) { sycl::half alpha(2.0); sycl::half beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 3, 8, 9, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 3, 8, 9, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmTests, RealDoublePrecision) { @@ -225,49 +225,49 @@ TEST_P(GemmTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmTests, ComplexDoublePrecision) { @@ -276,38 +276,38 @@ TEST_P(GemmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); } INSTANTIATE_TEST_SUITE_P(GemmTestSuite, GemmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/gemm_usm.cpp b/tests/unit_tests/blas/level3/gemm_usm.cpp index a07092bf3..82665e896 100644 --- a/tests/unit_tests/blas/level3/gemm_usm.cpp +++ b/tests/unit_tests/blas/level3/gemm_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, - oneapi::mkl::transpose transb, int m, int n, int k, int lda, int ldb, int ldc, Tc alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose transa, + oneapi::math::transpose transb, int m, int n, int k, int lda, int ldb, int ldc, Tc alpha, Tc beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -76,7 +76,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, vector C(ua); rand_matrix(A, layout, transa, m, k, lda); rand_matrix(B, layout, transb, k, n, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, m, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, m, n, ldc); auto C_ref = C; @@ -97,13 +97,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::gemm(main_queue, transa, transb, m, n, k, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::gemm(main_queue, transa, transb, m, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::gemm(main_queue, transa, transb, m, n, k, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::gemm(main_queue, transa, transb, m, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -112,13 +112,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::gemm, transa, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemm, transa, transb, m, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::gemm, transa, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm, transa, transb, m, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -132,7 +132,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -148,74 +148,74 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::transpose transa, } class GemmUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemmUsmTests, Bfloat16Bfloat16FloatPrecision) { float alpha(2.0); float beta(3.0); - EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); - EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); - EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); - EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + EXPECT_TRUEORSKIP((test( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + EXPECT_TRUEORSKIP((test( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + EXPECT_TRUEORSKIP((test( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + EXPECT_TRUEORSKIP((test( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmUsmTests, HalfHalfFloatPrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmUsmTests, RealHalfPrecision) { sycl::half alpha(2.0); sycl::half beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmUsmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmUsmTests, RealDoublePrecision) { @@ -224,49 +224,49 @@ TEST_P(GemmUsmTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); } TEST_P(GemmUsmTests, ComplexDoublePrecision) { @@ -275,38 +275,38 @@ TEST_P(GemmUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::nontrans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::trans, 79, 83, 91, 103, 105, 106, alpha, beta))); EXPECT_TRUEORSKIP((test, std::complex>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::transpose::conjtrans, - oneapi::mkl::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, + oneapi::math::transpose::conjtrans, 79, 83, 91, 103, 105, 106, alpha, beta))); } INSTANTIATE_TEST_SUITE_P(GemmUsmTestSuite, GemmUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/hemm.cpp b/tests/unit_tests/blas/level3/hemm.cpp index 5e661c81f..fc7b25858 100644 --- a/tests/unit_tests/blas/level3/hemm.cpp +++ b/tests/unit_tests/blas/level3/hemm.cpp @@ -48,17 +48,17 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, int m, int n, int lda, int ldb, int ldc, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, int m, int n, int lda, int ldb, int ldc, fp alpha, fp beta) { // Prepare data. vector> A, B, C, C_ref; - if (left_right == oneapi::mkl::side::left) - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, m, lda); + if (left_right == oneapi::math::side::left) + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, m, lda); else - rand_matrix(A, oneapi::mkl::transpose::nontrans, n, n, lda); - rand_matrix(B, layout, oneapi::mkl::transpose::nontrans, m, n, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, m, n, ldc); + rand_matrix(A, oneapi::math::transpose::nontrans, n, n, lda); + rand_matrix(B, layout, oneapi::math::transpose::nontrans, m, n, ldb); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, m, n, ldc); C_ref = C; // Call Reference HEMM. @@ -96,13 +96,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::hemm(main_queue, left_right, upper_lower, m, n, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::hemm(main_queue, left_right, upper_lower, m, n, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::hemm(main_queue, left_right, upper_lower, m, n, alpha, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::hemm(main_queue, left_right, upper_lower, m, n, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -110,13 +110,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hemm, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hemm, left_right, upper_lower, m, n, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hemm, left_right, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hemm, left_right, upper_lower, m, n, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -129,7 +129,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -145,23 +145,23 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, return (int)good; } -class HemmTests : public ::testing::TestWithParam> { +class HemmTests : public ::testing::TestWithParam> { }; TEST_P(HemmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(HemmTests, ComplexDoublePrecision) { @@ -170,23 +170,23 @@ TEST_P(HemmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(HemmTestSuite, HemmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/hemm_usm.cpp b/tests/unit_tests/blas/level3/hemm_usm.cpp index af0cf5217..6d3314720 100644 --- a/tests/unit_tests/blas/level3/hemm_usm.cpp +++ b/tests/unit_tests/blas/level3/hemm_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, int m, int n, int lda, int ldb, int ldc, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, int m, int n, int lda, int ldb, int ldc, fp alpha, fp beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -72,12 +72,12 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, // Prepare data. auto ua = usm_allocator(cxt, *dev); vector A(ua), B(ua), C(ua); - if (left_right == oneapi::mkl::side::left) - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, m, lda); + if (left_right == oneapi::math::side::left) + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, m, lda); else - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); - rand_matrix(B, layout, oneapi::mkl::transpose::nontrans, m, n, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, m, n, ldc); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); + rand_matrix(B, layout, oneapi::math::transpose::nontrans, m, n, ldb); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, m, n, ldc); auto C_ref = C; @@ -95,13 +95,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::hemm(main_queue, left_right, upper_lower, m, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::hemm(main_queue, left_right, upper_lower, m, n, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::hemm(main_queue, left_right, upper_lower, m, n, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::hemm(main_queue, left_right, upper_lower, m, n, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -110,13 +110,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::hemm, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::hemm, left_right, upper_lower, m, n, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::hemm, left_right, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hemm, left_right, upper_lower, m, n, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -130,7 +130,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -146,22 +146,22 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, } class HemmUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HemmUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(HemmUsmTests, ComplexDoublePrecision) { @@ -170,23 +170,23 @@ TEST_P(HemmUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(HemmUsmTestSuite, HemmUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/her2k.cpp b/tests/unit_tests/blas/level3/her2k.cpp index a2639426d..cfc3674d4 100644 --- a/tests/unit_tests/blas/level3/her2k.cpp +++ b/tests/unit_tests/blas/level3/her2k.cpp @@ -48,8 +48,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, fp_scalar beta) { fp alpha_row(alpha.real(), -alpha.imag()); @@ -57,7 +57,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, vector> A, B, C, C_ref; rand_matrix(A, layout, trans, n, k, lda); rand_matrix(B, layout, trans, n, k, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, n, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, n, n, ldc); C_ref = C; // Call Reference HER2K. @@ -97,13 +97,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::her2k(main_queue, upper_lower, trans, n, k, alpha, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::her2k(main_queue, upper_lower, trans, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::her2k(main_queue, upper_lower, trans, n, k, alpha, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::her2k(main_queue, upper_lower, trans, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -111,13 +111,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::her2k, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::her2k, upper_lower, trans, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::her2k, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::her2k, upper_lower, trans, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -130,7 +130,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -146,24 +146,24 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class Her2kTests : public ::testing::TestWithParam> { +class Her2kTests : public ::testing::TestWithParam> { }; TEST_P(Her2kTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); float beta(1.0); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); } TEST_P(Her2kTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -171,23 +171,23 @@ TEST_P(Her2kTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); double beta(1.0); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); } INSTANTIATE_TEST_SUITE_P(Her2kTestSuite, Her2kTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/her2k_usm.cpp b/tests/unit_tests/blas/level3/her2k_usm.cpp index 0bed61698..202e7c29d 100644 --- a/tests/unit_tests/blas/level3/her2k_usm.cpp +++ b/tests/unit_tests/blas/level3/her2k_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, fp_scalar beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -76,7 +76,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, vector A(ua), B(ua), C(ua); rand_matrix(A, layout, trans, n, k, lda); rand_matrix(B, layout, trans, n, k, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, n, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, n, n, ldc); auto C_ref = C; @@ -97,13 +97,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::her2k(main_queue, upper_lower, trans, n, k, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::her2k(main_queue, upper_lower, trans, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::her2k(main_queue, upper_lower, trans, n, k, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::her2k(main_queue, upper_lower, trans, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -112,13 +112,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::her2k, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::her2k, upper_lower, trans, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::her2k, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::her2k, upper_lower, trans, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -132,7 +132,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -148,23 +148,23 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class Her2kUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Her2kUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); float beta(1.0); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); } TEST_P(Her2kUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -172,23 +172,23 @@ TEST_P(Her2kUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); double beta(1.0); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, 72, 27, 101, 102, 103, alpha, beta))); } INSTANTIATE_TEST_SUITE_P(Her2kUsmTestSuite, Her2kUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/herk.cpp b/tests/unit_tests/blas/level3/herk.cpp index 8ba8dd041..eea240cd8 100644 --- a/tests/unit_tests/blas/level3/herk.cpp +++ b/tests/unit_tests/blas/level3/herk.cpp @@ -48,13 +48,13 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, int n, int k, int lda, int ldc, fp_scalar alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, int n, int k, int lda, int ldc, fp_scalar alpha, fp_scalar beta) { // Prepare data. vector> A, C, C_ref; rand_matrix(A, layout, trans, n, k, lda); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, n, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, n, n, ldc); C_ref = C; // Call Reference HERK. @@ -91,25 +91,25 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::herk(main_queue, upper_lower, trans, n, k, alpha, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::herk(main_queue, upper_lower, trans, n, k, alpha, A_buffer, lda, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::herk(main_queue, upper_lower, trans, n, k, alpha, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::herk(main_queue, upper_lower, trans, n, k, alpha, A_buffer, lda, beta, C_buffer, ldc); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::herk, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::herk, upper_lower, trans, n, k, alpha, A_buffer, lda, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::herk, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::herk, upper_lower, trans, n, k, alpha, A_buffer, lda, beta, C_buffer, ldc); break; default: break; @@ -121,7 +121,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -137,24 +137,24 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class HerkTests : public ::testing::TestWithParam> { +class HerkTests : public ::testing::TestWithParam> { }; TEST_P(HerkTests, ComplexSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); } TEST_P(HerkTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -162,23 +162,23 @@ TEST_P(HerkTests, ComplexDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); } INSTANTIATE_TEST_SUITE_P(HerkTestSuite, HerkTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/herk_usm.cpp b/tests/unit_tests/blas/level3/herk_usm.cpp index fec764421..c2c51b853 100644 --- a/tests/unit_tests/blas/level3/herk_usm.cpp +++ b/tests/unit_tests/blas/level3/herk_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, int n, int k, int lda, int ldc, fp_scalar alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, int n, int k, int lda, int ldc, fp_scalar alpha, fp_scalar beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -73,7 +73,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, auto ua = usm_allocator(cxt, *dev); vector A(ua), C(ua); rand_matrix(A, layout, trans, n, k, lda); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, n, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, n, n, ldc); auto C_ref = C; @@ -92,13 +92,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::herk(main_queue, upper_lower, trans, n, k, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::herk(main_queue, upper_lower, trans, n, k, alpha, A.data(), lda, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::herk(main_queue, upper_lower, trans, n, k, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::herk(main_queue, upper_lower, trans, n, k, alpha, A.data(), lda, beta, C.data(), ldc, dependencies); break; @@ -107,13 +107,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::herk, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::herk, upper_lower, trans, n, k, alpha, A.data(), lda, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::herk, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::herk, upper_lower, trans, n, k, alpha, A.data(), lda, beta, C.data(), ldc, dependencies); break; @@ -127,7 +127,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -143,23 +143,23 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class HerkUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HerkUsmTests, ComplexSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); } TEST_P(HerkUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -167,23 +167,23 @@ TEST_P(HerkUsmTests, ComplexDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, 72, 27, 101, 103, alpha, beta))); } INSTANTIATE_TEST_SUITE_P(HerkUsmTestSuite, HerkUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/symm.cpp b/tests/unit_tests/blas/level3/symm.cpp index 08a7dcd2e..13267aee0 100644 --- a/tests/unit_tests/blas/level3/symm.cpp +++ b/tests/unit_tests/blas/level3/symm.cpp @@ -48,17 +48,17 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, int m, int n, int lda, int ldb, int ldc, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, int m, int n, int lda, int ldb, int ldc, fp alpha, fp beta) { // Prepare data. vector> A, B, C, C_ref; - if (left_right == oneapi::mkl::side::left) - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, m, lda); + if (left_right == oneapi::math::side::left) + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, m, lda); else - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); - rand_matrix(B, layout, oneapi::mkl::transpose::nontrans, m, n, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, m, n, ldc); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); + rand_matrix(B, layout, oneapi::math::transpose::nontrans, m, n, ldb); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, m, n, ldc); C_ref = C; // Call Reference SYMM. @@ -96,13 +96,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::symm(main_queue, left_right, upper_lower, m, n, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::symm(main_queue, left_right, upper_lower, m, n, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::symm(main_queue, left_right, upper_lower, m, n, alpha, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::symm(main_queue, left_right, upper_lower, m, n, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -110,13 +110,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::symm, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::symm, left_right, upper_lower, m, n, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::symm, left_right, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::symm, left_right, upper_lower, m, n, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -129,7 +129,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -145,23 +145,23 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, return (int)good; } -class SymmTests : public ::testing::TestWithParam> { +class SymmTests : public ::testing::TestWithParam> { }; TEST_P(SymmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, 72, 27, 101, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, 72, 27, 101, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, 72, 27, 101, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, 72, 27, 101, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(SymmTests, RealDoublePrecision) { @@ -170,32 +170,32 @@ TEST_P(SymmTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, 72, 27, 101, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, 72, 27, 101, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, 72, 27, 101, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, 72, 27, 101, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(SymmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(SymmTests, ComplexDoublePrecision) { @@ -204,23 +204,23 @@ TEST_P(SymmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(SymmTestSuite, SymmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/symm_usm.cpp b/tests/unit_tests/blas/level3/symm_usm.cpp index 7785c922b..302dd1beb 100644 --- a/tests/unit_tests/blas/level3/symm_usm.cpp +++ b/tests/unit_tests/blas/level3/symm_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, int m, int n, int lda, int ldb, int ldc, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, int m, int n, int lda, int ldb, int ldc, fp alpha, fp beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { @@ -72,12 +72,12 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, // Prepare data. auto ua = usm_allocator(cxt, *dev); vector A(ua), B(ua), C(ua); - if (left_right == oneapi::mkl::side::left) - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, m, m, lda); + if (left_right == oneapi::math::side::left) + rand_matrix(A, layout, oneapi::math::transpose::nontrans, m, m, lda); else - rand_matrix(A, layout, oneapi::mkl::transpose::nontrans, n, n, lda); - rand_matrix(B, layout, oneapi::mkl::transpose::nontrans, m, n, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, m, n, ldc); + rand_matrix(A, layout, oneapi::math::transpose::nontrans, n, n, lda); + rand_matrix(B, layout, oneapi::math::transpose::nontrans, m, n, ldb); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, m, n, ldc); auto C_ref = C; @@ -95,13 +95,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::symm(main_queue, left_right, upper_lower, m, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::symm(main_queue, left_right, upper_lower, m, n, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::symm(main_queue, left_right, upper_lower, m, n, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::symm(main_queue, left_right, upper_lower, m, n, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -110,13 +110,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::symm, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::symm, left_right, upper_lower, m, n, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::symm, left_right, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::symm, left_right, upper_lower, m, n, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -130,7 +130,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -146,22 +146,22 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, } class SymmUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SymmUsmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, 72, 27, 101, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, 72, 27, 101, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, 72, 27, 101, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, 72, 27, 101, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(SymmUsmTests, RealDoublePrecision) { @@ -170,32 +170,32 @@ TEST_P(SymmUsmTests, RealDoublePrecision) { double alpha(2.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, 72, 27, 101, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, 72, 27, 101, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, 72, 27, 101, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, 72, 27, 101, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(SymmUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(SymmUsmTests, ComplexDoublePrecision) { @@ -204,23 +204,23 @@ TEST_P(SymmUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, + oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, + oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(SymmUsmTestSuite, SymmUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/syr2k.cpp b/tests/unit_tests/blas/level3/syr2k.cpp index 2709c9e87..e735f0ace 100644 --- a/tests/unit_tests/blas/level3/syr2k.cpp +++ b/tests/unit_tests/blas/level3/syr2k.cpp @@ -48,13 +48,13 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, fp beta) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, fp beta) { // Prepare data. vector> A, B, C, C_ref; rand_matrix(A, layout, trans, n, k, lda); rand_matrix(B, layout, trans, n, k, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, n, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, n, n, ldc); C_ref = C; // Call Reference SYR2K. @@ -92,13 +92,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::syr2k(main_queue, upper_lower, trans, n, k, alpha, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::syr2k(main_queue, upper_lower, trans, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::syr2k(main_queue, upper_lower, trans, n, k, alpha, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::syr2k(main_queue, upper_lower, trans, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -106,13 +106,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::syr2k, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syr2k, upper_lower, trans, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::syr2k, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syr2k, upper_lower, trans, n, k, alpha, A_buffer, lda, B_buffer, ldb, beta, C_buffer, ldc); break; @@ -125,7 +125,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -141,23 +141,23 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class Syr2kTests : public ::testing::TestWithParam> { +class Syr2kTests : public ::testing::TestWithParam> { }; TEST_P(Syr2kTests, RealSinglePrecision) { float alpha(3.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); } TEST_P(Syr2kTests, RealDoublePrecision) { @@ -166,33 +166,33 @@ TEST_P(Syr2kTests, RealDoublePrecision) { double alpha(3.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); } TEST_P(Syr2kTests, ComplexSinglePrecision) { std::complex alpha(3.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); } TEST_P(Syr2kTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -200,23 +200,23 @@ TEST_P(Syr2kTests, ComplexDoublePrecision) { std::complex alpha(3.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(Syr2kTestSuite, Syr2kTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/syr2k_usm.cpp b/tests/unit_tests/blas/level3/syr2k_usm.cpp index 403f5e89a..9a5899c6c 100644 --- a/tests/unit_tests/blas/level3/syr2k_usm.cpp +++ b/tests/unit_tests/blas/level3/syr2k_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, fp beta) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, fp beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -73,7 +73,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, vector A(ua), B(ua), C(ua); rand_matrix(A, layout, trans, n, k, lda); rand_matrix(B, layout, trans, n, k, ldb); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, n, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, n, n, ldc); auto C_ref = C; @@ -92,13 +92,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::syr2k(main_queue, upper_lower, trans, n, k, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::syr2k(main_queue, upper_lower, trans, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::syr2k(main_queue, upper_lower, trans, n, k, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::syr2k(main_queue, upper_lower, trans, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -107,13 +107,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::syr2k, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syr2k, upper_lower, trans, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::syr2k, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syr2k, upper_lower, trans, n, k, alpha, A.data(), lda, B.data(), ldb, beta, C.data(), ldc, dependencies); break; @@ -127,7 +127,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -143,22 +143,22 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class Syr2kUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Syr2kUsmTests, RealSinglePrecision) { float alpha(3.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); } TEST_P(Syr2kUsmTests, RealDoublePrecision) { @@ -167,33 +167,33 @@ TEST_P(Syr2kUsmTests, RealDoublePrecision) { double alpha(3.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); } TEST_P(Syr2kUsmTests, ComplexSinglePrecision) { std::complex alpha(3.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); } TEST_P(Syr2kUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -201,23 +201,23 @@ TEST_P(Syr2kUsmTests, ComplexDoublePrecision) { std::complex alpha(3.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, 73, 27, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(Syr2kUsmTestSuite, Syr2kUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/syrk.cpp b/tests/unit_tests/blas/level3/syrk.cpp index 74ec2be38..c0c281531 100644 --- a/tests/unit_tests/blas/level3/syrk.cpp +++ b/tests/unit_tests/blas/level3/syrk.cpp @@ -48,12 +48,12 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, int n, int k, int lda, int ldc, fp alpha, fp beta) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, int n, int k, int lda, int ldc, fp alpha, fp beta) { // Prepare data. vector> A, C, C_ref; rand_matrix(A, layout, trans, n, k, lda); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, n, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, n, n, ldc); C_ref = C; // Call Reference SYRK. @@ -90,25 +90,25 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::syrk(main_queue, upper_lower, trans, n, k, alpha, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::syrk(main_queue, upper_lower, trans, n, k, alpha, A_buffer, lda, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::syrk(main_queue, upper_lower, trans, n, k, alpha, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::syrk(main_queue, upper_lower, trans, n, k, alpha, A_buffer, lda, beta, C_buffer, ldc); break; default: break; } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::syrk, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syrk, upper_lower, trans, n, k, alpha, A_buffer, lda, beta, C_buffer, ldc); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::syrk, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syrk, upper_lower, trans, n, k, alpha, A_buffer, lda, beta, C_buffer, ldc); break; default: break; @@ -120,7 +120,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -136,23 +136,23 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, return (int)good; } -class SyrkTests : public ::testing::TestWithParam> { +class SyrkTests : public ::testing::TestWithParam> { }; TEST_P(SyrkTests, RealSinglePrecision) { float alpha(3.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); } TEST_P(SyrkTests, RealDoublePrecision) { @@ -161,33 +161,33 @@ TEST_P(SyrkTests, RealDoublePrecision) { double alpha(3.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); } TEST_P(SyrkTests, ComplexSinglePrecision) { std::complex alpha(3.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); } TEST_P(SyrkTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -195,23 +195,23 @@ TEST_P(SyrkTests, ComplexDoublePrecision) { std::complex alpha(3.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(SyrkTestSuite, SyrkTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/syrk_usm.cpp b/tests/unit_tests/blas/level3/syrk_usm.cpp index 00ed1cab4..6e912a9ac 100644 --- a/tests/unit_tests/blas/level3/syrk_usm.cpp +++ b/tests/unit_tests/blas/level3/syrk_usm.cpp @@ -47,8 +47,8 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, - oneapi::mkl::transpose trans, int n, int k, int lda, int ldc, fp alpha, fp beta) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, int n, int k, int lda, int ldc, fp alpha, fp beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -72,7 +72,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, auto ua = usm_allocator(cxt, *dev); vector A(ua), C(ua); rand_matrix(A, layout, trans, n, k, lda); - rand_matrix(C, layout, oneapi::mkl::transpose::nontrans, n, n, ldc); + rand_matrix(C, layout, oneapi::math::transpose::nontrans, n, n, ldc); auto C_ref = C; @@ -90,13 +90,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::syrk(main_queue, upper_lower, trans, n, k, + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::syrk(main_queue, upper_lower, trans, n, k, alpha, A.data(), lda, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::syrk(main_queue, upper_lower, trans, n, k, + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::syrk(main_queue, upper_lower, trans, n, k, alpha, A.data(), lda, beta, C.data(), ldc, dependencies); break; @@ -105,13 +105,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::syrk, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syrk, upper_lower, trans, n, k, alpha, A.data(), lda, beta, C.data(), ldc, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::syrk, upper_lower, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syrk, upper_lower, trans, n, k, alpha, A.data(), lda, beta, C.data(), ldc, dependencies); break; @@ -125,7 +125,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -141,22 +141,22 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::uplo upper_lower, } class SyrkUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SyrkUsmTests, RealSinglePrecision) { float alpha(3.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); } TEST_P(SyrkUsmTests, RealDoublePrecision) { @@ -165,33 +165,33 @@ TEST_P(SyrkUsmTests, RealDoublePrecision) { double alpha(3.0); double beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, 73, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, 73, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); } TEST_P(SyrkUsmTests, ComplexSinglePrecision) { std::complex alpha(3.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); } TEST_P(SyrkUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -199,23 +199,23 @@ TEST_P(SyrkUsmTests, ComplexDoublePrecision) { std::complex alpha(3.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, + oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, 73, 27, 101, 103, alpha, beta)); + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, + oneapi::math::transpose::trans, 73, 27, 101, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(SyrkUsmTestSuite, SyrkUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/trmm.cpp b/tests/unit_tests/blas/level3/trmm.cpp index 8629a58df..1dbaa4820 100644 --- a/tests/unit_tests/blas/level3/trmm.cpp +++ b/tests/unit_tests/blas/level3/trmm.cpp @@ -48,17 +48,17 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::diag unit_nonunit, int m, int n, int lda, int ldb, fp alpha) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::diag unit_nonunit, int m, int n, int lda, int ldb, fp alpha) { // Prepare data. vector> A, B, B_ref; - if (left_right == oneapi::mkl::side::right) + if (left_right == oneapi::math::side::right) rand_matrix(A, layout, transa, n, n, lda); else rand_matrix(A, layout, transa, m, m, lda); - rand_matrix(B, layout, oneapi::mkl::transpose::nontrans, m, n, ldb); + rand_matrix(B, layout, oneapi::math::transpose::nontrans, m, n, ldb); B_ref = B; // Call Reference TRMM. @@ -96,13 +96,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::trmm(main_queue, left_right, upper_lower, transa, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::trmm(main_queue, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::trmm(main_queue, left_right, upper_lower, transa, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::trmm(main_queue, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; @@ -110,13 +110,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::trmm, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trmm, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::trmm, left_right, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trmm, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; @@ -129,7 +129,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -145,50 +145,50 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, return (int)good; } -class TrmmTests : public ::testing::TestWithParam> { +class TrmmTests : public ::testing::TestWithParam> { }; TEST_P(TrmmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrmmTests, RealDoublePrecision) { @@ -196,103 +196,103 @@ TEST_P(TrmmTests, RealDoublePrecision) { double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrmmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrmmTests, ComplexDoublePrecision) { @@ -300,67 +300,67 @@ TEST_P(TrmmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } INSTANTIATE_TEST_SUITE_P(TrmmTestSuite, TrmmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/trmm_usm.cpp b/tests/unit_tests/blas/level3/trmm_usm.cpp index 3dec90f5b..0e885b37e 100644 --- a/tests/unit_tests/blas/level3/trmm_usm.cpp +++ b/tests/unit_tests/blas/level3/trmm_usm.cpp @@ -47,9 +47,9 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::diag unit_nonunit, int m, int n, int lda, int ldb, fp alpha) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::diag unit_nonunit, int m, int n, int lda, int ldb, fp alpha) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -72,12 +72,12 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, // Prepare data. auto ua = usm_allocator(cxt, *dev); vector A(ua), B(ua); - if (left_right == oneapi::mkl::side::right) + if (left_right == oneapi::math::side::right) rand_matrix(A, layout, transa, n, n, lda); else rand_matrix(A, layout, transa, m, m, lda); - rand_matrix(B, layout, oneapi::mkl::transpose::nontrans, m, n, ldb); + rand_matrix(B, layout, oneapi::math::transpose::nontrans, m, n, ldb); auto B_ref = B; @@ -97,13 +97,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::trmm( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::trmm( main_queue, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A.data(), lda, B.data(), ldb, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::trmm( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::trmm( main_queue, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A.data(), lda, B.data(), ldb, dependencies); break; @@ -112,13 +112,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::trmm, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trmm, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A.data(), lda, B.data(), ldb, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::trmm, left_right, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trmm, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A.data(), lda, B.data(), ldb, dependencies); break; @@ -132,7 +132,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -148,49 +148,49 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, } class TrmmUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TrmmUsmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrmmUsmTests, RealDoublePrecision) { @@ -198,103 +198,103 @@ TEST_P(TrmmUsmTests, RealDoublePrecision) { double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrmmUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrmmUsmTests, ComplexDoublePrecision) { @@ -302,67 +302,67 @@ TEST_P(TrmmUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } INSTANTIATE_TEST_SUITE_P(TrmmUsmTestSuite, TrmmUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/trsm.cpp b/tests/unit_tests/blas/level3/trsm.cpp index f8110ce2b..0eac2ff36 100644 --- a/tests/unit_tests/blas/level3/trsm.cpp +++ b/tests/unit_tests/blas/level3/trsm.cpp @@ -48,17 +48,17 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::diag unit_nonunit, int m, int n, int lda, int ldb, fp alpha) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::diag unit_nonunit, int m, int n, int lda, int ldb, fp alpha) { // Prepare data. vector> A, B, B_ref; - if (left_right == oneapi::mkl::side::right) + if (left_right == oneapi::math::side::right) rand_trsm_matrix(A, layout, transa, n, n, lda); else rand_trsm_matrix(A, layout, transa, m, m, lda); - rand_matrix(B, layout, oneapi::mkl::transpose::nontrans, m, n, ldb); + rand_matrix(B, layout, oneapi::math::transpose::nontrans, m, n, ldb); B_ref = B; // Call Reference TRSM. @@ -96,13 +96,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - oneapi::mkl::blas::column_major::trsm(main_queue, left_right, upper_lower, transa, + case oneapi::math::layout::col_major: + oneapi::math::blas::column_major::trsm(main_queue, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; - case oneapi::mkl::layout::row_major: - oneapi::mkl::blas::row_major::trsm(main_queue, left_right, upper_lower, transa, + case oneapi::math::layout::row_major: + oneapi::math::blas::row_major::trsm(main_queue, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; @@ -110,13 +110,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, } #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::trsm, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trsm, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::trsm, left_right, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsm, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A_buffer, lda, B_buffer, ldb); break; @@ -129,7 +129,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -145,74 +145,74 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, return (int)good; } -class TrsmTests : public ::testing::TestWithParam> { +class TrsmTests : public ::testing::TestWithParam> { }; TEST_P(TrsmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrsmTests, RealDoublePrecision) { @@ -220,167 +220,167 @@ TEST_P(TrsmTests, RealDoublePrecision) { double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrsmTests, ComplexDoublePrecision) { @@ -388,107 +388,107 @@ TEST_P(TrsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } INSTANTIATE_TEST_SUITE_P(TrsmTestSuite, TrsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/blas/level3/trsm_usm.cpp b/tests/unit_tests/blas/level3/trsm_usm.cpp index dd041f765..86b958d82 100644 --- a/tests/unit_tests/blas/level3/trsm_usm.cpp +++ b/tests/unit_tests/blas/level3/trsm_usm.cpp @@ -47,9 +47,9 @@ extern std::vector devices; namespace { template -int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, - oneapi::mkl::uplo upper_lower, oneapi::mkl::transpose transa, - oneapi::mkl::diag unit_nonunit, int m, int n, int lda, int ldb, fp alpha) { +int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + oneapi::math::diag unit_nonunit, int m, int n, int lda, int ldb, fp alpha) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -72,12 +72,12 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, // Prepare data. auto ua = usm_allocator(cxt, *dev); vector A(ua), B(ua); - if (left_right == oneapi::mkl::side::right) + if (left_right == oneapi::math::side::right) rand_trsm_matrix(A, layout, transa, n, n, lda); else rand_trsm_matrix(A, layout, transa, m, m, lda); - rand_matrix(B, layout, oneapi::mkl::transpose::nontrans, m, n, ldb); + rand_matrix(B, layout, oneapi::math::transpose::nontrans, m, n, ldb); auto B_ref = B; @@ -97,13 +97,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, try { #ifdef CALL_RT_API switch (layout) { - case oneapi::mkl::layout::col_major: - done = oneapi::mkl::blas::column_major::trsm( + case oneapi::math::layout::col_major: + done = oneapi::math::blas::column_major::trsm( main_queue, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A.data(), lda, B.data(), ldb, dependencies); break; - case oneapi::mkl::layout::row_major: - done = oneapi::mkl::blas::row_major::trsm( + case oneapi::math::layout::row_major: + done = oneapi::math::blas::row_major::trsm( main_queue, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A.data(), lda, B.data(), ldb, dependencies); break; @@ -112,13 +112,13 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, done.wait(); #else switch (layout) { - case oneapi::mkl::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::column_major::trsm, + case oneapi::math::layout::col_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trsm, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A.data(), lda, B.data(), ldb, dependencies); break; - case oneapi::mkl::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::mkl::blas::row_major::trsm, left_right, + case oneapi::math::layout::row_major: + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsm, left_right, upper_lower, transa, unit_nonunit, m, n, alpha, A.data(), lda, B.data(), ldb, dependencies); break; @@ -132,7 +132,7 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } @@ -149,73 +149,73 @@ int test(device* dev, oneapi::mkl::layout layout, oneapi::mkl::side left_right, } class TrsmUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TrsmUsmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrsmUsmTests, RealDoublePrecision) { @@ -223,167 +223,167 @@ TEST_P(TrsmUsmTests, RealDoublePrecision) { double alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::unit, 72, 27, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrsmUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } TEST_P(TrsmUsmTests, ComplexDoublePrecision) { @@ -391,107 +391,107 @@ TEST_P(TrsmUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::trans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::trans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::lower, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::left, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::left, oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::conjtrans, - oneapi::mkl::diag::unit, 72, 27, 101, 102, alpha)); + oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::conjtrans, + oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::trans, oneapi::mkl::diag::nonunit, 72, 27, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::lower, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::left, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::mkl::side::right, - oneapi::mkl::uplo::upper, oneapi::mkl::transpose::conjtrans, oneapi::mkl::diag::nonunit, 72, + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); } INSTANTIATE_TEST_SUITE_P(TrsmUsmTestSuite, TrsmUsmTests, ::testing::Combine(testing::ValuesIn(devices), - testing::Values(oneapi::mkl::layout::col_major, - oneapi::mkl::layout::row_major)), + testing::Values(oneapi::math::layout::col_major, + oneapi::math::layout::row_major)), ::LayoutDeviceNamePrint()); } // anonymous namespace diff --git a/tests/unit_tests/dft/include/compute_inplace.hpp b/tests/unit_tests/dft/include/compute_inplace.hpp index d0f3bbb0f..a423476fe 100644 --- a/tests/unit_tests/dft/include/compute_inplace.hpp +++ b/tests/unit_tests/dft/include/compute_inplace.hpp @@ -23,7 +23,7 @@ #include "compute_tester.hpp" #include -template +template int DFT_Test::test_in_place_buffer() { if (!init(MemoryAccessModel::buffer)) { return test_skipped; @@ -31,7 +31,7 @@ int DFT_Test::test_in_place_buffer() { auto modified_strides_fwd = this->strides_fwd; auto modified_strides_bwd = this->strides_bwd; - if (domain == oneapi::mkl::dft::domain::REAL) { + if (domain == oneapi::math::dft::domain::REAL) { // both input and output strides must be set auto default_conjuate_strides = get_conjugate_even_complex_strides(sizes); std::ptrdiff_t rank = static_cast(sizes.size()); @@ -57,30 +57,30 @@ int DFT_Test::test_in_place_buffer() { auto ref_distance = std::accumulate(sizes.begin(), sizes.end(), 1, std::multiplies<>()); descriptor_t descriptor{ sizes }; - descriptor.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::INPLACE); - if constexpr (domain == oneapi::mkl::dft::domain::REAL) { - descriptor.set_value(oneapi::mkl::dft::config_param::CONJUGATE_EVEN_STORAGE, - oneapi::mkl::dft::config_value::COMPLEX_COMPLEX); - descriptor.set_value(oneapi::mkl::dft::config_param::PACKED_FORMAT, - oneapi::mkl::dft::config_value::CCE_FORMAT); - } - descriptor.set_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, batches); - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_DISTANCE, forward_distance); - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_DISTANCE, backward_distance); + descriptor.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::INPLACE); + if constexpr (domain == oneapi::math::dft::domain::REAL) { + descriptor.set_value(oneapi::math::dft::config_param::CONJUGATE_EVEN_STORAGE, + oneapi::math::dft::config_value::COMPLEX_COMPLEX); + descriptor.set_value(oneapi::math::dft::config_param::PACKED_FORMAT, + oneapi::math::dft::config_value::CCE_FORMAT); + } + descriptor.set_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, batches); + descriptor.set_value(oneapi::math::dft::config_param::FWD_DISTANCE, forward_distance); + descriptor.set_value(oneapi::math::dft::config_param::BWD_DISTANCE, backward_distance); if (modified_strides_fwd.size()) { - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_STRIDES, + descriptor.set_value(oneapi::math::dft::config_param::FWD_STRIDES, modified_strides_fwd.data()); } if (modified_strides_bwd.size()) { - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_STRIDES, + descriptor.set_value(oneapi::math::dft::config_param::BWD_STRIDES, modified_strides_bwd.data()); } commit_descriptor(descriptor, sycl_queue); std::vector inout_host( strided_copy(input, sizes, modified_strides_fwd, batches, forward_distance)); - int real_multiplier = (domain == oneapi::mkl::dft::domain::REAL ? 2 : 1); + int real_multiplier = (domain == oneapi::math::dft::domain::REAL ? 2 : 1); inout_host.resize( cast_unsigned(std::max(forward_distance, real_multiplier * backward_distance) * batches + get_default(modified_strides_bwd, 0, 0L) * real_multiplier)); @@ -88,19 +88,19 @@ int DFT_Test::test_in_place_buffer() { { sycl::buffer inout_buf{ inout_host }; - oneapi::mkl::dft::compute_forward(descriptor, inout_buf); + oneapi::math::dft::compute_forward(descriptor, inout_buf); { auto acc_host = inout_buf.get_host_access(); auto ptr_host = reinterpret_cast(acc_host.get_pointer()); for (std::int64_t i = 0; i < batches; i++) { - EXPECT_TRUE(check_equal_strided( + EXPECT_TRUE(check_equal_strided( ptr_host + backward_distance * i, out_host_ref.data() + ref_distance * i, sizes, modified_strides_bwd, abs_error_margin, rel_error_margin, std::cout)); } } - oneapi::mkl::dft::compute_backward, + oneapi::math::dft::compute_backward, FwdInputType>(descriptor, inout_buf); } @@ -118,7 +118,7 @@ int DFT_Test::test_in_place_buffer() { return !::testing::Test::HasFailure(); } -template +template int DFT_Test::test_in_place_USM() { if (!init(MemoryAccessModel::usm)) { return test_skipped; @@ -126,7 +126,7 @@ int DFT_Test::test_in_place_USM() { auto modified_strides_fwd = this->strides_fwd; auto modified_strides_bwd = this->strides_bwd; - if (domain == oneapi::mkl::dft::domain::REAL) { + if (domain == oneapi::math::dft::domain::REAL) { // both input and output strides must be set auto default_conjuate_strides = get_conjugate_even_complex_strides(sizes); std::ptrdiff_t rank = static_cast(sizes.size()); @@ -152,23 +152,23 @@ int DFT_Test::test_in_place_USM() { auto ref_distance = std::accumulate(sizes.begin(), sizes.end(), 1, std::multiplies<>()); descriptor_t descriptor = { sizes }; - descriptor.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::INPLACE); - if constexpr (domain == oneapi::mkl::dft::domain::REAL) { - descriptor.set_value(oneapi::mkl::dft::config_param::CONJUGATE_EVEN_STORAGE, - oneapi::mkl::dft::config_value::COMPLEX_COMPLEX); - descriptor.set_value(oneapi::mkl::dft::config_param::PACKED_FORMAT, - oneapi::mkl::dft::config_value::CCE_FORMAT); - } - descriptor.set_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, batches); - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_DISTANCE, forward_distance); - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_DISTANCE, backward_distance); + descriptor.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::INPLACE); + if constexpr (domain == oneapi::math::dft::domain::REAL) { + descriptor.set_value(oneapi::math::dft::config_param::CONJUGATE_EVEN_STORAGE, + oneapi::math::dft::config_value::COMPLEX_COMPLEX); + descriptor.set_value(oneapi::math::dft::config_param::PACKED_FORMAT, + oneapi::math::dft::config_value::CCE_FORMAT); + } + descriptor.set_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, batches); + descriptor.set_value(oneapi::math::dft::config_param::FWD_DISTANCE, forward_distance); + descriptor.set_value(oneapi::math::dft::config_param::BWD_DISTANCE, backward_distance); if (modified_strides_fwd.size()) { - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_STRIDES, + descriptor.set_value(oneapi::math::dft::config_param::FWD_STRIDES, modified_strides_fwd.data()); } if (modified_strides_bwd.size()) { - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_STRIDES, + descriptor.set_value(oneapi::math::dft::config_param::BWD_STRIDES, modified_strides_bwd.data()); } commit_descriptor(descriptor, sycl_queue); @@ -177,25 +177,25 @@ int DFT_Test::test_in_place_USM() { std::vector inout( strided_copy(input, sizes, modified_strides_fwd, batches, forward_distance, ua_input), ua_input); - int real_multiplier = (domain == oneapi::mkl::dft::domain::REAL ? 2 : 1); + int real_multiplier = (domain == oneapi::math::dft::domain::REAL ? 2 : 1); inout.resize( cast_unsigned(std::max(forward_distance, real_multiplier * backward_distance) * batches + real_multiplier * get_default(modified_strides_bwd, 0, 0L))); std::vector no_dependencies; - oneapi::mkl::dft::compute_forward(descriptor, inout.data(), + oneapi::math::dft::compute_forward(descriptor, inout.data(), no_dependencies) .wait_and_throw(); for (std::int64_t i = 0; i < batches; i++) { - EXPECT_TRUE(check_equal_strided( + EXPECT_TRUE(check_equal_strided( reinterpret_cast(inout.data()) + backward_distance * i, out_host_ref.data() + ref_distance * i, sizes, modified_strides_bwd, abs_error_margin, rel_error_margin, std::cout)); } sycl::event done = - oneapi::mkl::dft::compute_backward, + oneapi::math::dft::compute_backward, FwdInputType>(descriptor, inout.data(), no_dependencies); done.wait_and_throw(); diff --git a/tests/unit_tests/dft/include/compute_inplace_real_real.hpp b/tests/unit_tests/dft/include/compute_inplace_real_real.hpp index 410440a02..0d9d726ee 100644 --- a/tests/unit_tests/dft/include/compute_inplace_real_real.hpp +++ b/tests/unit_tests/dft/include/compute_inplace_real_real.hpp @@ -22,12 +22,12 @@ #include "compute_tester.hpp" -template +template int DFT_Test::test_in_place_real_real_USM() { if (!init(MemoryAccessModel::usm)) { return test_skipped; } - if constexpr (domain == oneapi::mkl::dft::domain::REAL) { + if constexpr (domain == oneapi::math::dft::domain::REAL) { std::cout << "skipping real split tests as they are not supported" << std::endl; return test_skipped; @@ -35,14 +35,14 @@ int DFT_Test::test_in_place_real_real_USM() { else { descriptor_t descriptor{ sizes }; PrecisionType backward_scale = 1.f / static_cast(forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::INPLACE); - descriptor.set_value(oneapi::mkl::dft::config_param::COMPLEX_STORAGE, - oneapi::mkl::dft::config_value::REAL_REAL); - descriptor.set_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, batches); - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_DISTANCE, forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_DISTANCE, forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::BACKWARD_SCALE, backward_scale); + descriptor.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::INPLACE); + descriptor.set_value(oneapi::math::dft::config_param::COMPLEX_STORAGE, + oneapi::math::dft::config_value::REAL_REAL); + descriptor.set_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, batches); + descriptor.set_value(oneapi::math::dft::config_param::FWD_DISTANCE, forward_elements); + descriptor.set_value(oneapi::math::dft::config_param::BWD_DISTANCE, forward_elements); + descriptor.set_value(oneapi::math::dft::config_param::BACKWARD_SCALE, backward_scale); commit_descriptor(descriptor, sycl_queue); @@ -54,7 +54,7 @@ int DFT_Test::test_in_place_real_real_USM() { std::copy(input_im.begin(), input_im.end(), inout_im.begin()); std::vector no_dependencies; - oneapi::mkl::dft::compute_forward( + oneapi::math::dft::compute_forward( descriptor, inout_re.data(), inout_im.data(), no_dependencies) .wait_and_throw(); @@ -65,7 +65,7 @@ int DFT_Test::test_in_place_real_real_USM() { EXPECT_TRUE(check_equal_vector(output_data.data(), out_host_ref.data(), output_data.size(), abs_error_margin, rel_error_margin, std::cout)); - oneapi::mkl::dft::compute_backward, + oneapi::math::dft::compute_backward, PrecisionType>(descriptor, inout_re.data(), inout_im.data(), no_dependencies) .wait_and_throw(); @@ -81,13 +81,13 @@ int DFT_Test::test_in_place_real_real_USM() { } } -template +template int DFT_Test::test_in_place_real_real_buffer() { if (!init(MemoryAccessModel::buffer)) { return test_skipped; } - if constexpr (domain == oneapi::mkl::dft::domain::REAL) { + if constexpr (domain == oneapi::math::dft::domain::REAL) { std::cout << "skipping real split tests as they are not supported" << std::endl; return test_skipped; @@ -96,14 +96,14 @@ int DFT_Test::test_in_place_real_real_buffer() { descriptor_t descriptor{ sizes }; PrecisionType backward_scale = 1.f / static_cast(forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::INPLACE); - descriptor.set_value(oneapi::mkl::dft::config_param::COMPLEX_STORAGE, - oneapi::mkl::dft::config_value::REAL_REAL); - descriptor.set_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, batches); - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_DISTANCE, forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_DISTANCE, forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::BACKWARD_SCALE, backward_scale); + descriptor.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::INPLACE); + descriptor.set_value(oneapi::math::dft::config_param::COMPLEX_STORAGE, + oneapi::math::dft::config_value::REAL_REAL); + descriptor.set_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, batches); + descriptor.set_value(oneapi::math::dft::config_param::FWD_DISTANCE, forward_elements); + descriptor.set_value(oneapi::math::dft::config_param::BWD_DISTANCE, forward_elements); + descriptor.set_value(oneapi::math::dft::config_param::BACKWARD_SCALE, backward_scale); commit_descriptor(descriptor, sycl_queue); @@ -117,7 +117,7 @@ int DFT_Test::test_in_place_real_real_buffer() { sycl::buffer inout_im_buf{ host_inout_im.data(), sycl::range<1>(size_total) }; - oneapi::mkl::dft::compute_forward(descriptor, inout_re_buf, + oneapi::math::dft::compute_forward(descriptor, inout_re_buf, inout_im_buf); { @@ -132,7 +132,7 @@ int DFT_Test::test_in_place_real_real_buffer() { std::cout)); } - oneapi::mkl::dft::compute_backward, + oneapi::math::dft::compute_backward, PrecisionType>(descriptor, inout_re_buf, inout_im_buf); { diff --git a/tests/unit_tests/dft/include/compute_out_of_place.hpp b/tests/unit_tests/dft/include/compute_out_of_place.hpp index c99da10e1..30b5cb5e1 100644 --- a/tests/unit_tests/dft/include/compute_out_of_place.hpp +++ b/tests/unit_tests/dft/include/compute_out_of_place.hpp @@ -23,7 +23,7 @@ #include "compute_tester.hpp" #include -template +template int DFT_Test::test_out_of_place_buffer() { if (!init(MemoryAccessModel::buffer)) { return test_skipped; @@ -33,33 +33,33 @@ int DFT_Test::test_out_of_place_buffer() { auto strides_fwd_cpy = strides_fwd; auto strides_bwd_cpy = strides_bwd; if (strides_fwd_cpy.size()) { - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_STRIDES, strides_fwd_cpy.data()); + descriptor.set_value(oneapi::math::dft::config_param::FWD_STRIDES, strides_fwd_cpy.data()); } else { strides_fwd_cpy.resize(sizes.size() + 1); - descriptor.get_value(oneapi::mkl::dft::config_param::FWD_STRIDES, strides_fwd_cpy.data()); + descriptor.get_value(oneapi::math::dft::config_param::FWD_STRIDES, strides_fwd_cpy.data()); } if (strides_bwd_cpy.size()) { - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_STRIDES, strides_bwd_cpy.data()); + descriptor.set_value(oneapi::math::dft::config_param::BWD_STRIDES, strides_bwd_cpy.data()); } else { strides_bwd_cpy.resize(sizes.size() + 1); - descriptor.get_value(oneapi::mkl::dft::config_param::BWD_STRIDES, strides_bwd_cpy.data()); + descriptor.get_value(oneapi::math::dft::config_param::BWD_STRIDES, strides_bwd_cpy.data()); } auto [forward_distance, backward_distance] = get_default_distances(sizes, strides_fwd_cpy, strides_bwd_cpy); auto ref_distance = std::accumulate(sizes.begin(), sizes.end(), 1, std::multiplies<>()); - descriptor.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::NOT_INPLACE); - if constexpr (domain == oneapi::mkl::dft::domain::REAL) { - descriptor.set_value(oneapi::mkl::dft::config_param::CONJUGATE_EVEN_STORAGE, - oneapi::mkl::dft::config_value::COMPLEX_COMPLEX); - descriptor.set_value(oneapi::mkl::dft::config_param::PACKED_FORMAT, - oneapi::mkl::dft::config_value::CCE_FORMAT); - } - descriptor.set_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, batches); - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_DISTANCE, forward_distance); - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_DISTANCE, backward_distance); + descriptor.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::NOT_INPLACE); + if constexpr (domain == oneapi::math::dft::domain::REAL) { + descriptor.set_value(oneapi::math::dft::config_param::CONJUGATE_EVEN_STORAGE, + oneapi::math::dft::config_value::COMPLEX_COMPLEX); + descriptor.set_value(oneapi::math::dft::config_param::PACKED_FORMAT, + oneapi::math::dft::config_value::CCE_FORMAT); + } + descriptor.set_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, batches); + descriptor.set_value(oneapi::math::dft::config_param::FWD_DISTANCE, forward_distance); + descriptor.set_value(oneapi::math::dft::config_param::BWD_DISTANCE, backward_distance); commit_descriptor(descriptor, sycl_queue); std::vector fwd_data( strided_copy(input, sizes, strides_fwd_cpy, batches, forward_distance)); @@ -70,20 +70,20 @@ int DFT_Test::test_out_of_place_buffer() { sycl::buffer fwd_buf{ fwd_data }; sycl::buffer bwd_buf{ tmp }; - oneapi::mkl::dft::compute_forward( + oneapi::math::dft::compute_forward( descriptor, fwd_buf, bwd_buf); { auto acc_bwd = bwd_buf.get_host_access(); auto bwd_ptr = acc_bwd.get_pointer(); for (std::int64_t i = 0; i < batches; i++) { - EXPECT_TRUE(check_equal_strided( + EXPECT_TRUE(check_equal_strided( bwd_ptr + backward_distance * i, out_host_ref.data() + ref_distance * i, sizes, strides_bwd_cpy, abs_error_margin, rel_error_margin, std::cout)); } } - oneapi::mkl::dft::compute_backward, + oneapi::math::dft::compute_backward, FwdOutputType, FwdInputType>(descriptor, bwd_buf, fwd_buf); } @@ -101,7 +101,7 @@ int DFT_Test::test_out_of_place_buffer() { return !::testing::Test::HasFailure(); } -template +template int DFT_Test::test_out_of_place_USM() { if (!init(MemoryAccessModel::usm)) { return test_skipped; @@ -112,41 +112,41 @@ int DFT_Test::test_out_of_place_USM() { auto strides_fwd_cpy = strides_fwd; auto strides_bwd_cpy = strides_bwd; if (strides_fwd_cpy.size()) { - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_STRIDES, strides_fwd_cpy.data()); + descriptor.set_value(oneapi::math::dft::config_param::FWD_STRIDES, strides_fwd_cpy.data()); } else { strides_fwd_cpy.resize(sizes.size() + 1); - descriptor.get_value(oneapi::mkl::dft::config_param::FWD_STRIDES, strides_fwd_cpy.data()); + descriptor.get_value(oneapi::math::dft::config_param::FWD_STRIDES, strides_fwd_cpy.data()); } if (strides_bwd_cpy.size()) { - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_STRIDES, strides_bwd_cpy.data()); + descriptor.set_value(oneapi::math::dft::config_param::BWD_STRIDES, strides_bwd_cpy.data()); } else { strides_bwd_cpy.resize(sizes.size() + 1); - descriptor.get_value(oneapi::mkl::dft::config_param::BWD_STRIDES, strides_bwd_cpy.data()); + descriptor.get_value(oneapi::math::dft::config_param::BWD_STRIDES, strides_bwd_cpy.data()); } auto [forward_distance, backward_distance] = get_default_distances(sizes, strides_fwd_cpy, strides_bwd_cpy); auto ref_distance = std::accumulate(sizes.begin(), sizes.end(), 1, std::multiplies<>()); - descriptor.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::NOT_INPLACE); - if constexpr (domain == oneapi::mkl::dft::domain::REAL) { - descriptor.set_value(oneapi::mkl::dft::config_param::CONJUGATE_EVEN_STORAGE, - oneapi::mkl::dft::config_value::COMPLEX_COMPLEX); - descriptor.set_value(oneapi::mkl::dft::config_param::PACKED_FORMAT, - oneapi::mkl::dft::config_value::CCE_FORMAT); - } - descriptor.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::NOT_INPLACE); - if constexpr (domain == oneapi::mkl::dft::domain::REAL) { - descriptor.set_value(oneapi::mkl::dft::config_param::CONJUGATE_EVEN_STORAGE, - oneapi::mkl::dft::config_value::COMPLEX_COMPLEX); - descriptor.set_value(oneapi::mkl::dft::config_param::PACKED_FORMAT, - oneapi::mkl::dft::config_value::CCE_FORMAT); - } - descriptor.set_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, batches); - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_DISTANCE, forward_distance); - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_DISTANCE, backward_distance); + descriptor.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::NOT_INPLACE); + if constexpr (domain == oneapi::math::dft::domain::REAL) { + descriptor.set_value(oneapi::math::dft::config_param::CONJUGATE_EVEN_STORAGE, + oneapi::math::dft::config_value::COMPLEX_COMPLEX); + descriptor.set_value(oneapi::math::dft::config_param::PACKED_FORMAT, + oneapi::math::dft::config_value::CCE_FORMAT); + } + descriptor.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::NOT_INPLACE); + if constexpr (domain == oneapi::math::dft::domain::REAL) { + descriptor.set_value(oneapi::math::dft::config_param::CONJUGATE_EVEN_STORAGE, + oneapi::math::dft::config_value::COMPLEX_COMPLEX); + descriptor.set_value(oneapi::math::dft::config_param::PACKED_FORMAT, + oneapi::math::dft::config_value::CCE_FORMAT); + } + descriptor.set_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, batches); + descriptor.set_value(oneapi::math::dft::config_param::FWD_DISTANCE, forward_distance); + descriptor.set_value(oneapi::math::dft::config_param::BWD_DISTANCE, backward_distance); commit_descriptor(descriptor, sycl_queue); auto ua_input = usm_allocator_t(cxt, *dev); @@ -158,18 +158,18 @@ int DFT_Test::test_out_of_place_USM() { cast_unsigned(backward_distance * batches + get_default(strides_bwd_cpy, 0, 0L)), ua_output); - oneapi::mkl::dft::compute_forward( + oneapi::math::dft::compute_forward( descriptor, fwd.data(), bwd.data(), no_dependencies) .wait_and_throw(); auto bwd_ptr = &bwd[0]; for (std::int64_t i = 0; i < batches; i++) { - EXPECT_TRUE(check_equal_strided( + EXPECT_TRUE(check_equal_strided( bwd_ptr + backward_distance * i, out_host_ref.data() + ref_distance * i, sizes, strides_bwd_cpy, abs_error_margin, rel_error_margin, std::cout)); } - oneapi::mkl::dft::compute_backward, FwdOutputType, + oneapi::math::dft::compute_backward, FwdOutputType, FwdInputType>(descriptor, bwd.data(), fwd.data(), no_dependencies) .wait_and_throw(); diff --git a/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp b/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp index 6ca8da1aa..b1af7e560 100644 --- a/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp +++ b/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp @@ -22,13 +22,13 @@ #include "compute_tester.hpp" -template +template int DFT_Test::test_out_of_place_real_real_USM() { if (!init(MemoryAccessModel::usm)) { return test_skipped; } - if constexpr (domain == oneapi::mkl::dft::domain::REAL) { + if constexpr (domain == oneapi::math::dft::domain::REAL) { std::cout << "skipping real split tests as they are not supported" << std::endl; return test_skipped; @@ -37,14 +37,14 @@ int DFT_Test::test_out_of_place_real_real_USM() { descriptor_t descriptor{ sizes }; PrecisionType backward_scale = 1.f / static_cast(forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::NOT_INPLACE); - descriptor.set_value(oneapi::mkl::dft::config_param::COMPLEX_STORAGE, - oneapi::mkl::dft::config_value::REAL_REAL); - descriptor.set_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, batches); - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_DISTANCE, forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_DISTANCE, forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::BACKWARD_SCALE, backward_scale); + descriptor.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::NOT_INPLACE); + descriptor.set_value(oneapi::math::dft::config_param::COMPLEX_STORAGE, + oneapi::math::dft::config_value::REAL_REAL); + descriptor.set_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, batches); + descriptor.set_value(oneapi::math::dft::config_param::FWD_DISTANCE, forward_elements); + descriptor.set_value(oneapi::math::dft::config_param::BWD_DISTANCE, forward_elements); + descriptor.set_value(oneapi::math::dft::config_param::BACKWARD_SCALE, backward_scale); commit_descriptor(descriptor, sycl_queue); @@ -63,7 +63,7 @@ int DFT_Test::test_out_of_place_real_real_USM() { std::vector no_dependencies; - oneapi::mkl::dft::compute_forward( + oneapi::math::dft::compute_forward( descriptor, in_re.data(), in_im.data(), out_re.data(), out_im.data(), no_dependencies) .wait_and_throw(); std::vector output_data(size_total); @@ -73,7 +73,7 @@ int DFT_Test::test_out_of_place_real_real_USM() { EXPECT_TRUE(check_equal_vector(output_data.data(), out_host_ref.data(), output_data.size(), abs_error_margin, rel_error_margin, std::cout)); - oneapi::mkl::dft::compute_backward, + oneapi::math::dft::compute_backward, PrecisionType, PrecisionType>( descriptor, out_re.data(), out_im.data(), out_back_re.data(), out_back_im.data(), no_dependencies) @@ -90,13 +90,13 @@ int DFT_Test::test_out_of_place_real_real_USM() { return !::testing::Test::HasFailure(); } -template +template int DFT_Test::test_out_of_place_real_real_buffer() { if (!init(MemoryAccessModel::buffer)) { return test_skipped; } - if constexpr (domain == oneapi::mkl::dft::domain::REAL) { + if constexpr (domain == oneapi::math::dft::domain::REAL) { std::cout << "skipping real split tests as they are not supported" << std::endl; return test_skipped; @@ -105,14 +105,14 @@ int DFT_Test::test_out_of_place_real_real_buffer() { descriptor_t descriptor{ sizes }; PrecisionType backward_scale = 1.f / static_cast(forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::NOT_INPLACE); - descriptor.set_value(oneapi::mkl::dft::config_param::COMPLEX_STORAGE, - oneapi::mkl::dft::config_value::REAL_REAL); - descriptor.set_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, batches); - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_DISTANCE, forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_DISTANCE, forward_elements); - descriptor.set_value(oneapi::mkl::dft::config_param::BACKWARD_SCALE, backward_scale); + descriptor.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::NOT_INPLACE); + descriptor.set_value(oneapi::math::dft::config_param::COMPLEX_STORAGE, + oneapi::math::dft::config_value::REAL_REAL); + descriptor.set_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, batches); + descriptor.set_value(oneapi::math::dft::config_param::FWD_DISTANCE, forward_elements); + descriptor.set_value(oneapi::math::dft::config_param::BWD_DISTANCE, forward_elements); + descriptor.set_value(oneapi::math::dft::config_param::BACKWARD_SCALE, backward_scale); commit_descriptor(descriptor, sycl_queue); @@ -123,7 +123,7 @@ int DFT_Test::test_out_of_place_real_real_buffer() { sycl::buffer out_back_dev_re{ sycl::range<1>(size_total) }; sycl::buffer out_back_dev_im{ sycl::range<1>(size_total) }; - oneapi::mkl::dft::compute_forward( + oneapi::math::dft::compute_forward( descriptor, in_dev_re, in_dev_im, out_dev_re, out_dev_im); { @@ -138,7 +138,7 @@ int DFT_Test::test_out_of_place_real_real_buffer() { std::cout)); } - oneapi::mkl::dft::compute_backward, + oneapi::math::dft::compute_backward, PrecisionType, PrecisionType>( descriptor, out_dev_re, out_dev_im, out_back_dev_re, out_back_dev_im); diff --git a/tests/unit_tests/dft/include/compute_tester.hpp b/tests/unit_tests/dft/include/compute_tester.hpp index 8b0737a9e..96ad1ffa3 100644 --- a/tests/unit_tests/dft/include/compute_tester.hpp +++ b/tests/unit_tests/dft/include/compute_tester.hpp @@ -34,18 +34,18 @@ #include -template +template struct DFT_Test { - using descriptor_t = oneapi::mkl::dft::descriptor; + using descriptor_t = oneapi::math::dft::descriptor; template using usm_allocator_t = sycl::usm_allocator; using PrecisionType = - typename std::conditional_t; - using FwdInputType = typename std::conditional_t>; using FwdOutputType = std::complex; @@ -93,7 +93,7 @@ struct DFT_Test { out_host_ref = std::vector(size_total); rand_vector(input, size_total); - if constexpr (domain == oneapi::mkl::dft::domain::REAL) { + if constexpr (domain == oneapi::math::dft::domain::REAL) { for (std::size_t i = 0; i < input.size(); ++i) { input_re[i] = { input[i] }; input_im[i] = 0; @@ -108,7 +108,7 @@ struct DFT_Test { } bool skip_test(MemoryAccessModel mem_acc) { - if constexpr (precision == oneapi::mkl::dft::precision::DOUBLE) { + if constexpr (precision == oneapi::math::dft::precision::DOUBLE) { if (!sycl_queue.get_device().has(sycl::aspect::fp64)) { std::cout << "Device does not support double precision." << std::endl; return true; diff --git a/tests/unit_tests/dft/include/reference_dft.hpp b/tests/unit_tests/dft/include/reference_dft.hpp index 448fc3363..661829282 100644 --- a/tests/unit_tests/dft/include/reference_dft.hpp +++ b/tests/unit_tests/dft/include/reference_dft.hpp @@ -121,7 +121,7 @@ void reference_forward_dft(const std::vector &sizes, const TypeIn case 2: detail::reference::forward_dft(unsigned_sizes, in, out); break; case 3: detail::reference::forward_dft(unsigned_sizes, in, out); break; default: - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "reference_dft", "forward_dft", "dft with size " + std::to_string(unsigned_sizes.size())); } diff --git a/tests/unit_tests/dft/include/test_common.hpp b/tests/unit_tests/dft/include/test_common.hpp index 5b74d80c9..70f60d363 100644 --- a/tests/unit_tests/dft/include/test_common.hpp +++ b/tests/unit_tests/dft/include/test_common.hpp @@ -152,8 +152,8 @@ auto exception_handler = [](sycl::exception_list exceptions) { } }; -template -void commit_descriptor(oneapi::mkl::dft::descriptor &descriptor, +template +void commit_descriptor(oneapi::math::dft::descriptor &descriptor, sycl::queue queue) { #ifdef CALL_RT_API descriptor.commit(queue); @@ -170,7 +170,7 @@ inline std::array get_conjugate_even_complex_strides( case 2: return { 0, sizes[1] / 2 + 1, 1 }; case 3: return { 0, sizes[1] * (sizes[2] / 2 + 1), (sizes[2] / 2 + 1), 1 }; default: - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "dft/test_common", __FUNCTION__, "not implemented for " + std::to_string(sizes.size()) + " dimensions"); return {}; @@ -180,7 +180,7 @@ inline std::array get_conjugate_even_complex_strides( // is it assumed that the unused elements of the array are ignored inline std::array get_default_strides(const std::vector &sizes) { if (sizes.size() > 3) { - throw oneapi::mkl::unimplemented( + throw oneapi::math::unimplemented( "dft/test_common", __FUNCTION__, "not implemented for " + std::to_string(sizes.size()) + " dimensions"); } @@ -190,7 +190,7 @@ inline std::array get_default_strides(const std::vector vec, std::size_t idx, T default_) { return vec[idx]; } -template +template std::pair get_default_distances( const std::vector &sizes, const std::vector &strides_fwd, const std::vector &strides_bwd) { @@ -213,11 +213,11 @@ std::pair get_default_distances( std::int64_t size1 = get_default(sizes, 1, 1l); std::int64_t size2 = get_default(sizes, 2, 1l); std::int64_t size0_real = - domain == oneapi::mkl::dft::domain::REAL && sizes.size() == 1 ? size0 / 2 + 1 : size0; + domain == oneapi::math::dft::domain::REAL && sizes.size() == 1 ? size0 / 2 + 1 : size0; std::int64_t size1_real = - domain == oneapi::mkl::dft::domain::REAL && sizes.size() == 2 ? size1 / 2 + 1 : size1; + domain == oneapi::math::dft::domain::REAL && sizes.size() == 2 ? size1 / 2 + 1 : size1; std::int64_t size2_real = - domain == oneapi::mkl::dft::domain::REAL && sizes.size() == 3 ? size2 / 2 + 1 : size2; + domain == oneapi::math::dft::domain::REAL && sizes.size() == 3 ? size2 / 2 + 1 : size2; std::int64_t backward_distance = size0_real * size1_real * size2_real; std::int64_t forward_distance = size0 * size1 * size2; if (strides_fwd.size() > 1) { @@ -233,7 +233,7 @@ std::pair get_default_distances( if (in_place) { forward_distance = std::max(forward_distance, - backward_distance * (domain == oneapi::mkl::dft::domain::REAL ? 2L : 1L)); + backward_distance * (domain == oneapi::math::dft::domain::REAL ? 2L : 1L)); } return { forward_distance, backward_distance }; } diff --git a/tests/unit_tests/dft/source/compute_tests.cpp b/tests/unit_tests/dft/source/compute_tests.cpp index 7f737ef50..bea23c0d9 100644 --- a/tests/unit_tests/dft/source/compute_tests.cpp +++ b/tests/unit_tests/dft/source/compute_tests.cpp @@ -61,15 +61,15 @@ class ComputeTests_real_real_out_of_place_REAL TEST_P(ComputeTests##_##LAYOUT##PLACE##_##DOMAIN, \ DOMAIN##_##PRECISION##_##PLACE##_##LAYOUT##STORAGE) { \ try { \ - auto test = DFT_Test{ \ + auto test = DFT_Test{ \ std::get<0>(GetParam()), std::get<1>(GetParam()).sizes, \ std::get<1>(GetParam()).strides_fwd, std::get<1>(GetParam()).strides_bwd, \ std::get<1>(GetParam()).batches \ }; \ EXPECT_TRUEORSKIP(test.test_##PLACE##_##LAYOUT##STORAGE()); \ } \ - catch (oneapi::mkl::unimplemented & e) { \ + catch (oneapi::math::unimplemented & e) { \ std::cout << "Skipping test because: \"" << e.what() << "\"" << std::endl; \ GTEST_SKIP(); \ } \ diff --git a/tests/unit_tests/dft/source/descriptor_tests.cpp b/tests/unit_tests/dft/source/descriptor_tests.cpp index 155160149..da1f5c487 100644 --- a/tests/unit_tests/dft/source/descriptor_tests.cpp +++ b/tests/unit_tests/dft/source/descriptor_tests.cpp @@ -41,33 +41,33 @@ namespace { constexpr std::int64_t default_1d_lengths = 4; const std::vector default_3d_lengths{ 124, 5, 3 }; -template +template static void set_and_get_lengths() { /* Negative Testing */ { - oneapi::mkl::dft::descriptor descriptor{ default_3d_lengths }; - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::LENGTHS, nullptr), - oneapi::mkl::invalid_argument); + oneapi::math::dft::descriptor descriptor{ default_3d_lengths }; + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::LENGTHS, nullptr), + oneapi::math::invalid_argument); } /* 1D */ { const std::int64_t dimensions = 1; - oneapi::mkl::dft::descriptor descriptor{ default_1d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_1d_lengths }; const std::int64_t new_lengths{ 2345 }; std::int64_t lengths_value{ 0 }; std::int64_t dimensions_before_set{ 0 }; std::int64_t dimensions_after_set{ 0 }; - descriptor.get_value(oneapi::mkl::dft::config_param::LENGTHS, &lengths_value); - descriptor.get_value(oneapi::mkl::dft::config_param::DIMENSION, &dimensions_before_set); + descriptor.get_value(oneapi::math::dft::config_param::LENGTHS, &lengths_value); + descriptor.get_value(oneapi::math::dft::config_param::DIMENSION, &dimensions_before_set); EXPECT_EQ(default_1d_lengths, lengths_value); EXPECT_EQ(dimensions, dimensions_before_set); - descriptor.set_value(oneapi::mkl::dft::config_param::LENGTHS, new_lengths); - descriptor.get_value(oneapi::mkl::dft::config_param::LENGTHS, &lengths_value); - descriptor.get_value(oneapi::mkl::dft::config_param::DIMENSION, &dimensions_after_set); + descriptor.set_value(oneapi::math::dft::config_param::LENGTHS, new_lengths); + descriptor.get_value(oneapi::math::dft::config_param::LENGTHS, &lengths_value); + descriptor.get_value(oneapi::math::dft::config_param::DIMENSION, &dimensions_after_set); EXPECT_EQ(new_lengths, lengths_value); EXPECT_EQ(dimensions, dimensions_after_set); } @@ -76,22 +76,22 @@ static void set_and_get_lengths() { { const std::int64_t dimensions = 3; - oneapi::mkl::dft::descriptor descriptor{ default_3d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_3d_lengths }; std::vector lengths_value(3); std::vector new_lengths{ 1, 2, 7 }; std::int64_t dimensions_before_set{ 0 }; std::int64_t dimensions_after_set{ 0 }; - descriptor.get_value(oneapi::mkl::dft::config_param::LENGTHS, lengths_value.data()); - descriptor.get_value(oneapi::mkl::dft::config_param::DIMENSION, &dimensions_before_set); + descriptor.get_value(oneapi::math::dft::config_param::LENGTHS, lengths_value.data()); + descriptor.get_value(oneapi::math::dft::config_param::DIMENSION, &dimensions_before_set); EXPECT_EQ(default_3d_lengths, lengths_value); EXPECT_EQ(dimensions, dimensions_before_set); - descriptor.set_value(oneapi::mkl::dft::config_param::LENGTHS, new_lengths.data()); - descriptor.get_value(oneapi::mkl::dft::config_param::LENGTHS, lengths_value.data()); - descriptor.get_value(oneapi::mkl::dft::config_param::DIMENSION, &dimensions_after_set); + descriptor.set_value(oneapi::math::dft::config_param::LENGTHS, new_lengths.data()); + descriptor.get_value(oneapi::math::dft::config_param::LENGTHS, lengths_value.data()); + descriptor.get_value(oneapi::math::dft::config_param::DIMENSION, &dimensions_after_set); EXPECT_EQ(new_lengths, lengths_value); EXPECT_EQ(dimensions, dimensions_after_set); @@ -101,14 +101,14 @@ static void set_and_get_lengths() { // Test for deprecated functionality #pragma clang diagnostic push #pragma clang diagnostic ignored "-Wdeprecated-declarations" -template +template static void set_and_get_io_strides() { - oneapi::mkl::dft::descriptor descriptor{ default_3d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_3d_lengths }; - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::INPUT_STRIDES, nullptr), - oneapi::mkl::invalid_argument); - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::OUTPUT_STRIDES, nullptr), - oneapi::mkl::invalid_argument); + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::INPUT_STRIDES, nullptr), + oneapi::math::invalid_argument); + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::OUTPUT_STRIDES, nullptr), + oneapi::math::invalid_argument); constexpr std::int64_t strides_size = 4; const std::int64_t default_stride_d1 = default_3d_lengths[2] * default_3d_lengths[1]; @@ -120,7 +120,7 @@ static void set_and_get_io_strides() { std::vector input_strides_value; std::vector output_strides_value; - if constexpr (domain == oneapi::mkl::dft::domain::COMPLEX) { + if constexpr (domain == oneapi::math::dft::domain::COMPLEX) { input_strides_value = { 50, default_stride_d1 * 2, default_stride_d2 * 2, default_stride_d3 * 2 }; output_strides_value = { 50, default_stride_d1 * 2, default_stride_d2 * 2, @@ -138,38 +138,38 @@ static void set_and_get_io_strides() { std::vector fwd_strides_after_set(strides_size, -1); std::vector bwd_strides_after_set(strides_size, -1); - descriptor.get_value(oneapi::mkl::dft::config_param::INPUT_STRIDES, + descriptor.get_value(oneapi::math::dft::config_param::INPUT_STRIDES, input_strides_before_set.data()); EXPECT_EQ(std::vector(strides_size, 0), input_strides_before_set); - descriptor.set_value(oneapi::mkl::dft::config_param::INPUT_STRIDES, input_strides_value.data()); - descriptor.get_value(oneapi::mkl::dft::config_param::INPUT_STRIDES, + descriptor.set_value(oneapi::math::dft::config_param::INPUT_STRIDES, input_strides_value.data()); + descriptor.get_value(oneapi::math::dft::config_param::INPUT_STRIDES, input_strides_after_set.data()); - descriptor.get_value(oneapi::mkl::dft::config_param::FWD_STRIDES, fwd_strides_after_set.data()); - descriptor.get_value(oneapi::mkl::dft::config_param::BWD_STRIDES, bwd_strides_after_set.data()); + descriptor.get_value(oneapi::math::dft::config_param::FWD_STRIDES, fwd_strides_after_set.data()); + descriptor.get_value(oneapi::math::dft::config_param::BWD_STRIDES, bwd_strides_after_set.data()); EXPECT_EQ(input_strides_value, input_strides_after_set); EXPECT_EQ(std::vector(strides_size, 0), fwd_strides_after_set); EXPECT_EQ(std::vector(strides_size, 0), bwd_strides_after_set); std::vector output_strides_before_set(strides_size); std::vector output_strides_after_set(strides_size); - descriptor.get_value(oneapi::mkl::dft::config_param::OUTPUT_STRIDES, + descriptor.get_value(oneapi::math::dft::config_param::OUTPUT_STRIDES, output_strides_before_set.data()); EXPECT_EQ(std::vector(strides_size, 0), output_strides_before_set); - descriptor.set_value(oneapi::mkl::dft::config_param::OUTPUT_STRIDES, + descriptor.set_value(oneapi::math::dft::config_param::OUTPUT_STRIDES, output_strides_value.data()); - descriptor.get_value(oneapi::mkl::dft::config_param::OUTPUT_STRIDES, + descriptor.get_value(oneapi::math::dft::config_param::OUTPUT_STRIDES, output_strides_after_set.data()); EXPECT_EQ(output_strides_value, output_strides_after_set); } -template +template static void set_and_get_fwd_bwd_strides() { - oneapi::mkl::dft::descriptor descriptor{ default_3d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_3d_lengths }; - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::FWD_STRIDES, nullptr), - oneapi::mkl::invalid_argument); - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::BWD_STRIDES, nullptr), - oneapi::mkl::invalid_argument); + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::FWD_STRIDES, nullptr), + oneapi::math::invalid_argument); + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::BWD_STRIDES, nullptr), + oneapi::math::invalid_argument); constexpr std::int64_t strides_size = 4; const std::int64_t default_stride_d1 = default_3d_lengths[2] * default_3d_lengths[1]; @@ -178,7 +178,7 @@ static void set_and_get_fwd_bwd_strides() { std::vector fwd_strides_default_value; std::vector bwd_strides_default_value; - if constexpr (domain == oneapi::mkl::dft::domain::COMPLEX) { + if constexpr (domain == oneapi::math::dft::domain::COMPLEX) { fwd_strides_default_value = { 0, default_stride_d1, default_stride_d2, default_stride_d3 }; bwd_strides_default_value = { 0, default_stride_d1, default_stride_d2, default_stride_d3 }; } @@ -203,14 +203,14 @@ static void set_and_get_fwd_bwd_strides() { std::vector input_strides_after_set(strides_size, -1); std::vector output_strides_after_set(strides_size, -1); - descriptor.get_value(oneapi::mkl::dft::config_param::FWD_STRIDES, + descriptor.get_value(oneapi::math::dft::config_param::FWD_STRIDES, fwd_strides_before_set.data()); EXPECT_EQ(fwd_strides_default_value, fwd_strides_before_set); - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_STRIDES, fwd_strides_new_value.data()); - descriptor.get_value(oneapi::mkl::dft::config_param::FWD_STRIDES, fwd_strides_after_set.data()); - descriptor.get_value(oneapi::mkl::dft::config_param::INPUT_STRIDES, + descriptor.set_value(oneapi::math::dft::config_param::FWD_STRIDES, fwd_strides_new_value.data()); + descriptor.get_value(oneapi::math::dft::config_param::FWD_STRIDES, fwd_strides_after_set.data()); + descriptor.get_value(oneapi::math::dft::config_param::INPUT_STRIDES, input_strides_after_set.data()); - descriptor.get_value(oneapi::mkl::dft::config_param::OUTPUT_STRIDES, + descriptor.get_value(oneapi::math::dft::config_param::OUTPUT_STRIDES, output_strides_after_set.data()); EXPECT_EQ(fwd_strides_new_value, fwd_strides_after_set); EXPECT_EQ(std::vector(strides_size, 0), input_strides_after_set); @@ -218,21 +218,21 @@ static void set_and_get_fwd_bwd_strides() { std::vector bwd_strides_before_set(strides_size); std::vector bwd_strides_after_set(strides_size); - descriptor.get_value(oneapi::mkl::dft::config_param::BWD_STRIDES, + descriptor.get_value(oneapi::math::dft::config_param::BWD_STRIDES, bwd_strides_before_set.data()); EXPECT_EQ(bwd_strides_default_value, bwd_strides_before_set); - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_STRIDES, bwd_strides_new_value.data()); - descriptor.get_value(oneapi::mkl::dft::config_param::BWD_STRIDES, bwd_strides_after_set.data()); + descriptor.set_value(oneapi::math::dft::config_param::BWD_STRIDES, bwd_strides_new_value.data()); + descriptor.get_value(oneapi::math::dft::config_param::BWD_STRIDES, bwd_strides_after_set.data()); EXPECT_EQ(bwd_strides_new_value, bwd_strides_after_set); } #pragma clang diagnostic pop -template +template static void set_and_get_values() { - oneapi::mkl::dft::descriptor descriptor{ default_1d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_1d_lengths }; using Precision_Type = - typename std::conditional_t; { @@ -240,12 +240,12 @@ static void set_and_get_values() { Precision_Type forward_scale_before_set; Precision_Type forward_scale_after_set; - descriptor.get_value(oneapi::mkl::dft::config_param::FORWARD_SCALE, + descriptor.get_value(oneapi::math::dft::config_param::FORWARD_SCALE, &forward_scale_before_set); EXPECT_EQ(1.0, forward_scale_before_set); - descriptor.set_value(oneapi::mkl::dft::config_param::FORWARD_SCALE, + descriptor.set_value(oneapi::math::dft::config_param::FORWARD_SCALE, forward_scale_set_value); - descriptor.get_value(oneapi::mkl::dft::config_param::FORWARD_SCALE, + descriptor.get_value(oneapi::math::dft::config_param::FORWARD_SCALE, &forward_scale_after_set); EXPECT_EQ(forward_scale_set_value, forward_scale_after_set); } @@ -255,12 +255,12 @@ static void set_and_get_values() { Precision_Type backward_scale_before_set; Precision_Type backward_scale_after_set; - descriptor.get_value(oneapi::mkl::dft::config_param::BACKWARD_SCALE, + descriptor.get_value(oneapi::math::dft::config_param::BACKWARD_SCALE, &backward_scale_before_set); EXPECT_EQ(1.0, backward_scale_before_set); - descriptor.set_value(oneapi::mkl::dft::config_param::BACKWARD_SCALE, + descriptor.set_value(oneapi::math::dft::config_param::BACKWARD_SCALE, backward_scale_set_value); - descriptor.get_value(oneapi::mkl::dft::config_param::BACKWARD_SCALE, + descriptor.get_value(oneapi::math::dft::config_param::BACKWARD_SCALE, &backward_scale_after_set); EXPECT_EQ(backward_scale_set_value, backward_scale_after_set); } @@ -270,12 +270,12 @@ static void set_and_get_values() { std::int64_t n_transforms_before_set; std::int64_t n_transforms_after_set; - descriptor.get_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, + descriptor.get_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, &n_transforms_before_set); EXPECT_EQ(1, n_transforms_before_set); - descriptor.set_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, + descriptor.set_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, n_transforms_set_value); - descriptor.get_value(oneapi::mkl::dft::config_param::NUMBER_OF_TRANSFORMS, + descriptor.get_value(oneapi::math::dft::config_param::NUMBER_OF_TRANSFORMS, &n_transforms_after_set); EXPECT_EQ(n_transforms_set_value, n_transforms_after_set); } @@ -285,236 +285,236 @@ static void set_and_get_values() { std::int64_t fwd_distance_before_set; std::int64_t fwd_distance_after_set; - descriptor.get_value(oneapi::mkl::dft::config_param::FWD_DISTANCE, + descriptor.get_value(oneapi::math::dft::config_param::FWD_DISTANCE, &fwd_distance_before_set); EXPECT_EQ(1, fwd_distance_before_set); - descriptor.set_value(oneapi::mkl::dft::config_param::FWD_DISTANCE, fwd_distance_set_value); - descriptor.get_value(oneapi::mkl::dft::config_param::FWD_DISTANCE, &fwd_distance_after_set); + descriptor.set_value(oneapi::math::dft::config_param::FWD_DISTANCE, fwd_distance_set_value); + descriptor.get_value(oneapi::math::dft::config_param::FWD_DISTANCE, &fwd_distance_after_set); EXPECT_EQ(fwd_distance_set_value, fwd_distance_after_set); - std::int64_t bwd_distance_set_value{ domain == oneapi::mkl::dft::domain::REAL + std::int64_t bwd_distance_set_value{ domain == oneapi::math::dft::domain::REAL ? (fwd_distance_set_value / 2) + 1 : fwd_distance_set_value }; std::int64_t bwd_distance_before_set; std::int64_t bwd_distance_after_set; - descriptor.get_value(oneapi::mkl::dft::config_param::BWD_DISTANCE, + descriptor.get_value(oneapi::math::dft::config_param::BWD_DISTANCE, &bwd_distance_before_set); EXPECT_EQ(1, bwd_distance_before_set); - descriptor.set_value(oneapi::mkl::dft::config_param::BWD_DISTANCE, bwd_distance_set_value); - descriptor.get_value(oneapi::mkl::dft::config_param::BWD_DISTANCE, &bwd_distance_after_set); + descriptor.set_value(oneapi::math::dft::config_param::BWD_DISTANCE, bwd_distance_set_value); + descriptor.get_value(oneapi::math::dft::config_param::BWD_DISTANCE, &bwd_distance_after_set); EXPECT_EQ(bwd_distance_set_value, bwd_distance_after_set); } { - oneapi::mkl::dft::config_value value{ - oneapi::mkl::dft::config_value::COMMITTED + oneapi::math::dft::config_value value{ + oneapi::math::dft::config_value::COMMITTED }; // Initialize with invalid value - descriptor.get_value(oneapi::mkl::dft::config_param::PLACEMENT, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::INPLACE, value); - - descriptor.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::NOT_INPLACE); - descriptor.get_value(oneapi::mkl::dft::config_param::PLACEMENT, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::NOT_INPLACE, value); - - descriptor.set_value(oneapi::mkl::dft::config_param::PLACEMENT, - oneapi::mkl::dft::config_value::INPLACE); - descriptor.get_value(oneapi::mkl::dft::config_param::PLACEMENT, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::INPLACE, value); + descriptor.get_value(oneapi::math::dft::config_param::PLACEMENT, &value); + EXPECT_EQ(oneapi::math::dft::config_value::INPLACE, value); + + descriptor.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::NOT_INPLACE); + descriptor.get_value(oneapi::math::dft::config_param::PLACEMENT, &value); + EXPECT_EQ(oneapi::math::dft::config_value::NOT_INPLACE, value); + + descriptor.set_value(oneapi::math::dft::config_param::PLACEMENT, + oneapi::math::dft::config_value::INPLACE); + descriptor.get_value(oneapi::math::dft::config_param::PLACEMENT, &value); + EXPECT_EQ(oneapi::math::dft::config_value::INPLACE, value); } { - oneapi::mkl::dft::config_value value{ - oneapi::mkl::dft::config_value::COMMITTED + oneapi::math::dft::config_value value{ + oneapi::math::dft::config_value::COMMITTED }; // Initialize with invalid value - descriptor.get_value(oneapi::mkl::dft::config_param::COMPLEX_STORAGE, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::COMPLEX_COMPLEX, value); - - descriptor.set_value(oneapi::mkl::dft::config_param::COMPLEX_STORAGE, - oneapi::mkl::dft::config_value::REAL_REAL); - descriptor.get_value(oneapi::mkl::dft::config_param::COMPLEX_STORAGE, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::REAL_REAL, value); - - descriptor.set_value(oneapi::mkl::dft::config_param::COMPLEX_STORAGE, - oneapi::mkl::dft::config_value::COMPLEX_COMPLEX); - descriptor.get_value(oneapi::mkl::dft::config_param::COMPLEX_STORAGE, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::COMPLEX_COMPLEX, value); + descriptor.get_value(oneapi::math::dft::config_param::COMPLEX_STORAGE, &value); + EXPECT_EQ(oneapi::math::dft::config_value::COMPLEX_COMPLEX, value); + + descriptor.set_value(oneapi::math::dft::config_param::COMPLEX_STORAGE, + oneapi::math::dft::config_value::REAL_REAL); + descriptor.get_value(oneapi::math::dft::config_param::COMPLEX_STORAGE, &value); + EXPECT_EQ(oneapi::math::dft::config_value::REAL_REAL, value); + + descriptor.set_value(oneapi::math::dft::config_param::COMPLEX_STORAGE, + oneapi::math::dft::config_value::COMPLEX_COMPLEX); + descriptor.get_value(oneapi::math::dft::config_param::COMPLEX_STORAGE, &value); + EXPECT_EQ(oneapi::math::dft::config_value::COMPLEX_COMPLEX, value); } { - oneapi::mkl::dft::config_value value{ - oneapi::mkl::dft::config_value::COMMITTED + oneapi::math::dft::config_value value{ + oneapi::math::dft::config_value::COMMITTED }; // Initialize with invalid value - descriptor.get_value(oneapi::mkl::dft::config_param::CONJUGATE_EVEN_STORAGE, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::COMPLEX_COMPLEX, value); + descriptor.get_value(oneapi::math::dft::config_param::CONJUGATE_EVEN_STORAGE, &value); + EXPECT_EQ(oneapi::math::dft::config_value::COMPLEX_COMPLEX, value); - descriptor.set_value(oneapi::mkl::dft::config_param::CONJUGATE_EVEN_STORAGE, - oneapi::mkl::dft::config_value::COMPLEX_COMPLEX); + descriptor.set_value(oneapi::math::dft::config_param::CONJUGATE_EVEN_STORAGE, + oneapi::math::dft::config_value::COMPLEX_COMPLEX); - value = oneapi::mkl::dft::config_value::COMMITTED; // Initialize with invalid value - descriptor.get_value(oneapi::mkl::dft::config_param::CONJUGATE_EVEN_STORAGE, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::COMPLEX_COMPLEX, value); + value = oneapi::math::dft::config_value::COMMITTED; // Initialize with invalid value + descriptor.get_value(oneapi::math::dft::config_param::CONJUGATE_EVEN_STORAGE, &value); + EXPECT_EQ(oneapi::math::dft::config_value::COMPLEX_COMPLEX, value); } { - oneapi::mkl::dft::config_value value{ - oneapi::mkl::dft::config_value::COMMITTED + oneapi::math::dft::config_value value{ + oneapi::math::dft::config_value::COMMITTED }; // Initialize with invalid value - descriptor.get_value(oneapi::mkl::dft::config_param::REAL_STORAGE, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::REAL_REAL, value); + descriptor.get_value(oneapi::math::dft::config_param::REAL_STORAGE, &value); + EXPECT_EQ(oneapi::math::dft::config_value::REAL_REAL, value); - descriptor.set_value(oneapi::mkl::dft::config_param::REAL_STORAGE, - oneapi::mkl::dft::config_value::REAL_REAL); + descriptor.set_value(oneapi::math::dft::config_param::REAL_STORAGE, + oneapi::math::dft::config_value::REAL_REAL); - value = oneapi::mkl::dft::config_value::COMMITTED; // Initialize with invalid value - descriptor.get_value(oneapi::mkl::dft::config_param::REAL_STORAGE, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::REAL_REAL, value); + value = oneapi::math::dft::config_value::COMMITTED; // Initialize with invalid value + descriptor.get_value(oneapi::math::dft::config_param::REAL_STORAGE, &value); + EXPECT_EQ(oneapi::math::dft::config_value::REAL_REAL, value); } { - oneapi::mkl::dft::config_value value{ - oneapi::mkl::dft::config_value::COMMITTED + oneapi::math::dft::config_value value{ + oneapi::math::dft::config_value::COMMITTED }; // Initialize with invalid value - descriptor.get_value(oneapi::mkl::dft::config_param::ORDERING, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::ORDERED, value); - - descriptor.set_value(oneapi::mkl::dft::config_param::ORDERING, - oneapi::mkl::dft::config_value::BACKWARD_SCRAMBLED); - descriptor.get_value(oneapi::mkl::dft::config_param::ORDERING, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::BACKWARD_SCRAMBLED, value); - - descriptor.set_value(oneapi::mkl::dft::config_param::ORDERING, - oneapi::mkl::dft::config_value::ORDERED); - descriptor.get_value(oneapi::mkl::dft::config_param::ORDERING, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::ORDERED, value); + descriptor.get_value(oneapi::math::dft::config_param::ORDERING, &value); + EXPECT_EQ(oneapi::math::dft::config_value::ORDERED, value); + + descriptor.set_value(oneapi::math::dft::config_param::ORDERING, + oneapi::math::dft::config_value::BACKWARD_SCRAMBLED); + descriptor.get_value(oneapi::math::dft::config_param::ORDERING, &value); + EXPECT_EQ(oneapi::math::dft::config_value::BACKWARD_SCRAMBLED, value); + + descriptor.set_value(oneapi::math::dft::config_param::ORDERING, + oneapi::math::dft::config_value::ORDERED); + descriptor.get_value(oneapi::math::dft::config_param::ORDERING, &value); + EXPECT_EQ(oneapi::math::dft::config_value::ORDERED, value); } { bool value = true; - descriptor.get_value(oneapi::mkl::dft::config_param::TRANSPOSE, &value); + descriptor.get_value(oneapi::math::dft::config_param::TRANSPOSE, &value); EXPECT_EQ(false, value); - descriptor.set_value(oneapi::mkl::dft::config_param::TRANSPOSE, true); - descriptor.get_value(oneapi::mkl::dft::config_param::TRANSPOSE, &value); + descriptor.set_value(oneapi::math::dft::config_param::TRANSPOSE, true); + descriptor.get_value(oneapi::math::dft::config_param::TRANSPOSE, &value); EXPECT_EQ(true, value); /* Set value to false again because transpose is not implemented and will fail on commit * when using the MKLGPU backend */ - descriptor.set_value(oneapi::mkl::dft::config_param::TRANSPOSE, false); + descriptor.set_value(oneapi::math::dft::config_param::TRANSPOSE, false); } { /* Only value currently supported for PACKED_FORMAT is the config_value::CCE_FORMAT */ - oneapi::mkl::dft::config_value value{ - oneapi::mkl::dft::config_value::COMMITTED + oneapi::math::dft::config_value value{ + oneapi::math::dft::config_value::COMMITTED }; // Initialize with invalid value - descriptor.get_value(oneapi::mkl::dft::config_param::PACKED_FORMAT, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::CCE_FORMAT, value); + descriptor.get_value(oneapi::math::dft::config_param::PACKED_FORMAT, &value); + EXPECT_EQ(oneapi::math::dft::config_value::CCE_FORMAT, value); - descriptor.set_value(oneapi::mkl::dft::config_param::PACKED_FORMAT, - oneapi::mkl::dft::config_value::CCE_FORMAT); + descriptor.set_value(oneapi::math::dft::config_param::PACKED_FORMAT, + oneapi::math::dft::config_value::CCE_FORMAT); - value = oneapi::mkl::dft::config_value::COMMITTED; // Initialize with invalid value - descriptor.get_value(oneapi::mkl::dft::config_param::PACKED_FORMAT, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::CCE_FORMAT, value); + value = oneapi::math::dft::config_value::COMMITTED; // Initialize with invalid value + descriptor.get_value(oneapi::math::dft::config_param::PACKED_FORMAT, &value); + EXPECT_EQ(oneapi::math::dft::config_value::CCE_FORMAT, value); } { - oneapi::mkl::dft::config_value value{ - oneapi::mkl::dft::config_value::COMMITTED + oneapi::math::dft::config_value value{ + oneapi::math::dft::config_value::COMMITTED }; // Initialize with invalid value - descriptor.get_value(oneapi::mkl::dft::config_param::WORKSPACE_PLACEMENT, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::WORKSPACE_AUTOMATIC, value); + descriptor.get_value(oneapi::math::dft::config_param::WORKSPACE_PLACEMENT, &value); + EXPECT_EQ(oneapi::math::dft::config_value::WORKSPACE_AUTOMATIC, value); - descriptor.set_value(oneapi::mkl::dft::config_param::WORKSPACE_PLACEMENT, - oneapi::mkl::dft::config_value::WORKSPACE_EXTERNAL); + descriptor.set_value(oneapi::math::dft::config_param::WORKSPACE_PLACEMENT, + oneapi::math::dft::config_value::WORKSPACE_EXTERNAL); - value = oneapi::mkl::dft::config_value::COMMITTED; // Initialize with invalid value - descriptor.get_value(oneapi::mkl::dft::config_param::WORKSPACE_PLACEMENT, &value); - EXPECT_EQ(oneapi::mkl::dft::config_value::WORKSPACE_EXTERNAL, value); - descriptor.set_value(oneapi::mkl::dft::config_param::WORKSPACE_PLACEMENT, - oneapi::mkl::dft::config_value::WORKSPACE_AUTOMATIC); + value = oneapi::math::dft::config_value::COMMITTED; // Initialize with invalid value + descriptor.get_value(oneapi::math::dft::config_param::WORKSPACE_PLACEMENT, &value); + EXPECT_EQ(oneapi::math::dft::config_value::WORKSPACE_EXTERNAL, value); + descriptor.set_value(oneapi::math::dft::config_param::WORKSPACE_PLACEMENT, + oneapi::math::dft::config_value::WORKSPACE_AUTOMATIC); } } -template +template static void get_readonly_values() { - oneapi::mkl::dft::descriptor descriptor{ default_1d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_1d_lengths }; - oneapi::mkl::dft::domain domain_value; - descriptor.get_value(oneapi::mkl::dft::config_param::FORWARD_DOMAIN, &domain_value); + oneapi::math::dft::domain domain_value; + descriptor.get_value(oneapi::math::dft::config_param::FORWARD_DOMAIN, &domain_value); EXPECT_EQ(domain_value, domain); - oneapi::mkl::dft::precision precision_value; - descriptor.get_value(oneapi::mkl::dft::config_param::PRECISION, &precision_value); + oneapi::math::dft::precision precision_value; + descriptor.get_value(oneapi::math::dft::config_param::PRECISION, &precision_value); EXPECT_EQ(precision_value, precision); std::int64_t dimension_value; - descriptor.get_value(oneapi::mkl::dft::config_param::DIMENSION, &dimension_value); + descriptor.get_value(oneapi::math::dft::config_param::DIMENSION, &dimension_value); EXPECT_EQ(dimension_value, 1); - oneapi::mkl::dft::descriptor descriptor3D{ default_3d_lengths }; - descriptor3D.get_value(oneapi::mkl::dft::config_param::DIMENSION, &dimension_value); + oneapi::math::dft::descriptor descriptor3D{ default_3d_lengths }; + descriptor3D.get_value(oneapi::math::dft::config_param::DIMENSION, &dimension_value); EXPECT_EQ(dimension_value, 3); - oneapi::mkl::dft::config_value commit_status; - descriptor.get_value(oneapi::mkl::dft::config_param::COMMIT_STATUS, &commit_status); - EXPECT_EQ(commit_status, oneapi::mkl::dft::config_value::UNCOMMITTED); + oneapi::math::dft::config_value commit_status; + descriptor.get_value(oneapi::math::dft::config_param::COMMIT_STATUS, &commit_status); + EXPECT_EQ(commit_status, oneapi::math::dft::config_value::UNCOMMITTED); } -template +template static void set_readonly_values() { - oneapi::mkl::dft::descriptor descriptor{ default_1d_lengths }; - - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::FORWARD_DOMAIN, - oneapi::mkl::dft::domain::REAL), - oneapi::mkl::invalid_argument); - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::FORWARD_DOMAIN, - oneapi::mkl::dft::domain::COMPLEX), - oneapi::mkl::invalid_argument); - - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::PRECISION, - oneapi::mkl::dft::precision::SINGLE), - oneapi::mkl::invalid_argument); - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::PRECISION, - oneapi::mkl::dft::precision::DOUBLE), - oneapi::mkl::invalid_argument); + oneapi::math::dft::descriptor descriptor{ default_1d_lengths }; + + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::FORWARD_DOMAIN, + oneapi::math::dft::domain::REAL), + oneapi::math::invalid_argument); + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::FORWARD_DOMAIN, + oneapi::math::dft::domain::COMPLEX), + oneapi::math::invalid_argument); + + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::PRECISION, + oneapi::math::dft::precision::SINGLE), + oneapi::math::invalid_argument); + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::PRECISION, + oneapi::math::dft::precision::DOUBLE), + oneapi::math::invalid_argument); std::int64_t set_dimension{ 3 }; - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::DIMENSION, set_dimension), - oneapi::mkl::invalid_argument); - - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::COMMIT_STATUS, - oneapi::mkl::dft::config_value::COMMITTED), - oneapi::mkl::invalid_argument); - EXPECT_THROW(descriptor.set_value(oneapi::mkl::dft::config_param::COMMIT_STATUS, - oneapi::mkl::dft::config_value::UNCOMMITTED), - oneapi::mkl::invalid_argument); + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::DIMENSION, set_dimension), + oneapi::math::invalid_argument); + + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::COMMIT_STATUS, + oneapi::math::dft::config_value::COMMITTED), + oneapi::math::invalid_argument); + EXPECT_THROW(descriptor.set_value(oneapi::math::dft::config_param::COMMIT_STATUS, + oneapi::math::dft::config_value::UNCOMMITTED), + oneapi::math::invalid_argument); } -template +template static void get_commited(sycl::queue& sycl_queue) { - oneapi::mkl::dft::descriptor descriptor{ default_1d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_1d_lengths }; commit_descriptor(descriptor, sycl_queue); - oneapi::mkl::dft::config_value commit_status; - descriptor.get_value(oneapi::mkl::dft::config_param::COMMIT_STATUS, &commit_status); - EXPECT_EQ(commit_status, oneapi::mkl::dft::config_value::COMMITTED); + oneapi::math::dft::config_value commit_status; + descriptor.get_value(oneapi::math::dft::config_param::COMMIT_STATUS, &commit_status); + EXPECT_EQ(commit_status, oneapi::math::dft::config_value::COMMITTED); } -template +template inline void recommit_values(sycl::queue& sycl_queue) { - using oneapi::mkl::dft::config_param; - using oneapi::mkl::dft::config_value; + using oneapi::math::dft::config_param; + using oneapi::math::dft::config_value; using PrecisionType = - typename std::conditional_t; using value = std::variant; // this will hold a param to change and the value to change it to using test_params = std::vector>; - oneapi::mkl::dft::descriptor descriptor{ default_1d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_1d_lengths }; EXPECT_NO_THROW(commit_descriptor(descriptor, sycl_queue)); std::array strides{ 0, 1 }; @@ -551,17 +551,17 @@ inline void recommit_values(sycl::queue& sycl_queue) { try { commit_descriptor(descriptor, sycl_queue); } - catch (oneapi::mkl::unimplemented e) { + catch (oneapi::math::unimplemented e) { std::cout << "unimplemented exception at index " << i << " with error : " << e.what() << "\ncontinuing...\n"; } - catch (oneapi::mkl::exception& e) { + catch (oneapi::math::exception& e) { FAIL() << "exception at index " << i << " with error : " << e.what(); } } } -template +template inline void change_queue_causes_wait(sycl::queue& busy_queue) { // create a queue with work on it, and then show that work is waited on when the descriptor // is committed to a new queue. @@ -577,7 +577,7 @@ inline void change_queue_causes_wait(sycl::queue& busy_queue) { sycl::queue free_queue(busy_queue.get_device(), exception_handler); // commit the descriptor on the "busy" queue - oneapi::mkl::dft::descriptor descriptor{ default_1d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_1d_lengths }; EXPECT_NO_THROW(commit_descriptor(descriptor, busy_queue)); // add some work to the busy queue @@ -609,29 +609,29 @@ inline void change_queue_causes_wait(sycl::queue& busy_queue) { ASSERT_EQ(after_status, sycl::info::event_command_status::complete); } -template +template inline void swap_out_dead_queue(sycl::queue& sycl_queue) { // test that commit still works when the previously committed queue is no longer in scope // the queue is not actually dead (due to reference counting) // commit the descriptor on the "busy" queue - oneapi::mkl::dft::descriptor descriptor{ default_1d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_1d_lengths }; { sycl::queue transient_queue(sycl_queue.get_device(), exception_handler); EXPECT_NO_THROW(commit_descriptor(descriptor, transient_queue)); } EXPECT_NO_THROW(commit_descriptor(descriptor, sycl_queue)); - using ftype = typename std::conditional_t; - using forward_type = typename std::conditional_t>; // add two so that real-complex transforms have space for all the conjugate even components auto inout = sycl::malloc_device(default_1d_lengths + 2, sycl_queue); sycl_queue.wait(); - auto transform_event = oneapi::mkl::dft::compute_forward( + auto transform_event = oneapi::math::dft::compute_forward( descriptor, inout, std::vector{}); sycl_queue.wait(); @@ -641,17 +641,17 @@ inline void swap_out_dead_queue(sycl::queue& sycl_queue) { sycl::free(inout, sycl_queue); } -template +template static int test_move() { - using config_param = oneapi::mkl::dft::config_param; + using config_param = oneapi::math::dft::config_param; // Use forward distance to test an element copied by value (ie. not on heap) std::int64_t fwdDistanceRef(123); // Use the DFT dimensions to test heap allocated values. { // Move constructor - oneapi::mkl::dft::descriptor descriptor{ default_1d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_1d_lengths }; descriptor.set_value(config_param::FWD_DISTANCE, fwdDistanceRef); - oneapi::mkl::dft::descriptor descMoved{ std::move(descriptor) }; + oneapi::math::dft::descriptor descMoved{ std::move(descriptor) }; std::int64_t fwdDistance(0), dftLength(0); descMoved.get_value(config_param::FWD_DISTANCE, &fwdDistance); EXPECT_EQ(fwdDistance, fwdDistanceRef); @@ -660,9 +660,9 @@ static int test_move() { } { // Move assignment - oneapi::mkl::dft::descriptor descriptor{ default_1d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_1d_lengths }; descriptor.set_value(config_param::FWD_DISTANCE, fwdDistanceRef); - oneapi::mkl::dft::descriptor descMoved{ default_1d_lengths }; + oneapi::math::dft::descriptor descMoved{ default_1d_lengths }; descMoved = std::move(descriptor); std::int64_t fwdDistance(0), dftLength(0); descMoved.get_value(config_param::FWD_DISTANCE, &fwdDistance); @@ -674,7 +674,7 @@ static int test_move() { return !::testing::Test::HasFailure(); } -template +template static int test_getter_setter() { set_and_get_lengths(); set_and_get_io_strides(); @@ -686,11 +686,11 @@ static int test_getter_setter() { return !::testing::Test::HasFailure(); } -template +template int test_commit(sycl::device* dev) { sycl::queue sycl_queue(*dev, exception_handler); - if constexpr (precision == oneapi::mkl::dft::precision::DOUBLE) { + if constexpr (precision == oneapi::math::dft::precision::DOUBLE) { if (!dev->has(sycl::aspect::fp64)) { std::cout << "Device does not support double precision." << std::endl; return test_skipped; @@ -699,10 +699,10 @@ int test_commit(sycl::device* dev) { // test that descriptor is supported try { - oneapi::mkl::dft::descriptor descriptor{ default_1d_lengths }; + oneapi::math::dft::descriptor descriptor{ default_1d_lengths }; commit_descriptor(descriptor, sycl_queue); } - catch (oneapi::mkl::unimplemented& e) { + catch (oneapi::math::unimplemented& e) { std::cout << "Skipping because simple commit not supported. Reason: \"" << e.what() << "\"\n"; return test_skipped; @@ -717,66 +717,66 @@ int test_commit(sycl::device* dev) { } TEST(DescriptorTests, DescriptorMoveRealSingle) { - EXPECT_TRUE((test_move())); + EXPECT_TRUE((test_move())); } TEST(DescriptorTests, DescriptorMoveRealDouble) { - EXPECT_TRUE((test_move())); + EXPECT_TRUE((test_move())); } TEST(DescriptorTests, DescriptorMoveComplexSingle) { EXPECT_TRUE( - (test_move())); + (test_move())); } TEST(DescriptorTests, DescriptorMoveComplexDouble) { EXPECT_TRUE( - (test_move())); + (test_move())); } TEST(DescriptorTests, DescriptorTestsRealSingle) { EXPECT_TRUE(( - test_getter_setter())); + test_getter_setter())); } TEST(DescriptorTests, DescriptorTestsRealDouble) { EXPECT_TRUE(( - test_getter_setter())); + test_getter_setter())); } TEST(DescriptorTests, DescriptorTestsComplexSingle) { - EXPECT_TRUE((test_getter_setter())); + EXPECT_TRUE((test_getter_setter())); } TEST(DescriptorTests, DescriptorTestsComplexDouble) { - EXPECT_TRUE((test_getter_setter())); + EXPECT_TRUE((test_getter_setter())); } class DescriptorCommitTests : public ::testing::TestWithParam {}; TEST_P(DescriptorCommitTests, DescriptorCommitTestsRealSingle) { EXPECT_TRUEORSKIP( - (test_commit( + (test_commit( GetParam()))); } TEST_P(DescriptorCommitTests, DescriptorCommitTestsRealDouble) { EXPECT_TRUEORSKIP( - (test_commit( + (test_commit( GetParam()))); } TEST_P(DescriptorCommitTests, DescriptorCommitTestsComplexSingle) { EXPECT_TRUEORSKIP( - (test_commit( + (test_commit( GetParam()))); } TEST_P(DescriptorCommitTests, DescriptorCommitTestsComplexDouble) { EXPECT_TRUEORSKIP( - (test_commit( + (test_commit( GetParam()))); } diff --git a/tests/unit_tests/dft/source/workspace_external_tests.cpp b/tests/unit_tests/dft/source/workspace_external_tests.cpp index f96544a90..15b21cd65 100644 --- a/tests/unit_tests/dft/source/workspace_external_tests.cpp +++ b/tests/unit_tests/dft/source/workspace_external_tests.cpp @@ -35,9 +35,9 @@ extern std::vector devices; class WorkspaceExternalTests : public ::testing::TestWithParam {}; -template +template int test_workspace_external_usm_impl(std::size_t dft_size, sycl::device* dev) { - using namespace oneapi::mkl::dft; + using namespace oneapi::math::dft; using scalar_t = std::conditional_t; using forward_t = std::conditional_t, scalar_t>; using backward_t = std::complex; @@ -54,7 +54,7 @@ int test_workspace_external_usm_impl(std::size_t dft_size, sycl::device* dev) { try { commit_descriptor(desc, sycl_queue); } - catch (oneapi::mkl::unimplemented&) { + catch (oneapi::math::unimplemented&) { std::cout << "Test configuration not implemented." << std::endl; return test_skipped; } @@ -106,9 +106,9 @@ int test_workspace_external_usm_impl(std::size_t dft_size, sycl::device* dev) { return sanityCheckPasses ? !::testing::Test::HasFailure() : ::testing::Test::HasFailure(); } -template +template int test_workspace_external_buffer_impl(std::size_t dft_size, sycl::device* dev) { - using namespace oneapi::mkl::dft; + using namespace oneapi::math::dft; using scalar_t = std::conditional_t; using forward_t = std::conditional_t, scalar_t>; using backward_t = std::complex; @@ -125,7 +125,7 @@ int test_workspace_external_buffer_impl(std::size_t dft_size, sycl::device* dev) try { commit_descriptor(desc, sycl_queue); } - catch (oneapi::mkl::unimplemented&) { + catch (oneapi::math::unimplemented&) { std::cout << "Test configuration not implemented." << std::endl; return test_skipped; } @@ -170,49 +170,49 @@ int test_workspace_external_buffer_impl(std::size_t dft_size, sycl::device* dev) return sanityCheckPasses ? !::testing::Test::HasFailure() : ::testing::Test::HasFailure(); } -template +template void test_workspace_external_usm(sycl::device* dev) { EXPECT_TRUEORSKIP((test_workspace_external_usm_impl(2, dev))); EXPECT_TRUEORSKIP((test_workspace_external_usm_impl(1024 * 3 * 5 * 7 * 16, dev))); } -template +template void test_workspace_external_buffer(sycl::device* dev) { EXPECT_TRUEORSKIP((test_workspace_external_buffer_impl(2, dev))); EXPECT_TRUEORSKIP((test_workspace_external_buffer_impl(1024 * 3 * 5 * 7 * 16, dev))); } TEST_P(WorkspaceExternalTests, TestWorkspaceExternalSingleUsm) { - using precision = oneapi::mkl::dft::precision; - using domain = oneapi::mkl::dft::domain; + using precision = oneapi::math::dft::precision; + using domain = oneapi::math::dft::domain; test_workspace_external_usm(GetParam()); test_workspace_external_usm(GetParam()); } TEST_P(WorkspaceExternalTests, TestWorkspaceExternalDoubleUsm) { - using precision = oneapi::mkl::dft::precision; - using domain = oneapi::mkl::dft::domain; + using precision = oneapi::math::dft::precision; + using domain = oneapi::math::dft::domain; test_workspace_external_usm(GetParam()); test_workspace_external_usm(GetParam()); } TEST_P(WorkspaceExternalTests, TestWorkspaceExternalSingleBuffer) { - using precision = oneapi::mkl::dft::precision; - using domain = oneapi::mkl::dft::domain; + using precision = oneapi::math::dft::precision; + using domain = oneapi::math::dft::domain; test_workspace_external_buffer(GetParam()); test_workspace_external_buffer(GetParam()); } TEST_P(WorkspaceExternalTests, TestWorkspaceExternalDoubleBuffer) { - using precision = oneapi::mkl::dft::precision; - using domain = oneapi::mkl::dft::domain; + using precision = oneapi::math::dft::precision; + using domain = oneapi::math::dft::domain; test_workspace_external_buffer(GetParam()); test_workspace_external_buffer(GetParam()); } /// A test where set_workspace is called when an external workspace is not set. TEST_P(WorkspaceExternalTests, SetWorkspaceOnWorkspaceAutomatic) { - using namespace oneapi::mkl::dft; + using namespace oneapi::math::dft; sycl::queue sycl_queue(*GetParam()); const int dft_len = 1024 * 3 * 5 * 7 * 16; // A size likely to require an external workspace. float* fft_data_usm = sycl::malloc_device(dft_len * 2, sycl_queue); @@ -223,7 +223,7 @@ TEST_P(WorkspaceExternalTests, SetWorkspaceOnWorkspaceAutomatic) { commit_descriptor(desc_usm, sycl_queue); commit_descriptor(desc_buf, sycl_queue); } - catch (oneapi::mkl::unimplemented&) { + catch (oneapi::math::unimplemented&) { // The DFT size may not be supported. Use a size that is likely to be supported, even if // that means no external workspace is actually used. descriptor desc_usm2(2), desc_buf2(2); @@ -264,10 +264,10 @@ TEST_P(WorkspaceExternalTests, SetWorkspaceOnWorkspaceAutomatic) { sycl_queue.wait_and_throw(); // Should not work: - EXPECT_THROW(compute_forward(desc_usm, fft_data_buf), oneapi::mkl::invalid_argument); - EXPECT_THROW(compute_forward(desc_buf, fft_data_usm), oneapi::mkl::invalid_argument); - EXPECT_THROW(compute_backward(desc_usm, fft_data_buf), oneapi::mkl::invalid_argument); - EXPECT_THROW(compute_backward(desc_buf, fft_data_usm), oneapi::mkl::invalid_argument); + EXPECT_THROW(compute_forward(desc_usm, fft_data_buf), oneapi::math::invalid_argument); + EXPECT_THROW(compute_forward(desc_buf, fft_data_usm), oneapi::math::invalid_argument); + EXPECT_THROW(compute_backward(desc_usm, fft_data_buf), oneapi::math::invalid_argument); + EXPECT_THROW(compute_backward(desc_buf, fft_data_usm), oneapi::math::invalid_argument); sycl_queue.wait_and_throw(); // Free any allocations: @@ -277,7 +277,7 @@ TEST_P(WorkspaceExternalTests, SetWorkspaceOnWorkspaceAutomatic) { /// Test that the implementation throws as expected. TEST_P(WorkspaceExternalTests, ThrowOnBadCalls) { - using namespace oneapi::mkl::dft; + using namespace oneapi::math::dft; sycl::queue sycl_queue(*GetParam()); const int dft_len = 1024 * 3 * 5 * 7 * 16; // A size likely to require an external workspace. float* fft_data_usm = sycl::malloc_device(dft_len * 2, sycl_queue); @@ -289,13 +289,13 @@ TEST_P(WorkspaceExternalTests, ThrowOnBadCalls) { std::int64_t workspace_bytes = -10; float* usm_workspace = nullptr; EXPECT_THROW(desc_usm.get_value(config_param::WORKSPACE_EXTERNAL_BYTES, &workspace_bytes), - oneapi::mkl::invalid_argument); - EXPECT_THROW(desc_usm.set_workspace(usm_workspace), oneapi::mkl::uninitialized); + oneapi::math::invalid_argument); + EXPECT_THROW(desc_usm.set_workspace(usm_workspace), oneapi::math::uninitialized); try { commit_descriptor(desc_usm, sycl_queue); commit_descriptor(desc_buf, sycl_queue); } - catch (oneapi::mkl::unimplemented&) { + catch (oneapi::math::unimplemented&) { // DFT size may not be supported. Use a DFT size that probably will be, even if it // won't actually use an external workspace internally. descriptor desc_usm2(2), desc_buf2(2); @@ -311,16 +311,16 @@ TEST_P(WorkspaceExternalTests, ThrowOnBadCalls) { EXPECT_GE(workspace_bytes, 0); // We haven't set a workspace, so the following should fail; - EXPECT_THROW(compute_forward(desc_usm, fft_data_usm), oneapi::mkl::invalid_argument); + EXPECT_THROW(compute_forward(desc_usm, fft_data_usm), oneapi::math::invalid_argument); sycl_queue.wait_and_throw(); - EXPECT_THROW(compute_forward(desc_usm, fft_data_buf), oneapi::mkl::invalid_argument); + EXPECT_THROW(compute_forward(desc_usm, fft_data_buf), oneapi::math::invalid_argument); sycl_queue.wait_and_throw(); if (workspace_bytes > 0) { - EXPECT_THROW(desc_usm.set_workspace(nullptr), oneapi::mkl::invalid_argument); + EXPECT_THROW(desc_usm.set_workspace(nullptr), oneapi::math::invalid_argument); sycl::buffer undersize_workspace( static_cast(workspace_bytes) / sizeof(float) - 1); - EXPECT_THROW(desc_buf.set_workspace(undersize_workspace), oneapi::mkl::invalid_argument); + EXPECT_THROW(desc_buf.set_workspace(undersize_workspace), oneapi::math::invalid_argument); } usm_workspace = sycl::malloc_device( @@ -341,10 +341,10 @@ TEST_P(WorkspaceExternalTests, ThrowOnBadCalls) { sycl_queue.wait_and_throw(); // Should not work: - EXPECT_THROW(compute_forward(desc_usm, fft_data_buf), oneapi::mkl::invalid_argument); - EXPECT_THROW(compute_forward(desc_buf, fft_data_usm), oneapi::mkl::invalid_argument); - EXPECT_THROW(compute_backward(desc_usm, fft_data_buf), oneapi::mkl::invalid_argument); - EXPECT_THROW(compute_backward(desc_buf, fft_data_usm), oneapi::mkl::invalid_argument); + EXPECT_THROW(compute_forward(desc_usm, fft_data_buf), oneapi::math::invalid_argument); + EXPECT_THROW(compute_forward(desc_buf, fft_data_usm), oneapi::math::invalid_argument); + EXPECT_THROW(compute_backward(desc_usm, fft_data_buf), oneapi::math::invalid_argument); + EXPECT_THROW(compute_backward(desc_buf, fft_data_usm), oneapi::math::invalid_argument); sycl_queue.wait_and_throw(); // Free any allocations: @@ -353,7 +353,7 @@ TEST_P(WorkspaceExternalTests, ThrowOnBadCalls) { } TEST_P(WorkspaceExternalTests, RecommitBehaviour) { - using namespace oneapi::mkl::dft; + using namespace oneapi::math::dft; sycl::queue sycl_queue(*GetParam()); const int dft_len = 1024 * 3 * 5 * 7 * 16; // A size likely to require an external workspace. float* fft_data_usm = sycl::malloc_device(dft_len * 2, sycl_queue); @@ -362,7 +362,7 @@ TEST_P(WorkspaceExternalTests, RecommitBehaviour) { // WORKSPACE_EXTERNAL is NOT set. commit_descriptor(desc_usm, sycl_queue); } - catch (oneapi::mkl::unimplemented&) { + catch (oneapi::math::unimplemented&) { // DFT size may not be supported. Use a DFT size that probably will be, even if it // won't actually use an external workspace internally. descriptor desc_usm2(2); @@ -382,7 +382,7 @@ TEST_P(WorkspaceExternalTests, RecommitBehaviour) { commit_descriptor(desc_usm, sycl_queue); // No workspace, expect throw - EXPECT_THROW(compute_forward(desc_usm, fft_data_usm), oneapi::mkl::invalid_argument); + EXPECT_THROW(compute_forward(desc_usm, fft_data_usm), oneapi::math::invalid_argument); desc_usm.set_workspace(usm_workspace); @@ -391,7 +391,7 @@ TEST_P(WorkspaceExternalTests, RecommitBehaviour) { // Recommitting should require workspace to be set again. commit_descriptor(desc_usm, sycl_queue); - EXPECT_THROW(compute_forward(desc_usm, fft_data_usm), oneapi::mkl::invalid_argument); + EXPECT_THROW(compute_forward(desc_usm, fft_data_usm), oneapi::math::invalid_argument); sycl_queue.wait_and_throw(); // Free any allocations: diff --git a/tests/unit_tests/include/test_helper.hpp b/tests/unit_tests/include/test_helper.hpp index 01b277cc9..60702adf5 100644 --- a/tests/unit_tests/include/test_helper.hpp +++ b/tests/unit_tests/include/test_helper.hpp @@ -76,12 +76,12 @@ #if defined(ONEMATH_ENABLE_MKLCPU_BACKEND) || defined(ONEMATH_ENABLE_NETLIB_BACKEND) #ifdef ONEMATH_ENABLE_MKLCPU_BACKEND #define TEST_RUN_INTELCPU_SELECT_NO_ARGS(q, func) \ - func(oneapi::mkl::backend_selector{ q }) + func(oneapi::math::backend_selector{ q }) #define TEST_RUN_INTELCPU_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_INTELCPU_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #endif #else #define TEST_RUN_INTELCPU_SELECT_NO_ARGS(q, func) @@ -90,9 +90,9 @@ #ifdef ONEMATH_ENABLE_MKLGPU_BACKEND #define TEST_RUN_INTELGPU_SELECT_NO_ARGS(q, func) \ - func(oneapi::mkl::backend_selector{ q }) + func(oneapi::math::backend_selector{ q }) #define TEST_RUN_INTELGPU_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_INTELGPU_SELECT_NO_ARGS(q, func) #define TEST_RUN_INTELGPU_SELECT(q, func, ...) @@ -100,57 +100,57 @@ #ifdef ONEMATH_ENABLE_CUBLAS_BACKEND #define TEST_RUN_NVIDIAGPU_CUBLAS_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_NVIDIAGPU_CUBLAS_SELECT(q, func, ...) #endif #ifdef ONEMATH_ENABLE_CUSOLVER_BACKEND #define TEST_RUN_NVIDIAGPU_CUSOLVER_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_NVIDIAGPU_CUSOLVER_SELECT(q, func, ...) #endif #ifdef ONEMATH_ENABLE_ROCBLAS_BACKEND #define TEST_RUN_AMDGPU_ROCBLAS_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_AMDGPU_ROCBLAS_SELECT(q, func, ...) #endif #ifdef ONEMATH_ENABLE_CURAND_BACKEND #define TEST_RUN_NVIDIAGPU_CURAND_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_NVIDIAGPU_CURAND_SELECT(q, func, ...) #endif #ifdef ONEMATH_ENABLE_ROCRAND_BACKEND #define TEST_RUN_AMDGPU_ROCRAND_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_AMDGPU_ROCRAND_SELECT(q, func, ...) #endif #ifdef ONEMATH_ENABLE_ROCSOLVER_BACKEND #define TEST_RUN_AMDGPU_ROCSOLVER_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_AMDGPU_ROCSOLVER_SELECT(q, func, ...) #endif #ifdef ONEMATH_ENABLE_PORTBLAS_BACKEND #define TEST_RUN_PORTBLAS_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_PORTBLAS_SELECT(q, func, ...) #endif #ifdef ONEMATH_ENABLE_CUFFT_BACKEND #define TEST_RUN_NVIDIAGPU_CUFFT_SELECT_NO_ARGS(q, func) \ - func(oneapi::mkl::backend_selector{ q }) + func(oneapi::math::backend_selector{ q }) #define TEST_RUN_NVIDIAGPU_CUFFT_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_NVIDIAGPU_CUFFT_SELECT_NO_ARGS(q, func) #define TEST_RUN_NVIDIAGPU_CUFFT_SELECT(q, func, ...) @@ -158,9 +158,9 @@ #ifdef ONEMATH_ENABLE_ROCFFT_BACKEND #define TEST_RUN_AMDGPU_ROCFFT_SELECT_NO_ARGS(q, func) \ - func(oneapi::mkl::backend_selector{ q }) + func(oneapi::math::backend_selector{ q }) #define TEST_RUN_AMDGPU_ROCFFT_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_AMDGPU_ROCFFT_SELECT_NO_ARGS(q, func) #define TEST_RUN_AMDGPU_ROCFFT_SELECT(q, func, ...) @@ -168,9 +168,9 @@ #ifdef ONEMATH_ENABLE_PORTFFT_BACKEND #define TEST_RUN_PORTFFT_SELECT_NO_ARGS(q, func) \ - func(oneapi::mkl::backend_selector{ q }) + func(oneapi::math::backend_selector{ q }) #define TEST_RUN_PORTFFT_SELECT(q, func, ...) \ - func(oneapi::mkl::backend_selector{ q }, __VA_ARGS__) + func(oneapi::math::backend_selector{ q }, __VA_ARGS__) #else #define TEST_RUN_PORTFFT_SELECT_NO_ARGS(q, func) #define TEST_RUN_PORTFFT_SELECT(q, func, ...) @@ -290,9 +290,9 @@ class DeviceNamePrint { class LayoutDeviceNamePrint { public: std::string operator()( - testing::TestParamInfo> dev) const { + testing::TestParamInfo> dev) const { std::string layout_name = - std::get<1>(dev.param) == oneapi::mkl::layout::col_major ? "Column_Major" : "Row_Major"; + std::get<1>(dev.param) == oneapi::math::layout::col_major ? "Column_Major" : "Row_Major"; std::string dev_name = std::get<0>(dev.param)->get_info(); for (std::string::size_type i = 0; i < dev_name.size(); ++i) { if (!isalnum(dev_name[i])) @@ -304,9 +304,9 @@ class LayoutDeviceNamePrint { }; /* to accommodate Windows and Linux differences between alligned_alloc and - _aligned_malloc calls use oneapi::mkl::aligned_alloc and oneapi::mkl::aligned_free instead */ + _aligned_malloc calls use oneapi::math::aligned_alloc and oneapi::math::aligned_free instead */ namespace oneapi { -namespace mkl { +namespace math { static inline void *aligned_alloc(size_t align, size_t size) { #ifdef _WIN64 @@ -361,7 +361,7 @@ static inline void free_usm(void *p, sycl::context ctx) { sycl::free(p, ctx); } -} // namespace mkl +} // namespace math } // namespace oneapi #endif // _TEST_HELPER_HPP_ diff --git a/tests/unit_tests/lapack/include/lapack_accuracy_checks.hpp b/tests/unit_tests/lapack/include/lapack_accuracy_checks.hpp index 757541b11..8cf2ade94 100644 --- a/tests/unit_tests/lapack/include/lapack_accuracy_checks.hpp +++ b/tests/unit_tests/lapack/include/lapack_accuracy_checks.hpp @@ -127,9 +127,9 @@ bool check_geqrf_accuracy(int64_t m, int64_t n, const std::vector& A, int64_ /* |A - Q R| < |A| O(eps) */ std::vector R(m * n); int64_t ldr = m; - reference::laset(oneapi::mkl::uplo::lower, m, n, 0.0, 0.0, R.data(), ldr); - reference::lacpy(oneapi::mkl::uplo::upper, m, n, A.data(), lda, R.data(), ldr); - auto info = reference::or_un_mqr(oneapi::mkl::side::left, oneapi::mkl::transpose::nontrans, m, + reference::laset(oneapi::math::uplo::lower, m, n, 0.0, 0.0, R.data(), ldr); + reference::lacpy(oneapi::math::uplo::upper, m, n, A.data(), lda, R.data(), ldr); + auto info = reference::or_un_mqr(oneapi::math::side::left, oneapi::math::transpose::nontrans, m, n, std::min(m, n), A.data(), lda, tau.data(), R.data(), ldr); if (0 != info) { test_log::lout << "reference ormqr/unmqr failed with info = " << info << std::endl; @@ -153,7 +153,7 @@ bool check_geqrf_accuracy(int64_t m, int64_t n, const std::vector& A, int64_ } std::vector QQ(m * m); int64_t ldqq = m; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, m, m, m, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, m, m, m, 1.0, Q.data(), ldq, Q.data(), ldq, 0.0, QQ.data(), ldqq); if (!rel_id_err_check(m, QQ, ldqq)) { test_log::lout << "Orthogonality check failed" << std::endl; @@ -173,7 +173,7 @@ bool check_gerqf_accuracy(const std::vector& A, const std::vector& A_ini std::vector R(m * n); int64_t ldr = m; reference::lacpy('A', m, n, A.data(), lda, R.data(), ldr); - reference::laset(oneapi::mkl::uplo::lower, n - 1, n - 1, 0.0, 0.0, + reference::laset(oneapi::math::uplo::lower, n - 1, n - 1, 0.0, 0.0, R.data() + ((m - n + 1) + 0 * ldr), ldr); std::vector Q(lda * n); @@ -181,7 +181,7 @@ bool check_gerqf_accuracy(const std::vector& A, const std::vector& A_ini reference::lacpy('A', n, n, A.data() + ((m - n) + 0 * lda), lda, Q.data(), ldq); auto info = - reference::or_un_mrq(oneapi::mkl::side::right, oneapi::mkl::transpose::nontrans, m, n, + reference::or_un_mrq(oneapi::math::side::right, oneapi::math::transpose::nontrans, m, n, std::min(m, n), Q.data(), ldq, tau.data(), R.data(), ldr); if (0 != info) { test_log::lout << "reference ormqr/unmqr failed with info = " << info << std::endl; @@ -195,8 +195,8 @@ bool check_gerqf_accuracy(const std::vector& A, const std::vector& A_ini else { std::vector R(m * n); int64_t ldr = m; - reference::laset(oneapi::mkl::uplo::lower, m, m, 0.0, 0.0, R.data(), ldr); - reference::lacpy(oneapi::mkl::uplo::upper, m, m, A.data() + (0 + (n - m) * lda), lda, + reference::laset(oneapi::math::uplo::lower, m, m, 0.0, 0.0, R.data(), ldr); + reference::lacpy(oneapi::math::uplo::upper, m, m, A.data() + (0 + (n - m) * lda), lda, R.data() + (0 + (n - m) * ldr), ldr); std::vector Q(n * n); @@ -206,7 +206,7 @@ bool check_gerqf_accuracy(const std::vector& A, const std::vector& A_ini std::vector tau2(n); for (int64_t i = 0; i < std::min(m, n); i++) tau2[n - m + i] = tau[i]; - auto info = reference::or_un_mrq(oneapi::mkl::side::right, oneapi::mkl::transpose::nontrans, + auto info = reference::or_un_mrq(oneapi::math::side::right, oneapi::math::transpose::nontrans, m, n, n, Q.data(), ldq, tau2.data(), R.data(), ldr); if (0 != info) { test_log::lout << "reference ormqr/unmqr failed with info = " << info << std::endl; @@ -233,7 +233,7 @@ bool check_gerqf_accuracy(const std::vector& A, const std::vector& A_ini std::vector QQ(std::min(m, n) * std::min(m, n)); int64_t ldqq = std::min(m, n); - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, std::min(m, n), std::min(m, n), n, 1.0, Q.data(), ldq, Q.data(), ldq, 0.0, QQ.data(), ldqq); @@ -255,17 +255,17 @@ bool check_getrf_accuracy(int64_t m, int64_t n, const std::vector& A, int64_ /* Compute P L U */ reference::laset('A', m, n, 0.0, 0.0, residual.data(), m); if (m < n) { - reference::lacpy(oneapi::mkl::uplo::upper, m, n, A.data(), lda, residual.data(), m); - reference::trmm(oneapi::mkl::side::left, oneapi::mkl::uplo::lower, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::unit, m, n, 1.0, + reference::lacpy(oneapi::math::uplo::upper, m, n, A.data(), lda, residual.data(), m); + reference::trmm(oneapi::math::side::left, oneapi::math::uplo::lower, + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, m, n, 1.0, A.data(), lda, residual.data(), m); } else { - reference::lacpy(oneapi::mkl::uplo::lower, m, n, A.data(), lda, residual.data(), m); + reference::lacpy(oneapi::math::uplo::lower, m, n, A.data(), lda, residual.data(), m); for (int64_t diag = 0; diag < n; diag++) residual[diag + diag * m] = 1.0; - reference::trmm(oneapi::mkl::side::right, oneapi::mkl::uplo::upper, - oneapi::mkl::transpose::nontrans, oneapi::mkl::diag::nonunit, m, n, 1.0, + reference::trmm(oneapi::math::side::right, oneapi::math::uplo::upper, + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, m, n, 1.0, A.data(), lda, residual.data(), m); } reference::laswp(n, residual.data(), m, 1, std::min(m, n), ipiv.data(), -1); @@ -314,7 +314,7 @@ bool check_getri_accuracy(int64_t n, std::vector A, int64_t lda, std::vector /* Compute | I - inv(A)*A |. Store in residual array */ reference::laset('A', n, n, 0.0, 1.0, residual.data(), n); - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, n, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, -1.0, A.data(), lda, A_initial.data(), lda, 1.0, residual.data(), n); /* | I - inv(A)*A | / ( |A| * |inv(A)| * n * ulp ) */ @@ -331,7 +331,7 @@ bool check_getri_accuracy(int64_t n, std::vector A, int64_t lda, std::vector /* Compute | I - A*inv(A) |. Store in residual */ reference::laset('A', n, n, 0.0, 1.0, residual.data(), n); - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, n, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, -1.0, A_initial.data(), lda, A.data(), lda, 1.0, residual.data(), n); /* | I - A*inv(A) | / ( |A| * |inv(A)| * n * ulp ) */ @@ -350,13 +350,13 @@ bool check_getri_accuracy(int64_t n, std::vector A, int64_t lda, std::vector } template -bool check_getrs_accuracy(oneapi::mkl::transpose transa, int64_t n, int64_t nrhs, +bool check_getrs_accuracy(oneapi::math::transpose transa, int64_t n, int64_t nrhs, const std::vector& B, int64_t ldb, const std::vector& A_initial, int64_t lda, std::vector B_initial) { using fp_real = typename complex_info::real_type; // Compute A*X - B. Store result in B_initial - reference::gemm(transa, oneapi::mkl::transpose::nontrans, n, nrhs, n, -1.0, A_initial.data(), + reference::gemm(transa, oneapi::math::transpose::nontrans, n, nrhs, n, -1.0, A_initial.data(), lda, B.data(), ldb, 1.0, B_initial.data(), ldb); // Compute norm residual |A*X - B| @@ -384,25 +384,25 @@ bool check_getrs_accuracy(oneapi::mkl::transpose transa, int64_t n, int64_t nrhs } template -bool check_or_un_gbr_accuracy(oneapi::mkl::generate vect, int64_t m, int64_t n, int64_t k, +bool check_or_un_gbr_accuracy(oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, const std::vector& Q, int64_t ldq) { bool result = true; - if (vect == oneapi::mkl::generate::Q) { + if (vect == oneapi::math::generate::Q) { int64_t rows_Q = m; int64_t cols_Q = (m >= k) ? n : m; /* | I - Q'Q | < m O(eps) */ std::vector QQ(cols_Q * cols_Q); int64_t ldqq = cols_Q; - reference::gemm(oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, cols_Q, + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, cols_Q, cols_Q, rows_Q, 1.0, Q.data(), ldq, Q.data(), ldq, 0.0, QQ.data(), ldqq); if (!rel_id_err_check(cols_Q, QQ, ldqq)) { test_log::lout << "Q Orthogonality check failed" << std::endl; result = false; } } - else { /* vect == oneapi::mkl::generate::P */ + else { /* vect == oneapi::math::generate::P */ auto& P = Q; auto& ldp = ldq; int64_t rows_P = (k < n) ? m : n; @@ -411,7 +411,7 @@ bool check_or_un_gbr_accuracy(oneapi::mkl::generate vect, int64_t m, int64_t n, /* | I - (P')(P')' | < m O(eps) */ std::vector PP(rows_P * rows_P); int64_t ldpp = rows_P; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, rows_P, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, rows_P, rows_P, cols_P, 1.0, P.data(), ldp, P.data(), ldp, 0.0, PP.data(), ldpp); if (!rel_id_err_check(rows_P, PP, ldpp)) { test_log::lout << "P^t Orthogonality check failed" << std::endl; @@ -428,7 +428,7 @@ bool check_or_un_gqr_accuracy(int64_t m, int64_t n, const std::vector& Q, in /* | I - Q'Q | < m O(eps) */ std::vector QQ(n * n); int64_t ldqq = n; - reference::gemm(oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, n, n, m, + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, n, m, 1.0, Q.data(), ldq, Q.data(), ldq, 0.0, QQ.data(), ldqq); if (!rel_id_err_check(n, QQ, n)) { test_log::lout << "Orthogonality check failed" << std::endl; @@ -444,7 +444,7 @@ bool check_or_un_gtr_accuracy(int64_t n, const std::vector& Q, int64_t ldq) /* | I - Q'Q | < m O(eps) */ std::vector QQ(n * n); int64_t ldqq = n; - reference::gemm(oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, n, n, n, + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, Q.data(), ldq, Q.data(), ldq, 0.0, QQ.data(), ldqq); if (!rel_id_err_check(n, QQ, n)) { test_log::lout << "Orthogonality check failed" << std::endl; @@ -455,7 +455,7 @@ bool check_or_un_gtr_accuracy(int64_t n, const std::vector& Q, int64_t ldq) template bool check_potrf_accuracy(const std::vector& init, const std::vector& sol, - oneapi::mkl::uplo uplo, int64_t n, int64_t lda) { + oneapi::math::uplo uplo, int64_t n, int64_t lda) { using fp_real = typename complex_info::real_type; std::vector ref(init); @@ -464,7 +464,7 @@ bool check_potrf_accuracy(const std::vector& init, const std::vector& so fp_real eps = reference::lamch('e'); fp_real error, max_error = 0; bool lower = - (uplo == oneapi::mkl::uplo:: + (uplo == oneapi::math::uplo:: upper); // lower for row-major (which is this source) is upper for column major bool result = true; // Check solution values are inside allowed error bounds derived in: @@ -490,14 +490,14 @@ bool check_potrf_accuracy(const std::vector& init, const std::vector& so } template -bool check_potrs_accuracy(oneapi::mkl::uplo uplo, int64_t n, int64_t nrhs, const std::vector& B, +bool check_potrs_accuracy(oneapi::math::uplo uplo, int64_t n, int64_t nrhs, const std::vector& B, int64_t ldb, std::vector A_initial, int64_t lda, std::vector B_initial) { using fp_real = typename complex_info::real_type; hermitian_to_full(uplo, n, A_initial, lda); // Compute A*X - B. Store result in B_initial - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, nrhs, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, nrhs, n, -1.0, A_initial.data(), lda, B.data(), ldb, 1.0, B_initial.data(), ldb); // Compute norm residual |A*X - B| @@ -525,7 +525,7 @@ bool check_potrs_accuracy(oneapi::mkl::uplo uplo, int64_t n, int64_t nrhs, const } template -bool check_sy_he_evd_accuracy(oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, int64_t n, +bool check_sy_he_evd_accuracy(oneapi::math::job jobz, oneapi::math::uplo uplo, int64_t n, const std::vector& A, int64_t lda, const std::vector::real_type>& w, std::vector A_initial) { @@ -540,10 +540,10 @@ bool check_sy_he_evd_accuracy(oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, int std::vector D_ref(n); if constexpr (complex_info::is_complex) - reference::heevd(oneapi::mkl::job::novec, uplo, n, std::vector(A_initial).data(), lda, + reference::heevd(oneapi::math::job::novec, uplo, n, std::vector(A_initial).data(), lda, D_ref.data()); else - reference::syevd(oneapi::mkl::job::novec, uplo, n, std::vector(A_initial).data(), lda, + reference::syevd(oneapi::math::job::novec, uplo, n, std::vector(A_initial).data(), lda, D_ref.data()); if (!rel_vec_err_check(n, D_ref, D, 10.0)) { @@ -551,7 +551,7 @@ bool check_sy_he_evd_accuracy(oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, int result = false; } - if (oneapi::mkl::job::vec == jobz) { + if (oneapi::math::job::vec == jobz) { /* |A - Z D Z'| < |A| O(eps) */ std::vector ZD(n * n); int64_t ldzd = n; @@ -560,7 +560,7 @@ bool check_sy_he_evd_accuracy(oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, int for (int64_t col = 0; col < n; col++) for (int64_t row = 0; row < n; row++) ZD[row + col * ldzd] = Z[row + col * ldz] * D[col]; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, n, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, n, n, n, 1.0, ZD.data(), ldzd, Z.data(), ldz, 0.0, ZDZ.data(), ldzdz); if (!rel_mat_err_check(n, n, A_initial, lda, ZDZ, ldzdz)) { @@ -571,9 +571,9 @@ bool check_sy_he_evd_accuracy(oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, int /* |I - Z Z'| < n O(eps) */ std::vector ZZ(n * n); int64_t ldzz = n; - reference::sy_he_rk(oneapi::mkl::uplo::upper, oneapi::mkl::transpose::nontrans, n, n, 1.0, + reference::sy_he_rk(oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, n, n, 1.0, Z.data(), ldz, 0.0, ZZ.data(), ldzz); - hermitian_to_full(oneapi::mkl::uplo::upper, n, ZZ, ldzz); + hermitian_to_full(oneapi::math::uplo::upper, n, ZZ, ldzz); if (!rel_id_err_check(n, ZZ, ldzz)) { test_log::lout << "Orthogonality check failed" << std::endl; result = false; @@ -583,18 +583,18 @@ bool check_sy_he_evd_accuracy(oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, int } template -bool check_trtrs_accuracy(oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, int64_t n, int64_t nrhs, std::vector A, +bool check_trtrs_accuracy(oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, int64_t n, int64_t nrhs, std::vector A, int64_t lda, const std::vector& B, int64_t ldb, const std::vector& B_initial) { using fp_real = typename complex_info::real_type; fp_real threshold = 10.0; /* |A x - b| = |A (x-x_0)| < |A| |x-x0| < |A| |x| cond(A) O(eps) */ - if (diag == oneapi::mkl::diag::unit) + if (diag == oneapi::math::diag::unit) for (int64_t d = 0; d < n; d++) A[d + d * lda] = 1.0; - if (uplo == oneapi::mkl::uplo::upper) + if (uplo == oneapi::math::uplo::upper) for (int64_t col = 0; col < n; col++) for (int64_t row = col + 1; row < n; row++) A[row + col * lda] = 0.0; @@ -607,7 +607,7 @@ bool check_trtrs_accuracy(oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, auto norm_x = reference::lange('I', n, nrhs, B.data(), ldb); fp_real cond_A; - if (diag == oneapi::mkl::diag::unit) + if (diag == oneapi::math::diag::unit) cond_A = 1.0; else { fp_real min = std::abs(A[0]); @@ -626,7 +626,7 @@ bool check_trtrs_accuracy(oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, std::vector residual(n * nrhs); int64_t ldr = n; - reference::gemm(trans, oneapi::mkl::transpose::nontrans, n, nrhs, n, 1.0, A.data(), lda, + reference::gemm(trans, oneapi::math::transpose::nontrans, n, nrhs, n, 1.0, A.data(), lda, B.data(), ldb, 0.0, residual.data(), ldr); for (int64_t col = 0; col < nrhs; col++) for (int64_t row = 0; row < n; row++) diff --git a/tests/unit_tests/lapack/include/lapack_common.hpp b/tests/unit_tests/lapack/include/lapack_common.hpp index 1db0f9121..5df950392 100644 --- a/tests/unit_tests/lapack/include/lapack_common.hpp +++ b/tests/unit_tests/lapack/include/lapack_common.hpp @@ -123,9 +123,9 @@ inline std::complex rand_scalar(uint64_t& seed) { } template -void rand_matrix(uint64_t& seed, oneapi::mkl::transpose trans, int64_t m, int64_t n, +void rand_matrix(uint64_t& seed, oneapi::math::transpose trans, int64_t m, int64_t n, std::vector& M, int64_t ld, int64_t offset = 0) { - if (trans == oneapi::mkl::transpose::nontrans) + if (trans == oneapi::math::transpose::nontrans) for (int64_t col = 0; col < n; col++) for (int64_t row = 0; row < m; row++) M[offset + row + col * ld] = rand_scalar(seed); @@ -136,12 +136,12 @@ void rand_matrix(uint64_t& seed, oneapi::mkl::transpose trans, int64_t m, int64_ } template -void rand_matrix_diag_dom(uint64_t& seed, oneapi::mkl::transpose trans, int64_t m, int64_t n, +void rand_matrix_diag_dom(uint64_t& seed, oneapi::math::transpose trans, int64_t m, int64_t n, std::vector& M, int64_t ld, int64_t offset = 0) { using fp_real = typename complex_info::real_type; int64_t minsh; minsh = std::min(m, n); - if (trans == oneapi::mkl::transpose::nontrans) + if (trans == oneapi::math::transpose::nontrans) for (int64_t col = 0; col < n; col++) for (int64_t row = 0; row < m; row++) { M[offset + row + col * ld] = rand_scalar(seed); @@ -158,11 +158,11 @@ void rand_matrix_diag_dom(uint64_t& seed, oneapi::mkl::transpose trans, int64_t } template -void rand_symmetric_matrix(uint64_t& seed, oneapi::mkl::uplo uplo, int64_t n, std::vector& M, +void rand_symmetric_matrix(uint64_t& seed, oneapi::math::uplo uplo, int64_t n, std::vector& M, int64_t ld, int64_t offset = 0) { using fp_real = typename complex_info::real_type; - if (uplo == oneapi::mkl::uplo::upper) + if (uplo == oneapi::math::uplo::upper) for (int64_t col = 0; col < n; col++) for (int64_t row = 0; row <= col; row++) M[offset + row + col * ld] = rand_scalar(seed); @@ -173,7 +173,7 @@ void rand_symmetric_matrix(uint64_t& seed, oneapi::mkl::uplo uplo, int64_t n, st } template -void rand_hermitian_matrix(uint64_t& seed, oneapi::mkl::uplo uplo, int64_t n, std::vector& M, +void rand_hermitian_matrix(uint64_t& seed, oneapi::math::uplo uplo, int64_t n, std::vector& M, int64_t ld, int64_t offset = 0) { using fp_real = typename complex_info::real_type; @@ -183,7 +183,7 @@ void rand_hermitian_matrix(uint64_t& seed, oneapi::mkl::uplo uplo, int64_t n, st } template -void rand_pos_def_matrix(uint64_t& seed, oneapi::mkl::uplo uplo, int64_t n, std::vector& M, +void rand_pos_def_matrix(uint64_t& seed, oneapi::math::uplo uplo, int64_t n, std::vector& M, int64_t ld, int64_t offset = 0) { using fp_real = typename complex_info::real_type; @@ -194,8 +194,8 @@ void rand_pos_def_matrix(uint64_t& seed, oneapi::mkl::uplo uplo, int64_t n, std: } template -void symmetric_to_full(oneapi::mkl::uplo uplo, int64_t n, std::vector& A, int64_t lda) { - if (oneapi::mkl::uplo::upper == uplo) +void symmetric_to_full(oneapi::math::uplo uplo, int64_t n, std::vector& A, int64_t lda) { + if (oneapi::math::uplo::upper == uplo) for (int64_t col = 0; col < n; col++) for (int64_t row = col + 1; row < n; row++) A[row + col * lda] = A[col + row * lda]; @@ -207,10 +207,10 @@ void symmetric_to_full(oneapi::mkl::uplo uplo, int64_t n, std::vector& A, in } template -void hermitian_to_full(oneapi::mkl::uplo uplo, int64_t n, std::vector& A, int64_t lda) { +void hermitian_to_full(oneapi::math::uplo uplo, int64_t n, std::vector& A, int64_t lda) { for (int64_t diag = 0; diag < n; diag++) A[diag + diag * lda] = get_real(A[diag + diag * lda]); - if (oneapi::mkl::uplo::upper == uplo) + if (oneapi::math::uplo::upper == uplo) for (int64_t col = 0; col < n; col++) for (int64_t row = col + 1; row < n; row++) A[row + col * lda] = get_conj(A[col + row * lda]); diff --git a/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp b/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp index cb09ec16a..076a9e19f 100644 --- a/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp +++ b/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp @@ -39,136 +39,136 @@ static_assert(sizeof(lapack_int) == 8); static_assert(sizeof(CBLAS_INT) == 8); namespace reference { -inline CBLAS_TRANSPOSE cblas_trans(oneapi::mkl::transpose t) { - if (t == oneapi::mkl::transpose::nontrans) +inline CBLAS_TRANSPOSE cblas_trans(oneapi::math::transpose t) { + if (t == oneapi::math::transpose::nontrans) return CblasNoTrans; - if (t == oneapi::mkl::transpose::trans) + if (t == oneapi::math::transpose::trans) return CblasTrans; - if (t == oneapi::mkl::transpose::conjtrans) + if (t == oneapi::math::transpose::conjtrans) return CblasConjTrans; return CblasNoTrans; } -inline CBLAS_UPLO cblas_uplo(oneapi::mkl::uplo u) { - if (u == oneapi::mkl::uplo::upper) +inline CBLAS_UPLO cblas_uplo(oneapi::math::uplo u) { + if (u == oneapi::math::uplo::upper) return CblasUpper; - if (u == oneapi::mkl::uplo::lower) + if (u == oneapi::math::uplo::lower) return CblasLower; return CblasUpper; } -inline CBLAS_DIAG cblas_diag(oneapi::mkl::diag d) { - if (d == oneapi::mkl::diag::nonunit) +inline CBLAS_DIAG cblas_diag(oneapi::math::diag d) { + if (d == oneapi::math::diag::nonunit) return CblasNonUnit; - if (d == oneapi::mkl::diag::unit) + if (d == oneapi::math::diag::unit) return CblasUnit; return CblasNonUnit; } inline CBLAS_SIDE cblas_side(const char *c) { return *c == 'R' || *c == 'r' ? CblasRight : CblasLeft; } -inline CBLAS_SIDE cblas_side(oneapi::mkl::side s) { - if (s == oneapi::mkl::side::left) +inline CBLAS_SIDE cblas_side(oneapi::math::side s) { + if (s == oneapi::math::side::left) return CblasLeft; - if (s == oneapi::mkl::side::right) + if (s == oneapi::math::side::right) return CblasRight; return CblasLeft; } -inline char to_char(oneapi::mkl::transpose t) { - if (t == oneapi::mkl::transpose::nontrans) +inline char to_char(oneapi::math::transpose t) { + if (t == oneapi::math::transpose::nontrans) return 'N'; - if (t == oneapi::mkl::transpose::trans) + if (t == oneapi::math::transpose::trans) return 'T'; - if (t == oneapi::mkl::transpose::conjtrans) + if (t == oneapi::math::transpose::conjtrans) return 'C'; return 'N'; } -inline char to_char(oneapi::mkl::offset t) { - if (t == oneapi::mkl::offset::fix) +inline char to_char(oneapi::math::offset t) { + if (t == oneapi::math::offset::fix) return 'F'; - if (t == oneapi::mkl::offset::row) + if (t == oneapi::math::offset::row) return 'R'; - if (t == oneapi::mkl::offset::column) + if (t == oneapi::math::offset::column) return 'C'; return 'N'; } -inline char to_char(oneapi::mkl::uplo u) { - if (u == oneapi::mkl::uplo::upper) +inline char to_char(oneapi::math::uplo u) { + if (u == oneapi::math::uplo::upper) return 'U'; - if (u == oneapi::mkl::uplo::lower) + if (u == oneapi::math::uplo::lower) return 'L'; return 'U'; } -inline char to_char(oneapi::mkl::diag d) { - if (d == oneapi::mkl::diag::nonunit) +inline char to_char(oneapi::math::diag d) { + if (d == oneapi::math::diag::nonunit) return 'N'; - if (d == oneapi::mkl::diag::unit) + if (d == oneapi::math::diag::unit) return 'U'; return 'N'; } -inline char to_char(oneapi::mkl::side s) { - if (s == oneapi::mkl::side::left) +inline char to_char(oneapi::math::side s) { + if (s == oneapi::math::side::left) return 'L'; - if (s == oneapi::mkl::side::right) + if (s == oneapi::math::side::right) return 'R'; return 'L'; } -inline char to_char(oneapi::mkl::job j) { - if (j == oneapi::mkl::job::novec) +inline char to_char(oneapi::math::job j) { + if (j == oneapi::math::job::novec) return 'N'; - if (j == oneapi::mkl::job::vec) + if (j == oneapi::math::job::vec) return 'V'; - if (j == oneapi::mkl::job::updatevec) + if (j == oneapi::math::job::updatevec) return 'U'; - if (j == oneapi::mkl::job::allvec) + if (j == oneapi::math::job::allvec) return 'A'; - if (j == oneapi::mkl::job::somevec) + if (j == oneapi::math::job::somevec) return 'S'; - if (j == oneapi::mkl::job::overwritevec) + if (j == oneapi::math::job::overwritevec) return 'O'; return 'N'; } -inline char to_char(oneapi::mkl::jobsvd j) { - if (j == oneapi::mkl::jobsvd::novec) +inline char to_char(oneapi::math::jobsvd j) { + if (j == oneapi::math::jobsvd::novec) return 'N'; - if (j == oneapi::mkl::jobsvd::vectors) + if (j == oneapi::math::jobsvd::vectors) return 'A'; - if (j == oneapi::mkl::jobsvd::vectorsina) + if (j == oneapi::math::jobsvd::vectorsina) return 'O'; - if (j == oneapi::mkl::jobsvd::somevec) + if (j == oneapi::math::jobsvd::somevec) return 'S'; return 'N'; } -inline char to_char(oneapi::mkl::generate v) { - if (v == oneapi::mkl::generate::Q) +inline char to_char(oneapi::math::generate v) { + if (v == oneapi::math::generate::Q) return 'Q'; - if (v == oneapi::mkl::generate::P) + if (v == oneapi::math::generate::P) return 'P'; return 'Q'; } -inline void gemm(oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, int64_t m, int64_t n, +inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, int64_t n, int64_t k, float alpha, const float *a, int64_t lda, const float *b, int64_t ldb, float beta, float *c, int64_t ldc) { cblas_sgemm(CblasColMajor, cblas_trans(transa), cblas_trans(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -inline void gemm(oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, int64_t m, int64_t n, +inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, int64_t n, int64_t k, double alpha, const double *a, int64_t lda, const double *b, int64_t ldb, double beta, double *c, int64_t ldc) { cblas_dgemm(CblasColMajor, cblas_trans(transa), cblas_trans(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -inline void gemm(oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, int64_t m, int64_t n, +inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, int64_t n, int64_t k, std::complex alpha, const std::complex *a, int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, std::complex *c, int64_t ldc) { cblas_cgemm(CblasColMajor, cblas_trans(transa), cblas_trans(transb), m, n, k, (void *)&alpha, (void *)a, lda, (void *)(b), ldb, (void *)&beta, (void *)c, ldc); } -inline void gemm(oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, int64_t m, int64_t n, +inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, int64_t n, int64_t k, std::complex alpha, const std::complex *a, int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, std::complex *c, int64_t ldc) { @@ -176,98 +176,98 @@ inline void gemm(oneapi::mkl::transpose transa, oneapi::mkl::transpose transb, i (void *)a, lda, (void *)(b), ldb, (void *)&beta, (void *)c, ldc); } -inline int64_t syevd(oneapi::mkl::job j, oneapi::mkl::uplo u, int64_t n, float *a, int64_t lda, +inline int64_t syevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, float *a, int64_t lda, float *w) { return LAPACKE_ssyevd(LAPACK_COL_MAJOR, to_char(j), to_char(u), n, a, lda, w); } -inline int64_t syevd(oneapi::mkl::job j, oneapi::mkl::uplo u, int64_t n, double *a, int64_t lda, +inline int64_t syevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, double *a, int64_t lda, double *w) { return LAPACKE_dsyevd(LAPACK_COL_MAJOR, to_char(j), to_char(u), n, a, lda, w); } -inline int64_t sygvd(int64_t itype, oneapi::mkl::job j, oneapi::mkl::uplo u, int64_t n, float *a, +inline int64_t sygvd(int64_t itype, oneapi::math::job j, oneapi::math::uplo u, int64_t n, float *a, int64_t lda, float *b, int64_t ldb, float *w) { return LAPACKE_ssygvd(LAPACK_COL_MAJOR, itype, to_char(j), to_char(u), n, a, lda, b, ldb, w); } -inline int64_t sygvd(int64_t itype, oneapi::mkl::job j, oneapi::mkl::uplo u, int64_t n, double *a, +inline int64_t sygvd(int64_t itype, oneapi::math::job j, oneapi::math::uplo u, int64_t n, double *a, int64_t lda, double *b, int64_t ldb, double *w) { return LAPACKE_dsygvd(LAPACK_COL_MAJOR, itype, to_char(j), to_char(u), n, a, lda, b, ldb, w); } -inline void syrk(oneapi::mkl::uplo u, oneapi::mkl::transpose t, int64_t n, int64_t k, float alpha, +inline void syrk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, float alpha, const float *a, int64_t lda, float beta, float *c, int64_t ldc) { cblas_ssyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, c, ldc); } -inline void syrk(oneapi::mkl::uplo u, oneapi::mkl::transpose t, int64_t n, int64_t k, double alpha, +inline void syrk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, double alpha, const double *a, int64_t lda, double beta, double *c, int64_t ldc) { cblas_dsyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, c, ldc); } -inline void syrk(oneapi::mkl::uplo u, oneapi::mkl::transpose t, int64_t n, int64_t k, +inline void syrk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, std::complex alpha, const std::complex *a, int64_t lda, std::complex beta, std::complex *c, int64_t ldc) { cblas_csyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, (void *)&alpha, a, lda, (void *)&beta, (void *)c, ldc); } -inline void syrk(oneapi::mkl::uplo u, oneapi::mkl::transpose t, int64_t n, int64_t k, +inline void syrk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, std::complex alpha, const std::complex *a, int64_t lda, std::complex beta, std::complex *c, int64_t ldc) { cblas_zsyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, (void *)&alpha, a, lda, (void *)&beta, (void *)c, ldc); } -inline void herk(oneapi::mkl::uplo u, oneapi::mkl::transpose t, int64_t n, int64_t k, float alpha, +inline void herk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, float alpha, const std::complex *a, int64_t lda, float beta, std::complex *c, int64_t ldc) { cblas_cherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void *)c, ldc); } -inline void herk(oneapi::mkl::uplo u, oneapi::mkl::transpose t, int64_t n, int64_t k, double alpha, +inline void herk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, double alpha, const std::complex *a, int64_t lda, double beta, std::complex *c, int64_t ldc) { cblas_zherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void *)c, ldc); } -inline void sy_he_rk(oneapi::mkl::uplo u, oneapi::mkl::transpose t, int64_t n, int64_t k, +inline void sy_he_rk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, float alpha, const float *a, int64_t lda, float beta, float *c, int64_t ldc) { cblas_ssyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, c, ldc); } -inline void sy_he_rk(oneapi::mkl::uplo u, oneapi::mkl::transpose t, int64_t n, int64_t k, +inline void sy_he_rk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, double alpha, const double *a, int64_t lda, double beta, double *c, int64_t ldc) { cblas_dsyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, c, ldc); } -inline void sy_he_rk(oneapi::mkl::uplo u, oneapi::mkl::transpose t, int64_t n, int64_t k, +inline void sy_he_rk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, float alpha, const std::complex *a, int64_t lda, float beta, std::complex *c, int64_t ldc) { cblas_cherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void *)c, ldc); } -inline void sy_he_rk(oneapi::mkl::uplo u, oneapi::mkl::transpose t, int64_t n, int64_t k, +inline void sy_he_rk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, double alpha, const std::complex *a, int64_t lda, double beta, std::complex *c, int64_t ldc) { cblas_zherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void *)c, ldc); } -inline void trmm(oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose transa, - oneapi::mkl::diag diag, int64_t m, int64_t n, float alpha, const float *a, +inline void trmm(oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose transa, + oneapi::math::diag diag, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, float *b, int64_t ldb) { cblas_strmm(CblasColMajor, cblas_side(side), cblas_uplo(uplo), cblas_trans(transa), cblas_diag(diag), m, n, alpha, a, lda, b, ldb); } -inline void trmm(oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose transa, - oneapi::mkl::diag diag, int64_t m, int64_t n, double alpha, const double *a, +inline void trmm(oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose transa, + oneapi::math::diag diag, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, double *b, int64_t ldb) { cblas_dtrmm(CblasColMajor, cblas_side(side), cblas_uplo(uplo), cblas_trans(transa), cblas_diag(diag), m, n, alpha, a, lda, b, ldb); } -inline void trmm(oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose transa, - oneapi::mkl::diag diag, int64_t m, int64_t n, std::complex alpha, +inline void trmm(oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose transa, + oneapi::math::diag diag, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb) { cblas_ctrmm(CblasColMajor, cblas_side(side), cblas_uplo(uplo), cblas_trans(transa), cblas_diag(diag), m, n, (void *)&alpha, (void *)(a), lda, (void *)(b), ldb); } -inline void trmm(oneapi::mkl::side side, oneapi::mkl::uplo uplo, oneapi::mkl::transpose transa, - oneapi::mkl::diag diag, int64_t m, int64_t n, std::complex alpha, +inline void trmm(oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose transa, + oneapi::math::diag diag, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb) { cblas_ztrmm(CblasColMajor, cblas_side(side), cblas_uplo(uplo), cblas_trans(transa), cblas_diag(diag), m, n, (void *)&alpha, (void *)(a), lda, (void *)(b), ldb); @@ -314,29 +314,29 @@ inline double lange(char norm, int64_t m, int64_t n, const std::complex reinterpret_cast(a), lda); } -inline float lanhe(char norm, oneapi::mkl::uplo u, int64_t n, const std::complex *a, +inline float lanhe(char norm, oneapi::math::uplo u, int64_t n, const std::complex *a, int64_t lda) { return LAPACKE_clanhe(LAPACK_COL_MAJOR, norm, to_char(u), n, reinterpret_cast(a), lda); } -inline double lanhe(char norm, oneapi::mkl::uplo u, int64_t n, const std::complex *a, +inline double lanhe(char norm, oneapi::math::uplo u, int64_t n, const std::complex *a, int64_t lda) { return LAPACKE_zlanhe(LAPACK_COL_MAJOR, norm, to_char(u), n, reinterpret_cast(a), lda); } -inline float lansy(char norm, oneapi::mkl::uplo u, int64_t n, const std::complex *a, +inline float lansy(char norm, oneapi::math::uplo u, int64_t n, const std::complex *a, int64_t lda) { return LAPACKE_clansy(LAPACK_COL_MAJOR, norm, to_char(u), n, reinterpret_cast(a), lda); } -inline double lansy(char norm, oneapi::mkl::uplo u, int64_t n, const double *a, int64_t lda) { +inline double lansy(char norm, oneapi::math::uplo u, int64_t n, const double *a, int64_t lda) { return LAPACKE_dlansy(LAPACK_COL_MAJOR, norm, to_char(u), n, a, lda); } -inline float lansy(char norm, oneapi::mkl::uplo u, int64_t n, const float *a, int64_t lda) { +inline float lansy(char norm, oneapi::math::uplo u, int64_t n, const float *a, int64_t lda) { return LAPACKE_slansy(LAPACK_COL_MAJOR, norm, to_char(u), n, a, lda); } -inline double lansy(char norm, oneapi::mkl::uplo u, int64_t n, const std::complex *a, +inline double lansy(char norm, oneapi::math::uplo u, int64_t n, const std::complex *a, int64_t lda) { return LAPACKE_zlansy(LAPACK_COL_MAJOR, norm, to_char(u), n, reinterpret_cast(a), lda); @@ -362,43 +362,43 @@ inline int64_t lacpy(char u, int64_t m, int64_t n, const std::complex *a reinterpret_cast(a), lda, reinterpret_cast(b), ldb); } -inline int64_t lacpy(oneapi::mkl::uplo u, int64_t m, int64_t n, const std::complex *a, +inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb) { return LAPACKE_clacpy(LAPACK_COL_MAJOR, to_char(u), m, n, reinterpret_cast(a), lda, reinterpret_cast(b), ldb); } -inline int64_t lacpy(oneapi::mkl::uplo u, int64_t m, int64_t n, const double *a, int64_t lda, +inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const double *a, int64_t lda, double *b, int64_t ldb) { return LAPACKE_dlacpy(LAPACK_COL_MAJOR, to_char(u), m, n, a, lda, b, ldb); } -inline int64_t lacpy(oneapi::mkl::uplo u, int64_t m, int64_t n, const float *a, int64_t lda, +inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const float *a, int64_t lda, float *b, int64_t ldb) { return LAPACKE_slacpy(LAPACK_COL_MAJOR, to_char(u), m, n, a, lda, b, ldb); } -inline int64_t lacpy(oneapi::mkl::uplo u, int64_t m, int64_t n, const std::complex *a, +inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb) { return LAPACKE_zlacpy(LAPACK_COL_MAJOR, to_char(u), m, n, reinterpret_cast(a), lda, reinterpret_cast(b), ldb); } -inline int64_t laset(oneapi::mkl::uplo u, int64_t m, int64_t n, std::complex alpha, +inline int64_t laset(oneapi::math::uplo u, int64_t m, int64_t n, std::complex alpha, std::complex beta, std::complex *a, int64_t lda) { return LAPACKE_claset(LAPACK_COL_MAJOR, to_char(u), m, n, reinterpret_cast(alpha), reinterpret_cast(beta), reinterpret_cast(a), lda); } -inline int64_t laset(oneapi::mkl::uplo u, int64_t m, int64_t n, double alpha, double beta, +inline int64_t laset(oneapi::math::uplo u, int64_t m, int64_t n, double alpha, double beta, double *a, int64_t lda) { return LAPACKE_dlaset(LAPACK_COL_MAJOR, to_char(u), m, n, alpha, beta, a, lda); } -inline int64_t laset(oneapi::mkl::uplo u, int64_t m, int64_t n, float alpha, float beta, float *a, +inline int64_t laset(oneapi::math::uplo u, int64_t m, int64_t n, float alpha, float beta, float *a, int64_t lda) { return LAPACKE_slaset(LAPACK_COL_MAJOR, to_char(u), m, n, alpha, beta, a, lda); } -inline int64_t laset(oneapi::mkl::uplo u, int64_t m, int64_t n, std::complex alpha, +inline int64_t laset(oneapi::math::uplo u, int64_t m, int64_t n, std::complex alpha, std::complex beta, std::complex *a, int64_t lda) { return LAPACKE_zlaset(LAPACK_COL_MAJOR, to_char(u), m, n, reinterpret_cast(alpha), @@ -482,7 +482,7 @@ inline int64_t gerqf(int64_t m, int64_t n, std::complex *a, int64_t lda, reinterpret_cast(tau)); } -inline int64_t gesvd(oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, int64_t m, int64_t n, +inline int64_t gesvd(oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, int64_t m, int64_t n, std::complex *a, int64_t lda, float *s, std::complex *u, int64_t ldu, std::complex *vt, int64_t ldvt, float *superb) { return LAPACKE_cgesvd(LAPACK_COL_MAJOR, to_char(jobu), to_char(jobvt), m, n, @@ -490,19 +490,19 @@ inline int64_t gesvd(oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, int64_ reinterpret_cast(u), ldu, reinterpret_cast(vt), ldvt, superb); } -inline int64_t gesvd(oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, int64_t m, int64_t n, +inline int64_t gesvd(oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, int64_t m, int64_t n, double *a, int64_t lda, double *s, double *u, int64_t ldu, double *vt, int64_t ldvt, double *superb) { return LAPACKE_dgesvd(LAPACK_COL_MAJOR, to_char(jobu), to_char(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, superb); } -inline int64_t gesvd(oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, int64_t m, int64_t n, +inline int64_t gesvd(oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, int64_t m, int64_t n, float *a, int64_t lda, float *s, float *u, int64_t ldu, float *vt, int64_t ldvt, float *superb) { return LAPACKE_sgesvd(LAPACK_COL_MAJOR, to_char(jobu), to_char(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, superb); } -inline int64_t gesvd(oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, int64_t m, int64_t n, +inline int64_t gesvd(oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, int64_t m, int64_t n, std::complex *a, int64_t lda, double *s, std::complex *u, int64_t ldu, std::complex *vt, int64_t ldvt, double *superb) { return LAPACKE_zgesvd(LAPACK_COL_MAJOR, to_char(jobu), to_char(jobvt), m, n, @@ -526,25 +526,25 @@ inline int64_t getrf(int64_t m, int64_t n, std::complex *a, int64_t lda, reinterpret_cast(ipiv)); } -inline int64_t heevd(oneapi::mkl::job j, oneapi::mkl::uplo u, int64_t n, std::complex *a, +inline int64_t heevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, float *w) { return LAPACKE_cheevd(LAPACK_COL_MAJOR, to_char(j), to_char(u), n, reinterpret_cast(a), lda, w); } -inline int64_t heevd(oneapi::mkl::job j, oneapi::mkl::uplo u, int64_t n, std::complex *a, +inline int64_t heevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, double *w) { return LAPACKE_zheevd(LAPACK_COL_MAJOR, to_char(j), to_char(u), n, reinterpret_cast(a), lda, w); } -inline int64_t hegvd(int64_t itype, oneapi::mkl::job j, oneapi::mkl::uplo u, int64_t n, +inline int64_t hegvd(int64_t itype, oneapi::math::job j, oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, std::complex *b, int64_t ldb, float *w) { return LAPACKE_chegvd(LAPACK_COL_MAJOR, itype, to_char(j), to_char(u), n, reinterpret_cast(a), lda, reinterpret_cast(b), ldb, w); } -inline int64_t hegvd(int64_t itype, oneapi::mkl::job j, oneapi::mkl::uplo u, int64_t n, +inline int64_t hegvd(int64_t itype, oneapi::math::job j, oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, std::complex *b, int64_t ldb, double *w) { return LAPACKE_zhegvd(LAPACK_COL_MAJOR, itype, to_char(j), to_char(u), n, @@ -552,46 +552,46 @@ inline int64_t hegvd(int64_t itype, oneapi::mkl::job j, oneapi::mkl::uplo u, int reinterpret_cast(b), ldb, w); } -inline int64_t hetrd(oneapi::mkl::uplo u, int64_t n, std::complex *a, int64_t lda, float *d, +inline int64_t hetrd(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, float *d, float *e, std::complex *tau) { return LAPACKE_chetrd(LAPACK_COL_MAJOR, to_char(u), n, reinterpret_cast(a), lda, d, e, reinterpret_cast(tau)); } -inline int64_t hetrd(oneapi::mkl::uplo u, int64_t n, std::complex *a, int64_t lda, +inline int64_t hetrd(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, double *d, double *e, std::complex *tau) { return LAPACKE_zhetrd(LAPACK_COL_MAJOR, to_char(u), n, reinterpret_cast(a), lda, d, e, reinterpret_cast(tau)); } -inline int64_t hetrf(oneapi::mkl::uplo u, int64_t n, std::complex *a, int64_t lda, +inline int64_t hetrf(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, int64_t *ipiv) { return LAPACKE_chetrf(LAPACK_COL_MAJOR, to_char(u), n, reinterpret_cast(a), lda, reinterpret_cast(ipiv)); } -inline int64_t hetrf(oneapi::mkl::uplo u, int64_t n, std::complex *a, int64_t lda, +inline int64_t hetrf(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, int64_t *ipiv) { return LAPACKE_zhetrf(LAPACK_COL_MAJOR, to_char(u), n, reinterpret_cast(a), lda, reinterpret_cast(ipiv)); } -inline int64_t ungtr(oneapi::mkl::uplo u, int64_t n, std::complex *a, int64_t lda, +inline int64_t ungtr(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, const std::complex *tau) { return LAPACKE_cungtr(LAPACK_COL_MAJOR, to_char(u), n, reinterpret_cast(a), lda, reinterpret_cast(tau)); } -inline int64_t ungtr(oneapi::mkl::uplo u, int64_t n, std::complex *a, int64_t lda, +inline int64_t ungtr(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, const std::complex *tau) { return LAPACKE_zungtr(LAPACK_COL_MAJOR, to_char(u), n, reinterpret_cast(a), lda, reinterpret_cast(tau)); } -inline int64_t unmtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mkl::transpose trans, +inline int64_t unmtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, int64_t m, int64_t n, const std::complex *a, int64_t lda, const std::complex *tau, std::complex *c, int64_t ldc) { return LAPACKE_cunmtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, @@ -599,7 +599,7 @@ inline int64_t unmtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mkl::t reinterpret_cast(tau), reinterpret_cast(c), ldc); } -inline int64_t unmtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mkl::transpose trans, +inline int64_t unmtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, int64_t m, int64_t n, const std::complex *a, int64_t lda, const std::complex *tau, std::complex *c, int64_t ldc) { return LAPACKE_zunmtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, @@ -608,39 +608,39 @@ inline int64_t unmtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mkl::t reinterpret_cast(c), ldc); } -inline int64_t orgtr(oneapi::mkl::uplo u, int64_t n, double *a, int64_t lda, const double *tau) { +inline int64_t orgtr(oneapi::math::uplo u, int64_t n, double *a, int64_t lda, const double *tau) { return LAPACKE_dorgtr(LAPACK_COL_MAJOR, to_char(u), n, a, lda, tau); } -inline int64_t orgtr(oneapi::mkl::uplo u, int64_t n, float *a, int64_t lda, const float *tau) { +inline int64_t orgtr(oneapi::math::uplo u, int64_t n, float *a, int64_t lda, const float *tau) { return LAPACKE_sorgtr(LAPACK_COL_MAJOR, to_char(u), n, a, lda, tau); } -inline int64_t ormtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mkl::transpose trans, +inline int64_t ormtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, int64_t m, int64_t n, float *a, int64_t lda, const float *tau, float *c, int64_t ldc) { return LAPACKE_sormtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, a, lda, tau, c, ldc); } -inline int64_t ormtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mkl::transpose trans, +inline int64_t ormtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, int64_t m, int64_t n, double *a, int64_t lda, const double *tau, double *c, int64_t ldc) { return LAPACKE_dormtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, a, lda, tau, c, ldc); } -inline int64_t or_un_mtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mkl::transpose trans, +inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, int64_t m, int64_t n, float *a, int64_t lda, const float *tau, float *c, int64_t ldc) { return LAPACKE_sormtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, a, lda, tau, c, ldc); } -inline int64_t or_un_mtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mkl::transpose trans, +inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, int64_t m, int64_t n, double *a, int64_t lda, const double *tau, double *c, int64_t ldc) { return LAPACKE_dormtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, a, lda, tau, c, ldc); } -inline int64_t or_un_mtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mkl::transpose trans, +inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, int64_t m, int64_t n, std::complex *a, int64_t lda, std::complex *tau, std::complex *c, int64_t ldc) { return LAPACKE_cunmtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, @@ -648,7 +648,7 @@ inline int64_t or_un_mtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mk reinterpret_cast(tau), reinterpret_cast(c), ldc); } -inline int64_t or_un_mtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mkl::transpose trans, +inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, int64_t m, int64_t n, std::complex *a, int64_t lda, std::complex *tau, std::complex *c, int64_t ldc) { return LAPACKE_zunmtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, @@ -657,41 +657,41 @@ inline int64_t or_un_mtr(oneapi::mkl::side side, oneapi::mkl::uplo u, oneapi::mk reinterpret_cast(c), ldc); } -inline int64_t sytrd(oneapi::mkl::uplo u, int64_t n, float *a, int64_t lda, float *d, float *e, +inline int64_t sytrd(oneapi::math::uplo u, int64_t n, float *a, int64_t lda, float *d, float *e, float *tau) { return LAPACKE_ssytrd(LAPACK_COL_MAJOR, to_char(u), n, a, lda, d, e, tau); } -inline int64_t sytrd(oneapi::mkl::uplo u, int64_t n, double *a, int64_t lda, double *d, double *e, +inline int64_t sytrd(oneapi::math::uplo u, int64_t n, double *a, int64_t lda, double *d, double *e, double *tau) { return LAPACKE_dsytrd(LAPACK_COL_MAJOR, to_char(u), n, a, lda, d, e, tau); } -inline int64_t sytrf(oneapi::mkl::uplo u, int64_t n, float *a, int64_t lda, int64_t *ipiv) { +inline int64_t sytrf(oneapi::math::uplo u, int64_t n, float *a, int64_t lda, int64_t *ipiv) { return LAPACKE_ssytrf(LAPACK_COL_MAJOR, to_char(u), n, a, lda, reinterpret_cast(ipiv)); } -inline int64_t sytrf(oneapi::mkl::uplo u, int64_t n, double *a, int64_t lda, int64_t *ipiv) { +inline int64_t sytrf(oneapi::math::uplo u, int64_t n, double *a, int64_t lda, int64_t *ipiv) { return LAPACKE_dsytrf(LAPACK_COL_MAJOR, to_char(u), n, a, lda, reinterpret_cast(ipiv)); } -inline int64_t sytrf(oneapi::mkl::uplo u, int64_t n, std::complex *a, int64_t lda, +inline int64_t sytrf(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, int64_t *ipiv) { return LAPACKE_csytrf(LAPACK_COL_MAJOR, to_char(u), n, reinterpret_cast(a), lda, reinterpret_cast(ipiv)); } -inline int64_t sytrf(oneapi::mkl::uplo u, int64_t n, std::complex *a, int64_t lda, +inline int64_t sytrf(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, int64_t *ipiv) { return LAPACKE_zsytrf(LAPACK_COL_MAJOR, to_char(u), n, reinterpret_cast(a), lda, reinterpret_cast(ipiv)); } -inline void orgbr(oneapi::mkl::generate vect, int64_t m, int64_t n, int64_t k, double *a, +inline void orgbr(oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, double *a, int64_t lda, const double *tau) { LAPACKE_dorgbr(LAPACK_COL_MAJOR, to_char(vect), m, n, k, a, lda, tau); } -inline void orgbr(oneapi::mkl::generate vect, int64_t m, int64_t n, int64_t k, float *a, +inline void orgbr(oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, float *a, int64_t lda, const float *tau) { LAPACKE_sorgbr(LAPACK_COL_MAJOR, to_char(vect), m, n, k, a, lda, tau); } @@ -714,17 +714,17 @@ inline int64_t or_un_gqr(int64_t m, int64_t n, int64_t k, std::complex * lda, reinterpret_cast(tau)); } -inline int64_t or_un_mqr(oneapi::mkl::side s, oneapi::mkl::transpose t, int64_t m, int64_t n, +inline int64_t or_un_mqr(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, int64_t k, const float *a, int64_t lda, const float *tau, float *c, int64_t ldc) { return LAPACKE_sormqr(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, a, lda, tau, c, ldc); } -inline int64_t or_un_mqr(oneapi::mkl::side s, oneapi::mkl::transpose t, int64_t m, int64_t n, +inline int64_t or_un_mqr(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, int64_t k, const double *a, int64_t lda, const double *tau, double *c, int64_t ldc) { return LAPACKE_dormqr(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, a, lda, tau, c, ldc); } -inline int64_t or_un_mqr(oneapi::mkl::side s, oneapi::mkl::transpose t, int64_t m, int64_t n, +inline int64_t or_un_mqr(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, int64_t k, const std::complex *a, int64_t lda, const std::complex *tau, std::complex *c, int64_t ldc) { return LAPACKE_cunmqr(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, @@ -732,7 +732,7 @@ inline int64_t or_un_mqr(oneapi::mkl::side s, oneapi::mkl::transpose t, int64_t reinterpret_cast(tau), reinterpret_cast(c), ldc); } -inline int64_t or_un_mqr(oneapi::mkl::side s, oneapi::mkl::transpose t, int64_t m, int64_t n, +inline int64_t or_un_mqr(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, int64_t k, const std::complex *a, int64_t lda, const std::complex *tau, std::complex *c, int64_t ldc) { return LAPACKE_zunmqr(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, @@ -759,17 +759,17 @@ inline int64_t or_un_grq(int64_t m, int64_t n, int64_t k, std::complex * lda, reinterpret_cast(tau)); } -inline int64_t or_un_mrq(oneapi::mkl::side s, oneapi::mkl::transpose t, int64_t m, int64_t n, +inline int64_t or_un_mrq(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, int64_t k, const float *a, int64_t lda, const float *tau, float *c, int64_t ldc) { return LAPACKE_sormrq(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, a, lda, tau, c, ldc); } -inline int64_t or_un_mrq(oneapi::mkl::side s, oneapi::mkl::transpose t, int64_t m, int64_t n, +inline int64_t or_un_mrq(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, int64_t k, const double *a, int64_t lda, const double *tau, double *c, int64_t ldc) { return LAPACKE_dormrq(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, a, lda, tau, c, ldc); } -inline int64_t or_un_mrq(oneapi::mkl::side s, oneapi::mkl::transpose t, int64_t m, int64_t n, +inline int64_t or_un_mrq(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, int64_t k, const std::complex *a, int64_t lda, const std::complex *tau, std::complex *c, int64_t ldc) { return LAPACKE_cunmrq(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, @@ -777,7 +777,7 @@ inline int64_t or_un_mrq(oneapi::mkl::side s, oneapi::mkl::transpose t, int64_t reinterpret_cast(tau), reinterpret_cast(c), ldc); } -inline int64_t or_un_mrq(oneapi::mkl::side s, oneapi::mkl::transpose t, int64_t m, int64_t n, +inline int64_t or_un_mrq(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, int64_t k, const std::complex *a, int64_t lda, const std::complex *tau, std::complex *c, int64_t ldc) { return LAPACKE_zunmrq(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, @@ -786,39 +786,39 @@ inline int64_t or_un_mrq(oneapi::mkl::side s, oneapi::mkl::transpose t, int64_t reinterpret_cast(c), ldc); } -inline int64_t potrf(oneapi::mkl::uplo upper_lower, int64_t n, std::complex *a, +inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, std::complex *a, int64_t lda) { return LAPACKE_cpotrf(LAPACK_COL_MAJOR, to_char(upper_lower), n, reinterpret_cast(a), lda); } -inline int64_t potrf(oneapi::mkl::uplo upper_lower, int64_t n, double *a, int64_t lda) { +inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, double *a, int64_t lda) { return LAPACKE_dpotrf(LAPACK_COL_MAJOR, to_char(upper_lower), n, a, lda); } -inline int64_t potrf(oneapi::mkl::uplo upper_lower, int64_t n, float *a, int64_t lda) { +inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, float *a, int64_t lda) { return LAPACKE_spotrf(LAPACK_COL_MAJOR, to_char(upper_lower), n, a, lda); } -inline int64_t potrf(oneapi::mkl::uplo upper_lower, int64_t n, std::complex *a, +inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, std::complex *a, int64_t lda) { return LAPACKE_zpotrf(LAPACK_COL_MAJOR, to_char(upper_lower), n, reinterpret_cast(a), lda); } -inline int64_t potrs(oneapi::mkl::uplo upper_lower, int64_t n, int64_t nrhs, +inline int64_t potrs(oneapi::math::uplo upper_lower, int64_t n, int64_t nrhs, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb) { return LAPACKE_cpotrs(LAPACK_COL_MAJOR, to_char(upper_lower), n, nrhs, reinterpret_cast(a), lda, reinterpret_cast(b), ldb); } -inline int64_t potrs(oneapi::mkl::uplo upper_lower, int64_t n, int64_t nrhs, const double *a, +inline int64_t potrs(oneapi::math::uplo upper_lower, int64_t n, int64_t nrhs, const double *a, int64_t lda, double *b, int64_t ldb) { return LAPACKE_dpotrs(LAPACK_COL_MAJOR, to_char(upper_lower), n, nrhs, a, lda, b, ldb); } -inline int64_t potrs(oneapi::mkl::uplo upper_lower, int64_t n, int64_t nrhs, const float *a, +inline int64_t potrs(oneapi::math::uplo upper_lower, int64_t n, int64_t nrhs, const float *a, int64_t lda, float *b, int64_t ldb) { return LAPACKE_spotrs(LAPACK_COL_MAJOR, to_char(upper_lower), n, nrhs, a, lda, b, ldb); } -inline int64_t potrs(oneapi::mkl::uplo upper_lower, int64_t n, int64_t nrhs, +inline int64_t potrs(oneapi::math::uplo upper_lower, int64_t n, int64_t nrhs, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb) { return LAPACKE_zpotrs(LAPACK_COL_MAJOR, to_char(upper_lower), n, nrhs, @@ -826,18 +826,18 @@ inline int64_t potrs(oneapi::mkl::uplo upper_lower, int64_t n, int64_t nrhs, reinterpret_cast(b), ldb); } -inline int64_t potri(oneapi::mkl::uplo upper_lower, int64_t n, std::complex *a, +inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, std::complex *a, int64_t lda) { return LAPACKE_cpotri(LAPACK_COL_MAJOR, to_char(upper_lower), n, reinterpret_cast(a), lda); } -inline int64_t potri(oneapi::mkl::uplo upper_lower, int64_t n, double *a, int64_t lda) { +inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, double *a, int64_t lda) { return LAPACKE_dpotri(LAPACK_COL_MAJOR, to_char(upper_lower), n, a, lda); } -inline int64_t potri(oneapi::mkl::uplo upper_lower, int64_t n, float *a, int64_t lda) { +inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, float *a, int64_t lda) { return LAPACKE_spotri(LAPACK_COL_MAJOR, to_char(upper_lower), n, a, lda); } -inline int64_t potri(oneapi::mkl::uplo upper_lower, int64_t n, std::complex *a, +inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, std::complex *a, int64_t lda) { return LAPACKE_zpotri(LAPACK_COL_MAJOR, to_char(upper_lower), n, reinterpret_cast(a), lda); @@ -864,38 +864,38 @@ inline int64_t laswp(int64_t n, std::complex *a, int64_t lda, int64_t k1 k1, k2, reinterpret_cast(ipiv), incx); } -inline void ungbr(oneapi::mkl::generate vect, int64_t m, int64_t n, int64_t k, +inline void ungbr(oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, std::complex *a, int64_t lda, const std::complex *tau) { LAPACKE_cungbr(LAPACK_COL_MAJOR, to_char(vect), m, n, k, reinterpret_cast(a), lda, reinterpret_cast(tau)); } -inline void ungbr(oneapi::mkl::generate vect, int64_t m, int64_t n, int64_t k, +inline void ungbr(oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, std::complex *a, int64_t lda, const std::complex *tau) { LAPACKE_zungbr(LAPACK_COL_MAJOR, to_char(vect), m, n, k, reinterpret_cast(a), lda, reinterpret_cast(tau)); } -inline int64_t trtrs(oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, int64_t n, int64_t nrhs, const float *a, int64_t lda, float *b, int64_t ldb) { return LAPACKE_strtrs(LAPACK_COL_MAJOR, to_char(uplo), to_char(trans), to_char(diag), n, nrhs, a, lda, b, ldb); } -inline int64_t trtrs(oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, int64_t n, int64_t nrhs, const double *a, int64_t lda, double *b, int64_t ldb) { return LAPACKE_dtrtrs(LAPACK_COL_MAJOR, to_char(uplo), to_char(trans), to_char(diag), n, nrhs, a, lda, b, ldb); } -inline int64_t trtrs(oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, int64_t n, int64_t nrhs, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb) { return LAPACKE_ctrtrs(LAPACK_COL_MAJOR, to_char(uplo), to_char(trans), to_char(diag), n, nrhs, reinterpret_cast(a), lda, reinterpret_cast(b), ldb); } -inline int64_t trtrs(oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, int64_t n, int64_t nrhs, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb) { return LAPACKE_ztrtrs(LAPACK_COL_MAJOR, to_char(uplo), to_char(trans), to_char(diag), n, nrhs, diff --git a/tests/unit_tests/lapack/include/lapack_test_controller.hpp b/tests/unit_tests/lapack/include/lapack_test_controller.hpp index 807698cf8..3e21fa104 100644 --- a/tests/unit_tests/lapack/include/lapack_test_controller.hpp +++ b/tests/unit_tests/lapack/include/lapack_test_controller.hpp @@ -40,31 +40,31 @@ std::istream& operator>>(std::istream& is, T& t) { t = static_cast(i); return is; } -inline std::ostream& operator<<(std::ostream& os, const oneapi::mkl::job& t) { +inline std::ostream& operator<<(std::ostream& os, const oneapi::math::job& t) { os << static_cast(t); return os; } -inline std::ostream& operator<<(std::ostream& os, const oneapi::mkl::jobsvd& t) { +inline std::ostream& operator<<(std::ostream& os, const oneapi::math::jobsvd& t) { os << static_cast(t); return os; } -inline std::ostream& operator<<(std::ostream& os, const oneapi::mkl::transpose& t) { +inline std::ostream& operator<<(std::ostream& os, const oneapi::math::transpose& t) { os << static_cast(t); return os; } -inline std::ostream& operator<<(std::ostream& os, const oneapi::mkl::uplo& t) { +inline std::ostream& operator<<(std::ostream& os, const oneapi::math::uplo& t) { os << static_cast(t); return os; } -inline std::ostream& operator<<(std::ostream& os, const oneapi::mkl::side& t) { +inline std::ostream& operator<<(std::ostream& os, const oneapi::math::side& t) { os << static_cast(t); return os; } -inline std::ostream& operator<<(std::ostream& os, const oneapi::mkl::diag& t) { +inline std::ostream& operator<<(std::ostream& os, const oneapi::math::diag& t) { os << static_cast(t); return os; } -inline std::ostream& operator<<(std::ostream& os, const oneapi::mkl::generate& t) { +inline std::ostream& operator<<(std::ostream& os, const oneapi::math::generate& t) { os << static_cast(t); return os; } @@ -173,10 +173,10 @@ struct InputTestController { try { result = std::apply(tp, tp_args); } - catch (const oneapi::mkl::unsupported_device& e) { + catch (const oneapi::math::unsupported_device& e) { result = result_T{ e, result_T::result::pass }; } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { result = result_T{ e, result_T::result::pass }; } catch (const std::exception& e) { diff --git a/tests/unit_tests/lapack/source/gebrd.cpp b/tests/unit_tests/lapack/source/gebrd.cpp index b54b25da3..650a73145 100644 --- a/tests/unit_tests/lapack/source/gebrd.cpp +++ b/tests/unit_tests/lapack/source/gebrd.cpp @@ -49,7 +49,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 int64_t min_mn = std::min(m, n); std::vector A_initial(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); std::vector A = A_initial; std::vector d(min_mn); @@ -70,11 +70,11 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::gebrd_scratchpad_size(queue, m, n, lda); + oneapi::math::lapack::gebrd_scratchpad_size(queue, m, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::gebrd_scratchpad_size, m, n, lda); + queue, scratchpad_size = oneapi::math::lapack::gebrd_scratchpad_size, m, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -82,10 +82,10 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::gebrd(queue, m, n, A_dev, lda, d_dev, e_dev, tauq_dev, taup_dev, + oneapi::math::lapack::gebrd(queue, m, n, A_dev, lda, d_dev, e_dev, tauq_dev, taup_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::gebrd, m, n, A_dev, lda, d_dev, e_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::gebrd, m, n, A_dev, lda, d_dev, e_dev, tauq_dev, taup_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -122,7 +122,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, int64_t min_mn = std::min(m, n); std::vector A_initial(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); auto A = A_initial; std::vector d(min_mn); @@ -143,11 +143,11 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::gebrd_scratchpad_size(queue, m, n, lda); + oneapi::math::lapack::gebrd_scratchpad_size(queue, m, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::gebrd_scratchpad_size, m, n, lda); + queue, scratchpad_size = oneapi::math::lapack::gebrd_scratchpad_size, m, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -157,12 +157,12 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::gebrd( + sycl::event func_event = oneapi::math::lapack::gebrd( queue, m, n, A_dev, lda, d_dev, e_dev, tauq_dev, taup_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::gebrd, m, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::gebrd, m, n, A_dev, lda, d_dev, e_dev, tauq_dev, taup_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/geqrf.cpp b/tests/unit_tests/lapack/source/geqrf.cpp index 66c1bc8c8..570fb835e 100644 --- a/tests/unit_tests/lapack/source/geqrf.cpp +++ b/tests/unit_tests/lapack/source/geqrf.cpp @@ -49,7 +49,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 /* Initialize */ std::vector A_initial(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); std::vector A = A_initial; std::vector tau(std::min(m, n)); @@ -62,11 +62,11 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::geqrf_scratchpad_size(queue, m, n, lda); + oneapi::math::lapack::geqrf_scratchpad_size(queue, m, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::geqrf_scratchpad_size, m, n, lda); + queue, scratchpad_size = oneapi::math::lapack::geqrf_scratchpad_size, m, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -74,10 +74,10 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::geqrf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::geqrf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::geqrf, m, n, A_dev, lda, tau_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::geqrf, m, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -105,7 +105,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, /* Initialize */ std::vector A_initial(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); std::vector A = A_initial; std::vector tau(std::min(m, n)); @@ -119,11 +119,11 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::geqrf_scratchpad_size(queue, m, n, lda); + oneapi::math::lapack::geqrf_scratchpad_size(queue, m, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::geqrf_scratchpad_size, m, n, lda); + queue, scratchpad_size = oneapi::math::lapack::geqrf_scratchpad_size, m, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -134,11 +134,11 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::geqrf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::geqrf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::geqrf, m, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::geqrf, m, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/geqrf_batch_group.cpp b/tests/unit_tests/lapack/source/geqrf_batch_group.cpp index 78f0388a5..5f6f93b7c 100644 --- a/tests/unit_tests/lapack/source/geqrf_batch_group.cpp +++ b/tests/unit_tests/lapack/source/geqrf_batch_group.cpp @@ -68,7 +68,7 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_initial_list.emplace_back(lda * n); auto& A_initial = A_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); A_list.emplace_back(A_initial); tau_list.emplace_back(std::min(m, n)); @@ -95,12 +95,12 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::geqrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::geqrf_batch_scratchpad_size( queue, m_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::geqrf_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::geqrf_batch_scratchpad_size, m_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -120,11 +120,11 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::geqrf_batch(queue, m_vec.data(), n_vec.data(), A_dev_ptrs, + oneapi::math::lapack::geqrf_batch(queue, m_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::geqrf_batch, m_vec.data(), + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::geqrf_batch, m_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); @@ -205,7 +205,7 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_initial_list.emplace_back(lda * n); auto& A_initial = A_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); A_list.emplace_back(A_initial); tau_list.emplace_back(std::min(m, n)); @@ -233,12 +233,12 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::geqrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::geqrf_batch_scratchpad_size( queue, m_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::geqrf_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::geqrf_batch_scratchpad_size, m_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -260,13 +260,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::geqrf_batch( + sycl::event func_event = oneapi::math::lapack::geqrf_batch( queue, m_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::geqrf_batch, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::geqrf_batch, m_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); diff --git a/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp b/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp index 78dbcec3f..227fdc30f 100644 --- a/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp @@ -50,7 +50,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, int64_ std::vector tau(stride_tau * batch_size); for (int64_t i = 0; i < batch_size; i++) - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda, i * stride_a); std::vector A = A_initial; @@ -61,12 +61,12 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, int64_ auto A_dev = device_alloc(queue, A.size()); auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::geqrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::geqrf_batch_scratchpad_size( queue, m, n, lda, stride_a, stride_tau, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::geqrf_batch_scratchpad_size, m, n, + queue, scratchpad_size = oneapi::math::lapack::geqrf_batch_scratchpad_size, m, n, lda, stride_a, stride_tau, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -75,10 +75,10 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::geqrf_batch(queue, m, n, A_dev, lda, stride_a, tau_dev, stride_tau, + oneapi::math::lapack::geqrf_batch(queue, m, n, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::geqrf_batch, m, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::geqrf_batch, m, n, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size); #endif @@ -122,7 +122,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, std::vector tau(stride_tau * batch_size); for (int64_t i = 0; i < batch_size; i++) - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda, i * stride_a); std::vector A = A_initial; @@ -134,12 +134,12 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, auto A_dev = device_alloc(queue, A.size()); auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::geqrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::geqrf_batch_scratchpad_size( queue, m, n, lda, stride_a, stride_tau, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::geqrf_batch_scratchpad_size, m, n, + queue, scratchpad_size = oneapi::math::lapack::geqrf_batch_scratchpad_size, m, n, lda, stride_a, stride_tau, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -150,12 +150,12 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::geqrf_batch( + sycl::event func_event = oneapi::math::lapack::geqrf_batch( queue, m, n, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::geqrf_batch, m, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::geqrf_batch, m, n, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/gerqf.cpp b/tests/unit_tests/lapack/source/gerqf.cpp index ed46758df..091aa5d68 100644 --- a/tests/unit_tests/lapack/source/gerqf.cpp +++ b/tests/unit_tests/lapack/source/gerqf.cpp @@ -49,7 +49,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 /* Initialize */ std::vector A_initial(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); std::vector A = A_initial; std::vector tau(std::min(m, n)); @@ -62,11 +62,11 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::gerqf_scratchpad_size(queue, m, n, lda); + oneapi::math::lapack::gerqf_scratchpad_size(queue, m, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::gerqf_scratchpad_size, m, n, lda); + queue, scratchpad_size = oneapi::math::lapack::gerqf_scratchpad_size, m, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -74,10 +74,10 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::gerqf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::gerqf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::gerqf, m, n, A_dev, lda, tau_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::gerqf, m, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -105,7 +105,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, /* Initialize */ std::vector A_initial(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); std::vector A = A_initial; std::vector tau(std::min(m, n)); @@ -119,11 +119,11 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::gerqf_scratchpad_size(queue, m, n, lda); + oneapi::math::lapack::gerqf_scratchpad_size(queue, m, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::gerqf_scratchpad_size, m, n, lda); + queue, scratchpad_size = oneapi::math::lapack::gerqf_scratchpad_size, m, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -134,11 +134,11 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::gerqf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::gerqf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::gerqf, m, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::gerqf, m, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/gesvd.cpp b/tests/unit_tests/lapack/source/gesvd.cpp index 56f590605..2d793c1e8 100644 --- a/tests/unit_tests/lapack/source/gesvd.cpp +++ b/tests/unit_tests/lapack/source/gesvd.cpp @@ -41,7 +41,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +bool accuracy(const sycl::device& dev, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, int64_t m, int64_t n, int64_t lda, int64_t ldu, int64_t ldvt, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -49,10 +49,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::mkl::jo /* Initialize */ int64_t min_mn = std::min(m, n); int64_t ucols = min_mn; - if (jobu == oneapi::mkl::jobsvd::vectors) + if (jobu == oneapi::math::jobsvd::vectors) ucols = m; int64_t vtrows = min_mn; - if (jobvt == oneapi::mkl::jobsvd::vectors) + if (jobvt == oneapi::math::jobsvd::vectors) vtrows = n; std::vector A(lda * n); @@ -60,7 +60,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::mkl::jo std::vector Vt(ldvt * n); std::vector s(min_mn); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda); std::vector A_initial = A; /* Compute on device */ @@ -71,12 +71,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::mkl::jo auto Vt_dev = device_alloc(queue, Vt.size()); auto s_dev = device_alloc(queue, s.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::gesvd_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::gesvd_scratchpad_size( queue, jobu, jobvt, m, n, lda, ldu, ldvt); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::gesvd_scratchpad_size, + scratchpad_size = oneapi::math::lapack::gesvd_scratchpad_size, jobu, jobvt, m, n, lda, ldu, ldvt); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -85,10 +85,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::mkl::jo queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::gesvd(queue, jobu, jobvt, m, n, A_dev, lda, s_dev, U_dev, ldu, Vt_dev, + oneapi::math::lapack::gesvd(queue, jobu, jobvt, m, n, A_dev, lda, s_dev, U_dev, ldu, Vt_dev, ldvt, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::gesvd, jobu, jobvt, m, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::gesvd, jobu, jobvt, m, n, A_dev, lda, s_dev, U_dev, ldu, Vt_dev, ldvt, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -107,7 +107,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::mkl::jo } bool result = true; - if (jobu == oneapi::mkl::jobsvd::vectors && jobvt == oneapi::mkl::jobsvd::vectors) { + if (jobu == oneapi::math::jobsvd::vectors && jobvt == oneapi::math::jobsvd::vectors) { /* |A - U S V'| < |A| O(eps) */ std::vector US(m * n); int64_t ldus = m; @@ -116,7 +116,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::mkl::jo US[row + col * ldus] = U[row + col * ldu] * s[col]; std::vector USV(m * n); int64_t ldusv = m; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, m, n, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, m, n, n, 1.0, US.data(), ldus, Vt.data(), ldvt, 0.0, USV.data(), ldusv); if (!rel_mat_err_check(m, n, A_initial, lda, USV, ldusv)) { test_log::lout << "Factorization check failed" << std::endl; @@ -124,17 +124,17 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::mkl::jo } } - if (jobu == oneapi::mkl::jobsvd::vectorsina) + if (jobu == oneapi::math::jobsvd::vectorsina) reference::lacpy('A', m, ucols, A.data(), lda, U.data(), ldu); - if (jobvt == oneapi::mkl::jobsvd::vectorsina) + if (jobvt == oneapi::math::jobsvd::vectorsina) reference::lacpy('A', vtrows, n, A.data(), lda, Vt.data(), ldvt); - if (jobu == oneapi::mkl::jobsvd::vectors || jobu == oneapi::mkl::jobsvd::somevec || - jobu == oneapi::mkl::jobsvd::vectorsina) { + if (jobu == oneapi::math::jobsvd::vectors || jobu == oneapi::math::jobsvd::somevec || + jobu == oneapi::math::jobsvd::vectorsina) { /* |I - U' U| < n O(eps) */ std::vector UU(ucols * ucols); int64_t lduu = ucols; - reference::gemm(oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, ucols, + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, ucols, ucols, m, 1.0, U.data(), ldu, U.data(), ldu, 0.0, UU.data(), lduu); if (!rel_id_err_check(ucols, UU, lduu)) { test_log::lout << "U Orthogonality check failed" << std::endl; @@ -142,12 +142,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::mkl::jo } } - if (jobvt == oneapi::mkl::jobsvd::vectors || jobvt == oneapi::mkl::jobsvd::somevec || - jobvt == oneapi::mkl::jobsvd::vectorsina) { + if (jobvt == oneapi::math::jobsvd::vectors || jobvt == oneapi::math::jobsvd::somevec || + jobvt == oneapi::math::jobsvd::vectorsina) { /* |I - V' V| < n O(eps) */ std::vector VV(vtrows * vtrows); int64_t ldvv = vtrows; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, vtrows, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, vtrows, vtrows, n, 1.0, Vt.data(), ldvt, Vt.data(), ldvt, 0.0, VV.data(), ldvv); if (!rel_id_err_check(vtrows, VV, ldvv)) { test_log::lout << "V Orthogonality check failed" << std::endl; @@ -162,7 +162,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +bool usm_dependency(const sycl::device& dev, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, int64_t m, int64_t n, int64_t lda, int64_t ldu, int64_t ldvt, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -170,10 +170,10 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::m /* Initialize */ int64_t min_mn = std::min(m, n); int64_t ucols = min_mn; - if (jobu == oneapi::mkl::jobsvd::vectors) + if (jobu == oneapi::math::jobsvd::vectors) ucols = m; int64_t vtrows = min_mn; - if (jobvt == oneapi::mkl::jobsvd::vectors) + if (jobvt == oneapi::math::jobsvd::vectors) vtrows = n; std::vector A(lda * n); @@ -181,7 +181,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::m std::vector Vt(ldvt * n); std::vector s(min_mn); - rand_matrix_diag_dom(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda); + rand_matrix_diag_dom(seed, oneapi::math::transpose::nontrans, m, n, A, lda); std::vector A_initial = A; /* Compute on device */ @@ -193,12 +193,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::m auto Vt_dev = device_alloc(queue, Vt.size()); auto s_dev = device_alloc(queue, s.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::gesvd_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::gesvd_scratchpad_size( queue, jobu, jobvt, m, n, lda, ldu, ldvt); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::gesvd_scratchpad_size, + scratchpad_size = oneapi::math::lapack::gesvd_scratchpad_size, jobu, jobvt, m, n, lda, ldu, ldvt); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -209,12 +209,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::jobsvd jobu, oneapi::m /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::gesvd( + sycl::event func_event = oneapi::math::lapack::gesvd( queue, jobu, jobvt, m, n, A_dev, lda, s_dev, U_dev, ldu, Vt_dev, ldvt, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::gesvd, jobu, jobvt, m, n, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::gesvd, jobu, jobvt, m, n, A_dev, lda, s_dev, U_dev, ldu, Vt_dev, ldvt, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/getrf.cpp b/tests/unit_tests/lapack/source/getrf.cpp index 159d01e2d..165c7fed5 100644 --- a/tests/unit_tests/lapack/source/getrf.cpp +++ b/tests/unit_tests/lapack/source/getrf.cpp @@ -52,7 +52,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 /* Initialize */ std::vector A_initial(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); std::vector A = A_initial; std::vector ipiv(std::min(m, n)); @@ -65,11 +65,11 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::getrf_scratchpad_size(queue, m, n, lda); + oneapi::math::lapack::getrf_scratchpad_size(queue, m, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getrf_scratchpad_size, m, n, lda); + queue, scratchpad_size = oneapi::math::lapack::getrf_scratchpad_size, m, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -77,10 +77,10 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::getrf(queue, m, n, A_dev, lda, ipiv_dev, scratchpad_dev, + oneapi::math::lapack::getrf(queue, m, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::getrf, m, n, A_dev, lda, ipiv_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrf, m, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -108,7 +108,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, /* Initialize */ std::vector A_initial(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); std::vector A = A_initial; std::vector ipiv(std::min(m, n)); @@ -122,11 +122,11 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::getrf_scratchpad_size(queue, m, n, lda); + oneapi::math::lapack::getrf_scratchpad_size(queue, m, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getrf_scratchpad_size, m, n, lda); + queue, scratchpad_size = oneapi::math::lapack::getrf_scratchpad_size, m, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -137,11 +137,11 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::getrf(queue, m, n, A_dev, lda, ipiv_dev, scratchpad_dev, + oneapi::math::lapack::getrf(queue, m, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::getrf, m, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getrf, m, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/getrf_batch_group.cpp b/tests/unit_tests/lapack/source/getrf_batch_group.cpp index d16840ae5..0641698df 100644 --- a/tests/unit_tests/lapack/source/getrf_batch_group.cpp +++ b/tests/unit_tests/lapack/source/getrf_batch_group.cpp @@ -68,7 +68,7 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_initial_list.emplace_back(lda * n); auto& A_initial = A_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); A_list.emplace_back(A_initial); ipiv_list.emplace_back(std::min(m, n)); @@ -99,12 +99,12 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getrf_batch_scratchpad_size( queue, m_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getrf_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::getrf_batch_scratchpad_size, m_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -124,11 +124,11 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::getrf_batch(queue, m_vec.data(), n_vec.data(), A_dev_ptrs, + oneapi::math::lapack::getrf_batch(queue, m_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::getrf_batch, m_vec.data(), + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrf_batch, m_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); @@ -210,7 +210,7 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_initial_list.emplace_back(lda * n); auto& A_initial = A_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda); A_list.emplace_back(A_initial); ipiv_list.emplace_back(std::min(m, n)); @@ -242,12 +242,12 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getrf_batch_scratchpad_size( queue, m_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getrf_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::getrf_batch_scratchpad_size, m_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -269,14 +269,14 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::getrf_batch( + sycl::event func_event = oneapi::math::lapack::getrf_batch( queue, m_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT( - queue, func_event = oneapi::mkl::lapack::getrf_batch, m_vec.data(), n_vec.data(), + queue, func_event = oneapi::math::lapack::getrf_batch, m_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/getrf_batch_stride.cpp b/tests/unit_tests/lapack/source/getrf_batch_stride.cpp index 67f21654b..62c55d5fb 100644 --- a/tests/unit_tests/lapack/source/getrf_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/getrf_batch_stride.cpp @@ -48,7 +48,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, int64_ /* Initialize */ std::vector A_initial(stride_a * batch_size); for (int64_t i = 0; i < batch_size; i++) - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda, i * stride_a); std::vector A = A_initial; int64_t stride_ipiv = std::min(m, n); @@ -61,12 +61,12 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, int64_ auto A_dev = device_alloc(queue, A.size()); auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getrf_batch_scratchpad_size( queue, m, n, lda, stride_a, stride_ipiv, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getrf_batch_scratchpad_size, m, n, + queue, scratchpad_size = oneapi::math::lapack::getrf_batch_scratchpad_size, m, n, lda, stride_a, stride_ipiv, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -75,10 +75,10 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::getrf_batch(queue, m, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, + oneapi::math::lapack::getrf_batch(queue, m, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, batch_size, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::getrf_batch, m, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrf_batch, m, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, batch_size, scratchpad_dev, scratchpad_size); #endif @@ -120,7 +120,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, /* Initialize */ std::vector A_initial(stride_a * batch_size); for (int64_t i = 0; i < batch_size; i++) - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A_initial, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A_initial, lda, i * stride_a); std::vector A = A_initial; int64_t stride_ipiv = std::min(m, n); @@ -134,12 +134,12 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, auto A_dev = device_alloc(queue, A.size()); auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getrf_batch_scratchpad_size( queue, m, n, lda, stride_a, stride_ipiv, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getrf_batch_scratchpad_size, m, n, + queue, scratchpad_size = oneapi::math::lapack::getrf_batch_scratchpad_size, m, n, lda, stride_a, stride_ipiv, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -150,12 +150,12 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::getrf_batch( + sycl::event func_event = oneapi::math::lapack::getrf_batch( queue, m, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::getrf_batch, m, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getrf_batch, m, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/getri.cpp b/tests/unit_tests/lapack/source/getri.cpp index 9037d33c8..24d272520 100644 --- a/tests/unit_tests/lapack/source/getri.cpp +++ b/tests/unit_tests/lapack/source/getri.cpp @@ -52,7 +52,7 @@ bool accuracy(const sycl::device& dev, int64_t n, int64_t lda, uint64_t seed) { /* Initialize */ std::vector A_initial(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda); std::vector A = A_initial; std::vector ipiv(n); @@ -70,11 +70,11 @@ bool accuracy(const sycl::device& dev, int64_t n, int64_t lda, uint64_t seed) { auto A_dev = device_alloc(queue, A.size()); auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getri_scratchpad_size(queue, n, lda); + const auto scratchpad_size = oneapi::math::lapack::getri_scratchpad_size(queue, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getri_scratchpad_size, n, lda); + queue, scratchpad_size = oneapi::math::lapack::getri_scratchpad_size, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -83,9 +83,9 @@ bool accuracy(const sycl::device& dev, int64_t n, int64_t lda, uint64_t seed) { queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::getri(queue, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); + oneapi::math::lapack::getri(queue, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::getri, n, A_dev, lda, ipiv_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getri, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -112,7 +112,7 @@ bool usm_dependency(const sycl::device& dev, int64_t n, int64_t lda, uint64_t se /* Initialize */ std::vector A_initial(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda); std::vector A = A_initial; std::vector ipiv(n); @@ -132,11 +132,11 @@ bool usm_dependency(const sycl::device& dev, int64_t n, int64_t lda, uint64_t se auto A_dev = device_alloc(queue, A.size()); auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getri_scratchpad_size(queue, n, lda); + const auto scratchpad_size = oneapi::math::lapack::getri_scratchpad_size(queue, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getri_scratchpad_size, n, lda); + queue, scratchpad_size = oneapi::math::lapack::getri_scratchpad_size, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -148,11 +148,11 @@ bool usm_dependency(const sycl::device& dev, int64_t n, int64_t lda, uint64_t se auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::getri(queue, n, A_dev, lda, ipiv_dev, scratchpad_dev, + oneapi::math::lapack::getri(queue, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::getri, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getri, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/getri_batch_group.cpp b/tests/unit_tests/lapack/source/getri_batch_group.cpp index 6de30e512..d700a37f4 100644 --- a/tests/unit_tests/lapack/source/getri_batch_group.cpp +++ b/tests/unit_tests/lapack/source/getri_batch_group.cpp @@ -66,7 +66,7 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_initial_list.emplace_back(lda * n); auto& A_initial = A_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda); A_list.emplace_back(A_initial); auto& A = A_list.back(); @@ -106,12 +106,12 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getri_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getri_batch_scratchpad_size( queue, n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getri_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::getri_batch_scratchpad_size, n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -134,11 +134,11 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::getri_batch(queue, n_vec.data(), A_dev_ptrs, lda_vec.data(), + oneapi::math::lapack::getri_batch(queue, n_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::getri_batch, n_vec.data(), A_dev_ptrs, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getri_batch, n_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #endif @@ -213,7 +213,7 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_initial_list.emplace_back(lda * n); auto& A_initial = A_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda); A_list.emplace_back(A_initial); auto& A = A_list.back(); @@ -254,12 +254,12 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getri_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getri_batch_scratchpad_size( queue, n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getri_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::getri_batch_scratchpad_size, n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -284,13 +284,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::getri_batch( + sycl::event func_event = oneapi::math::lapack::getri_batch( queue, n_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::getri_batch, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getri_batch, n_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); diff --git a/tests/unit_tests/lapack/source/getri_batch_stride.cpp b/tests/unit_tests/lapack/source/getri_batch_stride.cpp index 22f709785..aee87f7b9 100644 --- a/tests/unit_tests/lapack/source/getri_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/getri_batch_stride.cpp @@ -47,7 +47,7 @@ bool accuracy(const sycl::device& dev, int64_t n, int64_t lda, int64_t stride_a, std::vector A_initial(stride_a * batch_size); std::vector ipiv(stride_ipiv * batch_size); for (int64_t i = 0; i < batch_size; i++) - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda, i * stride_a); std::vector A = A_initial; @@ -68,12 +68,12 @@ bool accuracy(const sycl::device& dev, int64_t n, int64_t lda, int64_t stride_a, auto A_dev = device_alloc(queue, A.size()); auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getri_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getri_batch_scratchpad_size( queue, n, lda, stride_a, stride_ipiv, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getri_batch_scratchpad_size, n, lda, + queue, scratchpad_size = oneapi::math::lapack::getri_batch_scratchpad_size, n, lda, stride_a, stride_ipiv, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -83,10 +83,10 @@ bool accuracy(const sycl::device& dev, int64_t n, int64_t lda, int64_t stride_a, queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::getri_batch(queue, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, + oneapi::math::lapack::getri_batch(queue, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, batch_size, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::getri_batch, n, A_dev, lda, stride_a, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getri_batch, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, batch_size, scratchpad_dev, scratchpad_size); #endif @@ -126,7 +126,7 @@ bool usm_dependency(const sycl::device& dev, int64_t n, int64_t lda, int64_t str std::vector A_initial(stride_a * batch_size); std::vector ipiv(stride_ipiv * batch_size); for (int64_t i = 0; i < batch_size; i++) - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda, i * stride_a); std::vector A = A_initial; @@ -148,12 +148,12 @@ bool usm_dependency(const sycl::device& dev, int64_t n, int64_t lda, int64_t str auto A_dev = device_alloc(queue, A.size()); auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getri_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getri_batch_scratchpad_size( queue, n, lda, stride_a, stride_ipiv, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getri_batch_scratchpad_size, n, lda, + queue, scratchpad_size = oneapi::math::lapack::getri_batch_scratchpad_size, n, lda, stride_a, stride_ipiv, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -165,12 +165,12 @@ bool usm_dependency(const sycl::device& dev, int64_t n, int64_t lda, int64_t str /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::getri_batch( + sycl::event func_event = oneapi::math::lapack::getri_batch( queue, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::getri_batch, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getri_batch, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/getrs.cpp b/tests/unit_tests/lapack/source/getrs.cpp index 767e9248d..4532e3dd9 100644 --- a/tests/unit_tests/lapack/source/getrs.cpp +++ b/tests/unit_tests/lapack/source/getrs.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::transpose trans, int64_t n, int64_t nrhs, +bool accuracy(const sycl::device& dev, oneapi::math::transpose trans, int64_t n, int64_t nrhs, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -48,8 +48,8 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::transpose trans, int64_t n, /* Initialize */ std::vector A_initial(lda * n); std::vector B_initial(ldb * nrhs); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B_initial, ldb); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B_initial, ldb); std::vector A = A_initial; std::vector B = B_initial; @@ -70,11 +70,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::transpose trans, int64_t n, auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::getrs_scratchpad_size(queue, trans, n, nrhs, lda, ldb); + oneapi::math::lapack::getrs_scratchpad_size(queue, trans, n, nrhs, lda, ldb); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::getrs_scratchpad_size, + scratchpad_size = oneapi::math::lapack::getrs_scratchpad_size, trans, n, nrhs, lda, ldb); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -85,10 +85,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::transpose trans, int64_t n, queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::getrs(queue, trans, n, nrhs, A_dev, lda, ipiv_dev, B_dev, ldb, + oneapi::math::lapack::getrs(queue, trans, n, nrhs, A_dev, lda, ipiv_dev, B_dev, ldb, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::getrs, trans, n, nrhs, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrs, trans, n, nrhs, A_dev, lda, ipiv_dev, B_dev, ldb, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -110,7 +110,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::transpose trans, int64_t n, int64_t nrhs, +bool usm_dependency(const sycl::device& dev, oneapi::math::transpose trans, int64_t n, int64_t nrhs, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -118,8 +118,8 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::transpose trans, int64 /* Initialize */ std::vector A_initial(lda * n); std::vector B_initial(ldb * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B_initial, ldb); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B_initial, ldb); std::vector A = A_initial; std::vector B = B_initial; @@ -142,11 +142,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::transpose trans, int64 auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::getrs_scratchpad_size(queue, trans, n, nrhs, lda, ldb); + oneapi::math::lapack::getrs_scratchpad_size(queue, trans, n, nrhs, lda, ldb); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::getrs_scratchpad_size, + scratchpad_size = oneapi::math::lapack::getrs_scratchpad_size, trans, n, nrhs, lda, ldb); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -159,12 +159,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::transpose trans, int64 /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::getrs( + sycl::event func_event = oneapi::math::lapack::getrs( queue, trans, n, nrhs, A_dev, lda, ipiv_dev, B_dev, ldb, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::getrs, trans, n, nrhs, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getrs, trans, n, nrhs, A_dev, lda, ipiv_dev, B_dev, ldb, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/getrs_batch_group.cpp b/tests/unit_tests/lapack/source/getrs_batch_group.cpp index 5149738d4..d2cd80e9c 100644 --- a/tests/unit_tests/lapack/source/getrs_batch_group.cpp +++ b/tests/unit_tests/lapack/source/getrs_batch_group.cpp @@ -46,8 +46,8 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { using fp_real = typename complex_info::real_type; /* Test Parameters */ - std::vector trans_vec = { oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans }; + std::vector trans_vec = { oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans }; std::vector n_vec = { 4, 5 }; std::vector nrhs_vec = { 9, 6 }; std::vector lda_vec = { 6, 6 }; @@ -75,14 +75,14 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_initial_list.emplace_back(lda * n); auto& A_initial = A_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda); A_list.emplace_back(A_initial); auto& A = A_list.back(); B_initial_list.emplace_back(ldb * nrhs); auto& B_initial = B_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B_initial, lda); B_list.emplace_back(B_initial); auto& B = B_list.back(); @@ -127,13 +127,13 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getrs_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size( queue, trans_vec.data(), n_vec.data(), nrhs_vec.data(), lda_vec.data(), ldb_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getrs_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size, trans_vec.data(), n_vec.data(), nrhs_vec.data(), lda_vec.data(), ldb_vec.data(), group_count, group_sizes_vec.data()); #endif @@ -162,12 +162,12 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::getrs_batch(queue, trans_vec.data(), n_vec.data(), nrhs_vec.data(), + oneapi::math::lapack::getrs_batch(queue, trans_vec.data(), n_vec.data(), nrhs_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, B_dev_ptrs, ldb_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::getrs_batch, trans_vec.data(), + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrs_batch, trans_vec.data(), n_vec.data(), nrhs_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, B_dev_ptrs, ldb_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); @@ -232,7 +232,7 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { using fp_real = typename complex_info::real_type; /* Test Parameters */ - std::vector trans_vec = { oneapi::mkl::transpose::nontrans }; + std::vector trans_vec = { oneapi::math::transpose::nontrans }; std::vector n_vec = { 1 }; std::vector nrhs_vec = { 1 }; std::vector lda_vec = { 1 }; @@ -260,14 +260,14 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_initial_list.emplace_back(lda * n); auto& A_initial = A_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda); A_list.emplace_back(A_initial); auto& A = A_list.back(); B_initial_list.emplace_back(ldb * nrhs); auto& B_initial = B_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B_initial, lda); B_list.emplace_back(B_initial); auto& B = B_list.back(); @@ -313,13 +313,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getrs_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size( queue, trans_vec.data(), n_vec.data(), nrhs_vec.data(), lda_vec.data(), ldb_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getrs_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size, trans_vec.data(), n_vec.data(), nrhs_vec.data(), lda_vec.data(), ldb_vec.data(), group_count, group_sizes_vec.data()); #endif @@ -350,13 +350,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::getrs_batch( + sycl::event func_event = oneapi::math::lapack::getrs_batch( queue, trans_vec.data(), n_vec.data(), nrhs_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, B_dev_ptrs, ldb_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::getrs_batch, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getrs_batch, trans_vec.data(), n_vec.data(), nrhs_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, B_dev_ptrs, ldb_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, diff --git a/tests/unit_tests/lapack/source/getrs_batch_stride.cpp b/tests/unit_tests/lapack/source/getrs_batch_stride.cpp index bd7c8f281..ff159416f 100644 --- a/tests/unit_tests/lapack/source/getrs_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/getrs_batch_stride.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::transpose trans, int64_t n, int64_t nrhs, +bool accuracy(const sycl::device& dev, oneapi::math::transpose trans, int64_t n, int64_t nrhs, int64_t lda, int64_t stride_a, int64_t stride_ipiv, int64_t ldb, int64_t stride_b, int64_t batch_size, uint64_t seed) { using fp = typename data_T_info::value_type; @@ -49,8 +49,8 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::transpose trans, int64_t n, std::vector B_initial(stride_b * batch_size); std::vector ipiv(stride_ipiv * batch_size); for (int64_t i = 0; i < batch_size; i++) { - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda, i * stride_a); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B_initial, ldb, i * stride_b); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B_initial, ldb, i * stride_b); } std::vector A = A_initial; @@ -74,12 +74,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::transpose trans, int64_t n, auto B_dev = device_alloc(queue, B.size()); auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getrs_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size( queue, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getrs_batch_scratchpad_size, trans, n, + queue, scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -90,11 +90,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::transpose trans, int64_t n, queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, A_dev, lda, stride_a, ipiv_dev, + oneapi::math::lapack::getrs_batch(queue, trans, n, nrhs, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, B_dev, ldb, stride_b, batch_size, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::getrs_batch, trans, n, nrhs, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrs_batch, trans, n, nrhs, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, B_dev, ldb, stride_b, batch_size, scratchpad_dev, scratchpad_size); #endif @@ -128,7 +128,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::transpose trans, int64_t n, int64_t nrhs, +bool usm_dependency(const sycl::device& dev, oneapi::math::transpose trans, int64_t n, int64_t nrhs, int64_t lda, int64_t stride_a, int64_t stride_ipiv, int64_t ldb, int64_t stride_b, int64_t batch_size, uint64_t seed) { using fp = typename data_T_info::value_type; @@ -137,8 +137,8 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::transpose trans, int64 std::vector B_initial(stride_b * batch_size); std::vector ipiv(stride_ipiv * batch_size); for (auto i = 0; i < batch_size; ++i) { - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A_initial, lda, i * stride_a); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, nrhs, n, B_initial, ldb, i * stride_b); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A_initial, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, nrhs, n, B_initial, ldb, i * stride_b); } std::vector A = A_initial; @@ -163,12 +163,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::transpose trans, int64 auto B_dev = device_alloc(queue, B.size()); auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::getrs_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size( queue, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::getrs_batch_scratchpad_size, trans, n, + queue, scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -181,13 +181,13 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::transpose trans, int64 /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::getrs_batch( + sycl::event func_event = oneapi::math::lapack::getrs_batch( queue, trans, n, nrhs, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, B_dev, ldb, stride_b, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::getrs_batch, trans, n, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getrs_batch, trans, n, nrhs, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, B_dev, ldb, stride_b, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); diff --git a/tests/unit_tests/lapack/source/heevd.cpp b/tests/unit_tests/lapack/source/heevd.cpp index 34f2c15d1..b2a4c906b 100644 --- a/tests/unit_tests/lapack/source/heevd.cpp +++ b/tests/unit_tests/lapack/source/heevd.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, int64_t n, +bool accuracy(const sycl::device& dev, oneapi::math::job jobz, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -59,11 +59,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl::uplo auto w_dev = device_alloc(queue, w.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::heevd_scratchpad_size(queue, jobz, uplo, n, lda); + oneapi::math::lapack::heevd_scratchpad_size(queue, jobz, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::heevd_scratchpad_size, + scratchpad_size = oneapi::math::lapack::heevd_scratchpad_size, jobz, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -72,10 +72,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl::uplo queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::heevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, + oneapi::math::lapack::heevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::heevd, jobz, uplo, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::heevd, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -97,7 +97,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +bool usm_dependency(const sycl::device& dev, oneapi::math::job jobz, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -117,11 +117,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl: auto w_dev = device_alloc(queue, w.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::heevd_scratchpad_size(queue, jobz, uplo, n, lda); + oneapi::math::lapack::heevd_scratchpad_size(queue, jobz, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::heevd_scratchpad_size, + scratchpad_size = oneapi::math::lapack::heevd_scratchpad_size, jobz, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -133,11 +133,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl: auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::heevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, + oneapi::math::lapack::heevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::heevd, jobz, uplo, n, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::heevd, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/hegvd.cpp b/tests/unit_tests/lapack/source/hegvd.cpp index 96585cc19..a072cd410 100644 --- a/tests/unit_tests/lapack/source/hegvd.cpp +++ b/tests/unit_tests/lapack/source/hegvd.cpp @@ -42,7 +42,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, int64_t n, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -65,11 +65,11 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one auto w_dev = device_alloc(queue, w.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::hegvd_scratchpad_size(queue, itype, jobz, uplo, n, lda, ldb); + oneapi::math::lapack::hegvd_scratchpad_size(queue, itype, jobz, uplo, n, lda, ldb); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::hegvd_scratchpad_size, + scratchpad_size = oneapi::math::lapack::hegvd_scratchpad_size, itype, jobz, uplo, n, lda, ldb); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -79,10 +79,10 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::hegvd(queue, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, + oneapi::math::lapack::hegvd(queue, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::hegvd, itype, jobz, uplo, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::hegvd, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -107,24 +107,24 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* |D_ref - D| < |D_ref| O(eps) */ std::vector D_ref(n); - reference::hegvd(itype, oneapi::mkl::job::novec, uplo, n, std::vector(A_initial).data(), + reference::hegvd(itype, oneapi::math::job::novec, uplo, n, std::vector(A_initial).data(), lda, std::vector(B_initial).data(), ldb, D_ref.data()); if (!rel_vec_err_check(n, D_ref, D, 10.0)) { test_log::lout << "Eigenvalue check failed" << std::endl; result = false; } - if (oneapi::mkl::job::vec == jobz) { + if (oneapi::math::job::vec == jobz) { if (itype == 1) { /* |A Z - B Z D| < |A Z| O(eps) */ std::vector AZ(n * n); int64_t ldaz = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, A_initial.data(), lda, Z.data(), ldz, 0.0, AZ.data(), ldaz); std::vector BZ(n * n); int64_t ldbz = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, B_initial.data(), ldb, Z.data(), ldz, 0.0, BZ.data(), ldbz); std::vector BZD(n * n); @@ -141,7 +141,7 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* |I - Z' B Z| < n O(eps) */ std::vector ZBZ(n * n); int64_t ldzbz = n; - reference::gemm(oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); if (!rel_id_err_check(n, ZBZ, ldzbz)) { test_log::lout << "Orthogonality check failed" << std::endl; @@ -152,12 +152,12 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* |A B Z - Z D| < |A B Z| O(eps) */ std::vector BZ(n * n); int64_t ldbz = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, B_initial.data(), ldb, Z.data(), ldz, 0.0, BZ.data(), ldbz); std::vector ABZ(n * n); int64_t ldabz = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, A_initial.data(), lda, BZ.data(), ldbz, 0.0, ABZ.data(), ldabz); @@ -175,7 +175,7 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* |I - Z' B Z| < n O(eps) */ std::vector ZBZ(n * n); int64_t ldzbz = n; - reference::gemm(oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); if (!rel_id_err_check(n, ZBZ, ldzbz)) { test_log::lout << "Orthogonality check failed" << std::endl; @@ -187,7 +187,7 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* C = B^-1 Z */ std::vector AZ(n * n); int64_t ldaz = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, A_initial.data(), lda, Z.data(), ldz, 0.0, AZ.data(), ldaz); std::vector C(n * n); @@ -213,7 +213,7 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* |I - Z' B^-1 Z| = |I - Z' C| < n O(eps) */ std::vector ZhC(n * n); int64_t ldzhc = n; - reference::gemm(oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, Z.data(), ldz, C.data(), ldc, 0.0, ZhC.data(), ldzhc); if (!rel_id_err_check(n, ZhC, ldzhc)) { test_log::lout << "Orthogonality check failed" << std::endl; @@ -229,8 +229,8 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, int64_t n, int64_t lda, int64_t ldb, uint64_t seed) { +bool usm_dependency(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, int64_t n, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -253,11 +253,11 @@ bool usm_dependency(const sycl::device& dev, int64_t itype, oneapi::mkl::job job auto w_dev = device_alloc(queue, w.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::hegvd_scratchpad_size(queue, itype, jobz, uplo, n, lda, ldb); + oneapi::math::lapack::hegvd_scratchpad_size(queue, itype, jobz, uplo, n, lda, ldb); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::hegvd_scratchpad_size, + scratchpad_size = oneapi::math::lapack::hegvd_scratchpad_size, itype, jobz, uplo, n, lda, ldb); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -269,12 +269,12 @@ bool usm_dependency(const sycl::device& dev, int64_t itype, oneapi::mkl::job job /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::hegvd( + sycl::event func_event = oneapi::math::lapack::hegvd( queue, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::hegvd, itype, jobz, uplo, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::hegvd, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/hetrd.cpp b/tests/unit_tests/lapack/source/hetrd.cpp index d6ccca707..4a5bb0345 100644 --- a/tests/unit_tests/lapack/source/hetrd.cpp +++ b/tests/unit_tests/lapack/source/hetrd.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -50,7 +50,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ std::vector d(n); std::vector e(n); std::vector tau(n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); std::vector A_initial = A; /* Compute on device */ @@ -63,11 +63,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::hetrd_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::hetrd_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::hetrd_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::hetrd_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -78,10 +78,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::hetrd(queue, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, + oneapi::math::lapack::hetrd(queue, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::hetrd, uplo, n, A_dev, lda, d_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::hetrd, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -114,9 +114,9 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ std::vector QTQ{ T }; int64_t ldqtq = n; - reference::or_un_mtr(oneapi::mkl::side::left, uplo, oneapi::mkl::transpose::nontrans, n, n, + reference::or_un_mtr(oneapi::math::side::left, uplo, oneapi::math::transpose::nontrans, n, n, A.data(), lda, tau.data(), QTQ.data(), ldqtq); - reference::or_un_mtr(oneapi::mkl::side::right, uplo, oneapi::mkl::transpose::conjtrans, n, n, + reference::or_un_mtr(oneapi::math::side::right, uplo, oneapi::math::transpose::conjtrans, n, n, A.data(), lda, tau.data(), QTQ.data(), ldqtq); if (!rel_mat_err_check(n, n, QTQ, ldqtq, A_initial, lda)) { @@ -127,7 +127,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ /* A[i, i] = d[i] */ for (int64_t diag = 0; diag < n; diag++) d[diag] -= A[diag + diag * lda].real(); - if (uplo == oneapi::mkl::uplo::upper) + if (uplo == oneapi::math::uplo::upper) for (int64_t diag = 0; diag < n - 1; diag++) e[diag] -= A[diag + (diag + 1) * lda].real(); else @@ -152,7 +152,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -162,7 +162,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, std::vector d(n); std::vector e(n); std::vector tau(n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); std::vector A_initial = A; /* Compute on device */ @@ -176,11 +176,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::hetrd_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::hetrd_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::hetrd_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::hetrd_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -193,12 +193,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::hetrd( + sycl::event func_event = oneapi::math::lapack::hetrd( queue, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::hetrd, uplo, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::hetrd, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/hetrf.cpp b/tests/unit_tests/lapack/source/hetrf.cpp index dc6d7676c..f22384b2c 100644 --- a/tests/unit_tests/lapack/source/hetrf.cpp +++ b/tests/unit_tests/lapack/source/hetrf.cpp @@ -43,7 +43,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -63,11 +63,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::hetrf_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::hetrf_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::hetrf_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::hetrf_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -76,10 +76,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::hetrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, + oneapi::math::lapack::hetrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::hetrf, uplo, n, A_dev, lda, ipiv_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::hetrf, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -104,7 +104,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ for (int64_t d = 0; d < n; d++) U[d + d * ldu] = 1.0; - if (uplo == oneapi::mkl::uplo::upper) { + if (uplo == oneapi::math::uplo::upper) { int64_t k = n - 1; while (k >= 0) { reference::laset('A', n, n, 0.0, 1.0, Uk.data(), ldu); @@ -117,7 +117,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ reference::swap(n, Uk.data() + (k + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), ldu); @@ -135,7 +135,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ reference::swap(n, Uk.data() + (k - 1 + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), ldu); @@ -160,7 +160,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ reference::swap(n, Uk.data() + (k + 0 * lda), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), ldu); @@ -178,7 +178,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ reference::swap(n, Uk.data() + (k + 1 + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), ldu); @@ -194,12 +194,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ /* |A - UDU'| < |A| O(eps) */ std::vector UD(n * n); int64_t ldud = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, n, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, U.data(), ldu, D.data(), ldd, 0.0, UD.data(), ldud); std::vector UDU(n * n); int64_t ldudu = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::conjtrans, n, n, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, n, n, n, 1.0, UD.data(), ldud, U.data(), ldu, 0.0, UDU.data(), ldudu); if (!rel_mat_err_check(n, n, UDU, ldudu, A_initial, lda)) { @@ -215,7 +215,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -236,11 +236,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::hetrf_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::hetrf_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::hetrf_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::hetrf_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -252,11 +252,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::hetrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, + oneapi::math::lapack::hetrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::hetrf, uplo, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::hetrf, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/orgbr.cpp b/tests/unit_tests/lapack/source/orgbr.cpp index d7fec952b..1ed2ab0fd 100644 --- a/tests/unit_tests/lapack/source/orgbr.cpp +++ b/tests/unit_tests/lapack/source/orgbr.cpp @@ -43,7 +43,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, int64_t n, int64_t k, +bool accuracy(const sycl::device& dev, oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -52,9 +52,9 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, in int64_t m_A = m; int64_t n_A = n; - if (vect == oneapi::mkl::generate::Q) + if (vect == oneapi::math::generate::Q) n_A = k; - else /* vect == oneapi::mkl::generate::P */ + else /* vect == oneapi::math::generate::P */ m_A = k; int64_t min_mn_A = std::min(m_A, n_A); @@ -65,10 +65,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, in std::vector tauq(min_mn_A); std::vector taup(min_mn_A); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m_A, n_A, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m_A, n_A, A, lda); reference::gebrd(m_A, n_A, A.data(), lda, d.data(), e.data(), tauq.data(), taup.data()); - auto& tau = (vect == oneapi::mkl::generate::Q) ? tauq : taup; + auto& tau = (vect == oneapi::math::generate::Q) ? tauq : taup; /* Compute on device */ { @@ -79,11 +79,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, in #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::orgbr_scratchpad_size(queue, vect, m, n, k, lda); + oneapi::math::lapack::orgbr_scratchpad_size(queue, vect, m, n, k, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::orgbr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::orgbr_scratchpad_size, vect, m, n, k, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -93,10 +93,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, in queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::orgbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::orgbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::orgbr, vect, m, n, k, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::orgbr, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -117,7 +117,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, int64_t n, +bool usm_dependency(const sycl::device& dev, oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -126,9 +126,9 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::generate vect, int64_t int64_t m_A = m; int64_t n_A = n; - if (vect == oneapi::mkl::generate::Q) + if (vect == oneapi::math::generate::Q) n_A = k; - else /* vect == oneapi::mkl::generate::P */ + else /* vect == oneapi::math::generate::P */ m_A = k; int64_t min_mn_A = std::min(m_A, n_A); @@ -139,10 +139,10 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::generate vect, int64_t std::vector tauq(min_mn_A); std::vector taup(min_mn_A); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m_A, n_A, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m_A, n_A, A, lda); reference::gebrd(m_A, n_A, A.data(), lda, d.data(), e.data(), tauq.data(), taup.data()); - auto& tau = (vect == oneapi::mkl::generate::Q) ? tauq : taup; + auto& tau = (vect == oneapi::math::generate::Q) ? tauq : taup; /* Compute on device */ bool result; @@ -154,11 +154,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::generate vect, int64_t #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::orgbr_scratchpad_size(queue, vect, m, n, k, lda); + oneapi::math::lapack::orgbr_scratchpad_size(queue, vect, m, n, k, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::orgbr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::orgbr_scratchpad_size, vect, m, n, k, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -171,11 +171,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::generate vect, int64_t auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::orgbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::orgbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::orgbr, vect, m, n, k, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::orgbr, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/orgqr.cpp b/tests/unit_tests/lapack/source/orgqr.cpp index 95d218992..f8ddc4f51 100644 --- a/tests/unit_tests/lapack/source/orgqr.cpp +++ b/tests/unit_tests/lapack/source/orgqr.cpp @@ -54,7 +54,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t std::vector A(lda * n); std::vector tau(k); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda); auto info = reference::geqrf(m, k, A.data(), lda, tau.data()); if (0 != info) { test_log::lout << "reference geqrf failed with info: " << info << std::endl; @@ -68,11 +68,11 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::orgqr_scratchpad_size(queue, m, n, k, lda); + oneapi::math::lapack::orgqr_scratchpad_size(queue, m, n, k, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::orgqr_scratchpad_size, m, n, k, lda); + queue, scratchpad_size = oneapi::math::lapack::orgqr_scratchpad_size, m, n, k, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -81,10 +81,10 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::orgqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::orgqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::orgqr, m, n, k, A_dev, lda, tau_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::orgqr, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -112,7 +112,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in /* Initialize */ std::vector A(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda); std::vector tau(k); auto info = reference::geqrf(m, k, A.data(), lda, tau.data()); @@ -129,11 +129,11 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::orgqr_scratchpad_size(queue, m, n, k, lda); + oneapi::math::lapack::orgqr_scratchpad_size(queue, m, n, k, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::orgqr_scratchpad_size, m, n, k, lda); + queue, scratchpad_size = oneapi::math::lapack::orgqr_scratchpad_size, m, n, k, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -145,11 +145,11 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::orgqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::orgqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::orgqr, m, n, k, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::orgqr, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/orgqr_batch_group.cpp b/tests/unit_tests/lapack/source/orgqr_batch_group.cpp index b08c0c9aa..4f44c9cf0 100644 --- a/tests/unit_tests/lapack/source/orgqr_batch_group.cpp +++ b/tests/unit_tests/lapack/source/orgqr_batch_group.cpp @@ -69,7 +69,7 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_list.emplace_back(lda * n); auto& A = A_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda); tau_list.emplace_back(k); auto& tau = tau_list.back(); @@ -101,13 +101,13 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::orgqr_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::orgqr_batch_scratchpad_size( queue, m_vec.data(), n_vec.data(), k_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::orgqr_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::orgqr_batch_scratchpad_size, m_vec.data(), n_vec.data(), k_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif @@ -130,11 +130,11 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::orgqr_batch(queue, m_vec.data(), n_vec.data(), k_vec.data(), + oneapi::math::lapack::orgqr_batch(queue, m_vec.data(), n_vec.data(), k_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::orgqr_batch, m_vec.data(), + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::orgqr_batch, m_vec.data(), n_vec.data(), k_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); @@ -213,7 +213,7 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_list.emplace_back(lda * n); auto& A = A_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda); tau_list.emplace_back(k); auto& tau = tau_list.back(); @@ -246,13 +246,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::orgqr_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::orgqr_batch_scratchpad_size( queue, m_vec.data(), n_vec.data(), k_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::orgqr_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::orgqr_batch_scratchpad_size, m_vec.data(), n_vec.data(), k_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif @@ -277,13 +277,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::orgqr_batch( + sycl::event func_event = oneapi::math::lapack::orgqr_batch( queue, m_vec.data(), n_vec.data(), k_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::orgqr_batch, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::orgqr_batch, m_vec.data(), n_vec.data(), k_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, diff --git a/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp b/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp index c78e7e82c..c88b6e88c 100644 --- a/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp @@ -50,7 +50,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t std::vector tau(stride_tau * batch_size); for (int64_t i = 0; i < batch_size; i++) { - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda, i * stride_a); auto info = reference::geqrf(m, k, A.data() + i * stride_a, lda, tau.data() + i * stride_tau); if (0 != info) { @@ -67,12 +67,12 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t auto A_dev = device_alloc(queue, A.size()); auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::orgqr_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::orgqr_batch_scratchpad_size( queue, m, n, k, lda, stride_a, stride_tau, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::orgqr_batch_scratchpad_size, m, n, k, + queue, scratchpad_size = oneapi::math::lapack::orgqr_batch_scratchpad_size, m, n, k, lda, stride_a, stride_tau, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -82,10 +82,10 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, + oneapi::math::lapack::orgqr_batch(queue, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::orgqr_batch, m, n, k, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::orgqr_batch, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size); #endif @@ -127,7 +127,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in std::vector tau(stride_tau * batch_size); for (int64_t i = 0; i < batch_size; i++) { - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda, i * stride_a); auto info = reference::geqrf(m, k, A.data() + i * stride_a, lda, tau.data() + i * stride_tau); if (0 != info) { @@ -145,12 +145,12 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in auto A_dev = device_alloc(queue, A.size()); auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::orgqr_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::orgqr_batch_scratchpad_size( queue, m, n, k, lda, stride_a, stride_tau, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::orgqr_batch_scratchpad_size, m, n, k, + queue, scratchpad_size = oneapi::math::lapack::orgqr_batch_scratchpad_size, m, n, k, lda, stride_a, stride_tau, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -162,12 +162,12 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::orgqr_batch( + sycl::event func_event = oneapi::math::lapack::orgqr_batch( queue, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::orgqr_batch, m, n, k, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::orgqr_batch, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); diff --git a/tests/unit_tests/lapack/source/orgtr.cpp b/tests/unit_tests/lapack/source/orgtr.cpp index 4021ccc0e..4a909d04e 100644 --- a/tests/unit_tests/lapack/source/orgtr.cpp +++ b/tests/unit_tests/lapack/source/orgtr.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -51,7 +51,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ std::vector e(n); std::vector tau(n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); auto info = reference::sytrd(uplo, n, A.data(), lda, d.data(), e.data(), tau.data()); if (0 != info) { test_log::lout << "reference sytrd failed with info = " << info << std::endl; @@ -66,11 +66,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::orgtr_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::orgtr_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::orgtr_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::orgtr_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -79,10 +79,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::orgtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::orgtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::orgtr, uplo, n, A_dev, lda, tau_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::orgtr, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -103,7 +103,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -114,7 +114,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, std::vector e(n); std::vector tau(n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); auto info = reference::sytrd(uplo, n, A.data(), lda, d.data(), e.data(), tau.data()); if (0 != info) { test_log::lout << "reference sytrd failed with info = " << info << std::endl; @@ -130,11 +130,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::orgtr_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::orgtr_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::orgtr_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::orgtr_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -146,11 +146,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::orgtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::orgtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::orgtr, uplo, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::orgtr, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/ormqr.cpp b/tests/unit_tests/lapack/source/ormqr.cpp index ea64e26be..5181d9a3b 100644 --- a/tests/unit_tests/lapack/source/ormqr.cpp +++ b/tests/unit_tests/lapack/source/ormqr.cpp @@ -47,19 +47,19 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl::transpose trans, +bool accuracy(const sycl::device& dev, oneapi::math::side left_right, oneapi::math::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ std::vector C_initial(ldc * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C_initial, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C_initial, ldc); std::vector C = C_initial; - int64_t nq = (left_right == oneapi::mkl::side::left) ? m : n; + int64_t nq = (left_right == oneapi::math::side::left) ? m : n; std::vector A(lda * k); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, nq, k, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, nq, k, A, lda); std::vector tau(k); auto info = reference::geqrf(nq, k, A.data(), lda, tau.data()); @@ -75,12 +75,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::ormqr_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::ormqr_scratchpad_size( queue, left_right, trans, m, n, k, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::ormqr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::ormqr_scratchpad_size, left_right, trans, m, n, k, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -91,10 +91,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::ormqr(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, + oneapi::math::lapack::ormqr(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::ormqr, left_right, trans, m, n, k, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ormqr, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -131,20 +131,20 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, - oneapi::mkl::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::side left_right, + oneapi::math::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ std::vector C_initial(ldc * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C_initial, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C_initial, ldc); std::vector C = C_initial; - int64_t nq = (left_right == oneapi::mkl::side::left) ? m : n; + int64_t nq = (left_right == oneapi::math::side::left) ? m : n; std::vector A(lda * k); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, nq, k, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, nq, k, A, lda); std::vector tau(k); auto info = reference::geqrf(nq, k, A.data(), lda, tau.data()); @@ -162,12 +162,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::ormqr_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::ormqr_scratchpad_size( queue, left_right, trans, m, n, k, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::ormqr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::ormqr_scratchpad_size, left_right, trans, m, n, k, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -180,12 +180,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::ormqr( + sycl::event func_event = oneapi::math::lapack::ormqr( queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::ormqr, left_right, trans, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ormqr, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/ormrq.cpp b/tests/unit_tests/lapack/source/ormrq.cpp index d9736da19..375632336 100644 --- a/tests/unit_tests/lapack/source/ormrq.cpp +++ b/tests/unit_tests/lapack/source/ormrq.cpp @@ -40,18 +40,18 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl::transpose trans, +bool accuracy(const sycl::device& dev, oneapi::math::side left_right, oneapi::math::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ std::vector C_initial(ldc * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C_initial, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C_initial, ldc); std::vector C = C_initial; int64_t nq; - if (left_right == oneapi::mkl::side::left) { + if (left_right == oneapi::math::side::left) { if (k > m) { test_log::lout << "Bad test input, side == left and k > m (" << k << " > " << m << ")" << std::endl; @@ -69,7 +69,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl } std::vector A(lda * k); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, nq, k, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, nq, k, A, lda); std::vector tau(k); auto info = reference::gerqf(nq, k, A.data(), lda, tau.data()); @@ -85,12 +85,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::ormrq_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::ormrq_scratchpad_size( queue, left_right, trans, m, n, k, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::ormrq_scratchpad_size, + scratchpad_size = oneapi::math::lapack::ormrq_scratchpad_size, left_right, trans, m, n, k, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -101,10 +101,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::ormrq(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, + oneapi::math::lapack::ormrq(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::ormrq, left_right, trans, m, n, k, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ormrq, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -141,20 +141,20 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, - oneapi::mkl::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::side left_right, + oneapi::math::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ std::vector C_initial(ldc * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C_initial, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C_initial, ldc); std::vector C = C_initial; - int64_t nq = (left_right == oneapi::mkl::side::left) ? m : n; + int64_t nq = (left_right == oneapi::math::side::left) ? m : n; std::vector A(lda * k); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, nq, k, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, nq, k, A, lda); std::vector tau(k); auto info = reference::gerqf(nq, k, A.data(), lda, tau.data()); @@ -171,12 +171,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::ormrq_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::ormrq_scratchpad_size( queue, left_right, trans, m, n, k, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::ormrq_scratchpad_size, + scratchpad_size = oneapi::math::lapack::ormrq_scratchpad_size, left_right, trans, m, n, k, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -189,12 +189,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::ormrq( + sycl::event func_event = oneapi::math::lapack::ormrq( queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::ormrq, left_right, trans, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ormrq, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/ormtr.cpp b/tests/unit_tests/lapack/source/ormtr.cpp index 5af36c851..fd37974ea 100644 --- a/tests/unit_tests/lapack/source/ormtr.cpp +++ b/tests/unit_tests/lapack/source/ormtr.cpp @@ -40,17 +40,17 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, int64_t n, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t m, int64_t n, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ - oneapi::mkl::side side = oneapi::mkl::side::right; - oneapi::mkl::transpose trans = oneapi::mkl::transpose::nontrans; + oneapi::math::side side = oneapi::math::side::right; + oneapi::math::transpose trans = oneapi::math::transpose::nontrans; std::vector A(n * lda); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); std::vector tau(n); std::vector d(n); @@ -62,7 +62,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, int64_ } std::vector C(n * ldc); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C, ldc); std::vector C_initial = C; /* Compute on device */ @@ -73,12 +73,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, int64_ auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::ormtr_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::ormtr_scratchpad_size( queue, side, uplo, trans, m, n, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::ormtr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::ormtr_scratchpad_size, side, uplo, trans, m, n, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -89,10 +89,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::ormtr(queue, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, + oneapi::math::lapack::ormtr(queue, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::ormtr, side, uplo, trans, m, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ormtr, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -127,17 +127,17 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, int64_t n, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t m, int64_t n, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ - oneapi::mkl::side side = oneapi::mkl::side::right; - oneapi::mkl::transpose trans = oneapi::mkl::transpose::nontrans; + oneapi::math::side side = oneapi::math::side::right; + oneapi::math::transpose trans = oneapi::math::transpose::nontrans; std::vector A(n * lda); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); std::vector tau(n); std::vector d(n); @@ -149,7 +149,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, } std::vector C(n * ldc); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C, ldc); std::vector C_initial = C; /* Compute on device */ @@ -161,12 +161,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::ormtr_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::ormtr_scratchpad_size( queue, side, uplo, trans, m, n, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::ormtr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::ormtr_scratchpad_size, side, uplo, trans, m, n, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -179,12 +179,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::ormtr( + sycl::event func_event = oneapi::math::lapack::ormtr( queue, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::ormtr, side, uplo, trans, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ormtr, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/potrf.cpp b/tests/unit_tests/lapack/source/potrf.cpp index d3b98d8c5..19063dfcf 100644 --- a/tests/unit_tests/lapack/source/potrf.cpp +++ b/tests/unit_tests/lapack/source/potrf.cpp @@ -43,7 +43,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -61,11 +61,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ auto A_dev = device_alloc(queue, A.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::potrf_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::potrf_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potrf_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::potrf_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -73,9 +73,9 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::potrf(queue, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size); + oneapi::math::lapack::potrf(queue, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::potrf, uplo, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potrf, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -95,7 +95,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -114,11 +114,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto A_dev = device_alloc(queue, A.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::potrf_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::potrf_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potrf_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::potrf_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -129,11 +129,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::potrf(queue, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size, + oneapi::math::lapack::potrf(queue, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::potrf, uplo, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::potrf, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/potrf_batch_group.cpp b/tests/unit_tests/lapack/source/potrf_batch_group.cpp index 97833e08b..ce2fbc718 100644 --- a/tests/unit_tests/lapack/source/potrf_batch_group.cpp +++ b/tests/unit_tests/lapack/source/potrf_batch_group.cpp @@ -46,8 +46,8 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { using fp_real = typename complex_info::real_type; /* Test Parameters */ - std::vector uplo_vec = { oneapi::mkl::uplo::upper, - oneapi::mkl::uplo::lower }; + std::vector uplo_vec = { oneapi::math::uplo::upper, + oneapi::math::uplo::lower }; std::vector n_vec = { 4, 4 }; std::vector lda_vec = { 5, 5 }; std::vector group_sizes_vec = { 2, 2 }; @@ -90,13 +90,13 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::potrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::potrf_batch_scratchpad_size( queue, uplo_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potrf_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::potrf_batch_scratchpad_size, uplo_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -113,11 +113,11 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::potrf_batch(queue, uplo_vec.data(), n_vec.data(), A_dev_ptrs, + oneapi::math::lapack::potrf_batch(queue, uplo_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::potrf_batch, uplo_vec.data(), + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potrf_batch, uplo_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #endif @@ -169,7 +169,7 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { using fp_real = typename complex_info::real_type; /* Test Parameters */ - std::vector uplo_vec = { oneapi::mkl::uplo::upper }; + std::vector uplo_vec = { oneapi::math::uplo::upper }; std::vector n_vec = { 1 }; std::vector lda_vec = { 1 }; std::vector group_sizes_vec = { 1 }; @@ -213,13 +213,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::potrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::potrf_batch_scratchpad_size( queue, uplo_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potrf_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::potrf_batch_scratchpad_size, uplo_vec.data(), n_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -238,13 +238,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::potrf_batch( + sycl::event func_event = oneapi::math::lapack::potrf_batch( queue, uplo_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::potrf_batch, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::potrf_batch, uplo_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); diff --git a/tests/unit_tests/lapack/source/potrf_batch_stride.cpp b/tests/unit_tests/lapack/source/potrf_batch_stride.cpp index f2a606e06..f893a0f56 100644 --- a/tests/unit_tests/lapack/source/potrf_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/potrf_batch_stride.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, int64_t stride_a, int64_t batch_size, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -58,12 +58,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ auto A_dev = device_alloc(queue, A.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::potrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::potrf_batch_scratchpad_size( queue, uplo, n, lda, stride_a, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potrf_batch_scratchpad_size, uplo, n, + queue, scratchpad_size = oneapi::math::lapack::potrf_batch_scratchpad_size, uplo, n, lda, stride_a, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -72,10 +72,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::potrf_batch(queue, uplo, n, A_dev, lda, stride_a, batch_size, + oneapi::math::lapack::potrf_batch(queue, uplo, n, A_dev, lda, stride_a, batch_size, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::potrf_batch, uplo, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potrf_batch, uplo, n, A_dev, lda, stride_a, batch_size, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -105,7 +105,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, int64_t stride_a, int64_t batch_size, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -124,12 +124,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto A_dev = device_alloc(queue, A.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::potrf_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::potrf_batch_scratchpad_size( queue, uplo, n, lda, stride_a, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potrf_batch_scratchpad_size, uplo, n, + queue, scratchpad_size = oneapi::math::lapack::potrf_batch_scratchpad_size, uplo, n, lda, stride_a, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -140,12 +140,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::potrf_batch( + sycl::event func_event = oneapi::math::lapack::potrf_batch( queue, uplo, n, A_dev, lda, stride_a, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::potrf_batch, uplo, n, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::potrf_batch, uplo, n, A_dev, lda, stride_a, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/potri.cpp b/tests/unit_tests/lapack/source/potri.cpp index 93387b4a2..3f10bad09 100644 --- a/tests/unit_tests/lapack/source/potri.cpp +++ b/tests/unit_tests/lapack/source/potri.cpp @@ -43,7 +43,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -65,11 +65,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ auto A_dev = device_alloc(queue, A.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::potri_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::potri_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potri_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::potri_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -77,9 +77,9 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::potri(queue, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size); + oneapi::math::lapack::potri(queue, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::potri, uplo, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potri, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -102,7 +102,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ int64_t ldr = n; for (int64_t diag = 0; diag < n; diag++) resid[diag + diag * ldr] = static_cast(1.0); - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, n, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, A_initial.data(), lda, A.data(), lda, -1.0, resid.data(), ldr); auto norm_resid = reference::lange('1', n, n, resid.data(), ldr); auto rel_err = norm_resid / (norm_A * norm_Ainv * n * ulp); @@ -125,7 +125,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -148,11 +148,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto A_dev = device_alloc(queue, A.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::potri_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::potri_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potri_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::potri_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -163,11 +163,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::potri(queue, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size, + oneapi::math::lapack::potri(queue, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::potri, uplo, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::potri, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/potrs.cpp b/tests/unit_tests/lapack/source/potrs.cpp index 209387345..d3b8abf0c 100644 --- a/tests/unit_tests/lapack/source/potrs.cpp +++ b/tests/unit_tests/lapack/source/potrs.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t nrhs, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t nrhs, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -49,7 +49,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ std::vector A_initial(lda * n); std::vector B_initial(ldb * nrhs); rand_pos_def_matrix(seed, uplo, n, A_initial, lda); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B_initial, ldb); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B_initial, ldb); std::vector A = A_initial; std::vector B = B_initial; @@ -68,11 +68,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ auto B_dev = device_alloc(queue, B.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::potrs_scratchpad_size(queue, uplo, n, nrhs, lda, ldb); + oneapi::math::lapack::potrs_scratchpad_size(queue, uplo, n, nrhs, lda, ldb); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::potrs_scratchpad_size, + scratchpad_size = oneapi::math::lapack::potrs_scratchpad_size, uplo, n, nrhs, lda, ldb); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -82,10 +82,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::potrs(queue, uplo, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, + oneapi::math::lapack::potrs(queue, uplo, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::potrs, uplo, n, nrhs, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potrs, uplo, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -106,7 +106,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t nrhs, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t nrhs, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -115,7 +115,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, std::vector A_initial(lda * n); std::vector B_initial(ldb * nrhs); rand_pos_def_matrix(seed, uplo, n, A_initial, lda); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B_initial, ldb); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B_initial, ldb); std::vector A = A_initial; std::vector B = B_initial; @@ -135,11 +135,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto B_dev = device_alloc(queue, B.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::potrs_scratchpad_size(queue, uplo, n, nrhs, lda, ldb); + oneapi::math::lapack::potrs_scratchpad_size(queue, uplo, n, nrhs, lda, ldb); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::potrs_scratchpad_size, + scratchpad_size = oneapi::math::lapack::potrs_scratchpad_size, uplo, n, nrhs, lda, ldb); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -152,11 +152,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::potrs(queue, uplo, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, + oneapi::math::lapack::potrs(queue, uplo, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::potrs, uplo, n, nrhs, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::potrs, uplo, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/potrs_batch_group.cpp b/tests/unit_tests/lapack/source/potrs_batch_group.cpp index d4d69edee..90359133d 100644 --- a/tests/unit_tests/lapack/source/potrs_batch_group.cpp +++ b/tests/unit_tests/lapack/source/potrs_batch_group.cpp @@ -46,8 +46,8 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { using fp_real = typename complex_info::real_type; /* Test Parameters */ - std::vector uplo_vec = { oneapi::mkl::uplo::upper, - oneapi::mkl::uplo::lower }; + std::vector uplo_vec = { oneapi::math::uplo::upper, + oneapi::math::uplo::lower }; std::vector n_vec = { 4, 5 }; std::vector nrhs_vec = { 9, 6 }; std::vector lda_vec = { 6, 6 }; @@ -81,7 +81,7 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { B_initial_list.emplace_back(ldb * nrhs); auto& B_initial = B_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B_initial, lda); B_list.emplace_back(B_initial); auto& B = B_list.back(); @@ -114,13 +114,13 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::potrs_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::potrs_batch_scratchpad_size( queue, uplo_vec.data(), n_vec.data(), nrhs_vec.data(), lda_vec.data(), ldb_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potrs_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::potrs_batch_scratchpad_size, uplo_vec.data(), n_vec.data(), nrhs_vec.data(), lda_vec.data(), ldb_vec.data(), group_count, group_sizes_vec.data()); #endif @@ -143,12 +143,12 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::potrs_batch(queue, uplo_vec.data(), n_vec.data(), nrhs_vec.data(), + oneapi::math::lapack::potrs_batch(queue, uplo_vec.data(), n_vec.data(), nrhs_vec.data(), A_dev_ptrs, lda_vec.data(), B_dev_ptrs, ldb_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::potrs_batch, uplo_vec.data(), + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potrs_batch, uplo_vec.data(), n_vec.data(), nrhs_vec.data(), A_dev_ptrs, lda_vec.data(), B_dev_ptrs, ldb_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); @@ -209,7 +209,7 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { using fp_real = typename complex_info::real_type; /* Test Parameters */ - std::vector uplo_vec = { oneapi::mkl::uplo::upper }; + std::vector uplo_vec = { oneapi::math::uplo::upper }; std::vector n_vec = { 1 }; std::vector nrhs_vec = { 1 }; std::vector lda_vec = { 1 }; @@ -243,7 +243,7 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { B_initial_list.emplace_back(ldb * nrhs); auto& B_initial = B_initial_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B_initial, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B_initial, lda); B_list.emplace_back(B_initial); auto& B = B_list.back(); @@ -277,13 +277,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::potrs_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::potrs_batch_scratchpad_size( queue, uplo_vec.data(), n_vec.data(), nrhs_vec.data(), lda_vec.data(), ldb_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potrs_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::potrs_batch_scratchpad_size, uplo_vec.data(), n_vec.data(), nrhs_vec.data(), lda_vec.data(), ldb_vec.data(), group_count, group_sizes_vec.data()); #endif @@ -308,13 +308,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::potrs_batch( + sycl::event func_event = oneapi::math::lapack::potrs_batch( queue, uplo_vec.data(), n_vec.data(), nrhs_vec.data(), A_dev_ptrs, lda_vec.data(), B_dev_ptrs, ldb_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::potrs_batch, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::potrs_batch, uplo_vec.data(), n_vec.data(), nrhs_vec.data(), A_dev_ptrs, lda_vec.data(), B_dev_ptrs, ldb_vec.data(), group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, diff --git a/tests/unit_tests/lapack/source/potrs_batch_stride.cpp b/tests/unit_tests/lapack/source/potrs_batch_stride.cpp index 2eedc3190..6affb7105 100644 --- a/tests/unit_tests/lapack/source/potrs_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/potrs_batch_stride.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t nrhs, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t nrhs, int64_t lda, int64_t stride_a, int64_t ldb, int64_t stride_b, int64_t batch_size, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -50,7 +50,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ std::vector B_initial(stride_b * batch_size); for (int64_t i = 0; i < batch_size; i++) { rand_pos_def_matrix(seed, uplo, n, A_initial, lda, i * stride_a); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B_initial, ldb, i * stride_b); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B_initial, ldb, i * stride_b); } std::vector A = A_initial; @@ -72,12 +72,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ auto A_dev = device_alloc(queue, A.size()); auto B_dev = device_alloc(queue, B.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::potrs_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::potrs_batch_scratchpad_size( queue, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potrs_batch_scratchpad_size, uplo, n, + queue, scratchpad_size = oneapi::math::lapack::potrs_batch_scratchpad_size, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -87,10 +87,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, A_dev, lda, stride_a, B_dev, ldb, + oneapi::math::lapack::potrs_batch(queue, uplo, n, nrhs, A_dev, lda, stride_a, B_dev, ldb, stride_b, batch_size, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::potrs_batch, uplo, n, nrhs, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potrs_batch, uplo, n, nrhs, A_dev, lda, stride_a, B_dev, ldb, stride_b, batch_size, scratchpad_dev, scratchpad_size); #endif @@ -123,7 +123,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t nrhs, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t nrhs, int64_t lda, int64_t stride_a, int64_t ldb, int64_t stride_b, int64_t batch_size, uint64_t seed) { using fp = typename data_T_info::value_type; @@ -134,7 +134,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, std::vector B_initial(stride_b * batch_size); for (int64_t i = 0; i < batch_size; i++) { rand_pos_def_matrix(seed, uplo, n, A_initial, lda, i * stride_a); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B_initial, ldb, i * stride_b); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B_initial, ldb, i * stride_b); } std::vector A = A_initial; @@ -157,12 +157,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto A_dev = device_alloc(queue, A.size()); auto B_dev = device_alloc(queue, B.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::potrs_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::potrs_batch_scratchpad_size( queue, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::potrs_batch_scratchpad_size, uplo, n, + queue, scratchpad_size = oneapi::math::lapack::potrs_batch_scratchpad_size, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -174,12 +174,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::potrs_batch( + sycl::event func_event = oneapi::math::lapack::potrs_batch( queue, uplo, n, nrhs, A_dev, lda, stride_a, B_dev, ldb, stride_b, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::potrs_batch, uplo, n, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::potrs_batch, uplo, n, nrhs, A_dev, lda, stride_a, B_dev, ldb, stride_b, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); diff --git a/tests/unit_tests/lapack/source/syevd.cpp b/tests/unit_tests/lapack/source/syevd.cpp index bf5f2d61b..c0de25b70 100644 --- a/tests/unit_tests/lapack/source/syevd.cpp +++ b/tests/unit_tests/lapack/source/syevd.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, int64_t n, +bool accuracy(const sycl::device& dev, oneapi::math::job jobz, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -59,11 +59,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl::uplo auto w_dev = device_alloc(queue, w.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::syevd_scratchpad_size(queue, jobz, uplo, n, lda); + oneapi::math::lapack::syevd_scratchpad_size(queue, jobz, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::syevd_scratchpad_size, + scratchpad_size = oneapi::math::lapack::syevd_scratchpad_size, jobz, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -72,10 +72,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl::uplo queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::syevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, + oneapi::math::lapack::syevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::syevd, jobz, uplo, n, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::syevd, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -97,7 +97,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +bool usm_dependency(const sycl::device& dev, oneapi::math::job jobz, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -117,11 +117,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl: auto w_dev = device_alloc(queue, w.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::syevd_scratchpad_size(queue, jobz, uplo, n, lda); + oneapi::math::lapack::syevd_scratchpad_size(queue, jobz, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::syevd_scratchpad_size, + scratchpad_size = oneapi::math::lapack::syevd_scratchpad_size, jobz, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -133,11 +133,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::job jobz, oneapi::mkl: auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::syevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, + oneapi::math::lapack::syevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::syevd, jobz, uplo, n, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::syevd, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/sygvd.cpp b/tests/unit_tests/lapack/source/sygvd.cpp index 8ea658299..ce671feb2 100644 --- a/tests/unit_tests/lapack/source/sygvd.cpp +++ b/tests/unit_tests/lapack/source/sygvd.cpp @@ -42,7 +42,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, int64_t n, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -65,11 +65,11 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one auto w_dev = device_alloc(queue, w.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, jobz, uplo, n, lda, ldb); + oneapi::math::lapack::sygvd_scratchpad_size(queue, itype, jobz, uplo, n, lda, ldb); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::sygvd_scratchpad_size, + scratchpad_size = oneapi::math::lapack::sygvd_scratchpad_size, itype, jobz, uplo, n, lda, ldb); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -79,10 +79,10 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::sygvd(queue, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, + oneapi::math::lapack::sygvd(queue, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::sygvd, itype, jobz, uplo, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::sygvd, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -108,7 +108,7 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* |D_ref - D| < |D_ref| O(eps) */ std::vector D_ref(n); auto info = - reference::sygvd(itype, oneapi::mkl::job::novec, uplo, n, std::vector(A_initial).data(), + reference::sygvd(itype, oneapi::math::job::novec, uplo, n, std::vector(A_initial).data(), lda, std::vector(B_initial).data(), ldb, D_ref.data()); if (0 != info) { test_log::lout << "reference sygvd failed with info = " << info << std::endl; @@ -119,17 +119,17 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one result = false; } - if (oneapi::mkl::job::vec == jobz) { + if (oneapi::math::job::vec == jobz) { if (itype == 1) { /* |A Z - B Z D| < |A Z| O(eps) */ std::vector AZ(n * n); int64_t ldaz = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, A_initial.data(), lda, Z.data(), ldz, 0.0, AZ.data(), ldaz); std::vector BZ(n * n); int64_t ldbz = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, B_initial.data(), ldb, Z.data(), ldz, 0.0, BZ.data(), ldbz); std::vector BZD(n * n); @@ -146,7 +146,7 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* |I - Z' B Z| < n O(eps) */ std::vector ZBZ(n * n); int64_t ldzbz = n; - reference::gemm(oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); if (!rel_id_err_check(n, ZBZ, ldzbz)) { test_log::lout << "Orthogonality check failed" << std::endl; @@ -157,12 +157,12 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* |A B Z - Z D| < |A B Z| O(eps) */ std::vector BZ(n * n); int64_t ldbz = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, B_initial.data(), ldb, Z.data(), ldz, 0.0, BZ.data(), ldbz); std::vector ABZ(n * n); int64_t ldabz = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, A_initial.data(), lda, BZ.data(), ldbz, 0.0, ABZ.data(), ldabz); @@ -180,7 +180,7 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* |I - Z' B Z| < n O(eps) */ std::vector ZBZ(n * n); int64_t ldzbz = n; - reference::gemm(oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); if (!rel_id_err_check(n, ZBZ, ldzbz)) { test_log::lout << "Orthogonality check failed" << std::endl; @@ -192,7 +192,7 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* C = B^-1 Z */ std::vector AZ(n * n); int64_t ldaz = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, A_initial.data(), lda, Z.data(), ldz, 0.0, AZ.data(), ldaz); std::vector C(n * n); @@ -218,7 +218,7 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, one /* |I - Z' B^-1 Z| = |I - Z' C| < n O(eps) */ std::vector ZhC(n * n); int64_t ldzhc = n; - reference::gemm(oneapi::mkl::transpose::conjtrans, oneapi::mkl::transpose::nontrans, n, + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, Z.data(), ldz, C.data(), ldc, 0.0, ZhC.data(), ldzhc); if (!rel_id_err_check(n, ZhC, ldzhc)) { test_log::lout << "Orthogonality check failed" << std::endl; @@ -234,8 +234,8 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, int64_t n, int64_t lda, int64_t ldb, uint64_t seed) { +bool usm_dependency(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, int64_t n, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -258,11 +258,11 @@ bool usm_dependency(const sycl::device& dev, int64_t itype, oneapi::mkl::job job auto w_dev = device_alloc(queue, w.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, jobz, uplo, n, lda, ldb); + oneapi::math::lapack::sygvd_scratchpad_size(queue, itype, jobz, uplo, n, lda, ldb); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::sygvd_scratchpad_size, + scratchpad_size = oneapi::math::lapack::sygvd_scratchpad_size, itype, jobz, uplo, n, lda, ldb); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -274,12 +274,12 @@ bool usm_dependency(const sycl::device& dev, int64_t itype, oneapi::mkl::job job /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::sygvd( + sycl::event func_event = oneapi::math::lapack::sygvd( queue, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::sygvd, itype, jobz, uplo, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::sygvd, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/sytrd.cpp b/tests/unit_tests/lapack/source/sytrd.cpp index e6864e09b..3b804d84b 100644 --- a/tests/unit_tests/lapack/source/sytrd.cpp +++ b/tests/unit_tests/lapack/source/sytrd.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -50,7 +50,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ std::vector d(n); std::vector e(n); std::vector tau(n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); std::vector A_initial = A; /* Compute on device */ @@ -63,11 +63,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::sytrd_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::sytrd_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::sytrd_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::sytrd_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -78,10 +78,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::sytrd(queue, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, + oneapi::math::lapack::sytrd(queue, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::sytrd, uplo, n, A_dev, lda, d_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::sytrd, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -114,9 +114,9 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ std::vector QTQ{ T }; int64_t ldqtq = n; - reference::or_un_mtr(oneapi::mkl::side::left, uplo, oneapi::mkl::transpose::nontrans, n, n, + reference::or_un_mtr(oneapi::math::side::left, uplo, oneapi::math::transpose::nontrans, n, n, A.data(), lda, tau.data(), QTQ.data(), ldqtq); - reference::or_un_mtr(oneapi::mkl::side::right, uplo, oneapi::mkl::transpose::trans, n, n, + reference::or_un_mtr(oneapi::math::side::right, uplo, oneapi::math::transpose::trans, n, n, A.data(), lda, tau.data(), QTQ.data(), ldqtq); if (!rel_mat_err_check(n, n, QTQ, ldqtq, A_initial, lda)) { @@ -127,7 +127,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ /* A[i, i] = d[i] */ for (int64_t diag = 0; diag < n; diag++) d[diag] -= A[diag + diag * lda]; - if (uplo == oneapi::mkl::uplo::upper) + if (uplo == oneapi::math::uplo::upper) for (int64_t diag = 0; diag < n - 1; diag++) e[diag] -= A[diag + (diag + 1) * lda]; else @@ -152,7 +152,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -162,7 +162,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, std::vector d(n); std::vector e(n); std::vector tau(n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); std::vector A_initial = A; /* Compute on device */ @@ -176,11 +176,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::sytrd_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::sytrd_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::sytrd_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::sytrd_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -193,12 +193,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::sytrd( + sycl::event func_event = oneapi::math::lapack::sytrd( queue, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::sytrd, uplo, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::sytrd, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/sytrf.cpp b/tests/unit_tests/lapack/source/sytrf.cpp index 3d04c528a..e24fba32e 100644 --- a/tests/unit_tests/lapack/source/sytrf.cpp +++ b/tests/unit_tests/lapack/source/sytrf.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -60,11 +60,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::sytrf_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::sytrf_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::sytrf_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::sytrf_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -73,10 +73,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::sytrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, + oneapi::math::lapack::sytrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::sytrf, uplo, n, A_dev, lda, ipiv_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::sytrf, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -101,7 +101,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ for (int64_t d = 0; d < n; d++) U[d + d * ldu] = 1.0; - if (uplo == oneapi::mkl::uplo::upper) { + if (uplo == oneapi::math::uplo::upper) { int64_t k = n - 1; while (k >= 0) { reference::laset('A', n, n, 0.0, 1.0, Uk.data(), ldu); @@ -114,7 +114,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ reference::swap(n, Uk.data() + (k + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), ldu); @@ -132,7 +132,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ reference::swap(n, Uk.data() + (k - 1 + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), ldu); @@ -157,7 +157,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ reference::swap(n, Uk.data() + (k + 0 * lda), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), ldu); @@ -175,7 +175,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ reference::swap(n, Uk.data() + (k + 1 + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), ldu); @@ -191,12 +191,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ /* |A - UDU'| < |A| O(eps) */ std::vector UD(n * n); int64_t ldud = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::nontrans, n, n, n, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, n, n, 1.0, U.data(), ldu, D.data(), ldd, 0.0, UD.data(), ldud); std::vector UDU(n * n); int64_t ldudu = n; - reference::gemm(oneapi::mkl::transpose::nontrans, oneapi::mkl::transpose::trans, n, n, n, 1.0, + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::trans, n, n, n, 1.0, UD.data(), ldud, U.data(), ldu, 0.0, UDU.data(), ldudu); if (!rel_mat_err_check(n, n, UDU, ldudu, A_initial, lda)) { @@ -212,7 +212,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -233,11 +233,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto ipiv_dev = device_alloc(queue, ipiv.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::sytrf_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::sytrf_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::sytrf_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::sytrf_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -249,11 +249,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::sytrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, + oneapi::math::lapack::sytrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::sytrf, uplo, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::sytrf, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/trtrs.cpp b/tests/unit_tests/lapack/source/trtrs.cpp index 70946ca11..4857e3892 100644 --- a/tests/unit_tests/lapack/source/trtrs.cpp +++ b/tests/unit_tests/lapack/source/trtrs.cpp @@ -47,8 +47,8 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, int64_t n, int64_t nrhs, int64_t lda, int64_t ldb, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, int64_t n, int64_t nrhs, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -58,8 +58,8 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, oneapi::mkl::tran std::vector B(ldb * nrhs); /* Initialize input data */ - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B, ldb); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B, ldb); std::vector B_initial = B; /* Compute on device */ @@ -70,12 +70,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, oneapi::mkl::tran auto B_dev = device_alloc(queue, B.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::trtrs_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::trtrs_scratchpad_size( queue, uplo, trans, diag, n, nrhs, lda, ldb); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::trtrs_scratchpad_size, + scratchpad_size = oneapi::math::lapack::trtrs_scratchpad_size, uplo, trans, diag, n, nrhs, lda, ldb); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -85,10 +85,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, oneapi::mkl::tran queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::trtrs(queue, uplo, trans, diag, n, nrhs, A_dev, lda, B_dev, ldb, + oneapi::math::lapack::trtrs(queue, uplo, trans, diag, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::trtrs, uplo, trans, diag, n, nrhs, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::trtrs, uplo, trans, diag, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -109,8 +109,8 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, int64_t n, int64_t nrhs, int64_t lda, int64_t ldb, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, int64_t n, int64_t nrhs, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -120,8 +120,8 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, oneapi::mkl std::vector B(ldb * nrhs); /* Initialize input data */ - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, nrhs, B, ldb); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, nrhs, B, ldb); std::vector B_initial = B; /* Compute on device */ @@ -133,12 +133,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, oneapi::mkl auto B_dev = device_alloc(queue, B.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::trtrs_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::trtrs_scratchpad_size( queue, uplo, trans, diag, n, nrhs, lda, ldb); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::trtrs_scratchpad_size, + scratchpad_size = oneapi::math::lapack::trtrs_scratchpad_size, uplo, trans, diag, n, nrhs, lda, ldb); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -150,12 +150,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, oneapi::mkl /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::trtrs( + sycl::event func_event = oneapi::math::lapack::trtrs( queue, uplo, trans, diag, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::trtrs, uplo, trans, diag, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::trtrs, uplo, trans, diag, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/ungbr.cpp b/tests/unit_tests/lapack/source/ungbr.cpp index 3fd840406..723429025 100644 --- a/tests/unit_tests/lapack/source/ungbr.cpp +++ b/tests/unit_tests/lapack/source/ungbr.cpp @@ -43,7 +43,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, int64_t n, int64_t k, +bool accuracy(const sycl::device& dev, oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -52,9 +52,9 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, in int64_t m_A = m; int64_t n_A = n; - if (vect == oneapi::mkl::generate::Q) + if (vect == oneapi::math::generate::Q) n_A = k; - else /* vect == oneapi::mkl::generate::P */ + else /* vect == oneapi::math::generate::P */ m_A = k; int64_t min_mn_A = std::min(m_A, n_A); @@ -65,10 +65,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, in std::vector tauq(min_mn_A); std::vector taup(min_mn_A); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m_A, n_A, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m_A, n_A, A, lda); reference::gebrd(m_A, n_A, A.data(), lda, d.data(), e.data(), tauq.data(), taup.data()); - auto& tau = (vect == oneapi::mkl::generate::Q) ? tauq : taup; + auto& tau = (vect == oneapi::math::generate::Q) ? tauq : taup; /* Compute on device */ { @@ -79,11 +79,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, in #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::ungbr_scratchpad_size(queue, vect, m, n, k, lda); + oneapi::math::lapack::ungbr_scratchpad_size(queue, vect, m, n, k, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::ungbr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::ungbr_scratchpad_size, vect, m, n, k, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -93,10 +93,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, in queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::ungbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::ungbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::ungbr, vect, m, n, k, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ungbr, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -117,7 +117,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::generate vect, int64_t m, int64_t n, +bool usm_dependency(const sycl::device& dev, oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -126,9 +126,9 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::generate vect, int64_t int64_t m_A = m; int64_t n_A = n; - if (vect == oneapi::mkl::generate::Q) + if (vect == oneapi::math::generate::Q) n_A = k; - else /* vect == oneapi::mkl::generate::P */ + else /* vect == oneapi::math::generate::P */ m_A = k; int64_t min_mn_A = std::min(m_A, n_A); @@ -139,10 +139,10 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::generate vect, int64_t std::vector tauq(min_mn_A); std::vector taup(min_mn_A); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m_A, n_A, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m_A, n_A, A, lda); reference::gebrd(m_A, n_A, A.data(), lda, d.data(), e.data(), tauq.data(), taup.data()); - auto& tau = (vect == oneapi::mkl::generate::Q) ? tauq : taup; + auto& tau = (vect == oneapi::math::generate::Q) ? tauq : taup; /* Compute on device */ bool result; @@ -154,11 +154,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::generate vect, int64_t #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::ungbr_scratchpad_size(queue, vect, m, n, k, lda); + oneapi::math::lapack::ungbr_scratchpad_size(queue, vect, m, n, k, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::ungbr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::ungbr_scratchpad_size, vect, m, n, k, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -171,11 +171,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::generate vect, int64_t auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::ungbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::ungbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::ungbr, vect, m, n, k, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ungbr, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/ungqr.cpp b/tests/unit_tests/lapack/source/ungqr.cpp index dce9a9308..63737914a 100644 --- a/tests/unit_tests/lapack/source/ungqr.cpp +++ b/tests/unit_tests/lapack/source/ungqr.cpp @@ -51,7 +51,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t /* Initialize */ std::vector A(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda); std::vector tau(k); auto info = reference::geqrf(m, k, A.data(), lda, tau.data()); @@ -67,11 +67,11 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::ungqr_scratchpad_size(queue, m, n, k, lda); + oneapi::math::lapack::ungqr_scratchpad_size(queue, m, n, k, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::ungqr_scratchpad_size, m, n, k, lda); + queue, scratchpad_size = oneapi::math::lapack::ungqr_scratchpad_size, m, n, k, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -80,10 +80,10 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::ungqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::ungqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::ungqr, m, n, k, A_dev, lda, tau_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ungqr, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -111,7 +111,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in /* Initialize */ std::vector A(lda * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda); std::vector tau(k); auto info = reference::geqrf(m, k, A.data(), lda, tau.data()); @@ -128,11 +128,11 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::ungqr_scratchpad_size(queue, m, n, k, lda); + oneapi::math::lapack::ungqr_scratchpad_size(queue, m, n, k, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::ungqr_scratchpad_size, m, n, k, lda); + queue, scratchpad_size = oneapi::math::lapack::ungqr_scratchpad_size, m, n, k, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -144,11 +144,11 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::ungqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::ungqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::ungqr, m, n, k, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ungqr, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/ungqr_batch_group.cpp b/tests/unit_tests/lapack/source/ungqr_batch_group.cpp index 99c7303b2..0b188ce1c 100644 --- a/tests/unit_tests/lapack/source/ungqr_batch_group.cpp +++ b/tests/unit_tests/lapack/source/ungqr_batch_group.cpp @@ -69,7 +69,7 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_list.emplace_back(lda * n); auto& A = A_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda); tau_list.emplace_back(k); auto& tau = tau_list.back(); @@ -101,13 +101,13 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::ungqr_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::ungqr_batch_scratchpad_size( queue, m_vec.data(), n_vec.data(), k_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::ungqr_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::ungqr_batch_scratchpad_size, m_vec.data(), n_vec.data(), k_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif @@ -130,11 +130,11 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::ungqr_batch(queue, m_vec.data(), n_vec.data(), k_vec.data(), + oneapi::math::lapack::ungqr_batch(queue, m_vec.data(), n_vec.data(), k_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::ungqr_batch, m_vec.data(), + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ungqr_batch, m_vec.data(), n_vec.data(), k_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); @@ -213,7 +213,7 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { for (int64_t local_id = 0; local_id < group_size; local_id++) { A_list.emplace_back(lda * n); auto& A = A_list.back(); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda); tau_list.emplace_back(k); auto& tau = tau_list.back(); @@ -246,13 +246,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { } #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::ungqr_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::ungqr_batch_scratchpad_size( queue, m_vec.data(), n_vec.data(), k_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::ungqr_batch_scratchpad_size, + queue, scratchpad_size = oneapi::math::lapack::ungqr_batch_scratchpad_size, m_vec.data(), n_vec.data(), k_vec.data(), lda_vec.data(), group_count, group_sizes_vec.data()); #endif @@ -277,13 +277,13 @@ bool usm_dependency(const sycl::device& dev, uint64_t seed) { /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::ungqr_batch( + sycl::event func_event = oneapi::math::lapack::ungqr_batch( queue, m_vec.data(), n_vec.data(), k_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::ungqr_batch, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ungqr_batch, m_vec.data(), n_vec.data(), k_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size, diff --git a/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp b/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp index 1da5a60df..782648c8d 100644 --- a/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp @@ -50,7 +50,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t std::vector tau(stride_tau * batch_size); for (int64_t i = 0; i < batch_size; i++) { - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda, i * stride_a); auto info = reference::geqrf(m, k, A.data() + i * stride_a, lda, tau.data() + i * stride_tau); if (0 != info) { @@ -67,12 +67,12 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t auto A_dev = device_alloc(queue, A.size()); auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::ungqr_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::ungqr_batch_scratchpad_size( queue, m, n, k, lda, stride_a, stride_tau, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::ungqr_batch_scratchpad_size, m, n, k, + queue, scratchpad_size = oneapi::math::lapack::ungqr_batch_scratchpad_size, m, n, k, lda, stride_a, stride_tau, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -82,10 +82,10 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, + oneapi::math::lapack::ungqr_batch(queue, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::ungqr_batch, m, n, k, A_dev, lda, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ungqr_batch, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size); #endif @@ -127,7 +127,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in std::vector tau(stride_tau * batch_size); for (int64_t i = 0; i < batch_size; i++) { - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, A, lda, i * stride_a); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, A, lda, i * stride_a); auto info = reference::geqrf(m, k, A.data() + i * stride_a, lda, tau.data() + i * stride_tau); if (0 != info) { @@ -145,12 +145,12 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in auto A_dev = device_alloc(queue, A.size()); auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::ungqr_batch_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::ungqr_batch_scratchpad_size( queue, m, n, k, lda, stride_a, stride_tau, batch_size); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::ungqr_batch_scratchpad_size, m, n, k, + queue, scratchpad_size = oneapi::math::lapack::ungqr_batch_scratchpad_size, m, n, k, lda, stride_a, stride_tau, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -162,12 +162,12 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::ungqr_batch( + sycl::event func_event = oneapi::math::lapack::ungqr_batch( queue, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::ungqr_batch, m, n, k, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ungqr_batch, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, scratchpad_size, std::vector{ in_event }); diff --git a/tests/unit_tests/lapack/source/ungtr.cpp b/tests/unit_tests/lapack/source/ungtr.cpp index b6204e606..628ec95c0 100644 --- a/tests/unit_tests/lapack/source/ungtr.cpp +++ b/tests/unit_tests/lapack/source/ungtr.cpp @@ -40,7 +40,7 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -51,7 +51,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ std::vector e(n); std::vector tau(n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); auto info = reference::hetrd(uplo, n, A.data(), lda, d.data(), e.data(), tau.data()); if (0 != info) { test_log::lout << "reference hetrd failed with info = " << info << std::endl; @@ -66,11 +66,11 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::ungtr_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::ungtr_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::ungtr_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::ungtr_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -79,10 +79,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::ungtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::ungtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::ungtr, uplo, n, A_dev, lda, tau_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ungtr, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -103,7 +103,7 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t lda, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -114,7 +114,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, std::vector e(n); std::vector tau(n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); auto info = reference::hetrd(uplo, n, A.data(), lda, d.data(), e.data(), tau.data()); if (0 != info) { test_log::lout << "reference hetrd failed with info = " << info << std::endl; @@ -130,11 +130,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto tau_dev = device_alloc(queue, tau.size()); #ifdef CALL_RT_API const auto scratchpad_size = - oneapi::mkl::lapack::ungtr_scratchpad_size(queue, uplo, n, lda); + oneapi::math::lapack::ungtr_scratchpad_size(queue, uplo, n, lda); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::mkl::lapack::ungtr_scratchpad_size, uplo, n, lda); + queue, scratchpad_size = oneapi::math::lapack::ungtr_scratchpad_size, uplo, n, lda); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -146,11 +146,11 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t n, auto in_event = create_dependency(queue); #ifdef CALL_RT_API sycl::event func_event = - oneapi::mkl::lapack::ungtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, + oneapi::math::lapack::ungtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::ungtr, uplo, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ungtr, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/unmqr.cpp b/tests/unit_tests/lapack/source/unmqr.cpp index 95f2ef2f1..d613751e3 100644 --- a/tests/unit_tests/lapack/source/unmqr.cpp +++ b/tests/unit_tests/lapack/source/unmqr.cpp @@ -47,19 +47,19 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl::transpose trans, +bool accuracy(const sycl::device& dev, oneapi::math::side left_right, oneapi::math::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ std::vector C_initial(ldc * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C_initial, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C_initial, ldc); std::vector C = C_initial; - int64_t nq = (left_right == oneapi::mkl::side::left) ? m : n; + int64_t nq = (left_right == oneapi::math::side::left) ? m : n; std::vector A(lda * k); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, nq, k, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, nq, k, A, lda); std::vector tau(k); auto info = reference::geqrf(nq, k, A.data(), lda, tau.data()); @@ -75,12 +75,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::unmqr_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::unmqr_scratchpad_size( queue, left_right, trans, m, n, k, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::unmqr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::unmqr_scratchpad_size, left_right, trans, m, n, k, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -91,10 +91,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::unmqr(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, + oneapi::math::lapack::unmqr(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::unmqr, left_right, trans, m, n, k, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::unmqr, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -131,20 +131,20 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, - oneapi::mkl::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::side left_right, + oneapi::math::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ std::vector C_initial(ldc * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C_initial, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C_initial, ldc); std::vector C = C_initial; - int64_t nq = (left_right == oneapi::mkl::side::left) ? m : n; + int64_t nq = (left_right == oneapi::math::side::left) ? m : n; std::vector A(lda * k); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, nq, k, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, nq, k, A, lda); std::vector tau(k); auto info = reference::geqrf(nq, k, A.data(), lda, tau.data()); @@ -162,12 +162,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::unmqr_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::unmqr_scratchpad_size( queue, left_right, trans, m, n, k, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::unmqr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::unmqr_scratchpad_size, left_right, trans, m, n, k, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -180,12 +180,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::unmqr( + sycl::event func_event = oneapi::math::lapack::unmqr( queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::unmqr, left_right, trans, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::unmqr, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/unmrq.cpp b/tests/unit_tests/lapack/source/unmrq.cpp index 8a26c7657..41d99a389 100644 --- a/tests/unit_tests/lapack/source/unmrq.cpp +++ b/tests/unit_tests/lapack/source/unmrq.cpp @@ -40,18 +40,18 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl::transpose trans, +bool accuracy(const sycl::device& dev, oneapi::math::side left_right, oneapi::math::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ std::vector C_initial(ldc * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C_initial, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C_initial, ldc); std::vector C = C_initial; int64_t nq; - if (left_right == oneapi::mkl::side::left) { + if (left_right == oneapi::math::side::left) { if (k > m) { test_log::lout << "Bad test input, side == left and k > m (" << k << " > " << m << ")" << std::endl; @@ -69,7 +69,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl } std::vector A(lda * k); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, nq, k, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, nq, k, A, lda); std::vector tau(k); auto info = reference::gerqf(nq, k, A.data(), lda, tau.data()); @@ -85,12 +85,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::unmrq_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::unmrq_scratchpad_size( queue, left_right, trans, m, n, k, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::unmrq_scratchpad_size, + scratchpad_size = oneapi::math::lapack::unmrq_scratchpad_size, left_right, trans, m, n, k, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -101,10 +101,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::side left_right, oneapi::mkl queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::unmrq(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, + oneapi::math::lapack::unmrq(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::unmrq, left_right, trans, m, n, k, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::unmrq, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -141,20 +141,20 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, - oneapi::mkl::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, +bool usm_dependency(const sycl::device& dev, oneapi::math::side left_right, + oneapi::math::transpose trans, int64_t m, int64_t n, int64_t k, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ std::vector C_initial(ldc * n); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C_initial, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C_initial, ldc); std::vector C = C_initial; - int64_t nq = (left_right == oneapi::mkl::side::left) ? m : n; + int64_t nq = (left_right == oneapi::math::side::left) ? m : n; std::vector A(lda * k); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, nq, k, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, nq, k, A, lda); std::vector tau(k); auto info = reference::gerqf(nq, k, A.data(), lda, tau.data()); @@ -171,12 +171,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::unmrq_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::unmrq_scratchpad_size( queue, left_right, trans, m, n, k, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::unmrq_scratchpad_size, + scratchpad_size = oneapi::math::lapack::unmrq_scratchpad_size, left_right, trans, m, n, k, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -189,12 +189,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::side left_right, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::unmrq( + sycl::event func_event = oneapi::math::lapack::unmrq( queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::unmrq, left_right, trans, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::unmrq, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/lapack/source/unmtr.cpp b/tests/unit_tests/lapack/source/unmtr.cpp index 31a2935d5..00761c469 100644 --- a/tests/unit_tests/lapack/source/unmtr.cpp +++ b/tests/unit_tests/lapack/source/unmtr.cpp @@ -40,17 +40,17 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, int64_t n, int64_t lda, +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t m, int64_t n, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ - oneapi::mkl::side side = oneapi::mkl::side::right; - oneapi::mkl::transpose trans = oneapi::mkl::transpose::nontrans; + oneapi::math::side side = oneapi::math::side::right; + oneapi::math::transpose trans = oneapi::math::transpose::nontrans; std::vector A(n * lda); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); std::vector tau(n); std::vector d(n); @@ -62,7 +62,7 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, int64_ } std::vector C(n * ldc); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C, ldc); std::vector C_initial = C; /* Compute on device */ @@ -73,12 +73,12 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, int64_ auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::unmtr_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::unmtr_scratchpad_size( queue, side, uplo, trans, m, n, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::unmtr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::unmtr_scratchpad_size, side, uplo, trans, m, n, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -89,10 +89,10 @@ bool accuracy(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, int64_ queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::mkl::lapack::unmtr(queue, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, + oneapi::math::lapack::unmtr(queue, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::mkl::lapack::unmtr, side, uplo, trans, m, n, A_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::unmtr, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -127,17 +127,17 @@ const char* dependency_input = R"( )"; template -bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, int64_t n, +bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t m, int64_t n, int64_t lda, int64_t ldc, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; /* Initialize */ - oneapi::mkl::side side = oneapi::mkl::side::right; - oneapi::mkl::transpose trans = oneapi::mkl::transpose::nontrans; + oneapi::math::side side = oneapi::math::side::right; + oneapi::math::transpose trans = oneapi::math::transpose::nontrans; std::vector A(n * lda); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, n, n, A, lda); + rand_matrix(seed, oneapi::math::transpose::nontrans, n, n, A, lda); std::vector tau(n); std::vector d(n); @@ -149,7 +149,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, } std::vector C(n * ldc); - rand_matrix(seed, oneapi::mkl::transpose::nontrans, m, n, C, ldc); + rand_matrix(seed, oneapi::math::transpose::nontrans, m, n, C, ldc); std::vector C_initial = C; /* Compute on device */ @@ -161,12 +161,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, auto tau_dev = device_alloc(queue, tau.size()); auto C_dev = device_alloc(queue, C.size()); #ifdef CALL_RT_API - const auto scratchpad_size = oneapi::mkl::lapack::unmtr_scratchpad_size( + const auto scratchpad_size = oneapi::math::lapack::unmtr_scratchpad_size( queue, side, uplo, trans, m, n, lda, ldc); #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT(queue, - scratchpad_size = oneapi::mkl::lapack::unmtr_scratchpad_size, + scratchpad_size = oneapi::math::lapack::unmtr_scratchpad_size, side, uplo, trans, m, n, lda, ldc); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -179,12 +179,12 @@ bool usm_dependency(const sycl::device& dev, oneapi::mkl::uplo uplo, int64_t m, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = oneapi::mkl::lapack::unmtr( + sycl::event func_event = oneapi::math::lapack::unmtr( queue, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::mkl::lapack::unmtr, side, uplo, trans, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::unmtr, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif diff --git a/tests/unit_tests/rng/device/include/moments.hpp b/tests/unit_tests/rng/device/include/moments.hpp index df1441ff6..661e85f32 100644 --- a/tests/unit_tests/rng/device/include/moments.hpp +++ b/tests/unit_tests/rng/device/include/moments.hpp @@ -20,7 +20,7 @@ /* * * Content: -* oneapi::mkl::rng::device:: distributions moments test (SYCL interface) +* oneapi::math::rng::device:: distributions moments test (SYCL interface) * *******************************************************************************/ @@ -47,19 +47,19 @@ class moments_test { // Note: the following methods of discrete distributions require double precision support if ((std::is_same_v< Distribution, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>> || + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>> || std::is_same_v< Distribution, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>> || - std::is_same_v> || + std::is_same_v> || + oneapi::math::rng::device::poisson_method::devroye>> || std::is_same_v< Distribution, - oneapi::mkl::rng::device::poisson< - std::int32_t, oneapi::mkl::rng::device::poisson_method::devroye>>)&&!queue + oneapi::math::rng::device::poisson< + std::int32_t, oneapi::math::rng::device::poisson_method::devroye>>)&&!queue .get_device() .has(sycl::aspect::fp64)) { status = test_skipped; @@ -79,11 +79,11 @@ class moments_test { size_t id = item.get_id(0); auto multiplier = Engine::vec_size; if constexpr (std::is_same_v>) + oneapi::math::rng::device::uniform_bits>) multiplier *= 2; Engine engine(SEED, id * multiplier); Distribution distr; - auto res = oneapi::mkl::rng::device::generate(distr, engine); + auto res = oneapi::math::rng::device::generate(distr, engine); if constexpr (Engine::vec_size == 1) { acc[id] = res; } @@ -94,7 +94,7 @@ class moments_test { }); event.wait_and_throw(); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { status = test_skipped; return; } @@ -108,7 +108,7 @@ class moments_test { // validation (statistics check is turned out for mcg59) if constexpr (!std::is_same>::value) { + oneapi::math::rng::device::mcg59>::value) { statistics_device stat; status = stat.check(r, Distribution{}); } diff --git a/tests/unit_tests/rng/device/include/rng_device_test_common.hpp b/tests/unit_tests/rng/device/include/rng_device_test_common.hpp index 33533255e..ac646fdeb 100644 --- a/tests/unit_tests/rng/device/include/rng_device_test_common.hpp +++ b/tests/unit_tests/rng/device/include/rng_device_test_common.hpp @@ -171,10 +171,10 @@ template struct statistics_device {}; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::uniform& distr) { + const oneapi::math::rng::device::uniform& distr) { double tM, tD, tQ; Fp a = distr.a(); Fp b = distr.b(); @@ -189,10 +189,10 @@ struct statistics_device> { }; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::uniform& distr) { + const oneapi::math::rng::device::uniform& distr) { double tM, tD, tQ; double a = distr.a(); double b = distr.b(); @@ -208,10 +208,10 @@ struct statistics_device }; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::uniform& distr) { + const oneapi::math::rng::device::uniform& distr) { double tM, tD, tQ; double a = distr.a(); double b = distr.b(); @@ -227,10 +227,10 @@ struct statistics_device -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::uniform& distr) { + const oneapi::math::rng::device::uniform& distr) { double tM, tD, tQ; double a = distr.a(); double b = distr.b(); @@ -246,10 +246,10 @@ struct statistics_device }; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::uniform& distr) { + const oneapi::math::rng::device::uniform& distr) { double tM, tD, tQ; double a = distr.a(); double b = distr.b(); @@ -265,10 +265,10 @@ struct statistics_device -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::gaussian& distr) { + const oneapi::math::rng::device::gaussian& distr) { double tM, tD, tQ; Fp a = distr.mean(); Fp sigma = distr.stddev(); @@ -283,10 +283,10 @@ struct statistics_device> { }; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::lognormal& distr) { + const oneapi::math::rng::device::lognormal& distr) { double tM, tD, tQ; Fp a = distr.m(); Fp b = distr.displ(); @@ -305,10 +305,10 @@ struct statistics_device> { }; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::exponential& distr) { + const oneapi::math::rng::device::exponential& distr) { double tM, tD, tQ; Fp a = distr.a(); Fp beta = distr.beta(); @@ -322,10 +322,10 @@ struct statistics_device> { }; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::poisson& distr) { + const oneapi::math::rng::device::poisson& distr) { double tM, tD, tQ; double lambda = distr.lambda(); @@ -338,10 +338,10 @@ struct statistics_device> { }; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::bernoulli& distr) { + const oneapi::math::rng::device::bernoulli& distr) { double tM, tD, tQ; double p = static_cast(distr.p()); @@ -354,10 +354,10 @@ struct statistics_device> { }; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::beta& distr) { + const oneapi::math::rng::device::beta& distr) { double tM, tD, tQ; double b, c, d, e, e2, b2, sum_pq; Fp p = distr.p(); @@ -382,10 +382,10 @@ struct statistics_device> { }; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::gamma& distr) { + const oneapi::math::rng::device::gamma& distr) { double tM, tD, tQ; Fp a = distr.a(); Fp alpha = distr.alpha(); @@ -400,19 +400,19 @@ struct statistics_device> { }; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::bits& distr) { + const oneapi::math::rng::device::bits& distr) { return true; } }; template -struct statistics_device> { +struct statistics_device> { template bool check(const std::vector& r, - const oneapi::mkl::rng::device::uniform_bits& distr) { + const oneapi::math::rng::device::uniform_bits& distr) { return true; } }; @@ -421,6 +421,6 @@ template struct is_mcg59 : std::false_type {}; template -struct is_mcg59> : std::true_type {}; +struct is_mcg59> : std::true_type {}; #endif // _RNG_DEVICE_TEST_COMMON_HPP__ diff --git a/tests/unit_tests/rng/device/include/skip_ahead_test.hpp b/tests/unit_tests/rng/device/include/skip_ahead_test.hpp index d1ff3dc8e..fb8648248 100644 --- a/tests/unit_tests/rng/device/include/skip_ahead_test.hpp +++ b/tests/unit_tests/rng/device/include/skip_ahead_test.hpp @@ -20,7 +20,7 @@ /* * * Content: -* oneapi::mkl::rng::device:: engines skip_ahead and skip_ahead_ex tests +* oneapi::math::rng::device:: engines skip_ahead and skip_ahead_ex tests * (SYCL interface) * *******************************************************************************/ @@ -61,9 +61,9 @@ class skip_ahead_test { cgh.parallel_for(range, [=](sycl::item<1> item) { size_t id = item.get_id(0); Engine engine(SEED); - oneapi::mkl::rng::device::skip_ahead(engine, id * Engine::vec_size); - oneapi::mkl::rng::device::bits distr; - auto res = oneapi::mkl::rng::device::generate(distr, engine); + oneapi::math::rng::device::skip_ahead(engine, id * Engine::vec_size); + oneapi::math::rng::device::bits distr; + auto res = oneapi::math::rng::device::generate(distr, engine); if constexpr (Engine::vec_size == 1) { acc[id] = res; } @@ -74,7 +74,7 @@ class skip_ahead_test { }); event.wait_and_throw(); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { status = test_skipped; return; } @@ -88,9 +88,9 @@ class skip_ahead_test { // validation Engine engine(SEED); - oneapi::mkl::rng::device::bits distr; + oneapi::math::rng::device::bits distr; for (int i = 0; i < N_GEN; i += Engine::vec_size) { - auto res = oneapi::mkl::rng::device::generate(distr, engine); + auto res = oneapi::math::rng::device::generate(distr, engine); if constexpr (Engine::vec_size == 1) { r_ref[i] = res; } @@ -125,10 +125,10 @@ class skip_ahead_ex_test { cgh.parallel_for(range, [=](sycl::item<1> item) { size_t id = item.get_id(0); Engine engine(SEED); - oneapi::mkl::rng::device::skip_ahead(engine, + oneapi::math::rng::device::skip_ahead(engine, { id * Engine::vec_size, skip_num }); - oneapi::mkl::rng::device::bits<> distr; - auto res = oneapi::mkl::rng::device::generate(distr, engine); + oneapi::math::rng::device::bits<> distr; + auto res = oneapi::math::rng::device::generate(distr, engine); if constexpr (Engine::vec_size == 1) { acc[id] = res; } @@ -139,7 +139,7 @@ class skip_ahead_ex_test { }); event.wait_and_throw(); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { status = test_skipped; return; } @@ -154,11 +154,11 @@ class skip_ahead_ex_test { // validation Engine engine(SEED); for (int j = 0; j < SKIP_TIMES; j++) { - oneapi::mkl::rng::device::skip_ahead(engine, N_SKIP); + oneapi::math::rng::device::skip_ahead(engine, N_SKIP); } - oneapi::mkl::rng::device::bits<> distr; + oneapi::math::rng::device::bits<> distr; for (int i = 0; i < N_GEN; i += Engine::vec_size) { - auto res = oneapi::mkl::rng::device::generate(distr, engine); + auto res = oneapi::math::rng::device::generate(distr, engine); if constexpr (Engine::vec_size == 1) { r_ref[i] = res; } diff --git a/tests/unit_tests/rng/device/moments/moments.cpp b/tests/unit_tests/rng/device/moments/moments.cpp index 3ae45d657..6449b2d0c 100644 --- a/tests/unit_tests/rng/device/moments/moments.cpp +++ b/tests/unit_tests/rng/device/moments/moments.cpp @@ -30,19 +30,19 @@ class Philox4x32x10UniformStdDeviceMomentsTests : public ::testing::TestWithPara class Philox4x32x10UniformAccDeviceMomentsTests : public ::testing::TestWithParam {}; TEST_P(Philox4x32x10UniformStdDeviceMomentsTests, RealSinglePrecision) { - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -50,61 +50,61 @@ TEST_P(Philox4x32x10UniformStdDeviceMomentsTests, RealSinglePrecision) { TEST_P(Philox4x32x10UniformStdDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10UniformStdDeviceMomentsTests, IntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10UniformStdDeviceMomentsTests, UnsignedIntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -113,21 +113,21 @@ TEST_P(Philox4x32x10UniformStdDeviceMomentsTests, Integer64Precision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int64_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int64_t, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int64_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int64_t, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int64_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int64_t, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -136,39 +136,39 @@ TEST_P(Philox4x32x10UniformStdDeviceMomentsTests, UnsignedInteger64Precision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint64_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint64_t, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint64_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint64_t, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint64_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint64_t, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10UniformAccDeviceMomentsTests, RealSinglePrecision) { - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -176,61 +176,61 @@ TEST_P(Philox4x32x10UniformAccDeviceMomentsTests, RealSinglePrecision) { TEST_P(Philox4x32x10UniformAccDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10UniformAccDeviceMomentsTests, IntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10UniformAccDeviceMomentsTests, UnsignedIntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -239,21 +239,21 @@ TEST_P(Philox4x32x10UniformAccDeviceMomentsTests, Integer64Precision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int64_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int64_t, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int64_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int64_t, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int64_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int64_t, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -262,21 +262,21 @@ TEST_P(Philox4x32x10UniformAccDeviceMomentsTests, UnsignedInteger64Precision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint64_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint64_t, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint64_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint64_t, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint64_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint64_t, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -294,19 +294,19 @@ class Mrg32k3aUniformStdDeviceMomentsTests : public ::testing::TestWithParam {}; TEST_P(Mrg32k3aUniformStdDeviceMomentsTests, RealSinglePrecision) { - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -314,79 +314,79 @@ TEST_P(Mrg32k3aUniformStdDeviceMomentsTests, RealSinglePrecision) { TEST_P(Mrg32k3aUniformStdDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mrg32k3aUniformStdDeviceMomentsTests, IntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mrg32k3aUniformStdDeviceMomentsTests, UnsignedIntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mrg32k3aUniformAccDeviceMomentsTests, RealSinglePrecision) { - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -394,61 +394,61 @@ TEST_P(Mrg32k3aUniformAccDeviceMomentsTests, RealSinglePrecision) { TEST_P(Mrg32k3aUniformAccDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mrg32k3aUniformAccDeviceMomentsTests, IntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mrg32k3aUniformAccDeviceMomentsTests, UnsignedIntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -466,19 +466,19 @@ class Mcg31m1UniformStdDeviceMomentsTests : public ::testing::TestWithParam {}; TEST_P(Mcg31m1UniformStdDeviceMomentsTests, RealSinglePrecision) { - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -486,79 +486,79 @@ TEST_P(Mcg31m1UniformStdDeviceMomentsTests, RealSinglePrecision) { TEST_P(Mcg31m1UniformStdDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mcg31m1UniformStdDeviceMomentsTests, IntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mcg31m1UniformStdDeviceMomentsTests, UnsignedIntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mcg31m1UniformAccDeviceMomentsTests, RealSinglePrecision) { - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -566,61 +566,61 @@ TEST_P(Mcg31m1UniformAccDeviceMomentsTests, RealSinglePrecision) { TEST_P(Mcg31m1UniformAccDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mcg31m1UniformAccDeviceMomentsTests, IntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mcg31m1UniformAccDeviceMomentsTests, UnsignedIntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -638,19 +638,19 @@ class Mcg59UniformStdDeviceMomentsTests : public ::testing::TestWithParam {}; TEST_P(Mcg59UniformStdDeviceMomentsTests, RealSinglePrecision) { - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -658,79 +658,79 @@ TEST_P(Mcg59UniformStdDeviceMomentsTests, RealSinglePrecision) { TEST_P(Mcg59UniformStdDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::standard>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mcg59UniformStdDeviceMomentsTests, IntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mcg59UniformStdDeviceMomentsTests, UnsignedIntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::standard>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::standard>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mcg59UniformAccDeviceMomentsTests, RealSinglePrecision) { - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - float, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + float, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -738,61 +738,61 @@ TEST_P(Mcg59UniformAccDeviceMomentsTests, RealSinglePrecision) { TEST_P(Mcg59UniformAccDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform< - double, oneapi::mkl::rng::device::uniform_method::accurate>>> + rng_device_test, + oneapi::math::rng::device::uniform< + double, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mcg59UniformAccDeviceMomentsTests, IntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::int32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::int32_t, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mcg59UniformAccDeviceMomentsTests, UnsignedIntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::uniform< - std::uint32_t, oneapi::mkl::rng::device::uniform_method::accurate>>> + moments_test, + oneapi::math::rng::device::uniform< + std::uint32_t, oneapi::math::rng::device::uniform_method::accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -806,16 +806,16 @@ INSTANTIATE_TEST_SUITE_P(Mcg59UniformAccDeviceMomentsTestsSuite, Mcg59UniformAcc class Philox4x32x10BitsDeviceMomentsTests : public ::testing::TestWithParam {}; TEST_P(Philox4x32x10BitsDeviceMomentsTests, UnsignedIntegerPrecision) { - rng_device_test, - oneapi::mkl::rng::device::bits>> + rng_device_test, + oneapi::math::rng::device::bits>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::bits>> + rng_device_test, + oneapi::math::rng::device::bits>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::bits>> + rng_device_test, + oneapi::math::rng::device::bits>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -828,31 +828,31 @@ class Philox4x32x10UniformBitsDeviceMomentsTests : public ::testing::TestWithPar }; TEST_P(Philox4x32x10UniformBitsDeviceMomentsTests, UnsignedIntegerPrecision) { - rng_device_test, - oneapi::mkl::rng::device::uniform_bits>> + rng_device_test, + oneapi::math::rng::device::uniform_bits>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform_bits>> + rng_device_test, + oneapi::math::rng::device::uniform_bits>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform_bits>> + rng_device_test, + oneapi::math::rng::device::uniform_bits>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10UniformBitsDeviceMomentsTests, UnsignedLongIntegerPrecision) { - rng_device_test, - oneapi::mkl::rng::device::uniform_bits>> + rng_device_test, + oneapi::math::rng::device::uniform_bits>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform_bits>> + rng_device_test, + oneapi::math::rng::device::uniform_bits>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::uniform_bits>> + rng_device_test, + oneapi::math::rng::device::uniform_bits>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -869,39 +869,39 @@ TEST_P(Philox4x32x10GaussianBoxMuller2DeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_device_test< - moments_test, - oneapi::mkl::rng::device::gaussian< - float, oneapi::mkl::rng::device::gaussian_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::gaussian< + float, oneapi::math::rng::device::gaussian_method::box_muller2>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::gaussian< - float, oneapi::mkl::rng::device::gaussian_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::gaussian< + float, oneapi::math::rng::device::gaussian_method::box_muller2>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::gaussian< - float, oneapi::mkl::rng::device::gaussian_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::gaussian< + float, oneapi::math::rng::device::gaussian_method::box_muller2>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::gaussian< - double, oneapi::mkl::rng::device::gaussian_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::gaussian< + double, oneapi::math::rng::device::gaussian_method::box_muller2>>> test4; EXPECT_TRUEORSKIP((test4(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::gaussian< - double, oneapi::mkl::rng::device::gaussian_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::gaussian< + double, oneapi::math::rng::device::gaussian_method::box_muller2>>> test5; EXPECT_TRUEORSKIP((test5(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::gaussian< - double, oneapi::mkl::rng::device::gaussian_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::gaussian< + double, oneapi::math::rng::device::gaussian_method::box_muller2>>> test6; EXPECT_TRUEORSKIP((test6(GetParam()))); } @@ -915,21 +915,21 @@ class Philox4x32x10LognormalBoxMuller2DeviceMomentsTests TEST_P(Philox4x32x10LognormalBoxMuller2DeviceMomentsTests, RealSinglePrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::lognormal< - float, oneapi::mkl::rng::device::lognormal_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::lognormal< + float, oneapi::math::rng::device::lognormal_method::box_muller2>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::lognormal< - float, oneapi::mkl::rng::device::lognormal_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::lognormal< + float, oneapi::math::rng::device::lognormal_method::box_muller2>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::lognormal< - float, oneapi::mkl::rng::device::lognormal_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::lognormal< + float, oneapi::math::rng::device::lognormal_method::box_muller2>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -938,21 +938,21 @@ TEST_P(Philox4x32x10LognormalBoxMuller2DeviceMomentsTests, RealDoublePrecision) CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_device_test< - moments_test, - oneapi::mkl::rng::device::lognormal< - double, oneapi::mkl::rng::device::lognormal_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::lognormal< + double, oneapi::math::rng::device::lognormal_method::box_muller2>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::lognormal< - double, oneapi::mkl::rng::device::lognormal_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::lognormal< + double, oneapi::math::rng::device::lognormal_method::box_muller2>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::lognormal< - double, oneapi::mkl::rng::device::lognormal_method::box_muller2>>> + moments_test, + oneapi::math::rng::device::lognormal< + double, oneapi::math::rng::device::lognormal_method::box_muller2>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -971,34 +971,34 @@ class Philox4x32x10ExponentialIcdfAccDeviceMomentsTests TEST_P(Philox4x32x10ExponentialIcdfDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_device_test, - oneapi::mkl::rng::device::exponential< - float, oneapi::mkl::rng::device::exponential_method::icdf>>> + rng_device_test, + oneapi::math::rng::device::exponential< + float, oneapi::math::rng::device::exponential_method::icdf>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::exponential< - float, oneapi::mkl::rng::device::exponential_method::icdf>>> + rng_device_test, + oneapi::math::rng::device::exponential< + float, oneapi::math::rng::device::exponential_method::icdf>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::exponential< - float, oneapi::mkl::rng::device::exponential_method::icdf>>> + rng_device_test, + oneapi::math::rng::device::exponential< + float, oneapi::math::rng::device::exponential_method::icdf>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::exponential< - double, oneapi::mkl::rng::device::exponential_method::icdf>>> + rng_device_test, + oneapi::math::rng::device::exponential< + double, oneapi::math::rng::device::exponential_method::icdf>>> test4; EXPECT_TRUEORSKIP((test4(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::exponential< - double, oneapi::mkl::rng::device::exponential_method::icdf>>> + rng_device_test, + oneapi::math::rng::device::exponential< + double, oneapi::math::rng::device::exponential_method::icdf>>> test5; EXPECT_TRUEORSKIP((test5(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::exponential< - double, oneapi::mkl::rng::device::exponential_method::icdf>>> + rng_device_test, + oneapi::math::rng::device::exponential< + double, oneapi::math::rng::device::exponential_method::icdf>>> test6; EXPECT_TRUEORSKIP((test6(GetParam()))); } @@ -1008,39 +1008,39 @@ TEST_P(Philox4x32x10ExponentialIcdfAccDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_device_test< - moments_test, - oneapi::mkl::rng::device::exponential< - float, oneapi::mkl::rng::device::exponential_method::icdf_accurate>>> + moments_test, + oneapi::math::rng::device::exponential< + float, oneapi::math::rng::device::exponential_method::icdf_accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::exponential< - float, oneapi::mkl::rng::device::exponential_method::icdf_accurate>>> + moments_test, + oneapi::math::rng::device::exponential< + float, oneapi::math::rng::device::exponential_method::icdf_accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::exponential< - float, oneapi::mkl::rng::device::exponential_method::icdf_accurate>>> + moments_test, + oneapi::math::rng::device::exponential< + float, oneapi::math::rng::device::exponential_method::icdf_accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::exponential< - double, oneapi::mkl::rng::device::exponential_method::icdf_accurate>>> + moments_test, + oneapi::math::rng::device::exponential< + double, oneapi::math::rng::device::exponential_method::icdf_accurate>>> test4; EXPECT_TRUEORSKIP((test4(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::exponential< - double, oneapi::mkl::rng::device::exponential_method::icdf_accurate>>> + moments_test, + oneapi::math::rng::device::exponential< + double, oneapi::math::rng::device::exponential_method::icdf_accurate>>> test5; EXPECT_TRUEORSKIP((test5(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::exponential< - double, oneapi::mkl::rng::device::exponential_method::icdf_accurate>>> + moments_test, + oneapi::math::rng::device::exponential< + double, oneapi::math::rng::device::exponential_method::icdf_accurate>>> test6; EXPECT_TRUEORSKIP((test6(GetParam()))); } @@ -1059,18 +1059,18 @@ class Philox4x32x10BetaCjaAccDeviceMomentsTests : public ::testing::TestWithPara TEST_P(Philox4x32x10BetaCjaDeviceMomentsTests, RealSinglePrecision) { rng_device_test, - oneapi::mkl::rng::device::beta>> + oneapi::math::rng::device::philox4x32x10<1>, + oneapi::math::rng::device::beta>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test, - oneapi::mkl::rng::device::beta>> + oneapi::math::rng::device::philox4x32x10<4>, + oneapi::math::rng::device::beta>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test, - oneapi::mkl::rng::device::beta>> + oneapi::math::rng::device::philox4x32x10<16>, + oneapi::math::rng::device::beta>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -1079,36 +1079,36 @@ TEST_P(Philox4x32x10BetaCjaDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_device_test, - oneapi::mkl::rng::device::beta>> + oneapi::math::rng::device::philox4x32x10<1>, + oneapi::math::rng::device::beta>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test, - oneapi::mkl::rng::device::beta>> + oneapi::math::rng::device::philox4x32x10<4>, + oneapi::math::rng::device::beta>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test, - oneapi::mkl::rng::device::beta>> + oneapi::math::rng::device::philox4x32x10<16>, + oneapi::math::rng::device::beta>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10BetaCjaAccDeviceMomentsTests, RealSinglePrecision) { rng_device_test, - oneapi::mkl::rng::device::beta>> + oneapi::math::rng::device::philox4x32x10<1>, + oneapi::math::rng::device::beta>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test, - oneapi::mkl::rng::device::beta>> + oneapi::math::rng::device::philox4x32x10<4>, + oneapi::math::rng::device::beta>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test, - oneapi::mkl::rng::device::beta>> + oneapi::math::rng::device::philox4x32x10<16>, + oneapi::math::rng::device::beta>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -1116,19 +1116,19 @@ TEST_P(Philox4x32x10BetaCjaAccDeviceMomentsTests, RealSinglePrecision) { TEST_P(Philox4x32x10BetaCjaAccDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_device_test, - oneapi::mkl::rng::device::beta< - double, oneapi::mkl::rng::device::beta_method::cja_accurate>>> + rng_device_test, + oneapi::math::rng::device::beta< + double, oneapi::math::rng::device::beta_method::cja_accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::beta< - double, oneapi::mkl::rng::device::beta_method::cja_accurate>>> + rng_device_test, + oneapi::math::rng::device::beta< + double, oneapi::math::rng::device::beta_method::cja_accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::beta< - double, oneapi::mkl::rng::device::beta_method::cja_accurate>>> + rng_device_test, + oneapi::math::rng::device::beta< + double, oneapi::math::rng::device::beta_method::cja_accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -1149,18 +1149,18 @@ class Philox4x32x10GammaMarsagliaAccDeviceMomentsTests TEST_P(Philox4x32x10GammaMarsagliaDeviceMomentsTests, RealSinglePrecision) { rng_device_test, - oneapi::mkl::rng::device::gamma>> + oneapi::math::rng::device::philox4x32x10<1>, + oneapi::math::rng::device::gamma>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test, - oneapi::mkl::rng::device::gamma>> + oneapi::math::rng::device::philox4x32x10<4>, + oneapi::math::rng::device::gamma>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test, - oneapi::mkl::rng::device::gamma>> + oneapi::math::rng::device::philox4x32x10<16>, + oneapi::math::rng::device::gamma>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -1169,39 +1169,39 @@ TEST_P(Philox4x32x10GammaMarsagliaDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_device_test, - oneapi::mkl::rng::device::gamma>> + oneapi::math::rng::device::philox4x32x10<1>, + oneapi::math::rng::device::gamma>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test, - oneapi::mkl::rng::device::gamma>> + oneapi::math::rng::device::philox4x32x10<4>, + oneapi::math::rng::device::gamma>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test, - oneapi::mkl::rng::device::gamma>> + oneapi::math::rng::device::philox4x32x10<16>, + oneapi::math::rng::device::gamma>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10GammaMarsagliaAccDeviceMomentsTests, RealSinglePrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::gamma< - float, oneapi::mkl::rng::device::gamma_method::marsaglia_accurate>>> + moments_test, + oneapi::math::rng::device::gamma< + float, oneapi::math::rng::device::gamma_method::marsaglia_accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::gamma< - float, oneapi::mkl::rng::device::gamma_method::marsaglia_accurate>>> + moments_test, + oneapi::math::rng::device::gamma< + float, oneapi::math::rng::device::gamma_method::marsaglia_accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::gamma< - float, oneapi::mkl::rng::device::gamma_method::marsaglia_accurate>>> + moments_test, + oneapi::math::rng::device::gamma< + float, oneapi::math::rng::device::gamma_method::marsaglia_accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -1210,21 +1210,21 @@ TEST_P(Philox4x32x10GammaMarsagliaAccDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_device_test< - moments_test, - oneapi::mkl::rng::device::gamma< - double, oneapi::mkl::rng::device::gamma_method::marsaglia_accurate>>> + moments_test, + oneapi::math::rng::device::gamma< + double, oneapi::math::rng::device::gamma_method::marsaglia_accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::gamma< - double, oneapi::mkl::rng::device::gamma_method::marsaglia_accurate>>> + moments_test, + oneapi::math::rng::device::gamma< + double, oneapi::math::rng::device::gamma_method::marsaglia_accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::gamma< - double, oneapi::mkl::rng::device::gamma_method::marsaglia_accurate>>> + moments_test, + oneapi::math::rng::device::gamma< + double, oneapi::math::rng::device::gamma_method::marsaglia_accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -1241,40 +1241,40 @@ class Philox4x32x10PoissonDevroyeDeviceMomentsTests : public ::testing::TestWithParam {}; TEST_P(Philox4x32x10PoissonDevroyeDeviceMomentsTests, IntegerPrecision) { - rng_device_test, - oneapi::mkl::rng::device::poisson< - int32_t, oneapi::mkl::rng::device::poisson_method::devroye>>> + rng_device_test, + oneapi::math::rng::device::poisson< + int32_t, oneapi::math::rng::device::poisson_method::devroye>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::poisson< - int32_t, oneapi::mkl::rng::device::poisson_method::devroye>>> + rng_device_test, + oneapi::math::rng::device::poisson< + int32_t, oneapi::math::rng::device::poisson_method::devroye>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::mkl::rng::device::poisson< - int32_t, oneapi::mkl::rng::device::poisson_method::devroye>>> + rng_device_test, + oneapi::math::rng::device::poisson< + int32_t, oneapi::math::rng::device::poisson_method::devroye>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10PoissonDevroyeDeviceMomentsTests, UnsignedIntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::poisson< - std::uint32_t, oneapi::mkl::rng::device::poisson_method::devroye>>> + moments_test, + oneapi::math::rng::device::poisson< + std::uint32_t, oneapi::math::rng::device::poisson_method::devroye>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::poisson< - std::uint32_t, oneapi::mkl::rng::device::poisson_method::devroye>>> + moments_test, + oneapi::math::rng::device::poisson< + std::uint32_t, oneapi::math::rng::device::poisson_method::devroye>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::poisson< - std::uint32_t, oneapi::mkl::rng::device::poisson_method::devroye>>> + moments_test, + oneapi::math::rng::device::poisson< + std::uint32_t, oneapi::math::rng::device::poisson_method::devroye>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -1288,126 +1288,126 @@ class Philox4x32x10BernoulliIcdfDeviceMomentsTests TEST_P(Philox4x32x10BernoulliIcdfDeviceMomentsTests, IntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::int32_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::int32_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::int32_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::int32_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::int32_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::int32_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10BernoulliIcdfDeviceMomentsTests, UnsignedIntegerPrecision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::uint32_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::uint32_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::uint32_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::uint32_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::uint32_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::uint32_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10BernoulliIcdfDeviceMomentsTests, Integer8Precision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::int8_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::int8_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::int8_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::int8_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::int8_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::int8_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10BernoulliIcdfDeviceMomentsTests, UnsignedInteger8Precision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::uint8_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::uint8_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::uint8_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::uint8_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::uint8_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::uint8_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10BernoulliIcdfDeviceMomentsTests, Integer16Precision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::int16_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::int16_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::int16_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::int16_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::int16_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::int16_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10BernoulliIcdfDeviceMomentsTests, UnsignedInteger16Precision) { rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::uint16_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::uint16_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::uint16_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::uint16_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); rng_device_test< - moments_test, - oneapi::mkl::rng::device::bernoulli< - std::uint16_t, oneapi::mkl::rng::device::bernoulli_method::icdf>>> + moments_test, + oneapi::math::rng::device::bernoulli< + std::uint16_t, oneapi::math::rng::device::bernoulli_method::icdf>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } diff --git a/tests/unit_tests/rng/device/service/skip_ahead.cpp b/tests/unit_tests/rng/device/service/skip_ahead.cpp index a5dfe0da8..662c56cff 100644 --- a/tests/unit_tests/rng/device/service/skip_ahead.cpp +++ b/tests/unit_tests/rng/device/service/skip_ahead.cpp @@ -30,20 +30,20 @@ class Philox4x32x10DeviceSkipAheadTests : public ::testing::TestWithParam {}; TEST_P(Philox4x32x10DeviceSkipAheadTests, BinaryPrecision) { - rng_device_test>> test1; + rng_device_test>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test>> test2; + rng_device_test>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test>> test3; + rng_device_test>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Philox4x32x10DeviceSkipAheadExTests, BinaryPrecision) { - rng_device_test>> test1; + rng_device_test>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test>> test2; + rng_device_test>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test>> test3; + rng_device_test>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -59,20 +59,20 @@ class Mrg32k3aDeviceSkipAheadTests : public ::testing::TestWithParam {}; TEST_P(Mrg32k3aDeviceSkipAheadTests, BinaryPrecision) { - rng_device_test>> test1; + rng_device_test>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test>> test2; + rng_device_test>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test>> test3; + rng_device_test>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } TEST_P(Mrg32k3aDeviceSkipAheadExTests, BinaryPrecision) { - rng_device_test>> test1; + rng_device_test>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test>> test2; + rng_device_test>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test>> test3; + rng_device_test>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -85,11 +85,11 @@ INSTANTIATE_TEST_SUITE_P(Mrg32k3aDeviceSkipAheadExTestsSuite, Mrg32k3aDeviceSkip class Mcg31m1DeviceSkipAheadTests : public ::testing::TestWithParam {}; TEST_P(Mcg31m1DeviceSkipAheadTests, BinaryPrecision) { - rng_device_test>> test1; + rng_device_test>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test>> test2; + rng_device_test>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test>> test3; + rng_device_test>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -99,11 +99,11 @@ INSTANTIATE_TEST_SUITE_P(Mcg31m1DeviceSkipAheadTestsSuite, Mcg31m1DeviceSkipAhea class Mcg59DeviceSkipAheadTests : public ::testing::TestWithParam {}; TEST_P(Mcg59DeviceSkipAheadTests, BinaryPrecision) { - rng_device_test>> test1; + rng_device_test>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test>> test2; + rng_device_test>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test>> test3; + rng_device_test>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } diff --git a/tests/unit_tests/rng/include/engines_api_tests.hpp b/tests/unit_tests/rng/include/engines_api_tests.hpp index 567707afc..0acdd4b98 100644 --- a/tests/unit_tests/rng/include/engines_api_tests.hpp +++ b/tests/unit_tests/rng/include/engines_api_tests.hpp @@ -52,20 +52,20 @@ class engines_constructors_test { Engine engine3(engine1); Engine engine4 = std::move(Engine(queue, SEED)); - oneapi::mkl::rng::bits distr; + oneapi::math::rng::bits distr; sycl::buffer r1_buffer(r1.data(), r1.size()); sycl::buffer r2_buffer(r2.data(), r2.size()); sycl::buffer r3_buffer(r3.data(), r3.size()); sycl::buffer r4_buffer(r4.data(), r4.size()); - oneapi::mkl::rng::generate(distr, engine1, N_GEN, r1_buffer); - oneapi::mkl::rng::generate(distr, engine2, N_GEN, r2_buffer); - oneapi::mkl::rng::generate(distr, engine3, N_GEN, r3_buffer); - oneapi::mkl::rng::generate(distr, engine4, N_GEN, r4_buffer); + oneapi::math::rng::generate(distr, engine1, N_GEN, r1_buffer); + oneapi::math::rng::generate(distr, engine2, N_GEN, r2_buffer); + oneapi::math::rng::generate(distr, engine3, N_GEN, r3_buffer); + oneapi::math::rng::generate(distr, engine4, N_GEN, r4_buffer); QUEUE_WAIT(queue); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { status = test_skipped; return; } @@ -99,13 +99,13 @@ class engines_copy_test { Engine engine1(queue, SEED); Engine engine2(engine1); - oneapi::mkl::rng::bits distr; + oneapi::math::rng::bits distr; { sycl::buffer r1_buffer(r1.data(), r1.size()); sycl::buffer r2_buffer(r2.data(), r2.size()); - oneapi::mkl::rng::generate(distr, engine1, N_GEN, r1_buffer); - oneapi::mkl::rng::generate(distr, engine2, N_GEN, r2_buffer); + oneapi::math::rng::generate(distr, engine1, N_GEN, r1_buffer); + oneapi::math::rng::generate(distr, engine2, N_GEN, r2_buffer); } Engine engine3 = engine1; @@ -115,13 +115,13 @@ class engines_copy_test { sycl::buffer r2_buffer(r2.data(), r2.size()); sycl::buffer r3_buffer(r3.data(), r3.size()); - oneapi::mkl::rng::generate(distr, engine1, N_GEN, r1_buffer); - oneapi::mkl::rng::generate(distr, engine3, N_GEN, r2_buffer); - oneapi::mkl::rng::generate(distr, engine4, N_GEN, r3_buffer); + oneapi::math::rng::generate(distr, engine1, N_GEN, r1_buffer); + oneapi::math::rng::generate(distr, engine3, N_GEN, r2_buffer); + oneapi::math::rng::generate(distr, engine4, N_GEN, r3_buffer); } QUEUE_WAIT(queue); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { status = test_skipped; return; } diff --git a/tests/unit_tests/rng/include/skip_ahead_test.hpp b/tests/unit_tests/rng/include/skip_ahead_test.hpp index 2cc657549..ecb65b92a 100644 --- a/tests/unit_tests/rng/include/skip_ahead_test.hpp +++ b/tests/unit_tests/rng/include/skip_ahead_test.hpp @@ -48,12 +48,12 @@ class skip_ahead_test { Engine engine(queue); std::vector engines; - oneapi::mkl::rng::bits distr; + oneapi::math::rng::bits distr; // Perform skip for (int i = 0; i < N_ENGINES; i++) { engines.push_back(new Engine(queue)); - oneapi::mkl::rng::skip_ahead(*(engines[i]), i * N_PORTION); + oneapi::math::rng::skip_ahead(*(engines[i]), i * N_PORTION); } sycl::buffer r_buffer(r1.data(), r1.size()); @@ -63,9 +63,9 @@ class skip_ahead_test { sycl::buffer(r2.data() + i * N_PORTION, N_PORTION)); } - oneapi::mkl::rng::generate(distr, engine, N_GEN_SERVICE, r_buffer); + oneapi::math::rng::generate(distr, engine, N_GEN_SERVICE, r_buffer); for (int i = 0; i < N_ENGINES; i++) { - oneapi::mkl::rng::generate(distr, *(engines[i]), N_PORTION, r_buffers[i]); + oneapi::math::rng::generate(distr, *(engines[i]), N_PORTION, r_buffers[i]); } QUEUE_WAIT(queue); @@ -74,7 +74,7 @@ class skip_ahead_test { delete engines[i]; } } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { status = test_skipped; return; } @@ -106,22 +106,22 @@ class skip_ahead_ex_test { Engine engine1(queue); Engine engine2(queue); - oneapi::mkl::rng::bits distr; + oneapi::math::rng::bits distr; // Perform skip for (int j = 0; j < SKIP_TIMES; j++) { - oneapi::mkl::rng::skip_ahead(engine1, N_SKIP); + oneapi::math::rng::skip_ahead(engine1, N_SKIP); } - oneapi::mkl::rng::skip_ahead(engine2, NUM_TO_SKIP); + oneapi::math::rng::skip_ahead(engine2, NUM_TO_SKIP); sycl::buffer r1_buffer(r1.data(), r1.size()); sycl::buffer r2_buffer(r2.data(), r2.size()); - oneapi::mkl::rng::generate(distr, engine1, N_GEN, r1_buffer); - oneapi::mkl::rng::generate(distr, engine2, N_GEN, r2_buffer); + oneapi::math::rng::generate(distr, engine1, N_GEN, r1_buffer); + oneapi::math::rng::generate(distr, engine2, N_GEN, r2_buffer); QUEUE_WAIT(queue); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { status = test_skipped; return; } diff --git a/tests/unit_tests/rng/include/statistics_check.hpp b/tests/unit_tests/rng/include/statistics_check.hpp index 2f8c68816..cb753460a 100644 --- a/tests/unit_tests/rng/include/statistics_check.hpp +++ b/tests/unit_tests/rng/include/statistics_check.hpp @@ -68,10 +68,10 @@ template struct statistics {}; template -struct statistics> { +struct statistics> { template bool check(std::vector& r, - const oneapi::mkl::rng::uniform& distr) { + const oneapi::math::rng::uniform& distr) { double tM, tD, tQ; Type a = distr.a(); Type b = distr.b(); @@ -86,10 +86,10 @@ struct statistics> { }; template -struct statistics> { +struct statistics> { template bool check(std::vector& r, - const oneapi::mkl::rng::uniform& distr) { + const oneapi::math::rng::uniform& distr) { double tM, tD, tQ; int32_t a = distr.a(); int32_t b = distr.b(); @@ -105,10 +105,10 @@ struct statistics> { }; template -struct statistics> { +struct statistics> { template bool check(std::vector& r, - const oneapi::mkl::rng::gaussian& distr) { + const oneapi::math::rng::gaussian& distr) { double tM, tD, tQ; Type a = distr.mean(); Type sigma = distr.stddev(); @@ -123,10 +123,10 @@ struct statistics> { }; template -struct statistics> { +struct statistics> { template bool check(std::vector& r, - const oneapi::mkl::rng::lognormal& distr) { + const oneapi::math::rng::lognormal& distr) { double tM, tD, tQ; Type a = distr.m(); Type b = distr.displ(); @@ -145,10 +145,10 @@ struct statistics> { }; template -struct statistics> { +struct statistics> { template bool check(std::vector& r, - const oneapi::mkl::rng::bernoulli& distr) { + const oneapi::math::rng::bernoulli& distr) { double tM, tD, tQ; double p = distr.p(); @@ -161,10 +161,10 @@ struct statistics> { }; template -struct statistics> { +struct statistics> { template bool check(std::vector& r, - const oneapi::mkl::rng::poisson& distr) { + const oneapi::math::rng::poisson& distr) { double tM, tD, tQ; double lambda = distr.lambda(); diff --git a/tests/unit_tests/rng/include/statistics_check_test.hpp b/tests/unit_tests/rng/include/statistics_check_test.hpp index 53ad61460..87493e09a 100644 --- a/tests/unit_tests/rng/include/statistics_check_test.hpp +++ b/tests/unit_tests/rng/include/statistics_check_test.hpp @@ -62,7 +62,7 @@ class statistics_test { Engine engine(queue, SEED); Distr distr(args...); - oneapi::mkl::rng::generate(distr, engine, n_gen, r_buffer); + oneapi::math::rng::generate(distr, engine, n_gen, r_buffer); QUEUE_WAIT(queue); } catch (sycl::exception const& e) { @@ -70,7 +70,7 @@ class statistics_test { << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { status = test_skipped; return; } @@ -101,7 +101,7 @@ class statistics_usm_test { try { Engine engine(queue, SEED); Distr distr(args...); - auto event = oneapi::mkl::rng::generate(distr, engine, n_gen, r.data()); + auto event = oneapi::math::rng::generate(distr, engine, n_gen, r.data()); event.wait_and_throw(); } catch (sycl::exception const& e) { @@ -109,7 +109,7 @@ class statistics_usm_test { << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::mkl::unimplemented& e) { + catch (const oneapi::math::unimplemented& e) { status = test_skipped; return; } diff --git a/tests/unit_tests/rng/service/engines_api_test.cpp b/tests/unit_tests/rng/service/engines_api_test.cpp index 500231703..9c17da89a 100644 --- a/tests/unit_tests/rng/service/engines_api_test.cpp +++ b/tests/unit_tests/rng/service/engines_api_test.cpp @@ -30,13 +30,13 @@ class Philox4x32x10ConstructorsTests : public ::testing::TestWithParam {}; TEST_P(Philox4x32x10ConstructorsTests, BinaryPrecision) { - rng_test> test; + rng_test> test; std::initializer_list seed_ex = { SEED, 0, 0 }; EXPECT_TRUEORSKIP((test(GetParam(), seed_ex))); } TEST_P(Philox4x32x10CopyTests, BinaryPrecision) { - rng_test> test; + rng_test> test; EXPECT_TRUEORSKIP((test(GetParam()))); } @@ -51,13 +51,13 @@ class Mrg32k3aConstructorsTests : public ::testing::TestWithParam class Mrg32k3aCopyTests : public ::testing::TestWithParam {}; TEST_P(Mrg32k3aConstructorsTests, BinaryPrecision) { - rng_test> test; + rng_test> test; std::initializer_list seed_ex = { SEED, 1, 1, 1, 1, 1 }; EXPECT_TRUEORSKIP((test(GetParam(), seed_ex))); } TEST_P(Mrg32k3aCopyTests, BinaryPrecision) { - rng_test> test; + rng_test> test; EXPECT_TRUEORSKIP((test(GetParam()))); } diff --git a/tests/unit_tests/rng/service/skip_ahead.cpp b/tests/unit_tests/rng/service/skip_ahead.cpp index 445b76abe..abd3d7ffd 100644 --- a/tests/unit_tests/rng/service/skip_ahead.cpp +++ b/tests/unit_tests/rng/service/skip_ahead.cpp @@ -30,12 +30,12 @@ class Philox4x32x10SkipAheadTests : public ::testing::TestWithParam {}; TEST_P(Philox4x32x10SkipAheadTests, BinaryPrecision) { - rng_test> test; + rng_test> test; EXPECT_TRUEORSKIP((test(GetParam()))); } TEST_P(Philox4x32x10SkipAheadExTests, BinaryPrecision) { - rng_test> test; + rng_test> test; EXPECT_TRUEORSKIP((test(GetParam()))); } @@ -50,12 +50,12 @@ class Mrg32k3aSkipAheadTests : public ::testing::TestWithParam {} class Mrg32k3aSkipAheadExTests : public ::testing::TestWithParam {}; TEST_P(Mrg32k3aSkipAheadTests, BinaryPrecision) { - rng_test> test; + rng_test> test; EXPECT_TRUEORSKIP((test(GetParam()))); } TEST_P(Mrg32k3aSkipAheadExTests, BinaryPrecision) { - rng_test> test; + rng_test> test; EXPECT_TRUEORSKIP((test(GetParam()))); } diff --git a/tests/unit_tests/rng/statistics_check/bernoulli.cpp b/tests/unit_tests/rng/statistics_check/bernoulli.cpp index b95d98118..87166dbf3 100755 --- a/tests/unit_tests/rng/statistics_check/bernoulli.cpp +++ b/tests/unit_tests/rng/statistics_check/bernoulli.cpp @@ -29,26 +29,26 @@ class BernoulliIcdfTests : public ::testing::TestWithParam {}; TEST_P(BernoulliIcdfTests, IntegerPrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::bernoulli, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, BERNOULLI_ARGS))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::bernoulli, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, BERNOULLI_ARGS))); } TEST_P(BernoulliIcdfTests, UnsignedIntegerPrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::bernoulli, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, BERNOULLI_ARGS))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::bernoulli, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, BERNOULLI_ARGS))); } diff --git a/tests/unit_tests/rng/statistics_check/bernoulli_usm.cpp b/tests/unit_tests/rng/statistics_check/bernoulli_usm.cpp index 9c8c934dd..ad0664500 100755 --- a/tests/unit_tests/rng/statistics_check/bernoulli_usm.cpp +++ b/tests/unit_tests/rng/statistics_check/bernoulli_usm.cpp @@ -29,26 +29,26 @@ class BernoulliIcdfUsmTests : public ::testing::TestWithParam {}; TEST_P(BernoulliIcdfUsmTests, IntegerPrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::bernoulli, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, BERNOULLI_ARGS))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::bernoulli, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, BERNOULLI_ARGS))); } TEST_P(BernoulliIcdfUsmTests, UnsignedIntegerPrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::bernoulli, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, BERNOULLI_ARGS))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::bernoulli, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, BERNOULLI_ARGS))); } diff --git a/tests/unit_tests/rng/statistics_check/gaussian.cpp b/tests/unit_tests/rng/statistics_check/gaussian.cpp index ed63f3221..9b693c44d 100644 --- a/tests/unit_tests/rng/statistics_check/gaussian.cpp +++ b/tests/unit_tests/rng/statistics_check/gaussian.cpp @@ -31,13 +31,13 @@ class GaussianIcdfTest : public ::testing::TestWithParam {}; TEST_P(GaussianIcdfTest, RealSinglePrecision) { rng_test< - statistics_test, - oneapi::mkl::rng::philox4x32x10>> + statistics_test, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, GAUSSIAN_ARGS_FLOAT))); rng_test< - statistics_test, - oneapi::mkl::rng::mrg32k3a>> + statistics_test, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, GAUSSIAN_ARGS_FLOAT))); } @@ -46,26 +46,26 @@ TEST_P(GaussianIcdfTest, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test< - statistics_test, - oneapi::mkl::rng::philox4x32x10>> + statistics_test, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, GAUSSIAN_ARGS_DOUBLE))); rng_test< - statistics_test, - oneapi::mkl::rng::mrg32k3a>> + statistics_test, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, GAUSSIAN_ARGS_DOUBLE))); } TEST_P(GaussianBoxmullerTest, RealSinglePrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::gaussian, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, GAUSSIAN_ARGS_FLOAT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::gaussian, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, GAUSSIAN_ARGS_FLOAT))); } @@ -74,13 +74,13 @@ TEST_P(GaussianBoxmullerTest, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::gaussian, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, GAUSSIAN_ARGS_DOUBLE))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::gaussian, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, GAUSSIAN_ARGS_DOUBLE))); } diff --git a/tests/unit_tests/rng/statistics_check/gaussian_usm.cpp b/tests/unit_tests/rng/statistics_check/gaussian_usm.cpp index a1d4d1b06..d387bbc6d 100644 --- a/tests/unit_tests/rng/statistics_check/gaussian_usm.cpp +++ b/tests/unit_tests/rng/statistics_check/gaussian_usm.cpp @@ -31,13 +31,13 @@ class GaussianIcdfUsmTest : public ::testing::TestWithParam {}; TEST_P(GaussianIcdfUsmTest, RealSinglePrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::gaussian, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, GAUSSIAN_ARGS_FLOAT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::gaussian, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, GAUSSIAN_ARGS_FLOAT))); } @@ -46,26 +46,26 @@ TEST_P(GaussianIcdfUsmTest, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::gaussian, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, GAUSSIAN_ARGS_DOUBLE))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::gaussian, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, GAUSSIAN_ARGS_DOUBLE))); } TEST_P(GaussianBoxmullerUsmTest, RealSinglePrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::gaussian, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, GAUSSIAN_ARGS_FLOAT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::gaussian, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, GAUSSIAN_ARGS_FLOAT))); } @@ -74,13 +74,13 @@ TEST_P(GaussianBoxmullerUsmTest, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::gaussian, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, GAUSSIAN_ARGS_DOUBLE))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::gaussian, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, GAUSSIAN_ARGS_DOUBLE))); } diff --git a/tests/unit_tests/rng/statistics_check/lognormal.cpp b/tests/unit_tests/rng/statistics_check/lognormal.cpp index 5486202bb..9ba225e20 100755 --- a/tests/unit_tests/rng/statistics_check/lognormal.cpp +++ b/tests/unit_tests/rng/statistics_check/lognormal.cpp @@ -31,13 +31,13 @@ class LognormalIcdfTest : public ::testing::TestWithParam {}; TEST_P(LognormalIcdfTest, RealSinglePrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::lognormal, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, LOGNORMAL_ARGS_FLOAT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::lognormal, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, LOGNORMAL_ARGS_FLOAT))); } @@ -46,26 +46,26 @@ TEST_P(LognormalIcdfTest, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::lognormal, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, LOGNORMAL_ARGS_DOUBLE))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::lognormal, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, LOGNORMAL_ARGS_DOUBLE))); } TEST_P(LognormalBoxmullerTest, RealSinglePrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::lognormal, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, LOGNORMAL_ARGS_FLOAT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::lognormal, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, LOGNORMAL_ARGS_FLOAT))); } @@ -74,13 +74,13 @@ TEST_P(LognormalBoxmullerTest, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::lognormal, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, LOGNORMAL_ARGS_DOUBLE))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::lognormal, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, LOGNORMAL_ARGS_DOUBLE))); } diff --git a/tests/unit_tests/rng/statistics_check/lognormal_usm.cpp b/tests/unit_tests/rng/statistics_check/lognormal_usm.cpp index d59d9458a..ad9dd48d7 100755 --- a/tests/unit_tests/rng/statistics_check/lognormal_usm.cpp +++ b/tests/unit_tests/rng/statistics_check/lognormal_usm.cpp @@ -31,13 +31,13 @@ class LognormalIcdfUsmTest : public ::testing::TestWithParam {}; TEST_P(LognormalIcdfUsmTest, RealSinglePrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::lognormal, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, LOGNORMAL_ARGS_FLOAT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::lognormal, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, LOGNORMAL_ARGS_FLOAT))); } @@ -46,26 +46,26 @@ TEST_P(LognormalIcdfUsmTest, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::lognormal, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, LOGNORMAL_ARGS_DOUBLE))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::lognormal, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, LOGNORMAL_ARGS_DOUBLE))); } TEST_P(LognormalBoxmullerUsmTest, RealSinglePrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::lognormal, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, LOGNORMAL_ARGS_FLOAT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::lognormal, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, LOGNORMAL_ARGS_FLOAT))); } @@ -74,13 +74,13 @@ TEST_P(LognormalBoxmullerUsmTest, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::lognormal, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, LOGNORMAL_ARGS_DOUBLE))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::lognormal, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, LOGNORMAL_ARGS_DOUBLE))); } diff --git a/tests/unit_tests/rng/statistics_check/poisson.cpp b/tests/unit_tests/rng/statistics_check/poisson.cpp index d39842e9f..238a3cb6e 100755 --- a/tests/unit_tests/rng/statistics_check/poisson.cpp +++ b/tests/unit_tests/rng/statistics_check/poisson.cpp @@ -29,30 +29,30 @@ class PoissonIcdfTests : public ::testing::TestWithParam {}; TEST_P(PoissonIcdfTests, IntegerPrecision) { rng_test< - statistics_test, - oneapi::mkl::rng::philox4x32x10>> + statistics_test, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, POISSON_ARGS))); rng_test< - statistics_test, - oneapi::mkl::rng::mrg32k3a>> + statistics_test, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, POISSON_ARGS))); } TEST_P(PoissonIcdfTests, UnsignedIntegerPrecision) { rng_test< - statistics_test, - oneapi::mkl::rng::philox4x32x10>> + statistics_test, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, POISSON_ARGS))); rng_test< - statistics_test, - oneapi::mkl::rng::mrg32k3a>> + statistics_test, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, POISSON_ARGS))); } diff --git a/tests/unit_tests/rng/statistics_check/poisson_usm.cpp b/tests/unit_tests/rng/statistics_check/poisson_usm.cpp index 052eff5a3..44a1e6b4c 100755 --- a/tests/unit_tests/rng/statistics_check/poisson_usm.cpp +++ b/tests/unit_tests/rng/statistics_check/poisson_usm.cpp @@ -29,30 +29,30 @@ class PoissonIcdfUsmTests : public ::testing::TestWithParam {}; TEST_P(PoissonIcdfUsmTests, IntegerPrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::poisson, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, POISSON_ARGS))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::poisson, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, POISSON_ARGS))); } TEST_P(PoissonIcdfUsmTests, UnsignedIntegerPrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::poisson, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, POISSON_ARGS))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::poisson, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, POISSON_ARGS))); } diff --git a/tests/unit_tests/rng/statistics_check/uniform.cpp b/tests/unit_tests/rng/statistics_check/uniform.cpp index eb11714e1..d21644708 100644 --- a/tests/unit_tests/rng/statistics_check/uniform.cpp +++ b/tests/unit_tests/rng/statistics_check/uniform.cpp @@ -31,13 +31,13 @@ class UniformAccurateTests : public ::testing::TestWithParam {}; TEST_P(UniformStdTests, RealSinglePrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::uniform, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, UNIFORM_ARGS_FLOAT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::uniform, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, UNIFORM_ARGS_FLOAT))); } @@ -46,39 +46,39 @@ TEST_P(UniformStdTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::uniform, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, UNIFORM_ARGS_DOUBLE))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::uniform, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, UNIFORM_ARGS_DOUBLE))); } TEST_P(UniformStdTests, IntegerPrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::uniform, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, UNIFORM_ARGS_INT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::uniform, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, UNIFORM_ARGS_INT))); } TEST_P(UniformAccurateTests, RealSinglePrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::uniform, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, UNIFORM_ARGS_FLOAT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::uniform, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, UNIFORM_ARGS_FLOAT))); } @@ -87,13 +87,13 @@ TEST_P(UniformAccurateTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::uniform, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, UNIFORM_ARGS_DOUBLE))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::uniform, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, UNIFORM_ARGS_DOUBLE))); } diff --git a/tests/unit_tests/rng/statistics_check/uniform_usm.cpp b/tests/unit_tests/rng/statistics_check/uniform_usm.cpp index df4f7a764..d406d0a59 100644 --- a/tests/unit_tests/rng/statistics_check/uniform_usm.cpp +++ b/tests/unit_tests/rng/statistics_check/uniform_usm.cpp @@ -31,13 +31,13 @@ class UniformAccurateUsmTests : public ::testing::TestWithParam { TEST_P(UniformStdUsmTests, RealSinglePrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::uniform, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, UNIFORM_ARGS_FLOAT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::uniform, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, UNIFORM_ARGS_FLOAT))); } @@ -46,39 +46,39 @@ TEST_P(UniformStdUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::uniform, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, UNIFORM_ARGS_DOUBLE))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::uniform, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, UNIFORM_ARGS_DOUBLE))); } TEST_P(UniformStdUsmTests, IntegerPrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::uniform, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, UNIFORM_ARGS_INT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::uniform, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, UNIFORM_ARGS_INT))); } TEST_P(UniformAccurateUsmTests, RealSinglePrecision) { rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::uniform, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, UNIFORM_ARGS_FLOAT))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::uniform, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, UNIFORM_ARGS_FLOAT))); } @@ -87,13 +87,13 @@ TEST_P(UniformAccurateUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); rng_test, - oneapi::mkl::rng::philox4x32x10>> + oneapi::math::rng::uniform, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, UNIFORM_ARGS_DOUBLE))); rng_test, - oneapi::mkl::rng::mrg32k3a>> + oneapi::math::rng::uniform, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, UNIFORM_ARGS_DOUBLE))); } diff --git a/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp b/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp index 777eb5e45..a02957905 100644 --- a/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp +++ b/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp @@ -55,10 +55,10 @@ inline T opVal(const T t, const bool isConj) { }; template -void do_csr_transpose(const oneapi::mkl::transpose opA, intType *ia_t, intType *ja_t, fpType *a_t, +void do_csr_transpose(const oneapi::math::transpose opA, intType *ia_t, intType *ja_t, fpType *a_t, intType a_nrows, intType a_ncols, intType indexing, accIntType &ia, accIntType &ja, accFpType &a, const bool structOnlyFlag = false) { - const bool isConj = (opA == oneapi::mkl::transpose::conjtrans); + const bool isConj = (opA == oneapi::math::transpose::conjtrans); // initialize ia_t to zero for (intType i = 0; i < a_ncols + 1; ++i) { @@ -107,17 +107,17 @@ void do_csr_transpose(const oneapi::mkl::transpose opA, intType *ia_t, intType * template auto sparse_transpose_if_needed(const intType *ia, const intType *ja, const fpType *a, intType a_nrows, intType a_ncols, std::size_t nnz, intType indexing, - oneapi::mkl::transpose transpose_val) { + oneapi::math::transpose transpose_val) { std::vector iopa; std::vector jopa; std::vector opa; - if (transpose_val == oneapi::mkl::transpose::nontrans) { + if (transpose_val == oneapi::math::transpose::nontrans) { iopa.assign(ia, ia + a_nrows + 1); jopa.assign(ja, ja + nnz); opa.assign(a, a + nnz); } - else if (transpose_val == oneapi::mkl::transpose::trans || - transpose_val == oneapi::mkl::transpose::conjtrans) { + else if (transpose_val == oneapi::math::transpose::trans || + transpose_val == oneapi::math::transpose::conjtrans) { iopa.resize(static_cast(a_ncols + 1)); jopa.resize(nnz); opa.resize(nnz); @@ -135,11 +135,11 @@ auto sparse_transpose_if_needed(const intType *ia, const intType *ja, const fpTy /// The outputted matrix always uses row major layout template auto extract_dense_matrix(const fpType *x, std::size_t nrows, std::size_t ncols, std::size_t ld, - oneapi::mkl::transpose transpose_val, - oneapi::mkl::layout dense_matrix_layout) { - const bool is_row_major = dense_matrix_layout == oneapi::mkl::layout::row_major; - const bool is_transposed = transpose_val != oneapi::mkl::transpose::nontrans; - const bool apply_conjugate = transpose_val == oneapi::mkl::transpose::conjtrans; + oneapi::math::transpose transpose_val, + oneapi::math::layout dense_matrix_layout) { + const bool is_row_major = dense_matrix_layout == oneapi::math::layout::row_major; + const bool is_transposed = transpose_val != oneapi::math::transpose::nontrans; + const bool apply_conjugate = transpose_val == oneapi::math::transpose::conjtrans; const bool swap_ld = is_row_major != is_transposed; if (swap_ld && ncols > ld) { throw std::runtime_error("Expected ncols <= ld"); @@ -164,31 +164,31 @@ template std::vector sparse_to_dense(sparse_matrix_format_t format, const intType *ia, const intType *ja, const fpType *a, std::size_t a_nrows, std::size_t a_ncols, std::size_t nnz, intType indexing, - oneapi::mkl::transpose transpose_val, - oneapi::mkl::sparse::matrix_view A_view) { - oneapi::mkl::sparse::matrix_descr type_view = A_view.type_view; - oneapi::mkl::uplo uplo_val = A_view.uplo_view; + oneapi::math::transpose transpose_val, + oneapi::math::sparse::matrix_view A_view) { + oneapi::math::sparse::matrix_descr type_view = A_view.type_view; + oneapi::math::uplo uplo_val = A_view.uplo_view; const bool is_symmetric_or_hermitian_view = - type_view == oneapi::mkl::sparse::matrix_descr::symmetric || - type_view == oneapi::mkl::sparse::matrix_descr::hermitian; - const bool apply_conjugate = transpose_val == oneapi::mkl::transpose::conjtrans; + type_view == oneapi::math::sparse::matrix_descr::symmetric || + type_view == oneapi::math::sparse::matrix_descr::hermitian; + const bool apply_conjugate = transpose_val == oneapi::math::transpose::conjtrans; std::vector dense_a(a_nrows * a_ncols, fpType(0)); auto write_to_dense_if_needed = [&](std::size_t a_idx, std::size_t row, std::size_t col) { - if ((type_view == oneapi::mkl::sparse::matrix_descr::triangular || + if ((type_view == oneapi::math::sparse::matrix_descr::triangular || is_symmetric_or_hermitian_view) && - ((uplo_val == oneapi::mkl::uplo::lower && col > row) || - (uplo_val == oneapi::mkl::uplo::upper && col < row))) { + ((uplo_val == oneapi::math::uplo::lower && col > row) || + (uplo_val == oneapi::math::uplo::upper && col < row))) { // Read only the upper or lower part of the sparse matrix return; } - if (type_view == oneapi::mkl::sparse::matrix_descr::diagonal && col != row) { + if (type_view == oneapi::math::sparse::matrix_descr::diagonal && col != row) { // Read only the diagonal return; } // Do not transpose symmetric matrices to simplify the propagation of the symmetric values std::size_t dense_a_idx = - (!is_symmetric_or_hermitian_view && transpose_val != oneapi::mkl::transpose::nontrans) + (!is_symmetric_or_hermitian_view && transpose_val != oneapi::math::transpose::nontrans) ? col * a_nrows + row : row * a_ncols + col; fpType val = opVal(a[a_idx], apply_conjugate); @@ -213,7 +213,7 @@ std::vector sparse_to_dense(sparse_matrix_format_t format, const intType } // Write unit diagonal - if (A_view.diag_view == oneapi::mkl::diag::unit && a_nrows == a_ncols) { + if (A_view.diag_view == oneapi::math::diag::unit && a_nrows == a_ncols) { for (std::size_t i = 0; i < a_nrows; i++) { dense_a[i * a_nrows + i] = fpType(1); } @@ -223,7 +223,7 @@ std::vector sparse_to_dense(sparse_matrix_format_t format, const intType if (is_symmetric_or_hermitian_view) { for (std::size_t i = 0; i < a_nrows; ++i) { for (std::size_t j = i + 1; j < a_ncols; ++j) { - if (uplo_val == oneapi::mkl::uplo::lower) { + if (uplo_val == oneapi::math::uplo::lower) { dense_a[i * a_ncols + j] = dense_a[j * a_nrows + i]; } else { diff --git a/tests/unit_tests/sparse_blas/include/test_common.hpp b/tests/unit_tests/sparse_blas/include/test_common.hpp index c11255a9a..a9b8841d9 100644 --- a/tests/unit_tests/sparse_blas/include/test_common.hpp +++ b/tests/unit_tests/sparse_blas/include/test_common.hpp @@ -59,11 +59,11 @@ enum sparse_matrix_format_t { COO, }; -static std::vector> test_matrix_properties{ - { oneapi::mkl::sparse::matrix_property::sorted }, - { oneapi::mkl::sparse::matrix_property::symmetric }, - { oneapi::mkl::sparse::matrix_property::sorted, - oneapi::mkl::sparse::matrix_property::symmetric } +static std::vector> test_matrix_properties{ + { oneapi::math::sparse::matrix_property::sorted }, + { oneapi::math::sparse::matrix_property::symmetric }, + { oneapi::math::sparse::matrix_property::sorted, + oneapi::math::sparse::matrix_property::symmetric } }; void print_error_code(sycl::exception const &e); @@ -129,12 +129,12 @@ auto swap_if_cond(bool swap, T x, T y) { } template -auto swap_if_transposed(oneapi::mkl::transpose op, XT x, YT y) { - return swap_if_cond(op != oneapi::mkl::transpose::nontrans, x, y); +auto swap_if_transposed(oneapi::math::transpose op, XT x, YT y) { + return swap_if_cond(op != oneapi::math::transpose::nontrans, x, y); } template -auto swap_if_transposed(oneapi::mkl::transpose op, T x, T y) { +auto swap_if_transposed(oneapi::math::transpose op, T x, T y) { return swap_if_transposed(op, x, y); } @@ -178,13 +178,13 @@ void rand_vector(std::vector &v, std::size_t n) { } template -void rand_matrix(std::vector &m, oneapi::mkl::layout layout_val, std::size_t nrows, +void rand_matrix(std::vector &m, oneapi::math::layout layout_val, std::size_t nrows, std::size_t ncols, std::size_t ld, - oneapi::mkl::transpose transpose_val = oneapi::mkl::transpose::nontrans) { + oneapi::math::transpose transpose_val = oneapi::math::transpose::nontrans) { using fpRealType = typename complex_info::real_type; auto [op_nrows, op_cols] = swap_if_transposed(transpose_val, nrows, ncols); auto [outer_size, inner_size] = - swap_if_cond(layout_val == oneapi::mkl::layout::row_major, op_cols, op_nrows); + swap_if_cond(layout_val == oneapi::math::layout::row_major, op_cols, op_nrows); if (inner_size > ld) { throw std::runtime_error("Expected inner_size <= ld"); } @@ -360,7 +360,7 @@ void shuffle_sparse_matrix(sparse_matrix_format_t format, intType indexing, intT } } else { - throw oneapi::mkl::exception("sparse_blas", "shuffle_sparse_matrix", + throw oneapi::math::exception("sparse_blas", "shuffle_sparse_matrix", "Internal error: unsupported format"); } } @@ -368,19 +368,19 @@ void shuffle_sparse_matrix(sparse_matrix_format_t format, intType indexing, intT /// Initialize a sparse matrix specified by the given format template void init_sparse_matrix(sycl::queue &queue, sparse_matrix_format_t format, - oneapi::mkl::sparse::matrix_handle_t *p_smhandle, std::int64_t num_rows, - std::int64_t num_cols, std::int64_t nnz, oneapi::mkl::index_base index, + oneapi::math::sparse::matrix_handle_t *p_smhandle, std::int64_t num_rows, + std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, ContainerIndexT rows, ContainerIndexT cols, ContainerValueT vals) { if (format == sparse_matrix_format_t::CSR) { - CALL_RT_OR_CT(oneapi::mkl::sparse::init_csr_matrix, queue, p_smhandle, num_rows, num_cols, + CALL_RT_OR_CT(oneapi::math::sparse::init_csr_matrix, queue, p_smhandle, num_rows, num_cols, nnz, index, rows, cols, vals); } else if (format == sparse_matrix_format_t::COO) { - CALL_RT_OR_CT(oneapi::mkl::sparse::init_coo_matrix, queue, p_smhandle, num_rows, num_cols, + CALL_RT_OR_CT(oneapi::math::sparse::init_coo_matrix, queue, p_smhandle, num_rows, num_cols, nnz, index, rows, cols, vals); } else { - throw oneapi::mkl::exception("sparse_blas", "init_sparse_matrix", + throw oneapi::math::exception("sparse_blas", "init_sparse_matrix", "Internal error: unsupported format"); } } @@ -388,19 +388,19 @@ void init_sparse_matrix(sycl::queue &queue, sparse_matrix_format_t format, /// Reset the data of a sparse matrix specified by the given format template void set_matrix_data(sycl::queue &queue, sparse_matrix_format_t format, - oneapi::mkl::sparse::matrix_handle_t smhandle, std::int64_t num_rows, - std::int64_t num_cols, std::int64_t nnz, oneapi::mkl::index_base index, + oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, + std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, ContainerIndexT rows, ContainerIndexT cols, ContainerValueT vals) { if (format == sparse_matrix_format_t::CSR) { - CALL_RT_OR_CT(oneapi::mkl::sparse::set_csr_matrix_data, queue, smhandle, num_rows, num_cols, + CALL_RT_OR_CT(oneapi::math::sparse::set_csr_matrix_data, queue, smhandle, num_rows, num_cols, nnz, index, rows, cols, vals); } else if (format == sparse_matrix_format_t::COO) { - CALL_RT_OR_CT(oneapi::mkl::sparse::set_coo_matrix_data, queue, smhandle, num_rows, num_cols, + CALL_RT_OR_CT(oneapi::math::sparse::set_coo_matrix_data, queue, smhandle, num_rows, num_cols, nnz, index, rows, cols, vals); } else { - throw oneapi::mkl::exception("sparse_blas", "set_matrix_data", + throw oneapi::math::exception("sparse_blas", "set_matrix_data", "Internal error: unsupported format"); } } @@ -416,18 +416,18 @@ inline void free_handles(sycl::queue &queue, const std::vector depe } sycl::event event; if constexpr (std::is_same_v) { - CALL_RT_OR_CT(event = oneapi::mkl::sparse::release_dense_vector, queue, handles, + oneapi::math::sparse::dense_vector_handle_t>) { + CALL_RT_OR_CT(event = oneapi::math::sparse::release_dense_vector, queue, handles, dependencies); } else if constexpr (std::is_same_v) { - CALL_RT_OR_CT(event = oneapi::mkl::sparse::release_dense_matrix, queue, handles, + oneapi::math::sparse::dense_matrix_handle_t>) { + CALL_RT_OR_CT(event = oneapi::math::sparse::release_dense_matrix, queue, handles, dependencies); } else if constexpr (std::is_same_v) { - CALL_RT_OR_CT(event = oneapi::mkl::sparse::release_sparse_matrix, queue, handles, + oneapi::math::sparse::matrix_handle_t>) { + CALL_RT_OR_CT(event = oneapi::math::sparse::release_sparse_matrix, queue, handles, dependencies); } event.wait(); @@ -447,12 +447,12 @@ inline void wait_and_free_handles(sycl::queue &queue, HandlesT &&... handles) { } inline bool require_square_matrix( - oneapi::mkl::sparse::matrix_view A_view, - const std::set &matrix_properties) { + oneapi::math::sparse::matrix_view A_view, + const std::set &matrix_properties) { const bool is_symmetric = - matrix_properties.find(oneapi::mkl::sparse::matrix_property::symmetric) != + matrix_properties.find(oneapi::math::sparse::matrix_property::symmetric) != matrix_properties.cend(); - return A_view.type_view != oneapi::mkl::sparse::matrix_descr::general || is_symmetric; + return A_view.type_view != oneapi::math::sparse::matrix_descr::general || is_symmetric; } template diff --git a/tests/unit_tests/sparse_blas/include/test_spmm.hpp b/tests/unit_tests/sparse_blas/include/test_spmm.hpp index 09eb24369..e34910521 100644 --- a/tests/unit_tests/sparse_blas/include/test_spmm.hpp +++ b/tests/unit_tests/sparse_blas/include/test_spmm.hpp @@ -55,26 +55,26 @@ template void test_helper_with_format_with_transpose( testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device *dev, sparse_matrix_format_t format, - const std::vector &non_default_algorithms, - oneapi::mkl::transpose transpose_A, oneapi::mkl::transpose transpose_B, int &num_passed, + const std::vector &non_default_algorithms, + oneapi::math::transpose transpose_A, oneapi::math::transpose transpose_B, int &num_passed, int &num_skipped) { double density_A_matrix = 0.8; fpType fp_zero = set_fp_value()(0.f, 0.f); fpType fp_one = set_fp_value()(1.f, 0.f); - oneapi::mkl::index_base index_zero = oneapi::mkl::index_base::zero; - oneapi::mkl::layout col_major = oneapi::mkl::layout::col_major; - oneapi::mkl::sparse::spmm_alg default_alg = oneapi::mkl::sparse::spmm_alg::default_alg; - oneapi::mkl::sparse::matrix_view default_A_view; - std::set no_properties; + oneapi::math::index_base index_zero = oneapi::math::index_base::zero; + oneapi::math::layout col_major = oneapi::math::layout::col_major; + oneapi::math::sparse::spmm_alg default_alg = oneapi::math::sparse::spmm_alg::default_alg; + oneapi::math::sparse::matrix_view default_A_view; + std::set no_properties; bool no_reset_data = false; bool no_scalars_on_device = false; { int m = 4, k = 6, n = 5; - int nrows_A = (transpose_A != oneapi::mkl::transpose::nontrans) ? k : m; - int ncols_A = (transpose_A != oneapi::mkl::transpose::nontrans) ? m : k; - int nrows_B = (transpose_B != oneapi::mkl::transpose::nontrans) ? n : k; - int ncols_B = (transpose_B != oneapi::mkl::transpose::nontrans) ? k : n; + int nrows_A = (transpose_A != oneapi::math::transpose::nontrans) ? k : m; + int ncols_A = (transpose_A != oneapi::math::transpose::nontrans) ? m : k; + int nrows_B = (transpose_B != oneapi::math::transpose::nontrans) ? n : k; + int ncols_B = (transpose_B != oneapi::math::transpose::nontrans) ? k : n; int nrows_C = m; int ncols_C = n; int ldb = nrows_B; @@ -103,7 +103,7 @@ void test_helper_with_format_with_transpose( // Test index_base 1 EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, nrows_A, ncols_A, ncols_C, density_A_matrix, - oneapi::mkl::index_base::one, col_major, transpose_A, transpose_B, + oneapi::math::index_base::one, col_major, transpose_A, transpose_B, fp_one, fp_zero, ldb, ldc, default_alg, default_A_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); @@ -152,7 +152,7 @@ void test_helper_with_format_with_transpose( // Test row major layout EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, nrows_A, ncols_A, ncols_C, density_A_matrix, index_zero, - oneapi::mkl::layout::row_major, transpose_A, transpose_B, fp_one, + oneapi::math::layout::row_major, transpose_A, transpose_B, fp_one, fp_zero, ncols_B, ncols_C, default_alg, default_A_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); @@ -187,9 +187,9 @@ void test_helper_with_format_with_transpose( { // Test different sizes int m = 6, k = 2, n = 5; - int nrows_A = (transpose_A != oneapi::mkl::transpose::nontrans) ? k : m; - int ncols_A = (transpose_A != oneapi::mkl::transpose::nontrans) ? m : k; - int nrows_B = (transpose_B != oneapi::mkl::transpose::nontrans) ? n : k; + int nrows_A = (transpose_A != oneapi::math::transpose::nontrans) ? k : m; + int ncols_A = (transpose_A != oneapi::math::transpose::nontrans) ? m : k; + int nrows_B = (transpose_B != oneapi::math::transpose::nontrans) ? n : k; int nrows_C = m; int ncols_C = n; int ldb = nrows_B; @@ -219,11 +219,11 @@ template void test_helper_with_format( testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device *dev, sparse_matrix_format_t format, - const std::vector &non_default_algorithms, int &num_passed, + const std::vector &non_default_algorithms, int &num_passed, int &num_skipped) { - std::vector transpose_vals{ oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::conjtrans }; + std::vector transpose_vals{ oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, + oneapi::math::transpose::conjtrans }; for (auto transpose_A : transpose_vals) { for (auto transpose_B : transpose_vals) { test_helper_with_format_with_transpose( @@ -248,14 +248,14 @@ void test_helper(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i6 sycl::device *dev, int &num_passed, int &num_skipped) { test_helper_with_format( test_functor_i32, test_functor_i64, dev, sparse_matrix_format_t::CSR, - { oneapi::mkl::sparse::spmm_alg::no_optimize_alg, oneapi::mkl::sparse::spmm_alg::csr_alg1, - oneapi::mkl::sparse::spmm_alg::csr_alg2, oneapi::mkl::sparse::spmm_alg::csr_alg3 }, + { oneapi::math::sparse::spmm_alg::no_optimize_alg, oneapi::math::sparse::spmm_alg::csr_alg1, + oneapi::math::sparse::spmm_alg::csr_alg2, oneapi::math::sparse::spmm_alg::csr_alg3 }, num_passed, num_skipped); test_helper_with_format( test_functor_i32, test_functor_i64, dev, sparse_matrix_format_t::COO, - { oneapi::mkl::sparse::spmm_alg::no_optimize_alg, oneapi::mkl::sparse::spmm_alg::coo_alg1, - oneapi::mkl::sparse::spmm_alg::coo_alg2, oneapi::mkl::sparse::spmm_alg::coo_alg3, - oneapi::mkl::sparse::spmm_alg::coo_alg4 }, + { oneapi::math::sparse::spmm_alg::no_optimize_alg, oneapi::math::sparse::spmm_alg::coo_alg1, + oneapi::math::sparse::spmm_alg::coo_alg2, oneapi::math::sparse::spmm_alg::coo_alg3, + oneapi::math::sparse::spmm_alg::coo_alg4 }, num_passed, num_skipped); } @@ -264,10 +264,10 @@ template void prepare_reference_spmm_data(sparse_matrix_format_t format, const intType *ia, const intType *ja, const fpType *a, intType a_nrows, intType a_ncols, intType c_ncols, intType a_nnz, intType indexing, - oneapi::mkl::layout dense_matrix_layout, - oneapi::mkl::transpose opA, oneapi::mkl::transpose opB, + oneapi::math::layout dense_matrix_layout, + oneapi::math::transpose opA, oneapi::math::transpose opB, fpType alpha, fpType beta, intType ldb, intType ldc, - const fpType *b, oneapi::mkl::sparse::matrix_view A_view, + const fpType *b, oneapi::math::sparse::matrix_view A_view, fpType *c_ref) { std::size_t a_nrows_u = static_cast(a_nrows); std::size_t a_ncols_u = static_cast(a_ncols); @@ -285,7 +285,7 @@ void prepare_reference_spmm_data(sparse_matrix_format_t format, const intType *i // Return the linear index to access a dense matrix from auto dense_linear_idx = [=](std::size_t row, std::size_t col, std::size_t ld) { - return (dense_matrix_layout == oneapi::mkl::layout::row_major) ? row * ld + col + return (dense_matrix_layout == oneapi::math::layout::row_major) ? row * ld + col : col * ld + row; }; diff --git a/tests/unit_tests/sparse_blas/include/test_spmv.hpp b/tests/unit_tests/sparse_blas/include/test_spmv.hpp index e6e4f60bc..848180c28 100644 --- a/tests/unit_tests/sparse_blas/include/test_spmv.hpp +++ b/tests/unit_tests/sparse_blas/include/test_spmv.hpp @@ -54,16 +54,16 @@ template void test_helper_with_format_with_transpose( testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device *dev, sparse_matrix_format_t format, - const std::vector &non_default_algorithms, - oneapi::mkl::transpose transpose_val, int &num_passed, int &num_skipped) { + const std::vector &non_default_algorithms, + oneapi::math::transpose transpose_val, int &num_passed, int &num_skipped) { double density_A_matrix = 0.8; fpType fp_zero = set_fp_value()(0.f, 0.f); fpType fp_one = set_fp_value()(1.f, 0.f); int nrows_A = 4, ncols_A = 6; - oneapi::mkl::index_base index_zero = oneapi::mkl::index_base::zero; - oneapi::mkl::sparse::spmv_alg default_alg = oneapi::mkl::sparse::spmv_alg::default_alg; - oneapi::mkl::sparse::matrix_view default_A_view; - std::set no_properties; + oneapi::math::index_base index_zero = oneapi::math::index_base::zero; + oneapi::math::sparse::spmv_alg default_alg = oneapi::math::sparse::spmv_alg::default_alg; + oneapi::math::sparse::matrix_view default_A_view; + std::set no_properties; bool no_reset_data = false; bool no_scalars_on_device = false; @@ -88,7 +88,7 @@ void test_helper_with_format_with_transpose( // Test index_base 1 EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, nrows_A, ncols_A, density_A_matrix, - oneapi::mkl::index_base::one, transpose_val, fp_one, fp_zero, default_alg, + oneapi::math::index_base::one, transpose_val, fp_one, fp_zero, default_alg, default_A_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); // Test non-default alpha @@ -122,59 +122,59 @@ void test_helper_with_format_with_transpose( no_scalars_on_device), num_passed, num_skipped); // Lower triangular - oneapi::mkl::sparse::matrix_view triangular_A_view( - oneapi::mkl::sparse::matrix_descr::triangular); + oneapi::math::sparse::matrix_view triangular_A_view( + oneapi::math::sparse::matrix_descr::triangular); EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, nrows_A, ncols_A, density_A_matrix, index_zero, transpose_val, fp_one, fp_zero, default_alg, triangular_A_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); // Upper triangular - triangular_A_view.uplo_view = oneapi::mkl::uplo::upper; + triangular_A_view.uplo_view = oneapi::math::uplo::upper; EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, nrows_A, ncols_A, density_A_matrix, index_zero, transpose_val, fp_one, fp_zero, default_alg, triangular_A_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); // Lower triangular unit diagonal - oneapi::mkl::sparse::matrix_view triangular_unit_A_view( - oneapi::mkl::sparse::matrix_descr::triangular); - triangular_unit_A_view.diag_view = oneapi::mkl::diag::unit; + oneapi::math::sparse::matrix_view triangular_unit_A_view( + oneapi::math::sparse::matrix_descr::triangular); + triangular_unit_A_view.diag_view = oneapi::math::diag::unit; EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, nrows_A, ncols_A, density_A_matrix, index_zero, transpose_val, fp_one, fp_zero, default_alg, triangular_unit_A_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); // Upper triangular unit diagonal - triangular_A_view.uplo_view = oneapi::mkl::uplo::upper; + triangular_A_view.uplo_view = oneapi::math::uplo::upper; EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, nrows_A, ncols_A, density_A_matrix, index_zero, transpose_val, fp_one, fp_zero, default_alg, triangular_unit_A_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); // Lower symmetric - oneapi::mkl::sparse::matrix_view symmetric_view(oneapi::mkl::sparse::matrix_descr::symmetric); + oneapi::math::sparse::matrix_view symmetric_view(oneapi::math::sparse::matrix_descr::symmetric); EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, nrows_A, ncols_A, density_A_matrix, index_zero, transpose_val, fp_one, fp_zero, default_alg, symmetric_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); // Upper symmetric - symmetric_view.uplo_view = oneapi::mkl::uplo::upper; + symmetric_view.uplo_view = oneapi::math::uplo::upper; EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, nrows_A, ncols_A, density_A_matrix, index_zero, transpose_val, fp_one, fp_zero, default_alg, symmetric_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); // Lower hermitian - oneapi::mkl::sparse::matrix_view hermitian_view(oneapi::mkl::sparse::matrix_descr::hermitian); + oneapi::math::sparse::matrix_view hermitian_view(oneapi::math::sparse::matrix_descr::hermitian); EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, nrows_A, ncols_A, density_A_matrix, index_zero, transpose_val, fp_one, fp_zero, default_alg, hermitian_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); // Upper hermitian - hermitian_view.uplo_view = oneapi::mkl::uplo::upper; + hermitian_view.uplo_view = oneapi::math::uplo::upper; EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, nrows_A, ncols_A, density_A_matrix, index_zero, transpose_val, fp_one, fp_zero, default_alg, hermitian_view, no_properties, no_reset_data, @@ -214,11 +214,11 @@ template void test_helper_with_format( testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device *dev, sparse_matrix_format_t format, - const std::vector &non_default_algorithms, int &num_passed, + const std::vector &non_default_algorithms, int &num_passed, int &num_skipped) { - std::vector transpose_vals{ oneapi::mkl::transpose::nontrans, - oneapi::mkl::transpose::trans, - oneapi::mkl::transpose::conjtrans }; + std::vector transpose_vals{ oneapi::math::transpose::nontrans, + oneapi::math::transpose::trans, + oneapi::math::transpose::conjtrans }; for (auto transpose_A : transpose_vals) { test_helper_with_format_with_transpose(test_functor_i32, test_functor_i64, dev, format, non_default_algorithms, transpose_A, @@ -241,13 +241,13 @@ void test_helper(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i6 sycl::device *dev, int &num_passed, int &num_skipped) { test_helper_with_format( test_functor_i32, test_functor_i64, dev, sparse_matrix_format_t::CSR, - { oneapi::mkl::sparse::spmv_alg::no_optimize_alg, oneapi::mkl::sparse::spmv_alg::csr_alg1, - oneapi::mkl::sparse::spmv_alg::csr_alg2, oneapi::mkl::sparse::spmv_alg::csr_alg3 }, + { oneapi::math::sparse::spmv_alg::no_optimize_alg, oneapi::math::sparse::spmv_alg::csr_alg1, + oneapi::math::sparse::spmv_alg::csr_alg2, oneapi::math::sparse::spmv_alg::csr_alg3 }, num_passed, num_skipped); test_helper_with_format( test_functor_i32, test_functor_i64, dev, sparse_matrix_format_t::COO, - { oneapi::mkl::sparse::spmv_alg::no_optimize_alg, oneapi::mkl::sparse::spmv_alg::coo_alg1, - oneapi::mkl::sparse::spmv_alg::coo_alg2 }, + { oneapi::math::sparse::spmv_alg::no_optimize_alg, oneapi::math::sparse::spmv_alg::coo_alg1, + oneapi::math::sparse::spmv_alg::coo_alg2 }, num_passed, num_skipped); } @@ -256,8 +256,8 @@ template void prepare_reference_spmv_data(sparse_matrix_format_t format, const intType *ia, const intType *ja, const fpType *a, intType a_nrows, intType a_ncols, intType a_nnz, intType indexing, - oneapi::mkl::transpose opA, fpType alpha, fpType beta, - const fpType *x, oneapi::mkl::sparse::matrix_view A_view, + oneapi::math::transpose opA, fpType alpha, fpType beta, + const fpType *x, oneapi::math::sparse::matrix_view A_view, fpType *y_ref) { std::size_t a_nrows_u = static_cast(a_nrows); std::size_t a_ncols_u = static_cast(a_ncols); diff --git a/tests/unit_tests/sparse_blas/include/test_spsv.hpp b/tests/unit_tests/sparse_blas/include/test_spsv.hpp index d301b0d48..16ffe4484 100644 --- a/tests/unit_tests/sparse_blas/include/test_spsv.hpp +++ b/tests/unit_tests/sparse_blas/include/test_spsv.hpp @@ -49,18 +49,18 @@ template void test_helper_with_format(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device *dev, sparse_matrix_format_t format, - oneapi::mkl::transpose transpose_val, int &num_passed, + oneapi::math::transpose transpose_val, int &num_passed, int &num_skipped) { double density_A_matrix = 0.144; fpType alpha = set_fp_value()(1.f, 0.f); int m = 277; - oneapi::mkl::index_base index_zero = oneapi::mkl::index_base::zero; - oneapi::mkl::sparse::spsv_alg default_alg = oneapi::mkl::sparse::spsv_alg::default_alg; - oneapi::mkl::sparse::spsv_alg no_optimize_alg = oneapi::mkl::sparse::spsv_alg::no_optimize_alg; - oneapi::mkl::sparse::matrix_view default_A_view(oneapi::mkl::sparse::matrix_descr::triangular); - oneapi::mkl::sparse::matrix_view upper_A_view(oneapi::mkl::sparse::matrix_descr::triangular); - upper_A_view.uplo_view = oneapi::mkl::uplo::upper; - std::set no_properties; + oneapi::math::index_base index_zero = oneapi::math::index_base::zero; + oneapi::math::sparse::spsv_alg default_alg = oneapi::math::sparse::spsv_alg::default_alg; + oneapi::math::sparse::spsv_alg no_optimize_alg = oneapi::math::sparse::spsv_alg::no_optimize_alg; + oneapi::math::sparse::matrix_view default_A_view(oneapi::math::sparse::matrix_descr::triangular); + oneapi::math::sparse::matrix_view upper_A_view(oneapi::math::sparse::matrix_descr::triangular); + upper_A_view.uplo_view = oneapi::math::uplo::upper; + std::set no_properties; bool no_reset_data = false; bool no_scalars_on_device = false; @@ -81,7 +81,7 @@ void test_helper_with_format(testFunctorI32 test_functor_i32, testFunctorI64 tes num_passed, num_skipped); // Test index_base 1 EXPECT_TRUE_OR_FUTURE_SKIP( - test_functor_i32(dev, format, m, density_A_matrix, oneapi::mkl::index_base::one, + test_functor_i32(dev, format, m, density_A_matrix, oneapi::math::index_base::one, transpose_val, alpha, default_alg, default_A_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); @@ -91,16 +91,16 @@ void test_helper_with_format(testFunctorI32 test_functor_i32, testFunctorI64 tes no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); // Test lower triangular unit diagonal matrix - oneapi::mkl::sparse::matrix_view triangular_unit_A_view( - oneapi::mkl::sparse::matrix_descr::triangular); - triangular_unit_A_view.diag_view = oneapi::mkl::diag::unit; + oneapi::math::sparse::matrix_view triangular_unit_A_view( + oneapi::math::sparse::matrix_descr::triangular); + triangular_unit_A_view.diag_view = oneapi::math::diag::unit; EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, m, density_A_matrix, index_zero, transpose_val, alpha, default_alg, triangular_unit_A_view, no_properties, no_reset_data, no_scalars_on_device), num_passed, num_skipped); // Test upper triangular unit diagonal matrix - triangular_unit_A_view.uplo_view = oneapi::mkl::uplo::upper; + triangular_unit_A_view.uplo_view = oneapi::math::uplo::upper; EXPECT_TRUE_OR_FUTURE_SKIP( test_functor_i32(dev, format, m, density_A_matrix, index_zero, transpose_val, alpha, default_alg, triangular_unit_A_view, no_properties, no_reset_data, @@ -158,7 +158,7 @@ void test_helper_with_format(testFunctorI32 test_functor_i32, testFunctorI64 tes */ template void test_helper(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, - sycl::device *dev, oneapi::mkl::transpose transpose_val, int &num_passed, + sycl::device *dev, oneapi::math::transpose transpose_val, int &num_passed, int &num_skipped) { test_helper_with_format(test_functor_i32, test_functor_i64, dev, sparse_matrix_format_t::CSR, transpose_val, num_passed, @@ -172,8 +172,8 @@ void test_helper(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i6 template void prepare_reference_spsv_data(sparse_matrix_format_t format, const intType *ia, const intType *ja, const fpType *a, intType m, intType nnz, - intType indexing, oneapi::mkl::transpose opA, const fpType *x, - fpType alpha, oneapi::mkl::sparse::matrix_view A_view, + intType indexing, oneapi::math::transpose opA, const fpType *x, + fpType alpha, oneapi::math::sparse::matrix_view A_view, fpType *y_ref) { std::size_t mu = static_cast(m); auto dense_opa = sparse_to_dense(format, ia, ja, a, mu, mu, static_cast(nnz), @@ -187,7 +187,7 @@ void prepare_reference_spsv_data(sparse_matrix_format_t format, const intType *i // Compute each element of the reference one after the other starting from 0 (resp. the end) for a lower (resp. upper) triangular matrix. // A matrix is considered lowered if it is lower and not transposed or upper and transposed. const bool is_lower = - (A_view.uplo_view == oneapi::mkl::uplo::lower) == (opA == oneapi::mkl::transpose::nontrans); + (A_view.uplo_view == oneapi::math::uplo::lower) == (opA == oneapi::math::transpose::nontrans); for (std::size_t row = 0; row < mu; row++) { std::size_t uplo_row = is_lower ? row : (mu - 1 - row); fpType rhs = alpha * x[uplo_row]; diff --git a/tests/unit_tests/sparse_blas/source/sparse_spmm_buffer.cpp b/tests/unit_tests/sparse_blas/source/sparse_spmm_buffer.cpp index b6f9e1185..330b06bba 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spmm_buffer.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spmm_buffer.cpp @@ -29,12 +29,12 @@ namespace { template int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, - intType ncols_C, double density_A_matrix, oneapi::mkl::index_base index, - oneapi::mkl::layout dense_matrix_layout, oneapi::mkl::transpose transpose_A, - oneapi::mkl::transpose transpose_B, fpType alpha, fpType beta, intType ldb, - intType ldc, oneapi::mkl::sparse::spmm_alg alg, - oneapi::mkl::sparse::matrix_view A_view, - const std::set &matrix_properties, + intType ncols_C, double density_A_matrix, oneapi::math::index_base index, + oneapi::math::layout dense_matrix_layout, oneapi::math::transpose transpose_A, + oneapi::math::transpose transpose_B, fpType alpha, fpType beta, intType ldb, + intType ldc, oneapi::math::sparse::spmm_alg alg, + oneapi::math::sparse::matrix_view A_view, + const std::set &matrix_properties, bool reset_data, bool test_scalar_on_device) { if (test_scalar_on_device) { // Scalars on the device is not planned to be supported with the buffer API @@ -50,11 +50,11 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, } auto [opa_nrows, opa_ncols] = swap_if_transposed(transpose_A, nrows_A, ncols_A); auto [opb_nrows, opb_ncols] = swap_if_transposed(transpose_B, opa_ncols, ncols_C); - intType indexing = (index == oneapi::mkl::index_base::zero) ? 0 : 1; - const bool is_sorted = matrix_properties.find(oneapi::mkl::sparse::matrix_property::sorted) != + intType indexing = (index == oneapi::math::index_base::zero) ? 0 : 1; + const bool is_sorted = matrix_properties.find(oneapi::math::sparse::matrix_property::sorted) != matrix_properties.cend(); const bool is_symmetric = - matrix_properties.find(oneapi::mkl::sparse::matrix_property::symmetric) != + matrix_properties.find(oneapi::math::sparse::matrix_property::symmetric) != matrix_properties.cend(); // Input matrix @@ -84,35 +84,35 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, auto b_buf = make_buffer(b_host); auto c_buf = make_buffer(c_host); - oneapi::mkl::sparse::matrix_handle_t A_handle = nullptr; - oneapi::mkl::sparse::dense_matrix_handle_t B_handle = nullptr; - oneapi::mkl::sparse::dense_matrix_handle_t C_handle = nullptr; - oneapi::mkl::sparse::spmm_descr_t descr = nullptr; + oneapi::math::sparse::matrix_handle_t A_handle = nullptr; + oneapi::math::sparse::dense_matrix_handle_t B_handle = nullptr; + oneapi::math::sparse::dense_matrix_handle_t C_handle = nullptr; + oneapi::math::sparse::spmm_descr_t descr = nullptr; try { init_sparse_matrix(main_queue, format, &A_handle, nrows_A, ncols_A, nnz, index, ia_buf, ja_buf, a_buf); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::mkl::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); } - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_matrix, main_queue, &B_handle, opb_nrows, + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_matrix, main_queue, &B_handle, opb_nrows, opb_ncols, ldb, dense_matrix_layout, b_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_matrix, main_queue, &C_handle, + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_matrix, main_queue, &C_handle, static_cast(opa_nrows), ncols_C, ldc, dense_matrix_layout, c_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_spmm_descr, main_queue, &descr); + CALL_RT_OR_CT(oneapi::math::sparse::init_spmm_descr, main_queue, &descr); std::size_t workspace_size = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spmm_buffer_size, main_queue, transpose_A, transpose_B, + CALL_RT_OR_CT(oneapi::math::sparse::spmm_buffer_size, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, workspace_size); sycl::buffer workspace_buf((sycl::range<1>(workspace_size))); - CALL_RT_OR_CT(oneapi::mkl::sparse::spmm_optimize, main_queue, transpose_A, transpose_B, + CALL_RT_OR_CT(oneapi::math::sparse::spmm_optimize, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, workspace_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::spmm, main_queue, transpose_A, transpose_B, &alpha, + CALL_RT_OR_CT(oneapi::math::sparse::spmm, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr); if (reset_data) { @@ -139,18 +139,18 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, ja_buf, a_buf); std::size_t workspace_size_2 = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spmm_buffer_size, main_queue, transpose_A, + CALL_RT_OR_CT(oneapi::math::sparse::spmm_buffer_size, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, workspace_size_2); if (workspace_size_2 > workspace_size) { workspace_buf = sycl::buffer((sycl::range<1>(workspace_size_2))); } - CALL_RT_OR_CT(oneapi::mkl::sparse::spmm_optimize, main_queue, transpose_A, transpose_B, + CALL_RT_OR_CT(oneapi::math::sparse::spmm_optimize, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, workspace_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::spmm, main_queue, transpose_A, transpose_B, &alpha, + CALL_RT_OR_CT(oneapi::math::sparse::spmm, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr); } } @@ -160,11 +160,11 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, print_error_code(e); return 0; } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { wait_and_free_handles(main_queue, A_handle, B_handle, C_handle); if (descr) { sycl::event ev_release_descr; - CALL_RT_OR_CT(ev_release_descr = oneapi::mkl::sparse::release_spmm_descr, main_queue, + CALL_RT_OR_CT(ev_release_descr = oneapi::math::sparse::release_spmm_descr, main_queue, descr); ev_release_descr.wait(); } @@ -174,7 +174,7 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, std::cout << "Error raised during execution of sparse SPMM:\n" << error.what() << std::endl; return 0; } - CALL_RT_OR_CT(oneapi::mkl::sparse::release_spmm_descr, main_queue, descr); + CALL_RT_OR_CT(oneapi::math::sparse::release_spmm_descr, main_queue, descr); free_handles(main_queue, A_handle, B_handle, C_handle); // Compute reference. diff --git a/tests/unit_tests/sparse_blas/source/sparse_spmm_usm.cpp b/tests/unit_tests/sparse_blas/source/sparse_spmm_usm.cpp index 5778430a6..334604a5a 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spmm_usm.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spmm_usm.cpp @@ -29,12 +29,12 @@ namespace { template int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, - intType ncols_C, double density_A_matrix, oneapi::mkl::index_base index, - oneapi::mkl::layout dense_matrix_layout, oneapi::mkl::transpose transpose_A, - oneapi::mkl::transpose transpose_B, fpType alpha, fpType beta, intType ldb, - intType ldc, oneapi::mkl::sparse::spmm_alg alg, - oneapi::mkl::sparse::matrix_view A_view, - const std::set &matrix_properties, + intType ncols_C, double density_A_matrix, oneapi::math::index_base index, + oneapi::math::layout dense_matrix_layout, oneapi::math::transpose transpose_A, + oneapi::math::transpose transpose_B, fpType alpha, fpType beta, intType ldb, + intType ldc, oneapi::math::sparse::spmm_alg alg, + oneapi::math::sparse::matrix_view A_view, + const std::set &matrix_properties, bool reset_data, bool test_scalar_on_device) { sycl::queue main_queue(*dev, exception_handler_t()); @@ -46,11 +46,11 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, } auto [opa_nrows, opa_ncols] = swap_if_transposed(transpose_A, nrows_A, ncols_A); auto [opb_nrows, opb_ncols] = swap_if_transposed(transpose_B, opa_ncols, ncols_C); - intType indexing = (index == oneapi::mkl::index_base::zero) ? 0 : 1; - const bool is_sorted = matrix_properties.find(oneapi::mkl::sparse::matrix_property::sorted) != + intType indexing = (index == oneapi::math::index_base::zero) ? 0 : 1; + const bool is_sorted = matrix_properties.find(oneapi::math::sparse::matrix_property::sorted) != matrix_properties.cend(); const bool is_symmetric = - matrix_properties.find(oneapi::mkl::sparse::matrix_property::symmetric) != + matrix_properties.find(oneapi::math::sparse::matrix_property::symmetric) != matrix_properties.cend(); // Input matrix @@ -113,38 +113,38 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, } sycl::event ev_copy, ev_spmm; - oneapi::mkl::sparse::matrix_handle_t A_handle = nullptr; - oneapi::mkl::sparse::dense_matrix_handle_t B_handle = nullptr; - oneapi::mkl::sparse::dense_matrix_handle_t C_handle = nullptr; - oneapi::mkl::sparse::spmm_descr_t descr = nullptr; + oneapi::math::sparse::matrix_handle_t A_handle = nullptr; + oneapi::math::sparse::dense_matrix_handle_t B_handle = nullptr; + oneapi::math::sparse::dense_matrix_handle_t C_handle = nullptr; + oneapi::math::sparse::spmm_descr_t descr = nullptr; std::unique_ptr workspace_usm(nullptr, UsmDeleter(main_queue)); try { init_sparse_matrix(main_queue, format, &A_handle, nrows_A, ncols_A, nnz, index, ia_usm, ja_usm, a_usm); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::mkl::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); } - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_matrix, main_queue, &B_handle, opb_nrows, + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_matrix, main_queue, &B_handle, opb_nrows, opb_ncols, ldb, dense_matrix_layout, b_usm); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_matrix, main_queue, &C_handle, + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_matrix, main_queue, &C_handle, static_cast(opa_nrows), ncols_C, ldc, dense_matrix_layout, c_usm); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_spmm_descr, main_queue, &descr); + CALL_RT_OR_CT(oneapi::math::sparse::init_spmm_descr, main_queue, &descr); std::size_t workspace_size = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spmm_buffer_size, main_queue, transpose_A, transpose_B, + CALL_RT_OR_CT(oneapi::math::sparse::spmm_buffer_size, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, workspace_size); workspace_usm = malloc_device_uptr(main_queue, workspace_size); sycl::event ev_opt; - CALL_RT_OR_CT(ev_opt = oneapi::mkl::sparse::spmm_optimize, main_queue, transpose_A, + CALL_RT_OR_CT(ev_opt = oneapi::math::sparse::spmm_optimize, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, workspace_usm.get(), mat_dependencies); spmm_dependencies.push_back(ev_opt); - CALL_RT_OR_CT(ev_spmm = oneapi::mkl::sparse::spmm, main_queue, transpose_A, transpose_B, + CALL_RT_OR_CT(ev_spmm = oneapi::math::sparse::spmm, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, spmm_dependencies); @@ -181,18 +181,18 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, ja_usm, a_usm); std::size_t workspace_size_2 = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spmm_buffer_size, main_queue, transpose_A, + CALL_RT_OR_CT(oneapi::math::sparse::spmm_buffer_size, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, workspace_size_2); if (workspace_size_2 > workspace_size) { workspace_usm = malloc_device_uptr(main_queue, workspace_size_2); } - CALL_RT_OR_CT(ev_opt = oneapi::mkl::sparse::spmm_optimize, main_queue, transpose_A, + CALL_RT_OR_CT(ev_opt = oneapi::math::sparse::spmm_optimize, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, workspace_usm.get(), mat_dependencies); - CALL_RT_OR_CT(ev_spmm = oneapi::mkl::sparse::spmm, main_queue, transpose_A, transpose_B, + CALL_RT_OR_CT(ev_spmm = oneapi::math::sparse::spmm, main_queue, transpose_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, { ev_opt }); } @@ -205,11 +205,11 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, print_error_code(e); return 0; } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { wait_and_free_handles(main_queue, A_handle, B_handle, C_handle); if (descr) { sycl::event ev_release_descr; - CALL_RT_OR_CT(ev_release_descr = oneapi::mkl::sparse::release_spmm_descr, main_queue, + CALL_RT_OR_CT(ev_release_descr = oneapi::math::sparse::release_spmm_descr, main_queue, descr); ev_release_descr.wait(); } @@ -220,7 +220,7 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, return 0; } sycl::event ev_release_descr; - CALL_RT_OR_CT(ev_release_descr = oneapi::mkl::sparse::release_spmm_descr, main_queue, descr, + CALL_RT_OR_CT(ev_release_descr = oneapi::math::sparse::release_spmm_descr, main_queue, descr, { ev_spmm }); ev_release_descr.wait_and_throw(); free_handles(main_queue, { ev_spmm }, A_handle, B_handle, C_handle); diff --git a/tests/unit_tests/sparse_blas/source/sparse_spmv_buffer.cpp b/tests/unit_tests/sparse_blas/source/sparse_spmv_buffer.cpp index 3d99f9e94..0c4136bcc 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spmv_buffer.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spmv_buffer.cpp @@ -29,10 +29,10 @@ namespace { template int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, - double density_A_matrix, oneapi::mkl::index_base index, - oneapi::mkl::transpose transpose_val, fpType alpha, fpType beta, - oneapi::mkl::sparse::spmv_alg alg, oneapi::mkl::sparse::matrix_view A_view, - const std::set &matrix_properties, + double density_A_matrix, oneapi::math::index_base index, + oneapi::math::transpose transpose_val, fpType alpha, fpType beta, + oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::matrix_view A_view, + const std::set &matrix_properties, bool reset_data, bool test_scalar_on_device) { if (test_scalar_on_device) { // Scalars on the device is not planned to be supported with the buffer API @@ -44,11 +44,11 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, ncols_A = nrows_A; } auto [opa_nrows, opa_ncols] = swap_if_transposed(transpose_val, nrows_A, ncols_A); - intType indexing = (index == oneapi::mkl::index_base::zero) ? 0 : 1; - const bool is_sorted = matrix_properties.find(oneapi::mkl::sparse::matrix_property::sorted) != + intType indexing = (index == oneapi::math::index_base::zero) ? 0 : 1; + const bool is_sorted = matrix_properties.find(oneapi::math::sparse::matrix_property::sorted) != matrix_properties.cend(); const bool is_symmetric = - matrix_properties.find(oneapi::mkl::sparse::matrix_property::symmetric) != + matrix_properties.find(oneapi::math::sparse::matrix_property::symmetric) != matrix_properties.cend(); // Input matrix @@ -77,32 +77,32 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, auto x_buf = make_buffer(x_host); auto y_buf = make_buffer(y_host); - oneapi::mkl::sparse::matrix_handle_t A_handle = nullptr; - oneapi::mkl::sparse::dense_vector_handle_t x_handle = nullptr; - oneapi::mkl::sparse::dense_vector_handle_t y_handle = nullptr; - oneapi::mkl::sparse::spmv_descr_t descr = nullptr; + oneapi::math::sparse::matrix_handle_t A_handle = nullptr; + oneapi::math::sparse::dense_vector_handle_t x_handle = nullptr; + oneapi::math::sparse::dense_vector_handle_t y_handle = nullptr; + oneapi::math::sparse::spmv_descr_t descr = nullptr; try { init_sparse_matrix(main_queue, format, &A_handle, nrows_A, ncols_A, nnz, index, ia_buf, ja_buf, a_buf); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::mkl::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); } - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_vector, main_queue, &x_handle, + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &x_handle, static_cast(x_host.size()), x_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_vector, main_queue, &y_handle, + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &y_handle, static_cast(y_host.size()), y_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_spmv_descr, main_queue, &descr); + CALL_RT_OR_CT(oneapi::math::sparse::init_spmv_descr, main_queue, &descr); std::size_t workspace_size = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spmv_buffer_size, main_queue, transpose_val, &alpha, + CALL_RT_OR_CT(oneapi::math::sparse::spmv_buffer_size, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr, workspace_size); sycl::buffer workspace_buf((sycl::range<1>(workspace_size))); - CALL_RT_OR_CT(oneapi::mkl::sparse::spmv_optimize, main_queue, transpose_val, &alpha, A_view, + CALL_RT_OR_CT(oneapi::math::sparse::spmv_optimize, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr, workspace_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::spmv, main_queue, transpose_val, &alpha, A_view, + CALL_RT_OR_CT(oneapi::math::sparse::spmv, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr); if (reset_data) { @@ -129,17 +129,17 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, ja_buf, a_buf); std::size_t workspace_size_2 = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spmv_buffer_size, main_queue, transpose_val, &alpha, + CALL_RT_OR_CT(oneapi::math::sparse::spmv_buffer_size, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr, workspace_size_2); if (workspace_size_2 > workspace_size) { workspace_buf = sycl::buffer((sycl::range<1>(workspace_size_2))); } - CALL_RT_OR_CT(oneapi::mkl::sparse::spmv_optimize, main_queue, transpose_val, &alpha, + CALL_RT_OR_CT(oneapi::math::sparse::spmv_optimize, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr, workspace_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::spmv, main_queue, transpose_val, &alpha, A_view, + CALL_RT_OR_CT(oneapi::math::sparse::spmv, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr); } } @@ -149,11 +149,11 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, print_error_code(e); return 0; } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { wait_and_free_handles(main_queue, A_handle, x_handle, y_handle); if (descr) { sycl::event ev_release_descr; - CALL_RT_OR_CT(ev_release_descr = oneapi::mkl::sparse::release_spmv_descr, main_queue, + CALL_RT_OR_CT(ev_release_descr = oneapi::math::sparse::release_spmv_descr, main_queue, descr); ev_release_descr.wait(); } @@ -163,7 +163,7 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, std::cout << "Error raised during execution of sparse SPMV:\n" << error.what() << std::endl; return 0; } - CALL_RT_OR_CT(oneapi::mkl::sparse::release_spmv_descr, main_queue, descr); + CALL_RT_OR_CT(oneapi::math::sparse::release_spmv_descr, main_queue, descr); free_handles(main_queue, A_handle, x_handle, y_handle); // Compute reference. diff --git a/tests/unit_tests/sparse_blas/source/sparse_spmv_usm.cpp b/tests/unit_tests/sparse_blas/source/sparse_spmv_usm.cpp index ded92a770..a3265ea70 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spmv_usm.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spmv_usm.cpp @@ -29,10 +29,10 @@ namespace { template int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, - double density_A_matrix, oneapi::mkl::index_base index, - oneapi::mkl::transpose transpose_val, fpType alpha, fpType beta, - oneapi::mkl::sparse::spmv_alg alg, oneapi::mkl::sparse::matrix_view A_view, - const std::set &matrix_properties, + double density_A_matrix, oneapi::math::index_base index, + oneapi::math::transpose transpose_val, fpType alpha, fpType beta, + oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::matrix_view A_view, + const std::set &matrix_properties, bool reset_data, bool test_scalar_on_device) { sycl::queue main_queue(*dev, exception_handler_t()); @@ -40,11 +40,11 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, ncols_A = nrows_A; } auto [opa_nrows, opa_ncols] = swap_if_transposed(transpose_val, nrows_A, ncols_A); - intType indexing = (index == oneapi::mkl::index_base::zero) ? 0 : 1; - const bool is_sorted = matrix_properties.find(oneapi::mkl::sparse::matrix_property::sorted) != + intType indexing = (index == oneapi::math::index_base::zero) ? 0 : 1; + const bool is_sorted = matrix_properties.find(oneapi::math::sparse::matrix_property::sorted) != matrix_properties.cend(); const bool is_symmetric = - matrix_properties.find(oneapi::mkl::sparse::matrix_property::symmetric) != + matrix_properties.find(oneapi::math::sparse::matrix_property::symmetric) != matrix_properties.cend(); // Input matrix @@ -106,37 +106,37 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, } sycl::event ev_copy, ev_spmv; - oneapi::mkl::sparse::matrix_handle_t A_handle = nullptr; - oneapi::mkl::sparse::dense_vector_handle_t x_handle = nullptr; - oneapi::mkl::sparse::dense_vector_handle_t y_handle = nullptr; - oneapi::mkl::sparse::spmv_descr_t descr = nullptr; + oneapi::math::sparse::matrix_handle_t A_handle = nullptr; + oneapi::math::sparse::dense_vector_handle_t x_handle = nullptr; + oneapi::math::sparse::dense_vector_handle_t y_handle = nullptr; + oneapi::math::sparse::spmv_descr_t descr = nullptr; std::unique_ptr workspace_usm(nullptr, UsmDeleter(main_queue)); try { init_sparse_matrix(main_queue, format, &A_handle, nrows_A, ncols_A, nnz, index, ia_usm, ja_usm, a_usm); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::mkl::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); } - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_vector, main_queue, &x_handle, + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &x_handle, static_cast(x_host.size()), x_usm); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_vector, main_queue, &y_handle, + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &y_handle, static_cast(y_host.size()), y_usm); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_spmv_descr, main_queue, &descr); + CALL_RT_OR_CT(oneapi::math::sparse::init_spmv_descr, main_queue, &descr); std::size_t workspace_size = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spmv_buffer_size, main_queue, transpose_val, + CALL_RT_OR_CT(oneapi::math::sparse::spmv_buffer_size, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, beta_host_or_usm_ptr, y_handle, alg, descr, workspace_size); workspace_usm = malloc_device_uptr(main_queue, workspace_size); sycl::event ev_opt; - CALL_RT_OR_CT(ev_opt = oneapi::mkl::sparse::spmv_optimize, main_queue, transpose_val, + CALL_RT_OR_CT(ev_opt = oneapi::math::sparse::spmv_optimize, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, beta_host_or_usm_ptr, y_handle, alg, descr, workspace_usm.get(), mat_dependencies); spmv_dependencies.push_back(ev_opt); - CALL_RT_OR_CT(ev_spmv = oneapi::mkl::sparse::spmv, main_queue, transpose_val, + CALL_RT_OR_CT(ev_spmv = oneapi::math::sparse::spmv, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, beta_host_or_usm_ptr, y_handle, alg, descr, spmv_dependencies); @@ -173,18 +173,18 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, ja_usm, a_usm); std::size_t workspace_size_2 = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spmv_buffer_size, main_queue, transpose_val, + CALL_RT_OR_CT(oneapi::math::sparse::spmv_buffer_size, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, beta_host_or_usm_ptr, y_handle, alg, descr, workspace_size_2); if (workspace_size_2 > workspace_size) { workspace_usm = malloc_device_uptr(main_queue, workspace_size_2); } - CALL_RT_OR_CT(ev_opt = oneapi::mkl::sparse::spmv_optimize, main_queue, transpose_val, + CALL_RT_OR_CT(ev_opt = oneapi::math::sparse::spmv_optimize, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, beta_host_or_usm_ptr, y_handle, alg, descr, workspace_usm.get(), mat_dependencies); - CALL_RT_OR_CT(ev_spmv = oneapi::mkl::sparse::spmv, main_queue, transpose_val, + CALL_RT_OR_CT(ev_spmv = oneapi::math::sparse::spmv, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, beta_host_or_usm_ptr, y_handle, alg, descr, { ev_opt }); } @@ -197,11 +197,11 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, print_error_code(e); return 0; } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { wait_and_free_handles(main_queue, A_handle, x_handle, y_handle); if (descr) { sycl::event ev_release_descr; - CALL_RT_OR_CT(ev_release_descr = oneapi::mkl::sparse::release_spmv_descr, main_queue, + CALL_RT_OR_CT(ev_release_descr = oneapi::math::sparse::release_spmv_descr, main_queue, descr); ev_release_descr.wait(); } @@ -212,7 +212,7 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, return 0; } sycl::event ev_release_descr; - CALL_RT_OR_CT(ev_release_descr = oneapi::mkl::sparse::release_spmv_descr, main_queue, descr, + CALL_RT_OR_CT(ev_release_descr = oneapi::math::sparse::release_spmv_descr, main_queue, descr, { ev_spmv }); ev_release_descr.wait_and_throw(); free_handles(main_queue, { ev_spmv }, A_handle, x_handle, y_handle); diff --git a/tests/unit_tests/sparse_blas/source/sparse_spsv_buffer.cpp b/tests/unit_tests/sparse_blas/source/sparse_spsv_buffer.cpp index 6b276dff4..62a84120a 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spsv_buffer.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spsv_buffer.cpp @@ -29,9 +29,9 @@ namespace { template int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, double density_A_matrix, - oneapi::mkl::index_base index, oneapi::mkl::transpose transpose_val, fpType alpha, - oneapi::mkl::sparse::spsv_alg alg, oneapi::mkl::sparse::matrix_view A_view, - const std::set &matrix_properties, + oneapi::math::index_base index, oneapi::math::transpose transpose_val, fpType alpha, + oneapi::math::sparse::spsv_alg alg, oneapi::math::sparse::matrix_view A_view, + const std::set &matrix_properties, bool reset_data, bool test_scalar_on_device) { if (test_scalar_on_device) { // Scalars on the device is not planned to be supported with the buffer API @@ -39,12 +39,12 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl } sycl::queue main_queue(*dev, exception_handler_t()); - intType indexing = (index == oneapi::mkl::index_base::zero) ? 0 : 1; + intType indexing = (index == oneapi::math::index_base::zero) ? 0 : 1; const std::size_t mu = static_cast(m); - const bool is_sorted = matrix_properties.find(oneapi::mkl::sparse::matrix_property::sorted) != + const bool is_sorted = matrix_properties.find(oneapi::math::sparse::matrix_property::sorted) != matrix_properties.cend(); const bool is_symmetric = - matrix_properties.find(oneapi::mkl::sparse::matrix_property::symmetric) != + matrix_properties.find(oneapi::math::sparse::matrix_property::symmetric) != matrix_properties.cend(); // Input matrix @@ -52,8 +52,8 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl std::vector a_host; // Set non-zero values to the diagonal, except if the matrix is viewed as a unit matrix. const bool require_diagonal = - !(A_view.type_view == oneapi::mkl::sparse::matrix_descr::diagonal && - A_view.diag_view == oneapi::mkl::diag::unit); + !(A_view.type_view == oneapi::math::sparse::matrix_descr::diagonal && + A_view.diag_view == oneapi::math::diag::unit); intType nnz = generate_random_matrix(format, m, m, density_A_matrix, indexing, ia_host, ja_host, a_host, is_symmetric, require_diagonal); @@ -80,29 +80,29 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl auto x_buf = make_buffer(x_host); auto y_buf = make_buffer(y_host); - oneapi::mkl::sparse::matrix_handle_t A_handle = nullptr; - oneapi::mkl::sparse::dense_vector_handle_t x_handle = nullptr; - oneapi::mkl::sparse::dense_vector_handle_t y_handle = nullptr; - oneapi::mkl::sparse::spsv_descr_t descr = nullptr; + oneapi::math::sparse::matrix_handle_t A_handle = nullptr; + oneapi::math::sparse::dense_vector_handle_t x_handle = nullptr; + oneapi::math::sparse::dense_vector_handle_t y_handle = nullptr; + oneapi::math::sparse::spsv_descr_t descr = nullptr; try { init_sparse_matrix(main_queue, format, &A_handle, m, m, nnz, index, ia_buf, ja_buf, a_buf); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::mkl::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); } - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_vector, main_queue, &x_handle, m, x_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_vector, main_queue, &y_handle, m, y_buf); + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &x_handle, m, x_buf); + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &y_handle, m, y_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_spsv_descr, main_queue, &descr); + CALL_RT_OR_CT(oneapi::math::sparse::init_spsv_descr, main_queue, &descr); std::size_t workspace_size = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spsv_buffer_size, main_queue, transpose_val, &alpha, + CALL_RT_OR_CT(oneapi::math::sparse::spsv_buffer_size, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, y_handle, alg, descr, workspace_size); sycl::buffer workspace_buf((sycl::range<1>(workspace_size))); - CALL_RT_OR_CT(oneapi::mkl::sparse::spsv_optimize, main_queue, transpose_val, &alpha, A_view, + CALL_RT_OR_CT(oneapi::math::sparse::spsv_optimize, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, y_handle, alg, descr, workspace_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::spsv, main_queue, transpose_val, &alpha, A_view, + CALL_RT_OR_CT(oneapi::math::sparse::spsv, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, y_handle, alg, descr); if (reset_data) { @@ -128,16 +128,16 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl set_matrix_data(main_queue, format, A_handle, m, m, nnz, index, ia_buf, ja_buf, a_buf); std::size_t workspace_size_2 = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spsv_buffer_size, main_queue, transpose_val, &alpha, + CALL_RT_OR_CT(oneapi::math::sparse::spsv_buffer_size, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, y_handle, alg, descr, workspace_size_2); if (workspace_size_2 > workspace_size) { workspace_buf = sycl::buffer((sycl::range<1>(workspace_size_2))); } - CALL_RT_OR_CT(oneapi::mkl::sparse::spsv_optimize, main_queue, transpose_val, &alpha, + CALL_RT_OR_CT(oneapi::math::sparse::spsv_optimize, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, y_handle, alg, descr, workspace_buf); - CALL_RT_OR_CT(oneapi::mkl::sparse::spsv, main_queue, transpose_val, &alpha, A_view, + CALL_RT_OR_CT(oneapi::math::sparse::spsv, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, y_handle, alg, descr); } } @@ -147,11 +147,11 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl print_error_code(e); return 0; } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { wait_and_free_handles(main_queue, A_handle, x_handle, y_handle); if (descr) { sycl::event ev_release_descr; - CALL_RT_OR_CT(ev_release_descr = oneapi::mkl::sparse::release_spsv_descr, main_queue, + CALL_RT_OR_CT(ev_release_descr = oneapi::math::sparse::release_spsv_descr, main_queue, descr); ev_release_descr.wait(); } @@ -161,7 +161,7 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl std::cout << "Error raised during execution of sparse SPSV:\n" << error.what() << std::endl; return 0; } - CALL_RT_OR_CT(oneapi::mkl::sparse::release_spsv_descr, main_queue, descr); + CALL_RT_OR_CT(oneapi::math::sparse::release_spsv_descr, main_queue, descr); free_handles(main_queue, A_handle, x_handle, y_handle); // Compute reference. @@ -182,9 +182,9 @@ TEST_P(SparseSpsvBufferTests, RealSinglePrecision) { using fpType = float; int num_passed = 0, num_skipped = 0; test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::nontrans, num_passed, num_skipped); + oneapi::math::transpose::nontrans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::trans, num_passed, num_skipped); + oneapi::math::transpose::trans, num_passed, num_skipped); if (num_skipped > 0) { // Mark that some tests were skipped GTEST_SKIP() << "Passed: " << num_passed << ", Skipped: " << num_skipped @@ -197,9 +197,9 @@ TEST_P(SparseSpsvBufferTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); int num_passed = 0, num_skipped = 0; test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::nontrans, num_passed, num_skipped); + oneapi::math::transpose::nontrans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::trans, num_passed, num_skipped); + oneapi::math::transpose::trans, num_passed, num_skipped); if (num_skipped > 0) { // Mark that some tests were skipped GTEST_SKIP() << "Passed: " << num_passed << ", Skipped: " << num_skipped @@ -211,11 +211,11 @@ TEST_P(SparseSpsvBufferTests, ComplexSinglePrecision) { using fpType = std::complex; int num_passed = 0, num_skipped = 0; test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::nontrans, num_passed, num_skipped); + oneapi::math::transpose::nontrans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::trans, num_passed, num_skipped); + oneapi::math::transpose::trans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::conjtrans, num_passed, num_skipped); + oneapi::math::transpose::conjtrans, num_passed, num_skipped); if (num_skipped > 0) { // Mark that some tests were skipped GTEST_SKIP() << "Passed: " << num_passed << ", Skipped: " << num_skipped @@ -228,11 +228,11 @@ TEST_P(SparseSpsvBufferTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); int num_passed = 0, num_skipped = 0; test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::nontrans, num_passed, num_skipped); + oneapi::math::transpose::nontrans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::trans, num_passed, num_skipped); + oneapi::math::transpose::trans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::conjtrans, num_passed, num_skipped); + oneapi::math::transpose::conjtrans, num_passed, num_skipped); if (num_skipped > 0) { // Mark that some tests were skipped GTEST_SKIP() << "Passed: " << num_passed << ", Skipped: " << num_skipped diff --git a/tests/unit_tests/sparse_blas/source/sparse_spsv_usm.cpp b/tests/unit_tests/sparse_blas/source/sparse_spsv_usm.cpp index 3b58db914..1555bd2c0 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spsv_usm.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spsv_usm.cpp @@ -29,18 +29,18 @@ namespace { template int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, double density_A_matrix, - oneapi::mkl::index_base index, oneapi::mkl::transpose transpose_val, fpType alpha, - oneapi::mkl::sparse::spsv_alg alg, oneapi::mkl::sparse::matrix_view A_view, - const std::set &matrix_properties, + oneapi::math::index_base index, oneapi::math::transpose transpose_val, fpType alpha, + oneapi::math::sparse::spsv_alg alg, oneapi::math::sparse::matrix_view A_view, + const std::set &matrix_properties, bool reset_data, bool test_scalar_on_device) { sycl::queue main_queue(*dev, exception_handler_t()); - intType indexing = (index == oneapi::mkl::index_base::zero) ? 0 : 1; + intType indexing = (index == oneapi::math::index_base::zero) ? 0 : 1; const std::size_t mu = static_cast(m); - const bool is_sorted = matrix_properties.find(oneapi::mkl::sparse::matrix_property::sorted) != + const bool is_sorted = matrix_properties.find(oneapi::math::sparse::matrix_property::sorted) != matrix_properties.cend(); const bool is_symmetric = - matrix_properties.find(oneapi::mkl::sparse::matrix_property::symmetric) != + matrix_properties.find(oneapi::math::sparse::matrix_property::symmetric) != matrix_properties.cend(); // Input matrix @@ -48,8 +48,8 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl std::vector a_host; // Set non-zero values to the diagonal, except if the matrix is viewed as a unit matrix. const bool require_diagonal = - !(A_view.type_view == oneapi::mkl::sparse::matrix_descr::diagonal && - A_view.diag_view == oneapi::mkl::diag::unit); + !(A_view.type_view == oneapi::math::sparse::matrix_descr::diagonal && + A_view.diag_view == oneapi::math::diag::unit); intType nnz = generate_random_matrix(format, m, m, density_A_matrix, indexing, ia_host, ja_host, a_host, is_symmetric, require_diagonal); @@ -105,34 +105,34 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl } sycl::event ev_copy, ev_spsv; - oneapi::mkl::sparse::matrix_handle_t A_handle = nullptr; - oneapi::mkl::sparse::dense_vector_handle_t x_handle = nullptr; - oneapi::mkl::sparse::dense_vector_handle_t y_handle = nullptr; - oneapi::mkl::sparse::spsv_descr_t descr = nullptr; + oneapi::math::sparse::matrix_handle_t A_handle = nullptr; + oneapi::math::sparse::dense_vector_handle_t x_handle = nullptr; + oneapi::math::sparse::dense_vector_handle_t y_handle = nullptr; + oneapi::math::sparse::spsv_descr_t descr = nullptr; std::unique_ptr workspace_usm(nullptr, UsmDeleter(main_queue)); try { init_sparse_matrix(main_queue, format, &A_handle, m, m, nnz, index, ia_usm, ja_usm, a_usm); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::mkl::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); } - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_vector, main_queue, &x_handle, m, x_usm); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_dense_vector, main_queue, &y_handle, m, y_usm); + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &x_handle, m, x_usm); + CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &y_handle, m, y_usm); - CALL_RT_OR_CT(oneapi::mkl::sparse::init_spsv_descr, main_queue, &descr); + CALL_RT_OR_CT(oneapi::math::sparse::init_spsv_descr, main_queue, &descr); std::size_t workspace_size = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spsv_buffer_size, main_queue, transpose_val, + CALL_RT_OR_CT(oneapi::math::sparse::spsv_buffer_size, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, y_handle, alg, descr, workspace_size); workspace_usm = malloc_device_uptr(main_queue, workspace_size); sycl::event ev_opt; - CALL_RT_OR_CT(ev_opt = oneapi::mkl::sparse::spsv_optimize, main_queue, transpose_val, + CALL_RT_OR_CT(ev_opt = oneapi::math::sparse::spsv_optimize, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, y_handle, alg, descr, workspace_usm.get(), mat_dependencies); spsv_dependencies.push_back(ev_opt); - CALL_RT_OR_CT(ev_spsv = oneapi::mkl::sparse::spsv, main_queue, transpose_val, + CALL_RT_OR_CT(ev_spsv = oneapi::math::sparse::spsv, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, y_handle, alg, descr, spsv_dependencies); @@ -168,18 +168,18 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl set_matrix_data(main_queue, format, A_handle, m, m, nnz, index, ia_usm, ja_usm, a_usm); std::size_t workspace_size_2 = 0; - CALL_RT_OR_CT(oneapi::mkl::sparse::spsv_buffer_size, main_queue, transpose_val, + CALL_RT_OR_CT(oneapi::math::sparse::spsv_buffer_size, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, y_handle, alg, descr, workspace_size_2); if (workspace_size_2 > workspace_size) { workspace_usm = malloc_device_uptr(main_queue, workspace_size_2); } - CALL_RT_OR_CT(ev_opt = oneapi::mkl::sparse::spsv_optimize, main_queue, transpose_val, + CALL_RT_OR_CT(ev_opt = oneapi::math::sparse::spsv_optimize, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, y_handle, alg, descr, workspace_usm.get(), mat_dependencies); - CALL_RT_OR_CT(ev_spsv = oneapi::mkl::sparse::spsv, main_queue, transpose_val, + CALL_RT_OR_CT(ev_spsv = oneapi::math::sparse::spsv, main_queue, transpose_val, alpha_host_or_usm_ptr, A_view, A_handle, x_handle, y_handle, alg, descr, { ev_opt }); } @@ -192,11 +192,11 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl print_error_code(e); return 0; } - catch (const oneapi::mkl::unimplemented &e) { + catch (const oneapi::math::unimplemented &e) { wait_and_free_handles(main_queue, A_handle, x_handle, y_handle); if (descr) { sycl::event ev_release_descr; - CALL_RT_OR_CT(ev_release_descr = oneapi::mkl::sparse::release_spsv_descr, main_queue, + CALL_RT_OR_CT(ev_release_descr = oneapi::math::sparse::release_spsv_descr, main_queue, descr); ev_release_descr.wait(); } @@ -207,7 +207,7 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl return 0; } sycl::event ev_release_descr; - CALL_RT_OR_CT(ev_release_descr = oneapi::mkl::sparse::release_spsv_descr, main_queue, descr, + CALL_RT_OR_CT(ev_release_descr = oneapi::math::sparse::release_spsv_descr, main_queue, descr, { ev_spsv }); ev_release_descr.wait_and_throw(); free_handles(main_queue, { ev_spsv }, A_handle, x_handle, y_handle); @@ -230,9 +230,9 @@ TEST_P(SparseSpsvUsmTests, RealSinglePrecision) { using fpType = float; int num_passed = 0, num_skipped = 0; test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::nontrans, num_passed, num_skipped); + oneapi::math::transpose::nontrans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::trans, num_passed, num_skipped); + oneapi::math::transpose::trans, num_passed, num_skipped); if (num_skipped > 0) { // Mark that some tests were skipped GTEST_SKIP() << "Passed: " << num_passed << ", Skipped: " << num_skipped @@ -245,9 +245,9 @@ TEST_P(SparseSpsvUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); int num_passed = 0, num_skipped = 0; test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::nontrans, num_passed, num_skipped); + oneapi::math::transpose::nontrans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::trans, num_passed, num_skipped); + oneapi::math::transpose::trans, num_passed, num_skipped); if (num_skipped > 0) { // Mark that some tests were skipped GTEST_SKIP() << "Passed: " << num_passed << ", Skipped: " << num_skipped @@ -259,11 +259,11 @@ TEST_P(SparseSpsvUsmTests, ComplexSinglePrecision) { using fpType = std::complex; int num_passed = 0, num_skipped = 0; test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::nontrans, num_passed, num_skipped); + oneapi::math::transpose::nontrans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::trans, num_passed, num_skipped); + oneapi::math::transpose::trans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::conjtrans, num_passed, num_skipped); + oneapi::math::transpose::conjtrans, num_passed, num_skipped); if (num_skipped > 0) { // Mark that some tests were skipped GTEST_SKIP() << "Passed: " << num_passed << ", Skipped: " << num_skipped @@ -276,11 +276,11 @@ TEST_P(SparseSpsvUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); int num_passed = 0, num_skipped = 0; test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::nontrans, num_passed, num_skipped); + oneapi::math::transpose::nontrans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::trans, num_passed, num_skipped); + oneapi::math::transpose::trans, num_passed, num_skipped); test_helper(test_spsv, test_spsv, GetParam(), - oneapi::mkl::transpose::conjtrans, num_passed, num_skipped); + oneapi::math::transpose::conjtrans, num_passed, num_skipped); if (num_skipped > 0) { // Mark that some tests were skipped GTEST_SKIP() << "Passed: " << num_passed << ", Skipped: " << num_skipped From c831ca08719754057994608ff005230bcb0da0d4 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Wed, 9 Oct 2024 18:53:45 +0200 Subject: [PATCH 17/41] Rename _MKL_ macro prefix to ONEMATH_ --- include/oneapi/math/rng/device.hpp | 6 +++--- include/oneapi/math/rng/device/detail/bernoulli_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/beta_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/bits_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/distribution_base.hpp | 6 +++--- include/oneapi/math/rng/device/detail/engine_base.hpp | 6 +++--- include/oneapi/math/rng/device/detail/exponential_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/gamma_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/gaussian_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/lognormal_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/mcg31m1_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/mcg59_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp | 6 +++--- .../math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp | 6 +++--- .../oneapi/math/rng/device/detail/philox4x32x10_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/poisson_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/uniform_bits_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/uniform_impl.hpp | 6 +++--- include/oneapi/math/rng/device/detail/vm_wrappers.hpp | 6 +++--- include/oneapi/math/rng/device/distributions.hpp | 6 +++--- include/oneapi/math/rng/device/engines.hpp | 6 +++--- include/oneapi/math/rng/device/functions.hpp | 6 +++--- include/oneapi/math/rng/device/types.hpp | 6 +++--- src/blas/backends/cublas/cublas_task.hpp | 6 +++--- src/lapack/backends/cusolver/cusolver_task.hpp | 6 +++--- src/lapack/backends/rocsolver/rocsolver_task.hpp | 6 +++--- src/rng/backends/curand/curand_helper.hpp | 6 +++--- src/rng/backends/curand/curand_task.hpp | 4 ++-- src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp | 4 ++-- src/rng/backends/rocrand/rocrand_helper.hpp | 6 +++--- src/rng/backends/rocrand/rocrand_task.hpp | 4 ++-- 31 files changed, 90 insertions(+), 90 deletions(-) diff --git a/include/oneapi/math/rng/device.hpp b/include/oneapi/math/rng/device.hpp index 88caa9fb9..fb0734175 100644 --- a/include/oneapi/math/rng/device.hpp +++ b/include/oneapi/math/rng/device.hpp @@ -17,12 +17,12 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_SYCL_DEVICE_HPP__ -#define _MKL_RNG_SYCL_DEVICE_HPP__ +#ifndef ONEMATH_RNG_SYCL_DEVICE_HPP__ +#define ONEMATH_RNG_SYCL_DEVICE_HPP__ #include "oneapi/math/rng/device/types.hpp" #include "oneapi/math/rng/device/functions.hpp" #include "oneapi/math/rng/device/distributions.hpp" #include "oneapi/math/rng/device/engines.hpp" -#endif // _MKL_RNG_SYCL_DEVICE_HPP__ +#endif // ONEMATH_RNG_SYCL_DEVICE_HPP__ diff --git a/include/oneapi/math/rng/device/detail/bernoulli_impl.hpp b/include/oneapi/math/rng/device/detail/bernoulli_impl.hpp index 9fb3d0559..0521d7f75 100644 --- a/include/oneapi/math/rng/device/detail/bernoulli_impl.hpp +++ b/include/oneapi/math/rng/device/detail/bernoulli_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_BERNOULLI_IMPL_HPP_ -#define _MKL_RNG_DEVICE_BERNOULLI_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_BERNOULLI_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_BERNOULLI_IMPL_HPP_ namespace oneapi::math::rng::device::detail { @@ -86,4 +86,4 @@ class distribution_base> { } // namespace oneapi::math::rng::device::detail -#endif // _MKL_RNG_DEVICE_BERNOULLI_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_BERNOULLI_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/beta_impl.hpp b/include/oneapi/math/rng/device/detail/beta_impl.hpp index fcb9db459..405b1c4d6 100644 --- a/include/oneapi/math/rng/device/detail/beta_impl.hpp +++ b/include/oneapi/math/rng/device/detail/beta_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_BETA_IMPL_HPP_ -#define _MKL_RNG_DEVICE_BETA_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_BETA_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_BETA_IMPL_HPP_ #include "vm_wrappers.hpp" @@ -465,4 +465,4 @@ class distribution_base> { } // namespace oneapi::math::rng::device::detail -#endif // _MKL_RNG_DEVICE_BETA_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_BETA_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/bits_impl.hpp b/include/oneapi/math/rng/device/detail/bits_impl.hpp index f48364638..398c25388 100644 --- a/include/oneapi/math/rng/device/detail/bits_impl.hpp +++ b/include/oneapi/math/rng/device/detail/bits_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_BITS_IMPL_HPP_ -#define _MKL_RNG_DEVICE_BITS_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_BITS_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_BITS_IMPL_HPP_ #include "engine_base.hpp" @@ -68,4 +68,4 @@ class distribution_base> { } // namespace oneapi::math::rng::device::detail -#endif // _MKL_RNG_DEVICE_BITS_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_BITS_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/distribution_base.hpp b/include/oneapi/math/rng/device/detail/distribution_base.hpp index b511e20b6..4faf6cd49 100644 --- a/include/oneapi/math/rng/device/detail/distribution_base.hpp +++ b/include/oneapi/math/rng/device/detail/distribution_base.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DISTRIBUTION_BASE_HPP_ -#define _MKL_RNG_DISTRIBUTION_BASE_HPP_ +#ifndef ONEMATH_RNG_DISTRIBUTION_BASE_HPP_ +#define ONEMATH_RNG_DISTRIBUTION_BASE_HPP_ #include @@ -78,4 +78,4 @@ class bernoulli; #include "oneapi/math/rng/device/detail/beta_impl.hpp" #include "oneapi/math/rng/device/detail/gamma_impl.hpp" -#endif // _MKL_RNG_DISTRIBUTION_BASE_HPP_ +#endif // ONEMATH_RNG_DISTRIBUTION_BASE_HPP_ diff --git a/include/oneapi/math/rng/device/detail/engine_base.hpp b/include/oneapi/math/rng/device/detail/engine_base.hpp index e104620fd..07efadb41 100644 --- a/include/oneapi/math/rng/device/detail/engine_base.hpp +++ b/include/oneapi/math/rng/device/detail/engine_base.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_ENGINE_BASE_HPP_ -#define _MKL_RNG_DEVICE_ENGINE_BASE_HPP_ +#ifndef ONEMATH_RNG_DEVICE_ENGINE_BASE_HPP_ +#define ONEMATH_RNG_DEVICE_ENGINE_BASE_HPP_ #include @@ -40,4 +40,4 @@ class engine_base {}; #include "oneapi/math/rng/device/detail/mcg31m1_impl.hpp" #include "oneapi/math/rng/device/detail/mcg59_impl.hpp" -#endif // _MKL_RNG_DEVICE_ENGINE_BASE_HPP_ +#endif // ONEMATH_RNG_DEVICE_ENGINE_BASE_HPP_ diff --git a/include/oneapi/math/rng/device/detail/exponential_impl.hpp b/include/oneapi/math/rng/device/detail/exponential_impl.hpp index 3db32fc13..f754fd548 100644 --- a/include/oneapi/math/rng/device/detail/exponential_impl.hpp +++ b/include/oneapi/math/rng/device/detail/exponential_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_EXPONENTIAL_IMPL_HPP_ -#define _MKL_RNG_DEVICE_EXPONENTIAL_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_EXPONENTIAL_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_EXPONENTIAL_IMPL_HPP_ #include "vm_wrappers.hpp" @@ -137,4 +137,4 @@ class distribution_base } // namespace oneapi::math::rng::device::detail -#endif // _MKL_RNG_DEVICE_EXPONENTIAL_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_EXPONENTIAL_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/gamma_impl.hpp b/include/oneapi/math/rng/device/detail/gamma_impl.hpp index 7371f8a8a..f338f2705 100644 --- a/include/oneapi/math/rng/device/detail/gamma_impl.hpp +++ b/include/oneapi/math/rng/device/detail/gamma_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_GAMMA_IMPL_HPP_ -#define _MKL_RNG_DEVICE_GAMMA_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_GAMMA_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_GAMMA_IMPL_HPP_ #include "vm_wrappers.hpp" @@ -284,4 +284,4 @@ class distribution_base> { } // namespace oneapi::math::rng::device::detail -#endif // _MKL_RNG_DEVICE_GAMMA_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_GAMMA_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/gaussian_impl.hpp b/include/oneapi/math/rng/device/detail/gaussian_impl.hpp index ff0b36471..a0de61e60 100644 --- a/include/oneapi/math/rng/device/detail/gaussian_impl.hpp +++ b/include/oneapi/math/rng/device/detail/gaussian_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_GAUSSIAN_IMPL_HPP_ -#define _MKL_RNG_DEVICE_GAUSSIAN_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_GAUSSIAN_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_GAUSSIAN_IMPL_HPP_ #include "vm_wrappers.hpp" @@ -267,4 +267,4 @@ class distribution_base> } // namespace oneapi::math::rng::device::detail -#endif // _MKL_RNG_DEVICE_LOGNORMAL_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_LOGNORMAL_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/mcg31m1_impl.hpp b/include/oneapi/math/rng/device/detail/mcg31m1_impl.hpp index 96a52c120..8542fec7c 100644 --- a/include/oneapi/math/rng/device/detail/mcg31m1_impl.hpp +++ b/include/oneapi/math/rng/device/detail/mcg31m1_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_MCG31M1_IMPL_HPP_ -#define _MKL_RNG_DEVICE_MCG31M1_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_MCG31M1_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_MCG31M1_IMPL_HPP_ namespace oneapi::math::rng::device { @@ -230,4 +230,4 @@ class engine_base> { } // namespace oneapi::math::rng::device -#endif // _MKL_RNG_DEVICE_MCG31M1_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_MCG31M1_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/mcg59_impl.hpp b/include/oneapi/math/rng/device/detail/mcg59_impl.hpp index 176330e1e..64f73fee4 100644 --- a/include/oneapi/math/rng/device/detail/mcg59_impl.hpp +++ b/include/oneapi/math/rng/device/detail/mcg59_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_MCG59_IMPL_HPP_ -#define _MKL_RNG_DEVICE_MCG59_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_MCG59_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_MCG59_IMPL_HPP_ namespace oneapi::math::rng::device { @@ -272,4 +272,4 @@ class engine_base> { } // namespace detail } // namespace oneapi::math::rng::device -#endif // _MKL_RNG_DEVICE_MCG59_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_MCG59_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp b/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp index 4ca6abdae..6bca38b19 100644 --- a/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp +++ b/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp @@ -22,8 +22,8 @@ // (2011). Parallelisation Techniques for Random Number Generators. // GPU Computing Gems Emerald Edition. 10.1016/B978-0-12-384988-5.00016-4 -#ifndef _MKL_RNG_DEVICE_MRG32K3A_IMPL_HPP_ -#define _MKL_RNG_DEVICE_MRG32K3A_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_MRG32K3A_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_MRG32K3A_IMPL_HPP_ #include "oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp" @@ -381,4 +381,4 @@ class engine_base> { } // namespace detail } // namespace oneapi::math::rng::device -#endif // _MKL_RNG_DEVICE_MRG32K3A_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_MRG32K3A_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp b/include/oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp index 2c69d2949..8a30d7727 100644 --- a/include/oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp +++ b/include/oneapi/math/rng/device/detail/mrg32k3a_skip_ahead_matrix.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_MRG32K3A_SKIP_AHEAD_MATRIX_HPP_ -#define _MKL_RNG_DEVICE_MRG32K3A_SKIP_AHEAD_MATRIX_HPP_ +#ifndef ONEMATH_RNG_DEVICE_MRG32K3A_SKIP_AHEAD_MATRIX_HPP_ +#define ONEMATH_RNG_DEVICE_MRG32K3A_SKIP_AHEAD_MATRIX_HPP_ namespace oneapi::math::rng::device::detail { namespace mrg32k3a_impl { @@ -3665,4 +3665,4 @@ static const std::uint32_t skip_ahead_matrix[2][quantity_of_3x3_matrices][3][3] } // namespace mrg32k3a_impl } // namespace oneapi::math::rng::device::detail -#endif // _MKL_RNG_DEVICE_MRG32K3A_SKIP_AHEAD_MATRIX_HPP_ +#endif // ONEMATH_RNG_DEVICE_MRG32K3A_SKIP_AHEAD_MATRIX_HPP_ diff --git a/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp b/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp index e3b0db9aa..7ec22d816 100644 --- a/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp +++ b/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_PHILOX4X32X10_IMPL_HPP_ -#define _MKL_RNG_DEVICE_PHILOX4X32X10_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_PHILOX4X32X10_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_PHILOX4X32X10_IMPL_HPP_ #include // std::pair @@ -549,4 +549,4 @@ class engine_base> { } // namespace detail } // namespace oneapi::math::rng::device -#endif // _MKL_RNG_DEVICE_PHILOX4X32X10_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_PHILOX4X32X10_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/poisson_impl.hpp b/include/oneapi/math/rng/device/detail/poisson_impl.hpp index bbcba7335..7c977eaee 100644 --- a/include/oneapi/math/rng/device/detail/poisson_impl.hpp +++ b/include/oneapi/math/rng/device/detail/poisson_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_POISSON_IMPL_HPP_ -#define _MKL_RNG_DEVICE_POISSON_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_POISSON_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_POISSON_IMPL_HPP_ #include @@ -352,4 +352,4 @@ class distribution_base> { } // namespace oneapi::math::rng::device::detail -#endif // _MKL_RNG_DEVICE_UNIFORM_BITS_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_UNIFORM_BITS_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/uniform_impl.hpp b/include/oneapi/math/rng/device/detail/uniform_impl.hpp index 8e8c55ad4..d5d1ecc78 100644 --- a/include/oneapi/math/rng/device/detail/uniform_impl.hpp +++ b/include/oneapi/math/rng/device/detail/uniform_impl.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_UNIFORM_IMPL_HPP_ -#define _MKL_RNG_DEVICE_UNIFORM_IMPL_HPP_ +#ifndef ONEMATH_RNG_DEVICE_UNIFORM_IMPL_HPP_ +#define ONEMATH_RNG_DEVICE_UNIFORM_IMPL_HPP_ #include #include @@ -320,4 +320,4 @@ class distribution_base> { } // namespace oneapi::math::rng::device::detail -#endif // _MKL_RNG_DEVICE_UNIFORM_IMPL_HPP_ +#endif // ONEMATH_RNG_DEVICE_UNIFORM_IMPL_HPP_ diff --git a/include/oneapi/math/rng/device/detail/vm_wrappers.hpp b/include/oneapi/math/rng/device/detail/vm_wrappers.hpp index acd730b66..ddb0035bf 100644 --- a/include/oneapi/math/rng/device/detail/vm_wrappers.hpp +++ b/include/oneapi/math/rng/device/detail/vm_wrappers.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_VM_WRAPPERS_HPP_ -#define _MKL_RNG_DEVICE_VM_WRAPPERS_HPP_ +#ifndef ONEMATH_RNG_DEVICE_VM_WRAPPERS_HPP_ +#define ONEMATH_RNG_DEVICE_VM_WRAPPERS_HPP_ #include @@ -72,4 +72,4 @@ static inline DataType exp_wrapper(DataType a) { } } // namespace oneapi::math::rng::device::detail -#endif // _MKL_RNG_DEVICE_VM_WRAPPERS_HPP_ +#endif // ONEMATH_RNG_DEVICE_VM_WRAPPERS_HPP_ diff --git a/include/oneapi/math/rng/device/distributions.hpp b/include/oneapi/math/rng/device/distributions.hpp index 6db4f70a7..6def09d45 100644 --- a/include/oneapi/math/rng/device/distributions.hpp +++ b/include/oneapi/math/rng/device/distributions.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_DISTRIBUTIONS_HPP_ -#define _MKL_RNG_DEVICE_DISTRIBUTIONS_HPP_ +#ifndef ONEMATH_RNG_DEVICE_DISTRIBUTIONS_HPP_ +#define ONEMATH_RNG_DEVICE_DISTRIBUTIONS_HPP_ #include @@ -634,4 +634,4 @@ class bernoulli : detail::distribution_base> { } // namespace oneapi::math::rng::device -#endif // _MKL_RNG_DEVICE_DISTRIBUTIONS_HPP_ +#endif // ONEMATH_RNG_DEVICE_DISTRIBUTIONS_HPP_ diff --git a/include/oneapi/math/rng/device/engines.hpp b/include/oneapi/math/rng/device/engines.hpp index 2bad9b692..c5bddbe94 100644 --- a/include/oneapi/math/rng/device/engines.hpp +++ b/include/oneapi/math/rng/device/engines.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_ENGINES_HPP_ -#define _MKL_RNG_DEVICE_ENGINES_HPP_ +#ifndef ONEMATH_RNG_DEVICE_ENGINES_HPP_ +#define ONEMATH_RNG_DEVICE_ENGINES_HPP_ #include @@ -167,4 +167,4 @@ class mcg59 : detail::engine_base> { } // namespace oneapi::math::rng::device -#endif // _MKL_RNG_DEVICE_ENGINES_HPP_ +#endif // ONEMATH_RNG_DEVICE_ENGINES_HPP_ diff --git a/include/oneapi/math/rng/device/functions.hpp b/include/oneapi/math/rng/device/functions.hpp index d8ed65337..ccd6d5cc8 100644 --- a/include/oneapi/math/rng/device/functions.hpp +++ b/include/oneapi/math/rng/device/functions.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_FUNCTIONS_HPP_ -#define _MKL_RNG_DEVICE_FUNCTIONS_HPP_ +#ifndef ONEMATH_RNG_DEVICE_FUNCTIONS_HPP_ +#define ONEMATH_RNG_DEVICE_FUNCTIONS_HPP_ #include @@ -49,4 +49,4 @@ void skip_ahead(Engine& engine, std::initializer_list num_to_skip } // namespace oneapi::math::rng::device -#endif // _MKL_RNG_DEVICE_FUNCTIONS_HPP_ +#endif // ONEMATH_RNG_DEVICE_FUNCTIONS_HPP_ diff --git a/include/oneapi/math/rng/device/types.hpp b/include/oneapi/math/rng/device/types.hpp index 1fecbc5b5..d2cb7ac37 100644 --- a/include/oneapi/math/rng/device/types.hpp +++ b/include/oneapi/math/rng/device/types.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_RNG_DEVICE_TYPES_HPP_ -#define _MKL_RNG_DEVICE_TYPES_HPP_ +#ifndef ONEMATH_RNG_DEVICE_TYPES_HPP_ +#define ONEMATH_RNG_DEVICE_TYPES_HPP_ namespace oneapi::math::rng::device { @@ -71,4 +71,4 @@ using by_default = marsaglia; } // namespace oneapi::math::rng::device -#endif // _MKL_RNG_DEVICE_TYPES_HPP_ +#endif // ONEMATH_RNG_DEVICE_TYPES_HPP_ diff --git a/src/blas/backends/cublas/cublas_task.hpp b/src/blas/backends/cublas/cublas_task.hpp index e8702fa62..58d5306ff 100644 --- a/src/blas/backends/cublas/cublas_task.hpp +++ b/src/blas/backends/cublas/cublas_task.hpp @@ -19,8 +19,8 @@ * **************************************************************************/ -#ifndef _MKL_BLAS_CUBLAS_TASK_HPP_ -#define _MKL_BLAS_CUBLAS_TASK_HPP_ +#ifndef ONEMATH_BLAS_CUBLAS_TASK_HPP_ +#define ONEMATH_BLAS_CUBLAS_TASK_HPP_ #include #include #include @@ -86,4 +86,4 @@ static inline void onemath_cublas_host_task(H &cgh, sycl::queue queue, F f) { } // namespace blas } // namespace math } // namespace oneapi -#endif // _MKL_BLAS_CUBLAS_TASK_HPP_ +#endif // ONEMATH_BLAS_CUBLAS_TASK_HPP_ diff --git a/src/lapack/backends/cusolver/cusolver_task.hpp b/src/lapack/backends/cusolver/cusolver_task.hpp index da008897e..497b29052 100644 --- a/src/lapack/backends/cusolver/cusolver_task.hpp +++ b/src/lapack/backends/cusolver/cusolver_task.hpp @@ -17,8 +17,8 @@ * **************************************************************************/ -#ifndef _MKL_LAPACK_CUSOLVER_TASK_HPP_ -#define _MKL_LAPACK_CUSOLVER_TASK_HPP_ +#ifndef ONEMATH_LAPACK_CUSOLVER_TASK_HPP_ +#define ONEMATH_LAPACK_CUSOLVER_TASK_HPP_ #include #include #include @@ -69,4 +69,4 @@ static inline void onemath_cusolver_host_task(H &cgh, sycl::queue queue, F f) { } // namespace lapack } // namespace math } // namespace oneapi -#endif // _MKL_LAPACK_CUSOLVER_TASK_HPP_ +#endif // ONEMATH_LAPACK_CUSOLVER_TASK_HPP_ diff --git a/src/lapack/backends/rocsolver/rocsolver_task.hpp b/src/lapack/backends/rocsolver/rocsolver_task.hpp index 75014c130..061e34504 100644 --- a/src/lapack/backends/rocsolver/rocsolver_task.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_task.hpp @@ -19,8 +19,8 @@ * **************************************************************************/ -#ifndef _MKL_LAPACK_ROCSOLVER_TASK_HPP_ -#define _MKL_LAPACK_ROCSOLVER_TASK_HPP_ +#ifndef ONEMATH_LAPACK_ROCSOLVER_TASK_HPP_ +#define ONEMATH_LAPACK_ROCSOLVER_TASK_HPP_ #include #include #include @@ -71,4 +71,4 @@ static inline void onemath_rocsolver_host_task(H &cgh, sycl::queue queue, F f) { } // namespace lapack } // namespace math } // namespace oneapi -#endif // _MKL_LAPACK_ROCSOLVER_TASK_HPP_ +#endif // ONEMATH_LAPACK_ROCSOLVER_TASK_HPP_ diff --git a/src/rng/backends/curand/curand_helper.hpp b/src/rng/backends/curand/curand_helper.hpp index 203f527c7..41f085862 100644 --- a/src/rng/backends/curand/curand_helper.hpp +++ b/src/rng/backends/curand/curand_helper.hpp @@ -60,8 +60,8 @@ * @file curand_helper.cpp : contains the implementation of all the routines * for CUDA backend */ -#ifndef _MKL_RNG_CURAND_HELPER_HPP_ -#define _MKL_RNG_CURAND_HELPER_HPP_ +#ifndef ONEMATH_RNG_CURAND_HELPER_HPP_ +#define ONEMATH_RNG_CURAND_HELPER_HPP_ #include #include @@ -323,4 +323,4 @@ static inline sycl::event sample_bernoulli_from_uniform(sycl::queue& queue, floa } // namespace math } // namespace oneapi -#endif // _MKL_RNG_CURAND_HELPER_HPP_ +#endif // ONEMATH_RNG_CURAND_HELPER_HPP_ diff --git a/src/rng/backends/curand/curand_task.hpp b/src/rng/backends/curand/curand_task.hpp index 58ce982de..6f9fa5613 100644 --- a/src/rng/backends/curand/curand_task.hpp +++ b/src/rng/backends/curand/curand_task.hpp @@ -1,5 +1,5 @@ -#ifndef _MKL_RNG_CURAND_TASK_HPP_ -#define _MKL_RNG_CURAND_TASK_HPP_ +#ifndef ONEMATH_RNG_CURAND_TASK_HPP_ +#define ONEMATH_RNG_CURAND_TASK_HPP_ #if __has_include() #include diff --git a/src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp b/src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp index d18b275fa..35923601a 100755 --- a/src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp +++ b/src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp @@ -17,8 +17,8 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _MKL_INTERNAL_RNG_GPU_HPP_ -#define _MKL_INTERNAL_RNG_GPU_HPP_ +#ifndef ONEMATH_INTERNAL_RNG_GPU_HPP_ +#define ONEMATH_INTERNAL_RNG_GPU_HPP_ #if __has_include() #include diff --git a/src/rng/backends/rocrand/rocrand_helper.hpp b/src/rng/backends/rocrand/rocrand_helper.hpp index a64da3236..594e8e122 100644 --- a/src/rng/backends/rocrand/rocrand_helper.hpp +++ b/src/rng/backends/rocrand/rocrand_helper.hpp @@ -62,8 +62,8 @@ * @file rocrand_helper.cpp : contains the implementation of all the routines * for HIP backend */ -#ifndef _MKL_RNG_ROCRAND_HELPER_HPP_ -#define _MKL_RNG_ROCRAND_HELPER_HPP_ +#ifndef ONEMATH_RNG_ROCRAND_HELPER_HPP_ +#define ONEMATH_RNG_ROCRAND_HELPER_HPP_ #include #include @@ -332,4 +332,4 @@ class rocm_error : virtual public std::runtime_error { } // namespace math } // namespace oneapi -#endif // _MKL_RNG_ROCRAND_HELPER_HPP_ +#endif // ONEMATH_RNG_ROCRAND_HELPER_HPP_ diff --git a/src/rng/backends/rocrand/rocrand_task.hpp b/src/rng/backends/rocrand/rocrand_task.hpp index 723721d7e..149c0a70d 100644 --- a/src/rng/backends/rocrand/rocrand_task.hpp +++ b/src/rng/backends/rocrand/rocrand_task.hpp @@ -1,5 +1,5 @@ -#ifndef _MKL_RNG_ROCRAND_TASK_HPP_ -#define _MKL_RNG_ROCRAND_TASK_HPP_ +#ifndef ONEMATH_RNG_ROCRAND_TASK_HPP_ +#define ONEMATH_RNG_ROCRAND_TASK_HPP_ #if __has_include() #include From 38b300a0da654864856508660e396cffec3ba94d Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 11 Oct 2024 11:28:51 +0200 Subject: [PATCH 18/41] Add deprecated CMake targets --- src/CMakeLists.txt | 16 ++++++++++++++++ src/blas/CMakeLists.txt | 1 + src/blas/backends/cublas/CMakeLists.txt | 1 + src/blas/backends/mklcpu/CMakeLists.txt | 1 + src/blas/backends/mklgpu/CMakeLists.txt | 1 + src/blas/backends/netlib/CMakeLists.txt | 1 + src/blas/backends/portblas/CMakeLists.txt | 1 + src/blas/backends/rocblas/CMakeLists.txt | 1 + src/dft/CMakeLists.txt | 1 + src/dft/backends/cufft/CMakeLists.txt | 1 + src/dft/backends/mklcpu/CMakeLists.txt | 1 + src/dft/backends/mklgpu/CMakeLists.txt | 1 + src/dft/backends/portfft/CMakeLists.txt | 1 + src/dft/backends/rocfft/CMakeLists.txt | 1 + src/lapack/CMakeLists.txt | 1 + src/lapack/backends/cusolver/CMakeLists.txt | 1 + src/lapack/backends/mklcpu/CMakeLists.txt | 1 + src/lapack/backends/mklgpu/CMakeLists.txt | 1 + src/lapack/backends/rocsolver/CMakeLists.txt | 1 + src/rng/CMakeLists.txt | 1 + src/rng/backends/curand/CMakeLists.txt | 1 + src/rng/backends/mklcpu/CMakeLists.txt | 1 + src/rng/backends/mklgpu/CMakeLists.txt | 1 + src/rng/backends/rocrand/CMakeLists.txt | 1 + src/sparse_blas/CMakeLists.txt | 1 + src/sparse_blas/backends/mklcpu/CMakeLists.txt | 1 + src/sparse_blas/backends/mklgpu/CMakeLists.txt | 1 + 27 files changed, 42 insertions(+) diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index 1e4cf917d..dfc28b703 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -33,6 +33,20 @@ set(ONEMATH_INCLUDE_DIRS $ ) +# Create a CMake library as a deprecated alias of target_name +# The deprecated target name is based on the target_name with "onemath" replaced to "onemkl" +# Deprecated targets are planned to be removed late 2025 +function(add_deprecated_library target_name) + string(REPLACE "onemath" "onemkl" deprecated_name "${target_name}") + if("${target_name}" EQUAL "${deprecated_name}") + message(FATAL_ERROR "Internal error: add_deprecated_library was not able to generate a deprecated target name") + endif() + # The INTERFACE IMPORTED target works like an alias which can have different properties + add_library(deprecated_name INTERFACE IMPORTED) + target_link_libraries(deprecated_name INTERFACE target_name) + set_target_properties(deprecated_name PROPERTIES DEPRECATION "${deprecated_name} target is deprecated, please use ${target_name} instead") +endfunction() + # Build loader and backends for each domain add_custom_target(onemath_backend_libs) foreach(domain ${TARGET_DOMAINS}) @@ -112,4 +126,6 @@ if(BUILD_SHARED_LIBS) ARCHIVE DESTINATION lib LIBRARY DESTINATION lib ) + + add_deprecated_library(onemkl onemath) endif() diff --git a/src/blas/CMakeLists.txt b/src/blas/CMakeLists.txt index fe0499f1e..2d0d796fa 100644 --- a/src/blas/CMakeLists.txt +++ b/src/blas/CMakeLists.txt @@ -23,6 +23,7 @@ add_subdirectory(backends) # Recipe for BLAS loader object if(BUILD_SHARED_LIBS) add_library(onemath_blas OBJECT) +add_deprecated_library(onemath_blas) target_sources(onemath_blas PRIVATE blas_loader.cpp) target_include_directories(onemath_blas PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/blas/backends/cublas/CMakeLists.txt b/src/blas/backends/cublas/CMakeLists.txt index 2eb20fd87..1cda3352f 100644 --- a/src/blas/backends/cublas/CMakeLists.txt +++ b/src/blas/backends/cublas/CMakeLists.txt @@ -29,6 +29,7 @@ set(SOURCES cublas_level1.cpp $<$:cublas_scope_handle_hipsycl.cpp > $<$: cublas_wrappers.cpp>) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) diff --git a/src/blas/backends/mklcpu/CMakeLists.txt b/src/blas/backends/mklcpu/CMakeLists.txt index 3db5a2fb3..7b211d508 100644 --- a/src/blas/backends/mklcpu/CMakeLists.txt +++ b/src/blas/backends/mklcpu/CMakeLists.txt @@ -24,6 +24,7 @@ set(SOURCES mklcpu_level1.cpp mklcpu_level2.cpp mklcpu_level3.cpp mklcpu_batch.cpp mklcpu_extensions.cpp $<$: mklcpu_wrappers.cpp>) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) diff --git a/src/blas/backends/mklgpu/CMakeLists.txt b/src/blas/backends/mklgpu/CMakeLists.txt index 6d69df3dc..84237366c 100644 --- a/src/blas/backends/mklgpu/CMakeLists.txt +++ b/src/blas/backends/mklgpu/CMakeLists.txt @@ -21,6 +21,7 @@ set(LIB_NAME onemath_blas_mklgpu) set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT mklgpu_level1.cpp mklgpu_level2.cpp mklgpu_level3.cpp mklgpu_batch.cpp mklgpu_extensions.cpp $<$: mklgpu_wrappers.cpp> diff --git a/src/blas/backends/netlib/CMakeLists.txt b/src/blas/backends/netlib/CMakeLists.txt index f6dadaac1..73b7f6ce5 100644 --- a/src/blas/backends/netlib/CMakeLists.txt +++ b/src/blas/backends/netlib/CMakeLists.txt @@ -28,6 +28,7 @@ set(SOURCES netlib_common.hpp $<$: netlib_wrappers.cpp> ) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) diff --git a/src/blas/backends/portblas/CMakeLists.txt b/src/blas/backends/portblas/CMakeLists.txt index 1aa6c965b..f2c38f9b3 100644 --- a/src/blas/backends/portblas/CMakeLists.txt +++ b/src/blas/backends/portblas/CMakeLists.txt @@ -176,6 +176,7 @@ set(SOURCES portblas_batch.cpp $<$: portblas_wrappers.cpp>) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) diff --git a/src/blas/backends/rocblas/CMakeLists.txt b/src/blas/backends/rocblas/CMakeLists.txt index 59427458b..c50e5349c 100644 --- a/src/blas/backends/rocblas/CMakeLists.txt +++ b/src/blas/backends/rocblas/CMakeLists.txt @@ -34,6 +34,7 @@ set(SOURCES rocblas_level1.cpp $<$:rocblas_scope_handle_hipsycl.cpp > $<$: rocblas_wrappers.cpp>) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) diff --git a/src/dft/CMakeLists.txt b/src/dft/CMakeLists.txt index 1b41cef72..40987cf38 100644 --- a/src/dft/CMakeLists.txt +++ b/src/dft/CMakeLists.txt @@ -23,6 +23,7 @@ add_subdirectory(backends) # Recipe for DFT loader object if(BUILD_SHARED_LIBS) add_library(onemath_dft OBJECT) +add_deprecated_library(onemath_dft) target_sources(onemath_dft PRIVATE backends/descriptor.cpp dft_loader.cpp) target_include_directories(onemath_dft PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/dft/backends/cufft/CMakeLists.txt b/src/dft/backends/cufft/CMakeLists.txt index dbbe92316..dfb28099f 100644 --- a/src/dft/backends/cufft/CMakeLists.txt +++ b/src/dft/backends/cufft/CMakeLists.txt @@ -22,6 +22,7 @@ set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT descriptor.cpp commit.cpp diff --git a/src/dft/backends/mklcpu/CMakeLists.txt b/src/dft/backends/mklcpu/CMakeLists.txt index b6eb2a674..e79308743 100644 --- a/src/dft/backends/mklcpu/CMakeLists.txt +++ b/src/dft/backends/mklcpu/CMakeLists.txt @@ -23,6 +23,7 @@ set(LIB_OBJ ${LIB_NAME}_obj) include(WarningsUtils) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT commit.cpp descriptor.cpp diff --git a/src/dft/backends/mklgpu/CMakeLists.txt b/src/dft/backends/mklgpu/CMakeLists.txt index 3617290a1..2d557ffa2 100644 --- a/src/dft/backends/mklgpu/CMakeLists.txt +++ b/src/dft/backends/mklgpu/CMakeLists.txt @@ -23,6 +23,7 @@ set(LIB_OBJ ${LIB_NAME}_obj) include(WarningsUtils) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT descriptor.cpp commit.cpp diff --git a/src/dft/backends/portfft/CMakeLists.txt b/src/dft/backends/portfft/CMakeLists.txt index 0c3d59b49..72ec70e24 100644 --- a/src/dft/backends/portfft/CMakeLists.txt +++ b/src/dft/backends/portfft/CMakeLists.txt @@ -62,6 +62,7 @@ set(LIB_NAME onemath_dft_portfft) set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT descriptor.cpp commit.cpp diff --git a/src/dft/backends/rocfft/CMakeLists.txt b/src/dft/backends/rocfft/CMakeLists.txt index 6e0bf6034..b234f70e3 100644 --- a/src/dft/backends/rocfft/CMakeLists.txt +++ b/src/dft/backends/rocfft/CMakeLists.txt @@ -22,6 +22,7 @@ set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT descriptor.cpp commit.cpp diff --git a/src/lapack/CMakeLists.txt b/src/lapack/CMakeLists.txt index 0bb134db6..17ee4cca1 100644 --- a/src/lapack/CMakeLists.txt +++ b/src/lapack/CMakeLists.txt @@ -23,6 +23,7 @@ add_subdirectory(backends) # Recipe for LAPACK loader object if(BUILD_SHARED_LIBS) add_library(onemath_lapack OBJECT) +add_deprecated_library(onemath_lapack) target_sources(onemath_lapack PRIVATE lapack_loader.cpp) target_include_directories(onemath_lapack PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/lapack/backends/cusolver/CMakeLists.txt b/src/lapack/backends/cusolver/CMakeLists.txt index 7340e6f7e..b49e9e1ff 100644 --- a/src/lapack/backends/cusolver/CMakeLists.txt +++ b/src/lapack/backends/cusolver/CMakeLists.txt @@ -26,6 +26,7 @@ set(SOURCES cusolver_lapack.cpp $<$:cusolver_scope_handle.cpp > $<$: cusolver_wrappers.cpp>) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_lapack ${LIB_NAME}) diff --git a/src/lapack/backends/mklcpu/CMakeLists.txt b/src/lapack/backends/mklcpu/CMakeLists.txt index 0821098cc..7099a68fb 100644 --- a/src/lapack/backends/mklcpu/CMakeLists.txt +++ b/src/lapack/backends/mklcpu/CMakeLists.txt @@ -21,6 +21,7 @@ set(LIB_NAME onemath_lapack_mklcpu) set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT mkl_lapack.cpp $<$: lapack_cpu_wrappers.cpp> diff --git a/src/lapack/backends/mklgpu/CMakeLists.txt b/src/lapack/backends/mklgpu/CMakeLists.txt index 24801ebc8..49bf4ac75 100644 --- a/src/lapack/backends/mklgpu/CMakeLists.txt +++ b/src/lapack/backends/mklgpu/CMakeLists.txt @@ -21,6 +21,7 @@ set(LIB_NAME onemath_lapack_mklgpu) set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT mkl_lapack.cpp $<$: lapack_gpu_wrappers.cpp> diff --git a/src/lapack/backends/rocsolver/CMakeLists.txt b/src/lapack/backends/rocsolver/CMakeLists.txt index 8488cc6ad..fe63454c1 100644 --- a/src/lapack/backends/rocsolver/CMakeLists.txt +++ b/src/lapack/backends/rocsolver/CMakeLists.txt @@ -30,6 +30,7 @@ set(SOURCES rocsolver_lapack.cpp $<$:rocsolver_scope_handle.cpp> $<$: rocsolver_wrappers.cpp>) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_lapack ${LIB_NAME}) diff --git a/src/rng/CMakeLists.txt b/src/rng/CMakeLists.txt index bc5161b1f..fc36a67c4 100644 --- a/src/rng/CMakeLists.txt +++ b/src/rng/CMakeLists.txt @@ -23,6 +23,7 @@ add_subdirectory(backends) # Recipe for RNG loader object if(BUILD_SHARED_LIBS) add_library(onemath_rng OBJECT) +add_deprecated_library(onemath_rng) target_sources(onemath_rng PRIVATE rng_loader.cpp) target_include_directories(onemath_rng PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/rng/backends/curand/CMakeLists.txt b/src/rng/backends/curand/CMakeLists.txt index f8415c4a2..818fa8637 100644 --- a/src/rng/backends/curand/CMakeLists.txt +++ b/src/rng/backends/curand/CMakeLists.txt @@ -65,6 +65,7 @@ set(SOURCES philox4x32x10.cpp $<$: curand_wrappers.cpp>) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) diff --git a/src/rng/backends/mklcpu/CMakeLists.txt b/src/rng/backends/mklcpu/CMakeLists.txt index 49aff884b..297206b9f 100644 --- a/src/rng/backends/mklcpu/CMakeLists.txt +++ b/src/rng/backends/mklcpu/CMakeLists.txt @@ -27,6 +27,7 @@ set(SOURCES cpu_common.hpp ) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) target_include_directories(${LIB_OBJ} diff --git a/src/rng/backends/mklgpu/CMakeLists.txt b/src/rng/backends/mklgpu/CMakeLists.txt index 92572d91b..9335f0d7d 100644 --- a/src/rng/backends/mklgpu/CMakeLists.txt +++ b/src/rng/backends/mklgpu/CMakeLists.txt @@ -21,6 +21,7 @@ set(LIB_NAME onemath_rng_mklgpu) set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT mkl_internal_rng_gpu.hpp philox4x32x10.cpp diff --git a/src/rng/backends/rocrand/CMakeLists.txt b/src/rng/backends/rocrand/CMakeLists.txt index 49de4054b..5958a2b10 100644 --- a/src/rng/backends/rocrand/CMakeLists.txt +++ b/src/rng/backends/rocrand/CMakeLists.txt @@ -62,6 +62,7 @@ set(SOURCES philox4x32x10.cpp mrg32k3a.cpp $<$: rocrand_wrappers.cpp>) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) diff --git a/src/sparse_blas/CMakeLists.txt b/src/sparse_blas/CMakeLists.txt index 3b1d3d1af..a2f99a7d3 100644 --- a/src/sparse_blas/CMakeLists.txt +++ b/src/sparse_blas/CMakeLists.txt @@ -21,6 +21,7 @@ add_subdirectory(backends) if(BUILD_SHARED_LIBS) add_library(onemath_sparse_blas OBJECT) + add_deprecated_library(onemath_sparse_blas) target_sources(onemath_sparse_blas PRIVATE sparse_blas_loader.cpp) target_include_directories(onemath_sparse_blas PRIVATE ${PROJECT_SOURCE_DIR}/include diff --git a/src/sparse_blas/backends/mklcpu/CMakeLists.txt b/src/sparse_blas/backends/mklcpu/CMakeLists.txt index 111d3f7ca..71d056f7c 100644 --- a/src/sparse_blas/backends/mklcpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklcpu/CMakeLists.txt @@ -23,6 +23,7 @@ set(LIB_OBJ ${LIB_NAME}_obj) include(WarningsUtils) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT mklcpu_handles.cpp mklcpu_operations.cpp diff --git a/src/sparse_blas/backends/mklgpu/CMakeLists.txt b/src/sparse_blas/backends/mklgpu/CMakeLists.txt index 4f4a9f82d..0ea8f6b93 100644 --- a/src/sparse_blas/backends/mklgpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklgpu/CMakeLists.txt @@ -23,6 +23,7 @@ set(LIB_OBJ ${LIB_NAME}_obj) include(WarningsUtils) add_library(${LIB_NAME}) +add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT mklgpu_handles.cpp mklgpu_operations.cpp From 228d214a5bf3c2c7ced8593d27abbb8797fb07a7 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 11 Oct 2024 12:02:18 +0200 Subject: [PATCH 19/41] Add deprecated headers and mkl namespace --- include/oneapi/mkl.hpp | 36 ++++++++++++++++++++++++++ include/oneapi/mkl/blas.hpp | 30 +++++++++++++++++++++ include/oneapi/mkl/dft.hpp | 30 +++++++++++++++++++++ include/oneapi/mkl/lapack.hpp | 30 +++++++++++++++++++++ include/oneapi/mkl/namespace_alias.hpp | 25 ++++++++++++++++++ include/oneapi/mkl/rng.hpp | 30 +++++++++++++++++++++ include/oneapi/mkl/rng/device.hpp | 30 +++++++++++++++++++++ include/oneapi/mkl/sparse_blas.hpp | 30 +++++++++++++++++++++ 8 files changed, 241 insertions(+) create mode 100644 include/oneapi/mkl.hpp create mode 100644 include/oneapi/mkl/blas.hpp create mode 100644 include/oneapi/mkl/dft.hpp create mode 100644 include/oneapi/mkl/lapack.hpp create mode 100644 include/oneapi/mkl/namespace_alias.hpp create mode 100644 include/oneapi/mkl/rng.hpp create mode 100644 include/oneapi/mkl/rng/device.hpp create mode 100644 include/oneapi/mkl/sparse_blas.hpp diff --git a/include/oneapi/mkl.hpp b/include/oneapi/mkl.hpp new file mode 100644 index 000000000..2eca1b82c --- /dev/null +++ b/include/oneapi/mkl.hpp @@ -0,0 +1,36 @@ +/******************************************************************************* +* Copyright 2020-2021 Intel Corporation +* +* Licensed under the Apache License, Version 2.0 (the "License"); +* you may not use this file except in compliance with the License. +* You may obtain a copy of the License at +* +* http://www.apache.org/licenses/LICENSE-2.0 +* +* Unless required by applicable law or agreed to in writing, +* software distributed under the License is distributed on an "AS IS" BASIS, +* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +* See the License for the specific language governing permissions +* and limitations under the License. +* +* +* SPDX-License-Identifier: Apache-2.0 +*******************************************************************************/ + +#ifndef ONEMATH_MKL_HPP +#define ONEMATH_MKL_HPP + +// Deprecated header is planned to be removed late 2025. +#pragma message("Header `oneapi/mkl.hpp` is deprecated, please use `oneapi/math.hpp` instead") + +#include "oneapi/math/types.hpp" + +#include "oneapi/math/blas.hpp" +#include "oneapi/math/dft.hpp" +#include "oneapi/math/lapack.hpp" +#include "oneapi/math/rng.hpp" +#include "oneapi/math/sparse_blas.hpp" + +#include "namespace_alias.hpp" + +#endif // ONEMATH_MKL_HPP diff --git a/include/oneapi/mkl/blas.hpp b/include/oneapi/mkl/blas.hpp new file mode 100644 index 000000000..a8befa71e --- /dev/null +++ b/include/oneapi/mkl/blas.hpp @@ -0,0 +1,30 @@ +/******************************************************************************* +* Copyright 2024 Intel Corporation +* +* Licensed under the Apache License, Version 2.0 (the "License"); +* you may not use this file except in compliance with the License. +* You may obtain a copy of the License at +* +* http://www.apache.org/licenses/LICENSE-2.0 +* +* Unless required by applicable law or agreed to in writing, +* software distributed under the License is distributed on an "AS IS" BASIS, +* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +* See the License for the specific language governing permissions +* and limitations under the License. +* +* +* SPDX-License-Identifier: Apache-2.0 +*******************************************************************************/ + +#ifndef ONEMATH_MKL_BLAS_HPP +#define ONEMATH_MKL_BLAS_HPP + +// Deprecated header is planned to be removed late 2025. +#pragma message("Header `oneapi/mkl/blas.hpp` is deprecated, please use `oneapi/math/blas.hpp` instead") + +#include "oneapi/math/blas.hpp" + +#include "namespace_alias.hpp" + +#endif // ONEMATH_MKL_BLAS_HPP diff --git a/include/oneapi/mkl/dft.hpp b/include/oneapi/mkl/dft.hpp new file mode 100644 index 000000000..09ec86cc7 --- /dev/null +++ b/include/oneapi/mkl/dft.hpp @@ -0,0 +1,30 @@ +/******************************************************************************* +* Copyright 2024 Intel Corporation +* +* Licensed under the Apache License, Version 2.0 (the "License"); +* you may not use this file except in compliance with the License. +* You may obtain a copy of the License at +* +* http://www.apache.org/licenses/LICENSE-2.0 +* +* Unless required by applicable law or agreed to in writing, +* software distributed under the License is distributed on an "AS IS" BASIS, +* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +* See the License for the specific language governing permissions +* and limitations under the License. +* +* +* SPDX-License-Identifier: Apache-2.0 +*******************************************************************************/ + +#ifndef ONEMATH_MKL_DFT_HPP +#define ONEMATH_MKL_DFT_HPP + +// Deprecated header is planned to be removed late 2025. +#pragma message("Header `oneapi/mkl/dft.hpp` is deprecated, please use `oneapi/math/dft.hpp` instead") + +#include "oneapi/math/dft.hpp" + +#include "namespace_alias.hpp" + +#endif // ONEMATH_MKL_DFT_HPP diff --git a/include/oneapi/mkl/lapack.hpp b/include/oneapi/mkl/lapack.hpp new file mode 100644 index 000000000..990511249 --- /dev/null +++ b/include/oneapi/mkl/lapack.hpp @@ -0,0 +1,30 @@ +/******************************************************************************* +* Copyright 2024 Intel Corporation +* +* Licensed under the Apache License, Version 2.0 (the "License"); +* you may not use this file except in compliance with the License. +* You may obtain a copy of the License at +* +* http://www.apache.org/licenses/LICENSE-2.0 +* +* Unless required by applicable law or agreed to in writing, +* software distributed under the License is distributed on an "AS IS" BASIS, +* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +* See the License for the specific language governing permissions +* and limitations under the License. +* +* +* SPDX-License-Identifier: Apache-2.0 +*******************************************************************************/ + +#ifndef ONEMATH_MKL_LAPACK_HPP +#define ONEMATH_MKL_LAPACK_HPP + +// Deprecated header is planned to be removed late 2025. +#pragma message("Header `oneapi/mkl/lapack.hpp` is deprecated, please use `oneapi/math/lapack.hpp` instead") + +#include "oneapi/math/lapack.hpp" + +#include "namespace_alias.hpp" + +#endif // ONEMATH_MKL_LAPACK_HPP diff --git a/include/oneapi/mkl/namespace_alias.hpp b/include/oneapi/mkl/namespace_alias.hpp new file mode 100644 index 000000000..fa52bad2c --- /dev/null +++ b/include/oneapi/mkl/namespace_alias.hpp @@ -0,0 +1,25 @@ +/*************************************************************************** +* Copyright (C) Codeplay Software Limited +* Licensed under the Apache License, Version 2.0 (the "License"); +* you may not use this file except in compliance with the License. +* You may obtain a copy of the License at +* +* http://www.apache.org/licenses/LICENSE-2.0 +* +* For your convenience, a copy of the License has been included in this +* repository. +* +* Unless required by applicable law or agreed to in writing, software +* distributed under the License is distributed on an "AS IS" BASIS, +* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +* See the License for the specific language governing permissions and +* limitations under the License. +* +**************************************************************************/ + +namespace oneapi { + // Deprecated namespace is planned to be removed late 2025. + namespace [[deprecated("Namespace `oneapi::mkl` is deprecated, please use `oneapi::math` instead")]] mkl { + using namespace math; + } +} diff --git a/include/oneapi/mkl/rng.hpp b/include/oneapi/mkl/rng.hpp new file mode 100644 index 000000000..094a446a1 --- /dev/null +++ b/include/oneapi/mkl/rng.hpp @@ -0,0 +1,30 @@ +/******************************************************************************* +* Copyright 2024 Intel Corporation +* +* Licensed under the Apache License, Version 2.0 (the "License"); +* you may not use this file except in compliance with the License. +* You may obtain a copy of the License at +* +* http://www.apache.org/licenses/LICENSE-2.0 +* +* Unless required by applicable law or agreed to in writing, +* software distributed under the License is distributed on an "AS IS" BASIS, +* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +* See the License for the specific language governing permissions +* and limitations under the License. +* +* +* SPDX-License-Identifier: Apache-2.0 +*******************************************************************************/ + +#ifndef ONEMATH_MKL_RNG_HPP +#define ONEMATH_MKL_RNG_HPP + +// Deprecated header is planned to be removed late 2025. +#pragma message("Header `oneapi/mkl/rng.hpp` is deprecated, please use `oneapi/math/rng.hpp` instead") + +#include "oneapi/math/rng.hpp" + +#include "namespace_alias.hpp" + +#endif // ONEMATH_MKL_RNG_HPP diff --git a/include/oneapi/mkl/rng/device.hpp b/include/oneapi/mkl/rng/device.hpp new file mode 100644 index 000000000..012649884 --- /dev/null +++ b/include/oneapi/mkl/rng/device.hpp @@ -0,0 +1,30 @@ +/******************************************************************************* +* Copyright 2024 Intel Corporation +* +* Licensed under the Apache License, Version 2.0 (the "License"); +* you may not use this file except in compliance with the License. +* You may obtain a copy of the License at +* +* http://www.apache.org/licenses/LICENSE-2.0 +* +* Unless required by applicable law or agreed to in writing, +* software distributed under the License is distributed on an "AS IS" BASIS, +* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +* See the License for the specific language governing permissions +* and limitations under the License. +* +* +* SPDX-License-Identifier: Apache-2.0 +*******************************************************************************/ + +#ifndef ONEMATH_MKL_RNG_DEVICE_HPP +#define ONEMATH_MKL_RNG_DEVICE_HPP + +// Deprecated header is planned to be removed late 2025. +#pragma message("Header `oneapi/mkl/rng/device.hpp` is deprecated, please use `oneapi/math/rng/math.hpp` instead") + +#include "oneapi/math/rng/device.hpp" + +#include "namespace_alias.hpp" + +#endif // ONEMATH_MKL_RNG_DEVICE_HPP diff --git a/include/oneapi/mkl/sparse_blas.hpp b/include/oneapi/mkl/sparse_blas.hpp new file mode 100644 index 000000000..805f55a0a --- /dev/null +++ b/include/oneapi/mkl/sparse_blas.hpp @@ -0,0 +1,30 @@ +/******************************************************************************* +* Copyright 2024 Intel Corporation +* +* Licensed under the Apache License, Version 2.0 (the "License"); +* you may not use this file except in compliance with the License. +* You may obtain a copy of the License at +* +* http://www.apache.org/licenses/LICENSE-2.0 +* +* Unless required by applicable law or agreed to in writing, +* software distributed under the License is distributed on an "AS IS" BASIS, +* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +* See the License for the specific language governing permissions +* and limitations under the License. +* +* +* SPDX-License-Identifier: Apache-2.0 +*******************************************************************************/ + +#ifndef ONEMATH_MKL_SPARSE_BLAS_HPP +#define ONEMATH_MKL_SPARSE_BLAS_HPP + +// Deprecated header is planned to be removed late 2025. +#pragma message("Header `oneapi/mkl/sparse_blas.hpp` is deprecated, please use `oneapi/math/sparse_blas.hpp` instead") + +#include "oneapi/math/sparse_blas.hpp" + +#include "namespace_alias.hpp" + +#endif // ONEMATH_MKL_SPARSE_BLAS_HPP From 8a733adfd1cec8c980e64a8746c4be936f4bb08e Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 11 Oct 2024 12:32:50 +0100 Subject: [PATCH 20/41] Rename oneMathConfig.cmake --- cmake/{oneMKLConfig.cmake => oneMathConfig.cmake} | 1 - 1 file changed, 1 deletion(-) rename cmake/{oneMKLConfig.cmake => oneMathConfig.cmake} (97%) diff --git a/cmake/oneMKLConfig.cmake b/cmake/oneMathConfig.cmake similarity index 97% rename from cmake/oneMKLConfig.cmake rename to cmake/oneMathConfig.cmake index b274c611a..bb261da47 100644 --- a/cmake/oneMKLConfig.cmake +++ b/cmake/oneMathConfig.cmake @@ -20,7 +20,6 @@ list(APPEND CMAKE_MODULE_PATH ${CMAKE_CURRENT_LIST_DIR}) include(CMakeFindDependencyMacro) -#find_dependency(MKL REQUIRED) # try to search for SYCLConfig first to find compiler. If it's not present, use local FindCompiler.cmake find_package(SYCL QUIET) if(NOT ${SYCL_FOUND}) From d44ff49371bb3d76ba7b2aa3912ee4f73af4719f Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 11 Oct 2024 12:45:39 +0100 Subject: [PATCH 21/41] Add include to Intel oneMKL lapack headers --- src/lapack/backends/mklcpu/mkl_lapack.cpp | 3 +++ src/lapack/backends/mklgpu/mkl_lapack.cpp | 3 +++ 2 files changed, 6 insertions(+) diff --git a/src/lapack/backends/mklcpu/mkl_lapack.cpp b/src/lapack/backends/mklcpu/mkl_lapack.cpp index 246722d15..5f76d3aac 100644 --- a/src/lapack/backends/mklcpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklcpu/mkl_lapack.cpp @@ -23,6 +23,9 @@ #include #endif +// Intel(R) oneMKL header +#include + #include "oneapi/math/types.hpp" #include "oneapi/math/lapack/types.hpp" #include "oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp" diff --git a/src/lapack/backends/mklgpu/mkl_lapack.cpp b/src/lapack/backends/mklgpu/mkl_lapack.cpp index 4650ccdd5..059fa9900 100644 --- a/src/lapack/backends/mklgpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklgpu/mkl_lapack.cpp @@ -23,6 +23,9 @@ #include #endif +// Intel(R) oneMKL header +#include + #include "oneapi/math/types.hpp" #include "oneapi/math/lapack/types.hpp" #include "oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp" From e85f0b31ee1c3034fa73003c348b378c35cd2372 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 11 Oct 2024 12:47:27 +0100 Subject: [PATCH 22/41] Fix spblas MKL include --- src/sparse_blas/backends/mkl_common/mkl_handles.hpp | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp index e7b2a44e0..4e99b85de 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp @@ -20,10 +20,8 @@ #ifndef _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HANDLES_HPP_ #define _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HANDLES_HPP_ -// MKLCPU and MKLGPU backends include -// This include defines its own oneapi::math::sparse namespace with some of the -// types that are used here: matrix_handle_t, index_base, transpose, uplo, diag. -#include +// Intel(R) oneMKL header +#include #include "sparse_blas/generic_container.hpp" From a7a8970c3ce738541791ecc743ab0d1dd691a9c9 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Mon, 14 Oct 2024 10:27:33 +0100 Subject: [PATCH 23/41] Fix compilation with MKL backends --- src/blas/backends/mkl_common/mkl_batch.cxx | 202 +-- .../backends/mkl_common/mkl_blas_backend.hpp | 6 +- .../backends/mkl_common/mkl_extensions.cxx | 80 +- src/blas/backends/mkl_common/mkl_level2.cxx | 240 ++-- src/blas/backends/mkl_common/mkl_level3.cxx | 132 +- src/blas/backends/mklcpu/CMakeLists.txt | 11 +- src/blas/backends/mklcpu/mklcpu_batch.cpp | 4 +- .../backends/mklcpu/mklcpu_extensions.cpp | 4 +- src/blas/backends/mklcpu/mklcpu_level1.cpp | 4 +- src/blas/backends/mklcpu/mklcpu_level2.cpp | 4 +- src/blas/backends/mklcpu/mklcpu_level3.cpp | 4 +- src/blas/backends/mklgpu/CMakeLists.txt | 11 +- src/blas/backends/mklgpu/mklgpu_batch.cpp | 4 +- .../backends/mklgpu/mklgpu_extensions.cpp | 4 +- src/blas/backends/mklgpu/mklgpu_level1.cpp | 4 +- src/blas/backends/mklgpu/mklgpu_level2.cpp | 4 +- src/blas/backends/mklgpu/mklgpu_level3.cpp | 4 +- src/dft/backends/mklgpu/backward.cpp | 6 +- src/dft/backends/mklgpu/commit.cpp | 15 +- src/dft/backends/mklgpu/forward.cpp | 6 +- src/dft/backends/mklgpu/mklgpu_helpers.hpp | 28 +- src/include/common_mkl_types_conversion.hpp | 72 ++ src/lapack/backends/mkl_common/mkl_lapack.cxx | 1090 ++++++++--------- .../mkl_common/mkl_lapack_backend.hpp | 1 + src/lapack/backends/mklcpu/CMakeLists.txt | 12 +- src/lapack/backends/mklgpu/CMakeLists.txt | 12 +- src/rng/backends/mklgpu/mrg32k3a.cpp | 2 +- src/rng/backends/mklgpu/philox4x32x10.cpp | 2 +- .../backends/mkl_common/mkl_handles.cxx | 16 +- .../backends/mkl_common/mkl_handles.hpp | 4 +- .../backends/mkl_common/mkl_helper.hpp | 3 + .../backends/mkl_common/mkl_spmm.cxx | 8 +- .../backends/mkl_common/mkl_spmv.cxx | 29 +- .../backends/mkl_common/mkl_spsv.cxx | 17 +- .../backends/mklcpu/CMakeLists.txt | 1 + .../backends/mklgpu/CMakeLists.txt | 1 + src/sparse_blas/generic_container.hpp | 1 + 37 files changed, 1091 insertions(+), 957 deletions(-) create mode 100644 src/include/common_mkl_types_conversion.hpp diff --git a/src/blas/backends/mkl_common/mkl_batch.cxx b/src/blas/backends/mkl_common/mkl_batch.cxx index 6358a3922..07bb70cde 100644 --- a/src/blas/backends/mkl_common/mkl_batch.cxx +++ b/src/blas/backends/mkl_common/mkl_batch.cxx @@ -73,7 +73,7 @@ void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, floa sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, int64_t incx, int64_t stride_x, float beta, sycl::buffer &y, int64_t incy, int64_t stride_y, int64_t batch_size) { - blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, + blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, incy, stride_y, batch_size); } @@ -81,7 +81,7 @@ void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, doub sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, int64_t incx, int64_t stride_x, double beta, sycl::buffer &y, int64_t incy, int64_t stride_y, int64_t batch_size) { - blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, + blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, incy, stride_y, batch_size); } @@ -90,7 +90,7 @@ void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, int64_t stride_a, sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, std::complex beta, sycl::buffer, 1> &y, int64_t incy, int64_t stride_y, int64_t batch_size) { - blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, + blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, incy, stride_y, batch_size); } @@ -100,7 +100,7 @@ void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, int64_t stride_x, std::complex beta, sycl::buffer, 1> &y, int64_t incy, int64_t stride_y, int64_t batch_size) { - blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, + blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, incy, stride_y, batch_size); } @@ -108,7 +108,7 @@ void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, int64_t incx, int64_t stride_x, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::dgmm_batch(queue, left_right, m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, + blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, stride_c, batch_size); } @@ -116,7 +116,7 @@ void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, int64_t incx, int64_t stride_x, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::dgmm_batch(queue, left_right, m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, + blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, stride_c, batch_size); } @@ -125,7 +125,7 @@ void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::dgmm_batch(queue, left_right, m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, + blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, stride_c, batch_size); } @@ -134,7 +134,7 @@ void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::dgmm_batch(queue, left_right, m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, + blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, stride_c, batch_size); } @@ -142,7 +142,7 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t k, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -150,7 +150,7 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t k, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -159,7 +159,7 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -169,7 +169,7 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t ldb, int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -178,7 +178,7 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, sycl::half beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -187,7 +187,7 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -205,7 +205,7 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } @@ -213,7 +213,7 @@ void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, + blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -221,7 +221,7 @@ void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, + blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -230,7 +230,7 @@ void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, + blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -239,14 +239,14 @@ void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, + blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, + blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -254,7 +254,7 @@ void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, + blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -262,7 +262,7 @@ void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n std::complex alpha, sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, + blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } @@ -271,21 +271,21 @@ void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, + blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -293,7 +293,7 @@ void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } @@ -301,39 +301,39 @@ void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size); } void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size); } void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size); } void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size); } void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, float beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::omatadd_batch(queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, + blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -341,7 +341,7 @@ void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64 double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, double beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::omatadd_batch(queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, + blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -351,7 +351,7 @@ void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64 sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::omatadd_batch(queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, + blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -361,7 +361,7 @@ void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64 sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::omatadd_batch(queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, + blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } @@ -489,7 +489,7 @@ sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t const float *a, int64_t lda, int64_t stride_a, const float *x, int64_t incx, int64_t stride_x, float beta, float *y, int64_t incy, int64_t stride_y, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, stride_a, x, incx, stride_x, + return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, incy, stride_y, batch_size, dependencies); } @@ -498,7 +498,7 @@ sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t int64_t incx, int64_t stride_x, double beta, double *y, int64_t incy, int64_t stride_y, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, stride_a, x, incx, stride_x, + return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, incy, stride_y, batch_size, dependencies); } @@ -508,7 +508,7 @@ sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t int64_t stride_x, std::complex beta, std::complex *y, int64_t incy, int64_t stride_y, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, stride_a, x, incx, stride_x, + return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, incy, stride_y, batch_size, dependencies); } @@ -518,7 +518,7 @@ sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t int64_t stride_x, std::complex beta, std::complex *y, int64_t incy, int64_t stride_y, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, stride_a, x, incx, stride_x, + return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, incy, stride_y, batch_size, dependencies); } @@ -526,7 +526,7 @@ sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_ const float **a, int64_t *lda, const float **x, int64_t *incx, float *beta, float **y, int64_t *incy, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, groupsize, dependencies); } @@ -534,7 +534,7 @@ sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_ const double **a, int64_t *lda, const double **x, int64_t *incx, double *beta, double **y, int64_t *incy, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, groupsize, dependencies); } @@ -543,7 +543,7 @@ sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_ const std::complex **x, int64_t *incx, std::complex *beta, std::complex **y, int64_t *incy, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, groupsize, dependencies); } @@ -552,7 +552,7 @@ sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_ const std::complex **x, int64_t *incx, std::complex *beta, std::complex **y, int64_t *incy, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, transa, m, n, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, groupsize, dependencies); } @@ -560,7 +560,7 @@ sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n int64_t lda, int64_t stride_a, const float *x, int64_t incx, int64_t stride_x, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, left_right, m, n, a, lda, stride_a, x, incx, stride_x, c, + return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, stride_c, batch_size, dependencies); } @@ -568,7 +568,7 @@ sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n int64_t lda, int64_t stride_a, const double *x, int64_t incx, int64_t stride_x, double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, left_right, m, n, a, lda, stride_a, x, incx, stride_x, c, + return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, stride_c, batch_size, dependencies); } @@ -577,7 +577,7 @@ sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n const std::complex *x, int64_t incx, int64_t stride_x, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, left_right, m, n, a, lda, stride_a, x, incx, stride_x, c, + return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, stride_c, batch_size, dependencies); } @@ -586,7 +586,7 @@ sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n const std::complex *x, int64_t incx, int64_t stride_x, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, left_right, m, n, a, lda, stride_a, x, incx, stride_x, c, + return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, stride_c, batch_size, dependencies); } @@ -594,7 +594,7 @@ sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t const float **a, int64_t *lda, const float **x, int64_t *incx, float **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, + return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, groupsize, dependencies); } @@ -602,7 +602,7 @@ sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t const double **a, int64_t *lda, const double **x, int64_t *incx, double **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, + return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, groupsize, dependencies); } @@ -610,7 +610,7 @@ sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t const std::complex **a, int64_t *lda, const std::complex **x, int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, + return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, groupsize, dependencies); } @@ -618,7 +618,7 @@ sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t const std::complex **a, int64_t *lda, const std::complex **x, int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, + return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, groupsize, dependencies); } @@ -627,7 +627,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i const float *b, int64_t ldb, int64_t stride_b, float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } @@ -636,7 +636,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i const double *b, int64_t ldb, int64_t stride_b, double beta, double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } @@ -646,7 +646,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i int64_t stride_b, std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } @@ -656,7 +656,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i int64_t stride_b, std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } @@ -665,7 +665,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i int64_t stride_a, const sycl::half *b, int64_t ldb, int64_t stride_b, sycl::half beta, sycl::half *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } @@ -674,7 +674,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i const sycl::half *b, int64_t ldb, int64_t stride_b, float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } @@ -692,7 +692,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i const std::int8_t *b, int64_t ldb, int64_t stride_b, float beta, std::int32_t *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } @@ -701,7 +701,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, const float **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } @@ -710,7 +710,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, const double **b, int64_t *ldb, double *beta, double **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } @@ -720,7 +720,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *ldb, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } @@ -730,7 +730,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *ldb, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } @@ -739,7 +739,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *lda, const sycl::half **b, int64_t *ldb, sycl::half *beta, sycl::half **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, groupsize, dependencies); } @@ -748,7 +748,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, const sycl::half **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, groupsize, dependencies); } @@ -766,7 +766,7 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, const std::int8_t **b, int64_t *ldb, float *beta, std::int32_t **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, + return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, groupsize, dependencies); } @@ -774,7 +774,7 @@ sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, tr diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, + return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } @@ -782,7 +782,7 @@ sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, tr diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, + return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } @@ -791,7 +791,7 @@ sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, tr const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, + return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } @@ -800,7 +800,7 @@ sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, tr const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, + return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } @@ -808,7 +808,7 @@ sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, diag *unit_diag, int64_t *m, int64_t *n, float *alpha, const float **a, int64_t *lda, float **b, int64_t *ldb, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, + return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } @@ -816,7 +816,7 @@ sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, diag *unit_diag, int64_t *m, int64_t *n, double *alpha, const double **a, int64_t *lda, double **b, int64_t *ldb, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, + return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } @@ -825,7 +825,7 @@ sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, const std::complex **a, int64_t *lda, std::complex **b, int64_t *ldb, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, + return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } @@ -834,7 +834,7 @@ sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, const std::complex **a, int64_t *lda, std::complex **b, int64_t *ldb, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, + return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } @@ -842,7 +842,7 @@ sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, in float alpha, const float *a, int64_t lda, int64_t stride_a, float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, + return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } @@ -850,7 +850,7 @@ sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, in double alpha, const double *a, int64_t lda, int64_t stride_a, double beta, double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, + return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } @@ -859,7 +859,7 @@ sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, in int64_t stride_a, std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, + return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } @@ -868,7 +868,7 @@ sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, in int64_t stride_a, std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, + return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } @@ -876,7 +876,7 @@ sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *k, float *alpha, const float **a, int64_t *lda, float *beta, float **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, + return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, group_count, groupsize, dependencies); } @@ -884,7 +884,7 @@ sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *k, double *alpha, const double **a, int64_t *lda, double *beta, double **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, + return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, group_count, groupsize, dependencies); } @@ -893,7 +893,7 @@ sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *lda, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, + return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, group_count, groupsize, dependencies); } @@ -902,7 +902,7 @@ sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *lda, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, + return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, group_count, groupsize, dependencies); } @@ -910,7 +910,7 @@ sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } @@ -918,7 +918,7 @@ sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } @@ -926,7 +926,7 @@ sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 std::complex alpha, const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } @@ -934,21 +934,21 @@ sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 std::complex alpha, const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, float *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, const std::vector &dependencies) { - return blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, + return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, double *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, const std::vector &dependencies) { - return blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, + return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } @@ -956,7 +956,7 @@ sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, const std::vector &dependencies) { - return blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, + return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } @@ -964,7 +964,7 @@ sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, const std::vector &dependencies) { - return blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, + return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } @@ -973,7 +973,7 @@ sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb float beta, const float *b, int64_t ldb, int64_t stride_b, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatadd_batch(queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, + return blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } @@ -982,7 +982,7 @@ sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb double beta, const double *b, int64_t ldb, int64_t stride_b, double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatadd_batch(queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, + return blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } @@ -992,7 +992,7 @@ sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb const std::complex *b, int64_t ldb, int64_t stride_b, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatadd_batch(queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, + return blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } @@ -1002,7 +1002,7 @@ sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb const std::complex *b, int64_t ldb, int64_t stride_b, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatadd_batch(queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, + return blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } @@ -1010,7 +1010,7 @@ sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int float* alpha, const float** a, int64_t* lda, float** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, b, ldb, group_count, + return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } @@ -1018,7 +1018,7 @@ sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int double* alpha, const double** a, int64_t* lda, double** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, b, ldb, group_count, + return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } @@ -1026,7 +1026,7 @@ sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int std::complex* alpha, const std::complex** a, int64_t* lda, std::complex** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, b, ldb, group_count, + return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } @@ -1035,7 +1035,7 @@ sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int int64_t* lda, std::complex** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::omatcopy_batch(queue, trans, m, n, alpha, a, lda, b, ldb, group_count, + return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } @@ -1043,7 +1043,7 @@ sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int float* alpha, float** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, group_count, + return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } @@ -1051,7 +1051,7 @@ sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int double* alpha, double** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, group_count, + return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } @@ -1059,7 +1059,7 @@ sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int std::complex* alpha, std::complex** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, group_count, + return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } @@ -1067,6 +1067,6 @@ sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int std::complex* alpha, std::complex** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::imatcopy_batch(queue, trans, m, n, alpha, ab, lda, ldb, group_count, + return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } diff --git a/src/blas/backends/mkl_common/mkl_blas_backend.hpp b/src/blas/backends/mkl_common/mkl_blas_backend.hpp index e753b4b68..2234772db 100644 --- a/src/blas/backends/mkl_common/mkl_blas_backend.hpp +++ b/src/blas/backends/mkl_common/mkl_blas_backend.hpp @@ -21,7 +21,11 @@ #include -#include "mkl_version.h" +// Intel(R) oneMKL headers +#include +#include + +#include "common_mkl_types_conversion.hpp" #include "oneapi/math/types.hpp" namespace oneapi { diff --git a/src/blas/backends/mkl_common/mkl_extensions.cxx b/src/blas/backends/mkl_common/mkl_extensions.cxx index 4672af5c7..faba97517 100644 --- a/src/blas/backends/mkl_common/mkl_extensions.cxx +++ b/src/blas/backends/mkl_common/mkl_extensions.cxx @@ -23,7 +23,7 @@ void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset of int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, int8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - blas_major::gemm_bias(queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, + blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -31,7 +31,7 @@ void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset of int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, int8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - blas_major::gemm_bias(queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, + blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -39,7 +39,7 @@ void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset of int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, uint8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - blas_major::gemm_bias(queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, + blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -47,7 +47,7 @@ void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset of int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, uint8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - blas_major::gemm_bias(queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, + blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } @@ -55,7 +55,7 @@ void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose tra int64_t k, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { - blas_major::gemmt(queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, + blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -63,7 +63,7 @@ void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose tra int64_t k, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, int64_t ldc) { - blas_major::gemmt(queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, + blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -71,7 +71,7 @@ void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose tra int64_t k, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - blas_major::gemmt(queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, + blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc); } @@ -79,30 +79,30 @@ void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose tra int64_t k, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - blas_major::gemmt(queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, + blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { - blas_major::omatcopy(queue, trans, m, n, alpha, a, lda, b, ldb); + blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb); } void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { - blas_major::omatcopy(queue, trans, m, n, alpha, a, lda, b, ldb); + blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb); } void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb) { - blas_major::omatcopy(queue, trans, m, n, alpha, a, lda, b, ldb); + blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb); } void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb) { - blas_major::omatcopy(queue, trans, m, n, alpha, a, lda, b, ldb); + blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb); } void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, @@ -132,48 +132,48 @@ void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, sycl::buffer &ab, int64_t lda, int64_t ldb) { - blas_major::imatcopy(queue, trans, m, n, alpha, ab, lda, ldb); + blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb); } void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, sycl::buffer &ab, int64_t lda, int64_t ldb) { - blas_major::imatcopy(queue, trans, m, n, alpha, ab, lda, ldb); + blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb); } void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { - blas_major::imatcopy(queue, trans, m, n, alpha, ab, lda, ldb); + blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb); } void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { - blas_major::imatcopy(queue, trans, m, n, alpha, ab, lda, ldb); + blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb); } void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, float beta, sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { - blas_major::omatadd(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); + blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc); } void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, double beta, sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { - blas_major::omatadd(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); + blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc); } void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, std::complex beta, sycl::buffer, 1> &b, int64_t ldb, sycl::buffer, 1> &c, int64_t ldc) { - blas_major::omatadd(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); + blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc); } void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, std::complex beta, sycl::buffer, 1> &b, int64_t ldb, sycl::buffer, 1> &c, int64_t ldc) { - blas_major::omatadd(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); + blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc); } // USM APIs @@ -183,7 +183,7 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, of int8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, const std::vector &dependencies) { - return blas_major::gemm_bias(queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, + return blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } @@ -192,7 +192,7 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, of int8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, const std::vector &dependencies) { - return blas_major::gemm_bias(queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, + return blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } @@ -201,7 +201,7 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, of uint8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, const std::vector &dependencies) { - return blas_major::gemm_bias(queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, + return blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } @@ -210,7 +210,7 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, of uint8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, const std::vector &dependencies) { - return blas_major::gemm_bias(queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, + return blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } @@ -218,7 +218,7 @@ sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transp int64_t n, int64_t k, float alpha, const float *a, int64_t lda, const float *b, int64_t ldb, float beta, float *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::gemmt(queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, + return blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -226,7 +226,7 @@ sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transp int64_t n, int64_t k, double alpha, const double *a, int64_t lda, const double *b, int64_t ldb, double beta, double *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::gemmt(queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, + return blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -235,7 +235,7 @@ sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transp int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::gemmt(queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, + return blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -244,34 +244,34 @@ sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transp int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::gemmt(queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, + return blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, float *b, int64_t ldb, const std::vector &dependencies) { - return blas_major::omatcopy(queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); + return blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies); } sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, double *b, int64_t ldb, const std::vector &dependencies) { - return blas_major::omatcopy(queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); + return blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies); } sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, const std::vector &dependencies) { - return blas_major::omatcopy(queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); + return blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies); } sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, const std::vector &dependencies) { - return blas_major::omatcopy(queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); + return blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies); } sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, @@ -303,32 +303,32 @@ sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, float *ab, int64_t lda, int64_t ldb, const std::vector &dependencies) { - return blas_major::imatcopy(queue, trans, m, n, alpha, ab, lda, ldb, dependencies); + return blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies); } sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, double *ab, int64_t lda, int64_t ldb, const std::vector &dependencies) { - return blas_major::imatcopy(queue, trans, m, n, alpha, ab, lda, ldb, dependencies); + return blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies); } sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, const std::vector &dependencies) { - return blas_major::imatcopy(queue, trans, m, n, alpha, ab, lda, ldb, dependencies); + return blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies); } sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, const std::vector &dependencies) { - return blas_major::imatcopy(queue, trans, m, n, alpha, ab, lda, ldb, dependencies); + return blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies); } sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, float beta, const float *b, int64_t ldb, float *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::omatadd(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, + return blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } @@ -336,7 +336,7 @@ sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int6 double alpha, const double *a, int64_t lda, double beta, const double *b, int64_t ldb, double *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::omatadd(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, + return blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } @@ -345,7 +345,7 @@ sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int6 std::complex beta, const std::complex *b, int64_t ldb, std::complex *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::omatadd(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, + return blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } @@ -354,6 +354,6 @@ sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int6 std::complex beta, const std::complex *b, int64_t ldb, std::complex *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::omatadd(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, + return blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } diff --git a/src/blas/backends/mkl_common/mkl_level2.cxx b/src/blas/backends/mkl_common/mkl_level2.cxx index 83494be12..68f3b3713 100644 --- a/src/blas/backends/mkl_common/mkl_level2.cxx +++ b/src/blas/backends/mkl_common/mkl_level2.cxx @@ -22,55 +22,55 @@ void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - blas_major::gemv(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); + blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy); } void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - blas_major::gemv(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); + blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy); } void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::gemv(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); + blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy); } void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::gemv(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); + blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy); } void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - blas_major::gbmv(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); + blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - blas_major::gbmv(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); + blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::gbmv(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); + blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::gbmv(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); + blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, @@ -117,314 +117,314 @@ void hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, std::co sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hbmv(queue, uplo, n, k, alpha, a, lda, x, incx, beta, y, incy); + blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy); } void hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hbmv(queue, uplo, n, k, alpha, a, lda, x, incx, beta, y, incy); + blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy); } void hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hemv(queue, uplo, n, alpha, a, lda, x, incx, beta, y, incy); + blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy); } void hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hemv(queue, uplo, n, alpha, a, lda, x, incx, beta, y, incy); + blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy); } void her(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::her(queue, uplo, n, alpha, x, incx, a, lda); + blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda); } void her(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::her(queue, uplo, n, alpha, x, incx, a, lda); + blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda); } void her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::her2(queue, uplo, n, alpha, x, incx, y, incy, a, lda); + blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda); } void her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::her2(queue, uplo, n, alpha, x, incx, y, incy, a, lda); + blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda); } void hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hpmv(queue, uplo, n, alpha, a, x, incx, beta, y, incy); + blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy); } void hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hpmv(queue, uplo, n, alpha, a, x, incx, beta, y, incy); + blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy); } void hpr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - blas_major::hpr(queue, uplo, n, alpha, x, incx, a); + blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a); } void hpr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - blas_major::hpr(queue, uplo, n, alpha, x, incx, a); + blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a); } void hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - blas_major::hpr2(queue, uplo, n, alpha, x, incx, y, incy, a); + blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a); } void hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - blas_major::hpr2(queue, uplo, n, alpha, x, incx, y, incy, a); + blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a); } void sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - blas_major::sbmv(queue, uplo, n, k, alpha, a, lda, x, incx, beta, y, incy); + blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy); } void sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - blas_major::sbmv(queue, uplo, n, k, alpha, a, lda, x, incx, beta, y, incy); + blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy); } void spmv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - blas_major::spmv(queue, uplo, n, alpha, a, x, incx, beta, y, incy); + blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy); } void spmv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - blas_major::spmv(queue, uplo, n, alpha, a, x, incx, beta, y, incy); + blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy); } void spr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - blas_major::spr(queue, uplo, n, alpha, x, incx, a); + blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a); } void spr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - blas_major::spr(queue, uplo, n, alpha, x, incx, a); + blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a); } void spr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - blas_major::spr2(queue, uplo, n, alpha, x, incx, y, incy, a); + blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a); } void spr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - blas_major::spr2(queue, uplo, n, alpha, x, incx, y, incy, a); + blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a); } void symv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - blas_major::symv(queue, uplo, n, alpha, a, lda, x, incx, beta, y, incy); + blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy); } void symv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - blas_major::symv(queue, uplo, n, alpha, a, lda, x, incx, beta, y, incy); + blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy); } void syr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - blas_major::syr(queue, uplo, n, alpha, x, incx, a, lda); + blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda); } void syr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - blas_major::syr(queue, uplo, n, alpha, x, incx, a, lda); + blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda); } void syr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - blas_major::syr2(queue, uplo, n, alpha, x, incx, y, incy, a, lda); + blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda); } void syr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - blas_major::syr2(queue, uplo, n, alpha, x, incx, y, incy, a, lda); + blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda); } void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::tbmv(queue, uplo, trans, diag, n, k, a, lda, x, incx); + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); } void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::tbmv(queue, uplo, trans, diag, n, k, a, lda, x, incx); + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); } void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tbmv(queue, uplo, trans, diag, n, k, a, lda, x, incx); + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); } void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tbmv(queue, uplo, trans, diag, n, k, a, lda, x, incx); + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); } void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::tbsv(queue, uplo, trans, diag, n, k, a, lda, x, incx); + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); } void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::tbsv(queue, uplo, trans, diag, n, k, a, lda, x, incx); + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); } void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tbsv(queue, uplo, trans, diag, n, k, a, lda, x, incx); + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); } void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tbsv(queue, uplo, trans, diag, n, k, a, lda, x, incx); + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); } void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - blas_major::tpmv(queue, uplo, trans, diag, n, a, x, incx); + blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); } void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - blas_major::tpmv(queue, uplo, trans, diag, n, a, x, incx); + blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); } void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tpmv(queue, uplo, trans, diag, n, a, x, incx); + blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); } void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tpmv(queue, uplo, trans, diag, n, a, x, incx); + blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); } void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - blas_major::tpsv(queue, uplo, trans, diag, n, a, x, incx); + blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); } void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - blas_major::tpsv(queue, uplo, trans, diag, n, a, x, incx); + blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); } void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tpsv(queue, uplo, trans, diag, n, a, x, incx); + blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); } void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tpsv(queue, uplo, trans, diag, n, a, x, incx); + blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); } void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::trmv(queue, uplo, trans, diag, n, a, lda, x, incx); + blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); } void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::trmv(queue, uplo, trans, diag, n, a, lda, x, incx); + blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); } void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::trmv(queue, uplo, trans, diag, n, a, lda, x, incx); + blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); } void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::trmv(queue, uplo, trans, diag, n, a, lda, x, incx); + blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); } void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::trsv(queue, uplo, trans, diag, n, a, lda, x, incx); + blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); } void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::trsv(queue, uplo, trans, diag, n, a, lda, x, incx); + blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); } void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::trsv(queue, uplo, trans, diag, n, a, lda, x, incx); + blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); } void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::trsv(queue, uplo, trans, diag, n, a, lda, x, incx); + blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); } // USM APIs @@ -432,14 +432,14 @@ void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_ sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gemv(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gemv(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } @@ -448,7 +448,7 @@ sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64 const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gemv(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } @@ -457,7 +457,7 @@ sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64 const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gemv(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } @@ -465,7 +465,7 @@ sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64 std::int64_t kl, std::int64_t ku, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gbmv(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } @@ -473,7 +473,7 @@ sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64 std::int64_t kl, std::int64_t ku, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gbmv(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } @@ -482,7 +482,7 @@ sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64 const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gbmv(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } @@ -491,7 +491,7 @@ sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64 const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gbmv(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, + return blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } @@ -540,7 +540,7 @@ sycl::event hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hbmv(queue, uplo, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); + return blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } sycl::event hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, @@ -548,315 +548,315 @@ sycl::event hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hbmv(queue, uplo, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); + return blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } sycl::event hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hemv(queue, uplo, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); + return blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } sycl::event hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hemv(queue, uplo, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); + return blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } sycl::event her(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::her(queue, uplo, n, alpha, x, incx, a, lda, dependencies); + return blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies); } sycl::event her(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::her(queue, uplo, n, alpha, x, incx, a, lda, dependencies); + return blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies); } sycl::event her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::her2(queue, uplo, n, alpha, x, incx, y, incy, a, lda, dependencies); + return blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies); } sycl::event her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::her2(queue, uplo, n, alpha, x, incx, y, incy, a, lda, dependencies); + return blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies); } sycl::event hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hpmv(queue, uplo, n, alpha, a, x, incx, beta, y, incy, dependencies); + return blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies); } sycl::event hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hpmv(queue, uplo, n, alpha, a, x, incx, beta, y, incy, dependencies); + return blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies); } sycl::event hpr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - return blas_major::hpr(queue, uplo, n, alpha, x, incx, a, dependencies); + return blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies); } sycl::event hpr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - return blas_major::hpr(queue, uplo, n, alpha, x, incx, a, dependencies); + return blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies); } sycl::event hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - return blas_major::hpr2(queue, uplo, n, alpha, x, incx, y, incy, a, dependencies); + return blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies); } sycl::event hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - return blas_major::hpr2(queue, uplo, n, alpha, x, incx, y, incy, a, dependencies); + return blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies); } sycl::event sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::sbmv(queue, uplo, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); + return blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } sycl::event sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::sbmv(queue, uplo, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); + return blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } sycl::event spmv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::spmv(queue, uplo, n, alpha, a, x, incx, beta, y, incy, dependencies); + return blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies); } sycl::event spmv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::spmv(queue, uplo, n, alpha, a, x, incx, beta, y, incy, dependencies); + return blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies); } sycl::event spr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies) { - return blas_major::spr(queue, uplo, n, alpha, x, incx, a, dependencies); + return blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies); } sycl::event spr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies) { - return blas_major::spr(queue, uplo, n, alpha, x, incx, a, dependencies); + return blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies); } sycl::event spr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies) { - return blas_major::spr2(queue, uplo, n, alpha, x, incx, y, incy, a, dependencies); + return blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies); } sycl::event spr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies) { - return blas_major::spr2(queue, uplo, n, alpha, x, incx, y, incy, a, dependencies); + return blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies); } sycl::event symv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::symv(queue, uplo, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); + return blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } sycl::event symv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::symv(queue, uplo, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); + return blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } sycl::event syr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::syr(queue, uplo, n, alpha, x, incx, a, lda, dependencies); + return blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies); } sycl::event syr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::syr(queue, uplo, n, alpha, x, incx, a, lda, dependencies); + return blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies); } sycl::event syr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::syr2(queue, uplo, n, alpha, x, incx, y, incy, a, lda, dependencies); + return blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies); } sycl::event syr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::syr2(queue, uplo, n, alpha, x, incx, y, incy, a, lda, dependencies); + return blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies); } sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbmv(queue, uplo, trans, diag, n, k, a, lda, x, incx, dependencies); + return blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); } sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbmv(queue, uplo, trans, diag, n, k, a, lda, x, incx, dependencies); + return blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); } sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbmv(queue, uplo, trans, diag, n, k, a, lda, x, incx, dependencies); + return blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); } sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbmv(queue, uplo, trans, diag, n, k, a, lda, x, incx, dependencies); + return blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); } sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbsv(queue, uplo, trans, diag, n, k, a, lda, x, incx, dependencies); + return blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); } sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbsv(queue, uplo, trans, diag, n, k, a, lda, x, incx, dependencies); + return blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); } sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbsv(queue, uplo, trans, diag, n, k, a, lda, x, incx, dependencies); + return blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); } sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbsv(queue, uplo, trans, diag, n, k, a, lda, x, incx, dependencies); + return blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); } sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpmv(queue, uplo, trans, diag, n, a, x, incx, dependencies); + return blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); } sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpmv(queue, uplo, trans, diag, n, a, x, incx, dependencies); + return blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); } sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpmv(queue, uplo, trans, diag, n, a, x, incx, dependencies); + return blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); } sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpmv(queue, uplo, trans, diag, n, a, x, incx, dependencies); + return blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); } sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpsv(queue, uplo, trans, diag, n, a, x, incx, dependencies); + return blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); } sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpsv(queue, uplo, trans, diag, n, a, x, incx, dependencies); + return blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); } sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpsv(queue, uplo, trans, diag, n, a, x, incx, dependencies); + return blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); } sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpsv(queue, uplo, trans, diag, n, a, x, incx, dependencies); + return blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); } sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trmv(queue, uplo, trans, diag, n, a, lda, x, incx, dependencies); + return blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); } sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trmv(queue, uplo, trans, diag, n, a, lda, x, incx, dependencies); + return blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); } sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trmv(queue, uplo, trans, diag, n, a, lda, x, incx, dependencies); + return blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); } sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trmv(queue, uplo, trans, diag, n, a, lda, x, incx, dependencies); + return blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); } sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trsv(queue, uplo, trans, diag, n, a, lda, x, incx, dependencies); + return blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); } sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trsv(queue, uplo, trans, diag, n, a, lda, x, incx, dependencies); + return blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); } sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trsv(queue, uplo, trans, diag, n, a, lda, x, incx, dependencies); + return blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); } sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trsv(queue, uplo, trans, diag, n, a, lda, x, incx, dependencies); + return blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); } diff --git a/src/blas/backends/mkl_common/mkl_level3.cxx b/src/blas/backends/mkl_common/mkl_level3.cxx index d52c710f1..68662f47f 100644 --- a/src/blas/backends/mkl_common/mkl_level3.cxx +++ b/src/blas/backends/mkl_common/mkl_level3.cxx @@ -23,178 +23,178 @@ void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::symm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); } void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::symm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); } void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::symm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); } void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::symm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); } void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::hemm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); } void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::hemm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); } void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::syrk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); + blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); } void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::syrk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); + blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); } void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::syrk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); + blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); } void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::syrk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); + blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); } void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::herk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); + blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); } void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::herk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); + blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); } void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::syr2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::syr2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::syr2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::syr2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::her2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::her2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); } void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - blas_major::trmm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, b, + blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb); } void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - blas_major::trmm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, b, + blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb); } @@ -202,7 +202,7 @@ void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - blas_major::trmm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, b, + blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb); } @@ -210,21 +210,21 @@ void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - blas_major::trmm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, b, + blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb); } void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - blas_major::trsm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, b, + blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb); } void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - blas_major::trsm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, b, + blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb); } @@ -232,7 +232,7 @@ void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - blas_major::trsm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, b, + blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb); } @@ -240,7 +240,7 @@ void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - blas_major::trsm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, b, + blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb); } @@ -250,7 +250,7 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::in std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -258,7 +258,7 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::in std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -267,7 +267,7 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::in const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -276,7 +276,7 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::in const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -284,7 +284,7 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::in std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -292,7 +292,7 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::in std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -300,7 +300,7 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::in std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -308,7 +308,7 @@ sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::symm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, + return blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -316,7 +316,7 @@ sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::symm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, + return blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -325,7 +325,7 @@ sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::symm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, + return blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -334,7 +334,7 @@ sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::symm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, + return blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -343,7 +343,7 @@ sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::hemm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, + return blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -352,21 +352,21 @@ sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::hemm(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, + return blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syrk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, + return blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, dependencies); } sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syrk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, + return blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, dependencies); } @@ -374,7 +374,7 @@ sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syrk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, + return blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, dependencies); } @@ -382,7 +382,7 @@ sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syrk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, + return blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, dependencies); } @@ -390,7 +390,7 @@ sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::herk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, + return blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, dependencies); } @@ -398,7 +398,7 @@ sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::herk(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, + return blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, dependencies); } @@ -406,7 +406,7 @@ sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::in std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syr2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -414,7 +414,7 @@ sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::in std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syr2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -423,7 +423,7 @@ sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::in std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syr2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -432,7 +432,7 @@ sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::in std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syr2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -441,7 +441,7 @@ sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::in std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::her2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -450,7 +450,7 @@ sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::in std::int64_t lda, const std::complex *b, std::int64_t ldb, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::her2k(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + return blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } @@ -458,7 +458,7 @@ sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpos diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trmm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, + return blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, dependencies); } @@ -466,7 +466,7 @@ sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpos diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trmm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, + return blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, dependencies); } @@ -474,7 +474,7 @@ sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpos diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trmm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, + return blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, dependencies); } @@ -482,7 +482,7 @@ sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpos diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trmm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, + return blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, dependencies); } @@ -490,7 +490,7 @@ sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpos diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trsm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, + return blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, dependencies); } @@ -498,7 +498,7 @@ sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpos diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trsm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, + return blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, dependencies); } @@ -506,7 +506,7 @@ sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpos diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trsm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, + return blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, dependencies); } @@ -514,6 +514,6 @@ sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpos diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trsm(queue, left_right, upper_lower, transa, unit_diag, m, n, alpha, a, lda, + return blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb, dependencies); } diff --git a/src/blas/backends/mklcpu/CMakeLists.txt b/src/blas/backends/mklcpu/CMakeLists.txt index 7b211d508..d59b2f316 100644 --- a/src/blas/backends/mklcpu/CMakeLists.txt +++ b/src/blas/backends/mklcpu/CMakeLists.txt @@ -31,9 +31,16 @@ if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET ${LIB_OBJ} SOURCES ${SOURCES}) endif() +# Due to using the same file name for different headers in this library and in +# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ +# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the +# correct header. +target_compile_options(${LIB_OBJ} + BEFORE PRIVATE -iquote $ +) + target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/include - ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} diff --git a/src/blas/backends/mklcpu/mklcpu_batch.cpp b/src/blas/backends/mklcpu/mklcpu_batch.cpp index 541aa80dc..cda964f7d 100644 --- a/src/blas/backends/mklcpu/mklcpu_batch.cpp +++ b/src/blas/backends/mklcpu/mklcpu_batch.cpp @@ -34,13 +34,13 @@ namespace blas { namespace mklcpu { namespace column_major { -namespace blas_major = ::oneapi::math::blas::column_major; +namespace blas_major = ::oneapi::mkl::blas::column_major; #include "../mkl_common/mkl_batch.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::math::blas::row_major; +namespace blas_major = ::oneapi::mkl::blas::row_major; #include "../mkl_common/mkl_batch.cxx" } // namespace row_major diff --git a/src/blas/backends/mklcpu/mklcpu_extensions.cpp b/src/blas/backends/mklcpu/mklcpu_extensions.cpp index 40accb82e..3ddf73c80 100644 --- a/src/blas/backends/mklcpu/mklcpu_extensions.cpp +++ b/src/blas/backends/mklcpu/mklcpu_extensions.cpp @@ -34,13 +34,13 @@ namespace blas { namespace mklcpu { namespace column_major { -namespace blas_major = ::oneapi::math::blas::column_major; +namespace blas_major = ::oneapi::mkl::blas::column_major; #include "../mkl_common/mkl_extensions.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::math::blas::row_major; +namespace blas_major = ::oneapi::mkl::blas::row_major; #include "../mkl_common/mkl_extensions.cxx" } // namespace row_major diff --git a/src/blas/backends/mklcpu/mklcpu_level1.cpp b/src/blas/backends/mklcpu/mklcpu_level1.cpp index 02cd4094e..846cb6068 100644 --- a/src/blas/backends/mklcpu/mklcpu_level1.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level1.cpp @@ -33,13 +33,13 @@ namespace blas { namespace mklcpu { namespace column_major { -namespace blas_major = ::oneapi::math::blas::column_major; +namespace blas_major = ::oneapi::mkl::blas::column_major; #include "../mkl_common/mkl_level1.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::math::blas::row_major; +namespace blas_major = ::oneapi::mkl::blas::row_major; #include "../mkl_common/mkl_level1.cxx" } // namespace row_major diff --git a/src/blas/backends/mklcpu/mklcpu_level2.cpp b/src/blas/backends/mklcpu/mklcpu_level2.cpp index 3eda04ee7..4b62d1bd0 100644 --- a/src/blas/backends/mklcpu/mklcpu_level2.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level2.cpp @@ -33,13 +33,13 @@ namespace blas { namespace mklcpu { namespace column_major { -namespace blas_major = ::oneapi::math::blas::column_major; +namespace blas_major = ::oneapi::mkl::blas::column_major; #include "../mkl_common/mkl_level2.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::math::blas::row_major; +namespace blas_major = ::oneapi::mkl::blas::row_major; #include "../mkl_common/mkl_level2.cxx" } // namespace row_major diff --git a/src/blas/backends/mklcpu/mklcpu_level3.cpp b/src/blas/backends/mklcpu/mklcpu_level3.cpp index 72450596a..4d4e5e915 100644 --- a/src/blas/backends/mklcpu/mklcpu_level3.cpp +++ b/src/blas/backends/mklcpu/mklcpu_level3.cpp @@ -33,13 +33,13 @@ namespace blas { namespace mklcpu { namespace column_major { -namespace blas_major = ::oneapi::math::blas::column_major; +namespace blas_major = ::oneapi::mkl::blas::column_major; #include "../mkl_common/mkl_level3.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::math::blas::row_major; +namespace blas_major = ::oneapi::mkl::blas::row_major; #include "../mkl_common/mkl_level3.cxx" } // namespace row_major diff --git a/src/blas/backends/mklgpu/CMakeLists.txt b/src/blas/backends/mklgpu/CMakeLists.txt index 84237366c..52010a10d 100644 --- a/src/blas/backends/mklgpu/CMakeLists.txt +++ b/src/blas/backends/mklgpu/CMakeLists.txt @@ -28,9 +28,16 @@ add_library(${LIB_OBJ} OBJECT ) add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) +# Due to using the same file name for different headers in this library and in +# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ +# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the +# correct header. +target_compile_options(${LIB_OBJ} + BEFORE PRIVATE -iquote $ +) + target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/include - ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} diff --git a/src/blas/backends/mklgpu/mklgpu_batch.cpp b/src/blas/backends/mklgpu/mklgpu_batch.cpp index b33a2b030..427f1d2f5 100644 --- a/src/blas/backends/mklgpu/mklgpu_batch.cpp +++ b/src/blas/backends/mklgpu/mklgpu_batch.cpp @@ -34,13 +34,13 @@ namespace blas { namespace mklgpu { namespace column_major { -namespace blas_major = ::oneapi::math::blas::column_major; +namespace blas_major = ::oneapi::mkl::blas::column_major; #include "../mkl_common/mkl_batch.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::math::blas::row_major; +namespace blas_major = ::oneapi::mkl::blas::row_major; #include "../mkl_common/mkl_batch.cxx" } // namespace row_major diff --git a/src/blas/backends/mklgpu/mklgpu_extensions.cpp b/src/blas/backends/mklgpu/mklgpu_extensions.cpp index 9ade59f76..709a65b27 100644 --- a/src/blas/backends/mklgpu/mklgpu_extensions.cpp +++ b/src/blas/backends/mklgpu/mklgpu_extensions.cpp @@ -34,13 +34,13 @@ namespace blas { namespace mklgpu { namespace column_major { -namespace blas_major = ::oneapi::math::blas::column_major; +namespace blas_major = ::oneapi::mkl::blas::column_major; #include "../mkl_common/mkl_extensions.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::math::blas::row_major; +namespace blas_major = ::oneapi::mkl::blas::row_major; #include "../mkl_common/mkl_extensions.cxx" } // namespace row_major diff --git a/src/blas/backends/mklgpu/mklgpu_level1.cpp b/src/blas/backends/mklgpu/mklgpu_level1.cpp index ca45cc6de..96e57ef45 100644 --- a/src/blas/backends/mklgpu/mklgpu_level1.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level1.cpp @@ -33,14 +33,14 @@ namespace blas { namespace mklgpu { namespace column_major { -namespace blas_major = ::oneapi::math::blas::column_major; +namespace blas_major = ::oneapi::mkl::blas::column_major; #include "../mkl_common/mkl_level1.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::math::blas::row_major; +namespace blas_major = ::oneapi::mkl::blas::row_major; #include "../mkl_common/mkl_level1.cxx" } // namespace row_major diff --git a/src/blas/backends/mklgpu/mklgpu_level2.cpp b/src/blas/backends/mklgpu/mklgpu_level2.cpp index 4bf1c08c5..383f8e6fb 100644 --- a/src/blas/backends/mklgpu/mklgpu_level2.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level2.cpp @@ -33,13 +33,13 @@ namespace blas { namespace mklgpu { namespace column_major { -namespace blas_major = ::oneapi::math::blas::column_major; +namespace blas_major = ::oneapi::mkl::blas::column_major; #include "../mkl_common/mkl_level2.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::math::blas::row_major; +namespace blas_major = ::oneapi::mkl::blas::row_major; #include "../mkl_common/mkl_level2.cxx" } // namespace row_major diff --git a/src/blas/backends/mklgpu/mklgpu_level3.cpp b/src/blas/backends/mklgpu/mklgpu_level3.cpp index bbb7d10fe..1307e440a 100644 --- a/src/blas/backends/mklgpu/mklgpu_level3.cpp +++ b/src/blas/backends/mklgpu/mklgpu_level3.cpp @@ -33,13 +33,13 @@ namespace blas { namespace mklgpu { namespace column_major { -namespace blas_major = ::oneapi::math::blas::column_major; +namespace blas_major = ::oneapi::mkl::blas::column_major; #include "../mkl_common/mkl_level3.cxx" } // namespace column_major namespace row_major { -namespace blas_major = ::oneapi::math::blas::row_major; +namespace blas_major = ::oneapi::mkl::blas::row_major; #include "../mkl_common/mkl_level3.cxx" } // namespace row_major diff --git a/src/dft/backends/mklgpu/backward.cpp b/src/dft/backends/mklgpu/backward.cpp index ec0e4ca08..0c248003f 100644 --- a/src/dft/backends/mklgpu/backward.cpp +++ b/src/dft/backends/mklgpu/backward.cpp @@ -45,7 +45,7 @@ namespace detail { /// Assumes backend descriptor values match those of the frontend. template inline auto compute_backward(dft::detail::descriptor &desc, ArgTs &&... args) { - using mklgpu_desc_t = dft::descriptor; + using mklgpu_desc_t = oneapi::mkl::dft::descriptor; using desc_shptr_t = std::shared_ptr; using handle_t = std::pair; auto commit_handle = dft::detail::get_commit(desc); @@ -56,7 +56,7 @@ inline auto compute_backward(dft::detail::descriptor &desc, ArgTs &&. auto handle = reinterpret_cast(commit_handle->get_handle()); auto mklgpu_desc = handle->second; // Second because backward DFT. int commit_status{ DFTI_UNCOMMITTED }; - mklgpu_desc->get_value(dft::config_param::COMMIT_STATUS, &commit_status); + mklgpu_desc->get_value(oneapi::mkl::dft::config_param::COMMIT_STATUS, &commit_status); if (commit_status != DFTI_COMMITTED) { throw math::invalid_argument("DFT", "compute_backward", "MKLGPU DFT descriptor was not successfully committed."); @@ -64,7 +64,7 @@ inline auto compute_backward(dft::detail::descriptor &desc, ArgTs &&. // The MKLGPU backend's interface contains fewer function signatures than in this // open-source library. Consequently, it is not required to forward template arguments // to resolve to the correct function. - return dft::compute_backward(*mklgpu_desc, std::forward(args)...); + return oneapi::mkl::dft::compute_backward(*mklgpu_desc, std::forward(args)...); } /// Throw an math::invalid_argument if the runtime param in the descriptor does not match diff --git a/src/dft/backends/mklgpu/commit.cpp b/src/dft/backends/mklgpu/commit.cpp index 4463eafe9..f7e1b2de8 100644 --- a/src/dft/backends/mklgpu/commit.cpp +++ b/src/dft/backends/mklgpu/commit.cpp @@ -44,7 +44,7 @@ #endif // Intel oneMKL 2024.1 deprecates input/output strides. -#include "mkl_version.h" +#include #if INTEL_MKL_VERSION < 20240001 #error MKLGPU requires oneMath 2024.1 or later #endif @@ -63,13 +63,13 @@ template class mklgpu_commit final : public dft::detail::commit_impl { private: // Equivalent MKLGPU precision and domain from oneMath's precision / domain. - static constexpr dft::precision mklgpu_prec = to_mklgpu(prec); - static constexpr dft::domain mklgpu_dom = to_mklgpu(dom); + static constexpr oneapi::mkl::dft::precision mklgpu_prec = to_mklgpu(prec); + static constexpr oneapi::mkl::dft::domain mklgpu_dom = to_mklgpu(dom); // A pair of descriptors are needed because of the [[deprecated]]IN/OUTPUT_STRIDES vs F/BWD_STRIDES API. // Of the pair [0] is fwd DFT, [1] is backward DFT. If possible, the pointers refer to the same descriptor. // Both pointers must be valid. - using mklgpu_descriptor_t = dft::descriptor; + using mklgpu_descriptor_t = oneapi::mkl::dft::descriptor; using descriptor_shptr_t = std::shared_ptr; using handle_t = std::pair; @@ -172,7 +172,7 @@ class mklgpu_commit final : public dft::detail::commit_impl { void set_value(mklgpu_descriptor_t& desc, const dft::detail::dft_values& config, bool assume_fwd_dft, dft::detail::stride_api stride_choice) { using onemath_param = dft::detail::config_param; - using backend_param = dft::config_param; + using backend_param = oneapi::mkl::dft::config_param; // The following are read-only: // Dimension, forward domain, precision, commit status. @@ -236,8 +236,9 @@ class mklgpu_commit final : public dft::detail::commit_impl { // This is called by the workspace_helper, and is not part of the user API. virtual std::int64_t get_workspace_external_bytes_impl() override { std::size_t workspaceSizeFwd = 0, workspaceSizeBwd = 0; - handle.first->get_value(dft::config_param::WORKSPACE_BYTES, &workspaceSizeFwd); - handle.second->get_value(dft::config_param::WORKSPACE_BYTES, &workspaceSizeBwd); + using backend_param = oneapi::mkl::dft::config_param; + handle.first->get_value(backend_param::WORKSPACE_BYTES, &workspaceSizeFwd); + handle.second->get_value(backend_param::WORKSPACE_BYTES, &workspaceSizeBwd); return static_cast(std::max(workspaceSizeFwd, workspaceSizeFwd)); } }; diff --git a/src/dft/backends/mklgpu/forward.cpp b/src/dft/backends/mklgpu/forward.cpp index 56421e110..209dc9295 100644 --- a/src/dft/backends/mklgpu/forward.cpp +++ b/src/dft/backends/mklgpu/forward.cpp @@ -51,7 +51,7 @@ namespace detail { /// Assumes backend descriptor values match those of the frontend. template inline auto compute_forward(dft::detail::descriptor &desc, ArgTs &&... args) { - using mklgpu_desc_t = dft::descriptor; + using mklgpu_desc_t = oneapi::mkl::dft::descriptor; using desc_shptr_t = std::shared_ptr; using handle_t = std::pair; auto commit_handle = dft::detail::get_commit(desc); @@ -62,7 +62,7 @@ inline auto compute_forward(dft::detail::descriptor &desc, ArgTs &&.. auto handle = reinterpret_cast(commit_handle->get_handle()); auto mklgpu_desc = handle->first; // First because forward DFT. int commit_status{ DFTI_UNCOMMITTED }; - mklgpu_desc->get_value(dft::config_param::COMMIT_STATUS, &commit_status); + mklgpu_desc->get_value(oneapi::mkl::dft::config_param::COMMIT_STATUS, &commit_status); if (commit_status != DFTI_COMMITTED) { throw math::invalid_argument("DFT", "compute_forward", "MKLGPU DFT descriptor was not successfully committed."); @@ -70,7 +70,7 @@ inline auto compute_forward(dft::detail::descriptor &desc, ArgTs &&.. // The MKLGPU backend's interface contains fewer function signatures than in this // open-source library. Consequently, it is not required to forward template arguments // to resolve to the correct function. - return dft::compute_forward(*mklgpu_desc, std::forward(args)...); + return oneapi::mkl::dft::compute_forward(*mklgpu_desc, std::forward(args)...); } /// Throw an math::invalid_argument if the runtime param in the descriptor does not match diff --git a/src/dft/backends/mklgpu/mklgpu_helpers.hpp b/src/dft/backends/mklgpu/mklgpu_helpers.hpp index 73c0d37f3..4554ae622 100644 --- a/src/dft/backends/mklgpu/mklgpu_helpers.hpp +++ b/src/dft/backends/mklgpu/mklgpu_helpers.hpp @@ -38,29 +38,29 @@ namespace mklgpu { namespace detail { /// Convert domain to equivalent backend native value. -inline constexpr dft::domain to_mklgpu(dft::detail::domain dom) { +inline constexpr oneapi::mkl::dft::domain to_mklgpu(dft::detail::domain dom) { if (dom == dft::detail::domain::REAL) { - return dft::domain::REAL; + return oneapi::mkl::dft::domain::REAL; } else { - return dft::domain::COMPLEX; + return oneapi::mkl::dft::domain::COMPLEX; } } /// Convert precision to equivalent backend native value. -inline constexpr dft::precision to_mklgpu(dft::detail::precision dom) { +inline constexpr oneapi::mkl::dft::precision to_mklgpu(dft::detail::precision dom) { if (dom == dft::detail::precision::SINGLE) { - return dft::precision::SINGLE; + return oneapi::mkl::dft::precision::SINGLE; } else { - return dft::precision::DOUBLE; + return oneapi::mkl::dft::precision::DOUBLE; } } /// Convert a config_param to equivalent backend native value. -inline constexpr dft::config_param to_mklgpu(dft::detail::config_param param) { +inline constexpr oneapi::mkl::dft::config_param to_mklgpu(dft::detail::config_param param) { using iparam = dft::detail::config_param; - using oparam = dft::config_param; + using oparam = oneapi::mkl::dft::config_param; switch (param) { case iparam::FORWARD_DOMAIN: return oparam::FORWARD_DOMAIN; case iparam::DIMENSION: return oparam::DIMENSION; @@ -151,23 +151,23 @@ inline constexpr int to_mklgpu( * @param value The config value to convert. **/ template -inline constexpr dft::config_value to_mklgpu_config_value(dft::detail::config_value value); +inline constexpr oneapi::mkl::dft::config_value to_mklgpu_config_value(dft::detail::config_value value); template <> -inline constexpr dft::config_value +inline constexpr oneapi::mkl::dft::config_value to_mklgpu_config_value( dft::detail::config_value value) { if (value == dft::detail::config_value::WORKSPACE_AUTOMATIC) { - // NB: dft::config_value != dft::detail::config_value - return dft::config_value::WORKSPACE_INTERNAL; + // NB: oneapi::mkl::dft::config_value != dft::detail::config_value + return oneapi::mkl::dft::config_value::WORKSPACE_INTERNAL; } else if (value == dft::detail::config_value::WORKSPACE_EXTERNAL) { - return dft::config_value::WORKSPACE_EXTERNAL; + return oneapi::mkl::dft::config_value::WORKSPACE_EXTERNAL; } else { throw math::invalid_argument("dft", "MKLGPU descriptor set_value()", "Invalid config value for workspace placement."); - return dft::config_value::WORKSPACE_INTERNAL; + return oneapi::mkl::dft::config_value::WORKSPACE_INTERNAL; } } } // namespace detail diff --git a/src/include/common_mkl_types_conversion.hpp b/src/include/common_mkl_types_conversion.hpp new file mode 100644 index 000000000..f2bc45c98 --- /dev/null +++ b/src/include/common_mkl_types_conversion.hpp @@ -0,0 +1,72 @@ +/******************************************************************************* +* Copyright Codeplay Software Ltd +* +* Licensed under the Apache License, Version 2.0 (the "License"); +* you may not use this file except in compliance with the License. +* You may obtain a copy of the License at +* +* http://www.apache.org/licenses/LICENSE-2.0 +* +* Unless required by applicable law or agreed to in writing, +* software distributed under the License is distributed on an "AS IS" BASIS, +* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +* See the License for the specific language governing permissions +* and limitations under the License. +* +* +* SPDX-License-Identifier: Apache-2.0 +*******************************************************************************/ + +#ifndef _ONEMATH_SRC_INCLUDE_COMMON_MKL_TYPES_CONVERSION_HPP_ +#define _ONEMATH_SRC_INCLUDE_COMMON_MKL_TYPES_CONVERSION_HPP_ + +// The file is used to convert oneMath types to Intel(R) oneMKL types for all the common types shared across domains. +// The file assumes that the common types are identical between the 2 libraries, except for their namespace. + +#include + +#include "oneapi/math/types.hpp" + +namespace oneapi { +namespace math { +namespace detail { + +inline auto get_onemkl_transpose(oneapi::math::transpose* param_ptr) { return reinterpret_cast(param_ptr); } +inline auto get_onemkl_transpose(oneapi::math::transpose param) { return *get_onemkl_transpose(¶m); } + +inline auto get_onemkl_uplo(oneapi::math::uplo* param_ptr) { return reinterpret_cast(param_ptr); } +inline auto get_onemkl_uplo(oneapi::math::uplo param) { return *get_onemkl_uplo(¶m); } + +inline auto get_onemkl_diag(oneapi::math::diag* param_ptr) { return reinterpret_cast(param_ptr); } +inline auto get_onemkl_diag(oneapi::math::diag param) { return *get_onemkl_diag(¶m); } + +inline auto get_onemkl_side(oneapi::math::side* param_ptr) { return reinterpret_cast(param_ptr); } +inline auto get_onemkl_side(oneapi::math::side param) { return *get_onemkl_side(¶m); } + +inline auto get_onemkl_offset(oneapi::math::offset param) { return *reinterpret_cast(¶m); } + +inline auto get_onemkl_layout(oneapi::math::layout param) { return *reinterpret_cast(¶m); } + +inline auto get_onemkl_index_base(oneapi::math::index_base param) { return *reinterpret_cast(¶m); } + +inline auto get_onemkl_job(oneapi::math::job param) { return *reinterpret_cast(¶m); } + +inline auto get_onemkl_jobsvd(oneapi::math::jobsvd param) { return *reinterpret_cast(¶m); } + +inline auto get_onemkl_generate(oneapi::math::generate param) { return *reinterpret_cast(¶m); } + +inline auto get_onemkl_compz(oneapi::math::compz param) { return *reinterpret_cast(¶m); } + +inline auto get_onemkl_direct(oneapi::math::direct param) { return *reinterpret_cast(¶m); } + +inline auto get_onemkl_storev(oneapi::math::storev param) { return *reinterpret_cast(¶m); } + +inline auto get_onemkl_rangev(oneapi::math::rangev param) { return *reinterpret_cast(¶m); } + +inline auto get_onemkl_order(oneapi::math::order param) { return *reinterpret_cast(¶m); } + +} // namespace detail +} // namespace math +} // namespace oneapi + +#endif // _ONEMATH_SRC_INCLUDE_COMMON_MKL_TYPES_CONVERSION_HPP_ diff --git a/src/lapack/backends/mkl_common/mkl_lapack.cxx b/src/lapack/backends/mkl_common/mkl_lapack.cxx index 8573bffd9..47dbc97c0 100644 --- a/src/lapack/backends/mkl_common/mkl_lapack.cxx +++ b/src/lapack/backends/mkl_common/mkl_lapack.cxx @@ -129,121 +129,121 @@ void getri(sycl::queue &queue, std::int64_t n, sycl::buffer sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, +void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gesvd(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, + ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gesvd(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, + ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gesvd(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, + ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, +void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gesvd(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, + ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::heevd(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size); } -void heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::heevd(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size); } -void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hegvd(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, + ::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hegvd(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, + ::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hetrd(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hetrd(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hetrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); } -void hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hetrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); } -void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgbr(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgbr(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, @@ -255,214 +255,214 @@ void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { ::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgtr(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size); } -void orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgtr(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size); } -void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, +void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormtr(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, +void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormtr(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormrq(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormrq(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormqr(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormqr(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf(queue, uplo, n, a, lda, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); } -void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf(queue, uplo, n, a, lda, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); } -void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf(queue, uplo, n, a, lda, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); } -void potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf(queue, uplo, n, a, lda, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); } -void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potri(queue, uplo, n, a, lda, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); } -void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potri(queue, uplo, n, a, lda, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); } -void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potri(queue, uplo, n, a, lda, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); } -void potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potri(queue, uplo, n, a, lda, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); } -void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::syevd(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size); } -void syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::syevd(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size); } -void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sygvd(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, + ::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, +void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sygvd(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, + ::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrd(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrd(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); } -void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); } -void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); } -void sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); } -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::trtrs(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, + ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::trtrs(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, + ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::trtrs(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, + ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::trtrs(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, + ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungbr(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungbr(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size); } void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, @@ -476,66 +476,66 @@ void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t scratchpad_size) { ::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungtr(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size); } -void ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungtr(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); + ::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size); } -void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmrq(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmrq(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmqr(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, +void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmqr(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmtr(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmtr(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, + ::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, @@ -660,132 +660,132 @@ sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, s return ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + return ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + return ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + return ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + return ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gesvd(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, + return ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gesvd(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, + return ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gesvd(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, + return ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(sycl::queue &queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, +sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gesvd(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, + return ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::heevd(queue, jobz, uplo, n, a, lda, w, scratchpad, + return ::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event heevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::heevd(queue, jobz, uplo, n, a, lda, w, scratchpad, + return ::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, +sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hegvd(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, + return ::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, +sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hegvd(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, + return ::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hetrd(queue, uplo, n, a, lda, d, e, tau, scratchpad, + return ::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hetrd(queue, uplo, n, a, lda, d, e, tau, scratchpad, + return ::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hetrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hetrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgbr(queue, vec, m, n, k, a, lda, tau, scratchpad, + return ::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgbr(queue, vec, m, n, k, a, lda, tau, scratchpad, + return ::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, @@ -800,240 +800,240 @@ sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64 return ::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgtr(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgtr(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, float *a, +sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormtr(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, double *a, +sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormtr(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormrq(queue, side, trans, m, n, k, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormrq(queue, side, trans, m, n, k, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormqr(queue, side, trans, m, n, k, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormqr(queue, side, trans, m, n, k, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf(queue, uplo, n, a, lda, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf(queue, uplo, n, a, lda, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf(queue, uplo, n, a, lda, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf(queue, uplo, n, a, lda, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potri(queue, uplo, n, a, lda, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potri(queue, uplo, n, a, lda, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potri(queue, uplo, n, a, lda, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potri(queue, uplo, n, a, lda, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, + return ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, + return ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, + return ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, + return ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::syevd(queue, jobz, uplo, n, a, lda, w, scratchpad, + return ::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event syevd(sycl::queue &queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::syevd(queue, jobz, uplo, n, a, lda, w, scratchpad, + return ::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, +sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sygvd(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, + return ::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, +sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sygvd(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, + return ::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrd(queue, uplo, n, a, lda, d, e, tau, scratchpad, + return ::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrd(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrd(queue, uplo, n, a, lda, d, e, tau, scratchpad, + return ::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrf(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::trtrs(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, + return ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, double *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::trtrs(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, + return ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, float *a, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::trtrs(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, + return ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(sycl::queue &queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, +sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::trtrs(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, + return ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungbr(queue, vec, m, n, k, a, lda, tau, scratchpad, + return ::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungbr(sycl::queue &queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, +sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungbr(queue, vec, m, n, k, a, lda, tau, scratchpad, + return ::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, @@ -1050,66 +1050,66 @@ sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64 return ::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungtr(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungtr(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungtr(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, + return ::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmrq(queue, side, trans, m, n, k, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmrq(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmrq(queue, side, trans, m, n, k, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmqr(queue, side, trans, m, n, k, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmqr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, +sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmqr(queue, side, trans, m, n, k, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmtr(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmtr(sycl::queue &queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, +sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmtr(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, + return ::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, @@ -1170,39 +1170,39 @@ void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, + ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, + ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, + ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, + ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, @@ -1249,63 +1249,63 @@ void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_ ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer &a, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, + ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, + ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, + ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, + ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, + ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, + ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, + ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, +void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, + ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, @@ -1520,83 +1520,83 @@ sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, stride_a, ipiv, + return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, stride_a, ipiv, + return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, stride_a, ipiv, + return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, stride_a, ipiv, + return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, + return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, + return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, + return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, +sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, + return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1632,139 +1632,139 @@ sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, st return ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, float *a, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, stride_a, batch_size, + return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, double *a, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, stride_a, batch_size, + return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, stride_a, batch_size, + return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, stride_a, batch_size, + return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, float **a, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, group_count, group_sizes, + return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, double **a, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, group_count, group_sizes, + return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, group_count, group_sizes, + return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, uplo, n, a, lda, group_count, group_sizes, + return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, + return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, + return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, + return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, + return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, b, ldb, group_count, + return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, b, ldb, group_count, + return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, b, ldb, group_count, + return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, +sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, uplo, n, nrhs, a, lda, b, ldb, group_count, + return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } @@ -1864,36 +1864,36 @@ std::int64_t geqrf_scratchpad_size>(sycl::queue &queue, std return ::oneapi::mkl::lapack::geqrf_scratchpad_size>(queue, m, n, lda); } template <> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, +std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return ::oneapi::mkl::lapack::gesvd_scratchpad_size(queue, jobu, jobvt, m, n, lda, ldu, + return ::oneapi::mkl::lapack::gesvd_scratchpad_size(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, ldvt); } template <> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, +std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return ::oneapi::mkl::lapack::gesvd_scratchpad_size(queue, jobu, jobvt, m, n, lda, ldu, + return ::oneapi::mkl::lapack::gesvd_scratchpad_size(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, ldvt); } template <> std::int64_t gesvd_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return ::oneapi::mkl::lapack::gesvd_scratchpad_size>(queue, jobu, jobvt, m, + return ::oneapi::mkl::lapack::gesvd_scratchpad_size>(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, ldvt); } template <> std::int64_t gesvd_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, + oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return ::oneapi::mkl::lapack::gesvd_scratchpad_size>(queue, jobu, jobvt, m, + return ::oneapi::mkl::lapack::gesvd_scratchpad_size>(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, ldvt); } template <> @@ -1935,104 +1935,104 @@ std::int64_t getri_scratchpad_size>(sycl::queue &queue, std return ::oneapi::mkl::lapack::getri_scratchpad_size>(queue, n, lda); } template <> -std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::getrs_scratchpad_size(queue, trans, n, nrhs, lda, ldb); + return ::oneapi::mkl::lapack::getrs_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb); } template <> -std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::getrs_scratchpad_size(queue, trans, n, nrhs, lda, ldb); + return ::oneapi::mkl::lapack::getrs_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb); } template <> std::int64_t getrs_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::transpose trans, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::getrs_scratchpad_size>(queue, trans, n, nrhs, + return ::oneapi::mkl::lapack::getrs_scratchpad_size>(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb); } template <> std::int64_t getrs_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::transpose trans, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::getrs_scratchpad_size>(queue, trans, n, nrhs, + return ::oneapi::mkl::lapack::getrs_scratchpad_size>(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb); } template <> -std::int64_t heevd_scratchpad_size>(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t heevd_scratchpad_size>(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::heevd_scratchpad_size>(queue, jobz, uplo, n, + return ::oneapi::mkl::lapack::heevd_scratchpad_size>(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t heevd_scratchpad_size>(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t heevd_scratchpad_size>(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::heevd_scratchpad_size>(queue, jobz, uplo, n, + return ::oneapi::mkl::lapack::heevd_scratchpad_size>(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda); } template <> std::int64_t hegvd_scratchpad_size>(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::hegvd_scratchpad_size>(queue, itype, jobz, - uplo, n, lda, ldb); + return ::oneapi::mkl::lapack::hegvd_scratchpad_size>(queue, itype, detail::get_onemkl_job(jobz), + detail::get_onemkl_uplo(uplo), n, lda, ldb); } template <> std::int64_t hegvd_scratchpad_size>(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::hegvd_scratchpad_size>(queue, itype, jobz, - uplo, n, lda, ldb); + return ::oneapi::mkl::lapack::hegvd_scratchpad_size>(queue, itype, detail::get_onemkl_job(jobz), + detail::get_onemkl_uplo(uplo), n, lda, ldb); } template <> -std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::hetrd_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::hetrd_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::hetrd_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::hetrd_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::hetrf_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::hetrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::hetrf_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::hetrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, +std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::orgbr_scratchpad_size(queue, vect, m, n, k, lda); + return ::oneapi::mkl::lapack::orgbr_scratchpad_size(queue, detail::get_onemkl_generate(vect), m, n, k, lda); } template <> -std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::mkl::generate vect, +std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::orgbr_scratchpad_size(queue, vect, m, n, k, lda); + return ::oneapi::mkl::lapack::orgbr_scratchpad_size(queue, detail::get_onemkl_generate(vect), m, n, k, lda); } template <> -std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::orgtr_scratchpad_size(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::orgtr_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::orgtr_scratchpad_size(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::orgtr_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, @@ -2045,223 +2045,223 @@ std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, s return ::oneapi::mkl::lapack::orgqr_scratchpad_size(queue, m, n, k, lda); } template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormrq_scratchpad_size(queue, side, trans, m, n, k, lda, + return ::oneapi::mkl::lapack::ormrq_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, ldc); } template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormrq_scratchpad_size(queue, side, trans, m, n, k, lda, + return ::oneapi::mkl::lapack::ormrq_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, ldc); } template <> -std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormqr_scratchpad_size(queue, side, trans, m, n, k, lda, + return ::oneapi::mkl::lapack::ormqr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, ldc); } template <> -std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, +std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormqr_scratchpad_size(queue, side, trans, m, n, k, lda, + return ::oneapi::mkl::lapack::ormqr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, ldc); } template <> -std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormtr_scratchpad_size(queue, side, uplo, trans, m, n, lda, + return ::oneapi::mkl::lapack::ormtr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, ldc); } template <> -std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, +std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormtr_scratchpad_size(queue, side, uplo, trans, m, n, lda, + return ::oneapi::mkl::lapack::ormtr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, ldc); } template <> -std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potrf_scratchpad_size(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::potrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potrf_scratchpad_size(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::potrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t potrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potrf_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::potrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t potrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potrf_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::potrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::potrs_scratchpad_size(queue, uplo, n, nrhs, lda, ldb); + return ::oneapi::mkl::lapack::potrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb); } template <> -std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::potrs_scratchpad_size(queue, uplo, n, nrhs, lda, ldb); + return ::oneapi::mkl::lapack::potrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb); } template <> -std::int64_t potrs_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::potrs_scratchpad_size>(queue, uplo, n, nrhs, + return ::oneapi::mkl::lapack::potrs_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb); } template <> -std::int64_t potrs_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::potrs_scratchpad_size>(queue, uplo, n, nrhs, + return ::oneapi::mkl::lapack::potrs_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb); } template <> -std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potri_scratchpad_size(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::potri_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potri_scratchpad_size(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::potri_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t potri_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potri_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potri_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::potri_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t potri_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potri_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potri_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::potri_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrf_scratchpad_size(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::sytrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrf_scratchpad_size(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::sytrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrf_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::sytrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrf_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::sytrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::syevd_scratchpad_size(queue, jobz, uplo, n, lda); + return ::oneapi::mkl::lapack::syevd_scratchpad_size(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, +std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::syevd_scratchpad_size(queue, jobz, uplo, n, lda); + return ::oneapi::mkl::lapack::syevd_scratchpad_size(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda); } template <> std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, jobz, uplo, n, lda, + return ::oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, ldb); } template <> std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, jobz, uplo, n, lda, + return ::oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, ldb); } template <> -std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrd_scratchpad_size(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::sytrd_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrd_scratchpad_size(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::sytrd_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::trtrs_scratchpad_size(queue, uplo, trans, diag, n, nrhs, + return ::oneapi::mkl::lapack::trtrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, lda, ldb); } template <> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, +std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::trtrs_scratchpad_size(queue, uplo, trans, diag, n, nrhs, + return ::oneapi::mkl::lapack::trtrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, lda, ldb); } template <> -std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return ::oneapi::mkl::lapack::trtrs_scratchpad_size>( - queue, uplo, trans, diag, n, nrhs, lda, ldb); + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, lda, ldb); } template <> -std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, +std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return ::oneapi::mkl::lapack::trtrs_scratchpad_size>( - queue, uplo, trans, diag, n, nrhs, lda, ldb); + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, lda, ldb); } template <> std::int64_t ungbr_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::ungbr_scratchpad_size>(queue, vect, m, n, k, + return ::oneapi::mkl::lapack::ungbr_scratchpad_size>(queue, detail::get_onemkl_generate(vect), m, n, k, lda); } template <> std::int64_t ungbr_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::generate vect, std::int64_t m, + oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::ungbr_scratchpad_size>(queue, vect, m, n, k, + return ::oneapi::mkl::lapack::ungbr_scratchpad_size>(queue, detail::get_onemkl_generate(vect), m, n, k, lda); } template <> @@ -2277,68 +2277,68 @@ std::int64_t ungqr_scratchpad_size>(sycl::queue &queue, std return ::oneapi::mkl::lapack::ungqr_scratchpad_size>(queue, m, n, k, lda); } template <> -std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::ungtr_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::ungtr_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::ungtr_scratchpad_size>(queue, uplo, n, lda); + return ::oneapi::mkl::lapack::ungtr_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); } template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::unmrq_scratchpad_size>(queue, side, trans, m, + return ::oneapi::mkl::lapack::unmrq_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, ldc); } template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::unmrq_scratchpad_size>(queue, side, trans, m, + return ::oneapi::mkl::lapack::unmrq_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, ldc); } template <> -std::int64_t unmqr_scratchpad_size>(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmqr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::unmqr_scratchpad_size>(queue, side, trans, m, + return ::oneapi::mkl::lapack::unmqr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, ldc); } template <> -std::int64_t unmqr_scratchpad_size>(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, +std::int64_t unmqr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::unmqr_scratchpad_size>(queue, side, trans, m, + return ::oneapi::mkl::lapack::unmqr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, ldc); } template <> -std::int64_t unmtr_scratchpad_size>(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, +std::int64_t unmtr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::unmtr_scratchpad_size>(queue, side, uplo, - trans, m, n, lda, ldc); + return ::oneapi::mkl::lapack::unmtr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, lda, ldc); } template <> -std::int64_t unmtr_scratchpad_size>(sycl::queue &queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, +std::int64_t unmtr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { return ::oneapi::mkl::lapack::unmtr_scratchpad_size>( - queue, side, uplo, trans, m, n, lda, ldc); + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, ldc); } template <> std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, @@ -2407,38 +2407,38 @@ std::int64_t getri_batch_scratchpad_size>(sycl::queue &queu queue, n, lda, stride_a, stride_ipiv, batch_size); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size( - queue, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size( - queue, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } template <> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, @@ -2473,67 +2473,67 @@ std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queu queue, m, n, lda, stride_a, stride_tau, batch_size); } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, uplo, n, lda, stride_a, + return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size); } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, uplo, n, lda, stride_a, + return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size); } template <> std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, uplo, n, lda, stride_a, batch_size); + queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size); } template <> std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo uplo, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, uplo, n, lda, stride_a, batch_size); + queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size( - queue, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo uplo, +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size( - queue, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, + sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template <> std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, @@ -2628,36 +2628,36 @@ std::int64_t getri_batch_scratchpad_size>(sycl::queue &queu queue, n, lda, group_count, group_sizes); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose *trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size(queue, trans, n, nrhs, lda, + return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::transpose *trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size( - queue, trans, n, nrhs, lda, ldb, group_count, group_sizes); + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, trans, n, nrhs, lda, ldb, group_count, group_sizes); + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::transpose *trans, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, trans, n, nrhs, lda, ldb, group_count, group_sizes); + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes); } template <> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, @@ -2708,70 +2708,70 @@ std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_ group_count, group_sizes); } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, uplo, n, lda, + return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes); } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, uplo, n, lda, + return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes); } template <> std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo *uplo, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, uplo, n, lda, group_count, group_sizes); + queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes); } template <> std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, - oneapi::mkl::uplo *uplo, + oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, uplo, n, lda, group_count, group_sizes); + queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size(queue, uplo, n, nrhs, lda, ldb, + return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mkl::uplo *uplo, +std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size( - queue, uplo, n, nrhs, lda, ldb, group_count, group_sizes); + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes); } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, uplo, n, nrhs, lda, ldb, group_count, group_sizes); + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes); } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::mkl::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, + sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, uplo, n, nrhs, lda, ldb, group_count, group_sizes); + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes); } template <> std::int64_t ungqr_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, diff --git a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp b/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp index cc31eb4c1..152426c9c 100644 --- a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp +++ b/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp @@ -21,6 +21,7 @@ #include +#include "common_mkl_types_conversion.hpp" #include "oneapi/math/types.hpp" #include "oneapi/math/lapack/types.hpp" diff --git a/src/lapack/backends/mklcpu/CMakeLists.txt b/src/lapack/backends/mklcpu/CMakeLists.txt index 7099a68fb..6aa4906e8 100644 --- a/src/lapack/backends/mklcpu/CMakeLists.txt +++ b/src/lapack/backends/mklcpu/CMakeLists.txt @@ -29,12 +29,20 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_lapack ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/include - ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/src + ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) +# Due to using the same file name for different headers in this library and in +# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ +# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the +# correct header. +target_compile_options(${LIB_OBJ} + BEFORE PRIVATE -iquote $ +) + target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::LAPACK) diff --git a/src/lapack/backends/mklgpu/CMakeLists.txt b/src/lapack/backends/mklgpu/CMakeLists.txt index 49bf4ac75..83bb2fe73 100644 --- a/src/lapack/backends/mklgpu/CMakeLists.txt +++ b/src/lapack/backends/mklgpu/CMakeLists.txt @@ -29,12 +29,20 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_lapack ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/include - ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/src + ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) +# Due to using the same file name for different headers in this library and in +# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ +# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the +# correct header. +target_compile_options(${LIB_OBJ} + BEFORE PRIVATE -iquote $ +) + target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::LAPACK) diff --git a/src/rng/backends/mklgpu/mrg32k3a.cpp b/src/rng/backends/mklgpu/mrg32k3a.cpp index 886598ca3..303ae7fdd 100644 --- a/src/rng/backends/mklgpu/mrg32k3a.cpp +++ b/src/rng/backends/mklgpu/mrg32k3a.cpp @@ -24,7 +24,7 @@ #include #endif -#include "mkl_version.h" +#include #include "oneapi/math/rng/detail/engine_impl.hpp" #include "oneapi/math/rng/engines.hpp" diff --git a/src/rng/backends/mklgpu/philox4x32x10.cpp b/src/rng/backends/mklgpu/philox4x32x10.cpp index 2bf91145a..e08fce53d 100644 --- a/src/rng/backends/mklgpu/philox4x32x10.cpp +++ b/src/rng/backends/mklgpu/philox4x32x10.cpp @@ -24,7 +24,7 @@ #include #endif -#include "mkl_version.h" +#include #include "oneapi/math/rng/detail/engine_impl.hpp" #include "oneapi/math/rng/engines.hpp" diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.cxx b/src/sparse_blas/backends/mkl_common/mkl_handles.cxx index ea3f3abb3..ee8be6004 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.cxx @@ -171,7 +171,7 @@ void init_coo_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t * // The backend handle must use the buffers from the internal handle as they will be kept alive until the handle is released. oneapi::mkl::sparse::set_coo_data(queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), static_cast(nnz), - index, internal_smhandle->row_container.get_buffer(), + detail::get_onemkl_index_base(index), internal_smhandle->row_container.get_buffer(), internal_smhandle->col_container.get_buffer(), internal_smhandle->value_container.get_buffer()); *p_smhandle = reinterpret_cast(internal_smhandle); @@ -187,7 +187,7 @@ void init_coo_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t * auto internal_smhandle = new detail::sparse_matrix_handle(mkl_handle, row_ind, col_ind, val); auto event = oneapi::mkl::sparse::set_coo_data( queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), - static_cast(nnz), index, row_ind, col_ind, val); + static_cast(nnz), detail::get_onemkl_index_base(index), row_ind, col_ind, val); event.wait_and_throw(); *p_smhandle = reinterpret_cast(internal_smhandle); } @@ -225,7 +225,7 @@ void set_coo_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle oneapi::mkl::sparse::set_coo_data(queue, internal_smhandle->backend_handle, static_cast(num_rows), static_cast(num_cols), static_cast(nnz), - index, internal_smhandle->row_container.get_buffer(), + detail::get_onemkl_index_base(index), internal_smhandle->row_container.get_buffer(), internal_smhandle->col_container.get_buffer(), internal_smhandle->value_container.get_buffer()); } @@ -242,7 +242,7 @@ void set_coo_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle internal_smhandle->value_container.set_usm_ptr(val); auto event = oneapi::mkl::sparse::set_coo_data( queue, internal_smhandle->backend_handle, static_cast(num_rows), - static_cast(num_cols), static_cast(nnz), index, row_ind, col_ind, val); + static_cast(num_cols), static_cast(nnz), detail::get_onemkl_index_base(index), row_ind, col_ind, val); event.wait_and_throw(); } @@ -280,7 +280,7 @@ void init_csr_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t * // The backend deduces nnz from row_ptr. // The backend handle must use the buffers from the internal handle as they will be kept alive until the handle is released. oneapi::mkl::sparse::set_csr_data(queue, mkl_handle, static_cast(num_rows), - static_cast(num_cols), index, + static_cast(num_cols), detail::get_onemkl_index_base(index), internal_smhandle->row_container.get_buffer(), internal_smhandle->col_container.get_buffer(), internal_smhandle->value_container.get_buffer()); @@ -297,7 +297,7 @@ void init_csr_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t * auto internal_smhandle = new detail::sparse_matrix_handle(mkl_handle, row_ptr, col_ind, val); // The backend deduces nnz from row_ptr. auto event = oneapi::mkl::sparse::set_csr_data( - queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), index, + queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), detail::get_onemkl_index_base(index), row_ptr, col_ind, val); event.wait_and_throw(); *p_smhandle = reinterpret_cast(internal_smhandle); @@ -317,7 +317,7 @@ void set_csr_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle // The backend handle must use the buffers from the internal handle as they will be kept alive until the handle is released. oneapi::mkl::sparse::set_csr_data(queue, internal_smhandle->backend_handle, static_cast(num_rows), - static_cast(num_cols), index, + static_cast(num_cols), detail::get_onemkl_index_base(index), internal_smhandle->row_container.get_buffer(), internal_smhandle->col_container.get_buffer(), internal_smhandle->value_container.get_buffer()); @@ -336,7 +336,7 @@ void set_csr_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle // The backend deduces nnz from row_ptr. auto event = oneapi::mkl::sparse::set_csr_data( queue, internal_smhandle->backend_handle, static_cast(num_rows), - static_cast(num_cols), index, row_ptr, col_ind, val); + static_cast(num_cols), detail::get_onemkl_index_base(index), row_ptr, col_ind, val); event.wait_and_throw(); } diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp index 4e99b85de..80c2b8893 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp @@ -66,10 +66,10 @@ namespace oneapi::math::sparse::detail { * oneapi::math::sparse::detail::sparse_matrix_handle which holds another * matrix_handle_t for the backend handle. */ -using sparse_matrix_handle = detail::generic_sparse_handle; +using sparse_matrix_handle = detail::generic_sparse_handle; /// Cast to oneMath's internal handle type -inline auto get_internal_handle(matrix_handle_t handle) { +inline auto get_internal_handle(oneapi::math::sparse::matrix_handle_t handle) { return reinterpret_cast(handle); } diff --git a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp index 6ab54d560..5ac48eb3f 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp @@ -29,11 +29,14 @@ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/sparse_blas/detail/helper_types.hpp" +#include "common_mkl_types_conversion.hpp" #include "sparse_blas/enum_data_types.hpp" #include "sparse_blas/macros.hpp" namespace oneapi::math::sparse::detail { +using namespace oneapi::math::detail; + /// Return whether a pointer is accessible on the host template inline bool is_ptr_accessible_on_host(sycl::queue &queue, const T *host_or_device_ptr) { diff --git a/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx b/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx index 0510aee31..e4a0abe1a 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx @@ -193,19 +193,21 @@ sycl::event internal_spmm( detail::get_scalar_on_host(queue, static_cast(beta), is_beta_host_accessible); auto internal_A_handle = detail::get_internal_handle(A_handle); internal_A_handle->can_be_reset = false; - auto layout = B_handle->dense_layout; + auto onemkl_layout = detail::get_onemkl_layout(B_handle->dense_layout); + auto onemkl_opa = detail::get_onemkl_transpose(opA); + auto onemkl_opb = detail::get_onemkl_transpose(opB); auto columns = C_handle->num_cols; auto ldb = B_handle->ld; auto ldc = C_handle->ld; if (internal_A_handle->all_use_buffer()) { - oneapi::mkl::sparse::gemm(queue, layout, opA, opB, host_alpha, + oneapi::mkl::sparse::gemm(queue, onemkl_layout, onemkl_opa, onemkl_opb, host_alpha, internal_A_handle->backend_handle, B_handle->get_buffer(), columns, ldb, host_beta, C_handle->get_buffer(), ldc); // Dependencies are not used for buffers return {}; } else { - return oneapi::mkl::sparse::gemm(queue, layout, opA, opB, host_alpha, + return oneapi::mkl::sparse::gemm(queue, onemkl_layout, onemkl_opa, onemkl_opb, host_alpha, internal_A_handle->backend_handle, B_handle->get_usm_ptr(), columns, ldb, host_beta, C_handle->get_usm_ptr(), ldc, dependencies); diff --git a/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx b/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx index 097c79c27..539581e1f 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx @@ -143,8 +143,11 @@ void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void * return; } internal_A_handle->can_be_reset = false; + auto onemkl_uplo = detail::get_onemkl_uplo(A_view.uplo_view); + auto onemkl_opa = detail::get_onemkl_transpose(opA); + auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); if (A_view.type_view == matrix_descr::triangular) { - oneapi::mkl::sparse::optimize_trmv(queue, A_view.uplo_view, opA, A_view.diag_view, + oneapi::mkl::sparse::optimize_trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, internal_A_handle->backend_handle); } else if (A_view.type_view == matrix_descr::symmetric || @@ -153,7 +156,7 @@ void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void * return; } else { - oneapi::mkl::sparse::optimize_gemv(queue, opA, internal_A_handle->backend_handle); + oneapi::mkl::sparse::optimize_gemv(queue, onemkl_opa, internal_A_handle->backend_handle); } } @@ -175,8 +178,11 @@ sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const return detail::collapse_dependencies(queue, dependencies); } internal_A_handle->can_be_reset = false; + auto onemkl_uplo = detail::get_onemkl_uplo(A_view.uplo_view); + auto onemkl_opa = detail::get_onemkl_transpose(opA); + auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); if (A_view.type_view == matrix_descr::triangular) { - return oneapi::mkl::sparse::optimize_trmv(queue, A_view.uplo_view, opA, A_view.diag_view, + return oneapi::mkl::sparse::optimize_trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, internal_A_handle->backend_handle, dependencies); } else if (A_view.type_view == matrix_descr::symmetric || @@ -184,7 +190,7 @@ sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const return detail::collapse_dependencies(queue, dependencies); } else { - return oneapi::mkl::sparse::optimize_gemv(queue, opA, internal_A_handle->backend_handle, + return oneapi::mkl::sparse::optimize_gemv(queue, onemkl_opa, internal_A_handle->backend_handle, dependencies); } } @@ -206,20 +212,23 @@ sycl::event internal_spmv(sycl::queue &queue, oneapi::math::transpose opA, const auto internal_A_handle = detail::get_internal_handle(A_handle); internal_A_handle->can_be_reset = false; auto backend_handle = internal_A_handle->backend_handle; + auto onemkl_uplo = detail::get_onemkl_uplo(A_view.uplo_view); + auto onemkl_opa = detail::get_onemkl_transpose(opA); + auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); if (internal_A_handle->all_use_buffer()) { auto x_buffer = x_handle->get_buffer(); auto y_buffer = y_handle->get_buffer(); if (A_view.type_view == matrix_descr::triangular) { - oneapi::mkl::sparse::trmv(queue, A_view.uplo_view, opA, A_view.diag_view, host_alpha, + oneapi::mkl::sparse::trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, backend_handle, x_buffer, host_beta, y_buffer); } else if (A_view.type_view == matrix_descr::symmetric || A_view.type_view == matrix_descr::hermitian) { - oneapi::mkl::sparse::symv(queue, A_view.uplo_view, host_alpha, backend_handle, x_buffer, + oneapi::mkl::sparse::symv(queue, onemkl_uplo, host_alpha, backend_handle, x_buffer, host_beta, y_buffer); } else { - oneapi::mkl::sparse::gemv(queue, opA, host_alpha, backend_handle, x_buffer, host_beta, + oneapi::mkl::sparse::gemv(queue, onemkl_opa, host_alpha, backend_handle, x_buffer, host_beta, y_buffer); } // Dependencies are not used for buffers @@ -229,17 +238,17 @@ sycl::event internal_spmv(sycl::queue &queue, oneapi::math::transpose opA, const auto x_usm = x_handle->get_usm_ptr(); auto y_usm = y_handle->get_usm_ptr(); if (A_view.type_view == matrix_descr::triangular) { - return oneapi::mkl::sparse::trmv(queue, A_view.uplo_view, opA, A_view.diag_view, + return oneapi::mkl::sparse::trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, backend_handle, x_usm, host_beta, y_usm, dependencies); } else if (A_view.type_view == matrix_descr::symmetric || A_view.type_view == matrix_descr::hermitian) { - return oneapi::mkl::sparse::symv(queue, A_view.uplo_view, host_alpha, backend_handle, + return oneapi::mkl::sparse::symv(queue, onemkl_uplo, host_alpha, backend_handle, x_usm, host_beta, y_usm, dependencies); } else { - return oneapi::mkl::sparse::gemv(queue, opA, host_alpha, backend_handle, x_usm, + return oneapi::mkl::sparse::gemv(queue, onemkl_opa, host_alpha, backend_handle, x_usm, host_beta, y_usm, dependencies); } } diff --git a/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx b/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx index 554281080..72253175b 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx @@ -138,7 +138,10 @@ void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void * return; } internal_A_handle->can_be_reset = false; - oneapi::mkl::sparse::optimize_trsv(queue, A_view.uplo_view, opA, A_view.diag_view, + auto onemkl_uplo = detail::get_onemkl_uplo(A_view.uplo_view); + auto onemkl_opa = detail::get_onemkl_transpose(opA); + auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); + oneapi::mkl::sparse::optimize_trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, internal_A_handle->backend_handle); } @@ -159,7 +162,10 @@ sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const return detail::collapse_dependencies(queue, dependencies); } internal_A_handle->can_be_reset = false; - return oneapi::mkl::sparse::optimize_trsv(queue, A_view.uplo_view, opA, A_view.diag_view, + auto onemkl_uplo = detail::get_onemkl_uplo(A_view.uplo_view); + auto onemkl_opa = detail::get_onemkl_transpose(opA); + auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); + return oneapi::mkl::sparse::optimize_trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, internal_A_handle->backend_handle, dependencies); } @@ -177,15 +183,18 @@ sycl::event internal_spsv(sycl::queue &queue, oneapi::math::transpose opA, const detail::get_scalar_on_host(queue, static_cast(alpha), is_alpha_host_accessible); auto internal_A_handle = detail::get_internal_handle(A_handle); internal_A_handle->can_be_reset = false; + auto onemkl_uplo = detail::get_onemkl_uplo(A_view.uplo_view); + auto onemkl_opa = detail::get_onemkl_transpose(opA); + auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); if (internal_A_handle->all_use_buffer()) { - oneapi::mkl::sparse::trsv(queue, A_view.uplo_view, opA, A_view.diag_view, host_alpha, + oneapi::mkl::sparse::trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, internal_A_handle->backend_handle, x_handle->get_buffer(), y_handle->get_buffer()); // Dependencies are not used for buffers return {}; } else { - return oneapi::mkl::sparse::trsv(queue, A_view.uplo_view, opA, A_view.diag_view, host_alpha, + return oneapi::mkl::sparse::trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, internal_A_handle->backend_handle, x_handle->get_usm_ptr(), y_handle->get_usm_ptr(), dependencies); diff --git a/src/sparse_blas/backends/mklcpu/CMakeLists.txt b/src/sparse_blas/backends/mklcpu/CMakeLists.txt index 71d056f7c..9b24beb41 100644 --- a/src/sparse_blas/backends/mklcpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklcpu/CMakeLists.txt @@ -34,6 +34,7 @@ add_dependencies(onemath_backend_libs_sparse_blas ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src + ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) diff --git a/src/sparse_blas/backends/mklgpu/CMakeLists.txt b/src/sparse_blas/backends/mklgpu/CMakeLists.txt index 0ea8f6b93..4adcf9861 100644 --- a/src/sparse_blas/backends/mklgpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklgpu/CMakeLists.txt @@ -34,6 +34,7 @@ add_dependencies(onemath_backend_libs_sparse_blas ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src + ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) diff --git a/src/sparse_blas/generic_container.hpp b/src/sparse_blas/generic_container.hpp index e2377b951..463956275 100644 --- a/src/sparse_blas/generic_container.hpp +++ b/src/sparse_blas/generic_container.hpp @@ -29,6 +29,7 @@ #include #endif +#include "oneapi/math/exceptions.hpp" #include "oneapi/math/sparse_blas/types.hpp" #include "enum_data_types.hpp" From 5e838895bb3c398f4cabd11f68685319bb9cc9cf Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Wed, 16 Oct 2024 16:56:44 +0100 Subject: [PATCH 24/41] Fix rng MKL backends --- src/rng/backends/mklcpu/CMakeLists.txt | 11 +- src/rng/backends/mklcpu/mrg32k3a.cpp | 3 +- src/rng/backends/mklcpu/philox4x32x10.cpp | 3 +- src/rng/backends/mklgpu/CMakeLists.txt | 12 +- .../backends/mklgpu/mkl_internal_rng_gpu.hpp | 81 ----------- src/rng/backends/mklgpu/mrg32k3a.cpp | 108 +++++++-------- .../mklgpu/onemkl_distribution_conversion.hpp | 131 ++++++++++++++++++ src/rng/backends/mklgpu/philox4x32x10.cpp | 108 +++++++-------- 8 files changed, 255 insertions(+), 202 deletions(-) delete mode 100755 src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp create mode 100644 src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp diff --git a/src/rng/backends/mklcpu/CMakeLists.txt b/src/rng/backends/mklcpu/CMakeLists.txt index 297206b9f..a58e699f9 100644 --- a/src/rng/backends/mklcpu/CMakeLists.txt +++ b/src/rng/backends/mklcpu/CMakeLists.txt @@ -31,12 +31,19 @@ add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/include - ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) +# Due to using the same file name for different headers in this library and in +# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ +# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the +# correct header. +target_compile_options(${LIB_OBJ} + BEFORE PRIVATE -iquote $ +) + target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET ${LIB_OBJ} SOURCES ${SOURCES}) diff --git a/src/rng/backends/mklcpu/mrg32k3a.cpp b/src/rng/backends/mklcpu/mrg32k3a.cpp index 4e9958211..89af18bd6 100644 --- a/src/rng/backends/mklcpu/mrg32k3a.cpp +++ b/src/rng/backends/mklcpu/mrg32k3a.cpp @@ -24,7 +24,8 @@ #include #endif -#include "mkl_vsl.h" +// Intel(R) oneMKL header +#include #include "oneapi/math/exceptions.hpp" #include "oneapi/math/rng/detail/engine_impl.hpp" diff --git a/src/rng/backends/mklcpu/philox4x32x10.cpp b/src/rng/backends/mklcpu/philox4x32x10.cpp index ecea512ce..58b4c3aee 100644 --- a/src/rng/backends/mklcpu/philox4x32x10.cpp +++ b/src/rng/backends/mklcpu/philox4x32x10.cpp @@ -24,7 +24,8 @@ #include #endif -#include "mkl_vsl.h" +// Intel(R) oneMKL header +#include #include "oneapi/math/exceptions.hpp" #include "oneapi/math/rng/detail/engine_impl.hpp" diff --git a/src/rng/backends/mklgpu/CMakeLists.txt b/src/rng/backends/mklgpu/CMakeLists.txt index 9335f0d7d..b008e1770 100644 --- a/src/rng/backends/mklgpu/CMakeLists.txt +++ b/src/rng/backends/mklgpu/CMakeLists.txt @@ -23,7 +23,6 @@ set(LIB_OBJ ${LIB_NAME}_obj) add_library(${LIB_NAME}) add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT - mkl_internal_rng_gpu.hpp philox4x32x10.cpp mrg32k3a.cpp $<$: mkl_rng_gpu_wrappers.cpp> @@ -31,12 +30,19 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/include - ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) +# Due to using the same file name for different headers in this library and in +# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ +# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the +# correct header. +target_compile_options(${LIB_OBJ} + BEFORE PRIVATE -iquote $ +) + target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::RNG) diff --git a/src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp b/src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp deleted file mode 100755 index 35923601a..000000000 --- a/src/rng/backends/mklgpu/mkl_internal_rng_gpu.hpp +++ /dev/null @@ -1,81 +0,0 @@ -/******************************************************************************* -* Copyright 2020-2021 Intel Corporation -* -* Licensed under the Apache License, Version 2.0 (the "License"); -* you may not use this file except in compliance with the License. -* You may obtain a copy of the License at -* -* http://www.apache.org/licenses/LICENSE-2.0 -* -* Unless required by applicable law or agreed to in writing, -* software distributed under the License is distributed on an "AS IS" BASIS, -* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -* See the License for the specific language governing permissions -* and limitations under the License. -* -* -* SPDX-License-Identifier: Apache-2.0 -*******************************************************************************/ - -#ifndef ONEMATH_INTERNAL_RNG_GPU_HPP_ -#define ONEMATH_INTERNAL_RNG_GPU_HPP_ - -#if __has_include() -#include -#else -#include -#endif - -namespace oneapi { -namespace math { -namespace rng { -namespace detail { - -template -class engine_base_impl; - -namespace gpu { - -template -engine_base_impl* create_engine(sycl::queue& queue, std::uint64_t seed); - -template -engine_base_impl* create_engine(sycl::queue& queue, std::int64_t n, - const unsigned int* seed_ptr); - -template -engine_base_impl* create_engine(sycl::queue& queue, - engine_base_impl* other_impl); - -template -void skip_ahead(sycl::queue& queue, engine_base_impl* impl, std::uint64_t num_to_skip); - -template -void skip_ahead(sycl::queue& queue, engine_base_impl* impl, - std::initializer_list num_to_skip); - -template -void leapfrog(sycl::queue& queue, engine_base_impl* impl, std::uint64_t idx, - std::uint64_t stride); - -template -void delete_engine(sycl::queue& queue, engine_base_impl* impl); - -template -sycl::event generate(sycl::queue& queue, const DistrType& distr, - engine_base_impl* engine, std::int64_t n, - sycl::buffer& r); - -template -sycl::event generate(sycl::queue& queue, const DistrType& distr, - engine_base_impl* engine, std::int64_t n, - typename DistrType::result_type* r, - const std::vector& dependencies = {}); - -} // namespace gpu -} // namespace detail -} // namespace rng -} // namespace math -} // namespace oneapi - -#endif //_MKL_INTERNAL_RNG_GPU_HPP_ diff --git a/src/rng/backends/mklgpu/mrg32k3a.cpp b/src/rng/backends/mklgpu/mrg32k3a.cpp index 303ae7fdd..f4e17145b 100644 --- a/src/rng/backends/mklgpu/mrg32k3a.cpp +++ b/src/rng/backends/mklgpu/mrg32k3a.cpp @@ -24,14 +24,16 @@ #include #endif +// Intel(R) oneMKL headers #include +#include #include "oneapi/math/rng/detail/engine_impl.hpp" #include "oneapi/math/rng/engines.hpp" #include "oneapi/math/exceptions.hpp" #include "oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp" -#include "mkl_internal_rng_gpu.hpp" +#include "onemkl_distribution_conversion.hpp" namespace oneapi { namespace math { @@ -41,21 +43,14 @@ namespace mklgpu { class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { public: mrg32k3a_impl(sycl::queue queue, std::uint32_t seed) - : oneapi::math::rng::detail::engine_impl(queue) { - engine_ = - oneapi::math::rng::detail::gpu::create_engine(queue, seed); + : oneapi::math::rng::detail::engine_impl(queue), engine_((oneapi::mkl::rng::mrg32k3a(queue, seed))) { } mrg32k3a_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::math::rng::detail::engine_impl(queue) { - engine_ = oneapi::math::rng::detail::gpu::create_engine( - queue, (std::int64_t)(seed.size()), (const unsigned int*)seed.begin()); + : oneapi::math::rng::detail::engine_impl(queue), engine_((oneapi::mkl::rng::mrg32k3a(queue, seed))) { } - mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::math::rng::detail::engine_impl(*other) { - sycl::queue queue(other->queue_); - engine_ = oneapi::math::rng::detail::gpu::create_engine( - queue, other->engine_); + mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::math::rng::detail::engine_impl(*other), engine_((oneapi::mkl::rng::mrg32k3a(other->engine_))) { } // Buffers API @@ -63,104 +58,104 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const oneapi::math::rng::uniform< std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const oneapi::math::rng::gaussian< float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const oneapi::math::rng::gaussian< double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const oneapi::math::rng::lognormal< float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const oneapi::math::rng::lognormal< double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } - virtual void generate(const bits& distr, std::int64_t n, - sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + virtual void generate(const bits& /*distr*/, std::int64_t /*n*/, + sycl::buffer& /*r*/) override { + throw unimplemented("rng/mklgpu", "mrg32k3a::generate", "bits distribution is not supported"); } // USM APIs @@ -168,13 +163,13 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( @@ -182,100 +177,101 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { ; - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } - virtual sycl::event generate(const bits& distr, std::int64_t n, std::uint32_t* r, - const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + virtual sycl::event generate(const bits& /*distr*/, std::int64_t /*n*/, std::uint32_t* /*r*/, + const std::vector& /*dependencies*/) override { + throw unimplemented("rng/mklgpu", "mrg32k3a::generate", "bits distribution is not supported"); + return {}; } virtual oneapi::math::rng::detail::engine_impl* copy_state() override { @@ -283,23 +279,21 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual void skip_ahead(std::uint64_t num_to_skip) override { - oneapi::math::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); + oneapi::mkl::rng::skip_ahead(engine_, num_to_skip); } virtual void skip_ahead(std::initializer_list num_to_skip) override { - oneapi::math::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); + oneapi::mkl::rng::skip_ahead(engine_, num_to_skip); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { throw oneapi::math::unimplemented("rng", "leapfrog"); } - virtual ~mrg32k3a_impl() override { - oneapi::math::rng::detail::gpu::delete_engine(queue_, engine_); - } + virtual ~mrg32k3a_impl() override {} private: - oneapi::math::rng::detail::engine_base_impl* engine_; + oneapi::mkl::rng::mrg32k3a engine_; }; oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std::uint32_t seed) { diff --git a/src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp b/src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp new file mode 100644 index 000000000..d1d16e293 --- /dev/null +++ b/src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp @@ -0,0 +1,131 @@ +/******************************************************************************* +* Copyright Codeplay Software Ltd +* +* Licensed under the Apache License, Version 2.0 (the "License"); +* you may not use this file except in compliance with the License. +* You may obtain a copy of the License at +* +* http://www.apache.org/licenses/LICENSE-2.0 +* +* Unless required by applicable law or agreed to in writing, +* software distributed under the License is distributed on an "AS IS" BASIS, +* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +* See the License for the specific language governing permissions +* and limitations under the License. +* +* +* SPDX-License-Identifier: Apache-2.0 +*******************************************************************************/ + +#ifndef _ONEMATH_SRC_RNG_ONEMKL_DISTRIBUTION_CONVERSION_HPP_ +#define _ONEMATH_SRC_RNG_ONEMKL_DISTRIBUTION_CONVERSION_HPP_ + +// Convert oneMath RNG distribution types to Intel(R) oneMKL equivalents + +#include + +#include "common_onemkl_conversion.hpp" +#include "oneapi/math/rng/distributions.hpp" + +namespace oneapi { +namespace math { +namespace rng { +namespace detail { + +template +struct convert_method_t; + +template <> +struct convert_method_t { + using type = oneapi::mkl::rng::uniform_method::standard; +}; + +template <> +struct convert_method_t { + using type = oneapi::mkl::rng::uniform_method::accurate; +}; + +template <> +struct convert_method_t { + using type = oneapi::mkl::rng::gaussian_method::icdf; +}; + +template <> +struct convert_method_t { + using type = oneapi::mkl::rng::gaussian_method::box_muller2; +}; + +template <> +struct convert_method_t { + using type = oneapi::mkl::rng::lognormal_method::icdf; +}; + +template <> +struct convert_method_t { + using type = oneapi::mkl::rng::lognormal_method::box_muller2; +}; + +template <> +struct convert_method_t { + using type = oneapi::mkl::rng::bernoulli_method::icdf; +}; + +template <> +struct convert_method_t { + using type = oneapi::mkl::rng::poisson_method::gaussian_icdf_based; +}; + +template +struct convert_distrib_t; + +template +struct convert_distrib_t> { +auto operator()(uniform distribution) { + using onemkl_method_t = typename convert_method_t::type; + return oneapi::mkl::rng::uniform(distribution.a(), distribution.b()); +} +}; + +template +struct convert_distrib_t> { +auto operator()(gaussian distribution) { + using onemkl_method_t = typename convert_method_t::type; + return oneapi::mkl::rng::gaussian(distribution.mean(), distribution.stddev()); +} +}; + +template +struct convert_distrib_t> { +auto operator()(lognormal distribution) { + using onemkl_method_t = typename convert_method_t::type; + return oneapi::mkl::rng::lognormal(distribution.m(), distribution.s(), distribution.displ(), distribution.scale()); +} +}; + +template +struct convert_distrib_t> { +auto operator()(bernoulli distribution) { + using onemkl_method_t = typename convert_method_t::type; + return oneapi::mkl::rng::bernoulli(distribution.p()); +} +}; + +template +struct convert_distrib_t> { +auto operator()(poisson distribution) { + using onemkl_method_t = typename convert_method_t::type; + return oneapi::mkl::rng::poisson(distribution.lambda()); +} +}; + +template +inline auto get_onemkl_distribution(DistributionT distribution) { + return convert_distrib_t()(distribution); +} + +} // namespace detail +} // namespace rng +} // namespace math +} // namespace oneapi + +#endif // _ONEMATH_SRC_RNG_ONEMKL_DISTRIBUTION_CONVERSION_HPP_ diff --git a/src/rng/backends/mklgpu/philox4x32x10.cpp b/src/rng/backends/mklgpu/philox4x32x10.cpp index e08fce53d..43a8a30b4 100644 --- a/src/rng/backends/mklgpu/philox4x32x10.cpp +++ b/src/rng/backends/mklgpu/philox4x32x10.cpp @@ -24,14 +24,16 @@ #include #endif +// Intel(R) oneMKL headers #include +#include #include "oneapi/math/rng/detail/engine_impl.hpp" #include "oneapi/math/rng/engines.hpp" #include "oneapi/math/exceptions.hpp" #include "oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp" -#include "mkl_internal_rng_gpu.hpp" +#include "onemkl_distribution_conversion.hpp" namespace oneapi { namespace math { @@ -41,22 +43,15 @@ namespace mklgpu { class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { public: philox4x32x10_impl(sycl::queue queue, std::uint64_t seed) - : oneapi::math::rng::detail::engine_impl(queue) { - engine_ = oneapi::math::rng::detail::gpu::create_engine( - queue, seed); + : oneapi::math::rng::detail::engine_impl(queue), engine_((oneapi::mkl::rng::philox4x32x10(queue, seed))) { } philox4x32x10_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::math::rng::detail::engine_impl(queue) { - engine_ = oneapi::math::rng::detail::gpu::create_engine( - queue, (std::int64_t)(seed.size() * 2), (const unsigned int*)seed.begin()); + : oneapi::math::rng::detail::engine_impl(queue), engine_((oneapi::mkl::rng::philox4x32x10(queue, seed))) { } philox4x32x10_impl(const philox4x32x10_impl* other) - : oneapi::math::rng::detail::engine_impl(*other) { - sycl::queue queue(other->queue_); - engine_ = oneapi::math::rng::detail::gpu::create_engine( - queue, other->engine_); + : oneapi::math::rng::detail::engine_impl(*other), engine_((oneapi::mkl::rng::philox4x32x10(other->engine_))) { } // Buffers API @@ -64,104 +59,104 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const oneapi::math::rng::uniform< std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const oneapi::math::rng::gaussian< float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const oneapi::math::rng::gaussian< double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const oneapi::math::rng::lognormal< float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const oneapi::math::rng::lognormal< double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); } - virtual void generate(const bits& distr, std::int64_t n, - sycl::buffer& r) override { - oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r); + virtual void generate(const bits& /*distr*/, std::int64_t /*n*/, + sycl::buffer& /*r*/) override { + throw unimplemented("rng/mklgpu", "philox4x32x10::generate", "bits distribution is not supported"); } // USM APIs @@ -169,13 +164,13 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( @@ -183,100 +178,101 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { ; - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); } - virtual sycl::event generate(const bits& distr, std::int64_t n, std::uint32_t* r, - const std::vector& dependencies) override { - return oneapi::math::rng::detail::gpu::generate(queue_, distr, engine_, n, r, dependencies); + virtual sycl::event generate(const bits& /*distr*/, std::int64_t /*n*/, std::uint32_t* /*r*/, + const std::vector& /*dependencies*/) override { + throw unimplemented("rng/mklgpu", "philox4x32x10::generate", "bits distribution is not supported"); + return {}; } virtual oneapi::math::rng::detail::engine_impl* copy_state() override { @@ -284,23 +280,21 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual void skip_ahead(std::uint64_t num_to_skip) override { - oneapi::math::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); + oneapi::mkl::rng::skip_ahead(engine_, num_to_skip); } virtual void skip_ahead(std::initializer_list num_to_skip) override { - oneapi::math::rng::detail::gpu::skip_ahead(queue_, engine_, num_to_skip); + oneapi::mkl::rng::skip_ahead(engine_, num_to_skip); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { throw oneapi::math::unimplemented("rng", "leapfrog"); } - virtual ~philox4x32x10_impl() override { - oneapi::math::rng::detail::gpu::delete_engine(queue_, engine_); - } + virtual ~philox4x32x10_impl() override {} private: - oneapi::math::rng::detail::engine_base_impl* engine_; + oneapi::mkl::rng::philox4x32x10 engine_; }; oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { From a499e4fa73cdcf9bf5842706d4575a04d5c18501 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Thu, 17 Oct 2024 12:13:23 +0100 Subject: [PATCH 25/41] Rethrow oneapi::mkl exceptions as oneapi::math ones --- include/oneapi/math/detail/exceptions.hpp | 10 + include/oneapi/math/exceptions.hpp | 18 +- src/blas/backends/mkl_common/mkl_batch.cxx | 476 +++--- .../backends/mkl_common/mkl_blas_backend.hpp | 2 +- .../backends/mkl_common/mkl_extensions.cxx | 120 +- src/blas/backends/mkl_common/mkl_level1.cxx | 232 +-- src/blas/backends/mkl_common/mkl_level2.cxx | 280 ++-- src/blas/backends/mkl_common/mkl_level3.cxx | 214 +-- src/dft/backends/mklcpu/CMakeLists.txt | 1 + src/dft/backends/mklgpu/CMakeLists.txt | 1 + src/dft/backends/mklgpu/backward.cpp | 3 +- src/dft/backends/mklgpu/commit.cpp | 35 +- src/dft/backends/mklgpu/forward.cpp | 3 +- ...rsion.hpp => common_onemkl_conversion.hpp} | 39 +- src/lapack/backends/mkl_common/mkl_lapack.cxx | 1376 ++++++++--------- .../mkl_common/mkl_lapack_backend.hpp | 2 +- src/rng/backends/mklgpu/CMakeLists.txt | 1 + src/rng/backends/mklgpu/mrg32k3a.cpp | 72 +- src/rng/backends/mklgpu/philox4x32x10.cpp | 72 +- .../backends/mkl_common/mkl_handles.cxx | 44 +- .../backends/mkl_common/mkl_helper.hpp | 2 +- .../backends/mkl_common/mkl_spmm.cxx | 8 +- .../backends/mkl_common/mkl_spmv.cxx | 38 +- .../backends/mkl_common/mkl_spsv.cxx | 16 +- .../blas/include/reference_blas_wrappers.hpp | 2 +- 25 files changed, 1547 insertions(+), 1520 deletions(-) rename src/include/{common_mkl_types_conversion.hpp => common_onemkl_conversion.hpp} (70%) diff --git a/include/oneapi/math/detail/exceptions.hpp b/include/oneapi/math/detail/exceptions.hpp index 0a1fc0a73..8a82db01a 100644 --- a/include/oneapi/math/detail/exceptions.hpp +++ b/include/oneapi/math/detail/exceptions.hpp @@ -45,6 +45,16 @@ class function_not_found : public oneapi::math::exception { } }; +class library_not_found : public oneapi::math::exception { +public: + library_not_found(const std::string& message) : exception(message) {} + library_not_found(const std::string &domain, const std::string &function, + const std::string &info = "") + : oneapi::math::exception( + domain, function, + "library not found" + ((info.length() != 0) ? (": " + info) : "")) {} +}; + class specification_mismatch : public oneapi::math::exception { public: specification_mismatch(const std::string &info = "") diff --git a/include/oneapi/math/exceptions.hpp b/include/oneapi/math/exceptions.hpp index babb3af0a..ef99a1fb9 100644 --- a/include/oneapi/math/exceptions.hpp +++ b/include/oneapi/math/exceptions.hpp @@ -38,6 +38,7 @@ class exception : public std::exception { std::string msg_; public: + exception(const std::string& message) : std::exception(), msg_(message) {} exception(const std::string &domain, const std::string &function, const std::string &info = "") : std::exception() { msg_ = std::string("oneMath: ") + domain + @@ -54,6 +55,7 @@ class exception : public std::exception { class unsupported_device : public oneapi::math::exception { public: + unsupported_device(const std::string& message) : exception(message) {} unsupported_device(const std::string &domain, const std::string &function, const sycl::device &device) : oneapi::math::exception( @@ -63,12 +65,14 @@ class unsupported_device : public oneapi::math::exception { class host_bad_alloc : public oneapi::math::exception { public: + host_bad_alloc(const std::string& message) : exception(message) {} host_bad_alloc(const std::string &domain, const std::string &function) : oneapi::math::exception(domain, function, "cannot allocate memory on host") {} }; class device_bad_alloc : public oneapi::math::exception { public: + device_bad_alloc(const std::string& message) : exception(message) {} device_bad_alloc(const std::string &domain, const std::string &function, const sycl::device &device) : oneapi::math::exception( @@ -78,6 +82,7 @@ class device_bad_alloc : public oneapi::math::exception { class unimplemented : public oneapi::math::exception { public: + unimplemented(const std::string& message) : exception(message) {} unimplemented(const std::string &domain, const std::string &function, const std::string &info = "") : oneapi::math::exception(domain, function, "function is not implemented " + info) {} @@ -85,6 +90,7 @@ class unimplemented : public oneapi::math::exception { class invalid_argument : public oneapi::math::exception { public: + invalid_argument(const std::string& message) : exception(message) {} invalid_argument(const std::string &domain, const std::string &function, const std::string &info = "") : oneapi::math::exception(domain, function, "invalid argument " + info) {} @@ -92,6 +98,7 @@ class invalid_argument : public oneapi::math::exception { class uninitialized : public oneapi::math::exception { public: + uninitialized(const std::string& message) : exception(message) {} uninitialized(const std::string &domain, const std::string &function, const std::string &info = "") : oneapi::math::exception(domain, function, @@ -100,6 +107,7 @@ class uninitialized : public oneapi::math::exception { class computation_error : public oneapi::math::exception { public: + computation_error(const std::string& message) : exception(message) {} computation_error(const std::string &domain, const std::string &function, const std::string &info = "") : oneapi::math::exception( @@ -109,21 +117,13 @@ class computation_error : public oneapi::math::exception { class batch_error : public oneapi::math::exception { public: + batch_error(const std::string& message) : exception(message) {} batch_error(const std::string &domain, const std::string &function, const std::string &info = "") : oneapi::math::exception(domain, function, "batch error" + ((info.length() != 0) ? (": " + info) : "")) {} }; -class library_not_found : public oneapi::math::exception { -public: - library_not_found(const std::string &domain, const std::string &function, - const std::string &info = "") - : oneapi::math::exception( - domain, function, - "library not found" + ((info.length() != 0) ? (": " + info) : "")) {} -}; - } // namespace math } // namespace oneapi diff --git a/src/blas/backends/mkl_common/mkl_batch.cxx b/src/blas/backends/mkl_common/mkl_batch.cxx index 07bb70cde..93bba84a1 100644 --- a/src/blas/backends/mkl_common/mkl_batch.cxx +++ b/src/blas/backends/mkl_common/mkl_batch.cxx @@ -22,67 +22,67 @@ void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, std::int64_t stridex, sycl::buffer &y, int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); } void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, std::int64_t stridex, sycl::buffer &y, int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); } void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); } void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); } void axpy_batch(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { - blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); } void axpy_batch(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { - blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); } void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx, int64_t stridex, sycl::buffer, 1> &y, int64_t incy, int64_t stridey, int64_t batch_size) { - blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); } void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, sycl::buffer, 1> &x, int64_t incx, int64_t stridex, sycl::buffer, 1> &y, int64_t incy, int64_t stridey, int64_t batch_size) { - blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); } void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, int64_t incx, int64_t stride_x, float beta, sycl::buffer &y, int64_t incy, int64_t stride_y, int64_t batch_size) { - blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, - incy, stride_y, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, + incy, stride_y, batch_size)); } void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, int64_t incx, int64_t stride_x, double beta, sycl::buffer &y, int64_t incy, int64_t stride_y, int64_t batch_size) { - blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, - incy, stride_y, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, + incy, stride_y, batch_size)); } void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, @@ -90,8 +90,8 @@ void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, int64_t stride_a, sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, std::complex beta, sycl::buffer, 1> &y, int64_t incy, int64_t stride_y, int64_t batch_size) { - blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, - incy, stride_y, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, + incy, stride_y, batch_size)); } void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, @@ -100,24 +100,24 @@ void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, int64_t stride_x, std::complex beta, sycl::buffer, 1> &y, int64_t incy, int64_t stride_y, int64_t batch_size) { - blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, - incy, stride_y, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, + incy, stride_y, batch_size)); } void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, int64_t incx, int64_t stride_x, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, - stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, + stride_c, batch_size)); } void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, int64_t incx, int64_t stride_x, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, - stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, + stride_c, batch_size)); } void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, @@ -125,8 +125,8 @@ void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, - stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, + stride_c, batch_size)); } void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, @@ -134,24 +134,24 @@ void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, - stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, + stride_c, batch_size)); } void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size)); } void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size)); } void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -159,8 +159,8 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size)); } void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -169,8 +169,8 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t ldb, int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size)); } void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -178,8 +178,8 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, sycl::half beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size)); } void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -187,8 +187,8 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size)); } void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -205,24 +205,24 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size)); } void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - stride_a, b, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + stride_a, b, ldb, stride_b, batch_size)); } void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - stride_a, b, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + stride_a, b, ldb, stride_b, batch_size)); } void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, @@ -230,8 +230,8 @@ void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - stride_a, b, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + stride_a, b, ldb, stride_b, batch_size)); } void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, @@ -239,31 +239,31 @@ void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - stride_a, b, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + stride_a, b, ldb, stride_b, batch_size)); } void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, - stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, + stride_c, batch_size)); } void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, - stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, + stride_c, batch_size)); } void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, - stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, + stride_c, batch_size)); } void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, @@ -271,78 +271,78 @@ void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, - stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, + stride_c, batch_size)); } void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + batch_size)); } void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + batch_size)); } void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + batch_size)); } void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + batch_size)); } void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); } void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); } void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); } void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); } void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, float beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size)); } void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, double beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size)); } void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -351,8 +351,8 @@ void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64 sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size)); } void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -361,8 +361,8 @@ void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64 sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size)); } // USM APIs @@ -370,127 +370,127 @@ void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64 sycl::event copy_batch(sycl::queue &queue, int64_t n, const float *x, int64_t incx, std::int64_t stridex, float *y, int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - return blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, + dependencies)); } sycl::event copy_batch(sycl::queue &queue, int64_t n, const double *x, int64_t incx, std::int64_t stridex, double *y, int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - return blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, + dependencies)); } sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, std::int64_t stridex, std::complex *y, int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - return blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, + dependencies)); } sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, std::int64_t stridex, std::complex *y, int64_t incy, std::int64_t stridey, std::int64_t batch_size, const std::vector &dependencies) { - return blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, + dependencies)); } sycl::event copy_batch(sycl::queue &queue, int64_t *n, const float **x, int64_t *incx, float **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, + dependencies)); } sycl::event copy_batch(sycl::queue &queue, int64_t *n, const double **x, int64_t *incx, double **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, + dependencies)); } sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, + dependencies)); } sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, + dependencies)); } sycl::event axpy_batch(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, int64_t stridex, float *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - return blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, + dependencies)); } sycl::event axpy_batch(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, int64_t stridex, double *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - return blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, + dependencies)); } sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, const std::complex *x, int64_t incx, int64_t stridex, std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - return blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, + dependencies)); } sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, const std::complex *x, int64_t incx, int64_t stridex, std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, const std::vector &dependencies) { - return blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, + dependencies)); } sycl::event axpy_batch(sycl::queue &queue, int64_t *n, float *alpha, const float **x, int64_t *incx, float **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, + dependencies)); } sycl::event axpy_batch(sycl::queue &queue, int64_t *n, double *alpha, const double **x, int64_t *incx, double **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, + dependencies)); } sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, const std::complex **x, int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, + dependencies)); } sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, const std::complex **x, int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, + dependencies)); } sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, const float *x, int64_t incx, int64_t stride_x, float beta, float *y, int64_t incy, int64_t stride_y, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, - beta, y, incy, stride_y, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, + beta, y, incy, stride_y, batch_size, dependencies)); } sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, double alpha, @@ -498,8 +498,8 @@ sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t int64_t incx, int64_t stride_x, double beta, double *y, int64_t incy, int64_t stride_y, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, - beta, y, incy, stride_y, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, + beta, y, incy, stride_y, batch_size, dependencies)); } sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, @@ -508,8 +508,8 @@ sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t int64_t stride_x, std::complex beta, std::complex *y, int64_t incy, int64_t stride_y, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, - beta, y, incy, stride_y, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, + beta, y, incy, stride_y, batch_size, dependencies)); } sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, @@ -518,24 +518,24 @@ sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t int64_t stride_x, std::complex beta, std::complex *y, int64_t incy, int64_t stride_y, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, - beta, y, incy, stride_y, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, + beta, y, incy, stride_y, batch_size, dependencies)); } sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, float *alpha, const float **a, int64_t *lda, const float **x, int64_t *incx, float *beta, float **y, int64_t *incy, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, - group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, + group_count, groupsize, dependencies)); } sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, double *alpha, const double **a, int64_t *lda, const double **x, int64_t *incx, double *beta, double **y, int64_t *incy, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, - group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, + group_count, groupsize, dependencies)); } sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, @@ -543,8 +543,8 @@ sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_ const std::complex **x, int64_t *incx, std::complex *beta, std::complex **y, int64_t *incy, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, - group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, + group_count, groupsize, dependencies)); } sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, @@ -552,24 +552,24 @@ sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_ const std::complex **x, int64_t *incx, std::complex *beta, std::complex **y, int64_t *incy, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, - group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, + group_count, groupsize, dependencies)); } sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, const float *a, int64_t lda, int64_t stride_a, const float *x, int64_t incx, int64_t stride_x, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, - ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, + ldc, stride_c, batch_size, dependencies)); } sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, const double *a, int64_t lda, int64_t stride_a, const double *x, int64_t incx, int64_t stride_x, double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, - ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, + ldc, stride_c, batch_size, dependencies)); } sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, @@ -577,8 +577,8 @@ sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n const std::complex *x, int64_t incx, int64_t stride_x, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, - ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, + ldc, stride_c, batch_size, dependencies)); } sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, @@ -586,40 +586,40 @@ sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n const std::complex *x, int64_t incx, int64_t stride_x, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, - ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, + ldc, stride_c, batch_size, dependencies)); } sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, const float **a, int64_t *lda, const float **x, int64_t *incx, float **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, + groupsize, dependencies)); } sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, const double **a, int64_t *lda, const double **x, int64_t *incx, double **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, + groupsize, dependencies)); } sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, const std::complex **a, int64_t *lda, const std::complex **x, int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, + groupsize, dependencies)); } sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, const std::complex **a, int64_t *lda, const std::complex **x, int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, + groupsize, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -627,8 +627,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i const float *b, int64_t ldb, int64_t stride_b, float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -636,8 +636,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i const double *b, int64_t ldb, int64_t stride_b, double beta, double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -646,8 +646,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i int64_t stride_b, std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -656,8 +656,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i int64_t stride_b, std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -665,8 +665,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i int64_t stride_a, const sycl::half *b, int64_t ldb, int64_t stride_b, sycl::half beta, sycl::half *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -674,8 +674,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i const sycl::half *b, int64_t ldb, int64_t stride_b, float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -692,8 +692,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i const std::int8_t *b, int64_t ldb, int64_t stride_b, float beta, std::int32_t *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, + stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, @@ -701,8 +701,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, const float **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, group_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, + ldc, group_count, group_size, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, @@ -710,8 +710,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, const double **b, int64_t *ldb, double *beta, double **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, group_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, + ldc, group_count, group_size, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, @@ -720,8 +720,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *ldb, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, group_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, + ldc, group_count, group_size, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, @@ -730,8 +730,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *ldb, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *group_size, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, group_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, + ldc, group_count, group_size, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, @@ -739,8 +739,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *lda, const sycl::half **b, int64_t *ldb, sycl::half *beta, sycl::half **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, + ldc, group_count, groupsize, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, @@ -748,8 +748,8 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, const sycl::half **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, + ldc, group_count, groupsize, dependencies)); } sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, @@ -766,24 +766,24 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, const std::int8_t **b, int64_t *ldb, float *beta, std::int32_t **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, + ldc, group_count, groupsize, dependencies)); } sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); } sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); } sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, @@ -791,8 +791,8 @@ sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, tr const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); } sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, @@ -800,24 +800,24 @@ sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, tr const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); } sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, float *alpha, const float **a, int64_t *lda, float **b, int64_t *ldb, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, b, ldb, group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, group_count, groupsize, dependencies)); } sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, double *alpha, const double **a, int64_t *lda, double **b, int64_t *ldb, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, b, ldb, group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, group_count, groupsize, dependencies)); } sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, @@ -825,8 +825,8 @@ sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, const std::complex **a, int64_t *lda, std::complex **b, int64_t *ldb, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, b, ldb, group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, group_count, groupsize, dependencies)); } sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, @@ -834,24 +834,24 @@ sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, const std::complex **a, int64_t *lda, std::complex **b, int64_t *ldb, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, b, ldb, group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, group_count, groupsize, dependencies)); } sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, const float *a, int64_t lda, int64_t stride_a, float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, - ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, + ldc, stride_c, batch_size, dependencies)); } sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, const double *a, int64_t lda, int64_t stride_a, double beta, double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, - ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, + ldc, stride_c, batch_size, dependencies)); } sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, @@ -859,8 +859,8 @@ sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, in int64_t stride_a, std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, - ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, + ldc, stride_c, batch_size, dependencies)); } sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, @@ -868,24 +868,24 @@ sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, in int64_t stride_a, std::complex beta, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, - ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, + ldc, stride_c, batch_size, dependencies)); } sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, int64_t *k, float *alpha, const float **a, int64_t *lda, float *beta, float **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, + group_count, groupsize, dependencies)); } sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, int64_t *k, double *alpha, const double **a, int64_t *lda, double *beta, double **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, + group_count, groupsize, dependencies)); } sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, @@ -893,8 +893,8 @@ sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *lda, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, + group_count, groupsize, dependencies)); } sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, @@ -902,70 +902,70 @@ sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *lda, std::complex *beta, std::complex **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, const std::vector &dependencies) { - return blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - group_count, groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, + group_count, groupsize, dependencies)); } sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + batch_size, dependencies)); } sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + batch_size, dependencies)); } sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + batch_size, dependencies)); } sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, + batch_size, dependencies)); } sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, float *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, const std::vector &dependencies) { - return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, + dependencies)); } sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, double *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, const std::vector &dependencies) { - return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, + dependencies)); } sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, const std::vector &dependencies) { - return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, + dependencies)); } sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, const std::vector &dependencies) { - return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, + dependencies)); } sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, @@ -973,8 +973,8 @@ sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb float beta, const float *b, int64_t ldb, int64_t stride_b, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, - ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, + ldb, stride_b, c, ldc, stride_c, batch_size, dependencies)); } sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, @@ -982,8 +982,8 @@ sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb double beta, const double *b, int64_t ldb, int64_t stride_b, double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, - ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, + ldb, stride_b, c, ldc, stride_c, batch_size, dependencies)); } sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, @@ -992,8 +992,8 @@ sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb const std::complex *b, int64_t ldb, int64_t stride_b, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, - ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, + ldb, stride_b, c, ldc, stride_c, batch_size, dependencies)); } sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, @@ -1002,32 +1002,32 @@ sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb const std::complex *b, int64_t ldb, int64_t stride_b, std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, const std::vector &dependencies) { - return blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, - ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, + ldb, stride_b, c, ldc, stride_c, batch_size, dependencies)); } sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, float* alpha, const float** a, int64_t* lda, float** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, + groupsize, dependencies)); } sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, double* alpha, const double** a, int64_t* lda, double** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, + groupsize, dependencies)); } sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, std::complex* alpha, const std::complex** a, int64_t* lda, std::complex** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, + groupsize, dependencies)); } sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, @@ -1035,38 +1035,38 @@ sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int int64_t* lda, std::complex** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, + groupsize, dependencies)); } sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, float* alpha, float** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, + groupsize, dependencies)); } sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, double* alpha, double** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, + groupsize, dependencies)); } sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, std::complex* alpha, std::complex** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, + groupsize, dependencies)); } sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, std::complex* alpha, std::complex** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - return blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, + groupsize, dependencies)); } diff --git a/src/blas/backends/mkl_common/mkl_blas_backend.hpp b/src/blas/backends/mkl_common/mkl_blas_backend.hpp index 2234772db..30be71602 100644 --- a/src/blas/backends/mkl_common/mkl_blas_backend.hpp +++ b/src/blas/backends/mkl_common/mkl_blas_backend.hpp @@ -25,7 +25,7 @@ #include #include -#include "common_mkl_types_conversion.hpp" +#include "common_onemkl_conversion.hpp" #include "oneapi/math/types.hpp" namespace oneapi { diff --git a/src/blas/backends/mkl_common/mkl_extensions.cxx b/src/blas/backends/mkl_common/mkl_extensions.cxx index faba97517..28500b63c 100644 --- a/src/blas/backends/mkl_common/mkl_extensions.cxx +++ b/src/blas/backends/mkl_common/mkl_extensions.cxx @@ -23,86 +23,86 @@ void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset of int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, int8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, - beta, c, ldc, co); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, + beta, c, ldc, co)); } void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, int8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, - beta, c, ldc, co); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, + beta, c, ldc, co)); } void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, uint8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, - beta, c, ldc, co); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, + beta, c, ldc, co)); } void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, uint8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, - beta, c, ldc, co); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, + beta, c, ldc, co)); } void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { - blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, - ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, + ldc)); } void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, int64_t ldc) { - blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, - ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, + ldc)); } void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, int64_t k, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, - ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, + ldc)); } void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, int64_t k, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb, std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, - ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, + ldc)); } void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { - blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb)); } void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { - blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb)); } void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb) { - blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb)); } void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, sycl::buffer, 1> &b, int64_t ldb) { - blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb)); } void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, @@ -132,48 +132,48 @@ void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, sycl::buffer &ab, int64_t lda, int64_t ldb) { - blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb)); } void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, sycl::buffer &ab, int64_t lda, int64_t ldb) { - blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb)); } void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { - blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb)); } void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { - blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb)); } void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, float beta, sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { - blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc)); } void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, double beta, sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { - blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc)); } void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, std::complex beta, sycl::buffer, 1> &b, int64_t ldb, sycl::buffer, 1> &c, int64_t ldc) { - blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc)); } void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, int64_t lda, std::complex beta, sycl::buffer, 1> &b, int64_t ldb, sycl::buffer, 1> &c, int64_t ldc) { - blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc)); } // USM APIs @@ -183,8 +183,8 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, of int8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, const std::vector &dependencies) { - return blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, - bo, beta, c, ldc, co, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, + bo, beta, c, ldc, co, dependencies)); } sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, @@ -192,8 +192,8 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, of int8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, const std::vector &dependencies) { - return blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, - bo, beta, c, ldc, co, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, + bo, beta, c, ldc, co, dependencies)); } sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, @@ -201,8 +201,8 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, of uint8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, const std::vector &dependencies) { - return blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, - bo, beta, c, ldc, co, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, + bo, beta, c, ldc, co, dependencies)); } sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, @@ -210,24 +210,24 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, of uint8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, const std::vector &dependencies) { - return blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, - bo, beta, c, ldc, co, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, + bo, beta, c, ldc, co, dependencies)); } sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, int64_t k, float alpha, const float *a, int64_t lda, const float *b, int64_t ldb, float beta, float *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies)); } sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, int64_t k, double alpha, const double *a, int64_t lda, const double *b, int64_t ldb, double beta, double *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies)); } sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, @@ -235,8 +235,8 @@ sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transp int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies)); } sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, @@ -244,34 +244,34 @@ sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transp int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies)); } sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, float *b, int64_t ldb, const std::vector &dependencies) { - return blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies)); } sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, double *b, int64_t ldb, const std::vector &dependencies) { - return blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies)); } sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, const std::vector &dependencies) { - return blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies)); } sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, const std::vector &dependencies) { - return blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies)); } sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, @@ -303,41 +303,41 @@ sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, float *ab, int64_t lda, int64_t ldb, const std::vector &dependencies) { - return blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies)); } sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, double *ab, int64_t lda, int64_t ldb, const std::vector &dependencies) { - return blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies)); } sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, const std::vector &dependencies) { - return blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies)); } sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, const std::vector &dependencies) { - return blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies)); } sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, float beta, const float *b, int64_t ldb, float *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies)); } sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, double beta, const double *b, int64_t ldb, double *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies)); } sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -345,8 +345,8 @@ sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int6 std::complex beta, const std::complex *b, int64_t ldb, std::complex *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies)); } sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, @@ -354,6 +354,6 @@ sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int6 std::complex beta, const std::complex *b, int64_t ldb, std::complex *c, int64_t ldc, const std::vector &dependencies) { - return blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies)); } diff --git a/src/blas/backends/mkl_common/mkl_level1.cxx b/src/blas/backends/mkl_common/mkl_level1.cxx index 85ccb0025..b6a985826 100644 --- a/src/blas/backends/mkl_common/mkl_level1.cxx +++ b/src/blas/backends/mkl_common/mkl_level1.cxx @@ -21,625 +21,625 @@ void asum(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - blas_major::asum(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::asum(queue, n, x, incx, result)); } void asum(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - blas_major::asum(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::asum(queue, n, x, incx, result)); } void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - blas_major::asum(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::asum(queue, n, x, incx, result)); } void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - blas_major::asum(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::asum(queue, n, x, incx, result)); } void axpy(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - blas_major::axpy(queue, n, alpha, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy(queue, n, alpha, x, incx, y, incy)); } void axpy(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - blas_major::axpy(queue, n, alpha, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy(queue, n, alpha, x, incx, y, incy)); } void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::axpy(queue, n, alpha, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy(queue, n, alpha, x, incx, y, incy)); } void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::axpy(queue, n, alpha, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy(queue, n, alpha, x, incx, y, incy)); } void axpby(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy)); } void axpby(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy)); } void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy)); } void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy)); } void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - blas_major::copy(queue, n, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy(queue, n, x, incx, y, incy)); } void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - blas_major::copy(queue, n, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy(queue, n, x, incx, y, incy)); } void copy(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::copy(queue, n, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy(queue, n, x, incx, y, incy)); } void copy(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::copy(queue, n, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy(queue, n, x, incx, y, incy)); } void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - blas_major::dot(queue, n, x, incx, y, incy, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dot(queue, n, x, incx, y, incy, result)); } void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - blas_major::dot(queue, n, x, incx, y, incy, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dot(queue, n, x, incx, y, incy, result)); } void sdsdot(sycl::queue &queue, std::int64_t n, float sb, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - blas_major::sdsdot(queue, n, sb, x, incx, y, incy, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::sdsdot(queue, n, sb, x, incx, y, incy, result)); } void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { - blas_major::dot(queue, n, x, incx, y, incy, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dot(queue, n, x, incx, y, incy, result)); } void dotc(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - blas_major::dotc(queue, n, x, incx, y, incy, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dotc(queue, n, x, incx, y, incy, result)); } void dotc(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - blas_major::dotc(queue, n, x, incx, y, incy, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dotc(queue, n, x, incx, y, incy, result)); } void dotu(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - blas_major::dotu(queue, n, x, incx, y, incy, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dotu(queue, n, x, incx, y, incy, result)); } void dotu(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &result) { - blas_major::dotu(queue, n, x, incx, y, incy, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dotu(queue, n, x, incx, y, incy, result)); } void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - blas_major::nrm2(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::nrm2(queue, n, x, incx, result)); } void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - blas_major::nrm2(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::nrm2(queue, n, x, incx, result)); } void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - blas_major::nrm2(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::nrm2(queue, n, x, incx, result)); } void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - blas_major::nrm2(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::nrm2(queue, n, x, incx, result)); } void rot(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { - blas_major::rot(queue, n, x, incx, y, incy, c, s); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rot(queue, n, x, incx, y, incy, c, s)); } void rot(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { - blas_major::rot(queue, n, x, incx, y, incy, c, s); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rot(queue, n, x, incx, y, incy, c, s)); } void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { - blas_major::rot(queue, n, x, incx, y, incy, c, s); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rot(queue, n, x, incx, y, incy, c, s)); } void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { - blas_major::rot(queue, n, x, incx, y, incy, c, s); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rot(queue, n, x, incx, y, incy, c, s)); } void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - blas_major::rotg(queue, a, b, c, s); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotg(queue, a, b, c, s)); } void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { - blas_major::rotg(queue, a, b, c, s); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotg(queue, a, b, c, s)); } void rotg(sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - blas_major::rotg(queue, a, b, c, s); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotg(queue, a, b, c, s)); } void rotg(sycl::queue &queue, sycl::buffer, 1> &a, sycl::buffer, 1> &b, sycl::buffer &c, sycl::buffer, 1> &s) { - blas_major::rotg(queue, a, b, c, s); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotg(queue, a, b, c, s)); } void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - blas_major::rotm(queue, n, x, incx, y, incy, param); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotm(queue, n, x, incx, y, incy, param)); } void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { - blas_major::rotm(queue, n, x, incx, y, incy, param); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotm(queue, n, x, incx, y, incy, param)); } void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, float y1, sycl::buffer ¶m) { - blas_major::rotmg(queue, d1, d2, x1, y1, param); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotmg(queue, d1, d2, x1, y1, param)); } void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, sycl::buffer &x1, double y1, sycl::buffer ¶m) { - blas_major::rotmg(queue, d1, d2, x1, y1, param); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotmg(queue, d1, d2, x1, y1, param)); } void scal(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx) { - blas_major::scal(queue, n, alpha, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } void scal(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx) { - blas_major::scal(queue, n, alpha, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::scal(queue, n, alpha, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::scal(queue, n, alpha, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } void scal(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::scal(queue, n, alpha, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } void scal(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::scal(queue, n, alpha, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - blas_major::swap(queue, n, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::swap(queue, n, x, incx, y, incy)); } void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy) { - blas_major::swap(queue, n, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::swap(queue, n, x, incx, y, incy)); } void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::swap(queue, n, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::swap(queue, n, x, incx, y, incy)); } void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::swap(queue, n, x, incx, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::swap(queue, n, x, incx, y, incy)); } void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - blas_major::iamax(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamax(queue, n, x, incx, result)); } void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - blas_major::iamax(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamax(queue, n, x, incx, result)); } void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - blas_major::iamax(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamax(queue, n, x, incx, result)); } void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - blas_major::iamax(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamax(queue, n, x, incx, result)); } void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - blas_major::iamin(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamin(queue, n, x, incx, result)); } void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { - blas_major::iamin(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamin(queue, n, x, incx, result)); } void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - blas_major::iamin(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamin(queue, n, x, incx, result)); } void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer &result) { - blas_major::iamin(queue, n, x, incx, result); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamin(queue, n, x, incx, result)); } // USM APIs sycl::event asum(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - return blas_major::asum(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::asum(queue, n, x, incx, result, dependencies)); } sycl::event asum(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - return blas_major::asum(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::asum(queue, n, x, incx, result, dependencies)); } sycl::event asum(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - return blas_major::asum(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::asum(queue, n, x, incx, result, dependencies)); } sycl::event asum(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - return blas_major::asum(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::asum(queue, n, x, incx, result, dependencies)); } sycl::event axpy(sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); } sycl::event axpy(sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); } sycl::event axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); } sycl::event axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); } sycl::event axpby(sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); } sycl::event axpby(sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); } sycl::event axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); } sycl::event axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); } sycl::event copy(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::copy(queue, n, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy(queue, n, x, incx, y, incy, dependencies)); } sycl::event copy(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::copy(queue, n, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy(queue, n, x, incx, y, incy, dependencies)); } sycl::event copy(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::copy(queue, n, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy(queue, n, x, incx, y, incy, dependencies)); } sycl::event copy(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::copy(queue, n, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy(queue, n, x, incx, y, incy, dependencies)); } sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - return blas_major::dot(queue, n, x, incx, y, incy, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dot(queue, n, x, incx, y, incy, result, dependencies)); } sycl::event dot(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *result, const std::vector &dependencies) { - return blas_major::dot(queue, n, x, incx, y, incy, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dot(queue, n, x, incx, y, incy, result, dependencies)); } sycl::event sdsdot(sycl::queue &queue, std::int64_t n, float sb, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *result, const std::vector &dependencies) { - return blas_major::sdsdot(queue, n, sb, x, incx, y, incy, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::sdsdot(queue, n, sb, x, incx, y, incy, result, dependencies)); } sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, const float *y, std::int64_t incy, double *result, const std::vector &dependencies) { - return blas_major::dot(queue, n, x, incx, y, incy, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dot(queue, n, x, incx, y, incy, result, dependencies)); } sycl::event dotc(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - return blas_major::dotc(queue, n, x, incx, y, incy, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dotc(queue, n, x, incx, y, incy, result, dependencies)); } sycl::event dotc(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - return blas_major::dotc(queue, n, x, incx, y, incy, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dotc(queue, n, x, incx, y, incy, result, dependencies)); } sycl::event dotu(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - return blas_major::dotu(queue, n, x, incx, y, incy, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dotu(queue, n, x, incx, y, incy, result, dependencies)); } sycl::event dotu(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *result, const std::vector &dependencies) { - return blas_major::dotu(queue, n, x, incx, y, incy, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dotu(queue, n, x, incx, y, incy, result, dependencies)); } sycl::event nrm2(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, float *result, const std::vector &dependencies) { - return blas_major::nrm2(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::nrm2(queue, n, x, incx, result, dependencies)); } sycl::event nrm2(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, double *result, const std::vector &dependencies) { - return blas_major::nrm2(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::nrm2(queue, n, x, incx, result, dependencies)); } sycl::event nrm2(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *result, const std::vector &dependencies) { - return blas_major::nrm2(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::nrm2(queue, n, x, incx, result, dependencies)); } sycl::event nrm2(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *result, const std::vector &dependencies) { - return blas_major::nrm2(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::nrm2(queue, n, x, incx, result, dependencies)); } sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - return blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies)); } sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - return blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies)); } sycl::event rot(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float c, float s, const std::vector &dependencies) { - return blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies)); } sycl::event rot(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double c, double s, const std::vector &dependencies) { - return blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies)); } sycl::event rotg(sycl::queue &queue, float *a, float *b, float *c, float *s, const std::vector &dependencies) { - return blas_major::rotg(queue, a, b, c, s, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotg(queue, a, b, c, s, dependencies)); } sycl::event rotg(sycl::queue &queue, double *a, double *b, double *c, double *s, const std::vector &dependencies) { - return blas_major::rotg(queue, a, b, c, s, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotg(queue, a, b, c, s, dependencies)); } sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, float *c, std::complex *s, const std::vector &dependencies) { - return blas_major::rotg(queue, a, b, c, s, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotg(queue, a, b, c, s, dependencies)); } sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, double *c, std::complex *s, const std::vector &dependencies) { - return blas_major::rotg(queue, a, b, c, s, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotg(queue, a, b, c, s, dependencies)); } sycl::event rotm(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, float *param, const std::vector &dependencies) { - return blas_major::rotm(queue, n, x, incx, y, incy, param, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotm(queue, n, x, incx, y, incy, param, dependencies)); } sycl::event rotm(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, double *param, const std::vector &dependencies) { - return blas_major::rotm(queue, n, x, incx, y, incy, param, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotm(queue, n, x, incx, y, incy, param, dependencies)); } sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, float y1, float *param, const std::vector &dependencies) { - return blas_major::rotmg(queue, d1, d2, x1, y1, param, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotmg(queue, d1, d2, x1, y1, param, dependencies)); } sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, double y1, double *param, const std::vector &dependencies) { - return blas_major::rotmg(queue, d1, d2, x1, y1, param, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotmg(queue, d1, d2, x1, y1, param, dependencies)); } sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::scal(queue, n, alpha, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::scal(queue, n, alpha, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } sycl::event scal(sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::scal(queue, n, alpha, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } sycl::event scal(sycl::queue &queue, std::int64_t n, std::complex alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::scal(queue, n, alpha, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::scal(queue, n, alpha, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::scal(queue, n, alpha, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } sycl::event swap(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::swap(queue, n, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::swap(queue, n, x, incx, y, incy, dependencies)); } sycl::event swap(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::swap(queue, n, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::swap(queue, n, x, incx, y, incy, dependencies)); } sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::swap(queue, n, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::swap(queue, n, x, incx, y, incy, dependencies)); } sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::swap(queue, n, x, incx, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::swap(queue, n, x, incx, y, incy, dependencies)); } sycl::event iamax(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - return blas_major::iamax(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamax(queue, n, x, incx, result, dependencies)); } sycl::event iamax(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - return blas_major::iamax(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamax(queue, n, x, incx, result, dependencies)); } sycl::event iamax(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - return blas_major::iamax(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamax(queue, n, x, incx, result, dependencies)); } sycl::event iamax(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - return blas_major::iamax(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamax(queue, n, x, incx, result, dependencies)); } sycl::event iamin(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - return blas_major::iamin(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamin(queue, n, x, incx, result, dependencies)); } sycl::event iamin(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - return blas_major::iamin(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamin(queue, n, x, incx, result, dependencies)); } sycl::event iamin(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - return blas_major::iamin(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamin(queue, n, x, incx, result, dependencies)); } sycl::event iamin(sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, std::int64_t *result, const std::vector &dependencies) { - return blas_major::iamin(queue, n, x, incx, result, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamin(queue, n, x, incx, result, dependencies)); } diff --git a/src/blas/backends/mkl_common/mkl_level2.cxx b/src/blas/backends/mkl_common/mkl_level2.cxx index 68f3b3713..f2d0a5561 100644 --- a/src/blas/backends/mkl_common/mkl_level2.cxx +++ b/src/blas/backends/mkl_common/mkl_level2.cxx @@ -22,409 +22,409 @@ void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy)); } void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy)); } void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy)); } void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy)); } void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy)); } void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy)); } void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy)); } void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy)); } void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda)); } void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda)); } void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda)); } void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda)); } void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda)); } void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda)); } void hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy)); } void hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy)); } void hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy)); } void hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy)); } void her(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); } void her(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); } void her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); } void her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a, std::int64_t lda) { - blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); } void hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy)); } void hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy)); } void hpr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); } void hpr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &a) { - blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); } void hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); } void hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, sycl::buffer, 1> &a) { - blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); } void sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy)); } void sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy)); } void spmv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy)); } void spmv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy)); } void spr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); } void spr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { - blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); } void spr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); } void spr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { - blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); } void symv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy)); } void symv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy)); } void syr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); } void syr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); } void syr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); } void syr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, std::int64_t lda) { - blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); } void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); } void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); } void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); } void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); } void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); } void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); } void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); } void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); } void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } // USM APIs @@ -432,15 +432,15 @@ void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_ sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, @@ -448,8 +448,8 @@ sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64 const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, @@ -457,24 +457,24 @@ sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64 const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, @@ -482,8 +482,8 @@ sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64 const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, @@ -491,48 +491,48 @@ sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64 const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } sycl::event hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, @@ -540,7 +540,7 @@ sycl::event hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies)); } sycl::event hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, @@ -548,315 +548,315 @@ sycl::event hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies)); } sycl::event hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies)); } sycl::event hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies)); } sycl::event her(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies)); } sycl::event her(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies)); } sycl::event her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies)); } sycl::event her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies)); } sycl::event hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies)); } sycl::event hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *a, const std::complex *x, std::int64_t incx, std::complex beta, std::complex *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies)); } sycl::event hpr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - return blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); } sycl::event hpr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const std::complex *x, std::int64_t incx, std::complex *a, const std::vector &dependencies) { - return blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); } sycl::event hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - return blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies)); } sycl::event hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, const std::complex *x, std::int64_t incx, const std::complex *y, std::int64_t incy, std::complex *a, const std::vector &dependencies) { - return blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies)); } sycl::event sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies)); } sycl::event sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies)); } sycl::event spmv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *a, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies)); } sycl::event spmv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *a, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies)); } sycl::event spr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, const std::vector &dependencies) { - return blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); } sycl::event spr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, const std::vector &dependencies) { - return blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); } sycl::event spr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, const std::vector &dependencies) { - return blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies)); } sycl::event spr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, const std::vector &dependencies) { - return blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies)); } sycl::event symv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies)); } sycl::event symv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, const std::vector &dependencies) { - return blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies)); } sycl::event syr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies)); } sycl::event syr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies)); } sycl::event syr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies)); } sycl::event syr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, const std::vector &dependencies) { - return blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies)); } sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); } sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); } sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); } sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); } sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const float *a, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); } sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const double *a, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); } sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); } sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); } sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); } sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); } sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); } sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); } sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); } sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); } sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); } sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, const std::complex *a, std::int64_t lda, std::complex *x, std::int64_t incx, const std::vector &dependencies) { - return blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); } diff --git a/src/blas/backends/mkl_common/mkl_level3.cxx b/src/blas/backends/mkl_common/mkl_level3.cxx index 68662f47f..62d26128e 100644 --- a/src/blas/backends/mkl_common/mkl_level3.cxx +++ b/src/blas/backends/mkl_common/mkl_level3.cxx @@ -23,225 +23,225 @@ void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); } void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); } void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); } void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); } void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); } void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); } void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); } void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); } void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); } void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); } void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); } void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); } void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, double beta, sycl::buffer, 1> &c, std::int64_t ldc) { - blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, + ldb)); } void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, + ldb)); } void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, + ldb)); } void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, + ldb)); } void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, + ldb)); } void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, + ldb)); } void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, + ldb)); } void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, + ldb)); } // USM APIs @@ -250,16 +250,16 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::in std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, @@ -267,8 +267,8 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::in const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, @@ -276,48 +276,48 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::in const std::complex *a, std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, sycl::half *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies)); } sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies)); } sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, @@ -325,8 +325,8 @@ sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies)); } sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, @@ -334,8 +334,8 @@ sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies)); } sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, @@ -343,8 +343,8 @@ sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies)); } sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, @@ -352,70 +352,70 @@ sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies)); } sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, + dependencies)); } sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, + dependencies)); } sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, + dependencies)); } sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, + dependencies)); } sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, + dependencies)); } sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, + dependencies)); } sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, @@ -423,8 +423,8 @@ sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::in std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, @@ -432,8 +432,8 @@ sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::in std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, @@ -441,8 +441,8 @@ sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::in std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, @@ -450,70 +450,70 @@ sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::in std::int64_t lda, const std::complex *b, std::int64_t ldb, double beta, std::complex *c, std::int64_t ldc, const std::vector &dependencies) { - return blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); } sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + b, ldb, dependencies)); } sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + b, ldb, dependencies)); } sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + b, ldb, dependencies)); } sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + b, ldb, dependencies)); } sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + b, ldb, dependencies)); } sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + b, ldb, dependencies)); } sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + b, ldb, dependencies)); } sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, const std::vector &dependencies) { - return blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, + b, ldb, dependencies)); } diff --git a/src/dft/backends/mklcpu/CMakeLists.txt b/src/dft/backends/mklcpu/CMakeLists.txt index e79308743..d79327087 100644 --- a/src/dft/backends/mklcpu/CMakeLists.txt +++ b/src/dft/backends/mklcpu/CMakeLists.txt @@ -42,6 +42,7 @@ target_include_directories(${LIB_NAME} target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src + ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) diff --git a/src/dft/backends/mklgpu/CMakeLists.txt b/src/dft/backends/mklgpu/CMakeLists.txt index 2d557ffa2..a60e7b5eb 100644 --- a/src/dft/backends/mklgpu/CMakeLists.txt +++ b/src/dft/backends/mklgpu/CMakeLists.txt @@ -47,6 +47,7 @@ target_compile_options(${LIB_OBJ} target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src + ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) diff --git a/src/dft/backends/mklgpu/backward.cpp b/src/dft/backends/mklgpu/backward.cpp index 0c248003f..ef0bf4836 100644 --- a/src/dft/backends/mklgpu/backward.cpp +++ b/src/dft/backends/mklgpu/backward.cpp @@ -28,6 +28,7 @@ #include "oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp" #include "oneapi/math/dft/detail/descriptor_impl.hpp" +#include "common_onemkl_conversion.hpp" #include "mklgpu_helpers.hpp" #include "mkl_version.h" @@ -64,7 +65,7 @@ inline auto compute_backward(dft::detail::descriptor &desc, ArgTs &&. // The MKLGPU backend's interface contains fewer function signatures than in this // open-source library. Consequently, it is not required to forward template arguments // to resolve to the correct function. - return oneapi::mkl::dft::compute_backward(*mklgpu_desc, std::forward(args)...); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::dft::compute_backward(*mklgpu_desc, std::forward(args)...)); } /// Throw an math::invalid_argument if the runtime param in the descriptor does not match diff --git a/src/dft/backends/mklgpu/commit.cpp b/src/dft/backends/mklgpu/commit.cpp index f7e1b2de8..a499ca163 100644 --- a/src/dft/backends/mklgpu/commit.cpp +++ b/src/dft/backends/mklgpu/commit.cpp @@ -32,6 +32,7 @@ #include "oneapi/math/dft/detail/descriptor_impl.hpp" #include "oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp" +#include "common_onemkl_conversion.hpp" #include "dft/backends/mklgpu/mklgpu_helpers.hpp" #include "../stride_helper.hpp" @@ -101,41 +102,19 @@ class mklgpu_commit final : public dft::detail::commit_impl { // A separate descriptor for each direction may not be required. bool one_descriptor = (stride_choice == dft::detail::stride_api::FB_STRIDES) || (config_values.input_strides == config_values.output_strides); - bool forward_good = true; // Make sure that second is always pointing to something new if this is a recommit. handle.second = handle.first; // Generate forward DFT descriptor. If using FWD/BWD_STRIDES API, only // one descriptor is needed. set_value(*handle.first, config_values, true, stride_choice); - try { - handle.first->commit(this->get_queue()); - } - catch (const std::exception& mkl_exception) { - // Catching the real Intel oneMKL exception causes headaches with naming - forward_good = false; - if (one_descriptor) { - throw math::exception("dft/backends/mklgpu" - "commit", - mkl_exception.what()); - } - } + RETHROW_ONEMKL_EXCEPTIONS(handle.first->commit(this->get_queue())); // Generate backward DFT descriptor only if required. if (!one_descriptor) { handle.second = std::make_shared(config_values.dimensions); set_value(*handle.second, config_values, false, stride_choice); - try { - handle.second->commit(this->get_queue()); - } - catch (const std::exception& mkl_exception) { - // Catching the real Intel oneMKL exception causes headaches with naming. - if (!forward_good) { - throw math::exception("dft/backends/mklgpu" - "commit", - mkl_exception.what()); - } - } + RETHROW_ONEMKL_EXCEPTIONS(handle.second->commit(this->get_queue())); } } @@ -147,17 +126,17 @@ class mklgpu_commit final : public dft::detail::commit_impl { virtual void set_workspace(scalar_type* usm_workspace) override { this->external_workspace_helper_.set_workspace_throw(*this, usm_workspace); - handle.first->set_workspace(usm_workspace); + RETHROW_ONEMKL_EXCEPTIONS(handle.first->set_workspace(usm_workspace)); if (handle.first != handle.second) { - handle.second->set_workspace(usm_workspace); + RETHROW_ONEMKL_EXCEPTIONS(handle.second->set_workspace(usm_workspace)); } } virtual void set_workspace(sycl::buffer& buffer_workspace) override { this->external_workspace_helper_.set_workspace_throw(*this, buffer_workspace); - handle.first->set_workspace(buffer_workspace); + RETHROW_ONEMKL_EXCEPTIONS(handle.first->set_workspace(buffer_workspace)); if (handle.first != handle.second) { - handle.second->set_workspace(buffer_workspace); + RETHROW_ONEMKL_EXCEPTIONS(handle.second->set_workspace(buffer_workspace)); } } diff --git a/src/dft/backends/mklgpu/forward.cpp b/src/dft/backends/mklgpu/forward.cpp index 209dc9295..dd6325459 100644 --- a/src/dft/backends/mklgpu/forward.cpp +++ b/src/dft/backends/mklgpu/forward.cpp @@ -29,6 +29,7 @@ #include "oneapi/math/dft/detail/mklgpu/onemath_dft_mklgpu.hpp" #include "oneapi/math/dft/detail/descriptor_impl.hpp" +#include "common_onemkl_conversion.hpp" #include "mklgpu_helpers.hpp" #include "mkl_version.h" @@ -70,7 +71,7 @@ inline auto compute_forward(dft::detail::descriptor &desc, ArgTs &&.. // The MKLGPU backend's interface contains fewer function signatures than in this // open-source library. Consequently, it is not required to forward template arguments // to resolve to the correct function. - return oneapi::mkl::dft::compute_forward(*mklgpu_desc, std::forward(args)...); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::dft::compute_forward(*mklgpu_desc, std::forward(args)...)); } /// Throw an math::invalid_argument if the runtime param in the descriptor does not match diff --git a/src/include/common_mkl_types_conversion.hpp b/src/include/common_onemkl_conversion.hpp similarity index 70% rename from src/include/common_mkl_types_conversion.hpp rename to src/include/common_onemkl_conversion.hpp index f2bc45c98..a54669e71 100644 --- a/src/include/common_mkl_types_conversion.hpp +++ b/src/include/common_onemkl_conversion.hpp @@ -17,15 +17,17 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -#ifndef _ONEMATH_SRC_INCLUDE_COMMON_MKL_TYPES_CONVERSION_HPP_ -#define _ONEMATH_SRC_INCLUDE_COMMON_MKL_TYPES_CONVERSION_HPP_ +#ifndef _ONEMATH_SRC_INCLUDE_COMMON_ONEMKL_TYPES_CONVERSION_HPP_ +#define _ONEMATH_SRC_INCLUDE_COMMON_ONEMKL_TYPES_CONVERSION_HPP_ // The file is used to convert oneMath types to Intel(R) oneMKL types for all the common types shared across domains. // The file assumes that the common types are identical between the 2 libraries, except for their namespace. +#include #include #include "oneapi/math/types.hpp" +#include "oneapi/math/exceptions.hpp" namespace oneapi { namespace math { @@ -65,8 +67,39 @@ inline auto get_onemkl_rangev(oneapi::math::rangev param) { return *reinterpret_ inline auto get_onemkl_order(oneapi::math::order param) { return *reinterpret_cast(¶m); } +// Rethrow Intel(R) oneMKL exceptions as oneMath exceptions +#define RETHROW_ONEMKL_EXCEPTIONS(EXPRESSION) \ +do { \ + try { \ + EXPRESSION; \ + } catch(const oneapi::mkl::unsupported_device& e) { \ + throw unsupported_device(e.what()); \ + } catch(const oneapi::mkl::host_bad_alloc& e) { \ + throw host_bad_alloc(e.what()); \ + } catch(const oneapi::mkl::device_bad_alloc& e) { \ + throw device_bad_alloc(e.what()); \ + } catch(const oneapi::mkl::unimplemented& e) { \ + throw unimplemented(e.what()); \ + } catch(const oneapi::mkl::invalid_argument& e) { \ + throw invalid_argument(e.what()); \ + } catch(const oneapi::mkl::uninitialized& e) { \ + throw uninitialized(e.what()); \ + } catch(const oneapi::mkl::computation_error& e) { \ + throw computation_error(e.what()); \ + } catch(const oneapi::mkl::batch_error& e) { \ + throw batch_error(e.what()); \ + } catch(const oneapi::mkl::exception& e) { \ + throw exception(e.what()); \ + } \ +} while (0) + +#define RETHROW_ONEMKL_EXCEPTIONS_RET(EXPRESSION) \ +do { \ +RETHROW_ONEMKL_EXCEPTIONS(return EXPRESSION); \ +} while(0) + } // namespace detail } // namespace math } // namespace oneapi -#endif // _ONEMATH_SRC_INCLUDE_COMMON_MKL_TYPES_CONVERSION_HPP_ +#endif // _ONEMATH_SRC_INCLUDE_COMMON_ONEMKL_TYPES_CONVERSION_HPP_ diff --git a/src/lapack/backends/mkl_common/mkl_lapack.cxx b/src/lapack/backends/mkl_common/mkl_lapack.cxx index 47dbc97c0..a0c85f82b 100644 --- a/src/lapack/backends/mkl_common/mkl_lapack.cxx +++ b/src/lapack/backends/mkl_common/mkl_lapack.cxx @@ -21,504 +21,504 @@ void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, + scratchpad_size)); } void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, + scratchpad_size)); } void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, + scratchpad_size)); } void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tauq, sycl::buffer> &taup, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, + scratchpad_size)); } void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); } void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); } void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); } void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); } void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); } void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); } void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); } void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); } void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + scratchpad_size)); } void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + scratchpad_size)); } void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + scratchpad_size)); } void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + scratchpad_size)); } void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, + scratchpad_size)); } void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, + scratchpad_size)); } void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, + scratchpad_size)); } void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, + scratchpad_size)); } void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size)); } void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size)); } void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, + scratchpad_size)); } void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, + scratchpad_size)); } void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); } void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); } void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); } void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); } void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size)); } void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size)); } void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); } void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); } void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); } void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); } void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); } void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); } void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); } void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); } void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); } void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); } void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); } void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); } void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); } void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); } void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); } void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); } void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size)); } void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size)); } void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, + scratchpad_size)); } void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, + scratchpad_size)); } void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); } void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); } void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); } void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); } void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); } void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); } void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size)); } void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size)); } void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size)); } void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size)); } void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size)); } void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size)); } void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); } void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); } void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); } void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); } void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, @@ -526,8 +526,8 @@ void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, @@ -535,648 +535,648 @@ void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, sycl::buffer> &tau, sycl::buffer> &c, std::int64_t ldc, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, + scratchpad_size, dependencies)); } sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tauq, double *taup, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, + scratchpad_size, dependencies)); } sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tauq, float *taup, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, + scratchpad_size, dependencies)); } sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tauq, std::complex *taup, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, + scratchpad_size, dependencies)); } sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + scratchpad_size, dependencies)); } sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + scratchpad_size, dependencies)); } sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + scratchpad_size, dependencies)); } sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, + scratchpad_size, dependencies)); } sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, + scratchpad, scratchpad_size, dependencies)); } sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, + scratchpad, scratchpad_size, dependencies)); } sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, + scratchpad, scratchpad_size, dependencies)); } sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, double *s, std::complex *u, std::int64_t ldu, std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, + scratchpad, scratchpad_size, dependencies)); } sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, + scratchpad_size, dependencies)); } sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, + scratchpad_size, dependencies)); } sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, float *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, + scratchpad_size, dependencies)); } sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, double *w, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, + scratchpad_size, dependencies)); } sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, + scratchpad_size, dependencies)); } sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, + scratchpad_size, dependencies)); } sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); } sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); } sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, + dependencies)); } sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, + dependencies)); } sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, + dependencies)); } sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, + dependencies)); } sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, + dependencies)); } sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, + dependencies)); } sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, + dependencies)); } sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, + dependencies)); } sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies)); } sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies)); } sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies)); } sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies)); } sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, + scratchpad_size, dependencies)); } sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, + scratchpad_size, dependencies)); } sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, + scratchpad_size, dependencies)); } sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, + scratchpad_size, dependencies)); } sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, + scratchpad_size, dependencies)); } sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, + scratchpad_size, dependencies)); } sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, + dependencies)); } sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, + scratchpad, scratchpad_size, dependencies)); } sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, + scratchpad, scratchpad_size, dependencies)); } sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, + scratchpad, scratchpad_size, dependencies)); } sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, std::complex *b, std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, + scratchpad, scratchpad_size, dependencies)); } sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); } sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); } sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, + dependencies)); } sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::complex *tau, std::complex *c, std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, + scratchpad, scratchpad_size, dependencies)); } void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size)); } void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size)); } void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size)); } void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size)); } void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size)); } void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size)); } void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size)); } void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size)); } void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, - b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, + b, ldb, stride_b, batch_size, scratchpad, scratchpad_size)); } void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, @@ -1184,8 +1184,8 @@ void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, - b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, + b, ldb, stride_b, batch_size, scratchpad, scratchpad_size)); } void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, @@ -1193,8 +1193,8 @@ void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, - b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, + b, ldb, stride_b, batch_size, scratchpad, scratchpad_size)); } void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, @@ -1202,323 +1202,323 @@ void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, - b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, + b, ldb, stride_b, batch_size, scratchpad, scratchpad_size)); } void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size)); } void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size)); } void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size)); } void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size)); } void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size)); } void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size)); } void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, + scratchpad_size)); } void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, + scratchpad_size)); } void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, + scratchpad_size)); } void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, - scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, + scratchpad_size)); } void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size)); } void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size)); } void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size)); } void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size)); } void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size)); } void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, sycl::buffer> &tau, std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - ::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size)); } sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, @@ -1526,9 +1526,9 @@ sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std:: std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies)); } sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, @@ -1536,9 +1536,9 @@ sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std:: std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies)); } sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, @@ -1547,9 +1547,9 @@ sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std:: std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies)); } sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, @@ -1558,27 +1558,27 @@ sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std:: std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies)); } sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, @@ -1586,9 +1586,9 @@ sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std: std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, @@ -1596,119 +1596,119 @@ sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std: std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, std::int64_t stride_a, float *tau, std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, std::int64_t stride_a, double *tau, std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, float **a, std::int64_t *lda, float **tau, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, double **a, std::int64_t *lda, double **tau, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, + scratchpad, scratchpad_size, dependencies)); } sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, + scratchpad, scratchpad_size, dependencies)); } sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, + scratchpad, scratchpad_size, dependencies)); } sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, + scratchpad, scratchpad_size, dependencies)); } sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::complex **a, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, @@ -1716,9 +1716,9 @@ sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_ std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, @@ -1726,27 +1726,27 @@ sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_ std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, @@ -1754,9 +1754,9 @@ sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64 std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, @@ -1764,119 +1764,119 @@ sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64 std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, - dependencies); + dependencies)); } sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex **a, std::int64_t *lda, std::complex **tau, std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, const std::vector &dependencies) { - return ::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } template <> std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::gebrd_scratchpad_size(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd_scratchpad_size(queue, m, n, lda)); } template <> std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::gebrd_scratchpad_size(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd_scratchpad_size(queue, m, n, lda)); } template <> std::int64_t gebrd_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::gebrd_scratchpad_size>(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd_scratchpad_size>(queue, m, n, lda)); } template <> std::int64_t gebrd_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::gebrd_scratchpad_size>(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd_scratchpad_size>(queue, m, n, lda)); } template <> std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::gerqf_scratchpad_size(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf_scratchpad_size(queue, m, n, lda)); } template <> std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::gerqf_scratchpad_size(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf_scratchpad_size(queue, m, n, lda)); } template <> std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::gerqf_scratchpad_size>(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf_scratchpad_size>(queue, m, n, lda)); } template <> std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::gerqf_scratchpad_size>(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf_scratchpad_size>(queue, m, n, lda)); } template <> std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::geqrf_scratchpad_size(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_scratchpad_size(queue, m, n, lda)); } template <> std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::geqrf_scratchpad_size(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_scratchpad_size(queue, m, n, lda)); } template <> std::int64_t geqrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::geqrf_scratchpad_size>(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_scratchpad_size>(queue, m, n, lda)); } template <> std::int64_t geqrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::geqrf_scratchpad_size>(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_scratchpad_size>(queue, m, n, lda)); } template <> std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return ::oneapi::mkl::lapack::gesvd_scratchpad_size(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, - ldvt); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd_scratchpad_size(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, + ldvt)); } template <> std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return ::oneapi::mkl::lapack::gesvd_scratchpad_size(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, - ldvt); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd_scratchpad_size(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, + ldvt)); } template <> std::int64_t gesvd_scratchpad_size>(sycl::queue &queue, @@ -1884,8 +1884,8 @@ std::int64_t gesvd_scratchpad_size>(sycl::queue &queue, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return ::oneapi::mkl::lapack::gesvd_scratchpad_size>(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, - n, lda, ldu, ldvt); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd_scratchpad_size>(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, + n, lda, ldu, ldvt)); } template <> std::int64_t gesvd_scratchpad_size>(sycl::queue &queue, @@ -1893,342 +1893,342 @@ std::int64_t gesvd_scratchpad_size>(sycl::queue &queue, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - return ::oneapi::mkl::lapack::gesvd_scratchpad_size>(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, - n, lda, ldu, ldvt); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd_scratchpad_size>(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, + n, lda, ldu, ldvt)); } template <> std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::getrf_scratchpad_size(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_scratchpad_size(queue, m, n, lda)); } template <> std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::getrf_scratchpad_size(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_scratchpad_size(queue, m, n, lda)); } template <> std::int64_t getrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::getrf_scratchpad_size>(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_scratchpad_size>(queue, m, n, lda)); } template <> std::int64_t getrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::getrf_scratchpad_size>(queue, m, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_scratchpad_size>(queue, m, n, lda)); } template <> std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::getri_scratchpad_size(queue, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_scratchpad_size(queue, n, lda)); } template <> std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::getri_scratchpad_size(queue, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_scratchpad_size(queue, n, lda)); } template <> std::int64_t getri_scratchpad_size>(sycl::queue &queue, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::getri_scratchpad_size>(queue, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_scratchpad_size>(queue, n, lda)); } template <> std::int64_t getri_scratchpad_size>(sycl::queue &queue, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::getri_scratchpad_size>(queue, n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_scratchpad_size>(queue, n, lda)); } template <> std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::getrs_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb)); } template <> std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::getrs_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb)); } template <> std::int64_t getrs_scratchpad_size>(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::getrs_scratchpad_size>(queue, detail::get_onemkl_transpose(trans), n, nrhs, - lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_scratchpad_size>(queue, detail::get_onemkl_transpose(trans), n, nrhs, + lda, ldb)); } template <> std::int64_t getrs_scratchpad_size>(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::getrs_scratchpad_size>(queue, detail::get_onemkl_transpose(trans), n, nrhs, - lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_scratchpad_size>(queue, detail::get_onemkl_transpose(trans), n, nrhs, + lda, ldb)); } template <> std::int64_t heevd_scratchpad_size>(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::heevd_scratchpad_size>(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, - lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::heevd_scratchpad_size>(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, + lda)); } template <> std::int64_t heevd_scratchpad_size>(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::heevd_scratchpad_size>(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, - lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::heevd_scratchpad_size>(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, + lda)); } template <> std::int64_t hegvd_scratchpad_size>(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::hegvd_scratchpad_size>(queue, itype, detail::get_onemkl_job(jobz), - detail::get_onemkl_uplo(uplo), n, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hegvd_scratchpad_size>(queue, itype, detail::get_onemkl_job(jobz), + detail::get_onemkl_uplo(uplo), n, lda, ldb)); } template <> std::int64_t hegvd_scratchpad_size>(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::hegvd_scratchpad_size>(queue, itype, detail::get_onemkl_job(jobz), - detail::get_onemkl_uplo(uplo), n, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hegvd_scratchpad_size>(queue, itype, detail::get_onemkl_job(jobz), + detail::get_onemkl_uplo(uplo), n, lda, ldb)); } template <> std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::hetrd_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrd_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::hetrd_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrd_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::hetrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::hetrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::orgbr_scratchpad_size(queue, detail::get_onemkl_generate(vect), m, n, k, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgbr_scratchpad_size(queue, detail::get_onemkl_generate(vect), m, n, k, lda)); } template <> std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::orgbr_scratchpad_size(queue, detail::get_onemkl_generate(vect), m, n, k, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgbr_scratchpad_size(queue, detail::get_onemkl_generate(vect), m, n, k, lda)); } template <> std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::orgtr_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::orgtr_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::orgqr_scratchpad_size(queue, m, n, k, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_scratchpad_size(queue, m, n, k, lda)); } template <> std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::orgqr_scratchpad_size(queue, m, n, k, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_scratchpad_size(queue, m, n, k, lda)); } template <> std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormrq_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, - ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormrq_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, - ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormqr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, - ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormqr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, - ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormtr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, - ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, + ldc)); } template <> std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::ormtr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, - ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, + ldc)); } template <> std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t potrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t potrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::potrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb)); } template <> std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::potrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb)); } template <> std::int64_t potrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::potrs_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, nrhs, - lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, nrhs, + lda, ldb)); } template <> std::int64_t potrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::potrs_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, nrhs, - lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, nrhs, + lda, ldb)); } template <> std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potri_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potri_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t potri_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potri_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t potri_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::potri_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::syevd_scratchpad_size(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd_scratchpad_size(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::syevd_scratchpad_size(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd_scratchpad_size(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, - ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, + ldb)); } template <> std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, - ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, + ldb)); } template <> std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrd_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::sytrd_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::trtrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, - lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, + lda, ldb)); } template <> std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::trtrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, - lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, + lda, ldb)); } template <> std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, @@ -2236,8 +2236,8 @@ std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, onea oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::trtrs_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, lda, ldb)); } template <> std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, @@ -2245,46 +2245,46 @@ std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, one oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return ::oneapi::mkl::lapack::trtrs_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, lda, ldb); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, lda, ldb)); } template <> std::int64_t ungbr_scratchpad_size>(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::ungbr_scratchpad_size>(queue, detail::get_onemkl_generate(vect), m, n, k, - lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungbr_scratchpad_size>(queue, detail::get_onemkl_generate(vect), m, n, k, + lda)); } template <> std::int64_t ungbr_scratchpad_size>(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::ungbr_scratchpad_size>(queue, detail::get_onemkl_generate(vect), m, n, k, - lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungbr_scratchpad_size>(queue, detail::get_onemkl_generate(vect), m, n, k, + lda)); } template <> std::int64_t ungqr_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::ungqr_scratchpad_size>(queue, m, n, k, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_scratchpad_size>(queue, m, n, k, lda)); } template <> std::int64_t ungqr_scratchpad_size>(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return ::oneapi::mkl::lapack::ungqr_scratchpad_size>(queue, m, n, k, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_scratchpad_size>(queue, m, n, k, lda)); } template <> std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::ungtr_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungtr_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return ::oneapi::mkl::lapack::ungtr_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungtr_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, @@ -2292,8 +2292,8 @@ std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, onea std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::unmrq_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, - n, k, lda, ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmrq_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, + n, k, lda, ldc)); } template <> std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, @@ -2301,8 +2301,8 @@ std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, one std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::unmrq_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, - n, k, lda, ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmrq_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, + n, k, lda, ldc)); } template <> std::int64_t unmqr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, @@ -2310,8 +2310,8 @@ std::int64_t unmqr_scratchpad_size>(sycl::queue &queue, onea std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::unmqr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, - n, k, lda, ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmqr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, + n, k, lda, ldc)); } template <> std::int64_t unmqr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, @@ -2319,8 +2319,8 @@ std::int64_t unmqr_scratchpad_size>(sycl::queue &queue, one std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::unmqr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, - n, k, lda, ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmqr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, + n, k, lda, ldc)); } template <> std::int64_t unmtr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, @@ -2328,8 +2328,8 @@ std::int64_t unmtr_scratchpad_size>(sycl::queue &queue, onea oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::unmtr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), - detail::get_onemkl_transpose(trans), m, n, lda, ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmtr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, lda, ldc)); } template <> std::int64_t unmtr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, @@ -2337,23 +2337,23 @@ std::int64_t unmtr_scratchpad_size>(sycl::queue &queue, one oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - return ::oneapi::mkl::lapack::unmtr_scratchpad_size>( - queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, ldc); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmtr_scratchpad_size>( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, ldc)); } template <> std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, stride_a, - stride_ipiv, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, stride_a, + stride_ipiv, batch_size)); } template <> std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, stride_a, - stride_ipiv, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, stride_a, + stride_ipiv, batch_size)); } template <> std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, @@ -2361,8 +2361,8 @@ std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( - queue, m, n, lda, stride_a, stride_ipiv, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( + queue, m, n, lda, stride_a, stride_ipiv, batch_size)); } template <> std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, @@ -2370,23 +2370,23 @@ std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queu std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( - queue, m, n, lda, stride_a, stride_ipiv, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( + queue, m, n, lda, stride_a, stride_ipiv, batch_size)); } template <> std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, stride_a, - stride_ipiv, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, stride_a, + stride_ipiv, batch_size)); } template <> std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, stride_a, - stride_ipiv, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, stride_a, + stride_ipiv, batch_size)); } template <> std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t n, @@ -2394,8 +2394,8 @@ std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getri_batch_scratchpad_size>( - queue, n, lda, stride_a, stride_ipiv, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size>( + queue, n, lda, stride_a, stride_ipiv, batch_size)); } template <> std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t n, @@ -2403,8 +2403,8 @@ std::int64_t getri_batch_scratchpad_size>(sycl::queue &queu std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getri_batch_scratchpad_size>( - queue, n, lda, stride_a, stride_ipiv, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size>( + queue, n, lda, stride_a, stride_ipiv, batch_size)); } template <> std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, @@ -2412,8 +2412,8 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size)); } template <> std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, @@ -2421,38 +2421,38 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mat std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size)); } template <> std::int64_t getrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size)); } template <> std::int64_t getrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size)); } template <> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, stride_a, - stride_tau, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, stride_a, + stride_tau, batch_size)); } template <> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, stride_a, - stride_tau, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, stride_a, + stride_tau, batch_size)); } template <> std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, @@ -2460,8 +2460,8 @@ std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( - queue, m, n, lda, stride_a, stride_tau, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( + queue, m, n, lda, stride_a, stride_tau, batch_size)); } template <> std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, @@ -2469,22 +2469,22 @@ std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queu std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( - queue, m, n, lda, stride_a, stride_tau, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( + queue, m, n, lda, stride_a, stride_tau, batch_size)); } template <> std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, - batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, + batch_size)); } template <> std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, - batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, + batch_size)); } template <> std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, @@ -2492,8 +2492,8 @@ std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size)); } template <> std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, @@ -2501,16 +2501,16 @@ std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queu std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size)); } template <> std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size( - queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size)); } template <> std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, @@ -2518,114 +2518,114 @@ std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mat std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size( - queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size)); } template <> std::int64_t potrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size)); } template <> std::int64_t potrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size)); } template <> std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, stride_a, - stride_tau, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, stride_a, + stride_tau, batch_size)); } template <> std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, stride_a, - stride_tau, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, stride_a, + stride_tau, batch_size)); } template <> std::int64_t ungqr_batch_scratchpad_size>( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( - queue, m, n, k, lda, stride_a, stride_tau, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( + queue, m, n, k, lda, stride_a, stride_tau, batch_size)); } template <> std::int64_t ungqr_batch_scratchpad_size>( sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - return ::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( - queue, m, n, k, lda, stride_a, stride_tau, batch_size); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( + queue, m, n, k, lda, stride_a, stride_tau, batch_size)); } template <> std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, group_count, - group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, group_count, + group_sizes)); } template <> std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, group_count, - group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, group_count, + group_sizes)); } template <> std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( - queue, m, n, lda, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( + queue, m, n, lda, group_count, group_sizes)); } template <> std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( - queue, m, n, lda, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( + queue, m, n, lda, group_count, group_sizes)); } template <> std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, group_count, - group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, group_count, + group_sizes)); } template <> std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, group_count, - group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, group_count, + group_sizes)); } template <> std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getri_batch_scratchpad_size>( - queue, n, lda, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size>( + queue, n, lda, group_count, group_sizes)); } template <> std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getri_batch_scratchpad_size>( - queue, n, lda, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size>( + queue, n, lda, group_count, group_sizes)); } template <> std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, @@ -2633,8 +2633,8 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, - ldb, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, + ldb, group_count, group_sizes)); } template <> std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, @@ -2642,86 +2642,86 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mat std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> std::int64_t getrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> std::int64_t getrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, group_count, - group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, group_count, + group_sizes)); } template <> std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, group_count, - group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, group_count, + group_sizes)); } template <> std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( - queue, m, n, lda, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( + queue, m, n, lda, group_count, group_sizes)); } template <> std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( - queue, m, n, lda, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( + queue, m, n, lda, group_count, group_sizes)); } template <> std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, - group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, + group_count, group_sizes)); } template <> std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, - group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, + group_count, group_sizes)); } template <> std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, - group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, + group_count, group_sizes)); } template <> std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, - group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, + group_count, group_sizes)); } template <> std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, @@ -2729,8 +2729,8 @@ std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes)); } template <> std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, @@ -2738,8 +2738,8 @@ std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queu std::int64_t *n, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes)); } template <> std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, @@ -2747,8 +2747,8 @@ std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, - group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, + group_count, group_sizes)); } template <> std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, @@ -2756,22 +2756,22 @@ std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mat std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size( - queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> std::int64_t potrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> std::int64_t potrs_batch_scratchpad_size>( sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> std::int64_t ungqr_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, @@ -2779,8 +2779,8 @@ std::int64_t ungqr_batch_scratchpad_size>(sycl::queue &queue std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( - queue, m, n, k, lda, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( + queue, m, n, k, lda, group_count, group_sizes)); } template <> std::int64_t ungqr_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, @@ -2788,6 +2788,6 @@ std::int64_t ungqr_batch_scratchpad_size>(sycl::queue &queu std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { - return ::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( - queue, m, n, k, lda, group_count, group_sizes); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( + queue, m, n, k, lda, group_count, group_sizes)); } diff --git a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp b/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp index 152426c9c..eee9e42e0 100644 --- a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp +++ b/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp @@ -21,7 +21,7 @@ #include -#include "common_mkl_types_conversion.hpp" +#include "common_onemkl_conversion.hpp" #include "oneapi/math/types.hpp" #include "oneapi/math/lapack/types.hpp" diff --git a/src/rng/backends/mklgpu/CMakeLists.txt b/src/rng/backends/mklgpu/CMakeLists.txt index b008e1770..d4bdc16ca 100644 --- a/src/rng/backends/mklgpu/CMakeLists.txt +++ b/src/rng/backends/mklgpu/CMakeLists.txt @@ -31,6 +31,7 @@ add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src + ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) diff --git a/src/rng/backends/mklgpu/mrg32k3a.cpp b/src/rng/backends/mklgpu/mrg32k3a.cpp index f4e17145b..849f78761 100644 --- a/src/rng/backends/mklgpu/mrg32k3a.cpp +++ b/src/rng/backends/mklgpu/mrg32k3a.cpp @@ -58,99 +58,99 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::uniform< std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::gaussian< float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::gaussian< double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::lognormal< float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::lognormal< double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bits& /*distr*/, std::int64_t /*n*/, @@ -163,13 +163,13 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( @@ -177,95 +177,95 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { ; - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate(const bits& /*distr*/, std::int64_t /*n*/, std::uint32_t* /*r*/, @@ -279,11 +279,11 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual void skip_ahead(std::uint64_t num_to_skip) override { - oneapi::mkl::rng::skip_ahead(engine_, num_to_skip); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::skip_ahead(engine_, num_to_skip)); } virtual void skip_ahead(std::initializer_list num_to_skip) override { - oneapi::mkl::rng::skip_ahead(engine_, num_to_skip); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::skip_ahead(engine_, num_to_skip)); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { diff --git a/src/rng/backends/mklgpu/philox4x32x10.cpp b/src/rng/backends/mklgpu/philox4x32x10.cpp index 43a8a30b4..36fa9e621 100644 --- a/src/rng/backends/mklgpu/philox4x32x10.cpp +++ b/src/rng/backends/mklgpu/philox4x32x10.cpp @@ -59,99 +59,99 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::uniform< std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::gaussian< float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::gaussian< double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::lognormal< float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::lognormal< double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bits& /*distr*/, std::int64_t /*n*/, @@ -164,13 +164,13 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( @@ -178,95 +178,95 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { ; - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - return oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate(const bits& /*distr*/, std::int64_t /*n*/, std::uint32_t* /*r*/, @@ -280,11 +280,11 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual void skip_ahead(std::uint64_t num_to_skip) override { - oneapi::mkl::rng::skip_ahead(engine_, num_to_skip); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::skip_ahead(engine_, num_to_skip)); } virtual void skip_ahead(std::initializer_list num_to_skip) override { - oneapi::mkl::rng::skip_ahead(engine_, num_to_skip); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::skip_ahead(engine_, num_to_skip)); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.cxx b/src/sparse_blas/backends/mkl_common/mkl_handles.cxx index ee8be6004..509b5a78a 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.cxx @@ -166,14 +166,14 @@ void init_coo_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t * oneapi::math::index_base index, sycl::buffer row_ind, sycl::buffer col_ind, sycl::buffer val) { oneapi::mkl::sparse::matrix_handle_t mkl_handle; - oneapi::mkl::sparse::init_matrix_handle(&mkl_handle); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::init_matrix_handle(&mkl_handle)); auto internal_smhandle = new detail::sparse_matrix_handle(mkl_handle, row_ind, col_ind, val); // The backend handle must use the buffers from the internal handle as they will be kept alive until the handle is released. - oneapi::mkl::sparse::set_coo_data(queue, mkl_handle, static_cast(num_rows), + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_coo_data(queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), static_cast(nnz), detail::get_onemkl_index_base(index), internal_smhandle->row_container.get_buffer(), internal_smhandle->col_container.get_buffer(), - internal_smhandle->value_container.get_buffer()); + internal_smhandle->value_container.get_buffer())); *p_smhandle = reinterpret_cast(internal_smhandle); } @@ -183,11 +183,11 @@ void init_coo_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t * oneapi::math::index_base index, intType *row_ind, intType *col_ind, fpType *val) { oneapi::mkl::sparse::matrix_handle_t mkl_handle; - oneapi::mkl::sparse::init_matrix_handle(&mkl_handle); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::init_matrix_handle(&mkl_handle)); auto internal_smhandle = new detail::sparse_matrix_handle(mkl_handle, row_ind, col_ind, val); - auto event = oneapi::mkl::sparse::set_coo_data( - queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), - static_cast(nnz), detail::get_onemkl_index_base(index), row_ind, col_ind, val); + sycl::event event; + RETHROW_ONEMKL_EXCEPTIONS(event = oneapi::mkl::sparse::set_coo_data(queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), + static_cast(nnz), detail::get_onemkl_index_base(index), row_ind, col_ind, val)); event.wait_and_throw(); *p_smhandle = reinterpret_cast(internal_smhandle); } @@ -222,12 +222,12 @@ void set_coo_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle internal_smhandle->col_container.set_buffer(col_ind); internal_smhandle->value_container.set_buffer(val); // The backend handle must use the buffers from the internal handle as they will be kept alive until the handle is released. - oneapi::mkl::sparse::set_coo_data(queue, internal_smhandle->backend_handle, + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_coo_data(queue, internal_smhandle->backend_handle, static_cast(num_rows), static_cast(num_cols), static_cast(nnz), detail::get_onemkl_index_base(index), internal_smhandle->row_container.get_buffer(), internal_smhandle->col_container.get_buffer(), - internal_smhandle->value_container.get_buffer()); + internal_smhandle->value_container.get_buffer())); } template @@ -275,15 +275,15 @@ void init_csr_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t * oneapi::math::index_base index, sycl::buffer row_ptr, sycl::buffer col_ind, sycl::buffer val) { oneapi::mkl::sparse::matrix_handle_t mkl_handle; - oneapi::mkl::sparse::init_matrix_handle(&mkl_handle); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::init_matrix_handle(&mkl_handle)); auto internal_smhandle = new detail::sparse_matrix_handle(mkl_handle, row_ptr, col_ind, val); // The backend deduces nnz from row_ptr. // The backend handle must use the buffers from the internal handle as they will be kept alive until the handle is released. - oneapi::mkl::sparse::set_csr_data(queue, mkl_handle, static_cast(num_rows), + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_csr_data(queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), detail::get_onemkl_index_base(index), internal_smhandle->row_container.get_buffer(), internal_smhandle->col_container.get_buffer(), - internal_smhandle->value_container.get_buffer()); + internal_smhandle->value_container.get_buffer())); *p_smhandle = reinterpret_cast(internal_smhandle); } @@ -293,12 +293,12 @@ void init_csr_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t * oneapi::math::index_base index, intType *row_ptr, intType *col_ind, fpType *val) { oneapi::mkl::sparse::matrix_handle_t mkl_handle; - oneapi::mkl::sparse::init_matrix_handle(&mkl_handle); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::init_matrix_handle(&mkl_handle)); auto internal_smhandle = new detail::sparse_matrix_handle(mkl_handle, row_ptr, col_ind, val); // The backend deduces nnz from row_ptr. - auto event = oneapi::mkl::sparse::set_csr_data( - queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), detail::get_onemkl_index_base(index), - row_ptr, col_ind, val); + sycl::event event; + RETHROW_ONEMKL_EXCEPTIONS(event = oneapi::mkl::sparse::set_csr_data(queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), detail::get_onemkl_index_base(index), + row_ptr, col_ind, val)); event.wait_and_throw(); *p_smhandle = reinterpret_cast(internal_smhandle); } @@ -315,12 +315,12 @@ void set_csr_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle internal_smhandle->value_container.set_buffer(val); // The backend deduces nnz from row_ptr. // The backend handle must use the buffers from the internal handle as they will be kept alive until the handle is released. - oneapi::mkl::sparse::set_csr_data(queue, internal_smhandle->backend_handle, + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_csr_data(queue, internal_smhandle->backend_handle, static_cast(num_rows), static_cast(num_cols), detail::get_onemkl_index_base(index), internal_smhandle->row_container.get_buffer(), internal_smhandle->col_container.get_buffer(), - internal_smhandle->value_container.get_buffer()); + internal_smhandle->value_container.get_buffer())); } template @@ -381,12 +381,12 @@ bool set_matrix_property(sycl::queue & /*queue*/, oneapi::math::sparse::matrix_h // Backend and oneMath types for the property don't match switch (property) { case oneapi::math::sparse::matrix_property::symmetric: - oneapi::mkl::sparse::set_matrix_property(internal_smhandle->backend_handle, - oneapi::mkl::sparse::property::symmetric); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_matrix_property(internal_smhandle->backend_handle, + oneapi::mkl::sparse::property::symmetric)); return true; case oneapi::math::sparse::matrix_property::sorted: - oneapi::mkl::sparse::set_matrix_property(internal_smhandle->backend_handle, - oneapi::mkl::sparse::property::sorted); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_matrix_property(internal_smhandle->backend_handle, + oneapi::mkl::sparse::property::sorted)); return true; default: return false; } diff --git a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp index 5ac48eb3f..8401d76c6 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp @@ -29,7 +29,7 @@ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/sparse_blas/detail/helper_types.hpp" -#include "common_mkl_types_conversion.hpp" +#include "common_onemkl_conversion.hpp" #include "sparse_blas/enum_data_types.hpp" #include "sparse_blas/macros.hpp" diff --git a/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx b/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx index e4a0abe1a..92cb6ec83 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx @@ -200,17 +200,17 @@ sycl::event internal_spmm( auto ldb = B_handle->ld; auto ldc = C_handle->ld; if (internal_A_handle->all_use_buffer()) { - oneapi::mkl::sparse::gemm(queue, onemkl_layout, onemkl_opa, onemkl_opb, host_alpha, + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::gemm(queue, onemkl_layout, onemkl_opa, onemkl_opb, host_alpha, internal_A_handle->backend_handle, B_handle->get_buffer(), - columns, ldb, host_beta, C_handle->get_buffer(), ldc); + columns, ldb, host_beta, C_handle->get_buffer(), ldc)); // Dependencies are not used for buffers return {}; } else { - return oneapi::mkl::sparse::gemm(queue, onemkl_layout, onemkl_opa, onemkl_opb, host_alpha, + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::gemm(queue, onemkl_layout, onemkl_opa, onemkl_opb, host_alpha, internal_A_handle->backend_handle, B_handle->get_usm_ptr(), columns, ldb, host_beta, - C_handle->get_usm_ptr(), ldc, dependencies); + C_handle->get_usm_ptr(), ldc, dependencies)); } } diff --git a/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx b/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx index 539581e1f..5d03466fd 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx @@ -147,8 +147,8 @@ void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void * auto onemkl_opa = detail::get_onemkl_transpose(opA); auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); if (A_view.type_view == matrix_descr::triangular) { - oneapi::mkl::sparse::optimize_trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, - internal_A_handle->backend_handle); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::optimize_trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, + internal_A_handle->backend_handle)); } else if (A_view.type_view == matrix_descr::symmetric || A_view.type_view == matrix_descr::hermitian) { @@ -156,7 +156,7 @@ void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void * return; } else { - oneapi::mkl::sparse::optimize_gemv(queue, onemkl_opa, internal_A_handle->backend_handle); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::optimize_gemv(queue, onemkl_opa, internal_A_handle->backend_handle)); } } @@ -182,16 +182,16 @@ sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const auto onemkl_opa = detail::get_onemkl_transpose(opA); auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); if (A_view.type_view == matrix_descr::triangular) { - return oneapi::mkl::sparse::optimize_trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, - internal_A_handle->backend_handle, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::optimize_trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, + internal_A_handle->backend_handle, dependencies)); } else if (A_view.type_view == matrix_descr::symmetric || A_view.type_view == matrix_descr::hermitian) { return detail::collapse_dependencies(queue, dependencies); } else { - return oneapi::mkl::sparse::optimize_gemv(queue, onemkl_opa, internal_A_handle->backend_handle, - dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::optimize_gemv(queue, onemkl_opa, internal_A_handle->backend_handle, + dependencies)); } } @@ -219,17 +219,17 @@ sycl::event internal_spmv(sycl::queue &queue, oneapi::math::transpose opA, const auto x_buffer = x_handle->get_buffer(); auto y_buffer = y_handle->get_buffer(); if (A_view.type_view == matrix_descr::triangular) { - oneapi::mkl::sparse::trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, - backend_handle, x_buffer, host_beta, y_buffer); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, + backend_handle, x_buffer, host_beta, y_buffer)); } else if (A_view.type_view == matrix_descr::symmetric || A_view.type_view == matrix_descr::hermitian) { - oneapi::mkl::sparse::symv(queue, onemkl_uplo, host_alpha, backend_handle, x_buffer, - host_beta, y_buffer); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::symv(queue, onemkl_uplo, host_alpha, backend_handle, x_buffer, + host_beta, y_buffer)); } else { - oneapi::mkl::sparse::gemv(queue, onemkl_opa, host_alpha, backend_handle, x_buffer, host_beta, - y_buffer); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::gemv(queue, onemkl_opa, host_alpha, backend_handle, x_buffer, host_beta, + y_buffer)); } // Dependencies are not used for buffers return {}; @@ -238,18 +238,18 @@ sycl::event internal_spmv(sycl::queue &queue, oneapi::math::transpose opA, const auto x_usm = x_handle->get_usm_ptr(); auto y_usm = y_handle->get_usm_ptr(); if (A_view.type_view == matrix_descr::triangular) { - return oneapi::mkl::sparse::trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, backend_handle, x_usm, host_beta, y_usm, - dependencies); + dependencies)); } else if (A_view.type_view == matrix_descr::symmetric || A_view.type_view == matrix_descr::hermitian) { - return oneapi::mkl::sparse::symv(queue, onemkl_uplo, host_alpha, backend_handle, - x_usm, host_beta, y_usm, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::symv(queue, onemkl_uplo, host_alpha, backend_handle, + x_usm, host_beta, y_usm, dependencies)); } else { - return oneapi::mkl::sparse::gemv(queue, onemkl_opa, host_alpha, backend_handle, x_usm, - host_beta, y_usm, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::gemv(queue, onemkl_opa, host_alpha, backend_handle, x_usm, + host_beta, y_usm, dependencies)); } } } diff --git a/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx b/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx index 72253175b..df998c59a 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx @@ -141,8 +141,8 @@ void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void * auto onemkl_uplo = detail::get_onemkl_uplo(A_view.uplo_view); auto onemkl_opa = detail::get_onemkl_transpose(opA); auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); - oneapi::mkl::sparse::optimize_trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, - internal_A_handle->backend_handle); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::optimize_trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, + internal_A_handle->backend_handle)); } sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, @@ -165,8 +165,8 @@ sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const auto onemkl_uplo = detail::get_onemkl_uplo(A_view.uplo_view); auto onemkl_opa = detail::get_onemkl_transpose(opA); auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); - return oneapi::mkl::sparse::optimize_trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, - internal_A_handle->backend_handle, dependencies); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::optimize_trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, + internal_A_handle->backend_handle, dependencies)); } template @@ -187,17 +187,17 @@ sycl::event internal_spsv(sycl::queue &queue, oneapi::math::transpose opA, const auto onemkl_opa = detail::get_onemkl_transpose(opA); auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); if (internal_A_handle->all_use_buffer()) { - oneapi::mkl::sparse::trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, internal_A_handle->backend_handle, x_handle->get_buffer(), - y_handle->get_buffer()); + y_handle->get_buffer())); // Dependencies are not used for buffers return {}; } else { - return oneapi::mkl::sparse::trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, internal_A_handle->backend_handle, x_handle->get_usm_ptr(), y_handle->get_usm_ptr(), - dependencies); + dependencies)); } } diff --git a/tests/unit_tests/blas/include/reference_blas_wrappers.hpp b/tests/unit_tests/blas/include/reference_blas_wrappers.hpp index 3b86e1d33..0ba5506c0 100644 --- a/tests/unit_tests/blas/include/reference_blas_wrappers.hpp +++ b/tests/unit_tests/blas/include/reference_blas_wrappers.hpp @@ -20,7 +20,7 @@ #ifndef _REFERENCE_BLAS_WRAPPERS_HPP__ #define _REFERENCE_BLAS_WRAPPERS_HPP__ -#include "oneapi/math/exceptions.hpp" +#include "oneapi/math/detail/exceptions.hpp" #include "oneapi/math/detail/config.hpp" #include #include "cblas.h" From 45158416507576645b8f671e1316e9d530520c27 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 18 Oct 2024 12:00:13 +0100 Subject: [PATCH 26/41] Remove MKL_INCLUDE from curand and rocrand backends --- src/rng/backends/curand/CMakeLists.txt | 1 - src/rng/backends/rocrand/CMakeLists.txt | 9 ++++++--- 2 files changed, 6 insertions(+), 4 deletions(-) diff --git a/src/rng/backends/curand/CMakeLists.txt b/src/rng/backends/curand/CMakeLists.txt index 818fa8637..566db6422 100644 --- a/src/rng/backends/curand/CMakeLists.txt +++ b/src/rng/backends/curand/CMakeLists.txt @@ -73,7 +73,6 @@ target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin - ${MKL_INCLUDE} ${ONEMATH_GENERATED_INCLUDE_PATH} ) diff --git a/src/rng/backends/rocrand/CMakeLists.txt b/src/rng/backends/rocrand/CMakeLists.txt index 5958a2b10..44dc98a51 100644 --- a/src/rng/backends/rocrand/CMakeLists.txt +++ b/src/rng/backends/rocrand/CMakeLists.txt @@ -66,9 +66,12 @@ add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) -target_include_directories( - ${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/include ${PROJECT_SOURCE_DIR}/src - ${CMAKE_BINARY_DIR}/bin ${MKL_INCLUDE} ${ONEMATH_GENERATED_INCLUDE_PATH}) +target_include_directories(${LIB_OBJ} + PRIVATE ${PROJECT_SOURCE_DIR}/include + ${PROJECT_SOURCE_DIR}/src + ${CMAKE_BINARY_DIR}/bin + ${ONEMATH_GENERATED_INCLUDE_PATH} +) target_link_libraries(${LIB_OBJ} PRIVATE roc::rocrand hip::host Threads::Threads) target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL) From c5c4e08e8bb8a79cdaad787a1ee39fe41d166978 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 18 Oct 2024 12:12:03 +0100 Subject: [PATCH 27/41] fp_scalar_mkl -> fp_scalar_mkl --- tests/unit_tests/blas/level1/scal.cpp | 4 ++-- tests/unit_tests/blas/level1/scal_usm.cpp | 4 ++-- tests/unit_tests/blas/level2/her.cpp | 4 ++-- tests/unit_tests/blas/level2/her_usm.cpp | 4 ++-- tests/unit_tests/blas/level2/hpr.cpp | 4 ++-- tests/unit_tests/blas/level2/hpr_usm.cpp | 4 ++-- tests/unit_tests/blas/level3/her2k.cpp | 4 ++-- tests/unit_tests/blas/level3/her2k_usm.cpp | 4 ++-- 8 files changed, 16 insertions(+), 16 deletions(-) diff --git a/tests/unit_tests/blas/level1/scal.cpp b/tests/unit_tests/blas/level1/scal.cpp index 1d7fad6d8..c345a6f55 100644 --- a/tests/unit_tests/blas/level1/scal.cpp +++ b/tests/unit_tests/blas/level1/scal.cpp @@ -55,11 +55,11 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, fp_scalar al // Call Reference SCAL. using fp_ref = typename ref_type_info::type; - using fp_scalar_mkl = typename ref_type_info::type; + using fp_scalar_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = std::abs(incx); - ::scal(&N_ref, (fp_scalar_mkl*)&alpha, (fp_ref*)x_ref.data(), &incx_ref); + ::scal(&N_ref, (fp_scalar_ref*)&alpha, (fp_ref*)x_ref.data(), &incx_ref); // Call DPC++ SCAL. diff --git a/tests/unit_tests/blas/level1/scal_usm.cpp b/tests/unit_tests/blas/level1/scal_usm.cpp index a62372331..6341533be 100644 --- a/tests/unit_tests/blas/level1/scal_usm.cpp +++ b/tests/unit_tests/blas/level1/scal_usm.cpp @@ -76,11 +76,11 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, fp_scalar al // Call Reference SCAL. using fp_ref = typename ref_type_info::type; - using fp_scalar_mkl = typename ref_type_info::type; + using fp_scalar_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = std::abs(incx); - ::scal(&N_ref, (fp_scalar_mkl*)&alpha, (fp_ref*)x_ref.data(), &incx_ref); + ::scal(&N_ref, (fp_scalar_ref*)&alpha, (fp_ref*)x_ref.data(), &incx_ref); // Call DPC++ SCAL. diff --git a/tests/unit_tests/blas/level2/her.cpp b/tests/unit_tests/blas/level2/her.cpp index d7817ba9d..9bc431053 100644 --- a/tests/unit_tests/blas/level2/her.cpp +++ b/tests/unit_tests/blas/level2/her.cpp @@ -58,10 +58,10 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe // Call Reference HER. const int n_ref = n, incx_ref = incx, lda_ref = lda; using fp_ref = typename ref_type_info::type; - using fp_scalar_mkl = typename ref_type_info::type; + using fp_scalar_ref = typename ref_type_info::type; ::her(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_scalar_mkl *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data(), &lda_ref); + (fp_scalar_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data(), &lda_ref); // Call DPC++ HER. diff --git a/tests/unit_tests/blas/level2/her_usm.cpp b/tests/unit_tests/blas/level2/her_usm.cpp index cc61fb93b..0a7f86afe 100644 --- a/tests/unit_tests/blas/level2/her_usm.cpp +++ b/tests/unit_tests/blas/level2/her_usm.cpp @@ -79,10 +79,10 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe // Call Reference HER. const int n_ref = n, incx_ref = incx, lda_ref = lda; using fp_ref = typename ref_type_info::type; - using fp_scalar_mkl = typename ref_type_info::type; + using fp_scalar_ref = typename ref_type_info::type; ::her(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_scalar_mkl *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data(), &lda_ref); + (fp_scalar_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data(), &lda_ref); // Call DPC++ HER. diff --git a/tests/unit_tests/blas/level2/hpr.cpp b/tests/unit_tests/blas/level2/hpr.cpp index 157efa989..3ea336bab 100644 --- a/tests/unit_tests/blas/level2/hpr.cpp +++ b/tests/unit_tests/blas/level2/hpr.cpp @@ -58,10 +58,10 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe // Call Reference HPR. const int n_ref = n, incx_ref = incx; using fp_ref = typename ref_type_info::type; - using fp_scalar_mkl = typename ref_type_info::type; + using fp_scalar_ref = typename ref_type_info::type; ::hpr(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_scalar_mkl *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data()); + (fp_scalar_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data()); // Call DPC++ HPR. diff --git a/tests/unit_tests/blas/level2/hpr_usm.cpp b/tests/unit_tests/blas/level2/hpr_usm.cpp index b7b38a80a..003d41f0b 100644 --- a/tests/unit_tests/blas/level2/hpr_usm.cpp +++ b/tests/unit_tests/blas/level2/hpr_usm.cpp @@ -79,10 +79,10 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe // Call Reference HPR. const int n_ref = n, incx_ref = incx; using fp_ref = typename ref_type_info::type; - using fp_scalar_mkl = typename ref_type_info::type; + using fp_scalar_ref = typename ref_type_info::type; ::hpr(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_scalar_mkl *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data()); + (fp_scalar_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data()); // Call DPC++ HPR. diff --git a/tests/unit_tests/blas/level3/her2k.cpp b/tests/unit_tests/blas/level3/her2k.cpp index cfc3674d4..73dbfa9ee 100644 --- a/tests/unit_tests/blas/level3/her2k.cpp +++ b/tests/unit_tests/blas/level3/her2k.cpp @@ -65,11 +65,11 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe const int lda_ref = lda, ldb_ref = ldb, ldc_ref = ldc; using fp_ref = typename ref_type_info::type; - using fp_scalar_mkl = typename ref_type_info::type; + using fp_scalar_ref = typename ref_type_info::type; ::her2k(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), &n_ref, &k_ref, (fp_ref*)&alpha, (fp_ref*)A.data(), - &lda_ref, (fp_ref*)B.data(), &ldb_ref, (fp_scalar_mkl*)&beta, (fp_ref*)C_ref.data(), + &lda_ref, (fp_ref*)B.data(), &ldb_ref, (fp_scalar_ref*)&beta, (fp_ref*)C_ref.data(), &ldc_ref); // Call DPC++ HER2K. diff --git a/tests/unit_tests/blas/level3/her2k_usm.cpp b/tests/unit_tests/blas/level3/her2k_usm.cpp index 202e7c29d..642fd6219 100644 --- a/tests/unit_tests/blas/level3/her2k_usm.cpp +++ b/tests/unit_tests/blas/level3/her2k_usm.cpp @@ -85,11 +85,11 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe const int lda_ref = lda, ldb_ref = ldb, ldc_ref = ldc; using fp_ref = typename ref_type_info::type; - using fp_scalar_mkl = typename ref_type_info::type; + using fp_scalar_ref = typename ref_type_info::type; ::her2k(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), &n_ref, &k_ref, (fp_ref*)&alpha, (fp_ref*)A.data(), - &lda_ref, (fp_ref*)B.data(), &ldb_ref, (fp_scalar_mkl*)&beta, (fp_ref*)C_ref.data(), + &lda_ref, (fp_ref*)B.data(), &ldb_ref, (fp_scalar_ref*)&beta, (fp_ref*)C_ref.data(), &ldc_ref); // Call DPC++ HER2K. From 213b84b43844009b01c48c08d9874f94c286fbe0 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 18 Oct 2024 12:18:15 +0100 Subject: [PATCH 28/41] Rename mkl_*_table --- include/oneapi/math/detail/backends_table.hpp | 10 +++++----- scripts/generate_wrappers.py | 2 +- src/blas/backends/cublas/cublas_wrappers.cpp | 2 +- src/blas/backends/mklcpu/mklcpu_wrappers.cpp | 2 +- src/blas/backends/mklgpu/mklgpu_wrappers.cpp | 2 +- src/blas/backends/netlib/netlib_wrappers.cpp | 2 +- src/blas/backends/portblas/portblas_wrappers.cpp | 2 +- src/blas/backends/rocblas/rocblas_wrappers.cpp | 2 +- src/dft/backends/backend_wrappers.cxx | 2 +- src/dft/backends/cufft/cufft_wrappers.cpp | 2 +- src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp | 2 +- src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp | 2 +- src/dft/backends/portfft/portfft_wrappers.cpp | 2 +- src/dft/backends/rocfft/rocfft_wrappers.cpp | 2 +- src/lapack/backends/cusolver/cusolver_wrappers.cpp | 2 +- src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp | 2 +- src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp | 2 +- src/lapack/backends/rocsolver/rocsolver_wrappers.cpp | 2 +- src/rng/backends/curand/curand_wrappers.cpp | 2 +- src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp | 2 +- src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp | 2 +- src/rng/backends/rocrand/rocrand_wrappers.cpp | 2 +- src/sparse_blas/backends/backend_wrappers.cxx | 2 +- src/sparse_blas/backends/mklcpu/mklcpu_wrappers.cpp | 2 +- src/sparse_blas/backends/mklgpu/mklgpu_wrappers.cpp | 2 +- 25 files changed, 29 insertions(+), 29 deletions(-) diff --git a/include/oneapi/math/detail/backends_table.hpp b/include/oneapi/math/detail/backends_table.hpp index 2b931527e..4fe71dbee 100644 --- a/include/oneapi/math/detail/backends_table.hpp +++ b/include/oneapi/math/detail/backends_table.hpp @@ -202,12 +202,12 @@ static std::map>> libraries = } } } }, }; -static std::map table_names = { { domain::blas, "mkl_blas_table" }, - { domain::lapack, "mkl_lapack_table" }, - { domain::dft, "mkl_dft_table" }, - { domain::rng, "mkl_rng_table" }, +static std::map table_names = { { domain::blas, "onemath_blas_table" }, + { domain::lapack, "onemath_lapack_table" }, + { domain::dft, "onemath_dft_table" }, + { domain::rng, "onemath_rng_table" }, { domain::sparse_blas, - "mkl_sparse_blas_table" } }; + "onemath_sparse_blas_table" } }; } //namespace math } //namespace oneapi diff --git a/scripts/generate_wrappers.py b/scripts/generate_wrappers.py index 61602128e..8f449bf2e 100755 --- a/scripts/generate_wrappers.py +++ b/scripts/generate_wrappers.py @@ -144,7 +144,7 @@ def print_funcs(func_list): #define WRAPPER_VERSION 1 -extern "C" function_table_t mkl_blas_table = {{ +extern "C" function_table_t onemath_blas_table = {{ WRAPPER_VERSION, """.format(table=in_table.strip('src/'), header=in_filename.strip('include/'))) diff --git a/src/blas/backends/cublas/cublas_wrappers.cpp b/src/blas/backends/cublas/cublas_wrappers.cpp index ec89f9c58..993abf84a 100644 --- a/src/blas/backends/cublas/cublas_wrappers.cpp +++ b/src/blas/backends/cublas/cublas_wrappers.cpp @@ -21,7 +21,7 @@ #define WRAPPER_VERSION 1 -extern "C" blas_function_table_t mkl_blas_table = { +extern "C" blas_function_table_t onemath_blas_table = { WRAPPER_VERSION, oneapi::math::blas::cublas::column_major::asum, oneapi::math::blas::cublas::column_major::asum, diff --git a/src/blas/backends/mklcpu/mklcpu_wrappers.cpp b/src/blas/backends/mklcpu/mklcpu_wrappers.cpp index 6b5caea4e..13f2d89fb 100644 --- a/src/blas/backends/mklcpu/mklcpu_wrappers.cpp +++ b/src/blas/backends/mklcpu/mklcpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT blas_function_table_t mkl_blas_table = { +extern "C" ONEMATH_EXPORT blas_function_table_t onemath_blas_table = { WRAPPER_VERSION, #define BACKEND mklcpu #define MAJOR column_major diff --git a/src/blas/backends/mklgpu/mklgpu_wrappers.cpp b/src/blas/backends/mklgpu/mklgpu_wrappers.cpp index 6328196b8..c47019f8c 100644 --- a/src/blas/backends/mklgpu/mklgpu_wrappers.cpp +++ b/src/blas/backends/mklgpu/mklgpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT blas_function_table_t mkl_blas_table = { +extern "C" ONEMATH_EXPORT blas_function_table_t onemath_blas_table = { WRAPPER_VERSION, #define BACKEND mklgpu #define MAJOR column_major diff --git a/src/blas/backends/netlib/netlib_wrappers.cpp b/src/blas/backends/netlib/netlib_wrappers.cpp index 6d067e904..291b8e4ed 100644 --- a/src/blas/backends/netlib/netlib_wrappers.cpp +++ b/src/blas/backends/netlib/netlib_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT blas_function_table_t mkl_blas_table = { +extern "C" ONEMATH_EXPORT blas_function_table_t onemath_blas_table = { WRAPPER_VERSION, #define BACKEND netlib #define MAJOR column_major diff --git a/src/blas/backends/portblas/portblas_wrappers.cpp b/src/blas/backends/portblas/portblas_wrappers.cpp index bf9d2591b..6777f358a 100644 --- a/src/blas/backends/portblas/portblas_wrappers.cpp +++ b/src/blas/backends/portblas/portblas_wrappers.cpp @@ -8,7 +8,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT blas_function_table_t mkl_blas_table = { +extern "C" ONEMATH_EXPORT blas_function_table_t onemath_blas_table = { WRAPPER_VERSION, #define BACKEND portblas #define MAJOR column_major diff --git a/src/blas/backends/rocblas/rocblas_wrappers.cpp b/src/blas/backends/rocblas/rocblas_wrappers.cpp index 3c2fb8c38..111944283 100644 --- a/src/blas/backends/rocblas/rocblas_wrappers.cpp +++ b/src/blas/backends/rocblas/rocblas_wrappers.cpp @@ -23,7 +23,7 @@ #define WRAPPER_VERSION 1 -extern "C" blas_function_table_t mkl_blas_table = { +extern "C" blas_function_table_t onemath_blas_table = { WRAPPER_VERSION, oneapi::math::blas::rocblas::column_major::asum, oneapi::math::blas::rocblas::column_major::asum, diff --git a/src/dft/backends/backend_wrappers.cxx b/src/dft/backends/backend_wrappers.cxx index d265bc91c..1beb8f988 100644 --- a/src/dft/backends/backend_wrappers.cxx +++ b/src/dft/backends/backend_wrappers.cxx @@ -26,7 +26,7 @@ To use this: #define WRAPPER_VERSION #define BACKEND -extern "C" dft_function_table_t mkl_dft_table = { +extern "C" dft_function_table_t onemath_dft_table = { WRAPPER_VERSION, #include "dft/backends/backend_wrappers.cxx" }; diff --git a/src/dft/backends/cufft/cufft_wrappers.cpp b/src/dft/backends/cufft/cufft_wrappers.cpp index fed14d6fc..6cbb9cc99 100644 --- a/src/dft/backends/cufft/cufft_wrappers.cpp +++ b/src/dft/backends/cufft/cufft_wrappers.cpp @@ -23,7 +23,7 @@ #define WRAPPER_VERSION 1 #define BACKEND cufft -extern "C" dft_function_table_t mkl_dft_table = { +extern "C" dft_function_table_t onemath_dft_table = { WRAPPER_VERSION, #include "dft/backends/backend_wrappers.cxx" }; diff --git a/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp b/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp index 10f86017c..3c8193f3b 100644 --- a/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp +++ b/src/dft/backends/mklcpu/mkl_dft_cpu_wrappers.cpp @@ -23,7 +23,7 @@ #define WRAPPER_VERSION 1 #define BACKEND mklcpu -extern "C" dft_function_table_t mkl_dft_table = { +extern "C" dft_function_table_t onemath_dft_table = { WRAPPER_VERSION, #include "dft/backends/backend_wrappers.cxx" }; diff --git a/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp b/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp index bcadc5464..a6b50fe39 100644 --- a/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp +++ b/src/dft/backends/mklgpu/mkl_dft_gpu_wrappers.cpp @@ -23,7 +23,7 @@ #define WRAPPER_VERSION 1 #define BACKEND mklgpu -extern "C" dft_function_table_t mkl_dft_table = { +extern "C" dft_function_table_t onemath_dft_table = { WRAPPER_VERSION, #include "dft/backends/backend_wrappers.cxx" }; diff --git a/src/dft/backends/portfft/portfft_wrappers.cpp b/src/dft/backends/portfft/portfft_wrappers.cpp index c702d3696..a628e3e25 100644 --- a/src/dft/backends/portfft/portfft_wrappers.cpp +++ b/src/dft/backends/portfft/portfft_wrappers.cpp @@ -23,7 +23,7 @@ #define WRAPPER_VERSION 1 #define BACKEND portfft -extern "C" dft_function_table_t mkl_dft_table = { +extern "C" dft_function_table_t onemath_dft_table = { WRAPPER_VERSION, #include "dft/backends/backend_wrappers.cxx" }; diff --git a/src/dft/backends/rocfft/rocfft_wrappers.cpp b/src/dft/backends/rocfft/rocfft_wrappers.cpp index 7554f0829..d2d198b78 100644 --- a/src/dft/backends/rocfft/rocfft_wrappers.cpp +++ b/src/dft/backends/rocfft/rocfft_wrappers.cpp @@ -23,7 +23,7 @@ #define WRAPPER_VERSION 1 #define BACKEND rocfft -extern "C" dft_function_table_t mkl_dft_table = { +extern "C" dft_function_table_t onemath_dft_table = { WRAPPER_VERSION, #include "dft/backends/backend_wrappers.cxx" }; diff --git a/src/lapack/backends/cusolver/cusolver_wrappers.cpp b/src/lapack/backends/cusolver/cusolver_wrappers.cpp index 52a4eec26..13db62901 100644 --- a/src/lapack/backends/cusolver/cusolver_wrappers.cpp +++ b/src/lapack/backends/cusolver/cusolver_wrappers.cpp @@ -21,7 +21,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT lapack_function_table_t mkl_lapack_table = { +extern "C" ONEMATH_EXPORT lapack_function_table_t onemath_lapack_table = { WRAPPER_VERSION, #define LAPACK_BACKEND cusolver oneapi::math::lapack::cusolver::gebrd, diff --git a/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp b/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp index 30195d3a1..abdfcda46 100644 --- a/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp +++ b/src/lapack/backends/mklcpu/lapack_cpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT lapack_function_table_t mkl_lapack_table = { +extern "C" ONEMATH_EXPORT lapack_function_table_t onemath_lapack_table = { WRAPPER_VERSION, #define LAPACK_BACKEND mklcpu #include "../mkl_common/lapack_wrappers.cxx" diff --git a/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp b/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp index e6435b898..dd2d39e42 100644 --- a/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp +++ b/src/lapack/backends/mklgpu/lapack_gpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT lapack_function_table_t mkl_lapack_table = { +extern "C" ONEMATH_EXPORT lapack_function_table_t onemath_lapack_table = { WRAPPER_VERSION, #define LAPACK_BACKEND mklgpu #include "../mkl_common/lapack_wrappers.cxx" diff --git a/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp b/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp index 506963182..ac383582f 100644 --- a/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_wrappers.cpp @@ -23,7 +23,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT lapack_function_table_t mkl_lapack_table = { +extern "C" ONEMATH_EXPORT lapack_function_table_t onemath_lapack_table = { WRAPPER_VERSION, #define LAPACK_BACKEND rocsolver oneapi::math::lapack::rocsolver::gebrd, diff --git a/src/rng/backends/curand/curand_wrappers.cpp b/src/rng/backends/curand/curand_wrappers.cpp index c3e25a63c..8ebc8225e 100644 --- a/src/rng/backends/curand/curand_wrappers.cpp +++ b/src/rng/backends/curand/curand_wrappers.cpp @@ -61,7 +61,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { +extern "C" ONEMATH_EXPORT rng_function_table_t onemath_rng_table = { WRAPPER_VERSION, oneapi::math::rng::curand::create_philox4x32x10, oneapi::math::rng::curand::create_philox4x32x10, oneapi::math::rng::curand::create_mrg32k3a, oneapi::math::rng::curand::create_mrg32k3a diff --git a/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp b/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp index 36b8d8169..dce8d4830 100644 --- a/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp +++ b/src/rng/backends/mklcpu/mkl_rng_cpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { +extern "C" ONEMATH_EXPORT rng_function_table_t onemath_rng_table = { WRAPPER_VERSION, oneapi::math::rng::mklcpu::create_philox4x32x10, oneapi::math::rng::mklcpu::create_philox4x32x10, oneapi::math::rng::mklcpu::create_mrg32k3a, oneapi::math::rng::mklcpu::create_mrg32k3a diff --git a/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp b/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp index 47b36245e..678b6397b 100644 --- a/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp +++ b/src/rng/backends/mklgpu/mkl_rng_gpu_wrappers.cpp @@ -22,7 +22,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { +extern "C" ONEMATH_EXPORT rng_function_table_t onemath_rng_table = { WRAPPER_VERSION, oneapi::math::rng::mklgpu::create_philox4x32x10, oneapi::math::rng::mklgpu::create_philox4x32x10, oneapi::math::rng::mklgpu::create_mrg32k3a, oneapi::math::rng::mklgpu::create_mrg32k3a diff --git a/src/rng/backends/rocrand/rocrand_wrappers.cpp b/src/rng/backends/rocrand/rocrand_wrappers.cpp index 821909121..365f1b99e 100644 --- a/src/rng/backends/rocrand/rocrand_wrappers.cpp +++ b/src/rng/backends/rocrand/rocrand_wrappers.cpp @@ -63,7 +63,7 @@ #define WRAPPER_VERSION 1 -extern "C" ONEMATH_EXPORT rng_function_table_t mkl_rng_table = { +extern "C" ONEMATH_EXPORT rng_function_table_t onemath_rng_table = { WRAPPER_VERSION, oneapi::math::rng::rocrand::create_philox4x32x10, oneapi::math::rng::rocrand::create_philox4x32x10, oneapi::math::rng::rocrand::create_mrg32k3a, oneapi::math::rng::rocrand::create_mrg32k3a diff --git a/src/sparse_blas/backends/backend_wrappers.cxx b/src/sparse_blas/backends/backend_wrappers.cxx index 99ef75a82..07a0e54fe 100644 --- a/src/sparse_blas/backends/backend_wrappers.cxx +++ b/src/sparse_blas/backends/backend_wrappers.cxx @@ -26,7 +26,7 @@ To use this: #define WRAPPER_VERSION #define BACKEND -extern "C" sparse_blas_function_table_t mkl_sparse_blas_table = { +extern "C" sparse_blas_function_table_t onemath_sparse_blas_table = { WRAPPER_VERSION, #include "sparse_blas/backends/backend_wrappers.cxx" }; diff --git a/src/sparse_blas/backends/mklcpu/mklcpu_wrappers.cpp b/src/sparse_blas/backends/mklcpu/mklcpu_wrappers.cpp index d67d1d239..ea5b7af32 100644 --- a/src/sparse_blas/backends/mklcpu/mklcpu_wrappers.cpp +++ b/src/sparse_blas/backends/mklcpu/mklcpu_wrappers.cpp @@ -26,7 +26,7 @@ #define WRAPPER_VERSION 1 #define BACKEND mklcpu -extern "C" sparse_blas_function_table_t mkl_sparse_blas_table = { +extern "C" sparse_blas_function_table_t onemath_sparse_blas_table = { WRAPPER_VERSION, #include "sparse_blas/backends/backend_wrappers.cxx" }; diff --git a/src/sparse_blas/backends/mklgpu/mklgpu_wrappers.cpp b/src/sparse_blas/backends/mklgpu/mklgpu_wrappers.cpp index 17e433663..ed9cfd801 100644 --- a/src/sparse_blas/backends/mklgpu/mklgpu_wrappers.cpp +++ b/src/sparse_blas/backends/mklgpu/mklgpu_wrappers.cpp @@ -26,7 +26,7 @@ #define WRAPPER_VERSION 1 #define BACKEND mklgpu -extern "C" sparse_blas_function_table_t mkl_sparse_blas_table = { +extern "C" sparse_blas_function_table_t onemath_sparse_blas_table = { WRAPPER_VERSION, #include "sparse_blas/backends/backend_wrappers.cxx" }; From 462b5cf848aa0a9d59255ec9a5b00911d61e119a Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 18 Oct 2024 12:18:33 +0100 Subject: [PATCH 29/41] Rename mkl after merge --- src/dft/execute_helper_generic.hpp | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/dft/execute_helper_generic.hpp b/src/dft/execute_helper_generic.hpp index b857328df..ba4cc709a 100644 --- a/src/dft/execute_helper_generic.hpp +++ b/src/dft/execute_helper_generic.hpp @@ -26,7 +26,7 @@ #include #endif -namespace oneapi::mkl::dft::detail { +namespace oneapi::math::dft::detail { /** Wrap interop API to launch interop host task. * @@ -48,6 +48,6 @@ static inline void fft_enqueue_task(HandlerT&& cgh, FnT&& f) { }); } -} // namespace oneapi::mkl::dft::detail +} // namespace oneapi::math::dft::detail #endif // _ONEMATH_DFT_SRC_EXECUTE_HELPER_GENERIC_HPP_ From 14a82163cd0b40e8aba2f6c16fd45f8608e42c84 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 18 Oct 2024 16:13:03 +0100 Subject: [PATCH 30/41] Remove old workaround catch FFT_UNIMPLEMENTED --- tests/unit_tests/dft/source/compute_tests.cpp | 9 --------- 1 file changed, 9 deletions(-) diff --git a/tests/unit_tests/dft/source/compute_tests.cpp b/tests/unit_tests/dft/source/compute_tests.cpp index bea23c0d9..6375a5e50 100644 --- a/tests/unit_tests/dft/source/compute_tests.cpp +++ b/tests/unit_tests/dft/source/compute_tests.cpp @@ -73,15 +73,6 @@ class ComputeTests_real_real_out_of_place_REAL std::cout << "Skipping test because: \"" << e.what() << "\"" << std::endl; \ GTEST_SKIP(); \ } \ - catch (std::exception & e) { \ - std::string msg = e.what(); \ - if ((msg.find("FFT_UNIMPLEMENTED") != std::string::npos) || \ - (msg.find("Unimplemented") != std::string::npos)) { \ - std::cout << "Skipping test because: \"" << msg << "\"" << std::endl; \ - GTEST_SKIP(); \ - } \ - throw; \ - } \ } #define INSTANTIATE_TEST_DIMENSIONS_PRECISION_DOMAIN(PLACE, LAYOUT, STORAGE) \ From 09136c0874f765829967cc5a2998172c1eb468c1 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 18 Oct 2024 16:31:55 +0100 Subject: [PATCH 31/41] Factorize target_add_onemath_include --- src/CMakeLists.txt | 11 +++++++++++ src/blas/backends/mklcpu/CMakeLists.txt | 8 +------- src/blas/backends/mklgpu/CMakeLists.txt | 8 +------- src/dft/backends/mklgpu/CMakeLists.txt | 8 +------- src/lapack/backends/mklgpu/CMakeLists.txt | 8 +------- src/rng/backends/mklcpu/CMakeLists.txt | 8 +------- src/rng/backends/mklgpu/CMakeLists.txt | 8 +------- 7 files changed, 17 insertions(+), 42 deletions(-) diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index dfc28b703..b66666953 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -47,6 +47,17 @@ function(add_deprecated_library target_name) set_target_properties(deprecated_name PROPERTIES DEPRECATION "${deprecated_name} target is deprecated, please use ${target_name} instead") endfunction() + +# Due to using the same file name for different headers in this library and in +# the Intel(R) oneAPI Math Kernel Library, the compiler may not include the +# expected header. Using `-iquote` lets us include this project's headers when +# double quotes are used and external headers when angle brackets are used. +function(target_add_onemath_include target_name) + target_compile_options(${target_name} + BEFORE PRIVATE -iquote $ + ) +endfunction() + # Build loader and backends for each domain add_custom_target(onemath_backend_libs) foreach(domain ${TARGET_DOMAINS}) diff --git a/src/blas/backends/mklcpu/CMakeLists.txt b/src/blas/backends/mklcpu/CMakeLists.txt index d59b2f316..99c69bbf4 100644 --- a/src/blas/backends/mklcpu/CMakeLists.txt +++ b/src/blas/backends/mklcpu/CMakeLists.txt @@ -31,13 +31,7 @@ if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET ${LIB_OBJ} SOURCES ${SOURCES}) endif() -# Due to using the same file name for different headers in this library and in -# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ -# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the -# correct header. -target_compile_options(${LIB_OBJ} - BEFORE PRIVATE -iquote $ -) +target_add_onemath_include(${LIB_OBJ}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src diff --git a/src/blas/backends/mklgpu/CMakeLists.txt b/src/blas/backends/mklgpu/CMakeLists.txt index 52010a10d..84b1004ab 100644 --- a/src/blas/backends/mklgpu/CMakeLists.txt +++ b/src/blas/backends/mklgpu/CMakeLists.txt @@ -28,13 +28,7 @@ add_library(${LIB_OBJ} OBJECT ) add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) -# Due to using the same file name for different headers in this library and in -# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ -# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the -# correct header. -target_compile_options(${LIB_OBJ} - BEFORE PRIVATE -iquote $ -) +target_add_onemath_include(${LIB_OBJ}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src diff --git a/src/dft/backends/mklgpu/CMakeLists.txt b/src/dft/backends/mklgpu/CMakeLists.txt index a60e7b5eb..101963360 100644 --- a/src/dft/backends/mklgpu/CMakeLists.txt +++ b/src/dft/backends/mklgpu/CMakeLists.txt @@ -37,13 +37,7 @@ target_include_directories(${LIB_NAME} PUBLIC ${ONEMATH_INCLUDE_DIRS} ) -# Due to using the same file name for different headers in this library and in -# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ -# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the -# correct header. -target_compile_options(${LIB_OBJ} - BEFORE PRIVATE -iquote $ -) +target_add_onemath_include(${LIB_OBJ}) target_include_directories(${LIB_OBJ} PRIVATE ${PROJECT_SOURCE_DIR}/src diff --git a/src/lapack/backends/mklgpu/CMakeLists.txt b/src/lapack/backends/mklgpu/CMakeLists.txt index 83bb2fe73..7cc0eff11 100644 --- a/src/lapack/backends/mklgpu/CMakeLists.txt +++ b/src/lapack/backends/mklgpu/CMakeLists.txt @@ -35,13 +35,7 @@ target_include_directories(${LIB_OBJ} ${ONEMATH_GENERATED_INCLUDE_PATH} ) -# Due to using the same file name for different headers in this library and in -# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ -# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the -# correct header. -target_compile_options(${LIB_OBJ} - BEFORE PRIVATE -iquote $ -) +target_add_onemath_include(${LIB_OBJ}) target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) diff --git a/src/rng/backends/mklcpu/CMakeLists.txt b/src/rng/backends/mklcpu/CMakeLists.txt index a58e699f9..03ac7e9c5 100644 --- a/src/rng/backends/mklcpu/CMakeLists.txt +++ b/src/rng/backends/mklcpu/CMakeLists.txt @@ -36,13 +36,7 @@ target_include_directories(${LIB_OBJ} ${ONEMATH_GENERATED_INCLUDE_PATH} ) -# Due to using the same file name for different headers in this library and in -# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ -# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the -# correct header. -target_compile_options(${LIB_OBJ} - BEFORE PRIVATE -iquote $ -) +target_add_onemath_include(${LIB_OBJ}) target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) diff --git a/src/rng/backends/mklgpu/CMakeLists.txt b/src/rng/backends/mklgpu/CMakeLists.txt index d4bdc16ca..68bde3a64 100644 --- a/src/rng/backends/mklgpu/CMakeLists.txt +++ b/src/rng/backends/mklgpu/CMakeLists.txt @@ -36,13 +36,7 @@ target_include_directories(${LIB_OBJ} ${ONEMATH_GENERATED_INCLUDE_PATH} ) -# Due to using the same file name for different headers in this library and in -# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ -# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the -# correct header. -target_compile_options(${LIB_OBJ} - BEFORE PRIVATE -iquote $ -) +target_add_onemath_include(${LIB_OBJ}) target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) From 6690035e67003d051865b8c777867681d740b642 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 18 Oct 2024 16:39:38 +0100 Subject: [PATCH 32/41] Use angle brackets to include Intel oneMKL headers --- src/dft/backends/mklcpu/backward.cpp | 4 ++-- src/dft/backends/mklcpu/commit.cpp | 6 ++++-- src/dft/backends/mklcpu/commit_derived_impl.hpp | 4 ++-- src/dft/backends/mklcpu/forward.cpp | 4 ++-- src/dft/backends/mklcpu/mklcpu_helpers.hpp | 4 ++-- src/dft/backends/mklgpu/backward.cpp | 4 ++-- src/dft/backends/mklgpu/commit.cpp | 4 ++-- src/dft/backends/mklgpu/forward.cpp | 4 ++-- src/dft/backends/mklgpu/mklgpu_helpers.hpp | 4 ++-- .../unit_tests/lapack/include/lapack_reference_wrappers.hpp | 4 ++-- 10 files changed, 22 insertions(+), 20 deletions(-) diff --git a/src/dft/backends/mklcpu/backward.cpp b/src/dft/backends/mklcpu/backward.cpp index 063648167..cf62db084 100644 --- a/src/dft/backends/mklcpu/backward.cpp +++ b/src/dft/backends/mklcpu/backward.cpp @@ -31,8 +31,8 @@ #include "dft/backends/mklcpu/commit_derived_impl.hpp" -// MKLCPU header -#include "mkl_dfti.h" +// Intel(R) oneMKL header +#include namespace oneapi::math::dft::mklcpu { namespace detail { diff --git a/src/dft/backends/mklcpu/commit.cpp b/src/dft/backends/mklcpu/commit.cpp index 2668761aa..ca5ecaa87 100644 --- a/src/dft/backends/mklcpu/commit.cpp +++ b/src/dft/backends/mklcpu/commit.cpp @@ -34,8 +34,10 @@ #include "dft/backends/mklcpu/commit_derived_impl.hpp" #include "../stride_helper.hpp" -#include "mkl_service.h" -#include "mkl_dfti.h" + +// Intel(R) oneMKL header +#include +#include namespace oneapi { namespace math { diff --git a/src/dft/backends/mklcpu/commit_derived_impl.hpp b/src/dft/backends/mklcpu/commit_derived_impl.hpp index ee4a4e53f..6e35b5735 100644 --- a/src/dft/backends/mklcpu/commit_derived_impl.hpp +++ b/src/dft/backends/mklcpu/commit_derived_impl.hpp @@ -24,8 +24,8 @@ #include "oneapi/math/dft/detail/types_impl.hpp" #include "dft/backends/mklcpu/mklcpu_helpers.hpp" -// MKLCPU header -#include "mkl_dfti.h" +// Intel(R) oneMKL header +#include namespace oneapi { namespace math { diff --git a/src/dft/backends/mklcpu/forward.cpp b/src/dft/backends/mklcpu/forward.cpp index f915c9ce5..8c8b17f85 100644 --- a/src/dft/backends/mklcpu/forward.cpp +++ b/src/dft/backends/mklcpu/forward.cpp @@ -31,8 +31,8 @@ #include "dft/backends/mklcpu/commit_derived_impl.hpp" -// MKLCPU header -#include "mkl_dfti.h" +// Intel(R) oneMKL header +#include namespace oneapi::math::dft::mklcpu { namespace detail { diff --git a/src/dft/backends/mklcpu/mklcpu_helpers.hpp b/src/dft/backends/mklcpu/mklcpu_helpers.hpp index 89c3b6158..e8f9ff951 100644 --- a/src/dft/backends/mklcpu/mklcpu_helpers.hpp +++ b/src/dft/backends/mklcpu/mklcpu_helpers.hpp @@ -23,8 +23,8 @@ #include "oneapi/math/exceptions.hpp" #include "oneapi/math/dft/detail/types_impl.hpp" -// MKLCPU header -#include "mkl_dfti.h" +// Intel(R) oneMKL header +#include namespace oneapi::math::dft::mklcpu::detail { diff --git a/src/dft/backends/mklgpu/backward.cpp b/src/dft/backends/mklgpu/backward.cpp index ef0bf4836..f9d315d2f 100644 --- a/src/dft/backends/mklgpu/backward.cpp +++ b/src/dft/backends/mklgpu/backward.cpp @@ -31,8 +31,8 @@ #include "common_onemkl_conversion.hpp" #include "mklgpu_helpers.hpp" -#include "mkl_version.h" -// MKLGPU header +// Intel(R) oneMKL headers +#include #if INTEL_MKL_VERSION < 20250000 #include #else diff --git a/src/dft/backends/mklgpu/commit.cpp b/src/dft/backends/mklgpu/commit.cpp index a499ca163..9742baded 100644 --- a/src/dft/backends/mklgpu/commit.cpp +++ b/src/dft/backends/mklgpu/commit.cpp @@ -36,8 +36,8 @@ #include "dft/backends/mklgpu/mklgpu_helpers.hpp" #include "../stride_helper.hpp" -#include "mkl_version.h" -// MKLGPU header +// Intel(R) oneMKL headers +#include #if INTEL_MKL_VERSION < 20250000 #include #else diff --git a/src/dft/backends/mklgpu/forward.cpp b/src/dft/backends/mklgpu/forward.cpp index dd6325459..3ff8e9c9b 100644 --- a/src/dft/backends/mklgpu/forward.cpp +++ b/src/dft/backends/mklgpu/forward.cpp @@ -32,8 +32,8 @@ #include "common_onemkl_conversion.hpp" #include "mklgpu_helpers.hpp" -#include "mkl_version.h" -// MKLGPU header +// Intel(R) oneMKL headers +#include #if INTEL_MKL_VERSION < 20250000 #include #else diff --git a/src/dft/backends/mklgpu/mklgpu_helpers.hpp b/src/dft/backends/mklgpu/mklgpu_helpers.hpp index 4554ae622..0687b26e7 100644 --- a/src/dft/backends/mklgpu/mklgpu_helpers.hpp +++ b/src/dft/backends/mklgpu/mklgpu_helpers.hpp @@ -23,8 +23,8 @@ #include "oneapi/math/detail/exceptions.hpp" #include "oneapi/math/dft/detail/types_impl.hpp" -#include "mkl_version.h" -// MKLGPU header +// Intel(R) oneMKL headers +#include #if INTEL_MKL_VERSION < 20250000 #include #else diff --git a/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp b/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp index 076a9e19f..37fe92dad 100644 --- a/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp +++ b/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp @@ -28,8 +28,8 @@ #define WeirdNEC extern "C" { #ifdef USE_MKLREF -#include "mkl_cblas.h" -#include "mkl_lapacke.h" +#include +#include #else #include "cblas.h" #include "lapacke.h" From e13449c5bda2da0a76043b2521cde18cd57851a9 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 18 Oct 2024 16:53:46 +0100 Subject: [PATCH 33/41] Update CODEOWNERS --- .github/CODEOWNERS | 123 ++++++++++++++++++++++++--------------------- 1 file changed, 65 insertions(+), 58 deletions(-) diff --git a/.github/CODEOWNERS b/.github/CODEOWNERS index 2c2615e2b..5fb7c42d6 100644 --- a/.github/CODEOWNERS +++ b/.github/CODEOWNERS @@ -1,71 +1,78 @@ # Arch files -/.github/ @oneapi-src/onemkl-arch-write -/.gitignore/ @oneapi-src/onemkl-arch-write -/cmake/ @oneapi-src/onemkl-arch-write -/deps/ @oneapi-src/onemkl-arch-write -/docs/ @oneapi-src/onemkl-arch-write -/examples/include/ @oneapi-src/onemkl-arch-write -/examples/CMakeList.txt @oneapi-src/onemkl-arch-write -/examples/README.md @oneapi-src/onemkl-arch-write -/include/oneapi/mkl.hpp @oneapi-src/onemkl-arch-write -/include/oneapi/mkl/detail/ @oneapi-src/onemkl-arch-write -/include/oneapi/mkl/exceptions.hpp @oneapi-src/onemkl-arch-write -/scripts/ @oneapi-src/onemkl-arch-write -/src/include/ @oneapi-src/onemkl-arch-write -/src/CMakeLists.txt @oneapi-src/onemkl-arch-write -/src/config.hpp.in @oneapi-src/onemkl-arch-write -/tests/CMakeLists.txt @oneapi-src/onemkl-arch-write -/tests/README.md @oneapi-src/onemkl-arch-write -/tests/unit_tests/include/ @oneapi-src/onemkl-arch-write -/tests/unit_tests/CMakeLists.txt @oneapi-src/onemkl-arch-write -/tests/unit_tests/main_test.cpp @oneapi-src/onemkl-arch-write -/third-party-programs/ @oneapi-src/onemkl-arch-write -/CMakeLists.txt @oneapi-src/onemkl-arch-write -/CODE_OF_CONDUCT.md @oneapi-src/onemkl-arch-write -/CONTRIBUTING.md @oneapi-src/onemkl-arch-write -/LICENSE @oneapi-src/onemkl-arch-write -/MAINTAINERS.md @oneapi-src/onemkl-arch-write -/README.md @oneapi-src/onemkl-arch-write -/SECURITY.md @oneapi-src/onemkl-arch-write -/_clang-format @oneapi-src/onemkl-arch-write -/legal_information.md @oneapi-src/onemkl-arch-write +/.github/ @uxlfoundation/onemath-arch-write +/.gitignore/ @uxlfoundation/onemath-arch-write +/cmake/ @uxlfoundation/onemath-arch-write +/deps/ @uxlfoundation/onemath-arch-write +/docs/ @uxlfoundation/onemath-arch-write +/examples/include/ @uxlfoundation/onemath-arch-write +/examples/CMakeList.txt @uxlfoundation/onemath-arch-write +/examples/README.md @uxlfoundation/onemath-arch-write +/include/oneapi/math.hpp @uxlfoundation/onemath-arch-write +/include/oneapi/math/detail/ @uxlfoundation/onemath-arch-write +/include/oneapi/mkl.hpp @uxlfoundation/onemath-arch-write +/include/oneapi/mkl/namespace_alias.hpp @uxlfoundation/onemath-arch-write +/scripts/ @uxlfoundation/onemath-arch-write +/src/include/ @uxlfoundation/onemath-arch-write +/src/CMakeLists.txt @uxlfoundation/onemath-arch-write +/src/config.hpp.in @uxlfoundation/onemath-arch-write +/tests/CMakeLists.txt @uxlfoundation/onemath-arch-write +/tests/README.md @uxlfoundation/onemath-arch-write +/tests/unit_tests/include/ @uxlfoundation/onemath-arch-write +/tests/unit_tests/CMakeLists.txt @uxlfoundation/onemath-arch-write +/tests/unit_tests/main_test.cpp @uxlfoundation/onemath-arch-write +/third-party-programs/ @uxlfoundation/onemath-arch-write +/CMakeLists.txt @uxlfoundation/onemath-arch-write +/CODE_OF_CONDUCT.md @uxlfoundation/onemath-arch-write +/CONTRIBUTING.md @uxlfoundation/onemath-arch-write +/LICENSE @uxlfoundation/onemath-arch-write +/MAINTAINERS.md @uxlfoundation/onemath-arch-write +/README.md @uxlfoundation/onemath-arch-write +/SECURITY.md @uxlfoundation/onemath-arch-write +/_clang-format @uxlfoundation/onemath-arch-write +/legal_information.md @uxlfoundation/onemath-arch-write # BLAS files -/examples/blas/ @oneapi-src/onemkl-blas-write -/include/oneapi/mkl/blas/ @oneapi-src/onemkl-blas-write -/include/oneapi/mkl/blas.hxx @oneapi-src/onemkl-blas-write -/include/oneapi/mkl/blas.hpp @oneapi-src/onemkl-blas-write -/include/oneapi/mkl/bfloat16.hpp @oneapi-src/onemkl-blas-write -/src/blas/ @oneapi-src/onemkl-blas-write -/tests/unit_tests/blas/ @oneapi-src/onemkl-blas-write +/examples/blas/ @uxlfoundation/onemath-blas-write +/include/oneapi/math/blas/ @uxlfoundation/onemath-blas-write +/include/oneapi/math/blas.hpp @uxlfoundation/onemath-blas-write +/include/oneapi/math/blas.hxx @uxlfoundation/onemath-blas-write +/include/oneapi/math/bfloat16.hpp @uxlfoundation/onemath-blas-write +/include/oneapi/mkl/blas.hpp @uxlfoundation/onemath-blas-write +/src/blas/ @uxlfoundation/onemath-blas-write +/tests/unit_tests/blas/ @uxlfoundation/onemath-blas-write # DFT files -/examples/dft/ @oneapi-src/onemkl-dft-write -/include/oneapi/mkl/dft/ @oneapi-src/onemkl-dft-write -/include/oneapi/mkl/dft.hpp @oneapi-src/onemkl-dft-write -/src/dft/ @oneapi-src/onemkl-dft-write -/tests/unit_tests/dft/ @oneapi-src/onemkl-dft-write +/examples/dft/ @uxlfoundation/onemath-dft-write +/include/oneapi/math/dft/ @uxlfoundation/onemath-dft-write +/include/oneapi/math/dft.hpp @uxlfoundation/onemath-dft-write +/include/oneapi/mkl/dft.hpp @uxlfoundation/onemath-dft-write +/src/dft/ @uxlfoundation/onemath-dft-write +/tests/unit_tests/dft/ @uxlfoundation/onemath-dft-write # LAPACK files -/examples/lapack/ @oneapi-src/onemkl-lapack-write -/include/oneapi/mkl/lapack/ @oneapi-src/onemkl-lapack-write -/include/oneapi/mkl/lapack.hpp @oneapi-src/onemkl-lapack-write -/src/lapack/ @oneapi-src/onemkl-lapack-write -/tests/unit_tests/lapack/ @oneapi-src/onemkl-lapack-write +/examples/lapack/ @uxlfoundation/onemath-lapack-write +/include/oneapi/math/lapack/ @uxlfoundation/onemath-lapack-write +/include/oneapi/math/lapack.hpp @uxlfoundation/onemath-lapack-write +/include/oneapi/mkl/lapack.hpp @uxlfoundation/onemath-lapack-write +/src/lapack/ @uxlfoundation/onemath-lapack-write +/tests/unit_tests/lapack/ @uxlfoundation/onemath-lapack-write # RNG files -/examples/rng/ @oneapi-src/onemkl-rng-write -/include/oneapi/mkl/rng/ @oneapi-src/onemkl-rng-write -/include/oneapi/mkl/rng.hpp @oneapi-src/onemkl-rng-write -/src/rng/ @oneapi-src/onemkl-rng-write -/tests/unit_tests/rng/ @oneapi-src/onemkl-rng-write +/examples/rng/ @uxlfoundation/onemath-rng-write +/include/oneapi/math/rng/ @uxlfoundation/onemath-rng-write +/include/oneapi/math/rng.hpp @uxlfoundation/onemath-rng-write +/include/oneapi/mkl/rng.hpp @uxlfoundation/onemath-rng-write +/include/oneapi/mkl/rng/device.hpp @uxlfoundation/onemath-rng-write +/src/rng/ @uxlfoundation/onemath-rng-write +/tests/unit_tests/rng/ @uxlfoundation/onemath-rng-write # Sparse BLAS files -/examples/sparse_blas/ @oneapi-src/onemkl-sparse-write -/include/oneapi/mkl/sparse_blas/ @oneapi-src/onemkl-sparse-write -/include/oneapi/mkl/sparse_blas.hpp @oneapi-src/onemkl-sparse-write -/src/sparse_blas/ @oneapi-src/onemkl-sparse-write -/tests/unit_tests/sparse_blas/ @oneapi-src/onemkl-sparse-write +/examples/sparse_blas/ @uxlfoundation/onemath-sparse-write +/include/oneapi/math/sparse_blas/ @uxlfoundation/onemath-sparse-write +/include/oneapi/math/sparse_blas.hpp @uxlfoundation/onemath-sparse-write +/include/oneapi/mkl/sparse_blas.hpp @uxlfoundation/onemath-sparse-write +/src/sparse_blas/ @uxlfoundation/onemath-sparse-write +/tests/unit_tests/sparse_blas/ @uxlfoundation/onemath-sparse-write # Shared files -/include/oneapi/mkl/types.hpp @oneapi-src/onemkl-blas-write @oneapi-src/onemkl-lapack-write +/include/oneapi/math/types.hpp @uxlfoundation/onemath-blas-write @uxlfoundation/onemath-lapack-write From 0f677b420b6eed834a75adece84bc0722b1a3dc1 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 18 Oct 2024 17:04:42 +0100 Subject: [PATCH 34/41] Use oneAPI Math Library (oneMath) --- CONTRIBUTING.md | 2 +- README.md | 6 +++--- docs/create_new_backend.rst | 4 ++-- third-party-programs/THIRD-PARTY-PROGRAMS | 2 +- 4 files changed, 7 insertions(+), 7 deletions(-) diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 8f90769d9..e33e319a0 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -1,5 +1,5 @@ # Contributing Guidelines -If you have improvements, new libraries integrated under oneMath, or new interfaces to contribute to the oneMath Specification, please send us your pull requests! For getting started, see GitHub [howto](https://help.github.com/en/articles/about-pull-requests). +If you have improvements, new libraries integrated under oneAPI Math Library (oneMath), or new interfaces to contribute to the oneMath Specification, please send us your pull requests! For getting started, see GitHub [howto](https://help.github.com/en/articles/about-pull-requests). For how to enable a new third-party library, see the [guidelines](docs/create_new_backend.rst). diff --git a/README.md b/README.md index 7ab3d1c0e..de1a60fbd 100644 --- a/README.md +++ b/README.md @@ -1,6 +1,6 @@ UXL Foundation Logo -# oneMath +# oneAPI Math Library (oneMath) oneMath is an open-source implementation of the [oneMath specification](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemath/source/). It can work with multiple devices using multiple libraries (backends) underneath. The oneMath project was previously referred to as oneMKL Interface. @@ -573,13 +573,13 @@ Distributed under the Apache license 2.0. See [LICENSE](LICENSE) for more inform **Q: What is the difference between the following items?** - The [oneAPI Specification for oneMath](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemath/source/) - - The [oneMath](https://github.com/uxlfoundation/oneMath) Project + - The [oneAPI Math Library (oneMath)](https://github.com/uxlfoundation/oneMath) project - The [Intel(R) oneAPI Math Kernel Library (oneMKL)](https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onemkl.html) Product **A:** - The [oneAPI Specification for oneMath](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemath/source/) defines the SYCL interfaces for performance math library functions. The oneMath specification can evolve faster and more frequently than implementations of the specification. -- The [oneMath](https://github.com/uxlfoundation/oneMath) project is an open source implementation of the specification. The project goal is to demonstrate how the SYCL interfaces documented in the oneMath specification can be implemented for any math library and work for any target hardware. While the implementation provided here may not yet be the full implementation of the specification, the goal is to build it out over time. We encourage the community to contribute to this project and help to extend support to multiple hardware targets and other math libraries. +- The [oneAPI Math Library (oneMath)](https://github.com/uxlfoundation/oneMath) project is an open source implementation of the specification. The project goal is to demonstrate how the SYCL interfaces documented in the oneMath specification can be implemented for any math library and work for any target hardware. While the implementation provided here may not yet be the full implementation of the specification, the goal is to build it out over time. We encourage the community to contribute to this project and help to extend support to multiple hardware targets and other math libraries. - The [Intel(R) oneAPI Math Kernel Library (oneMKL)](https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onemkl.html) project is an Intel product provided as part of the Intel(R) oneAPI Base Toolkit. It is used for the Intel backends of oneMath. Its C++ API is very similar to the oneMath specification. It is highly optimized for Intel CPU and Intel GPU hardware. diff --git a/docs/create_new_backend.rst b/docs/create_new_backend.rst index 8b0d10055..374debc3b 100644 --- a/docs/create_new_backend.rst +++ b/docs/create_new_backend.rst @@ -3,8 +3,8 @@ .. _create_backend_wrappers: -Integrating a Third-Party Library to oneMath -============================================ +Integrating a Third-Party Library to oneAPI Math Library (oneMath) +================================================================== This step-by-step tutorial provides examples for enabling new third-party libraries in oneMath. diff --git a/third-party-programs/THIRD-PARTY-PROGRAMS b/third-party-programs/THIRD-PARTY-PROGRAMS index 1de26e969..a75bbd2fe 100644 --- a/third-party-programs/THIRD-PARTY-PROGRAMS +++ b/third-party-programs/THIRD-PARTY-PROGRAMS @@ -1,4 +1,4 @@ -oneMath +oneAPI Math Library (oneMath) This file contains the list of third party software (“third party programs”) contained in the Intel software and their required notices and/or license terms. From ca37c3860fd8d6f81fb441c9a0fee78a9dc95bbe Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 18 Oct 2024 17:18:31 +0100 Subject: [PATCH 35/41] clang-format-19.1 --- .pre-commit-config.yaml | 8 + _clang-format | 96 +- .../level3/gemm_usm_mklcpu_cublas.cpp | 36 +- .../run_time_dispatching/level3/gemm_usm.cpp | 2 +- .../complex_fwd_usm_mklcpu_cufft.cpp | 2 +- .../dft/run_time_dispatching/real_fwd_usm.cpp | 2 +- examples/include/example_helper.hpp | 6 +- .../getrs_usm_mklcpu_cusolver.cpp | 12 +- .../lapack/run_time_dispatching/getrs_usm.cpp | 4 +- .../sparse_blas_spmv_usm_mklcpu.cpp | 42 +- .../sparse_blas_spmv_usm.cpp | 42 +- include/oneapi/math/bfloat16.hpp | 50 +- include/oneapi/math/blas.hxx | 4028 +++++---- .../math/blas/detail/blas_ct_backends.hxx | 3471 ++++---- .../oneapi/math/blas/detail/blas_loader.hxx | 5164 ++++++------ .../math/blas/detail/cublas/blas_ct.hxx | 3865 +++++---- .../detail/cublas/onemath_blas_cublas.hxx | 3959 +++++---- .../math/blas/detail/mklcpu/blas_ct.hxx | 3865 +++++---- .../math/blas/detail/mklgpu/blas_ct.hxx | 3865 +++++---- .../math/blas/detail/netlib/blas_ct.hxx | 3870 +++++---- .../blas/detail/onemath_blas_backends.hxx | 5616 +++++++------ .../math/blas/detail/portblas/blas_ct.hxx | 3612 ++++---- .../math/blas/detail/rocblas/blas_ct.hxx | 3412 ++++---- .../detail/rocblas/onemath_blas_rocblas.hxx | 3426 ++++---- include/oneapi/math/detail/exceptions.hpp | 13 +- include/oneapi/math/detail/get_device_id.hpp | 2 +- include/oneapi/math/dft/backward.hpp | 54 +- .../oneapi/math/dft/detail/commit_impl.hpp | 106 +- include/oneapi/math/dft/detail/dft_ct.hxx | 108 +- include/oneapi/math/dft/detail/dft_loader.hpp | 2 +- .../dft/detail/external_workspace_helper.hpp | 4 +- include/oneapi/math/dft/detail/types_impl.hpp | 2 +- include/oneapi/math/dft/forward.hpp | 52 +- include/oneapi/math/exceptions.hpp | 40 +- .../math/lapack/detail/cusolver/lapack_ct.hxx | 2780 +++---- .../cusolver/onemath_lapack_cusolver.hxx | 3436 ++++---- .../math/lapack/detail/lapack_loader.hpp | 3654 +++++---- .../oneapi/math/lapack/detail/lapack_rt.hpp | 2279 +++--- .../lapack/detail/mkl_common/lapack_ct.hxx | 2648 +++--- .../mkl_common/onemath_lapack_backends.hxx | 3422 ++++---- .../lapack/detail/rocsolver/lapack_ct.hxx | 2891 +++---- .../rocsolver/onemath_lapack_rocsolver.hxx | 3445 ++++---- include/oneapi/math/lapack/exceptions.hpp | 16 +- .../rng/detail/curand/onemath_rng_curand.hpp | 4 +- .../rng/detail/mklcpu/onemath_rng_mklcpu.hpp | 4 +- .../rng/detail/mklgpu/onemath_rng_mklgpu.hpp | 4 +- include/oneapi/math/rng/detail/rng_loader.hpp | 8 +- .../detail/rocrand/onemath_rng_rocrand.hpp | 4 +- .../rng/device/detail/exponential_impl.hpp | 3 +- .../math/rng/device/detail/gamma_impl.hpp | 6 +- .../math/rng/device/detail/mrg32k3a_impl.hpp | 5 +- .../rng/device/detail/philox4x32x10_impl.hpp | 10 +- include/oneapi/math/rng/distributions.hpp | 6 +- .../math/sparse_blas/detail/helper_types.hpp | 2 +- .../detail/onemath_sparse_blas_backends.hxx | 315 +- .../sparse_blas/detail/sparse_blas_ct.hxx | 122 +- .../sparse_blas/detail/sparse_blas_rt.hpp | 146 +- include/oneapi/mkl/blas.hpp | 3 +- include/oneapi/mkl/dft.hpp | 3 +- include/oneapi/mkl/lapack.hpp | 3 +- include/oneapi/mkl/namespace_alias.hpp | 9 +- include/oneapi/mkl/rng.hpp | 3 +- include/oneapi/mkl/rng/device.hpp | 3 +- include/oneapi/mkl/sparse_blas.hpp | 3 +- src/blas/backends/cublas/cublas_batch.cpp | 1646 ++-- .../backends/cublas/cublas_extensions.cpp | 592 +- src/blas/backends/cublas/cublas_handle.hpp | 4 +- src/blas/backends/cublas/cublas_helper.hpp | 21 +- src/blas/backends/cublas/cublas_level1.cpp | 874 +- src/blas/backends/cublas/cublas_level2.cpp | 1596 ++-- src/blas/backends/cublas/cublas_level3.cpp | 820 +- .../backends/cublas/cublas_scope_handle.cpp | 12 +- .../backends/cublas/cublas_scope_handle.hpp | 14 +- .../cublas/cublas_scope_handle_hipsycl.cpp | 6 +- .../cublas/cublas_scope_handle_hipsycl.hpp | 8 +- src/blas/backends/cublas/cublas_task.hpp | 8 +- src/blas/backends/mkl_common/mkl_batch.cxx | 1648 ++-- .../backends/mkl_common/mkl_blas_backend.hxx | 3876 ++++----- .../backends/mkl_common/mkl_extensions.cxx | 580 +- src/blas/backends/mkl_common/mkl_level1.cxx | 592 +- src/blas/backends/mkl_common/mkl_level2.cxx | 1334 +-- src/blas/backends/mkl_common/mkl_level3.cxx | 985 ++- src/blas/backends/netlib/netlib_batch.cxx | 970 ++- src/blas/backends/netlib/netlib_common.hpp | 6 +- .../backends/netlib/netlib_extensions.cxx | 346 +- src/blas/backends/netlib/netlib_level1.cpp | 20 +- src/blas/backends/netlib/netlib_level1.cxx | 746 +- src/blas/backends/netlib/netlib_level2.cxx | 1186 +-- src/blas/backends/netlib/netlib_level3.cxx | 700 +- src/blas/backends/portblas/portblas_batch.cxx | 993 +-- .../backends/portblas/portblas_common.hpp | 6 +- .../backends/portblas/portblas_gemm_bias.cxx | 72 +- .../backends/portblas/portblas_level1.cxx | 298 +- .../backends/portblas/portblas_level2.cxx | 456 +- .../backends/portblas/portblas_level3.cxx | 403 +- .../portblas/portblas_level3_bfloat16.cpp | 32 +- .../portblas/portblas_level3_half.cpp | 56 +- src/blas/backends/rocblas/rocblas_batch.cpp | 1444 ++-- .../backends/rocblas/rocblas_extensions.cpp | 544 +- src/blas/backends/rocblas/rocblas_handle.hpp | 4 +- src/blas/backends/rocblas/rocblas_helper.hpp | 12 +- src/blas/backends/rocblas/rocblas_level1.cpp | 830 +- src/blas/backends/rocblas/rocblas_level2.cpp | 2028 +++-- src/blas/backends/rocblas/rocblas_level3.cpp | 900 +- .../backends/rocblas/rocblas_scope_handle.cpp | 14 +- .../backends/rocblas/rocblas_scope_handle.hpp | 14 +- .../rocblas/rocblas_scope_handle_hipsycl.cpp | 8 +- .../rocblas/rocblas_scope_handle_hipsycl.hpp | 10 +- src/blas/backends/rocblas/rocblas_task.hpp | 8 +- src/blas/blas_loader.cpp | 7018 ++++++++-------- src/blas/function_table.hpp | 7286 +++++++++-------- .../backend_backward_instantiations.cxx | 38 +- .../backends/backend_compute_signature.cxx | 7 +- .../backend_forward_instantiations.cxx | 38 +- src/dft/backends/cufft/backward.cpp | 86 +- src/dft/backends/cufft/commit.cpp | 14 +- src/dft/backends/cufft/execute_helper.hpp | 66 +- src/dft/backends/cufft/forward.cpp | 81 +- src/dft/backends/descriptor.cpp | 10 +- src/dft/backends/mklcpu/backward.cpp | 107 +- src/dft/backends/mklcpu/commit.cpp | 8 +- src/dft/backends/mklcpu/forward.cpp | 105 +- src/dft/backends/mklcpu/mklcpu_helpers.hpp | 12 +- src/dft/backends/mklgpu/backward.cpp | 74 +- src/dft/backends/mklgpu/commit.cpp | 14 +- src/dft/backends/mklgpu/forward.cpp | 70 +- src/dft/backends/mklgpu/mklgpu_helpers.hpp | 15 +- src/dft/backends/portfft/commit.cpp | 30 +- src/dft/backends/portfft/portfft_helper.hpp | 10 +- src/dft/backends/rocfft/backward.cpp | 160 +- src/dft/backends/rocfft/commit.cpp | 44 +- src/dft/backends/rocfft/execute_helper.hpp | 28 +- src/dft/backends/rocfft/forward.cpp | 157 +- src/dft/backends/stride_helper.hpp | 4 +- src/dft/descriptor.cxx | 6 +- src/dft/descriptor_config_helper.hpp | 12 +- src/dft/execute_helper_generic.hpp | 4 +- src/dft/function_table.hpp | 18 +- src/include/allocator_helper.hpp | 4 +- src/include/common_onemkl_conversion.hpp | 179 +- src/include/function_table_initializer.hpp | 10 +- .../backends/cusolver/cusolver_batch.cpp | 1014 +-- .../backends/cusolver/cusolver_handle.hpp | 4 +- .../backends/cusolver/cusolver_helper.hpp | 27 +- .../backends/cusolver/cusolver_lapack.cpp | 2417 +++--- .../cusolver/cusolver_scope_handle.cpp | 12 +- .../cusolver/cusolver_scope_handle.hpp | 14 +- .../backends/cusolver/cusolver_task.hpp | 6 +- src/lapack/backends/mkl_common/mkl_lapack.cxx | 4520 +++++----- .../mkl_common/mkl_lapack_backend.hpp | 2032 ++--- .../backends/rocsolver/rocsolver_batch.cpp | 932 +-- .../backends/rocsolver/rocsolver_handle.hpp | 4 +- .../backends/rocsolver/rocsolver_helper.hpp | 17 +- .../backends/rocsolver/rocsolver_lapack.cpp | 1763 ++-- .../rocsolver/rocsolver_scope_handle.cpp | 12 +- .../rocsolver/rocsolver_scope_handle.hpp | 12 +- .../backends/rocsolver/rocsolver_task.hpp | 6 +- src/lapack/function_table.hpp | 2672 +++--- src/lapack/lapack_loader.cpp | 2513 +++--- src/rng/backends/curand/curand_task.hpp | 20 +- src/rng/backends/curand/mrg32k3a.cpp | 70 +- src/rng/backends/curand/philox4x32x10.cpp | 71 +- src/rng/backends/mklcpu/cpu_common.hpp | 8 +- src/rng/backends/mklcpu/mrg32k3a.cpp | 2 +- src/rng/backends/mklcpu/philox4x32x10.cpp | 3 +- src/rng/backends/mklgpu/mrg32k3a.cpp | 157 +- .../mklgpu/onemkl_distribution_conversion.hpp | 50 +- src/rng/backends/mklgpu/philox4x32x10.cpp | 157 +- src/rng/backends/rocrand/mrg32k3a.cpp | 68 +- src/rng/backends/rocrand/philox4x32x10.cpp | 69 +- src/rng/backends/rocrand/rocrand_task.hpp | 20 +- src/rng/function_table.hpp | 4 +- .../backends/mkl_common/mkl_handles.cxx | 276 +- .../backends/mkl_common/mkl_helper.hpp | 18 +- .../backends/mkl_common/mkl_spmm.cxx | 83 +- .../backends/mkl_common/mkl_spmv.cxx | 101 +- .../backends/mkl_common/mkl_spsv.cxx | 68 +- src/sparse_blas/function_table.hpp | 225 +- src/sparse_blas/generic_container.hpp | 8 +- src/sparse_blas/macros.hpp | 6 +- src/sparse_blas/sparse_blas_loader.cpp | 152 +- .../blas/batch/axpy_batch_stride.cpp | 28 +- .../blas/batch/axpy_batch_stride_usm.cpp | 24 +- .../unit_tests/blas/batch/axpy_batch_usm.cpp | 64 +- .../blas/batch/copy_batch_stride.cpp | 26 +- .../blas/batch/copy_batch_stride_usm.cpp | 28 +- .../unit_tests/blas/batch/copy_batch_usm.cpp | 66 +- .../blas/batch/dgmm_batch_stride.cpp | 33 +- .../blas/batch/dgmm_batch_stride_usm.cpp | 23 +- .../unit_tests/blas/batch/dgmm_batch_usm.cpp | 75 +- .../blas/batch/gemm_batch_stride.cpp | 27 +- .../blas/batch/gemm_batch_stride_usm.cpp | 35 +- .../unit_tests/blas/batch/gemm_batch_usm.cpp | 91 +- .../blas/batch/gemv_batch_stride.cpp | 27 +- .../blas/batch/gemv_batch_stride_usm.cpp | 24 +- .../unit_tests/blas/batch/gemv_batch_usm.cpp | 76 +- .../blas/batch/imatcopy_batch_stride.cpp | 22 +- .../blas/batch/imatcopy_batch_stride_usm.cpp | 30 +- .../blas/batch/imatcopy_batch_usm.cpp | 38 +- .../blas/batch/omatadd_batch_stride.cpp | 16 +- .../blas/batch/omatadd_batch_stride_usm.cpp | 24 +- .../blas/batch/omatcopy_batch_stride.cpp | 30 +- .../blas/batch/omatcopy_batch_stride_usm.cpp | 28 +- .../blas/batch/omatcopy_batch_usm.cpp | 48 +- .../blas/batch/syrk_batch_stride.cpp | 42 +- .../blas/batch/syrk_batch_stride_usm.cpp | 36 +- .../unit_tests/blas/batch/syrk_batch_usm.cpp | 68 +- .../blas/batch/trsm_batch_stride.cpp | 23 +- .../blas/batch/trsm_batch_stride_usm.cpp | 23 +- .../unit_tests/blas/batch/trsm_batch_usm.cpp | 71 +- .../unit_tests/blas/extensions/gemm_bias.cpp | 23 +- .../blas/extensions/gemm_bias_usm.cpp | 13 +- tests/unit_tests/blas/extensions/gemmt.cpp | 76 +- .../unit_tests/blas/extensions/gemmt_usm.cpp | 68 +- tests/unit_tests/blas/extensions/imatcopy.cpp | 26 +- .../blas/extensions/imatcopy_usm.cpp | 24 +- tests/unit_tests/blas/extensions/omatadd.cpp | 40 +- .../blas/extensions/omatadd_usm.cpp | 24 +- tests/unit_tests/blas/extensions/omatcopy.cpp | 20 +- .../unit_tests/blas/extensions/omatcopy2.cpp | 27 +- .../blas/extensions/omatcopy2_usm.cpp | 24 +- .../blas/extensions/omatcopy_usm.cpp | 20 +- .../blas/include/onemath_blas_helper.hpp | 2 +- .../blas/include/reference_blas_templates.hpp | 1302 ++- .../blas/include/reference_blas_wrappers.hpp | 1470 ++-- tests/unit_tests/blas/include/test_common.hpp | 79 +- tests/unit_tests/blas/level1/asum.cpp | 7 +- tests/unit_tests/blas/level1/asum_usm.cpp | 8 +- tests/unit_tests/blas/level1/axpby.cpp | 24 +- tests/unit_tests/blas/level1/axpby_usm.cpp | 24 +- tests/unit_tests/blas/level1/axpy.cpp | 27 +- tests/unit_tests/blas/level1/axpy_usm.cpp | 27 +- tests/unit_tests/blas/level1/copy.cpp | 6 +- tests/unit_tests/blas/level1/copy_usm.cpp | 8 +- tests/unit_tests/blas/level1/dot.cpp | 7 +- tests/unit_tests/blas/level1/dot_usm.cpp | 6 +- tests/unit_tests/blas/level1/dotc.cpp | 28 +- tests/unit_tests/blas/level1/dotc_usm.cpp | 26 +- tests/unit_tests/blas/level1/dotu.cpp | 28 +- tests/unit_tests/blas/level1/dotu_usm.cpp | 26 +- tests/unit_tests/blas/level1/iamax.cpp | 6 +- tests/unit_tests/blas/level1/iamax_usm.cpp | 4 +- tests/unit_tests/blas/level1/iamin.cpp | 6 +- tests/unit_tests/blas/level1/iamin_usm.cpp | 4 +- tests/unit_tests/blas/level1/nrm2.cpp | 7 +- tests/unit_tests/blas/level1/nrm2_usm.cpp | 8 +- tests/unit_tests/blas/level1/rot.cpp | 24 +- tests/unit_tests/blas/level1/rot_usm.cpp | 26 +- tests/unit_tests/blas/level1/rotg.cpp | 26 +- tests/unit_tests/blas/level1/rotg_usm.cpp | 44 +- tests/unit_tests/blas/level1/rotm.cpp | 30 +- tests/unit_tests/blas/level1/rotm_usm.cpp | 26 +- tests/unit_tests/blas/level1/rotmg.cpp | 10 +- tests/unit_tests/blas/level1/rotmg_usm.cpp | 38 +- tests/unit_tests/blas/level1/scal.cpp | 4 +- tests/unit_tests/blas/level1/scal_usm.cpp | 8 +- tests/unit_tests/blas/level1/sdsdot.cpp | 26 +- tests/unit_tests/blas/level1/sdsdot_usm.cpp | 26 +- tests/unit_tests/blas/level1/swap.cpp | 6 +- tests/unit_tests/blas/level1/swap_usm.cpp | 8 +- tests/unit_tests/blas/level2/gbmv.cpp | 52 +- tests/unit_tests/blas/level2/gbmv_usm.cpp | 52 +- tests/unit_tests/blas/level2/gemv.cpp | 65 +- tests/unit_tests/blas/level2/gemv_usm.cpp | 51 +- tests/unit_tests/blas/level2/ger.cpp | 26 +- tests/unit_tests/blas/level2/ger_usm.cpp | 28 +- tests/unit_tests/blas/level2/gerc.cpp | 28 +- tests/unit_tests/blas/level2/gerc_usm.cpp | 34 +- tests/unit_tests/blas/level2/geru.cpp | 28 +- tests/unit_tests/blas/level2/geru_usm.cpp | 34 +- tests/unit_tests/blas/level2/hbmv.cpp | 48 +- tests/unit_tests/blas/level2/hbmv_usm.cpp | 50 +- tests/unit_tests/blas/level2/hemv.cpp | 32 +- tests/unit_tests/blas/level2/hemv_usm.cpp | 34 +- tests/unit_tests/blas/level2/her.cpp | 24 +- tests/unit_tests/blas/level2/her2.cpp | 29 +- tests/unit_tests/blas/level2/her2_usm.cpp | 34 +- tests/unit_tests/blas/level2/her_usm.cpp | 18 +- tests/unit_tests/blas/level2/hpmv.cpp | 35 +- tests/unit_tests/blas/level2/hpmv_usm.cpp | 40 +- tests/unit_tests/blas/level2/hpr.cpp | 64 +- tests/unit_tests/blas/level2/hpr2.cpp | 29 +- tests/unit_tests/blas/level2/hpr2_usm.cpp | 34 +- tests/unit_tests/blas/level2/hpr_usm.cpp | 64 +- tests/unit_tests/blas/level2/sbmv.cpp | 32 +- tests/unit_tests/blas/level2/sbmv_usm.cpp | 34 +- tests/unit_tests/blas/level2/spmv.cpp | 29 +- tests/unit_tests/blas/level2/spmv_usm.cpp | 34 +- tests/unit_tests/blas/level2/spr.cpp | 24 +- tests/unit_tests/blas/level2/spr2.cpp | 29 +- tests/unit_tests/blas/level2/spr2_usm.cpp | 34 +- tests/unit_tests/blas/level2/spr_usm.cpp | 24 +- tests/unit_tests/blas/level2/symv.cpp | 32 +- tests/unit_tests/blas/level2/symv_usm.cpp | 34 +- tests/unit_tests/blas/level2/syr.cpp | 24 +- tests/unit_tests/blas/level2/syr2.cpp | 29 +- tests/unit_tests/blas/level2/syr2_usm.cpp | 34 +- tests/unit_tests/blas/level2/syr_usm.cpp | 18 +- tests/unit_tests/blas/level2/tbmv.cpp | 14 +- tests/unit_tests/blas/level2/tbmv_usm.cpp | 14 +- tests/unit_tests/blas/level2/tbsv.cpp | 14 +- tests/unit_tests/blas/level2/tbsv_usm.cpp | 14 +- tests/unit_tests/blas/level2/tpmv.cpp | 13 +- tests/unit_tests/blas/level2/tpmv_usm.cpp | 14 +- tests/unit_tests/blas/level2/tpsv.cpp | 13 +- tests/unit_tests/blas/level2/tpsv_usm.cpp | 14 +- tests/unit_tests/blas/level2/trmv.cpp | 17 +- tests/unit_tests/blas/level2/trmv_usm.cpp | 17 +- tests/unit_tests/blas/level2/trsv.cpp | 17 +- tests/unit_tests/blas/level2/trsv_usm.cpp | 17 +- tests/unit_tests/blas/level3/gemm.cpp | 8 +- tests/unit_tests/blas/level3/gemm_usm.cpp | 8 +- tests/unit_tests/blas/level3/hemm.cpp | 46 +- tests/unit_tests/blas/level3/hemm_usm.cpp | 48 +- tests/unit_tests/blas/level3/her2k.cpp | 12 +- tests/unit_tests/blas/level3/her2k_usm.cpp | 8 +- tests/unit_tests/blas/level3/herk.cpp | 9 +- tests/unit_tests/blas/level3/herk_usm.cpp | 14 +- tests/unit_tests/blas/level3/symm.cpp | 54 +- tests/unit_tests/blas/level3/symm_usm.cpp | 56 +- tests/unit_tests/blas/level3/syr2k.cpp | 23 +- tests/unit_tests/blas/level3/syr2k_usm.cpp | 19 +- tests/unit_tests/blas/level3/syrk.cpp | 17 +- tests/unit_tests/blas/level3/syrk_usm.cpp | 22 +- tests/unit_tests/blas/level3/trmm.cpp | 174 +- tests/unit_tests/blas/level3/trmm_usm.cpp | 166 +- tests/unit_tests/blas/level3/trsm.cpp | 286 +- tests/unit_tests/blas/level3/trsm_usm.cpp | 278 +- .../dft/include/compute_inplace.hpp | 25 +- .../dft/include/compute_inplace_real_real.hpp | 8 +- .../dft/include/compute_out_of_place.hpp | 29 +- .../compute_out_of_place_real_real.hpp | 4 +- .../unit_tests/dft/include/reference_dft.hpp | 10 +- tests/unit_tests/dft/include/test_common.hpp | 33 +- tests/unit_tests/dft/source/compute_tests.cpp | 24 +- .../dft/source/descriptor_tests.cpp | 41 +- tests/unit_tests/include/test_helper.hpp | 23 +- .../lapack/include/lapack_accuracy_checks.hpp | 25 +- .../include/lapack_reference_wrappers.hpp | 753 +- tests/unit_tests/lapack/source/gebrd.cpp | 6 +- tests/unit_tests/lapack/source/geqrf.cpp | 4 +- .../lapack/source/geqrf_batch_group.cpp | 4 +- .../lapack/source/geqrf_batch_stride.cpp | 9 +- tests/unit_tests/lapack/source/gerqf.cpp | 4 +- tests/unit_tests/lapack/source/gesvd.cpp | 19 +- tests/unit_tests/lapack/source/getrf.cpp | 4 +- .../lapack/source/getrf_batch_group.cpp | 4 +- .../lapack/source/getrf_batch_stride.cpp | 9 +- tests/unit_tests/lapack/source/getri.cpp | 5 +- .../lapack/source/getri_batch_group.cpp | 8 +- .../lapack/source/getri_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/getrs.cpp | 2 +- .../lapack/source/getrs_batch_group.cpp | 8 +- .../lapack/source/getrs_batch_stride.cpp | 12 +- tests/unit_tests/lapack/source/heevd.cpp | 4 +- tests/unit_tests/lapack/source/hegvd.cpp | 24 +- tests/unit_tests/lapack/source/hetrd.cpp | 2 +- tests/unit_tests/lapack/source/hetrf.cpp | 28 +- tests/unit_tests/lapack/source/orgbr.cpp | 4 +- tests/unit_tests/lapack/source/orgqr.cpp | 4 +- .../lapack/source/orgqr_batch_group.cpp | 4 +- .../lapack/source/orgqr_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/orgtr.cpp | 4 +- tests/unit_tests/lapack/source/ormqr.cpp | 6 +- tests/unit_tests/lapack/source/ormrq.cpp | 6 +- tests/unit_tests/lapack/source/ormtr.cpp | 10 +- tests/unit_tests/lapack/source/potrf.cpp | 2 +- .../lapack/source/potrf_batch_group.cpp | 6 +- .../lapack/source/potrf_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/potri.cpp | 2 +- tests/unit_tests/lapack/source/potrs.cpp | 12 +- .../lapack/source/potrs_batch_group.cpp | 8 +- .../lapack/source/potrs_batch_stride.cpp | 7 +- tests/unit_tests/lapack/source/syevd.cpp | 4 +- tests/unit_tests/lapack/source/sygvd.cpp | 30 +- tests/unit_tests/lapack/source/sytrd.cpp | 2 +- tests/unit_tests/lapack/source/sytrf.cpp | 28 +- tests/unit_tests/lapack/source/trtrs.cpp | 8 +- tests/unit_tests/lapack/source/ungbr.cpp | 4 +- tests/unit_tests/lapack/source/ungqr.cpp | 4 +- .../lapack/source/ungqr_batch_group.cpp | 4 +- .../lapack/source/ungqr_batch_stride.cpp | 2 +- tests/unit_tests/lapack/source/ungtr.cpp | 4 +- tests/unit_tests/lapack/source/unmqr.cpp | 6 +- tests/unit_tests/lapack/source/unmrq.cpp | 6 +- tests/unit_tests/lapack/source/unmtr.cpp | 10 +- tests/unit_tests/main_test.cpp | 22 +- .../unit_tests/rng/device/include/moments.hpp | 12 +- .../device/include/rng_device_test_common.hpp | 7 +- .../rng/device/include/skip_ahead_test.hpp | 2 +- .../unit_tests/rng/device/moments/moments.cpp | 54 +- .../rng/include/rng_test_common.hpp | 7 +- .../rng/statistics_check/gaussian.cpp | 24 +- .../rng/statistics_check/poisson_usm.cpp | 8 +- .../include/common_sparse_reference.hpp | 14 +- .../sparse_blas/include/test_common.hpp | 72 +- .../sparse_blas/include/test_spmm.hpp | 32 +- .../sparse_blas/include/test_spmv.hpp | 26 +- .../sparse_blas/include/test_spsv.hpp | 28 +- .../sparse_blas/source/sparse_spmm_buffer.cpp | 17 +- .../sparse_blas/source/sparse_spmm_usm.cpp | 37 +- .../sparse_blas/source/sparse_spmv_buffer.cpp | 21 +- .../sparse_blas/source/sparse_spmv_usm.cpp | 31 +- .../sparse_blas/source/sparse_spsv_buffer.cpp | 21 +- .../sparse_blas/source/sparse_spsv_usm.cpp | 29 +- 405 files changed, 73086 insertions(+), 72570 deletions(-) create mode 100644 .pre-commit-config.yaml mode change 100755 => 100644 tests/unit_tests/rng/statistics_check/poisson_usm.cpp diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml new file mode 100644 index 000000000..1be15eecd --- /dev/null +++ b/.pre-commit-config.yaml @@ -0,0 +1,8 @@ +repos: + +- repo: https://github.com/pre-commit/mirrors-clang-format + rev: v19.1.0 + hooks: + - id: clang-format + files: \.(c|cxx|cpp|h|hxx|hpp)$ + exclude: ^deps/ diff --git a/_clang-format b/_clang-format index 37a50f367..aef7d8e0f 100644 --- a/_clang-format +++ b/_clang-format @@ -21,31 +21,31 @@ Language: Cpp AccessModifierOffset: -4 AlignAfterOpenBracket: Align -AlignConsecutiveAssignments: false -AlignConsecutiveDeclarations: false -AlignConsecutiveMacros: true +AlignConsecutiveAssignments: + Enabled: false +AlignConsecutiveDeclarations: + Enabled: false +AlignConsecutiveMacros: + Enabled: true AlignEscapedNewlines: Left -AlignOperands: true -AlignTrailingComments: false +AlignOperands: Align +AlignTrailingComments: + Kind: Never AllowAllArgumentsOnNextLine: true -AllowAllConstructorInitializersOnNextLine: false AllowAllParametersOfDeclarationOnNextLine: false -AllowShortBlocksOnASingleLine: false +AllowShortBlocksOnASingleLine: Never AllowShortCaseLabelsOnASingleLine: true AllowShortFunctionsOnASingleLine: Empty AllowShortIfStatementsOnASingleLine: Never AllowShortLambdasOnASingleLine: Inline AllowShortLoopsOnASingleLine: false -AlwaysBreakAfterDefinitionReturnType: None -AlwaysBreakAfterReturnType: None AlwaysBreakBeforeMultilineStrings: false -AlwaysBreakTemplateDeclarations: Yes BinPackArguments: true BinPackParameters: true BraceWrapping: AfterCaseLabel: false AfterClass: false - AfterControlStatement: false + AfterControlStatement: Never AfterEnum: false AfterFunction: false AfterNamespace: false @@ -59,46 +59,33 @@ BraceWrapping: SplitEmptyFunction: false SplitEmptyRecord: false SplitEmptyNamespace: false +BreakAfterReturnType: Automatic BreakBeforeBinaryOperators: None BreakBeforeBraces: Custom BreakBeforeTernaryOperators: true BreakConstructorInitializers: BeforeColon BreakInheritanceList: BeforeColon BreakStringLiterals: false +BreakTemplateDeclarations: Yes ColumnLimit: 100 -CommentPragmas: '^ IWYU pragma:' CompactNamespaces: false -ConstructorInitializerAllOnOneLineOrOnePerLine: true ConstructorInitializerIndentWidth: 8 ContinuationIndentWidth: 4 Cpp11BracedListStyle: false -DerivePointerAlignment: true +DerivePointerAlignment: false DisableFormat: false FixNamespaceComments: true -ForEachMacros: - - foreach - - Q_FOREACH - - BOOST_FOREACH -IncludeBlocks: Preserve -IncludeCategories: - - Regex: '^' - Priority: 2 - - Regex: '^<.*\.h>' - Priority: 1 - - Regex: '^<.*' - Priority: 2 - - Regex: '.*' - Priority: 3 -IncludeIsMainRegex: '([-_](test|unittest))?$' IndentCaseLabels: true IndentPPDirectives: None IndentWidth: 4 IndentWrappedFunctionNames: false -KeepEmptyLinesAtTheStartOfBlocks: false -MacroBlockBegin: '' -MacroBlockEnd: '' +KeepEmptyLines: + AtEndOfFile: false + AtStartOfBlock: false + AtStartOfFile: false MaxEmptyLinesToKeep: 1 NamespaceIndentation: None +PackConstructorInitializers: CurrentLine PenaltyBreakAssignment: 2 PenaltyBreakBeforeFirstCallParameter: 1 PenaltyBreakComment: 300 @@ -108,57 +95,24 @@ PenaltyBreakTemplateDeclaration: 10 PenaltyExcessCharacter: 1000000 PenaltyReturnTypeOnItsOwnLine: 200 PointerAlignment: Left -RawStringFormats: - - Language: Cpp - Delimiters: - - cc - - CC - - cpp - - Cpp - - CPP - - 'c++' - - 'C++' - CanonicalDelimiter: '' - BasedOnStyle: google - - Language: TextProto - Delimiters: - - pb - - PB - - proto - - PROTO - EnclosingFunctions: - - EqualsProto - - EquivToProto - - PARSE_PARTIAL_TEXT_PROTO - - PARSE_TEST_PROTO - - PARSE_TEXT_PROTO - - ParseTextOrDie - - ParseTextProtoOrDie - CanonicalDelimiter: '' - BasedOnStyle: google ReflowComments: false -SortIncludes: false -SortUsingDeclarations: false +SortIncludes: Never +SortUsingDeclarations: Never SpaceAfterCStyleCast: false SpaceAfterLogicalNot: false SpaceAfterTemplateKeyword: true SpaceBeforeAssignmentOperators: true +SpaceBeforeCaseColon: false SpaceBeforeCpp11BracedList: false SpaceBeforeCtorInitializerColon: true SpaceBeforeInheritanceColon: true SpaceBeforeParens: ControlStatements SpaceBeforeRangeBasedForLoopColon: true -SpaceInEmptyParentheses: false SpacesBeforeTrailingComments: 1 -SpacesInAngles: false +SpacesInAngles: Never SpacesInContainerLiterals: false -SpacesInCStyleCastParentheses: false -SpacesInParentheses: false +SpacesInParens: Never SpacesInSquareBrackets: false -Standard: Cpp11 -StatementMacros: - - Q_UNUSED - - QT_REQUIRE_VERSION -TabWidth: 1 +Standard: c++17 UseTab: Never ... diff --git a/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp b/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp index 641ee5ef4..295cc4dea 100644 --- a/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp +++ b/examples/blas/compile_time_dispatching/level3/gemm_usm_mklcpu_cublas.cpp @@ -61,7 +61,7 @@ // // is performed and finally the results are post processed. // -void run_gemm_example(const sycl::device &cpu_dev, const sycl::device &gpu_dev) { +void run_gemm_example(const sycl::device& cpu_dev, const sycl::device& gpu_dev) { // // Initialize data for Gemm // @@ -89,11 +89,11 @@ void run_gemm_example(const sycl::device &cpu_dev, const sycl::device &gpu_dev) // Catch asynchronous exceptions for CPU and GPU auto cpu_exception_handler = [](sycl::exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (sycl::exception const &e) { + catch (sycl::exception const& e) { std::cerr << "Caught asynchronous SYCL exception on CPU device during GEMM:" << std::endl; std::cerr << "\t" << e.what() << std::endl; @@ -102,11 +102,11 @@ void run_gemm_example(const sycl::device &cpu_dev, const sycl::device &gpu_dev) std::exit(2); }; auto gpu_exception_handler = [](sycl::exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (sycl::exception const &e) { + catch (sycl::exception const& e) { std::cerr << "Caught asynchronous SYCL exception on GPU device during GEMM:" << std::endl; std::cerr << "\t" << e.what() << std::endl; @@ -141,9 +141,9 @@ void run_gemm_example(const sycl::device &cpu_dev, const sycl::device &gpu_dev) sycl::context cpu_cxt = cpu_queue.get_context(); // allocate on CPU device and copy data from host to SYCL CPU device - float *cpu_A = sycl::malloc_device(sizea * sizeof(float), cpu_queue); - float *cpu_B = sycl::malloc_device(sizeb * sizeof(float), cpu_queue); - float *cpu_C = sycl::malloc_device(sizec * sizeof(float), cpu_queue); + float* cpu_A = sycl::malloc_device(sizea * sizeof(float), cpu_queue); + float* cpu_B = sycl::malloc_device(sizeb * sizeof(float), cpu_queue); + float* cpu_C = sycl::malloc_device(sizec * sizeof(float), cpu_queue); if (!cpu_A || !cpu_B || !cpu_C) { throw std::runtime_error("Failed to allocate USM memory."); } @@ -159,9 +159,9 @@ void run_gemm_example(const sycl::device &cpu_dev, const sycl::device &gpu_dev) sycl::context gpu_cxt = gpu_queue.get_context(); // allocate on GPU device and copy data from host to SYCL GPU device - float *gpu_A = sycl::malloc_device(sizea * sizeof(float), gpu_queue); - float *gpu_B = sycl::malloc_device(sizeb * sizeof(float), gpu_queue); - float *gpu_C = sycl::malloc_device(sizec * sizeof(float), gpu_queue); + float* gpu_A = sycl::malloc_device(sizea * sizeof(float), gpu_queue); + float* gpu_B = sycl::malloc_device(sizeb * sizeof(float), gpu_queue); + float* gpu_C = sycl::malloc_device(sizec * sizeof(float), gpu_queue); if (!gpu_A || !gpu_B || !gpu_C) { throw std::runtime_error("Failed to allocate USM memory."); } @@ -174,11 +174,11 @@ void run_gemm_example(const sycl::device &cpu_dev, const sycl::device &gpu_dev) // // add oneapi::math::blas::gemm to execution queue cpu_gemm_done = oneapi::math::blas::column_major::gemm( - oneapi::math::backend_selector{ cpu_queue }, transA, transB, m, - n, k, alpha, cpu_A, ldA, cpu_B, ldB, beta, cpu_C, ldC); + oneapi::math::backend_selector{ cpu_queue }, transA, transB, + m, n, k, alpha, cpu_A, ldA, cpu_B, ldB, beta, cpu_C, ldC); gpu_gemm_done = oneapi::math::blas::column_major::gemm( - oneapi::math::backend_selector{ gpu_queue }, transA, transB, m, - n, k, alpha, gpu_A, ldA, gpu_B, ldB, beta, gpu_C, ldC); + oneapi::math::backend_selector{ gpu_queue }, transA, transB, + m, n, k, alpha, gpu_A, ldA, gpu_B, ldB, beta, gpu_C, ldC); // Wait until calculations are done cpu_gemm_done.wait_and_throw(); @@ -260,7 +260,7 @@ void print_example_banner() { // // Main entry point for example. // -int main(int argc, char **argv) { +int main(int argc, char** argv) { print_example_banner(); try { @@ -279,13 +279,13 @@ int main(int argc, char **argv) { run_gemm_example(cpu_dev, gpu_dev); std::cout << "BLAS GEMM USM example ran OK on MKLCPU and CUBLAS" << std::endl; } - catch (sycl::exception const &e) { + catch (sycl::exception const& e) { std::cerr << "Caught synchronous SYCL exception during GEMM:" << std::endl; std::cerr << "\t" << e.what() << std::endl; std::cerr << "\tSYCL error code: " << e.code().value() << std::endl; return 1; } - catch (std::exception const &e) { + catch (std::exception const& e) { std::cerr << "Caught std::exception during GEMM:"; std::cerr << "\t" << e.what() << std::endl; return 1; diff --git a/examples/blas/run_time_dispatching/level3/gemm_usm.cpp b/examples/blas/run_time_dispatching/level3/gemm_usm.cpp index 3df710101..1cfef28fb 100644 --- a/examples/blas/run_time_dispatching/level3/gemm_usm.cpp +++ b/examples/blas/run_time_dispatching/level3/gemm_usm.cpp @@ -136,7 +136,7 @@ void run_gemm_example(const sycl::device& dev) { // // add oneapi::math::blas::gemm to execution queue gemm_done = oneapi::math::blas::column_major::gemm(main_queue, transA, transB, m, n, k, alpha, - dev_A, ldA, dev_B, ldB, beta, dev_C, ldC); + dev_A, ldA, dev_B, ldB, beta, dev_C, ldC); // Wait until calculations are done main_queue.wait_and_throw(); diff --git a/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp b/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp index fdad04de2..864e073eb 100644 --- a/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp +++ b/examples/dft/compile_time_dispatching/complex_fwd_usm_mklcpu_cufft.cpp @@ -83,7 +83,7 @@ void run_example(const sycl::device& cpu_device, const sycl::device& gpu_device) // enabling // 1. create descriptors oneapi::math::dft::descriptor + oneapi::math::dft::domain::COMPLEX> desc(static_cast(N)); // 2. variadic set_value diff --git a/examples/dft/run_time_dispatching/real_fwd_usm.cpp b/examples/dft/run_time_dispatching/real_fwd_usm.cpp index 5f1f53acf..f674a5c42 100644 --- a/examples/dft/run_time_dispatching/real_fwd_usm.cpp +++ b/examples/dft/run_time_dispatching/real_fwd_usm.cpp @@ -54,7 +54,7 @@ void run_example(const sycl::device& dev) { // 1. create descriptors oneapi::math::dft::descriptor + oneapi::math::dft::domain::REAL> desc(static_cast(N)); // 2. variadic set_value diff --git a/examples/include/example_helper.hpp b/examples/include/example_helper.hpp index 9ed99940c..681ab9142 100644 --- a/examples/include/example_helper.hpp +++ b/examples/include/example_helper.hpp @@ -88,7 +88,7 @@ fp rand_scalar() { } template -void rand_matrix(vec &M, oneapi::math::transpose trans, int m, int n, int ld) { +void rand_matrix(vec& M, oneapi::math::transpose trans, int m, int n, int ld) { using fp = typename vec::value_type; if (trans == oneapi::math::transpose::nontrans) { @@ -104,7 +104,7 @@ void rand_matrix(vec &M, oneapi::math::transpose trans, int m, int n, int ld) { } template -intType generate_sparse_matrix(const intType nx, intType *ia, intType *ja, fp *a, +intType generate_sparse_matrix(const intType nx, intType* ia, intType* ja, fp* a, const intType index = 0) { intType nz = nx, ny = nx; intType nnz = 0; @@ -172,7 +172,7 @@ bool check_result(fp res, fp ref, intType nFlops, intType index) { } template -void free_vec(std::vector &ptr_vec, sycl::queue queue) { +void free_vec(std::vector& ptr_vec, sycl::queue queue) { for (auto ptr : ptr_vec) { sycl::free(ptr, queue); } diff --git a/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp b/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp index f8ee084de..658006938 100644 --- a/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp +++ b/examples/lapack/compile_time_dispatching/getrs_usm_mklcpu_cusolver.cpp @@ -178,8 +178,8 @@ void run_getrs_example(const sycl::device& cpu_device, const sycl::device& gpu_d std::int64_t gpu_getrf_scratchpad_size = oneapi::math::lapack::getrf_scratchpad_size( oneapi::math::backend_selector{ gpu_queue }, m, n, lda); std::int64_t gpu_getrs_scratchpad_size = oneapi::math::lapack::getrs_scratchpad_size( - oneapi::math::backend_selector{ gpu_queue }, trans, n, nrhs, - lda, ldb); + oneapi::math::backend_selector{ gpu_queue }, trans, n, + nrhs, lda, ldb); float* gpu_getrf_scratchpad = sycl::malloc_device( gpu_getrf_scratchpad_size * sizeof(float), gpu_device, gpu_context); float* gpu_getrs_scratchpad = sycl::malloc_device( @@ -197,8 +197,8 @@ void run_getrs_example(const sycl::device& cpu_device, const sycl::device& gpu_d // cpu_getrf_done = oneapi::math::lapack::getrf( - oneapi::math::backend_selector{ cpu_queue }, m, n, cpu_A, lda, - cpu_ipiv, cpu_getrf_scratchpad, cpu_getrf_scratchpad_size); + oneapi::math::backend_selector{ cpu_queue }, m, n, cpu_A, + lda, cpu_ipiv, cpu_getrf_scratchpad, cpu_getrf_scratchpad_size); cpu_getrs_done = oneapi::math::lapack::getrs( oneapi::math::backend_selector{ cpu_queue }, trans, n, nrhs, cpu_A, lda, cpu_ipiv, cpu_B, ldb, cpu_getrs_scratchpad, cpu_getrs_scratchpad_size, @@ -207,8 +207,8 @@ void run_getrs_example(const sycl::device& cpu_device, const sycl::device& gpu_d oneapi::math::backend_selector{ gpu_queue }, m, n, gpu_A, lda, gpu_ipiv, gpu_getrf_scratchpad, gpu_getrf_scratchpad_size); gpu_getrs_done = oneapi::math::lapack::getrs( - oneapi::math::backend_selector{ gpu_queue }, trans, n, nrhs, - gpu_A, lda, gpu_ipiv, gpu_B, ldb, gpu_getrs_scratchpad, gpu_getrs_scratchpad_size, + oneapi::math::backend_selector{ gpu_queue }, trans, n, + nrhs, gpu_A, lda, gpu_ipiv, gpu_B, ldb, gpu_getrs_scratchpad, gpu_getrs_scratchpad_size, { gpu_getrf_done }); // Wait until calculations are done diff --git a/examples/lapack/run_time_dispatching/getrs_usm.cpp b/examples/lapack/run_time_dispatching/getrs_usm.cpp index 68a7836e0..c31d96d6e 100644 --- a/examples/lapack/run_time_dispatching/getrs_usm.cpp +++ b/examples/lapack/run_time_dispatching/getrs_usm.cpp @@ -146,10 +146,10 @@ void run_getrs_example(const sycl::device& device) { // Execute on device getrf_done = oneapi::math::lapack::getrf(queue, m, n, dev_A, lda, dev_ipiv, getrf_scratchpad, - getrf_scratchpad_size); + getrf_scratchpad_size); getrs_done = oneapi::math::lapack::getrs(queue, trans, n, nrhs, dev_A, lda, dev_ipiv, dev_B, ldb, - getrs_scratchpad, getrs_scratchpad_size, { getrf_done }); + getrs_scratchpad, getrs_scratchpad_size, { getrf_done }); // Wait until calculations are done queue.wait_and_throw(); diff --git a/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp b/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp index a023c0a69..20fc2f3a2 100644 --- a/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp +++ b/examples/sparse_blas/compile_time_dispatching/sparse_blas_spmv_usm_mklcpu.cpp @@ -60,7 +60,7 @@ // is performed and finally the results are post processed. // template -int run_sparse_matrix_vector_multiply_example(const sycl::device &cpu_dev) { +int run_sparse_matrix_vector_multiply_example(const sycl::device& cpu_dev) { // Matrix data size intType size = 4; intType nrows = size * size * size; @@ -71,11 +71,11 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &cpu_dev) { // Catch asynchronous exceptions auto exception_handler = [](sycl::exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (sycl::exception const &e) { + catch (sycl::exception const& e) { std::cout << "Caught asynchronous SYCL " "exception during sparse::spmv:\n" << e.what() << std::endl; @@ -94,12 +94,12 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &cpu_dev) { std::size_t sizeia = static_cast(nrows + 1); std::size_t sizevec = static_cast(nrows); - ia = (intType *)sycl::malloc_shared(sizeia * sizeof(intType), cpu_queue); - ja = (intType *)sycl::malloc_shared(sizeja * sizeof(intType), cpu_queue); - a = (fp *)sycl::malloc_shared(sizea * sizeof(fp), cpu_queue); - x = (fp *)sycl::malloc_shared(sizevec * sizeof(fp), cpu_queue); - y = (fp *)sycl::malloc_shared(sizevec * sizeof(fp), cpu_queue); - z = (fp *)sycl::malloc_shared(sizevec * sizeof(fp), cpu_queue); + ia = (intType*)sycl::malloc_shared(sizeia * sizeof(intType), cpu_queue); + ja = (intType*)sycl::malloc_shared(sizeja * sizeof(intType), cpu_queue); + a = (fp*)sycl::malloc_shared(sizea * sizeof(fp), cpu_queue); + x = (fp*)sycl::malloc_shared(sizevec * sizeof(fp), cpu_queue); + y = (fp*)sycl::malloc_shared(sizevec * sizeof(fp), cpu_queue); + z = (fp*)sycl::malloc_shared(sizevec * sizeof(fp), cpu_queue); if (!ia || !ja || !a || !x || !y || !z) { throw std::runtime_error("Failed to allocate USM memory"); @@ -114,10 +114,10 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &cpu_dev) { z[i] = set_fp_value(fp(0.0)); } - std::vector int_ptr_vec; + std::vector int_ptr_vec; int_ptr_vec.push_back(ia); int_ptr_vec.push_back(ja); - std::vector fp_ptr_vec; + std::vector fp_ptr_vec; fp_ptr_vec.push_back(a); fp_ptr_vec.push_back(x); fp_ptr_vec.push_back(y); @@ -143,7 +143,7 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &cpu_dev) { // Create and initialize handle for a Sparse Matrix in CSR format oneapi::math::sparse::matrix_handle_t A_handle = nullptr; oneapi::math::sparse::init_csr_matrix(cpu_selector, &A_handle, nrows, nrows, nnz, - oneapi::math::index_base::zero, ia, ja, a); + oneapi::math::index_base::zero, ia, ja, a); // Create and initialize dense vector handles oneapi::math::sparse::dense_vector_handle_t x_handle = nullptr; @@ -158,17 +158,17 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &cpu_dev) { // Allocate external workspace std::size_t workspace_size = 0; oneapi::math::sparse::spmv_buffer_size(cpu_selector, transA, &alpha, A_view, A_handle, x_handle, - &beta, y_handle, alg, descr, workspace_size); - void *workspace = sycl::malloc_device(workspace_size, cpu_queue); + &beta, y_handle, alg, descr, workspace_size); + void* workspace = sycl::malloc_device(workspace_size, cpu_queue); // Optimize spmv auto ev_opt = - oneapi::math::sparse::spmv_optimize(cpu_selector, transA, &alpha, A_view, A_handle, x_handle, - &beta, y_handle, alg, descr, workspace); + oneapi::math::sparse::spmv_optimize(cpu_selector, transA, &alpha, A_view, A_handle, + x_handle, &beta, y_handle, alg, descr, workspace); // Run spmv auto ev_spmv = oneapi::math::sparse::spmv(cpu_selector, transA, &alpha, A_view, A_handle, - x_handle, &beta, y_handle, alg, descr, { ev_opt }); + x_handle, &beta, y_handle, alg, descr, { ev_opt }); // Release handles and descriptor std::vector release_events; @@ -188,7 +188,7 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &cpu_dev) { // Post Processing // - fp *res = y; + fp* res = y; const bool isConj = (transA == oneapi::math::transpose::conjtrans); for (intType row = 0; row < nrows; row++) { z[row] *= beta; @@ -254,7 +254,7 @@ void print_example_banner() { // // Main entry point for example // -int main(int /*argc*/, char ** /*argv*/) { +int main(int /*argc*/, char** /*argv*/) { print_example_banner(); try { @@ -269,13 +269,13 @@ int main(int /*argc*/, char ** /*argv*/) { run_sparse_matrix_vector_multiply_example(cpu_dev); std::cout << "Sparse BLAS SPMV USM example ran OK." << std::endl; } - catch (sycl::exception const &e) { + catch (sycl::exception const& e) { std::cerr << "Caught synchronous SYCL exception during Sparse SPMV:" << std::endl; std::cerr << "\t" << e.what() << std::endl; std::cerr << "\tSYCL error code: " << e.code().value() << std::endl; return 1; } - catch (std::exception const &e) { + catch (std::exception const& e) { std::cerr << "Caught std::exception during Sparse SPMV:" << std::endl; std::cerr << "\t" << e.what() << std::endl; return 1; diff --git a/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp b/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp index ba51a08e8..a6ff30354 100644 --- a/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp +++ b/examples/sparse_blas/run_time_dispatching/sparse_blas_spmv_usm.cpp @@ -61,7 +61,7 @@ // is performed and finally the results are post processed. // template -int run_sparse_matrix_vector_multiply_example(const sycl::device &dev) { +int run_sparse_matrix_vector_multiply_example(const sycl::device& dev) { // Matrix data size intType size = 4; intType nrows = size * size * size; @@ -72,11 +72,11 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &dev) { // Catch asynchronous exceptions auto exception_handler = [](sycl::exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (sycl::exception const &e) { + catch (sycl::exception const& e) { std::cout << "Caught asynchronous SYCL " "exception during sparse::spmv:\n" << e.what() << std::endl; @@ -95,12 +95,12 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &dev) { std::size_t sizevec = static_cast(nrows); auto sizevec_i64 = static_cast(sizevec); - ia = (intType *)sycl::malloc_shared(sizeia * sizeof(intType), main_queue); - ja = (intType *)sycl::malloc_shared(sizeja * sizeof(intType), main_queue); - a = (fp *)sycl::malloc_shared(sizea * sizeof(fp), main_queue); - x = (fp *)sycl::malloc_shared(sizevec * sizeof(fp), main_queue); - y = (fp *)sycl::malloc_shared(sizevec * sizeof(fp), main_queue); - z = (fp *)sycl::malloc_shared(sizevec * sizeof(fp), main_queue); + ia = (intType*)sycl::malloc_shared(sizeia * sizeof(intType), main_queue); + ja = (intType*)sycl::malloc_shared(sizeja * sizeof(intType), main_queue); + a = (fp*)sycl::malloc_shared(sizea * sizeof(fp), main_queue); + x = (fp*)sycl::malloc_shared(sizevec * sizeof(fp), main_queue); + y = (fp*)sycl::malloc_shared(sizevec * sizeof(fp), main_queue); + z = (fp*)sycl::malloc_shared(sizevec * sizeof(fp), main_queue); if (!ia || !ja || !a || !x || !y || !z) { throw std::runtime_error("Failed to allocate USM memory"); @@ -115,10 +115,10 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &dev) { z[i] = set_fp_value(fp(0.0)); } - std::vector int_ptr_vec; + std::vector int_ptr_vec; int_ptr_vec.push_back(ia); int_ptr_vec.push_back(ja); - std::vector fp_ptr_vec; + std::vector fp_ptr_vec; fp_ptr_vec.push_back(a); fp_ptr_vec.push_back(x); fp_ptr_vec.push_back(y); @@ -144,7 +144,7 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &dev) { // Create and initialize handle for a Sparse Matrix in CSR format oneapi::math::sparse::matrix_handle_t A_handle = nullptr; oneapi::math::sparse::init_csr_matrix(main_queue, &A_handle, nrows, nrows, nnz, - oneapi::math::index_base::zero, ia, ja, a); + oneapi::math::index_base::zero, ia, ja, a); // Create and initialize dense vector handles oneapi::math::sparse::dense_vector_handle_t x_handle = nullptr; @@ -159,17 +159,17 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &dev) { // Allocate external workspace std::size_t workspace_size = 0; oneapi::math::sparse::spmv_buffer_size(main_queue, transA, &alpha, A_view, A_handle, x_handle, - &beta, y_handle, alg, descr, workspace_size); - void *workspace = sycl::malloc_device(workspace_size, main_queue); + &beta, y_handle, alg, descr, workspace_size); + void* workspace = sycl::malloc_device(workspace_size, main_queue); // Optimize spmv auto ev_opt = oneapi::math::sparse::spmv_optimize(main_queue, transA, &alpha, A_view, A_handle, x_handle, - &beta, y_handle, alg, descr, workspace); + &beta, y_handle, alg, descr, workspace); // Run spmv - auto ev_spmv = oneapi::math::sparse::spmv(main_queue, transA, &alpha, A_view, A_handle, x_handle, - &beta, y_handle, alg, descr, { ev_opt }); + auto ev_spmv = oneapi::math::sparse::spmv(main_queue, transA, &alpha, A_view, A_handle, + x_handle, &beta, y_handle, alg, descr, { ev_opt }); // Release handles and descriptor std::vector release_events; @@ -189,7 +189,7 @@ int run_sparse_matrix_vector_multiply_example(const sycl::device &dev) { // Post Processing // - fp *res = y; + fp* res = y; const bool isConj = (transA == oneapi::math::transpose::conjtrans); for (intType row = 0; row < nrows; row++) { z[row] *= beta; @@ -258,7 +258,7 @@ void print_example_banner() { // // Main entry point for example // -int main(int /*argc*/, char ** /*argv*/) { +int main(int /*argc*/, char** /*argv*/) { print_example_banner(); try { @@ -279,13 +279,13 @@ int main(int /*argc*/, char ** /*argv*/) { run_sparse_matrix_vector_multiply_example(dev); std::cout << "Sparse BLAS SPMV USM example ran OK." << std::endl; } - catch (sycl::exception const &e) { + catch (sycl::exception const& e) { std::cerr << "Caught synchronous SYCL exception during Sparse SPMV:" << std::endl; std::cerr << "\t" << e.what() << std::endl; std::cerr << "\tSYCL error code: " << e.code().value() << std::endl; return 1; } - catch (std::exception const &e) { + catch (std::exception const& e) { std::cerr << "Caught std::exception during Sparse SPMV:" << std::endl; std::cerr << "\t" << e.what() << std::endl; return 1; diff --git a/include/oneapi/math/bfloat16.hpp b/include/oneapi/math/bfloat16.hpp index ebf267283..21a84eab7 100644 --- a/include/oneapi/math/bfloat16.hpp +++ b/include/oneapi/math/bfloat16.hpp @@ -70,7 +70,7 @@ struct bfloat16 { inline bfloat16(float f); bfloat16(double d) : bfloat16(float(d)) {} template - bfloat16(T i, typename std::enable_if::value>::type *_ = nullptr) + bfloat16(T i, typename std::enable_if::value>::type* _ = nullptr) : bfloat16(float(i)) {} inline operator float() const; @@ -101,115 +101,115 @@ struct bfloat16 { return h; } - friend float operator+(const bfloat16 &h1, const bfloat16 &h2) { + friend float operator+(const bfloat16& h1, const bfloat16& h2) { return float(h1) + float(h2); } - friend float operator-(const bfloat16 &h1, const bfloat16 &h2) { + friend float operator-(const bfloat16& h1, const bfloat16& h2) { return float(h1) - float(h2); } - friend float operator*(const bfloat16 &h1, const bfloat16 &h2) { + friend float operator*(const bfloat16& h1, const bfloat16& h2) { return float(h1) * float(h2); } - friend float operator/(const bfloat16 &h1, const bfloat16 &h2) { + friend float operator/(const bfloat16& h1, const bfloat16& h2) { return float(h1) / float(h2); } template friend typename std::enable_if::value, float>::type operator+( - const bfloat16 &h, const T &o) { + const bfloat16& h, const T& o) { return float(h) + float(o); } template friend typename std::enable_if::value, float>::type operator-( - const bfloat16 &h, const T &o) { + const bfloat16& h, const T& o) { return float(h) - float(o); } template friend typename std::enable_if::value, float>::type operator*( - const bfloat16 &h, const T &o) { + const bfloat16& h, const T& o) { return float(h) * float(o); } template friend typename std::enable_if::value, float>::type operator/( - const bfloat16 &h, const T &o) { + const bfloat16& h, const T& o) { return float(h) / float(o); } template friend typename std::enable_if::value, float>::type operator+( - const T &o, const bfloat16 &h) { + const T& o, const bfloat16& h) { return float(o) + float(h); } template friend typename std::enable_if::value, float>::type operator-( - const T &o, const bfloat16 &h) { + const T& o, const bfloat16& h) { return float(o) - float(h); } template friend typename std::enable_if::value, float>::type operator*( - const T &o, const bfloat16 &h) { + const T& o, const bfloat16& h) { return float(o) * float(h); } template friend typename std::enable_if::value, float>::type operator/( - const T &o, const bfloat16 &h) { + const T& o, const bfloat16& h) { return float(o) / float(h); } template friend typename std::enable_if::value, T>::type operator+( - const bfloat16 &h, const T &o) { + const bfloat16& h, const T& o) { return float(h) + o; } template friend typename std::enable_if::value, T>::type operator-( - const bfloat16 &h, const T &o) { + const bfloat16& h, const T& o) { return float(h) - o; } template friend typename std::enable_if::value, T>::type operator*( - const bfloat16 &h, const T &o) { + const bfloat16& h, const T& o) { return float(h) * o; } template friend typename std::enable_if::value, T>::type operator/( - const bfloat16 &h, const T &o) { + const bfloat16& h, const T& o) { return float(h) / o; } template friend typename std::enable_if::value, T>::type operator+( - const T &o, const bfloat16 &h) { + const T& o, const bfloat16& h) { return o + float(h); } template friend typename std::enable_if::value, T>::type operator-( - const T &o, const bfloat16 &h) { + const T& o, const bfloat16& h) { return o - float(h); } template friend typename std::enable_if::value, T>::type operator*( - const T &o, const bfloat16 &h) { + const T& o, const bfloat16& h) { return o * float(h); } template friend typename std::enable_if::value, T>::type operator/( - const T &o, const bfloat16 &h) { + const T& o, const bfloat16& h) { return o / float(h); } template - bfloat16 operator+=(const T &o) { + bfloat16 operator+=(const T& o) { return *this = bfloat16(*this + o); } template - bfloat16 operator-=(const T &o) { + bfloat16 operator-=(const T& o) { return *this = bfloat16(*this - o); } template - bfloat16 operator*=(const T &o) { + bfloat16 operator*=(const T& o) { return *this = bfloat16(*this * o); } template - bfloat16 operator/=(const T &o) { + bfloat16 operator/=(const T& o) { return *this = bfloat16(*this / o); } }; diff --git a/include/oneapi/math/blas.hxx b/include/oneapi/math/blas.hxx index 374585912..cb89703fc 100644 --- a/include/oneapi/math/blas.hxx +++ b/include/oneapi/math/blas.hxx @@ -19,1723 +19,1679 @@ // Buffer APIs -static inline void asum(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +static inline void asum(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { detail::asum(get_device_id(queue), queue, n, x, incx, result); } -static inline void asum(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +static inline void asum(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { detail::asum(get_device_id(queue), queue, n, x, incx, result); } -static inline void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +static inline void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { detail::asum(get_device_id(queue), queue, n, x, incx, result); } -static inline void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +static inline void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { detail::asum(get_device_id(queue), queue, n, x, incx, result); } -static inline void axpy(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy) { +static inline void axpy(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { detail::axpy(get_device_id(queue), queue, n, alpha, x, incx, y, incy); } -static inline void axpy(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy) { +static inline void axpy(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy) { detail::axpy(get_device_id(queue), queue, n, alpha, x, incx, y, incy); } -static inline void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +static inline void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { detail::axpy(get_device_id(queue), queue, n, alpha, x, incx, y, incy); } -static inline void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +static inline void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { detail::axpy(get_device_id(queue), queue, n, alpha, x, incx, y, incy); } -static inline void axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { +static inline void axpy_batch(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -static inline void axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { +static inline void axpy_batch(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -static inline void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, +static inline void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -static inline void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, +static inline void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -static inline void axpby(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { +static inline void axpby(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { detail::axpby(get_device_id(queue), queue, n, alpha, x, incx, beta, y, incy); } -static inline void axpby(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { +static inline void axpby(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy) { detail::axpby(get_device_id(queue), queue, n, alpha, x, incx, beta, y, incy); } -static inline void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, +static inline void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { detail::axpby(get_device_id(queue), queue, n, alpha, x, incx, beta, y, incy); } -static inline void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, +static inline void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { detail::axpby(get_device_id(queue), queue, n, alpha, x, incx, beta, y, incy); } -static inline void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +static inline void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { detail::copy(get_device_id(queue), queue, n, x, incx, y, incy); } -static inline void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +static inline void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { detail::copy(get_device_id(queue), queue, n, x, incx, y, incy); } -static inline void copy(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +static inline void copy(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy) { detail::copy(get_device_id(queue), queue, n, x, incx, y, incy); } -static inline void copy(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +static inline void copy(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { detail::copy(get_device_id(queue), queue, n, x, incx, y, incy); } -static inline void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { +static inline void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { detail::copy_batch(get_device_id(queue), queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -static inline void copy_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, +static inline void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { detail::copy_batch(get_device_id(queue), queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -static inline void copy_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, +static inline void copy_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { detail::copy_batch(get_device_id(queue), queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -static inline void copy_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, +static inline void copy_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { detail::copy_batch(get_device_id(queue), queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -static inline void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +static inline void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { detail::dot(get_device_id(queue), queue, n, x, incx, y, incy, result); } -static inline void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +static inline void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { detail::dot(get_device_id(queue), queue, n, x, incx, y, incy, result); } -static inline void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +static inline void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { detail::dot(get_device_id(queue), queue, n, x, incx, y, incy, result); } -static inline void dotc(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +static inline void dotc(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& result) { detail::dotc(get_device_id(queue), queue, n, x, incx, y, incy, result); } -static inline void dotc(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +static inline void dotc(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { detail::dotc(get_device_id(queue), queue, n, x, incx, y, incy, result); } -static inline void dotu(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +static inline void dotu(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& result) { detail::dotu(get_device_id(queue), queue, n, x, incx, y, incy, result); } -static inline void dotu(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +static inline void dotu(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { detail::dotu(get_device_id(queue), queue, n, x, incx, y, incy, result); } -static inline void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { +static inline void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { detail::gbmv(get_device_id(queue), queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -static inline void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { +static inline void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy) { detail::gbmv(get_device_id(queue), queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -static inline void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, +static inline void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { detail::gbmv(get_device_id(queue), queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -static inline void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, +static inline void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { detail::gbmv(get_device_id(queue), queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -static inline void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - float beta, sycl::buffer &c, std::int64_t ldc) { +static inline void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { +static inline void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +static inline void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +static inline void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +static inline void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::half beta, - sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::half beta, + sycl::buffer& c, std::int64_t ldc) { detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { +static inline void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + float beta, sycl::buffer& c, std::int64_t ldc) { detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { +static inline void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + float beta, sycl::buffer& c, std::int64_t ldc) { detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -static inline void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, double beta, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + double beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -static inline void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -static inline void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -static inline void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -static inline void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -static inline void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -static inline void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -static inline void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - int8_t ao, sycl::buffer &b, std::int64_t ldb, - uint8_t bo, float beta, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &co) { +static inline void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, int8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co) { detail::gemm_bias(get_device_id(queue), queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -static inline void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - int8_t ao, sycl::buffer &b, std::int64_t ldb, int8_t bo, - float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { +static inline void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, int8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co) { detail::gemm_bias(get_device_id(queue), queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -static inline void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - uint8_t ao, sycl::buffer &b, std::int64_t ldb, - int8_t bo, float beta, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &co) { +static inline void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co) { detail::gemm_bias(get_device_id(queue), queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -static inline void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - uint8_t ao, sycl::buffer &b, std::int64_t ldb, - uint8_t bo, float beta, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &co) { +static inline void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co) { detail::gemm_bias(get_device_id(queue), queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -static inline void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { +static inline void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { detail::gemmt(get_device_id(queue), queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { +static inline void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + double beta, sycl::buffer& c, std::int64_t ldc) { detail::gemmt(get_device_id(queue), queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { +static inline void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + std::int64_t n, std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc) { detail::gemmt(get_device_id(queue), queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { +static inline void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + std::int64_t n, std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc) { detail::gemmt(get_device_id(queue), queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { +static inline void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { detail::gemv(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -static inline void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { +static inline void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy) { detail::gemv(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -static inline void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, +static inline void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& y, std::int64_t incy) { detail::gemv(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -static inline void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, +static inline void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& y, std::int64_t incy) { detail::gemv(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -static inline void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, float beta, - sycl::buffer &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { +static inline void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, float beta, sycl::buffer& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -static inline void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, double beta, sycl::buffer &y, +static inline void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, double beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -static inline void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, +static inline void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -static inline void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, +static inline void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -static inline void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, +static inline void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -static inline void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size) { +static inline void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size) { detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -static inline void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { +static inline void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size) { detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -static inline void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { +static inline void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size) { detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -static inline void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda) { +static inline void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { detail::ger(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda); } -static inline void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda) { +static inline void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { detail::ger(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda); } -static inline void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, +static inline void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda) { detail::gerc(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda); } -static inline void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, +static inline void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda) { detail::gerc(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda); } -static inline void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, +static inline void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda) { detail::geru(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda); } -static inline void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, +static inline void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda) { detail::geru(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda); } -static inline void hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, +static inline void hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& y, std::int64_t incy) { detail::hbmv(get_device_id(queue), queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -static inline void hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, +static inline void hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& y, std::int64_t incy) { detail::hbmv(get_device_id(queue), queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -static inline void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, +static inline void hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::hemm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, +static inline void hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::hemm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, +static inline void hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& y, std::int64_t incy) { detail::hemv(get_device_id(queue), queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -static inline void hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, +static inline void hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& y, std::int64_t incy) { detail::hemv(get_device_id(queue), queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -static inline void her(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { +static inline void her(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { detail::her(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, lda); } -static inline void her(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { +static inline void her(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { detail::her(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, lda); } -static inline void her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, +static inline void her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda) { detail::her2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -static inline void her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, +static inline void her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda) { detail::her2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -static inline void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, +static inline void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc) { detail::her2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, +static inline void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - double beta, sycl::buffer, 1> &c, - std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, double beta, + sycl::buffer, 1>& c, std::int64_t ldc) { detail::her2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer, 1> &a, - std::int64_t lda, float beta, sycl::buffer, 1> &c, +static inline void herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer, 1>& a, + std::int64_t lda, float beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::herk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -static inline void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer, 1> &a, - std::int64_t lda, double beta, sycl::buffer, 1> &c, +static inline void herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer, 1>& a, + std::int64_t lda, double beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::herk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -static inline void hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, +static inline void hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { detail::hpmv(get_device_id(queue), queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -static inline void hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, +static inline void hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { detail::hpmv(get_device_id(queue), queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -static inline void hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { +static inline void hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { detail::hpr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a); } -static inline void hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { +static inline void hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { detail::hpr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a); } -static inline void hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a) { +static inline void hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a) { detail::hpr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a); } -static inline void hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a) { +static inline void hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a) { detail::hpr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a); } -static inline void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +static inline void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { detail::iamax(get_device_id(queue), queue, n, x, incx, result); } -static inline void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +static inline void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { detail::iamax(get_device_id(queue), queue, n, x, incx, result); } -static inline void iamax(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +static inline void iamax(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { detail::iamax(get_device_id(queue), queue, n, x, incx, result); } -static inline void iamax(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +static inline void iamax(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { detail::iamax(get_device_id(queue), queue, n, x, incx, result); } -static inline void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +static inline void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { detail::iamin(get_device_id(queue), queue, n, x, incx, result); } -static inline void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +static inline void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { detail::iamin(get_device_id(queue), queue, n, x, incx, result); } -static inline void iamin(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +static inline void iamin(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { detail::iamin(get_device_id(queue), queue, n, x, incx, result); } -static inline void iamin(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +static inline void iamin(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { detail::iamin(get_device_id(queue), queue, n, x, incx, result); } -static inline void nrm2(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +static inline void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { detail::nrm2(get_device_id(queue), queue, n, x, incx, result); } -static inline void nrm2(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +static inline void nrm2(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { detail::nrm2(get_device_id(queue), queue, n, x, incx, result); } -static inline void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +static inline void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { detail::nrm2(get_device_id(queue), queue, n, x, incx, result); } -static inline void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +static inline void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { detail::nrm2(get_device_id(queue), queue, n, x, incx, result); } -static inline void rot(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, - float s) { +static inline void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, float c, float s) { detail::rot(get_device_id(queue), queue, n, x, incx, y, incy, c, s); } -static inline void rot(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, double c, - double s) { +static inline void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, double c, double s) { detail::rot(get_device_id(queue), queue, n, x, incx, y, incy, c, s); } -static inline void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, +static inline void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, float c, float s) { detail::rot(get_device_id(queue), queue, n, x, incx, y, incy, c, s); } -static inline void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - double c, double s) { +static inline void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, double c, + double s) { detail::rot(get_device_id(queue), queue, n, x, incx, y, incy, c, s); } -static inline void rotg(sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s) { +static inline void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s) { detail::rotg(get_device_id(queue), queue, a, b, c, s); } -static inline void rotg(sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s) { +static inline void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s) { detail::rotg(get_device_id(queue), queue, a, b, c, s); } -static inline void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +static inline void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { detail::rotg(get_device_id(queue), queue, a, b, c, s); } -static inline void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, - sycl::buffer &c, - sycl::buffer, 1> &s) { +static inline void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { detail::rotg(get_device_id(queue), queue, a, b, c, s); } -static inline void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m) { +static inline void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { detail::rotm(get_device_id(queue), queue, n, x, incx, y, incy, param); } -static inline void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m) { +static inline void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { detail::rotm(get_device_id(queue), queue, n, x, incx, y, incy, param); } -static inline void rotmg(sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, float y1, - sycl::buffer ¶m) { +static inline void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, float y1, sycl::buffer& param) { detail::rotmg(get_device_id(queue), queue, d1, d2, x1, y1, param); } -static inline void rotmg(sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, - double y1, sycl::buffer ¶m) { +static inline void rotmg(sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, double y1, + sycl::buffer& param) { detail::rotmg(get_device_id(queue), queue, d1, d2, x1, y1, param); } -static inline void sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { +static inline void sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { detail::sbmv(get_device_id(queue), queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -static inline void sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { +static inline void sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy) { detail::sbmv(get_device_id(queue), queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -static inline void scal(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx) { +static inline void scal(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx) { detail::scal(get_device_id(queue), queue, n, alpha, x, incx); } -static inline void scal(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx) { +static inline void scal(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx) { detail::scal(get_device_id(queue), queue, n, alpha, x, incx); } -static inline void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +static inline void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx) { detail::scal(get_device_id(queue), queue, n, alpha, x, incx); } -static inline void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +static inline void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx) { detail::scal(get_device_id(queue), queue, n, alpha, x, incx); } -static inline void scal(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +static inline void scal(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx) { detail::scal(get_device_id(queue), queue, n, alpha, x, incx); } -static inline void scal(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +static inline void scal(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx) { detail::scal(get_device_id(queue), queue, n, alpha, x, incx); } -static inline void sdsdot(sycl::queue &queue, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +static inline void sdsdot(sycl::queue& queue, std::int64_t n, float sb, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { detail::sdsdot(get_device_id(queue), queue, n, sb, x, incx, y, incy, result); } -static inline void spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, - std::int64_t incy) { +static inline void spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy) { detail::spmv(get_device_id(queue), queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -static inline void spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, - std::int64_t incy) { +static inline void spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy) { detail::spmv(get_device_id(queue), queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -static inline void spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &a) { +static inline void spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { detail::spr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a); } -static inline void spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &a) { +static inline void spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { detail::spr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a); } -static inline void spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a) { +static inline void spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { detail::spr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a); } -static inline void spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a) { +static inline void spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { detail::spr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a); } -static inline void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +static inline void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { detail::swap(get_device_id(queue), queue, n, x, incx, y, incy); } -static inline void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +static inline void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { detail::swap(get_device_id(queue), queue, n, x, incx, y, incy); } -static inline void swap(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +static inline void swap(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy) { detail::swap(get_device_id(queue), queue, n, x, incx, y, incy); } -static inline void swap(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +static inline void swap(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { detail::swap(get_device_id(queue), queue, n, x, incx, y, incy); } -static inline void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - float beta, sycl::buffer &c, std::int64_t ldc) { +static inline void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { detail::symm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - double beta, sycl::buffer &c, std::int64_t ldc) { +static inline void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { detail::symm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, +static inline void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::symm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, +static inline void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::symm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { +static inline void symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { detail::symv(get_device_id(queue), queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -static inline void symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { +static inline void symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { detail::symv(get_device_id(queue), queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -static inline void syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &a, std::int64_t lda) { +static inline void syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, + std::int64_t lda) { detail::syr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, lda); } -static inline void syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &a, std::int64_t lda) { +static inline void syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, + std::int64_t lda) { detail::syr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, lda); } -static inline void syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda) { +static inline void syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { detail::syr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -static inline void syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda) { +static inline void syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { detail::syr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -static inline void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - float beta, sycl::buffer &c, std::int64_t ldc) { +static inline void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { detail::syr2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - double beta, sycl::buffer &c, std::int64_t ldc) { +static inline void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { detail::syr2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, +static inline void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::syr2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, +static inline void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::syr2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -static inline void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &c, - std::int64_t ldc) { +static inline void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& c, std::int64_t ldc) { detail::syrk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -static inline void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &c, - std::int64_t ldc) { +static inline void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + double beta, sycl::buffer& c, std::int64_t ldc) { detail::syrk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -static inline void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, +static inline void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::syrk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -static inline void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, +static inline void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { detail::syrk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -static inline void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, float beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { +static inline void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size) { detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -static inline void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { +static inline void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size) { detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -static inline void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, +static inline void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -static inline void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, +static inline void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -static inline void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +static inline void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { detail::tbmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -static inline void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +static inline void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { detail::tbmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -static inline void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, +static inline void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { detail::tbmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -static inline void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { +static inline void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx) { detail::tbmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -static inline void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +static inline void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { detail::tbsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -static inline void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +static inline void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { detail::tbsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -static inline void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, +static inline void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { detail::tbsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -static inline void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { +static inline void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx) { detail::tbsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -static inline void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { +static inline void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { detail::tpmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx); } -static inline void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { +static inline void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { detail::tpmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx); } -static inline void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +static inline void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { detail::tpmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx); } -static inline void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +static inline void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { detail::tpmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx); } -static inline void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { +static inline void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { detail::tpsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx); } -static inline void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { +static inline void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { detail::tpsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx); } -static inline void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +static inline void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { detail::tpsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx); } -static inline void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +static inline void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { detail::tpsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx); } -static inline void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +static inline void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb) { detail::trmm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -static inline void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +static inline void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb) { detail::trmm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -static inline void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +static inline void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { detail::trmm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -static inline void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +static inline void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { detail::trmm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -static inline void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +static inline void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { detail::trmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -static inline void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +static inline void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { detail::trmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -static inline void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx) { +static inline void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { detail::trmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -static inline void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx) { +static inline void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { detail::trmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -static inline void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +static inline void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb) { detail::trsm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -static inline void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +static inline void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb) { detail::trsm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -static inline void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +static inline void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { detail::trsm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -static inline void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +static inline void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { detail::trsm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -static inline void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, +static inline void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size) { detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -static inline void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, +static inline void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { + double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size) { detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -static inline void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, +static inline void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size) { detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -static inline void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, +static inline void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size) { detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -static inline void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +static inline void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { detail::trsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -static inline void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +static inline void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { detail::trsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -static inline void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx) { +static inline void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { detail::trsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -static inline void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx) { +static inline void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { detail::trsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -static inline void omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, +static inline void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { detail::omatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -static inline void omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, +static inline void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { detail::omatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -static inline void omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, +static inline void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { detail::omatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -static inline void omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, +static inline void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { detail::omatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -static inline void imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, +static inline void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { detail::imatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -static inline void imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, +static inline void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { detail::imatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -static inline void imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, +static inline void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { detail::imatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -static inline void imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, +static inline void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { detail::imatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -static inline void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - float beta, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + float beta, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::omatadd_batch(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -static inline void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, double beta, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::omatadd_batch(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -static inline void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::omatadd_batch(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -static inline void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { detail::omatadd_batch(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -static inline void omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { +static inline void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { detail::omatcopy(get_device_id(queue), queue, trans, m, n, alpha, a, lda, b, ldb); } -static inline void omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { +static inline void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { detail::omatcopy(get_device_id(queue), queue, trans, m, n, alpha, a, lda, b, ldb); } -static inline void omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, +static inline void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { detail::omatcopy(get_device_id(queue), queue, trans, m, n, alpha, a, lda, b, ldb); } -static inline void omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, +static inline void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { detail::omatcopy(get_device_id(queue), queue, trans, m, n, alpha, a, lda, b, ldb); } -static inline void omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, +static inline void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { detail::omatcopy2(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -static inline void omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, +static inline void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { detail::omatcopy2(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -static inline void omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, +static inline void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { detail::omatcopy2(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -static inline void omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, +static inline void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { detail::omatcopy2(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -static inline void imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &ab, std::int64_t lda, +static inline void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { detail::imatcopy(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb); } -static inline void imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &ab, std::int64_t lda, +static inline void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { detail::imatcopy(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb); } -static inline void imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +static inline void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { detail::imatcopy(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb); } -static inline void imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +static inline void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { detail::imatcopy(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb); } -static inline void omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - float beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { +static inline void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { detail::omatadd(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -static inline void omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &c, std::int64_t ldc) { +static inline void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& c, std::int64_t ldc) { detail::omatadd(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -static inline void omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +static inline void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& b, + std::int64_t ldb, sycl::buffer, 1>& c, std::int64_t ldc) { detail::omatadd(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -static inline void omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +static inline void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& b, + std::int64_t ldb, sycl::buffer, 1>& c, std::int64_t ldc) { detail::omatadd(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); @@ -1743,603 +1699,590 @@ static inline void omatadd(sycl::queue &queue, transpose transa, transpose trans // USM APIs -static inline sycl::event asum(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies = {}) { +static inline sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}) { auto done = detail::asum(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event asum(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies = {}) { +static inline sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}) { auto done = detail::asum(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event asum(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies = {}) { +static inline sycl::event asum(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}) { auto done = detail::asum(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event asum(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies = {}) { +static inline sycl::event asum(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}) { auto done = detail::asum(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event axpy(sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event axpy(sycl::queue& queue, std::int64_t n, float alpha, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::axpy(get_device_id(queue), queue, n, alpha, x, incx, y, incy, dependencies); return done; } -static inline sycl::event axpy(sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event axpy(sycl::queue& queue, std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::axpy(get_device_id(queue), queue, n, alpha, x, incx, y, incy, dependencies); return done; } -static inline sycl::event axpy(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::axpy(get_device_id(queue), queue, n, alpha, x, incx, y, incy, dependencies); return done; } -static inline sycl::event axpy(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::axpy(get_device_id(queue), queue, n, alpha, x, incx, y, incy, dependencies); return done; } -static inline sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, double *alpha, - const double **x, std::int64_t *incx, double **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, double* alpha, + const double** x, std::int64_t* incx, double** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, float *alpha, - const float **x, std::int64_t *incx, float **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, float* alpha, + const float** x, std::int64_t* incx, float** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, - std::complex *alpha, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, std::int64_t stridex, - float *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, std::int64_t stridex, + float* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -static inline sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, std::int64_t stridex, - double *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, double alpha, + const double* x, std::int64_t incx, std::int64_t stridex, + double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -static inline sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -static inline sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::axpy_batch(get_device_id(queue), queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -static inline sycl::event axpby(sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float beta, float *y, - std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event axpby(sycl::queue& queue, std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::axpby(get_device_id(queue), queue, n, alpha, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event axpby(sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double beta, - double *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event axpby(sycl::queue& queue, std::int64_t n, double alpha, const double* x, + std::int64_t incx, const double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::axpby(get_device_id(queue), queue, n, alpha, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event axpby(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::axpby(get_device_id(queue), queue, n, alpha, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event axpby(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::axpby(get_device_id(queue), queue, n, alpha, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event copy(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event copy(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::copy(get_device_id(queue), queue, n, x, incx, y, incy, dependencies); return done; } -static inline sycl::event copy(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event copy(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::copy(get_device_id(queue), queue, n, x, incx, y, incy, dependencies); return done; } -static inline sycl::event copy(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::copy(get_device_id(queue), queue, n, x, incx, y, incy, dependencies); return done; } -static inline sycl::event copy(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::copy(get_device_id(queue), queue, n, x, incx, y, incy, dependencies); return done; } -static inline sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const float **x, - std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const float** x, + std::int64_t* incx, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::copy_batch(get_device_id(queue), queue, n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const double **x, - std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const double** x, + std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::copy_batch(get_device_id(queue), queue, n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, + const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::copy_batch(get_device_id(queue), queue, n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, + const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::copy_batch(get_device_id(queue), queue, n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t stridex, float* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::copy_batch(get_device_id(queue), queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -static inline sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t stridex, double* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::copy_batch(get_device_id(queue), queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -static inline sycl::event copy_batch(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event copy_batch(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::copy_batch(get_device_id(queue), queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -static inline sycl::event copy_batch(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event copy_batch(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::copy_batch(get_device_id(queue), queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -static inline sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - float *result, - const std::vector &dependencies = {}) { +static inline sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* result, + const std::vector& dependencies = {}) { auto done = detail::dot(get_device_id(queue), queue, n, x, incx, y, incy, result, dependencies); return done; } -static inline sycl::event dot(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, - double *result, - const std::vector &dependencies = {}) { +static inline sycl::event dot(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* result, + const std::vector& dependencies = {}) { auto done = detail::dot(get_device_id(queue), queue, n, x, incx, y, incy, result, dependencies); return done; } -static inline sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - double *result, - const std::vector &dependencies = {}) { +static inline sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, double* result, + const std::vector& dependencies = {}) { auto done = detail::dot(get_device_id(queue), queue, n, x, incx, y, incy, result, dependencies); return done; } -static inline sycl::event dotc(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}) { +static inline sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}) { auto done = detail::dotc(get_device_id(queue), queue, n, x, incx, y, incy, result, dependencies); return done; } -static inline sycl::event dotc(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}) { +static inline sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}) { auto done = detail::dotc(get_device_id(queue), queue, n, x, incx, y, incy, result, dependencies); return done; } -static inline sycl::event dotu(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}) { +static inline sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}) { auto done = detail::dotu(get_device_id(queue), queue, n, x, incx, y, incy, result, dependencies); return done; } -static inline sycl::event dotu(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}) { +static inline sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}) { auto done = detail::dotu(get_device_id(queue), queue, n, x, incx, y, incy, result, dependencies); return done; } -static inline sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, - const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::gbmv(get_device_id(queue), queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::gbmv(get_device_id(queue), queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::gbmv(get_device_id(queue), queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::gbmv(get_device_id(queue), queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, const double* b, std::int64_t ldb, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, - std::int64_t ldb, sycl::half beta, sycl::half *c, - std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, + const sycl::half* a, std::int64_t lda, const sycl::half* b, + std::int64_t ldb, sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const sycl::half* a, std::int64_t lda, const sycl::half* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const bfloat16 *a, std::int64_t lda, const bfloat16 *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const bfloat16* a, std::int64_t lda, const bfloat16* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::gemm(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float *alpha, const float **a, - std::int64_t *lda, const float **b, std::int64_t *ldb, - float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const float** a, std::int64_t* lda, + const float** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, double *alpha, const double **a, - std::int64_t *lda, const double **b, std::int64_t *ldb, - double *beta, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + double* alpha, const double** a, std::int64_t* lda, + const double** b, std::int64_t* ldb, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event gemm_batch( - sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **b, std::int64_t *ldb, std::complex *beta, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** b, + std::int64_t* ldb, std::complex* beta, + std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event gemm_batch( - sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **b, std::int64_t *ldb, std::complex *beta, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** b, + std::int64_t* ldb, std::complex* beta, + std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, sycl::half *alpha, const sycl::half **a, - std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, - sycl::half *beta, sycl::half **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + sycl::half* alpha, const sycl::half** a, std::int64_t* lda, + const sycl::half** b, std::int64_t* ldb, sycl::half* beta, + sycl::half** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, float *beta, - float **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const sycl::half** a, std::int64_t* lda, + const sycl::half** b, std::int64_t* ldb, float* beta, + float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, - float **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, + float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, - std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, + std::int32_t** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, const float *b, std::int64_t ldb, - std::int64_t stride_b, float beta, float *c, - std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, std::int64_t stride_a, + const float* b, std::int64_t ldb, std::int64_t stride_b, + float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, const double *b, std::int64_t ldb, - std::int64_t stride_b, double beta, double *c, - std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, std::int64_t stride_a, + const double* b, std::int64_t ldb, std::int64_t stride_b, + double beta, double* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2347,11 +2290,11 @@ static inline sycl::event gemm_batch(sycl::queue &queue, transpose transa, trans } static inline sycl::event gemm_batch( - sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}) { + sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2359,226 +2302,219 @@ static inline sycl::event gemm_batch( } static inline sycl::event gemm_batch( - sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}) { + sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, const sycl::half *a, std::int64_t lda, - std::int64_t stride_a, const sycl::half *b, - std::int64_t ldb, std::int64_t stride_b, sycl::half beta, - sycl::half *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, + sycl::half alpha, const sycl::half* a, std::int64_t lda, + std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, sycl::half* c, + std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, - float beta, float *c, std::int64_t ldc, std::int64_t stride_c, + const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, + float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, - float beta, float *c, std::int64_t ldc, std::int64_t stride_c, + const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, + float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, - float beta, std::int32_t *c, std::int64_t ldc, + const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, + float beta, std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::gemm_batch(get_device_id(queue), queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::gemmt(get_device_id(queue), queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::gemmt(get_device_id(queue), queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::gemmt(get_device_id(queue), queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::gemmt(get_device_id(queue), queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int8_t ao, const std::uint8_t *b, - std::int64_t ldb, std::uint8_t bo, float beta, - std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, + offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::uint8_t* b, std::int64_t ldb, + std::uint8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}) { auto done = detail::gemm_bias(get_device_id(queue), queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; } -static inline sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int8_t ao, const std::int8_t *b, - std::int64_t ldb, std::int8_t bo, float beta, - std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, + offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::int8_t* b, std::int64_t ldb, + std::int8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}) { auto done = detail::gemm_bias(get_device_id(queue), queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; } -static inline sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, - std::int64_t lda, std::uint8_t ao, const std::int8_t *b, - std::int64_t ldb, std::int8_t bo, float beta, - std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, + offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::int8_t* b, std::int64_t ldb, + std::int8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}) { auto done = detail::gemm_bias(get_device_id(queue), queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; } -static inline sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, - std::int64_t lda, std::uint8_t ao, const std::uint8_t *b, - std::int64_t ldb, std::uint8_t bo, float beta, - std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies = {}) { +static inline sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, + offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::uint8_t* b, std::int64_t ldb, + std::uint8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}) { auto done = detail::gemm_bias(get_device_id(queue), queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); return done; } -static inline sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::gemv(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::gemv(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::gemv(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::gemv(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stridea, const float *x, - std::int64_t incx, std::int64_t stridex, float beta, - float *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float beta, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -static inline sycl::event gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, const double *x, - std::int64_t incx, std::int64_t stridex, double beta, - double *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stridea, const double* x, + std::int64_t incx, std::int64_t stridex, double beta, + double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); @@ -2586,11 +2522,11 @@ static inline sycl::event gemv_batch(sycl::queue &queue, transpose trans, std::i } static inline sycl::event gemv_batch( - sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, - std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}) { + sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, std::complex beta, + std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); @@ -2598,1538 +2534,1508 @@ static inline sycl::event gemv_batch( } static inline sycl::event gemv_batch( - sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies = {}) { + sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies = {}) { auto done = detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -static inline sycl::event gemv_batch(sycl::queue &queue, transpose *trans, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, const float **x, std::int64_t *incx, - float *beta, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemv_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, const float** x, std::int64_t* incx, + float* beta, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event gemv_batch(sycl::queue &queue, transpose *trans, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, const double **x, std::int64_t *incx, - double *beta, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemv_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, const double** x, std::int64_t* incx, + double* beta, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event gemv_batch(sycl::queue &queue, transpose *trans, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemv_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, + std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event gemv_batch(sycl::queue &queue, transpose *trans, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event gemv_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, + std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::gemv_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; } -static inline sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float *c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, const float* a, std::int64_t lda, + std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -static inline sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double *c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, const double* a, std::int64_t lda, + std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -static inline sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *c, - std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, + std::complex* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -static inline sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *c, - std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, + std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, + std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -static inline sycl::event dgmm_batch(sycl::queue &queue, side *left_right, std::int64_t *m, - std::int64_t *n, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event dgmm_batch(sycl::queue& queue, side* left_right, std::int64_t* m, + std::int64_t* n, const float** a, std::int64_t* lda, + const float** x, std::int64_t* incx, float** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event dgmm_batch(sycl::queue &queue, side *left_right, std::int64_t *m, - std::int64_t *n, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event dgmm_batch(sycl::queue& queue, side* left_right, std::int64_t* m, + std::int64_t* n, const double** a, std::int64_t* lda, + const double** x, std::int64_t* incx, double** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event dgmm_batch(sycl::queue &queue, side *left_right, std::int64_t *m, - std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, - std::int64_t *incx, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event dgmm_batch(sycl::queue& queue, side* left_right, std::int64_t* m, + std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, + std::int64_t* incx, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event dgmm_batch(sycl::queue &queue, side *left_right, std::int64_t *m, - std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, - std::int64_t *incx, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event dgmm_batch(sycl::queue& queue, side* left_right, std::int64_t* m, + std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, + std::int64_t* incx, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::dgmm_batch(get_device_id(queue), queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::ger(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } -static inline sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, + const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::ger(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } -static inline sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::gerc(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } -static inline sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::gerc(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } -static inline sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::geru(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } -static inline sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::geru(get_device_id(queue), queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } -static inline sycl::event hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::hbmv(get_device_id(queue), queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::hbmv(get_device_id(queue), queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::hemm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::hemm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::hemv(get_device_id(queue), queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::hemv(get_device_id(queue), queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event her(sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event her(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::her(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } -static inline sycl::event her(sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event her(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::her(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } -static inline sycl::event her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::her2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } -static inline sycl::event her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::her2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } -static inline sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, float beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, float beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::her2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, double beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, double beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::her2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, - const std::complex *a, std::int64_t lda, float beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, + const std::complex* a, std::int64_t lda, float beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::herk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -static inline sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, - const std::complex *a, std::int64_t lda, double beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, + const std::complex* a, std::int64_t lda, double beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::herk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -static inline sycl::event hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::hpmv(get_device_id(queue), queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::hpmv(get_device_id(queue), queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, - const std::vector &dependencies = {}) { +static inline sycl::event hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + const std::complex* x, std::int64_t incx, + std::complex* a, + const std::vector& dependencies = {}) { auto done = detail::hpr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, dependencies); return done; } -static inline sycl::event hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, - const std::vector &dependencies = {}) { +static inline sycl::event hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const std::complex* x, std::int64_t incx, + std::complex* a, + const std::vector& dependencies = {}) { auto done = detail::hpr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, dependencies); return done; } -static inline sycl::event hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, - const std::vector &dependencies = {}) { +static inline sycl::event hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies = {}) { auto done = detail::hpr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } -static inline sycl::event hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, - const std::vector &dependencies = {}) { +static inline sycl::event hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies = {}) { auto done = detail::hpr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } -static inline sycl::event iamax(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}) { +static inline sycl::event iamax(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}) { auto done = detail::iamax(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event iamax(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}) { +static inline sycl::event iamax(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}) { auto done = detail::iamax(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event iamax(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}) { +static inline sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}) { auto done = detail::iamax(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event iamax(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}) { +static inline sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}) { auto done = detail::iamax(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event iamin(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}) { +static inline sycl::event iamin(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}) { auto done = detail::iamin(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event iamin(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}) { +static inline sycl::event iamin(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}) { auto done = detail::iamin(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event iamin(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}) { +static inline sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}) { auto done = detail::iamin(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event iamin(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}) { +static inline sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}) { auto done = detail::iamin(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event nrm2(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies = {}) { +static inline sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}) { auto done = detail::nrm2(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event nrm2(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies = {}) { +static inline sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}) { auto done = detail::nrm2(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event nrm2(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies = {}) { +static inline sycl::event nrm2(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}) { auto done = detail::nrm2(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event nrm2(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies = {}) { +static inline sycl::event nrm2(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}) { auto done = detail::nrm2(get_device_id(queue), queue, n, x, incx, result, dependencies); return done; } -static inline sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - float c, float s, - const std::vector &dependencies = {}) { +static inline sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, float c, + float s, const std::vector& dependencies = {}) { auto done = detail::rot(get_device_id(queue), queue, n, x, incx, y, incy, c, s, dependencies); return done; } -static inline sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - double c, double s, - const std::vector &dependencies = {}) { +static inline sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + double c, double s, + const std::vector& dependencies = {}) { auto done = detail::rot(get_device_id(queue), queue, n, x, incx, y, incy, c, s, dependencies); return done; } -static inline sycl::event rot(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float c, float s, - const std::vector &dependencies = {}) { +static inline sycl::event rot(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, + float* y, std::int64_t incy, float c, float s, + const std::vector& dependencies = {}) { auto done = detail::rot(get_device_id(queue), queue, n, x, incx, y, incy, c, s, dependencies); return done; } -static inline sycl::event rot(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double c, - double s, const std::vector &dependencies = {}) { +static inline sycl::event rot(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, + double* y, std::int64_t incy, double c, double s, + const std::vector& dependencies = {}) { auto done = detail::rot(get_device_id(queue), queue, n, x, incx, y, incy, c, s, dependencies); return done; } -static inline sycl::event rotg(sycl::queue &queue, float *a, float *b, float *c, float *s, - const std::vector &dependencies = {}) { +static inline sycl::event rotg(sycl::queue& queue, float* a, float* b, float* c, float* s, + const std::vector& dependencies = {}) { auto done = detail::rotg(get_device_id(queue), queue, a, b, c, s, dependencies); return done; } -static inline sycl::event rotg(sycl::queue &queue, double *a, double *b, double *c, - double *s, - const std::vector &dependencies = {}) { +static inline sycl::event rotg(sycl::queue& queue, double* a, double* b, double* c, double* s, + const std::vector& dependencies = {}) { auto done = detail::rotg(get_device_id(queue), queue, a, b, c, s, dependencies); return done; } -static inline sycl::event rotg(sycl::queue &queue, std::complex *a, - std::complex *b, float *c, std::complex *s, - const std::vector &dependencies = {}) { +static inline sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, + float* c, std::complex* s, + const std::vector& dependencies = {}) { auto done = detail::rotg(get_device_id(queue), queue, a, b, c, s, dependencies); return done; } -static inline sycl::event rotg(sycl::queue &queue, std::complex *a, - std::complex *b, double *c, std::complex *s, - const std::vector &dependencies = {}) { +static inline sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, + double* c, std::complex* s, + const std::vector& dependencies = {}) { auto done = detail::rotg(get_device_id(queue), queue, a, b, c, s, dependencies); return done; } -static inline sycl::event rotm(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float *param, - const std::vector &dependencies = {}) { +static inline sycl::event rotm(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, + float* y, std::int64_t incy, float* param, + const std::vector& dependencies = {}) { auto done = detail::rotm(get_device_id(queue), queue, n, x, incx, y, incy, param, dependencies); return done; } -static inline sycl::event rotm(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double *param, - const std::vector &dependencies = {}) { +static inline sycl::event rotm(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, + double* y, std::int64_t incy, double* param, + const std::vector& dependencies = {}) { auto done = detail::rotm(get_device_id(queue), queue, n, x, incx, y, incy, param, dependencies); return done; } -static inline sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, - float y1, float *param, - const std::vector &dependencies = {}) { +static inline sycl::event rotmg(sycl::queue& queue, float* d1, float* d2, float* x1, float y1, + float* param, const std::vector& dependencies = {}) { auto done = detail::rotmg(get_device_id(queue), queue, d1, d2, x1, y1, param, dependencies); return done; } -static inline sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, - double y1, double *param, - const std::vector &dependencies = {}) { +static inline sycl::event rotmg(sycl::queue& queue, double* d1, double* d2, double* x1, double y1, + double* param, const std::vector& dependencies = {}) { auto done = detail::rotmg(get_device_id(queue), queue, d1, d2, x1, y1, param, dependencies); return done; } -static inline sycl::event sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::sbmv(get_device_id(queue), queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::sbmv(get_device_id(queue), queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, float *x, - std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event scal(sycl::queue& queue, std::int64_t n, float alpha, float* x, + std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::scal(get_device_id(queue), queue, n, alpha, x, incx, dependencies); return done; } -static inline sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, double *x, - std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event scal(sycl::queue& queue, std::int64_t n, double alpha, double* x, + std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::scal(get_device_id(queue), queue, n, alpha, x, incx, dependencies); return done; } -static inline sycl::event scal(sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::scal(get_device_id(queue), queue, n, alpha, x, incx, dependencies); return done; } -static inline sycl::event scal(sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::scal(get_device_id(queue), queue, n, alpha, x, incx, dependencies); return done; } -static inline sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event scal(sycl::queue& queue, std::int64_t n, float alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::scal(get_device_id(queue), queue, n, alpha, x, incx, dependencies); return done; } -static inline sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event scal(sycl::queue& queue, std::int64_t n, double alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::scal(get_device_id(queue), queue, n, alpha, x, incx, dependencies); return done; } -static inline sycl::event sdsdot(sycl::queue &queue, std::int64_t n, float sb, - const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *result, - const std::vector &dependencies = {}) { +static inline sycl::event sdsdot(sycl::queue& queue, std::int64_t n, float sb, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, + float* result, const std::vector& dependencies = {}) { auto done = detail::sdsdot(get_device_id(queue), queue, n, sb, x, incx, y, incy, result, dependencies); return done; } -static inline sycl::event spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *a, const float *x, std::int64_t incx, - float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + const float* a, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::spmv(get_device_id(queue), queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *a, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const double* a, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::spmv(get_device_id(queue), queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, - const std::vector &dependencies = {}) { +static inline sycl::event spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + const float* x, std::int64_t incx, float* a, + const std::vector& dependencies = {}) { auto done = detail::spr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, dependencies); return done; } -static inline sycl::event spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, - const std::vector &dependencies = {}) { +static inline sycl::event spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const double* x, std::int64_t incx, double* a, + const std::vector& dependencies = {}) { auto done = detail::spr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, dependencies); return done; } -static inline sycl::event spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, - const std::vector &dependencies = {}) { +static inline sycl::event spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, const std::vector& dependencies = {}) { auto done = detail::spr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } -static inline sycl::event spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *a, - const std::vector &dependencies = {}) { +static inline sycl::event spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, + const std::vector& dependencies = {}) { auto done = detail::spr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); return done; } -static inline sycl::event swap(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event swap(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, + float* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::swap(get_device_id(queue), queue, n, x, incx, y, incy, dependencies); return done; } -static inline sycl::event swap(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event swap(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, + double* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::swap(get_device_id(queue), queue, n, x, incx, y, incy, dependencies); return done; } -static inline sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::swap(get_device_id(queue), queue, n, x, incx, y, incy, dependencies); return done; } -static inline sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::swap(get_device_id(queue), queue, n, x, incx, y, incy, dependencies); return done; } -static inline sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, - float *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, + float* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::symm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, - double *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, + double* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::symm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::symm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::symm(get_device_id(queue), queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + const float* a, std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::symv(get_device_id(queue), queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}) { +static inline sycl::event symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}) { auto done = detail::symv(get_device_id(queue), queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -static inline sycl::event syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, - std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + const float* x, std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::syr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } -static inline sycl::event syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, - std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const double* x, std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::syr(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); return done; } -static inline sycl::event syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::syr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } -static inline sycl::event syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies = {}) { +static inline sycl::event syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies = {}) { auto done = detail::syr2(get_device_id(queue), queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } -static inline sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, - float *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, + float* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::syr2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, + double* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::syr2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::syr2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::syr2k(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -static inline sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::syrk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -static inline sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::syrk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -static inline sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::syrk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -static inline sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::syrk(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -static inline sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - float *alpha, const float **a, std::int64_t *lda, - float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, + std::int64_t* n, std::int64_t* k, float* alpha, + const float** a, std::int64_t* lda, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - double *alpha, const double **a, std::int64_t *lda, - double *beta, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, + std::int64_t* n, std::int64_t* k, double* alpha, + const double** a, std::int64_t* lda, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, - std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, + std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, + std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -static inline sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, - float beta, float *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float beta, float* c, + std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, - double beta, double *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, + std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, + std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::syrk_batch(get_device_id(queue), queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, + std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tbmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } -static inline sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, + std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tbmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } -static inline sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tbmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } -static inline sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tbmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } -static inline sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, + std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tbsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } -static inline sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, + std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tbsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } -static inline sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tbsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } -static inline sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tbsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } -static inline sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, - std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, float* x, + std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tpmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } -static inline sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, - std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, double* x, + std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tpmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } -static inline sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tpmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } -static inline sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tpmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } -static inline sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, - std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, float* x, + std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tpsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } -static inline sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, - std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, double* x, + std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tpsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } -static inline sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tpsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } -static inline sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::tpsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); return done; } -static inline sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, float *b, - std::int64_t ldb, - const std::vector &dependencies = {}) { +static inline sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, float* b, + std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::trmm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, double *b, - std::int64_t ldb, - const std::vector &dependencies = {}) { +static inline sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, double* b, + std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::trmm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}) { +static inline sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::trmm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}) { +static inline sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::trmm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, - float *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, + float* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::trmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -static inline sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, + double* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::trmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -static inline sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::trmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -static inline sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::trmv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -static inline sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, float *b, - std::int64_t ldb, - const std::vector &dependencies = {}) { +static inline sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, float* b, + std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::trsm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, double *b, - std::int64_t ldb, - const std::vector &dependencies = {}) { +static inline sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, double* b, + std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::trsm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}) { +static inline sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::trsm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}) { +static inline sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::trsm(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -static inline sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -static inline sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -static inline sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}) { +static inline sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, + const std::vector& dependencies = {}) { auto done = detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -static inline sycl::event trsm_batch(sycl::queue &queue, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, - std::int64_t *m, std::int64_t *n, float *alpha, - const float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, + transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -static inline sycl::event trsm_batch(sycl::queue &queue, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, - std::int64_t *m, std::int64_t *n, double *alpha, - const double **a, std::int64_t *lda, double **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}) { +static inline sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, + transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -static inline sycl::event trsm_batch( - sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, - std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies = {}) { +static inline sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, + transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -static inline sycl::event trsm_batch( - sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, diag *unit_diag, - std::int64_t *m, std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies = {}) { +static inline sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, + transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}) { auto done = detail::trsm_batch(get_device_id(queue), queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -static inline sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, - float *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, + float* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::trsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -static inline sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, + double* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::trsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -static inline sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::trsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -static inline sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}) { +static inline sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}) { auto done = detail::trsv(get_device_id(queue), queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -static inline sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float *b, +static inline sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::omatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -static inline sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double *b, +static inline sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::omatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -static inline sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, +static inline sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, + const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::omatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -static inline sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, +static inline sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, + const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::omatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -static inline sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, float *ab, std::int64_t lda, +static inline sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::imatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } -static inline sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, double *ab, std::int64_t lda, +static inline sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::imatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } -static inline sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, +static inline sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::imatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } -static inline sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, +static inline sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::imatcopy_batch(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } -static inline sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, const float *a, +static inline sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, - float *c, std::int64_t ldc, std::int64_t stride_c, + const float* b, std::int64_t ldb, std::int64_t stride_b, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::omatadd_batch(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, - double beta, const double *b, std::int64_t ldb, - std::int64_t stride_b, double *c, std::int64_t ldc, + const double* a, std::int64_t lda, std::int64_t stride_a, + double beta, const double* b, std::int64_t ldb, + std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::omatadd_batch(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4137,185 +4043,185 @@ static inline sycl::event omatadd_batch(sycl::queue &queue, transpose transa, tr } static inline sycl::event omatadd_batch( - sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}) { + sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies = {}) { auto done = detail::omatadd_batch(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, +static inline sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::complex *c, + const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::omatadd_batch(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); return done; } -static inline sycl::event omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - float *b, std::int64_t ldb, - const std::vector &dependencies = {}) { +static inline sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + float* b, std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::omatcopy(get_device_id(queue), queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - double *b, std::int64_t ldb, - const std::vector &dependencies = {}) { +static inline sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + double* b, std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::omatcopy(get_device_id(queue), queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, +static inline sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}) { + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::omatcopy(get_device_id(queue), queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, +static inline sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}) { + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::omatcopy(get_device_id(queue), queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); return done; } -static inline sycl::event omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, float *b, std::int64_t ldb, +static inline sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::omatcopy2(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } -static inline sycl::event omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stridea, double *b, std::int64_t ldb, +static inline sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, double* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::omatcopy2(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } -static inline sycl::event omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, +static inline sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, std::int64_t ldb, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::omatcopy2(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } -static inline sycl::event omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, +static inline sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, std::int64_t ldb, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::omatcopy2(get_device_id(queue), queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } -static inline sycl::event imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, float *ab, std::int64_t lda, +static inline sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::imatcopy(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } -static inline sycl::event imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, double *ab, std::int64_t lda, +static inline sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { auto done = detail::imatcopy(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } -static inline sycl::event imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, +static inline sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}) { + std::complex* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::imatcopy(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } -static inline sycl::event imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, +static inline sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}) { + std::complex* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}) { auto done = detail::imatcopy(get_device_id(queue), queue, trans, m, n, alpha, ab, lda, ldb, dependencies); return done; } -static inline sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float beta, const float *b, std::int64_t ldb, - float *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, float beta, const float* b, std::int64_t ldb, + float* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::omatadd(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } -static inline sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double beta, const double *b, std::int64_t ldb, - double *c, std::int64_t ldc, - const std::vector &dependencies = {}) { +static inline sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double beta, const double* b, std::int64_t ldb, + double* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::omatadd(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } -static inline sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, +static inline sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex beta, const std::complex *b, - std::int64_t ldb, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}) { + const std::complex* a, std::int64_t lda, + std::complex beta, const std::complex* b, + std::int64_t ldb, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::omatadd(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; } -static inline sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, +static inline sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex beta, const std::complex *b, - std::int64_t ldb, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}) { + const std::complex* a, std::int64_t lda, + std::complex beta, const std::complex* b, + std::int64_t ldb, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}) { auto done = detail::omatadd(get_device_id(queue), queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); return done; diff --git a/include/oneapi/math/blas/detail/blas_ct_backends.hxx b/include/oneapi/math/blas/detail/blas_ct_backends.hxx index afebb93c3..cd03497d6 100644 --- a/include/oneapi/math/blas/detail/blas_ct_backends.hxx +++ b/include/oneapi/math/blas/detail/blas_ct_backends.hxx @@ -20,2892 +20,2813 @@ // Buffer APIs static inline void syr2(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); + std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a, + std::int64_t lda); static inline void syr2(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); + std::int64_t n, double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a, + std::int64_t lda); static inline void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx); + sycl::buffer& x, std::int64_t incx); static inline void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx); + sycl::buffer& x, std::int64_t incx); static inline void scal(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx); static inline void scal(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx); static inline void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& x, std::int64_t incx); static inline void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& x, std::int64_t incx); static inline void trmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); static inline void trmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); static inline void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); static inline void trmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); static inline void tpmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); static inline void tpmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); static inline void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); static inline void tpmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); static inline void spr(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a); + std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a); static inline void spr(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a); + std::int64_t n, double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a); static inline void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); static inline void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); static inline void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, float beta, - sycl::buffer &c, std::int64_t ldc); + sycl::buffer& a, std::int64_t lda, float beta, + sycl::buffer& c, std::int64_t ldc); static inline void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, double beta, - sycl::buffer &c, std::int64_t ldc); + sycl::buffer& a, std::int64_t lda, double beta, + sycl::buffer& c, std::int64_t ldc); static inline void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); static inline void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); static inline void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, float beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + float beta, sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); static inline void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + double beta, sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); static inline void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void her2(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); static inline void her2(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); static inline void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); static inline void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); static inline void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, float c, float s); static inline void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, double c, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, double s); static inline void rot(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, float c, float s); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, float c, float s); static inline void rot(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, double c, double s); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, double c, double s); static inline void axpy(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy); static inline void axpy(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy); static inline void axpy(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); static inline void axpy(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); static inline void axpy_batch(backend_selector selector, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + float alpha, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); static inline void axpy_batch(backend_selector selector, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + double alpha, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); static inline void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); static inline void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); static inline void axpby(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); static inline void axpby(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); static inline void axpby(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); + sycl::buffer, 1>& y, std::int64_t incy); static inline void axpby(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); + sycl::buffer, 1>& y, std::int64_t incy); static inline void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda); static inline void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda); static inline void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); static inline void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, + std::int64_t ldc); static inline void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); static inline void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& c, std::int64_t ldc); static inline void gemv(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); static inline void gemv(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy); static inline void gemv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); static inline void gemv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); static inline void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, float beta, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + float beta, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); static inline void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, double beta, - sycl::buffer &y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + double beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); static inline void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); static inline void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); static inline void dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size); + std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); static inline void dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &c, + std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); static inline void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); static inline void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); static inline void her(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a, + std::int64_t n, float alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a, std::int64_t lda); static inline void her(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a, + std::int64_t n, double alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a, std::int64_t lda); static inline void hpr(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a); + std::int64_t n, float alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a); static inline void hpr(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a); + std::int64_t n, double alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a); static inline void iamin(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); static inline void iamin(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); static inline void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); static inline void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); static inline void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, float beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, double beta, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); + double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, double beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, - sycl::buffer &c, std::int64_t ldc, + sycl::half alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::half beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void spmv(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); + std::int64_t n, float alpha, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); static inline void spmv(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); + std::int64_t n, double alpha, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); static inline void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co); static inline void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co); static inline void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, uint8_t ao, sycl::buffer &b, - std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, uint8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co); static inline void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, uint8_t ao, sycl::buffer &b, - std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, uint8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co); static inline void swap(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy); static inline void swap(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy); static inline void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); static inline void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); static inline void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda); static inline void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda); static inline void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); static inline void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); static inline void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); static inline void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); static inline void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); static inline void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); static inline void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc); static inline void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& c, std::int64_t ldc); static inline void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::half beta, - sycl::buffer &c, std::int64_t ldc); + sycl::half alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::half beta, + sycl::buffer& c, std::int64_t ldc); static inline void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); static inline void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); static inline void herk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer, 1> &a, std::int64_t lda, float beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& a, std::int64_t lda, float beta, + sycl::buffer, 1>& c, std::int64_t ldc); static inline void herk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer, 1> &a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& a, std::int64_t lda, double beta, + sycl::buffer, 1>& c, std::int64_t ldc); static inline void ger(backend_selector selector, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); + float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a, + std::int64_t lda); static inline void ger(backend_selector selector, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); + double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a, + std::int64_t lda); static inline void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); static inline void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); static inline void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); static inline void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); static inline void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); static inline void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); static inline void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); static inline void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& c, std::int64_t ldc); static inline void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a); static inline void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a); static inline void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); static inline void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); static inline void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); + sycl::buffer, 1>& y, std::int64_t incy); static inline void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); + sycl::buffer, 1>& y, std::int64_t incy); static inline void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); static inline void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); static inline void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); static inline void tbmv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); static inline void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); static inline void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, + std::int64_t ldc); static inline void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); static inline void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& c, std::int64_t ldc); static inline void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); static inline void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); static inline void syr(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a, std::int64_t lda); + std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a, std::int64_t lda); static inline void syr(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a, std::int64_t lda); + std::int64_t n, double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a, std::int64_t lda); static inline void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); static inline void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); static inline void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); static inline void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); -static inline void rotmg(backend_selector selector, - sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, float y1, - sycl::buffer ¶m); +static inline void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, float y1, + sycl::buffer& param); -static inline void rotmg(backend_selector selector, - sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, double y1, - sycl::buffer ¶m); +static inline void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, double y1, + sycl::buffer& param); static inline void tpsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); static inline void tpsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); static inline void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); static inline void tpsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); static inline void trsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); static inline void trsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); static inline void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); static inline void trsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); static inline void copy(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy); static inline void copy(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy); static inline void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); static inline void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); static inline void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size); static inline void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size); static inline void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); static inline void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); static inline void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); static inline void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); static inline void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); static inline void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); static inline void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& c, std::int64_t ldc); static inline void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& c, std::int64_t ldc); static inline void sbmv(backend_selector selector, uplo upper_lower, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); static inline void sbmv(backend_selector selector, uplo upper_lower, - std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy); static inline void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); static inline void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); static inline void asum(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); static inline void asum(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); static inline void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); static inline void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); static inline void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); static inline void tbsv(backend_selector selector, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); static inline void spr2(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a); + std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a); static inline void spr2(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a); + std::int64_t n, double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a); static inline void iamax(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); static inline void iamax(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); static inline void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); static inline void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); static inline void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size); + std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); static inline void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size); + std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); static inline void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); static inline void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); static inline void rotm(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& param); static inline void rotm(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& param); -static inline void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s); +static inline void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, + sycl::buffer& s); -static inline void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s); +static inline void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, + sycl::buffer& s); static inline void rotg(backend_selector selector, - sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s); + sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); static inline void rotg(backend_selector selector, - sycl::buffer, 1> &a, - sycl::buffer, 1> &b, - sycl::buffer &c, - sycl::buffer, 1> &s); + sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); static inline void sdsdot(backend_selector selector, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result); static inline void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc); static inline void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc); + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, double beta, sycl::buffer, 1>& c, + std::int64_t ldc); static inline void dot(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result); static inline void dot(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result); static inline void dot(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result); static inline void symv(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); static inline void symv(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy); + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); static inline void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); static inline void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); static inline void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); static inline void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); static inline void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); static inline void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); static inline void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); static inline void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); static inline void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - float beta, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + float beta, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, double beta, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); static inline void omatcopy(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb); static inline void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); static inline void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); static inline void omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); static inline void omatcopy2(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer &b, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); static inline void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); static inline void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb); static inline void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb); static inline void imatcopy(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); static inline void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); static inline void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb); static inline void imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb); static inline void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, float beta, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, + sycl::buffer& a, std::int64_t lda, float beta, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& c, std::int64_t ldc); static inline void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, double beta, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, + sycl::buffer& a, std::int64_t lda, double beta, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& c, std::int64_t ldc); static inline void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); static inline void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); // USM APIs static inline sycl::event syr2(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, float alpha, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event syr2(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, double alpha, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event scal(backend_selector selector, std::int64_t n, - float alpha, float *x, std::int64_t incx, - const std::vector &dependencies = {}); + float alpha, float* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event scal(backend_selector selector, std::int64_t n, - double alpha, double *x, std::int64_t incx, - const std::vector &dependencies = {}); + double alpha, double* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}); + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}); + std::complex alpha, std::complex* x, + std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event scal(backend_selector selector, std::int64_t n, - float alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); + float alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event scal(backend_selector selector, std::int64_t n, - double alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); + double alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event trmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, const float* a, + std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event trmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, const double* a, + std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event trmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event trmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tpmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, const float *a, - float *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, const float* a, + float* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tpmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, const double *a, - double *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, const double* a, + double* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tpmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - const std::complex *a, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::complex* x, + std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tpmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - const std::complex *a, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::complex* x, + std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event spr(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, const float *x, std::int64_t incx, - float *a, const std::vector &dependencies = {}); + std::int64_t n, float alpha, const float* x, std::int64_t incx, + float* a, const std::vector& dependencies = {}); static inline sycl::event spr(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, const double *x, std::int64_t incx, - double *a, const std::vector &dependencies = {}); + std::int64_t n, double alpha, const double* x, std::int64_t incx, + double* a, const std::vector& dependencies = {}); static inline sycl::event hpmv(backend_selector selector, uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *a, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* a, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event hpmv(backend_selector selector, uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *a, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* a, const std::complex* x, + std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event syrk(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, float beta, float *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies = {}); static inline sycl::event syrk(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, double beta, double *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies = {}); static inline sycl::event syrk(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); static inline sycl::event syrk(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -static inline sycl::event syrk_batch(backend_selector selector, - uplo *upper_lower, transpose *trans, std::int64_t *n, - std::int64_t *k, float *alpha, const float **a, - std::int64_t *lda, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event syrk_batch(backend_selector selector, - uplo *upper_lower, transpose *trans, std::int64_t *n, - std::int64_t *k, double *alpha, const double **a, - std::int64_t *lda, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event syrk_batch(backend_selector selector, - uplo *upper_lower, transpose *trans, std::int64_t *n, - std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event syrk_batch(backend_selector selector, - uplo *upper_lower, transpose *trans, std::int64_t *n, - std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event syrk_batch(backend_selector selector, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +static inline sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + float* alpha, const float** a, std::int64_t* lda, float* beta, + float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + double* alpha, const double** a, std::int64_t* lda, + double* beta, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, + std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, + std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event syrk_batch(backend_selector selector, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, std::int64_t stride_a, + float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event syrk_batch(backend_selector selector, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, std::int64_t stride_a, + double beta, double* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event syrk_batch(backend_selector selector, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); -static inline sycl::event syrk_batch(backend_selector selector, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event syrk_batch(backend_selector selector, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event syrk_batch(backend_selector selector, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); +static inline sycl::event syrk_batch(backend_selector selector, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); static inline sycl::event her2(backend_selector selector, uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event her2(backend_selector selector, uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event hbmv(backend_selector selector, uplo upper_lower, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event hbmv(backend_selector selector, uplo upper_lower, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, float c, float s, - const std::vector &dependencies = {}); + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, float c, float s, + const std::vector& dependencies = {}); static inline sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, double c, double s, - const std::vector &dependencies = {}); + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, double c, double s, + const std::vector& dependencies = {}); -static inline sycl::event rot(backend_selector selector, std::int64_t n, - float *x, std::int64_t incx, float *y, std::int64_t incy, float c, - float s, const std::vector &dependencies = {}); +static inline sycl::event rot(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float c, float s, + const std::vector& dependencies = {}); static inline sycl::event rot(backend_selector selector, std::int64_t n, - double *x, std::int64_t incx, double *y, std::int64_t incy, - double c, double s, - const std::vector &dependencies = {}); + double* x, std::int64_t incx, double* y, std::int64_t incy, double c, + double s, const std::vector& dependencies = {}); static inline sycl::event axpy(backend_selector selector, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *y, - std::int64_t incy, - const std::vector &dependencies = {}); + float alpha, const float* x, std::int64_t incx, float* y, + std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event axpy(backend_selector selector, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *y, - std::int64_t incy, - const std::vector &dependencies = {}); + double alpha, const double* x, std::int64_t incx, double* y, + std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::complex alpha, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -static inline sycl::event axpy_batch(backend_selector selector, - std::int64_t *n, float *alpha, const float **x, - std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event axpy_batch(backend_selector selector, - std::int64_t *n, double *alpha, const double **x, - std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event axpy_batch(backend_selector selector, - std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event axpy_batch(backend_selector selector, - std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event axpy_batch(backend_selector selector, - std::int64_t n, float alpha, const float *x, - std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event axpy_batch(backend_selector selector, - std::int64_t n, double alpha, const double *x, - std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event axpy_batch(backend_selector selector, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event axpy_batch(backend_selector selector, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); + std::complex alpha, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +static inline sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + float* alpha, const float** x, std::int64_t* incx, float** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + double* alpha, const double** x, std::int64_t* incx, + double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event axpy_batch(backend_selector selector, std::int64_t n, + float alpha, const float* x, std::int64_t incx, + std::int64_t stridex, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event axpy_batch(backend_selector selector, std::int64_t n, + double alpha, const double* x, std::int64_t incx, + std::int64_t stridex, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event axpy_batch(backend_selector selector, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event axpy_batch(backend_selector selector, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); static inline sycl::event axpby(backend_selector selector, std::int64_t n, - float alpha, const float *x, std::int64_t incx, - const float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); + float alpha, const float* x, std::int64_t incx, const float beta, + float* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event axpby(backend_selector selector, std::int64_t n, - double alpha, const double *x, std::int64_t incx, - const double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); + double alpha, const double* x, std::int64_t incx, const double beta, + double* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event gerc(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event gerc(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event syr2k(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event syr2k(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event syr2k(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event syr2k(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event gemv(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event gemv(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event gemv(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event gemv(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -static inline sycl::event gemv_batch(backend_selector selector, - transpose trans, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float beta, float *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +static inline sycl::event gemv_batch(backend_selector selector, transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stridea, const float* x, + std::int64_t incx, std::int64_t stridex, float beta, float* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); -static inline sycl::event gemv_batch(backend_selector selector, - transpose trans, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double beta, double *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); +static inline sycl::event gemv_batch(backend_selector selector, transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stridea, const double* x, + std::int64_t incx, std::int64_t stridex, double beta, + double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); static inline sycl::event gemv_batch( backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, - std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, std::complex beta, + std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); static inline sycl::event gemv_batch( backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies = {}); - -static inline sycl::event gemv_batch(backend_selector selector, - transpose *trans, std::int64_t *m, std::int64_t *n, - float *alpha, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float *beta, - float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event gemv_batch(backend_selector selector, - transpose *trans, std::int64_t *m, std::int64_t *n, - double *alpha, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double *beta, - double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event gemv_batch(backend_selector selector, - transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **x, - std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event gemv_batch( - backend_selector selector, transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies = {}); - -static inline sycl::event dgmm_batch(backend_selector selector, - side left_right, std::int64_t m, std::int64_t n, - const float *a, std::int64_t lda, std::int64_t stridea, - const float *x, std::int64_t incx, std::int64_t stridex, - float *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}); + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies = {}); + +static inline sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, + const float** a, std::int64_t* lda, const float** x, + std::int64_t* incx, float* beta, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, + const double** a, std::int64_t* lda, const double** x, + std::int64_t* incx, double* beta, double** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event dgmm_batch(backend_selector selector, side left_right, + std::int64_t m, std::int64_t n, const float* a, + std::int64_t lda, std::int64_t stridea, const float* x, + std::int64_t incx, std::int64_t stridex, float* c, + std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, + const std::vector& dependencies = {}); -static inline sycl::event dgmm_batch(backend_selector selector, - side left_right, std::int64_t m, std::int64_t n, - const double *a, std::int64_t lda, std::int64_t stridea, - const double *x, std::int64_t incx, std::int64_t stridex, - double *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event dgmm_batch(backend_selector selector, - side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event dgmm_batch(backend_selector selector, - side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event dgmm_batch(backend_selector selector, - side *left_right, std::int64_t *m, std::int64_t *n, - const float **a, std::int64_t *lda, const float **x, - std::int64_t *incx, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event dgmm_batch(backend_selector selector, - side *left_right, std::int64_t *m, std::int64_t *n, - const double **a, std::int64_t *lda, const double **x, - std::int64_t *incx, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event dgmm_batch(backend_selector selector, - side *left_right, std::int64_t *m, std::int64_t *n, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event dgmm_batch(backend_selector selector, - side *left_right, std::int64_t *m, std::int64_t *n, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); +static inline sycl::event dgmm_batch(backend_selector selector, side left_right, + std::int64_t m, std::int64_t n, const double* a, + std::int64_t lda, std::int64_t stridea, const double* x, + std::int64_t incx, std::int64_t stridex, double* c, + std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event dgmm_batch(backend_selector selector, side left_right, + std::int64_t m, std::int64_t n, const std::complex* a, + std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event dgmm_batch(backend_selector selector, side left_right, + std::int64_t m, std::int64_t n, const std::complex* a, + std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, + std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const float** a, + std::int64_t* lda, const float** x, std::int64_t* incx, + float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const double** a, + std::int64_t* lda, const double** x, std::int64_t* incx, + double** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); static inline sycl::event her(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, const std::complex *x, - std::int64_t incx, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, float alpha, const std::complex* x, + std::int64_t incx, std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event her(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, const std::complex *x, - std::int64_t incx, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, double alpha, const std::complex* x, + std::int64_t incx, std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event hpr(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, const std::complex *x, - std::int64_t incx, std::complex *a, - const std::vector &dependencies = {}); + std::int64_t n, float alpha, const std::complex* x, + std::int64_t incx, std::complex* a, + const std::vector& dependencies = {}); static inline sycl::event hpr(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, const std::complex *x, - std::int64_t incx, std::complex *a, - const std::vector &dependencies = {}); + std::int64_t n, double alpha, const std::complex* x, + std::int64_t incx, std::complex* a, + const std::vector& dependencies = {}); static inline sycl::event iamin(backend_selector selector, std::int64_t n, - const float *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); + const float* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); static inline sycl::event iamin(backend_selector selector, std::int64_t n, - const double *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); + const double* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); static inline sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies = {}); static inline sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); - -static inline sycl::event gemm_batch(backend_selector selector, - transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, - const float **a, std::int64_t *lda, const float **b, - std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event gemm_batch(backend_selector selector, - transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, double *alpha, - const double **a, std::int64_t *lda, const double **b, - std::int64_t *ldb, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event gemm_batch(backend_selector selector, - transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, - std::int64_t *ldb, std::complex *beta, - std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float* alpha, const float** a, + std::int64_t* lda, const float** b, std::int64_t* ldb, + float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, double* alpha, const double** a, + std::int64_t* lda, const double** b, std::int64_t* ldb, + double* beta, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, sycl::half* alpha, const sycl::half** a, + std::int64_t* lda, const sycl::half** b, std::int64_t* ldb, + sycl::half* beta, sycl::half** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float* alpha, const sycl::half** a, + std::int64_t* lda, const sycl::half** b, std::int64_t* ldb, + float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float* alpha, const std::int8_t** a, + std::int64_t* lda, const std::int8_t** b, std::int64_t* ldb, + float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float* alpha, const std::int8_t** a, + std::int64_t* lda, const std::int8_t** b, std::int64_t* ldb, + float* beta, std::int32_t** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); -static inline sycl::event gemm_batch( - backend_selector selector, transpose *transa, transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, const std::complex **b, - std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event gemm_batch(backend_selector selector, - transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, sycl::half *alpha, - const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, sycl::half *beta, - sycl::half **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float *alpha, const sycl::half **a, - std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, - float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float *alpha, const std::int8_t **a, - std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, - float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float *alpha, const std::int8_t **a, - std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, - float *beta, std::int32_t **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event gemm_batch(backend_selector selector, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, - const float *b, std::int64_t ldb, std::int64_t stride_b, - float beta, float *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event gemm_batch(backend_selector selector, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, - const double *b, std::int64_t ldb, std::int64_t stride_b, - double beta, double *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); +static inline sycl::event gemm_batch(backend_selector selector, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, const float* b, std::int64_t ldb, + std::int64_t stride_b, float beta, float* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_batch(backend_selector selector, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, const double* b, + std::int64_t ldb, std::int64_t stride_b, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); static inline sycl::event gemm_batch( backend_selector selector, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, + std::int64_t n, std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event gemm_batch( backend_selector selector, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::complex beta, std::complex *c, std::int64_t ldc, + std::int64_t n, std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -static inline sycl::event gemm_batch( - backend_selector selector, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, - std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, - sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}); +static inline sycl::event gemm_batch(backend_selector selector, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::half alpha, const sycl::half* a, + std::int64_t lda, std::int64_t stride_a, const sycl::half* b, + std::int64_t ldb, std::int64_t stride_b, sycl::half beta, + sycl::half* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); static inline sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, std::int64_t stride_a, const sycl::half *b, - std::int64_t ldb, std::int64_t stride_b, float beta, float *c, + std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, std::int64_t stride_a, const sycl::half* b, + std::int64_t ldb, std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, - std::int64_t ldb, std::int64_t stride_b, float beta, float *c, + std::int64_t k, float alpha, const std::int8_t* a, + std::int64_t lda, std::int64_t stride_a, const std::int8_t* b, + std::int64_t ldb, std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, + std::int64_t k, float alpha, const std::int8_t* a, + std::int64_t lda, std::int64_t stride_a, const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, - std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, + std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event spmv(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, const float *a, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, float alpha, const float* a, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event spmv(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, const double *a, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, double alpha, const double* a, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event swap(backend_selector selector, std::int64_t n, - float *x, std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies = {}); + float* x, std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event swap(backend_selector selector, std::int64_t n, - double *x, std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies = {}); + double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event geru(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event geru(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies = {}); static inline sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies = {}); static inline sycl::event nrm2(backend_selector selector, std::int64_t n, - const float *x, std::int64_t incx, float *result, - const std::vector &dependencies = {}); + const float* x, std::int64_t incx, float* result, + const std::vector& dependencies = {}); static inline sycl::event nrm2(backend_selector selector, std::int64_t n, - const double *x, std::int64_t incx, double *result, - const std::vector &dependencies = {}); + const double* x, std::int64_t incx, double* result, + const std::vector& dependencies = {}); static inline sycl::event gemm(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event gemm(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event gemm(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event gemm(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event gemm(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, const sycl::half *a, std::int64_t lda, - const sycl::half *b, std::int64_t ldb, sycl::half beta, - sycl::half *c, std::int64_t ldc, - const std::vector &dependencies = {}); + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::half alpha, const sycl::half* a, std::int64_t lda, + const sycl::half* b, std::int64_t ldb, sycl::half beta, + sycl::half* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event gemm(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const sycl::half *a, std::int64_t lda, - const sycl::half *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const sycl::half* a, std::int64_t lda, + const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies = {}); static inline sycl::event gemm(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const bfloat16 *a, std::int64_t lda, - const bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const bfloat16* a, std::int64_t lda, const bfloat16* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event herk(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, - const std::complex *a, std::int64_t lda, float beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, float alpha, + const std::complex* a, std::int64_t lda, float beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event herk(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, double alpha, - const std::complex *a, std::int64_t lda, double beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, double alpha, + const std::complex* a, std::int64_t lda, double beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event ger(backend_selector selector, std::int64_t m, - std::int64_t n, float alpha, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, float alpha, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event ger(backend_selector selector, std::int64_t m, - std::int64_t n, double alpha, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, double alpha, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event trsm(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies = {}); + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + float* b, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event trsm(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies = {}); + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + double* b, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event trsm(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event trsm(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -static inline sycl::event trsm_batch(backend_selector selector, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, float *b, - int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event trsm_batch(backend_selector selector, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, double *b, - int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event trsm_batch(backend_selector selector, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, - int64_t lda, int64_t stride_a, std::complex *b, - int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event trsm_batch(backend_selector selector, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, - int64_t lda, int64_t stride_a, std::complex *b, - int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event trsm_batch(backend_selector selector, - side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, float *alpha, - const float **a, int64_t *lda, float **b, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event trsm_batch(backend_selector selector, - side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, double *alpha, - const double **a, int64_t *lda, double **b, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event trsm_batch(backend_selector selector, - side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex **b, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event trsm_batch(backend_selector selector, - side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, - std::complex *alpha, - const std::complex **a, int64_t *lda, - std::complex **b, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +static inline sycl::event trsm_batch(backend_selector selector, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, int64_t m, + int64_t n, float alpha, const float* a, int64_t lda, + int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event trsm_batch(backend_selector selector, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, int64_t m, + int64_t n, double alpha, const double* a, int64_t lda, + int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event trsm_batch(backend_selector selector, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, int64_t m, + int64_t n, std::complex alpha, + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event trsm_batch(backend_selector selector, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, int64_t m, + int64_t n, std::complex alpha, + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, + int64_t* m, int64_t* n, float* alpha, const float** a, + int64_t* lda, float** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, + int64_t* m, int64_t* n, double* alpha, const double** a, + int64_t* lda, double** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, + int64_t* m, int64_t* n, std::complex* alpha, + const std::complex** a, int64_t* lda, + std::complex** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, + int64_t* m, int64_t* n, std::complex* alpha, + const std::complex** a, int64_t* lda, + std::complex** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, + const std::vector& dependencies = {}); static inline sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); static inline sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); static inline sycl::event hemm(backend_selector selector, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event hemm(backend_selector selector, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event hpr2(backend_selector selector, uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies = {}); static inline sycl::event hpr2(backend_selector selector, uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies = {}); static inline sycl::event gbmv(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event gbmv(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event gbmv(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event gbmv(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event tbmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - const float *a, std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tbmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - const double *a, std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tbmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tbmv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event symm(backend_selector selector, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); + uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event symm(backend_selector selector, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); + uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, const double* b, std::int64_t ldb, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event symm(backend_selector selector, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event symm(backend_selector selector, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); static inline sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); static inline sycl::event syr(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, const float *x, std::int64_t incx, - float *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, float alpha, const float* x, std::int64_t incx, + float* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event syr(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, const double *x, std::int64_t incx, - double *a, std::int64_t lda, - const std::vector &dependencies = {}); + std::int64_t n, double alpha, const double* x, std::int64_t incx, + double* a, std::int64_t lda, + const std::vector& dependencies = {}); static inline sycl::event trmm(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies = {}); + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + float* b, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event trmm(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies = {}); + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + double* b, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event trmm(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event trmm(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); -static inline sycl::event rotmg(backend_selector selector, float *d1, - float *d2, float *x1, float y1, float *param, - const std::vector &dependencies = {}); +static inline sycl::event rotmg(backend_selector selector, float* d1, float* d2, + float* x1, float y1, float* param, + const std::vector& dependencies = {}); -static inline sycl::event rotmg(backend_selector selector, double *d1, - double *d2, double *x1, double y1, double *param, - const std::vector &dependencies = {}); +static inline sycl::event rotmg(backend_selector selector, double* d1, double* d2, + double* x1, double y1, double* param, + const std::vector& dependencies = {}); static inline sycl::event tpsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, const float *a, - float *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, const float* a, + float* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tpsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, const double *a, - double *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, const double* a, + double* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tpsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - const std::complex *a, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::complex* x, + std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tpsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - const std::complex *a, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::complex* x, + std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event trsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, const float* a, + std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event trsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, const double* a, + std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event trsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event trsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event copy(backend_selector selector, std::int64_t n, - const float *x, std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies = {}); + const float* x, std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event copy(backend_selector selector, std::int64_t n, - const double *x, std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies = {}); + const double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -static inline sycl::event copy_batch(backend_selector selector, - std::int64_t *n, const float **x, std::int64_t *incx, - float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event copy_batch(backend_selector selector, - std::int64_t *n, const double **x, std::int64_t *incx, - double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event copy_batch(backend_selector selector, - std::int64_t *n, const std::complex **x, - std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event copy_batch(backend_selector selector, - std::int64_t *n, const std::complex **x, - std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -static inline sycl::event copy_batch(backend_selector selector, - std::int64_t n, const float *x, std::int64_t incx, - std::int64_t stridex, float *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event copy_batch(backend_selector selector, - std::int64_t n, const double *x, std::int64_t incx, - std::int64_t stridex, double *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event copy_batch(backend_selector selector, - std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -static inline sycl::event copy_batch(backend_selector selector, - std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +static inline sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const float** x, std::int64_t* incx, float** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const double** x, std::int64_t* incx, double** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +static inline sycl::event copy_batch(backend_selector selector, std::int64_t n, + const float* x, std::int64_t incx, std::int64_t stridex, + float* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event copy_batch(backend_selector selector, std::int64_t n, + const double* x, std::int64_t incx, std::int64_t stridex, + double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event copy_batch(backend_selector selector, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +static inline sycl::event copy_batch(backend_selector selector, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); static inline sycl::event hemv(backend_selector selector, uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event hemv(backend_selector selector, uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -static inline sycl::event gemm_bias( - backend_selector selector, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const std::int8_t *a, int64_t lda, std::int8_t ao, - const std::uint8_t *b, int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, int64_t ldc, - const std::int32_t *co, const std::vector &dependencies = {}); - -static inline sycl::event gemm_bias( - backend_selector selector, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const std::int8_t *a, int64_t lda, std::int8_t ao, - const std::int8_t *b, int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, int64_t ldc, - const std::int32_t *co, const std::vector &dependencies = {}); - -static inline sycl::event gemm_bias( - backend_selector selector, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const std::uint8_t *a, int64_t lda, - std::uint8_t ao, const std::int8_t *b, int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, - int64_t ldc, const std::int32_t *co, const std::vector &dependencies = {}); - -static inline sycl::event gemm_bias(backend_selector selector, - transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, - const std::uint8_t *a, int64_t lda, std::uint8_t ao, - const std::uint8_t *b, int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, int64_t ldc, - const std::int32_t *co, - const std::vector &dependencies = {}); + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_bias(backend_selector selector, transpose transa, + transpose transb, offset offsetc, int64_t m, int64_t n, + int64_t k, float alpha, const std::int8_t* a, int64_t lda, + std::int8_t ao, const std::uint8_t* b, int64_t ldb, + std::uint8_t bo, float beta, std::int32_t* c, int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_bias(backend_selector selector, transpose transa, + transpose transb, offset offsetc, int64_t m, int64_t n, + int64_t k, float alpha, const std::int8_t* a, int64_t lda, + std::int8_t ao, const std::int8_t* b, int64_t ldb, + std::int8_t bo, float beta, std::int32_t* c, int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_bias(backend_selector selector, transpose transa, + transpose transb, offset offsetc, int64_t m, int64_t n, + int64_t k, float alpha, const std::uint8_t* a, int64_t lda, + std::uint8_t ao, const std::int8_t* b, int64_t ldb, + std::int8_t bo, float beta, std::int32_t* c, int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}); + +static inline sycl::event gemm_bias(backend_selector selector, transpose transa, + transpose transb, offset offsetc, int64_t m, int64_t n, + int64_t k, float alpha, const std::uint8_t* a, int64_t lda, + std::uint8_t ao, const std::uint8_t* b, int64_t ldb, + std::uint8_t bo, float beta, std::int32_t* c, int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}); static inline sycl::event sbmv(backend_selector selector, uplo upper_lower, - std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event sbmv(backend_selector selector, uplo upper_lower, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies = {}); static inline sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies = {}); static inline sycl::event asum(backend_selector selector, std::int64_t n, - const float *x, std::int64_t incx, float *result, - const std::vector &dependencies = {}); + const float* x, std::int64_t incx, float* result, + const std::vector& dependencies = {}); static inline sycl::event asum(backend_selector selector, std::int64_t n, - const double *x, std::int64_t incx, double *result, - const std::vector &dependencies = {}); + const double* x, std::int64_t incx, double* result, + const std::vector& dependencies = {}); static inline sycl::event tbsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - const float *a, std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tbsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - const double *a, std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tbsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event tbsv(backend_selector selector, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); static inline sycl::event spr2(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *a, - const std::vector &dependencies = {}); + std::int64_t n, float alpha, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* a, + const std::vector& dependencies = {}); static inline sycl::event spr2(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *a, - const std::vector &dependencies = {}); + std::int64_t n, double alpha, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* a, + const std::vector& dependencies = {}); static inline sycl::event iamax(backend_selector selector, std::int64_t n, - const float *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); + const float* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); static inline sycl::event iamax(backend_selector selector, std::int64_t n, - const double *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); + const double* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); static inline sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies = {}); static inline sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies = {}); static inline sycl::event rotm(backend_selector selector, std::int64_t n, - float *x, std::int64_t incx, float *y, std::int64_t incy, - float *param, - const std::vector &dependencies = {}); + float* x, std::int64_t incx, float* y, std::int64_t incy, + float* param, const std::vector& dependencies = {}); static inline sycl::event rotm(backend_selector selector, std::int64_t n, - double *x, std::int64_t incx, double *y, std::int64_t incy, - double *param, - const std::vector &dependencies = {}); + double* x, std::int64_t incx, double* y, std::int64_t incy, + double* param, const std::vector& dependencies = {}); -static inline sycl::event rotg(backend_selector selector, float *a, float *b, - float *c, float *s, - const std::vector &dependencies = {}); +static inline sycl::event rotg(backend_selector selector, float* a, float* b, + float* c, float* s, + const std::vector& dependencies = {}); -static inline sycl::event rotg(backend_selector selector, double *a, - double *b, double *c, double *s, - const std::vector &dependencies = {}); +static inline sycl::event rotg(backend_selector selector, double* a, double* b, + double* c, double* s, + const std::vector& dependencies = {}); -static inline sycl::event rotg(backend_selector selector, - std::complex *a, std::complex *b, float *c, - std::complex *s, - const std::vector &dependencies = {}); +static inline sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, float* c, std::complex* s, + const std::vector& dependencies = {}); -static inline sycl::event rotg(backend_selector selector, - std::complex *a, std::complex *b, double *c, - std::complex *s, - const std::vector &dependencies = {}); +static inline sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, double* c, std::complex* s, + const std::vector& dependencies = {}); static inline sycl::event sdsdot(backend_selector selector, std::int64_t n, - float sb, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *result, - const std::vector &dependencies = {}); + float sb, const float* x, std::int64_t incx, const float* y, + std::int64_t incy, float* result, + const std::vector& dependencies = {}); static inline sycl::event her2k(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, float beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event her2k(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, double beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event dot(backend_selector selector, std::int64_t n, - const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *result, - const std::vector &dependencies = {}); + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* result, const std::vector& dependencies = {}); static inline sycl::event dot(backend_selector selector, std::int64_t n, - const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *result, - const std::vector &dependencies = {}); + const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* result, + const std::vector& dependencies = {}); static inline sycl::event dot(backend_selector selector, std::int64_t n, - const float *x, std::int64_t incx, const float *y, - std::int64_t incy, double *result, - const std::vector &dependencies = {}); + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + double* result, const std::vector& dependencies = {}); static inline sycl::event symv(backend_selector selector, uplo upper_lower, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, float alpha, const float* a, std::int64_t lda, + const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event symv(backend_selector selector, uplo upper_lower, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, - const std::vector &dependencies = {}); + std::int64_t n, double alpha, const double* a, std::int64_t lda, + const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, + const std::vector& dependencies = {}); static inline sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, + float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, + double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, - float alpha, float *ab, std::int64_t lda, std::int64_t ldb, + float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, - double alpha, double *ab, std::int64_t lda, + double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, std::complex *ab, + std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, std::complex *ab, + std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, float alpha, const float *a, + std::int64_t n, float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, - float *c, std::int64_t ldc, std::int64_t stride_c, + const float* b, std::int64_t ldb, std::int64_t stride_b, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, double alpha, const double *a, + std::int64_t n, double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, double beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, - double *c, std::int64_t ldc, std::int64_t stride_c, + const double* b, std::int64_t ldb, std::int64_t stride_b, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event omatadd_batch( backend_selector selector, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, const std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}); + std::int64_t n, std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies = {}); static inline sycl::event omatadd_batch( backend_selector selector, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, + std::int64_t n, std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, const std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event omatcopy(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies = {}); + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event omatcopy(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies = {}); + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event omatcopy2(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stridea, float *b, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event omatcopy2(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, double *b, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stridea, double* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, std::int64_t ldb, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, std::int64_t ldb, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event imatcopy(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, + std::int64_t m, std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event imatcopy(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, + std::int64_t m, std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); static inline sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); + std::complex* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); + std::complex* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); static inline sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float beta, const float *b, - std::int64_t ldb, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); + const float* a, std::int64_t lda, float beta, const float* b, + std::int64_t ldb, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double beta, const double *b, - std::int64_t ldb, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); + const double* a, std::int64_t lda, double beta, const double* b, + std::int64_t ldb, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + const std::complex* b, std::int64_t ldb, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + const std::complex* b, std::int64_t ldb, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); static inline sycl::event omatcopy_batch(backend_selector selector, transpose* trans, std::int64_t* m, std::int64_t* n, diff --git a/include/oneapi/math/blas/detail/blas_loader.hxx b/include/oneapi/math/blas/detail/blas_loader.hxx index 9732218de..f37cc32a9 100644 --- a/include/oneapi/math/blas/detail/blas_loader.hxx +++ b/include/oneapi/math/blas/detail/blas_loader.hxx @@ -19,2681 +19,2571 @@ // Buffer APIs -ONEMATH_EXPORT void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer, 1> &a, std::int64_t lda, float beta, - sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer, 1> &a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx); -ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx); -ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - float alpha, sycl::buffer, 1> &x, - std::int64_t incx); -ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - double alpha, sycl::buffer, 1> &x, - std::int64_t incx); - -ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); -ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); -ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a); -ONEMATH_EXPORT void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a); - -ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, double beta, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size); -ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, float beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, double beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc); -ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc); - -ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, float beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -ONEMATH_EXPORT void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); -ONEMATH_EXPORT void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, - float s); -ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, double c, - double s); -ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, float c, float s); -ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, double c, double s); - -ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy); -ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -ONEMATH_EXPORT void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer, 1>& a, std::int64_t lda, float beta, + sycl::buffer, 1>& c, std::int64_t ldc); +ONEMATH_EXPORT void herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, double alpha, + sycl::buffer, 1>& a, std::int64_t lda, double beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx); +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx); +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx); +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx); +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float alpha, sycl::buffer, 1>& x, std::int64_t incx); +ONEMATH_EXPORT void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double alpha, sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); +ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); +ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); +ONEMATH_EXPORT void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); +ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx); +ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); +ONEMATH_EXPORT void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a); +ONEMATH_EXPORT void spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a); + +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, float beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, double beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::half alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::half beta, + sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size); +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size); +ONEMATH_EXPORT void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, + sycl::buffer& a, std::int64_t lda, float beta, + sycl::buffer& c, std::int64_t ldc); +ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, + sycl::buffer& a, std::int64_t lda, double beta, + sycl::buffer& c, std::int64_t ldc); +ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); -ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +ONEMATH_EXPORT void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + float beta, sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); +ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, double alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + double beta, sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); +ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); +ONEMATH_EXPORT void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); +ONEMATH_EXPORT void her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +ONEMATH_EXPORT void hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); +ONEMATH_EXPORT void hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); + +ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, float c, + float s); +ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, + double s); +ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, float c, float s); +ONEMATH_EXPORT void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, double c, double s); + +ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); +ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); +ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy); +ONEMATH_EXPORT void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy); + +ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); +ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); +ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); +ONEMATH_EXPORT void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); + +ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); +ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); +ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); +ONEMATH_EXPORT void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +ONEMATH_EXPORT void gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); +ONEMATH_EXPORT void gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc); +ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, + std::int64_t ldc); +ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); +ONEMATH_EXPORT void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); +ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy); +ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); +ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); + +ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + float beta, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); + +ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + double beta, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); + +ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); + +ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); + +ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); + +ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); + +ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size); + +ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size); + +ONEMATH_EXPORT void her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a, + std::int64_t lda); +ONEMATH_EXPORT void her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a, + std::int64_t lda); + +ONEMATH_EXPORT void hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a); +ONEMATH_EXPORT void hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a); + +ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co); +ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co); +ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, uint8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co); +ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, uint8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co); + +ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); +ONEMATH_EXPORT void hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); + +ONEMATH_EXPORT void spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); +ONEMATH_EXPORT void spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + +ONEMATH_EXPORT void rotmg(oneapi::math::device libkey, sycl::queue& queue, + sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, float y1, sycl::buffer& param); +ONEMATH_EXPORT void rotmg(oneapi::math::device libkey, sycl::queue& queue, + sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, double y1, sycl::buffer& param); + +ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy); +ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy); +ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); +ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + +ONEMATH_EXPORT void geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); +ONEMATH_EXPORT void geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); +ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); +ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); +ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); -ONEMATH_EXPORT void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, float beta, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + sycl::buffer, 1>& c, std::int64_t ldc); +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::half alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::half beta, + sycl::buffer& c, std::int64_t ldc); +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); +ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a, + std::int64_t lda); +ONEMATH_EXPORT void syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda); + +ONEMATH_EXPORT void ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a, + std::int64_t lda); +ONEMATH_EXPORT void ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a, + std::int64_t lda); + +ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); +ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); +ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); +ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + +ONEMATH_EXPORT void dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); +ONEMATH_EXPORT void dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + +ONEMATH_EXPORT void hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); +ONEMATH_EXPORT void hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a); +ONEMATH_EXPORT void hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a); + +ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); +ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); +ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); +ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); +ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); +ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); +ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); +ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, + std::int64_t ldc); +ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); +ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); +ONEMATH_EXPORT void dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + +ONEMATH_EXPORT void syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a, std::int64_t lda); +ONEMATH_EXPORT void syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a, std::int64_t lda); + +ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); +ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); +ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); +ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + +ONEMATH_EXPORT void symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); +ONEMATH_EXPORT void symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + +ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); +ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx); +ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); +ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); +ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); +ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); +ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy); +ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy); +ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); +ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + +ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size); + +ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size); + +ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + +ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + +ONEMATH_EXPORT void hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); +ONEMATH_EXPORT void hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); + +ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); +ONEMATH_EXPORT void sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy); + +ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); +ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); +ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); +ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); +ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a); +ONEMATH_EXPORT void spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a); + +ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); +ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); +ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); +ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); + +ONEMATH_EXPORT void rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& param); +ONEMATH_EXPORT void rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& param); + +ONEMATH_EXPORT void dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result); +ONEMATH_EXPORT void dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result); +ONEMATH_EXPORT void dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result); + +ONEMATH_EXPORT void sdsdot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float sb, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + +ONEMATH_EXPORT void her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, float beta, sycl::buffer, 1>& c, + std::int64_t ldc); +ONEMATH_EXPORT void her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, double beta, sycl::buffer, 1>& c, + std::int64_t ldc); + +ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, + sycl::buffer& s); +ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue& queue, + sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s); +ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue& queue, + sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); +ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue& queue, + sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); + +ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); +ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); +ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); +ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); + +ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, + std::int64_t stride, std::int64_t batch_size); +ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, + std::int64_t stride, std::int64_t batch_size); +ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, + std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); +ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, + std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, double beta, - sycl::buffer &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size); +ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, float beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); +ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, double beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); +ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); +ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb); +ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); +ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); +ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); -ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); +ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); +ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t strideb); +ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size); - -ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); - -ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size); - -ONEMATH_EXPORT void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size); - -ONEMATH_EXPORT void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a, - std::int64_t lda); -ONEMATH_EXPORT void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a, - std::int64_t lda); - -ONEMATH_EXPORT void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a); -ONEMATH_EXPORT void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a); - -ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); -ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); -ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, uint8_t ao, sycl::buffer &b, - std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); -ONEMATH_EXPORT void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, uint8_t ao, sycl::buffer &b, - std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); - -ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - -ONEMATH_EXPORT void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); -ONEMATH_EXPORT void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void rotmg(oneapi::math::device libkey, sycl::queue &queue, - sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, float y1, - sycl::buffer ¶m); -ONEMATH_EXPORT void rotmg(oneapi::math::device libkey, sycl::queue &queue, - sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, double y1, - sycl::buffer ¶m); - -ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - -ONEMATH_EXPORT void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); -ONEMATH_EXPORT void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); - -ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::half beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - -ONEMATH_EXPORT void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - -ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - -ONEMATH_EXPORT void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); -ONEMATH_EXPORT void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - -ONEMATH_EXPORT void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc); -ONEMATH_EXPORT void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a); -ONEMATH_EXPORT void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a); - -ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); -ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc); -ONEMATH_EXPORT void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); -ONEMATH_EXPORT void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - -ONEMATH_EXPORT void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a, std::int64_t lda); -ONEMATH_EXPORT void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a, std::int64_t lda); - -ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - -ONEMATH_EXPORT void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); -ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); -ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); -ONEMATH_EXPORT void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - -ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); -ONEMATH_EXPORT void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - -ONEMATH_EXPORT void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); -ONEMATH_EXPORT void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); -ONEMATH_EXPORT void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); - -ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); -ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); -ONEMATH_EXPORT void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a); -ONEMATH_EXPORT void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a); - -ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - -ONEMATH_EXPORT void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m); -ONEMATH_EXPORT void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m); - -ONEMATH_EXPORT void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); -ONEMATH_EXPORT void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); -ONEMATH_EXPORT void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); - -ONEMATH_EXPORT void sdsdot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - float sb, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); - -ONEMATH_EXPORT void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue &queue, - sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s); -ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue &queue, - sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s); -ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue &queue, - sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s); -ONEMATH_EXPORT void rotg(oneapi::math::device libkey, sycl::queue &queue, - sycl::buffer, 1> &a, - sycl::buffer, 1> &b, - sycl::buffer &c, - sycl::buffer, 1> &s); - -ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -ONEMATH_EXPORT void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - -ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, - std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, - std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - -ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - float beta, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); -ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); -ONEMATH_EXPORT void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - -ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer &b, - std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t strideb); + +ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, + sycl::buffer, 1>& ab, std::int64_t lda, + std::int64_t ldb); +ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t strideb); - -ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &ab, - std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, - std::int64_t ldb); -ONEMATH_EXPORT void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, - std::int64_t ldb); - -ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, float beta, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, - std::int64_t ldc); -ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, double beta, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, - std::int64_t ldc); -ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); -ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& ab, std::int64_t lda, + std::int64_t ldb); + +ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, float beta, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& c, + std::int64_t ldc); +ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, double beta, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc); +ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); +ONEMATH_EXPORT void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); // USM APIs -ONEMATH_EXPORT sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, const std::complex *a, - std::int64_t lda, float beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, const std::complex *a, - std::int64_t lda, double beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, float alpha, float *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, double alpha, double *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, float alpha, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, double alpha, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, float alpha, const float *x, - std::int64_t incx, float *a, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, double alpha, const double *x, - std::int64_t incx, double *a, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, - const float **a, std::int64_t *lda, const float **b, - std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, double *alpha, - const double **a, std::int64_t *lda, const double **b, - std::int64_t *ldb, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, - std::int64_t *ldb, std::complex *beta, - std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, float alpha, + const std::complex* a, std::int64_t lda, float beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, double alpha, + const std::complex* a, std::int64_t lda, double beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float alpha, float* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double alpha, double* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, std::complex* x, + std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, std::complex* x, + std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, const float* a, + std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, const double* a, + std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, const float* a, + float* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, const double* a, + double* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::complex* x, + std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::complex* x, + std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, const float* x, std::int64_t incx, + float* a, const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, const double* x, std::int64_t incx, + double* a, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, + const float** a, std::int64_t* lda, const float** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, double* alpha, + const double** a, std::int64_t* lda, const double** b, + std::int64_t* ldb, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, sycl::half* alpha, + const sycl::half** a, std::int64_t* lda, const sycl::half** b, + std::int64_t* ldb, sycl::half* beta, sycl::half** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, + const sycl::half** a, std::int64_t* lda, const sycl::half** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, + const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, + float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, + const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, + std::int32_t** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, const float* b, + std::int64_t ldb, std::int64_t stride_b, float beta, float* c, + std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, const double* b, + std::int64_t ldb, std::int64_t stride_b, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event gemm_batch( - oneapi::math::device libkey, sycl::queue &queue, transpose *transa, transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, const std::complex **b, - std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, sycl::half *alpha, - const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, sycl::half *beta, - sycl::half **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, - const sycl::half **a, std::int64_t *lda, const sycl::half **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, - const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, - float **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, - const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, - std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, - const float *b, std::int64_t ldb, std::int64_t stride_b, - float beta, float *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, - const double *b, std::int64_t ldb, std::int64_t stride_b, - double beta, double *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch( - oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, + oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event gemm_batch( - oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, + oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch( - oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, - std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, - const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, - float beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, - const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, - float beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, - const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, - float beta, std::int32_t *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, - uplo *upper_lower, transpose *trans, std::int64_t *n, - std::int64_t *k, float *alpha, const float **a, - std::int64_t *lda, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, - uplo *upper_lower, transpose *trans, std::int64_t *n, - std::int64_t *k, double *alpha, const double **a, - std::int64_t *lda, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, - uplo *upper_lower, transpose *trans, std::int64_t *n, - std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, - uplo *upper_lower, transpose *trans, std::int64_t *n, - std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, float c, float s, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, double c, double s, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, float *x, std::int64_t incx, float *y, - std::int64_t incy, float c, float s, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, double *x, std::int64_t incx, double *y, - std::int64_t incy, double c, double s, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, float alpha, const float *x, std::int64_t incx, - float *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, double alpha, const double *x, std::int64_t incx, - double *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, float *alpha, const float **x, - std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, double *alpha, const double **x, - std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, float alpha, const float *x, - std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, double alpha, const double *x, - std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, float alpha, const float *x, std::int64_t incx, - const float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double beta, double *y, - std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float beta, float *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double beta, double *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch( - oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::half alpha, + const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, + sycl::half beta, sycl::half* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, + const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, + float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, + const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, + float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, + const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, + float beta, std::int32_t* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, float beta, float* c, + std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, double beta, double* c, + std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, + uplo* upper_lower, transpose* trans, std::int64_t* n, + std::int64_t* k, float* alpha, const float** a, + std::int64_t* lda, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, + uplo* upper_lower, transpose* trans, std::int64_t* n, + std::int64_t* k, double* alpha, const double** a, + std::int64_t* lda, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, + uplo* upper_lower, transpose* trans, std::int64_t* n, + std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, + uplo* upper_lower, transpose* trans, std::int64_t* n, + std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, + uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float beta, float* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, + uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, + uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, + std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, + uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, + std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, float c, float s, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, double c, double s, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float* x, std::int64_t incx, float* y, std::int64_t incy, float c, + float s, const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double* x, std::int64_t incx, double* y, std::int64_t incy, double c, + double s, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float alpha, const float* x, std::int64_t incx, float* y, + std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double alpha, const double* x, std::int64_t incx, double* y, + std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, float* alpha, const float** x, + std::int64_t* incx, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, double* alpha, const double** x, + std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, std::complex* alpha, + const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, std::complex* alpha, + const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, float alpha, const float* x, + std::int64_t incx, std::int64_t stridex, float* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, double alpha, const double* x, + std::int64_t incx, std::int64_t stridex, double* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float alpha, const float* x, std::int64_t incx, const float beta, + float* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double alpha, const double* x, std::int64_t incx, + const double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, std::int64_t stridea, + const float* x, std::int64_t incx, std::int64_t stridex, + float beta, float* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, std::int64_t stridea, + const double* x, std::int64_t incx, std::int64_t stridex, + double beta, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event gemv_batch( - oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose *trans, std::int64_t *m, std::int64_t *n, - float *alpha, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float *beta, - float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose *trans, std::int64_t *m, std::int64_t *n, - double *alpha, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double *beta, - double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **x, - std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); + oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event gemv_batch( - oneapi::math::device libkey, sycl::queue &queue, transpose *trans, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, - side left_right, std::int64_t m, std::int64_t n, - const float *a, std::int64_t lda, std::int64_t stridea, - const float *x, std::int64_t incx, std::int64_t stridex, - float *c, std::int64_t ldc, std::int64_t stridec, + oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* trans, std::int64_t* m, std::int64_t* n, + float* alpha, const float** a, std::int64_t* lda, + const float** x, std::int64_t* incx, float* beta, float** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* trans, std::int64_t* m, std::int64_t* n, + double* alpha, const double** a, std::int64_t* lda, + const double** x, std::int64_t* incx, double* beta, + double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** x, + std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** x, + std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, + side left_right, std::int64_t m, std::int64_t n, + const float* a, std::int64_t lda, std::int64_t stridea, + const float* x, std::int64_t incx, std::int64_t stridex, + float* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, + side left_right, std::int64_t m, std::int64_t n, + const double* a, std::int64_t lda, std::int64_t stridea, + const double* x, std::int64_t incx, std::int64_t stridex, + double* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, + side left_right, std::int64_t m, std::int64_t n, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, + std::complex* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, + side left_right, std::int64_t m, std::int64_t n, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, + std::complex* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, + side* left_right, std::int64_t* m, std::int64_t* n, + const float** a, std::int64_t* lda, const float** x, + std::int64_t* incx, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, + side* left_right, std::int64_t* m, std::int64_t* n, + const double** a, std::int64_t* lda, const double** x, + std::int64_t* incx, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, + side* left_right, std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, + side* left_right, std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, const std::complex* x, + std::int64_t incx, std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, const std::complex* x, + std::int64_t incx, std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, const std::complex* x, + std::int64_t incx, std::complex* a, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, const std::complex* x, + std::int64_t incx, std::complex* a, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const float* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const double* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + const std::complex* a, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + const std::complex* a, const std::complex* x, + std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, const float* a, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, const double* a, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rotmg(oneapi::math::device libkey, sycl::queue& queue, float* d1, + float* d2, float* x1, float y1, float* param, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event rotmg(oneapi::math::device libkey, sycl::queue& queue, double* d1, + double* d2, double* x1, double y1, double* param, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float* x, std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const float* x, std::int64_t incx, float* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const double* x, std::int64_t incx, double* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose transa, transpose transb, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::half alpha, const sycl::half* a, std::int64_t lda, + const sycl::half* b, std::int64_t ldb, sycl::half beta, + sycl::half* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const sycl::half* a, std::int64_t lda, + const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const bfloat16* a, std::int64_t lda, const bfloat16* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, const std::int8_t* a, + int64_t lda, std::int8_t ao, const std::uint8_t* b, + int64_t ldb, std::uint8_t bo, float beta, std::int32_t* c, + int64_t ldc, const std::int32_t* co, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, const std::int8_t* a, + int64_t lda, std::int8_t ao, const std::int8_t* b, int64_t ldb, + std::int8_t bo, float beta, std::int32_t* c, int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, const std::uint8_t* a, + int64_t lda, std::uint8_t ao, const std::int8_t* b, + int64_t ldb, std::int8_t bo, float beta, std::int32_t* c, + int64_t ldc, const std::int32_t* co, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, const std::uint8_t* a, + int64_t lda, std::uint8_t ao, const std::uint8_t* b, + int64_t ldb, std::uint8_t bo, float beta, std::int32_t* c, + int64_t ldc, const std::int32_t* co, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* a, std::int64_t lda, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, float alpha, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* a, std::int64_t lda, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, double alpha, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + float* b, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + double* b, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, + side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stride_a, float* b, + int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, + side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stride_a, double* b, + int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, + side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, + int64_t lda, int64_t stride_a, std::complex* b, + int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, + side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, + int64_t lda, int64_t stride_a, std::complex* b, + int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, + side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, float* alpha, + const float** a, int64_t* lda, float** b, int64_t* ldb, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, + side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, double* alpha, + const double** a, int64_t* lda, double** b, int64_t* ldb, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, + side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex** b, int64_t* ldb, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, + side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex** b, int64_t* ldb, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, const float* x, std::int64_t incx, + float* a, std::int64_t lda, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, const double* x, std::int64_t incx, + double* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + float* b, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + double* b, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, const float* a, + float* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, const double* a, + double* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::complex* x, + std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::complex* x, + std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, const float* a, + std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, const double* a, + std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const float* x, std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, const float** x, std::int64_t* incx, + float** y, std::int64_t* incy, int64_t group_count, + int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, const double** x, std::int64_t* incx, + double** y, std::int64_t* incy, int64_t group_count, + int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, const std::complex** x, + std::int64_t* incx, std::complex** y, + std::int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, const std::complex** x, + std::int64_t* incx, std::complex** y, + std::int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, const float* x, std::int64_t incx, + std::int64_t stridex, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, const double* x, std::int64_t incx, + std::int64_t stridex, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const float* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const double* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const float* x, std::int64_t incx, float* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const double* x, std::int64_t incx, double* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, std::int64_t n, std::int64_t k, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, float alpha, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* a, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + std::int64_t n, double alpha, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* a, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float* x, std::int64_t incx, float* y, std::int64_t incy, + float* param, const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double* x, std::int64_t incx, double* y, std::int64_t incy, + double* param, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* result, const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* result, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + double* result, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sdsdot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float sb, const float* x, std::int64_t incx, const float* y, + std::int64_t incy, float* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, float* a, float* b, + float* c, float* s, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, double* a, + double* b, double* c, double* s, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, + std::complex* a, std::complex* b, float* c, + std::complex* s, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, + std::complex* a, std::complex* b, double* c, + std::complex* s, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + float alpha, float* ab, std::int64_t lda, + std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + double alpha, double* ab, std::int64_t lda, + std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, std::complex* ab, + std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, std::complex* ab, + std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, std::int64_t m, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float beta, + const float* b, std::int64_t ldb, std::int64_t stride_b, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, - side left_right, std::int64_t m, std::int64_t n, - const double *a, std::int64_t lda, std::int64_t stridea, - const double *x, std::int64_t incx, std::int64_t stridex, - double *c, std::int64_t ldc, std::int64_t stridec, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, std::int64_t m, + std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double beta, + const double* b, std::int64_t ldb, std::int64_t stride_b, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, - side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, - side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, - side *left_right, std::int64_t *m, std::int64_t *n, - const float **a, std::int64_t *lda, const float **x, - std::int64_t *incx, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, - side *left_right, std::int64_t *m, std::int64_t *n, - const double **a, std::int64_t *lda, const double **x, - std::int64_t *incx, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, - side *left_right, std::int64_t *m, std::int64_t *n, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, - side *left_right, std::int64_t *m, std::int64_t *n, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event her(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, float alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, double alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, float alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, double alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const float *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const double *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, float alpha, const float *a, - const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, double alpha, const double *a, - const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, float *d1, - float *d2, float *x1, float y1, float *param, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, double *d1, - double *d2, double *x1, double y1, double *param, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, float *x, std::int64_t incx, float *y, - std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, double *x, std::int64_t incx, double *y, - std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - float *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - double *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const float *x, std::int64_t incx, float *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const double *x, std::int64_t incx, - double *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, - float *c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, - float *c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, - double *c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatadd_batch( + oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatadd_batch( + oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, std::int64_t stridea, + float* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, std::int64_t stridea, + double* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, + std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, + std::complex* b, std::int64_t ldb, + std::int64_t strideb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, float alpha, + float* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, double alpha, + double* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, std::complex* ab, + std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, + transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, std::complex* ab, + std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + float beta, const float* b, std::int64_t ldb, float* c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, - std::int64_t ldb, sycl::half beta, sycl::half *c, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + double beta, const double* b, std::int64_t ldb, double* c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, - std::int64_t lda, const bfloat16 *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, - const std::int8_t *a, int64_t lda, std::int8_t ao, - const std::uint8_t *b, int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, int64_t ldc, - const std::int32_t *co, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_bias( - oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, const std::int8_t *a, int64_t lda, - std::int8_t ao, const std::int8_t *b, int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, - int64_t ldc, const std::int32_t *co, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, - const std::uint8_t *a, int64_t lda, std::uint8_t ao, - const std::int8_t *b, int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, int64_t ldc, - const std::int32_t *co, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, - const std::uint8_t *a, int64_t lda, std::uint8_t ao, - const std::uint8_t *b, int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, int64_t ldc, - const std::int32_t *co, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, float alpha, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *a, - std::int64_t lda, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *a, - std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, float alpha, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *a, - std::int64_t lda, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *a, - std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, float *b, - int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, double *b, - int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, - int64_t lda, int64_t stride_a, std::complex *b, - int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, - int64_t lda, int64_t stride_a, std::complex *b, - int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, - side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, float *alpha, - const float **a, int64_t *lda, float **b, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, - side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, double *alpha, - const double **a, int64_t *lda, double **b, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, - side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex **b, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, - side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, - std::complex *alpha, - const std::complex **a, int64_t *lda, - std::complex **b, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, float alpha, const float *a, std::int64_t lda, - const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, std::int64_t m, + const std::complex* a, std::int64_t lda, + std::complex beta, const std::complex* b, + std::int64_t ldb, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, + transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, float alpha, const float *x, - std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, double alpha, const double *x, - std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, - side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const float *x, std::int64_t incx, float *y, - std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const double *x, std::int64_t incx, double *y, - std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, const float **x, std::int64_t *incx, - float **y, std::int64_t *incy, int64_t group_count, - int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, const double **x, std::int64_t *incx, - double **y, std::int64_t *incy, int64_t group_count, - int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, const std::complex **x, - std::int64_t *incx, std::complex **y, - std::int64_t *incy, int64_t group_count, - int64_t *group_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, const std::complex **x, - std::int64_t *incx, std::complex **y, - std::int64_t *incy, int64_t group_count, - int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const float *x, std::int64_t incx, - std::int64_t stridex, float *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const double *x, std::int64_t incx, - std::int64_t stridex, double *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const float *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const double *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - float *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const std::complex *x, std::int64_t incx, - double *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const float *x, std::int64_t incx, float *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const double *x, std::int64_t incx, - double *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, float alpha, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *a, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *a, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, float *x, std::int64_t incx, float *y, - std::int64_t incy, float *param, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, double *x, std::int64_t incx, double *y, - std::int64_t incy, double *param, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *result, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, double *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sdsdot(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, float sb, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, float beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, - uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, double beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, float *a, - float *b, float *c, float *s, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, double *a, - double *b, double *c, double *s, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, - std::complex *a, std::complex *b, float *c, - std::complex *s, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, - std::complex *a, std::complex *b, double *c, - std::complex *s, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - float alpha, float *ab, std::int64_t lda, std::int64_t ldb, - std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - double alpha, double *ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, std::complex *ab, - std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, std::complex *ab, - std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, - transpose transa, transpose transb, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd_batch( - oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd_batch( - oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stridea, float *b, - std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, double *b, - std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, std::int64_t ldb, - std::int64_t strideb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, std::int64_t ldb, - std::int64_t strideb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, - std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, - std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float beta, const float *b, - std::int64_t ldb, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double beta, const double *b, - std::int64_t ldb, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, + std::complex beta, const std::complex* b, + std::int64_t ldb, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, - transpose* trans, std::int64_t* m, std::int64_t* n, - float* alpha, const float** a, std::int64_t* lda, - float** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies = {}); + transpose* trans, std::int64_t* m, std::int64_t* n, + float* alpha, const float** a, std::int64_t* lda, + float** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, - transpose* trans, std::int64_t* m, std::int64_t* n, - double* alpha, const double** a, std::int64_t* lda, - double** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies = {}); + transpose* trans, std::int64_t* m, std::int64_t* n, + double* alpha, const double** a, std::int64_t* lda, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, - transpose* trans, std::int64_t* m, std::int64_t* n, - std::complex* alpha, const std::complex** a, - std::int64_t* lda, std::complex** b, - std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies = {}); + transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, - transpose* trans, std::int64_t* m, std::int64_t* n, - std::complex* alpha, - const std::complex** a, std::int64_t* lda, - std::complex** b, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* groupsize, - const std::vector& dependencies = {}); + transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, - transpose* trans, std::int64_t* m, std::int64_t* n, - float* alpha, float** ab, std::int64_t* lda, - std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies = {}); + transpose* trans, std::int64_t* m, std::int64_t* n, + float* alpha, float** ab, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, - transpose* trans, std::int64_t* m, std::int64_t* n, - double* alpha, double** ab, std::int64_t* lda, - std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies = {}); + transpose* trans, std::int64_t* m, std::int64_t* n, + double* alpha, double** ab, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, - transpose* trans, std::int64_t* m, std::int64_t* n, - std::complex* alpha, std::complex** ab, - std::int64_t* lda, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* groupsize, - const std::vector& dependencies = {}); + transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, std::complex** ab, + std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, - transpose* trans, std::int64_t* m, std::int64_t* n, - std::complex* alpha, std::complex** ab, - std::int64_t* lda, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* groupsize, - const std::vector& dependencies = {}); + transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, std::complex** ab, + std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies = {}); diff --git a/include/oneapi/math/blas/detail/cublas/blas_ct.hxx b/include/oneapi/math/blas/detail/cublas/blas_ct.hxx index c9ac759ab..4a0898f9a 100644 --- a/include/oneapi/math/blas/detail/cublas/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/cublas/blas_ct.hxx @@ -18,1991 +18,1955 @@ **************************************************************************/ void herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, - std::int64_t lda, float beta, sycl::buffer, 1> &c, - std::int64_t ldc) { - oneapi::math::blas::cublas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1>& a, + std::int64_t lda, float beta, sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::cublas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, - sycl::buffer, 1> &a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::cublas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1>& a, + std::int64_t lda, double beta, sycl::buffer, 1>& c, + std::int64_t ldc) { + oneapi::math::blas::cublas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx) { + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx) { + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { oneapi::math::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { oneapi::math::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - double beta, sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, double beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::half beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::half beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::cublas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size) { + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, float c, float s) { oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, double s) { oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { +void rot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, float c, float s) { oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { +void rot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, double c, double s) { oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpby(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { + oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { + oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { - oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void sdsdot(backend_selector selector, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &result) { - oneapi::math::blas::cublas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result) { + oneapi::math::blas::cublas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, + result); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, float beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, double beta, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, double beta, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size) { + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { oneapi::math::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { oneapi::math::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } -void iamin(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } -void iamin(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy) { + sycl::buffer& a, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { + double alpha, sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::cublas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } -void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void swap(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } -void swap(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy) { +void swap(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::half beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &a, - std::int64_t lda) { + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { oneapi::math::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { oneapi::math::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { - oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, float y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, float y1, + sycl::buffer& param) { oneapi::math::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, double y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, double y1, + sycl::buffer& param) { oneapi::math::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } -void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void copy(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } -void copy(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy) { +void copy(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { + oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { + oneapi::math::blas::cublas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc) { - oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::cublas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { oneapi::math::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { oneapi::math::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } -void iamax(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void iamax(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m) { +void rotm(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { oneapi::math::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void rotm(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer ¶m) { +void rotm(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { oneapi::math::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, double beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::cublas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::cublas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::cublas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, float beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } // USM APIs sycl::event syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } -sycl::event scal(backend_selector selector, std::int64_t n, float alpha, - float *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(backend_selector selector, std::int64_t n, float alpha, float* x, + std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, double alpha, - double *x, std::int64_t incx, - const std::vector &dependencies) { + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, float alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, double alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, float* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, double* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, - const float **a, std::int64_t *lda, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, float* alpha, + const float** a, std::int64_t* lda, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, - const double **a, std::int64_t *lda, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, double* alpha, + const double** a, std::int64_t* lda, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); @@ -2010,11 +1974,10 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float beta, float* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2022,11 +1985,10 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2034,12 +1996,11 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2047,12 +2008,11 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2060,679 +2020,665 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { + const float* x, std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { + const double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - float *alpha, const float **x, std::int64_t *incx, float **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, float* alpha, + const float** x, std::int64_t* incx, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - double *alpha, const double **x, std::int64_t *incx, double **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, double* alpha, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const float* x, std::int64_t incx, std::int64_t stridex, float* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const double* x, std::int64_t incx, std::int64_t stridex, double* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { + const float* x, std::int64_t incx, const float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double beta, double *y, - std::int64_t incy, const std::vector &dependencies) { + const double* x, std::int64_t incx, const double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stridea, const float *x, - std::int64_t incx, std::int64_t stridex, float beta, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float beta, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, const double *x, - std::int64_t incx, std::int64_t stridex, double beta, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double beta, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, - float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, const float** x, std::int64_t* incx, float* beta, + float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, - double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, const double** x, std::int64_t* incx, double* beta, + double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const float* a, std::int64_t lda, std::int64_t stridea, + const float* x, std::int64_t incx, std::int64_t stridex, float* c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const double* a, std::int64_t lda, std::int64_t stridea, + const double* x, std::int64_t incx, std::int64_t stridex, double* c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const float** a, std::int64_t* lda, + const float** x, std::int64_t* incx, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const double** a, std::int64_t* lda, + const double** x, std::int64_t* incx, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } sycl::event her(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event her(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } -sycl::event iamin(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamin(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const float **a, std::int64_t *lda, const float **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const float** a, std::int64_t* lda, const float** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - double *alpha, const double **a, std::int64_t *lda, const double **b, - std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + double* alpha, const double** a, std::int64_t* lda, const double** b, + std::int64_t* ldb, double* beta, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, std::complex* beta, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - sycl::half *alpha, const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, sycl::half *beta, - sycl::half **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + sycl::half* alpha, const sycl::half** a, std::int64_t* lda, + const sycl::half** b, std::int64_t* ldb, sycl::half* beta, sycl::half** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const sycl::half** a, std::int64_t* lda, const sycl::half** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, std::int32_t** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); @@ -2740,12 +2686,11 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, - const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, + const float* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2753,12 +2698,11 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, - const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, + const double* b, std::int64_t ldb, std::int64_t stride_b, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2766,13 +2710,12 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2780,13 +2723,12 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2794,12 +2736,12 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, const sycl::half *a, std::int64_t lda, - std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::half alpha, const sycl::half* a, std::int64_t lda, + std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, sycl::half* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2808,10 +2750,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2820,10 +2762,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2832,10 +2774,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2843,185 +2785,180 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *a, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, - const std::vector &dependencies) { + float alpha, const float* a, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *a, const double *x, std::int64_t incx, double beta, - double *y, std::int64_t incy, - const std::vector &dependencies) { + double alpha, const double* a, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - sycl::half beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = - oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, - std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16* a, + std::int64_t lda, const bfloat16* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::cublas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3029,11 +2966,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3041,11 +2978,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3053,11 +2990,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3065,89 +3002,87 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const std::complex *a, - std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const std::complex* a, + std::int64_t lda, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const std::complex *a, - std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const std::complex* a, + std::int64_t lda, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event ger(backend_selector selector, std::int64_t m, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, - incy, a, lda, dependencies); + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + y, incy, a, lda, dependencies); return done; } sycl::event ger(backend_selector selector, std::int64_t m, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, - incy, a, lda, dependencies); + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + y, incy, a, lda, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3155,11 +3090,10 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3167,11 +3101,11 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3179,62 +3113,57 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, - float **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, - double **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); @@ -3242,779 +3171,759 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri } sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - const double *a, std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } -sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, - float y1, float *param, const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, - dependencies); +sycl::event rotmg(backend_selector selector, float* d1, float* d2, float* x1, + float y1, float* param, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + param, dependencies); return done; } -sycl::event rotmg(backend_selector selector, double *d1, double *d2, - double *x1, double y1, double *param, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, - dependencies); +sycl::event rotmg(backend_selector selector, double* d1, double* d2, double* x1, + double y1, double* param, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + param, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -sycl::event copy(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const float **x, std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, const float** x, + std::int64_t* incx, float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const double **x, std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t n, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t n, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t stridex, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event asum(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event asum(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } -sycl::event iamax(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamax(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event rotm(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotm(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotg(backend_selector selector, float *a, float *b, float *c, - float *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, float* a, float* b, float* c, float* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, double *a, double *b, double *c, - double *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, double* a, double* b, double* c, + double* s, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, float *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, float* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, double *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, double* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event sdsdot(backend_selector selector, std::int64_t n, float sb, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *result, const std::vector &dependencies) { + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, - incy, result, dependencies); + incy, result, dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, double beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + float alpha, const float* a, std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4022,10 +3931,10 @@ sycl::event omatcopy_batch(backend_selector selector, transpose } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4034,9 +3943,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4045,9 +3954,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4055,18 +3964,18 @@ sycl::event omatcopy_batch(backend_selector selector, transpose } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, + std::int64_t m, std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, + std::int64_t m, std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4074,9 +3983,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4084,9 +3993,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4094,10 +4003,10 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, + const float* a, std::int64_t lda, std::int64_t stride_a, float beta, + const float* b, std::int64_t ldb, std::int64_t stride_b, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4106,10 +4015,10 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + const double* b, std::int64_t ldb, std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4118,11 +4027,11 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4131,11 +4040,11 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4143,150 +4052,150 @@ sycl::event omatadd_batch(backend_selector selector, transpose } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, - std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, float alpha, const float* a, std::int64_t lda, float* b, + std::int64_t ldb, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, - std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, double alpha, const double* a, std::int64_t lda, double* b, + std::int64_t ldb, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, double* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + float beta, const float* b, std::int64_t ldb, float* c, std::int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, - double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + double beta, const double* b, std::int64_t ldb, double* c, std::int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } @@ -4295,9 +4204,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4306,9 +4215,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4316,11 +4225,10 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + std::int64_t* groupsize, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4328,33 +4236,30 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { - auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + std::int64_t* groupsize, const std::vector& dependencies) { + auto done = oneapi::math::blas::cublas::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose* trans, std::int64_t* m, std::int64_t* n, float* alpha, float** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose* trans, std::int64_t* m, std::int64_t* n, double* alpha, double** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } @@ -4364,8 +4269,8 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } @@ -4375,7 +4280,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::cublas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } diff --git a/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hxx b/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hxx index 1141eb238..caa75a646 100644 --- a/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hxx +++ b/include/oneapi/math/blas/detail/cublas/onemath_blas_cublas.hxx @@ -19,2314 +19,2211 @@ // Buffer APIs -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void axpy(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); +void axpy(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); -void axpy(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); +void axpy(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); -void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); +void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); -void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); +void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); -void axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer &y, +void axpy_batch(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size); - -void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, +void axpy_batch(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); +void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size); -void axpby(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); +void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size); -void axpby(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); +void axpby(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy); -void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); +void axpby(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); -void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); +void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); +void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); -void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); -void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); -void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void copy_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &result); +void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& result); -void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &result); +void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& result); -void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &result); +void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& result); -void dotc(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); +void dotc(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); -void dotc(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); +void dotc(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); -void dotu(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); +void dotu(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); -void dotu(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); +void dotu(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, float c, float s); -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - double c, double s); +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, double c, + double s); -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, float c, float s); +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, float c, float s); -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, double c, double s); +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, double c, double s); -void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s); +void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s); -void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s); +void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s); -void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s); +void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); -void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s); +void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); -void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m); +void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& param); -void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m); +void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& param); -void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, float y1, sycl::buffer ¶m); +void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, float y1, sycl::buffer& param); -void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, double y1, sycl::buffer ¶m); +void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, double y1, sycl::buffer& param); -void scal(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, +void scal(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx); -void scal(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, +void scal(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, std::int64_t incx); -void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx); - -void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx); - -void scal(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx); - -void scal(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx); +void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx); -void sdsdot(sycl::queue &queue, std::int64_t n, float sb, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); +void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx); -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); +void scal(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer, 1>& x, + std::int64_t incx); -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); +void scal(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx); -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); +void sdsdot(sycl::queue& queue, std::int64_t n, float sb, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy); -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); - -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); - -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); -void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); + +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); + +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy); + +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); + +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, float beta, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, double beta, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); +void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, double beta, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size); -void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, +void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size); - -void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size); - -void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, +void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size); - -void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size); - -void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda); - -void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda); - -void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -void hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -void hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -void hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -void hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); +void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size); -void her(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda); +void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size); -void her(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda); +void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size); -void her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); +void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size); -void her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); +void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, sycl::buffer& a, + std::int64_t lda); -void hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, +void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda); + +void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +void hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +void hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +void hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +void her(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda); + +void her(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda); + +void her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); -void hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, +void hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); -void hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a); +void hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a); -void hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a); +void hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a); -void hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a); +void hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a); -void hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a); +void hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a); -void sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); +void sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy); -void sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); +void sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); -void spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); +void spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); -void spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy); +void spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); -void spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a); +void spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a); -void spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a); +void spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a); -void spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a); +void spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a); -void spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a); +void spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a); -void symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); +void symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy); -void symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); +void symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); -void syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, - std::int64_t lda); +void syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda); -void syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, +void syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda); -void syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda); +void syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda); -void syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda); +void syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx); +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx); +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, std::int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, + std::int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx); +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx); +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, std::int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, + std::int64_t incx); -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::half beta, - sycl::buffer &c, std::int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - -void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); -void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - float beta, sycl::buffer, 1> &c, std::int64_t ldc); - -void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - double beta, sycl::buffer, 1> &c, std::int64_t ldc); - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc); -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, float beta, - sycl::buffer &c, std::int64_t ldc); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, double beta, - sycl::buffer &c, std::int64_t ldc); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::half alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::half beta, + sycl::buffer& c, std::int64_t ldc); -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::complex beta, sycl::buffer, 1> &c, +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, float beta, sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc); -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, +void hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer, 1>& a, std::int64_t lda, float beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer, 1>& a, std::int64_t lda, double beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, double beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); + +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc); + +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, float beta, + sycl::buffer& c, std::int64_t ldc); + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, double beta, + sycl::buffer& c, std::int64_t ldc); + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); + +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, double beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc); +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size); -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb); - -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb); - -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb); - -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb); - -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, double beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size); +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); + +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc); + +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb); + +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb); + +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); + +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::half beta, sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::half beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); + +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc); + +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc); -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); + +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); + +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co); -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co); -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co); -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co); -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size); +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size); -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size); +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size); -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size); -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - float beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size); +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + float beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size); +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb); -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb); -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb); -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb); -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb); +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb); -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb); +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb); -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &b, int64_t ldb, std::int64_t strideb); +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& b, int64_t ldb, std::int64_t strideb); -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, int64_t ldb, +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, int64_t ldb, std::int64_t strideb); -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb); -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb); -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb); -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, float beta, - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, sycl::buffer& a, int64_t lda, float beta, + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, double beta, - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, sycl::buffer& a, int64_t lda, double beta, + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &b, int64_t ldb, - sycl::buffer, 1> &c, int64_t ldc); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& b, int64_t ldb, + sycl::buffer, 1>& c, int64_t ldc); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &b, int64_t ldb, - sycl::buffer, 1> &c, int64_t ldc); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& b, int64_t ldb, + sycl::buffer, 1>& c, int64_t ldc); // USM APIs -sycl::event asum(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, float *result, - const std::vector &dependencies = {}); - -sycl::event asum(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, double *result, - const std::vector &dependencies = {}); - -sycl::event asum(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - float *result, const std::vector &dependencies = {}); - -sycl::event asum(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - double *result, const std::vector &dependencies = {}); - -sycl::event axpy(sycl::queue &queue, std::int64_t n, float alpha, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event axpy(sycl::queue &queue, std::int64_t n, double alpha, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, float *alpha, const float **x, - std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, double *alpha, const double **x, - std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, const float *x, - std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, const double *x, - std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event axpby(sycl::queue &queue, std::int64_t n, float alpha, const float *x, - std::int64_t incx, const float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event axpby(sycl::queue &queue, std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event copy(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event copy(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event copy(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event copy(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const float **x, - std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const double **x, - std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *result, - const std::vector &dependencies = {}); - -sycl::event dot(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *result, - const std::vector &dependencies = {}); - -sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, double *result, - const std::vector &dependencies = {}); - -sycl::event dotc(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); - -sycl::event dotc(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); - -sycl::event dotu(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); - -sycl::event dotu(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); +sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}); + +sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}); + +sycl::event asum(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + float* result, const std::vector& dependencies = {}); + +sycl::event asum(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + double* result, const std::vector& dependencies = {}); + +sycl::event axpy(sycl::queue& queue, std::int64_t n, float alpha, const float* x, std::int64_t incx, + float* y, std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event axpy(sycl::queue& queue, std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, float* alpha, const float** x, + std::int64_t* incx, float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, double* alpha, const double** x, + std::int64_t* incx, double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, std::complex* alpha, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, std::complex* alpha, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, float alpha, const float* x, + std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, double alpha, const double* x, + std::int64_t incx, std::int64_t stridex, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event axpby(sycl::queue& queue, std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event axpby(sycl::queue& queue, std::int64_t n, double alpha, const double* x, + std::int64_t incx, const double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event copy(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, float* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event copy(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, double* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const float** x, std::int64_t* incx, + float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const double** x, std::int64_t* incx, + double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies = {}); +sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* result, + const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); +sycl::event dot(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* result, + const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); +sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, double* result, + const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, float *result, - const std::vector &dependencies = {}); +sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, double *result, - const std::vector &dependencies = {}); +sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - float *result, const std::vector &dependencies = {}); +sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - double *result, const std::vector &dependencies = {}); +sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, float c, float s, - const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, double c, - double s, const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, - std::int64_t incy, float c, float s, - const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, - std::int64_t incy, double c, double s, - const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, float *a, float *b, float *c, float *s, - const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, double *a, double *b, double *c, double *s, - const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, - float *c, std::complex *s, - const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, - double *c, std::complex *s, - const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); -sycl::event rotm(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, - std::int64_t incy, float *param, - const std::vector &dependencies = {}); +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}); -sycl::event rotm(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, - double *y, std::int64_t incy, double *param, - const std::vector &dependencies = {}); - -sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, float y1, - float *param, const std::vector &dependencies = {}); - -sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, double y1, - double *param, const std::vector &dependencies = {}); - -sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, float *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, double *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event sdsdot(sycl::queue &queue, std::int64_t n, float sb, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *result, - const std::vector &dependencies = {}); - -sycl::event swap(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event swap(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, - double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, std::int64_t stridea, - const float *x, std::int64_t incx, std::int64_t stridex, float beta, - float *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, std::int64_t stridea, - const double *x, std::int64_t incx, std::int64_t stridex, double beta, - double *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose *trans, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float *beta, float **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose *trans, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double *beta, double **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose *trans, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose *trans, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - const float *a, std::int64_t lda, std::int64_t stridea, const float *x, - std::int64_t incx, std::int64_t stridex, float *c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - const double *a, std::int64_t lda, std::int64_t stridea, const double *x, - std::int64_t incx, std::int64_t stridex, double *c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, std::int64_t *m, - std::int64_t *n, const float **a, std::int64_t *lda, const float **x, - std::int64_t *incx, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, std::int64_t *m, - std::int64_t *n, const double **a, std::int64_t *lda, const double **x, - std::int64_t *incx, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, std::int64_t *m, - std::int64_t *n, const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, std::int64_t *m, - std::int64_t *n, const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double *y, std::int64_t incy, - double *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event her(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event her(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - const std::vector &dependencies = {}); - -sycl::event hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - const std::vector &dependencies = {}); - -sycl::event hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies = {}); - -sycl::event hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies = {}); - -sycl::event sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - const float *a, const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - const double *a, const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *a, - const std::vector &dependencies = {}); - -sycl::event spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *a, - const std::vector &dependencies = {}); - -sycl::event spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, - const std::vector &dependencies = {}); - -sycl::event spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double *y, std::int64_t incy, - double *a, const std::vector &dependencies = {}); - -sycl::event symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - const float *a, std::int64_t lda, const float *x, std::int64_t incx, - float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - const double *a, std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double *y, std::int64_t incy, - double *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, - sycl::half *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, - std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, - float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, - double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, - float *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, - double *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, - std::int64_t *n, std::int64_t *k, float *alpha, const float **a, - std::int64_t *lda, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, - std::int64_t *n, std::int64_t *k, double *alpha, const double **a, - std::int64_t *lda, double *beta, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, - std::int64_t *n, std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, - std::int64_t *n, std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float beta, float *c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double beta, double *c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, std::complex *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, std::complex *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies = {}); - -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies = {}); - -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies = {}); - -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, - float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, - double *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, - transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, - float *alpha, const float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, - transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, - double *alpha, const double **a, std::int64_t *lda, double **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, - transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, - transpose *trans, diag *unit_diag, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, - const float **a, std::int64_t *lda, const float **b, std::int64_t *ldb, - float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, double *alpha, - const double **a, std::int64_t *lda, const double **b, std::int64_t *ldb, - double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, - const sycl::half **a, std::int64_t *lda, const sycl::half **b, - std::int64_t *ldb, sycl::half *beta, sycl::half **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, - std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, float *beta, - float **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, - std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, - float **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, - std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, - std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, const float *b, - std::int64_t ldb, std::int64_t stride_b, float beta, float *c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, - const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, - sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, std::int64_t stride_a, const sycl::half *b, - std::int64_t ldb, std::int64_t stride_b, float beta, float *c, +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}); + +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + float* result, const std::vector& dependencies = {}); + +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + double* result, const std::vector& dependencies = {}); + +sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, float c, float s, + const std::vector& dependencies = {}); + +sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, double c, double s, + const std::vector& dependencies = {}); + +sycl::event rot(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, + std::int64_t incy, float c, float s, + const std::vector& dependencies = {}); + +sycl::event rot(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, + std::int64_t incy, double c, double s, + const std::vector& dependencies = {}); + +sycl::event rotg(sycl::queue& queue, float* a, float* b, float* c, float* s, + const std::vector& dependencies = {}); + +sycl::event rotg(sycl::queue& queue, double* a, double* b, double* c, double* s, + const std::vector& dependencies = {}); + +sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, float* c, + std::complex* s, const std::vector& dependencies = {}); + +sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, double* c, + std::complex* s, const std::vector& dependencies = {}); + +sycl::event rotm(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, + std::int64_t incy, float* param, + const std::vector& dependencies = {}); + +sycl::event rotm(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, + std::int64_t incy, double* param, + const std::vector& dependencies = {}); + +sycl::event rotmg(sycl::queue& queue, float* d1, float* d2, float* x1, float y1, float* param, + const std::vector& dependencies = {}); + +sycl::event rotmg(sycl::queue& queue, double* d1, double* d2, double* x1, double y1, double* param, + const std::vector& dependencies = {}); + +sycl::event scal(sycl::queue& queue, std::int64_t n, float alpha, float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event scal(sycl::queue& queue, std::int64_t n, double alpha, double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event scal(sycl::queue& queue, std::int64_t n, float alpha, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event scal(sycl::queue& queue, std::int64_t n, double alpha, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event sdsdot(sycl::queue& queue, std::int64_t n, float sb, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* result, + const std::vector& dependencies = {}); + +sycl::event swap(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event swap(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, float alpha, const float* a, std::int64_t lda, + const float* x, std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, double alpha, const double* a, std::int64_t lda, + const double* x, std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, std::int64_t stridea, + const float* x, std::int64_t incx, std::int64_t stridex, float beta, + float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, std::int64_t stridea, + const double* x, std::int64_t incx, std::int64_t stridex, double beta, + double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex beta, std::complex* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex beta, std::complex* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, + float* alpha, const float** a, std::int64_t* lda, const float** x, + std::int64_t* incx, float* beta, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, + double* alpha, const double** a, std::int64_t* lda, const double** x, + std::int64_t* incx, double* beta, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + const float* a, std::int64_t lda, std::int64_t stridea, const float* x, + std::int64_t incx, std::int64_t stridex, float* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + const double* a, std::int64_t lda, std::int64_t stridea, const double* x, + std::int64_t incx, std::int64_t stridex, double* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, std::int64_t* m, std::int64_t* n, + const float** a, std::int64_t* lda, const float** x, std::int64_t* incx, + float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, std::int64_t* m, std::int64_t* n, + const double** a, std::int64_t* lda, const double** x, std::int64_t* incx, + double** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event her(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + std::int64_t lda, const std::vector& dependencies = {}); + +sycl::event her(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + std::int64_t lda, const std::vector& dependencies = {}); + +sycl::event her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + const std::complex* a, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + const std::complex* a, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + const std::vector& dependencies = {}); + +sycl::event hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + const std::vector& dependencies = {}); + +sycl::event hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, + const std::vector& dependencies = {}); + +sycl::event hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, + const std::vector& dependencies = {}); + +sycl::event sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, const float* a, + const float* x, std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const double* a, const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, const float* x, + std::int64_t incx, float* a, const std::vector& dependencies = {}); + +sycl::event spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* a, const std::vector& dependencies = {}); + +sycl::event spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* a, + const std::vector& dependencies = {}); + +sycl::event spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const double* x, std::int64_t incx, const double* y, std::int64_t incy, double* a, + const std::vector& dependencies = {}); + +sycl::event symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const double* a, std::int64_t lda, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, const float* x, + std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + const double* x, std::int64_t incx, const double* y, std::int64_t incy, double* a, + std::int64_t lda, const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, sycl::half beta, + sycl::half* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, std::int64_t lda, + const sycl::half* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const bfloat16* a, std::int64_t lda, + const bfloat16* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, const std::complex* a, std::int64_t lda, + float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, const std::complex* a, std::int64_t lda, + double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, float beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, double beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, const float* a, std::int64_t lda, float beta, + float* c, std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, const double* a, std::int64_t lda, double beta, + double* c, std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, std::int64_t* n, + std::int64_t* k, float* alpha, const float** a, std::int64_t* lda, + float* beta, float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, std::int64_t* n, + std::int64_t* k, double* alpha, const double** a, std::int64_t* lda, + double* beta, double** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, std::int64_t* n, + std::int64_t* k, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, std::int64_t* n, + std::int64_t* k, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float beta, float* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double beta, double* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies = {}); + +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies = {}); + +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies = {}); + +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, std::int64_t stride_a, double* b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, std::int64_t* m, std::int64_t* n, float* alpha, + const float** a, std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, std::int64_t* m, std::int64_t* n, double* alpha, + const double** a, std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, const float** a, + std::int64_t* lda, const float** b, std::int64_t* ldb, float* beta, + float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, double* alpha, const double** a, + std::int64_t* lda, const double** b, std::int64_t* ldb, double* beta, + double** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, std::complex* beta, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, sycl::half* alpha, const sycl::half** a, + std::int64_t* lda, const sycl::half** b, std::int64_t* ldb, sycl::half* beta, + sycl::half** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, const sycl::half** a, + std::int64_t* lda, const sycl::half** b, std::int64_t* ldb, float* beta, + float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, const std::int8_t** a, + std::int64_t* lda, const std::int8_t** b, std::int64_t* ldb, float* beta, + float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, const std::int8_t** a, + std::int64_t* lda, const std::int8_t** b, std::int64_t* ldb, float* beta, + std::int32_t** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, const float* b, std::int64_t ldb, + std::int64_t stride_b, float beta, float* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, const double* b, std::int64_t ldb, + std::int64_t stride_b, double beta, double* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half* a, + std::int64_t lda, std::int64_t stride_a, const sycl::half* b, + std::int64_t ldb, std::int64_t stride_b, sycl::half beta, sycl::half* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, std::int64_t stride_a, const sycl::half* b, + std::int64_t ldb, std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, - std::int64_t ldb, std::int64_t stride_b, float beta, float *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const std::int8_t* a, + std::int64_t lda, std::int64_t stride_a, const std::int8_t* b, + std::int64_t ldb, std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, - std::int64_t ldb, std::int64_t stride_b, float beta, std::int32_t *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const std::int8_t* a, + std::int64_t lda, std::int64_t stride_a, const std::int8_t* b, + std::int64_t ldb, std::int64_t stride_b, float beta, std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, - const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, - std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies = {}); - -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int8_t ao, - const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, - std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies = {}); - -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, - const std::int8_t *b, std::int64_t ldb, std::int8_t bo, float beta, - std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies = {}); - -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, - const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, float beta, - std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const std::int8_t* a, std::int64_t lda, std::int8_t ao, const std::uint8_t* b, + std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t* c, + std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies = {}); + +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const std::int8_t* a, std::int64_t lda, std::int8_t ao, const std::int8_t* b, + std::int64_t ldb, std::int8_t bo, float beta, std::int32_t* c, + std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies = {}); + +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, + const std::int8_t* b, std::int64_t ldb, std::int8_t bo, float beta, + std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies = {}); + +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, + const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, float beta, + std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies = {}); + +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies = {}); +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies = {}); +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, - float beta, const float *b, int64_t ldb, int64_t stride_b, float *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, float alpha, const float* a, int64_t lda, int64_t stride_a, + float beta, const float* b, int64_t ldb, int64_t stride_b, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, double alpha, const double *a, int64_t lda, int64_t stride_a, - double beta, const double *b, int64_t ldb, int64_t stride_b, double *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, double alpha, const double* a, int64_t lda, int64_t stride_a, + double beta, const double* b, int64_t ldb, int64_t stride_b, double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, float *b, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, double *b, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, std::int64_t stridea, float *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies = {}); - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, std::int64_t stridea, double *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies = {}); - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::int64_t stridea, std::complex *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies = {}); - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::int64_t stridea, std::complex *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies = {}); - -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, const float *a, int64_t lda, float beta, const float *b, - int64_t ldb, float *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, const double *a, int64_t lda, double beta, const double *b, - int64_t ldb, double *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, const std::complex *b, int64_t ldb, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, const std::complex *b, int64_t ldb, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, float* b, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, double* b, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex* b, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex* b, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, std::int64_t stridea, float* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies = {}); + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, std::int64_t stridea, double* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies = {}); + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::int64_t stridea, std::complex* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies = {}); + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::int64_t stridea, std::complex* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies = {}); + +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, const float* a, int64_t lda, float beta, const float* b, + int64_t ldb, float* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, const double* a, int64_t lda, double beta, const double* b, + int64_t ldb, double* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, const std::complex* b, int64_t ldb, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, const std::complex* b, int64_t ldb, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, float* alpha, const float** a, int64_t* lda, float** b, int64_t* ldb, diff --git a/include/oneapi/math/blas/detail/mklcpu/blas_ct.hxx b/include/oneapi/math/blas/detail/mklcpu/blas_ct.hxx index 0d34723db..dbc1554fe 100644 --- a/include/oneapi/math/blas/detail/mklcpu/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/mklcpu/blas_ct.hxx @@ -20,1991 +20,1955 @@ // Buffer APIs void herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, - std::int64_t lda, float beta, sycl::buffer, 1> &c, - std::int64_t ldc) { - oneapi::math::blas::mklcpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1>& a, + std::int64_t lda, float beta, sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::mklcpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, - sycl::buffer, 1> &a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::mklcpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1>& a, + std::int64_t lda, double beta, sycl::buffer, 1>& c, + std::int64_t ldc) { + oneapi::math::blas::mklcpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx) { + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx) { + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { oneapi::math::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { oneapi::math::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - double beta, sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, double beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::half beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::half beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::mklcpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size) { + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, float c, float s) { oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, double s) { oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { +void rot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, float c, float s) { oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { +void rot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, double c, double s) { oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpby(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { + oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { + oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { - oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void sdsdot(backend_selector selector, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &result) { - oneapi::math::blas::mklcpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result) { + oneapi::math::blas::mklcpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, + result); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, float beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, double beta, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, double beta, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size) { + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { oneapi::math::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { oneapi::math::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } -void iamin(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } -void iamin(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy) { + sycl::buffer& a, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { + double alpha, sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::mklcpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } -void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void swap(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } -void swap(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy) { +void swap(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::half beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &a, - std::int64_t lda) { + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { oneapi::math::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { oneapi::math::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { - oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, float y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, float y1, + sycl::buffer& param) { oneapi::math::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, double y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, double y1, + sycl::buffer& param) { oneapi::math::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } -void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void copy(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } -void copy(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy) { +void copy(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { + oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { + oneapi::math::blas::mklcpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc) { - oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklcpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { oneapi::math::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { oneapi::math::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } -void iamax(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void iamax(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m) { +void rotm(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { oneapi::math::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void rotm(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer ¶m) { +void rotm(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { oneapi::math::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, double beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklcpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklcpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::mklcpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, float beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } // USM APIs sycl::event syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } -sycl::event scal(backend_selector selector, std::int64_t n, float alpha, - float *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(backend_selector selector, std::int64_t n, float alpha, float* x, + std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, double alpha, - double *x, std::int64_t incx, - const std::vector &dependencies) { + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, float alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, double alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, float* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, double* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, - const float **a, std::int64_t *lda, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, float* alpha, + const float** a, std::int64_t* lda, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, - const double **a, std::int64_t *lda, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, double* alpha, + const double** a, std::int64_t* lda, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); @@ -2012,11 +1976,10 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float beta, float* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2024,11 +1987,10 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2036,12 +1998,11 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2049,12 +2010,11 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2062,679 +2022,665 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { + const float* x, std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { + const double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - float *alpha, const float **x, std::int64_t *incx, float **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, float* alpha, + const float** x, std::int64_t* incx, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - double *alpha, const double **x, std::int64_t *incx, double **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, double* alpha, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const float* x, std::int64_t incx, std::int64_t stridex, float* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const double* x, std::int64_t incx, std::int64_t stridex, double* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { + const float* x, std::int64_t incx, const float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double beta, double *y, - std::int64_t incy, const std::vector &dependencies) { + const double* x, std::int64_t incx, const double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stridea, const float *x, - std::int64_t incx, std::int64_t stridex, float beta, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float beta, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, const double *x, - std::int64_t incx, std::int64_t stridex, double beta, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double beta, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, - float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, const float** x, std::int64_t* incx, float* beta, + float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, - double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, const double** x, std::int64_t* incx, double* beta, + double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const float* a, std::int64_t lda, std::int64_t stridea, + const float* x, std::int64_t incx, std::int64_t stridex, float* c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const double* a, std::int64_t lda, std::int64_t stridea, + const double* x, std::int64_t incx, std::int64_t stridex, double* c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const float** a, std::int64_t* lda, + const float** x, std::int64_t* incx, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const double** a, std::int64_t* lda, + const double** x, std::int64_t* incx, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } sycl::event her(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event her(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } -sycl::event iamin(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamin(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const float **a, std::int64_t *lda, const float **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const float** a, std::int64_t* lda, const float** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - double *alpha, const double **a, std::int64_t *lda, const double **b, - std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + double* alpha, const double** a, std::int64_t* lda, const double** b, + std::int64_t* ldb, double* beta, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, std::complex* beta, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - sycl::half *alpha, const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, sycl::half *beta, - sycl::half **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + sycl::half* alpha, const sycl::half** a, std::int64_t* lda, + const sycl::half** b, std::int64_t* ldb, sycl::half* beta, sycl::half** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const sycl::half** a, std::int64_t* lda, const sycl::half** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, std::int32_t** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); @@ -2742,12 +2688,11 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, - const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, + const float* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2755,12 +2700,11 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, - const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, + const double* b, std::int64_t ldb, std::int64_t stride_b, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2768,13 +2712,12 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2782,13 +2725,12 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2796,12 +2738,12 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, const sycl::half *a, std::int64_t lda, - std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::half alpha, const sycl::half* a, std::int64_t lda, + std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, sycl::half* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2810,10 +2752,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2822,10 +2764,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2834,10 +2776,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2845,185 +2787,180 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *a, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, - const std::vector &dependencies) { + float alpha, const float* a, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *a, const double *x, std::int64_t incx, double beta, - double *y, std::int64_t incy, - const std::vector &dependencies) { + double alpha, const double* a, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - sycl::half beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = - oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, - std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16* a, + std::int64_t lda, const bfloat16* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklcpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3031,11 +2968,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3043,11 +2980,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3055,11 +2992,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3067,89 +3004,87 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const std::complex *a, - std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const std::complex* a, + std::int64_t lda, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const std::complex *a, - std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const std::complex* a, + std::int64_t lda, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event ger(backend_selector selector, std::int64_t m, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, - incy, a, lda, dependencies); + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + y, incy, a, lda, dependencies); return done; } sycl::event ger(backend_selector selector, std::int64_t m, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, - incy, a, lda, dependencies); + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + y, incy, a, lda, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3157,11 +3092,10 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3169,11 +3103,11 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3181,62 +3115,57 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, - float **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, - double **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); @@ -3244,779 +3173,759 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri } sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - const double *a, std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } -sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, - float y1, float *param, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, - dependencies); +sycl::event rotmg(backend_selector selector, float* d1, float* d2, float* x1, + float y1, float* param, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + param, dependencies); return done; } -sycl::event rotmg(backend_selector selector, double *d1, double *d2, - double *x1, double y1, double *param, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, - dependencies); +sycl::event rotmg(backend_selector selector, double* d1, double* d2, double* x1, + double y1, double* param, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + param, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -sycl::event copy(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const float **x, std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, const float** x, + std::int64_t* incx, float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const double **x, std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t n, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t n, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t stridex, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event asum(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event asum(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } -sycl::event iamax(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamax(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event rotm(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotm(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotg(backend_selector selector, float *a, float *b, float *c, - float *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, float* a, float* b, float* c, float* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, double *a, double *b, double *c, - double *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, double* a, double* b, double* c, + double* s, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, float *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, float* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, double *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, double* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event sdsdot(backend_selector selector, std::int64_t n, float sb, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *result, const std::vector &dependencies) { + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, - incy, result, dependencies); + incy, result, dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, double beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + float alpha, const float* a, std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4024,10 +3933,10 @@ sycl::event omatcopy_batch(backend_selector selector, transpose } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4036,9 +3945,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4047,9 +3956,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4057,18 +3966,18 @@ sycl::event omatcopy_batch(backend_selector selector, transpose } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, + std::int64_t m, std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, + std::int64_t m, std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4076,9 +3985,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4086,9 +3995,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4096,10 +4005,10 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, + const float* a, std::int64_t lda, std::int64_t stride_a, float beta, + const float* b, std::int64_t ldb, std::int64_t stride_b, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4108,10 +4017,10 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + const double* b, std::int64_t ldb, std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4120,11 +4029,11 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4133,11 +4042,11 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4145,150 +4054,150 @@ sycl::event omatadd_batch(backend_selector selector, transpose } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, - std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, float alpha, const float* a, std::int64_t lda, float* b, + std::int64_t ldb, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, - std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, double alpha, const double* a, std::int64_t lda, double* b, + std::int64_t ldb, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, double* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + float beta, const float* b, std::int64_t ldb, float* c, std::int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, - double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + double beta, const double* b, std::int64_t ldb, double* c, std::int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } @@ -4297,9 +4206,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4308,9 +4217,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4318,11 +4227,10 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + std::int64_t* groupsize, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4330,33 +4238,30 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { - auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + std::int64_t* groupsize, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklcpu::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose* trans, std::int64_t* m, std::int64_t* n, float* alpha, float** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose* trans, std::int64_t* m, std::int64_t* n, double* alpha, double** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } @@ -4366,8 +4271,8 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } @@ -4377,7 +4282,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::mklcpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } diff --git a/include/oneapi/math/blas/detail/mklgpu/blas_ct.hxx b/include/oneapi/math/blas/detail/mklgpu/blas_ct.hxx index b00317329..7ab1f887c 100644 --- a/include/oneapi/math/blas/detail/mklgpu/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/mklgpu/blas_ct.hxx @@ -20,1991 +20,1955 @@ // Buffer APIs void herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, - std::int64_t lda, float beta, sycl::buffer, 1> &c, - std::int64_t ldc) { - oneapi::math::blas::mklgpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1>& a, + std::int64_t lda, float beta, sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::mklgpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, - sycl::buffer, 1> &a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::mklgpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1>& a, + std::int64_t lda, double beta, sycl::buffer, 1>& c, + std::int64_t ldc) { + oneapi::math::blas::mklgpu::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx) { + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx) { + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { oneapi::math::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { oneapi::math::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - double beta, sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, double beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::half beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::half beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::mklgpu::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size) { + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, float c, float s) { oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, double s) { oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { +void rot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, float c, float s) { oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { +void rot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, double c, double s) { oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpby(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { + oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { + oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { - oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void sdsdot(backend_selector selector, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &result) { - oneapi::math::blas::mklgpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result) { + oneapi::math::blas::mklgpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, + result); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, float beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, double beta, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, double beta, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size) { + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { oneapi::math::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { oneapi::math::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } -void iamin(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } -void iamin(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy) { + sycl::buffer& a, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { + double alpha, sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::mklgpu::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } -void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void swap(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } -void swap(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy) { +void swap(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::half beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &a, - std::int64_t lda) { + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { oneapi::math::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { oneapi::math::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { - oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, float y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, float y1, + sycl::buffer& param) { oneapi::math::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, double y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, double y1, + sycl::buffer& param) { oneapi::math::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } -void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void copy(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } -void copy(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy) { +void copy(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { + oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { + oneapi::math::blas::mklgpu::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc) { - oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::mklgpu::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { oneapi::math::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { oneapi::math::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } -void iamax(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void iamax(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m) { +void rotm(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { oneapi::math::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void rotm(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer ¶m) { +void rotm(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { oneapi::math::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, double beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::mklgpu::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::mklgpu::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::mklgpu::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, float beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } // USM APIs sycl::event syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } -sycl::event scal(backend_selector selector, std::int64_t n, float alpha, - float *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(backend_selector selector, std::int64_t n, float alpha, float* x, + std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, double alpha, - double *x, std::int64_t incx, - const std::vector &dependencies) { + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, float alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, double alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, float* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, double* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, - const float **a, std::int64_t *lda, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, float* alpha, + const float** a, std::int64_t* lda, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, - const double **a, std::int64_t *lda, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, double* alpha, + const double** a, std::int64_t* lda, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); @@ -2012,11 +1976,10 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float beta, float* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2024,11 +1987,10 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2036,12 +1998,11 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2049,12 +2010,11 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2062,679 +2022,665 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { + const float* x, std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { + const double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - float *alpha, const float **x, std::int64_t *incx, float **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, float* alpha, + const float** x, std::int64_t* incx, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - double *alpha, const double **x, std::int64_t *incx, double **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, double* alpha, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const float* x, std::int64_t incx, std::int64_t stridex, float* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const double* x, std::int64_t incx, std::int64_t stridex, double* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { + const float* x, std::int64_t incx, const float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double beta, double *y, - std::int64_t incy, const std::vector &dependencies) { + const double* x, std::int64_t incx, const double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stridea, const float *x, - std::int64_t incx, std::int64_t stridex, float beta, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float beta, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, const double *x, - std::int64_t incx, std::int64_t stridex, double beta, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double beta, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, - float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, const float** x, std::int64_t* incx, float* beta, + float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, - double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, const double** x, std::int64_t* incx, double* beta, + double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const float* a, std::int64_t lda, std::int64_t stridea, + const float* x, std::int64_t incx, std::int64_t stridex, float* c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const double* a, std::int64_t lda, std::int64_t stridea, + const double* x, std::int64_t incx, std::int64_t stridex, double* c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const float** a, std::int64_t* lda, + const float** x, std::int64_t* incx, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const double** a, std::int64_t* lda, + const double** x, std::int64_t* incx, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } sycl::event her(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event her(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } -sycl::event iamin(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamin(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - sycl::half *alpha, const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, sycl::half *beta, - sycl::half **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + sycl::half* alpha, const sycl::half** a, std::int64_t* lda, + const sycl::half** b, std::int64_t* ldb, sycl::half* beta, sycl::half** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const sycl::half** a, std::int64_t* lda, const sycl::half** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, std::int32_t** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const float **a, std::int64_t *lda, const float **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const float** a, std::int64_t* lda, const float** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - double *alpha, const double **a, std::int64_t *lda, const double **b, - std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + double* alpha, const double** a, std::int64_t* lda, const double** b, + std::int64_t* ldb, double* beta, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, std::complex* beta, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); @@ -2742,12 +2688,12 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, const sycl::half *a, std::int64_t lda, - std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::half alpha, const sycl::half* a, std::int64_t lda, + std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, sycl::half* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2756,10 +2702,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2768,10 +2714,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2780,10 +2726,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2791,12 +2737,11 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, - const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, + const float* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2804,12 +2749,11 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, - const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, + const double* b, std::int64_t ldb, std::int64_t stride_b, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2817,13 +2761,12 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2831,13 +2774,12 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2845,185 +2787,180 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *a, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, - const std::vector &dependencies) { + float alpha, const float* a, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *a, const double *x, std::int64_t incx, double beta, - double *y, std::int64_t incy, - const std::vector &dependencies) { + double alpha, const double* a, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - sycl::half beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = - oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, - std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16* a, + std::int64_t lda, const bfloat16* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::mklgpu::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3031,11 +2968,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3043,11 +2980,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3055,11 +2992,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3067,89 +3004,87 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const std::complex *a, - std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const std::complex* a, + std::int64_t lda, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const std::complex *a, - std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const std::complex* a, + std::int64_t lda, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event ger(backend_selector selector, std::int64_t m, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, - incy, a, lda, dependencies); + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + y, incy, a, lda, dependencies); return done; } sycl::event ger(backend_selector selector, std::int64_t m, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, - incy, a, lda, dependencies); + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + y, incy, a, lda, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3157,11 +3092,10 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3169,11 +3103,11 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3181,62 +3115,57 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, - float **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, - double **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); @@ -3244,779 +3173,759 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri } sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - const double *a, std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } -sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, - float y1, float *param, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, - dependencies); +sycl::event rotmg(backend_selector selector, float* d1, float* d2, float* x1, + float y1, float* param, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + param, dependencies); return done; } -sycl::event rotmg(backend_selector selector, double *d1, double *d2, - double *x1, double y1, double *param, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, - dependencies); +sycl::event rotmg(backend_selector selector, double* d1, double* d2, double* x1, + double y1, double* param, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + param, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -sycl::event copy(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const float **x, std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, const float** x, + std::int64_t* incx, float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const double **x, std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t n, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t n, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t stridex, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event asum(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event asum(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } -sycl::event iamax(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamax(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event rotm(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotm(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotg(backend_selector selector, float *a, float *b, float *c, - float *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, float* a, float* b, float* c, float* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, double *a, double *b, double *c, - double *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, double* a, double* b, double* c, + double* s, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, float *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, float* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, double *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, double* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event sdsdot(backend_selector selector, std::int64_t n, float sb, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *result, const std::vector &dependencies) { + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, - incy, result, dependencies); + incy, result, dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, double beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + float alpha, const float* a, std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4024,10 +3933,10 @@ sycl::event omatcopy_batch(backend_selector selector, transpose } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4036,9 +3945,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4047,9 +3956,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4057,18 +3966,18 @@ sycl::event omatcopy_batch(backend_selector selector, transpose } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, + std::int64_t m, std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, + std::int64_t m, std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4076,9 +3985,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4086,9 +3995,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4096,10 +4005,10 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, + const float* a, std::int64_t lda, std::int64_t stride_a, float beta, + const float* b, std::int64_t ldb, std::int64_t stride_b, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4108,10 +4017,10 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + const double* b, std::int64_t ldb, std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4120,11 +4029,11 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4133,11 +4042,11 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4145,150 +4054,150 @@ sycl::event omatadd_batch(backend_selector selector, transpose } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, - std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, float alpha, const float* a, std::int64_t lda, float* b, + std::int64_t ldb, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, - std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, double alpha, const double* a, std::int64_t lda, double* b, + std::int64_t ldb, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, double* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + float beta, const float* b, std::int64_t ldb, float* c, std::int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, - double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + double beta, const double* b, std::int64_t ldb, double* c, std::int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } @@ -4297,9 +4206,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4308,9 +4217,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4318,11 +4227,10 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + std::int64_t* groupsize, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4330,33 +4238,30 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { - auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + std::int64_t* groupsize, const std::vector& dependencies) { + auto done = oneapi::math::blas::mklgpu::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose* trans, std::int64_t* m, std::int64_t* n, float* alpha, float** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose* trans, std::int64_t* m, std::int64_t* n, double* alpha, double** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } @@ -4366,8 +4271,8 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } @@ -4377,7 +4282,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::mklgpu::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } diff --git a/include/oneapi/math/blas/detail/netlib/blas_ct.hxx b/include/oneapi/math/blas/detail/netlib/blas_ct.hxx index 1e55eaa94..c1ff9b629 100644 --- a/include/oneapi/math/blas/detail/netlib/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/netlib/blas_ct.hxx @@ -20,1991 +20,1955 @@ // Buffer APIs void herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, - std::int64_t lda, float beta, sycl::buffer, 1> &c, - std::int64_t ldc) { - oneapi::math::blas::netlib::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1>& a, + std::int64_t lda, float beta, sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::netlib::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, - sycl::buffer, 1> &a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::netlib::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1>& a, + std::int64_t lda, double beta, sycl::buffer, 1>& c, + std::int64_t ldc) { + oneapi::math::blas::netlib::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx) { + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx) { + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { oneapi::math::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { oneapi::math::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - double beta, sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, double beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::half beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::half beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, - ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, a, - lda, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::netlib::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, + a, lda, beta, c, ldc); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size) { + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, float c, float s) { oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, double s) { oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { +void rot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, float c, float s) { oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { +void rot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, double c, double s) { oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpby(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { + oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { + oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { - oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, incy); + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, + incy); } void sdsdot(backend_selector selector, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &result) { - oneapi::math::blas::netlib::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, result); + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result) { + oneapi::math::blas::netlib::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, + result); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, float beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, double beta, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, double beta, sycl::buffer& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size) { + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, stridey, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size) { + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { oneapi::math::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { oneapi::math::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); } -void iamin(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } -void iamin(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy) { + sycl::buffer& a, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { + double alpha, sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + beta, y, incy); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co) { - oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, n, - k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { + oneapi::math::blas::netlib::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } -void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void swap(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } -void swap(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy) { +void swap(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::half beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &a, - std::int64_t lda) { + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { oneapi::math::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { oneapi::math::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { - oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, lda, - x, incx, beta, y, incy); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, + lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, float y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, float y1, + sycl::buffer& param) { oneapi::math::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, double y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, double y1, + sycl::buffer& param) { oneapi::math::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } -void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void copy(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } -void copy(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy) { +void copy(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { + oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, incy, - stridey, batch_size); + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size) { + oneapi::math::blas::netlib::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, + incy, stridey, batch_size); } void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc) { - oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, k, - alpha, a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, + k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::netlib::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { oneapi::math::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &a) { + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { oneapi::math::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + incy, a); } -void iamax(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void iamax(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m) { +void rotm(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { oneapi::math::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void rotm(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer ¶m) { +void rotm(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { oneapi::math::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, b, - ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, double beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::netlib::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::netlib::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, stride_b, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + ldb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::netlib::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, float beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + lda, beta, b, ldb, c, ldc); } // USM APIs sycl::event syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } -sycl::event scal(backend_selector selector, std::int64_t n, float alpha, - float *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(backend_selector selector, std::int64_t n, float alpha, float* x, + std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, double alpha, - double *x, std::int64_t incx, - const std::vector &dependencies) { + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, float alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, double alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::trmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::trmv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, float* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, double* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, - const float **a, std::int64_t *lda, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, float* alpha, + const float** a, std::int64_t* lda, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, - const double **a, std::int64_t *lda, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, double* alpha, + const double** a, std::int64_t* lda, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); @@ -2012,11 +1976,10 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_l } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float beta, float* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2024,11 +1987,10 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2036,12 +1998,11 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2049,12 +2010,11 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2062,679 +2022,665 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_lo } sycl::event her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + x, incx, y, incy, a, lda, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); + s, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { + const float* x, std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { + const double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - float *alpha, const float **x, std::int64_t *incx, float **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, float* alpha, + const float** x, std::int64_t* incx, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - double *alpha, const double **x, std::int64_t *incx, double **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, double* alpha, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const float* x, std::int64_t incx, std::int64_t stridex, float* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const double* x, std::int64_t incx, std::int64_t stridex, double* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { + const float* x, std::int64_t incx, const float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double beta, double *y, - std::int64_t incy, const std::vector &dependencies) { + const double* x, std::int64_t incx, const double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stridea, const float *x, - std::int64_t incx, std::int64_t stridex, float beta, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float beta, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, const double *x, - std::int64_t incx, std::int64_t stridex, double beta, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double beta, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, - float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, const float** x, std::int64_t* incx, float* beta, + float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, - double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, const double** x, std::int64_t* incx, double* beta, + double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, x, incx, beta, y, incy, - group_count, group_size, dependencies); +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::gemv_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, + group_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const float* a, std::int64_t lda, std::int64_t stridea, + const float* x, std::int64_t incx, std::int64_t stridex, float* c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const double* a, std::int64_t lda, std::int64_t stridea, + const double* x, std::int64_t incx, std::int64_t stridex, double* c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const float** a, std::int64_t* lda, + const float** x, std::int64_t* incx, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const double** a, std::int64_t* lda, + const double** x, std::int64_t* incx, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, - a, lda, x, incx, c, ldc, group_count, - group_size, dependencies); +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::dgmm_batch( + selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, + dependencies); return done; } sycl::event her(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event her(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } -sycl::event iamin(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamin(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const float **a, std::int64_t *lda, const float **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const float** a, std::int64_t* lda, const float** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - double *alpha, const double **a, std::int64_t *lda, const double **b, - std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + double* alpha, const double** a, std::int64_t* lda, const double** b, + std::int64_t* ldb, double* beta, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, std::complex* beta, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - sycl::half *alpha, const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, sycl::half *beta, - sycl::half **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + sycl::half* alpha, const sycl::half** a, std::int64_t* lda, + const sycl::half** b, std::int64_t* ldb, sycl::half* beta, sycl::half** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const sycl::half** a, std::int64_t* lda, const sycl::half** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, std::int32_t** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); @@ -2742,12 +2688,11 @@ sycl::event gemm_batch(backend_selector selector, transpose *tr } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, - const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, + const float* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2755,12 +2700,11 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, - const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, + const double* b, std::int64_t ldb, std::int64_t stride_b, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2768,13 +2712,12 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2782,13 +2725,12 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2796,12 +2738,12 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event gemm_batch(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, const sycl::half *a, std::int64_t lda, - std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::half alpha, const sycl::half* a, std::int64_t lda, + std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, sycl::half* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2810,10 +2752,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2822,10 +2764,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2834,10 +2776,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tra sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2845,185 +2787,180 @@ sycl::event gemm_batch(backend_selector selector, transpose tra } sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *a, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, - const std::vector &dependencies) { + float alpha, const float* a, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *a, const double *x, std::int64_t incx, double beta, - double *y, std::int64_t incy, - const std::vector &dependencies) { + double alpha, const double* a, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + a, x, incx, beta, y, incy, dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - sycl::half beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = - oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, - std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16* a, + std::int64_t lda, const bfloat16* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = - oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc, dependencies); + oneapi::math::blas::netlib::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3031,11 +2968,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3043,11 +2980,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3055,11 +2992,11 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event gemm_bias(backend_selector selector, transpose transa, - transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + transpose transb, offset offsetc, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3067,89 +3004,87 @@ sycl::event gemm_bias(backend_selector selector, transpose tran } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const std::complex *a, - std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const std::complex* a, + std::int64_t lda, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const std::complex *a, - std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const std::complex* a, + std::int64_t lda, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event ger(backend_selector selector, std::int64_t m, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, - incy, a, lda, dependencies); + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + y, incy, a, lda, dependencies); return done; } sycl::event ger(backend_selector selector, std::int64_t m, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, - incy, a, lda, dependencies); + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, + y, incy, a, lda, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3157,11 +3092,10 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3169,11 +3103,11 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3181,62 +3115,57 @@ sycl::event trsm_batch(backend_selector selector, side left_rig } sycl::event trsm_batch(backend_selector selector, side left_right, - uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies) { + uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, - float **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, - double **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); @@ -3244,779 +3173,759 @@ sycl::event trsm_batch(backend_selector selector, side *left_ri } sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - const double *a, std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } -sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, - float y1, float *param, const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, - dependencies); +sycl::event rotmg(backend_selector selector, float* d1, float* d2, float* x1, + float y1, float* param, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + param, dependencies); return done; } -sycl::event rotmg(backend_selector selector, double *d1, double *d2, - double *x1, double y1, double *param, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param, - dependencies); +sycl::event rotmg(backend_selector selector, double* d1, double* d2, double* x1, + double y1, double* param, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, + param, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::trsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, lda, x, incx, dependencies); + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::trsv( + selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -sycl::event copy(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const float **x, std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, const float** x, + std::int64_t* incx, float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const double **x, std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t n, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t n, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t stridex, double* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } -sycl::event gemmt(backend_selector selector, uplo upper_lower, - transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event asum(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event asum(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, - std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, - const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + x, incx, y, incy, a, dependencies); return done; } -sycl::event iamax(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamax(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event rotm(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotm(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotg(backend_selector selector, float *a, float *b, float *c, - float *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, float* a, float* b, float* c, float* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, double *a, double *b, double *c, - double *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, double* a, double* b, double* c, + double* s, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, float *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, float* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, double *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, double* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event sdsdot(backend_selector selector, std::int64_t n, float sb, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *result, const std::vector &dependencies) { + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, - incy, result, dependencies); + incy, result, dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, double beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + float alpha, const float* a, std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4024,10 +3933,10 @@ sycl::event omatcopy_batch(backend_selector selector, transpose } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4036,10 +3945,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4048,10 +3956,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -4059,19 +3966,18 @@ sycl::event omatcopy_batch(backend_selector selector, transpose } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, + std::int64_t m, std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, + std::int64_t m, std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, - const std::vector &dependencies) { + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4079,9 +3985,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4089,9 +3995,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4099,10 +4005,10 @@ sycl::event imatcopy_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, + const float* a, std::int64_t lda, std::int64_t stride_a, float beta, + const float* b, std::int64_t ldb, std::int64_t stride_b, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4111,10 +4017,10 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + const double* b, std::int64_t ldb, std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4123,12 +4029,11 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4137,12 +4042,11 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4150,150 +4054,150 @@ sycl::event omatadd_batch(backend_selector selector, transpose } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, - std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, float alpha, const float* a, std::int64_t lda, float* b, + std::int64_t ldb, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, - std::int64_t ldb, const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, double alpha, const double* a, std::int64_t lda, double* b, + std::int64_t ldb, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, - a, lda, b, ldb, dependencies); + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy(selector.get_queue(), trans, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, double* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, - ab, lda, ldb, dependencies); + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::imatcopy(selector.get_queue(), trans, m, n, + alpha, ab, lda, ldb, dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + float beta, const float* b, std::int64_t ldb, float* c, std::int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, - double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + double beta, const double* b, std::int64_t ldb, double* c, std::int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc, dependencies); + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::omatadd(selector.get_queue(), transa, transb, m, + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } @@ -4302,9 +4206,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, float** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4313,9 +4217,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* lda, double** b, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4323,11 +4227,10 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + std::int64_t* groupsize, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } @@ -4335,33 +4238,30 @@ sycl::event omatcopy_batch(backend_selector selector, transpose std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { - auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, group_count, - groupsize, dependencies); + std::int64_t* groupsize, const std::vector& dependencies) { + auto done = oneapi::math::blas::netlib::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, + dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose* trans, std::int64_t* m, std::int64_t* n, float* alpha, float** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose* trans, std::int64_t* m, std::int64_t* n, double* alpha, double** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } @@ -4371,8 +4271,8 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } @@ -4382,7 +4282,7 @@ sycl::event imatcopy_batch(backend_selector selector, transpose std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::netlib::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + alpha, ab, lda, ldb, group_count, + groupsize, dependencies); return done; } diff --git a/include/oneapi/math/blas/detail/onemath_blas_backends.hxx b/include/oneapi/math/blas/detail/onemath_blas_backends.hxx index e9fcbbf66..06ea19b75 100644 --- a/include/oneapi/math/blas/detail/onemath_blas_backends.hxx +++ b/include/oneapi/math/blas/detail/onemath_blas_backends.hxx @@ -19,2928 +19,2880 @@ // Buffer APIs -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - float beta, sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - double beta, sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc); - -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc); - -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::half alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::half beta, sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - float beta, sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - float beta, sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void symm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void hemm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void hemm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, float beta, - sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, double beta, - sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, float beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void herk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer, 1> &a, std::int64_t lda, float beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void herk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer, 1> &a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); +ONEMATH_EXPORT void gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc); -ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); +ONEMATH_EXPORT void gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc); -ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, +ONEMATH_EXPORT void gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::half alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::half beta, sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + float beta, sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + float beta, sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void symm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void symm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void symm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& c, std::int64_t ldc); -ONEMATH_EXPORT void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, +ONEMATH_EXPORT void symm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void hemm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void hemm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, float beta, + sycl::buffer& c, std::int64_t ldc); -ONEMATH_EXPORT void her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, +ONEMATH_EXPORT void syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, float beta, sycl::buffer, 1> &c, - std::int64_t ldc); + double alpha, sycl::buffer& a, std::int64_t lda, double beta, + sycl::buffer& c, std::int64_t ldc); -ONEMATH_EXPORT void her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, +ONEMATH_EXPORT void syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - -ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - -ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb); - -ONEMATH_EXPORT void trmm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb); - -ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - -ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - -ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb); - -ONEMATH_EXPORT void trsm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, - std::int64_t ldb); - -ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, float beta, - sycl::buffer &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, double beta, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size); + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); -ONEMATH_EXPORT void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size); - -ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size); - -ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); - -ONEMATH_EXPORT void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &c, - std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); - -ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -ONEMATH_EXPORT void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - -ONEMATH_EXPORT void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - -ONEMATH_EXPORT void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, - std::int64_t lda); - -ONEMATH_EXPORT void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, - std::int64_t lda); - -ONEMATH_EXPORT void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, - std::int64_t lda); - -ONEMATH_EXPORT void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, - std::int64_t lda); - -ONEMATH_EXPORT void hbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void hbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -ONEMATH_EXPORT void hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -ONEMATH_EXPORT void her(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda); - -ONEMATH_EXPORT void her(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a, - std::int64_t lda); - -ONEMATH_EXPORT void her2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, - std::int64_t lda); - -ONEMATH_EXPORT void her2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, - std::int64_t lda); - -ONEMATH_EXPORT void hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, - std::int64_t incy); - -ONEMATH_EXPORT void hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a); - -ONEMATH_EXPORT void hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a); - -ONEMATH_EXPORT void hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a); - -ONEMATH_EXPORT void hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a); - -ONEMATH_EXPORT void sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void symv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void symv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void syr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &a, std::int64_t lda); - -ONEMATH_EXPORT void syr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &a, std::int64_t lda); - -ONEMATH_EXPORT void syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - -ONEMATH_EXPORT void syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - -ONEMATH_EXPORT void spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &a, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, - std::int64_t incy); - -ONEMATH_EXPORT void spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void spr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &a); - -ONEMATH_EXPORT void spr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &a); - -ONEMATH_EXPORT void spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a); - -ONEMATH_EXPORT void spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a); - -ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - -ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - -ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx); - -ONEMATH_EXPORT void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx); - -ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - -ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - -ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx); - -ONEMATH_EXPORT void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, - std::int64_t incx); - -ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); - -ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); - -ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); - -ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, - std::int64_t incx); - -ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - -ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - -ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - -ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - -ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void dotc(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - -ONEMATH_EXPORT void dotc(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - -ONEMATH_EXPORT void dotu(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - -ONEMATH_EXPORT void dotu(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - -ONEMATH_EXPORT void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - -ONEMATH_EXPORT void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - -ONEMATH_EXPORT void iamax(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - -ONEMATH_EXPORT void iamax(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - -ONEMATH_EXPORT void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - -ONEMATH_EXPORT void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - -ONEMATH_EXPORT void iamin(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - -ONEMATH_EXPORT void iamin(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - -ONEMATH_EXPORT void asum(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - -ONEMATH_EXPORT void asum(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - -ONEMATH_EXPORT void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - -ONEMATH_EXPORT void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - -ONEMATH_EXPORT void axpy(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void axpy(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - -ONEMATH_EXPORT void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - -ONEMATH_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -ONEMATH_EXPORT void axpby(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void axpby(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, +ONEMATH_EXPORT void syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, float beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, double beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void herk(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer, 1>& a, std::int64_t lda, float beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void herk(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer, 1>& a, std::int64_t lda, + double beta, sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void her2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, float beta, sycl::buffer, 1>& c, + std::int64_t ldc); + +ONEMATH_EXPORT void her2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb, double beta, sycl::buffer, 1>& c, + std::int64_t ldc); + +ONEMATH_EXPORT void trmm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb); + +ONEMATH_EXPORT void trmm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb); + +ONEMATH_EXPORT void trmm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb); + +ONEMATH_EXPORT void trmm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb); + +ONEMATH_EXPORT void trsm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb); + +ONEMATH_EXPORT void trsm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb); + +ONEMATH_EXPORT void trsm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb); + +ONEMATH_EXPORT void trsm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, + std::int64_t ldb); + +ONEMATH_EXPORT void gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); + +ONEMATH_EXPORT void gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + +ONEMATH_EXPORT void gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); -ONEMATH_EXPORT void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, +ONEMATH_EXPORT void gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); -ONEMATH_EXPORT void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); +ONEMATH_EXPORT void gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, float beta, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size); + +ONEMATH_EXPORT void gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, double beta, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size); + +ONEMATH_EXPORT void gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); + +ONEMATH_EXPORT void gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size); + +ONEMATH_EXPORT void dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size); + +ONEMATH_EXPORT void dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& c, std::int64_t ldc, + std::int64_t stridec, std::int64_t batch_size); + +ONEMATH_EXPORT void dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, + std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); + +ONEMATH_EXPORT void dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, + std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); + +ONEMATH_EXPORT void gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy); -ONEMATH_EXPORT void copy(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); +ONEMATH_EXPORT void gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy); -ONEMATH_EXPORT void copy(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); +ONEMATH_EXPORT void gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +ONEMATH_EXPORT void gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +ONEMATH_EXPORT void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda); + +ONEMATH_EXPORT void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda); + +ONEMATH_EXPORT void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, + std::int64_t lda); + +ONEMATH_EXPORT void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, + std::int64_t lda); + +ONEMATH_EXPORT void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, + std::int64_t lda); + +ONEMATH_EXPORT void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, + std::int64_t lda); + +ONEMATH_EXPORT void hbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); -ONEMATH_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size); +ONEMATH_EXPORT void hbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); -ONEMATH_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); +ONEMATH_EXPORT void hemv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +ONEMATH_EXPORT void hemv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +ONEMATH_EXPORT void her(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda); + +ONEMATH_EXPORT void her(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda); + +ONEMATH_EXPORT void her2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, + std::int64_t lda); + +ONEMATH_EXPORT void her2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, + std::int64_t lda); + +ONEMATH_EXPORT void hpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); -ONEMATH_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); +ONEMATH_EXPORT void hpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); -ONEMATH_EXPORT void copy_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); +ONEMATH_EXPORT void hpr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a); + +ONEMATH_EXPORT void hpr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a); + +ONEMATH_EXPORT void hpr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a); + +ONEMATH_EXPORT void hpr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a); + +ONEMATH_EXPORT void sbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); + +ONEMATH_EXPORT void sbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + +ONEMATH_EXPORT void symv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); + +ONEMATH_EXPORT void symv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + +ONEMATH_EXPORT void syr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a, std::int64_t lda); + +ONEMATH_EXPORT void syr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a, std::int64_t lda); + +ONEMATH_EXPORT void syr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a, + std::int64_t lda); + +ONEMATH_EXPORT void syr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a, + std::int64_t lda); + +ONEMATH_EXPORT void spmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx, float beta, sycl::buffer& y, + std::int64_t incy); -ONEMATH_EXPORT void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); +ONEMATH_EXPORT void spmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, + std::int64_t incy); -ONEMATH_EXPORT void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); +ONEMATH_EXPORT void spr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a); + +ONEMATH_EXPORT void spr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a); + +ONEMATH_EXPORT void spr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a); + +ONEMATH_EXPORT void spr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& a); + +ONEMATH_EXPORT void tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); + +ONEMATH_EXPORT void tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); + +ONEMATH_EXPORT void tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx); + +ONEMATH_EXPORT void tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx); + +ONEMATH_EXPORT void tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); + +ONEMATH_EXPORT void tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); + +ONEMATH_EXPORT void tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx); + +ONEMATH_EXPORT void tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, + std::int64_t incx); + +ONEMATH_EXPORT void tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx); + +ONEMATH_EXPORT void tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx); + +ONEMATH_EXPORT void tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx); + +ONEMATH_EXPORT void tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx); + +ONEMATH_EXPORT void tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + +ONEMATH_EXPORT void trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + +ONEMATH_EXPORT void trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + +ONEMATH_EXPORT void trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + +ONEMATH_EXPORT void trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void dotc(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + +ONEMATH_EXPORT void dotc(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + +ONEMATH_EXPORT void dotu(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + +ONEMATH_EXPORT void dotu(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + +ONEMATH_EXPORT void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + +ONEMATH_EXPORT void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + +ONEMATH_EXPORT void iamax(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void iamax(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + +ONEMATH_EXPORT void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + +ONEMATH_EXPORT void iamin(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void iamin(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void asum(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void asum(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + +ONEMATH_EXPORT void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + +ONEMATH_EXPORT void axpy(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); -ONEMATH_EXPORT void sdsdot(sycl::queue &queue, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); +ONEMATH_EXPORT void axpy(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy); -ONEMATH_EXPORT void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); +ONEMATH_EXPORT void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + +ONEMATH_EXPORT void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); -ONEMATH_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); +ONEMATH_EXPORT void axpy_batch(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size); -ONEMATH_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); +ONEMATH_EXPORT void axpy_batch(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size); -ONEMATH_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); +ONEMATH_EXPORT void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); + +ONEMATH_EXPORT void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); +ONEMATH_EXPORT void axpby(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); -ONEMATH_EXPORT void rot(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, - float s); - -ONEMATH_EXPORT void rot(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, double c, - double s); - -ONEMATH_EXPORT void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, - float s); - -ONEMATH_EXPORT void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - double c, double s); - -ONEMATH_EXPORT void rotg(sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s); - -ONEMATH_EXPORT void rotg(sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s); - -ONEMATH_EXPORT void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s); - -ONEMATH_EXPORT void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, - sycl::buffer &c, - sycl::buffer, 1> &s); - -ONEMATH_EXPORT void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m); - -ONEMATH_EXPORT void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m); - -ONEMATH_EXPORT void rotmg(sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, float y1, - sycl::buffer ¶m); - -ONEMATH_EXPORT void rotmg(sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, - double y1, sycl::buffer ¶m); - -ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx); - -ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx); - -ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void scal(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx); - -ONEMATH_EXPORT void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); - -ONEMATH_EXPORT void swap(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - -ONEMATH_EXPORT void swap(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, float beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, double beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::half alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, float beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, float beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, float beta, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - -ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - -ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - -ONEMATH_EXPORT void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - -ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - float beta, sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - -ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc); +ONEMATH_EXPORT void axpby(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); -ONEMATH_EXPORT void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, - std::int64_t ldc); +ONEMATH_EXPORT void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); -ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, oneapi::math::offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, - float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); - -ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, oneapi::math::offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, - float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); - -ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, oneapi::math::offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, - float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); - -ONEMATH_EXPORT void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, oneapi::math::offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, - float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); - -ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size); - -ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size); - -ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - -ONEMATH_EXPORT void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - -ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, - std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size); - -ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, - std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size); - -ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - -ONEMATH_EXPORT void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - -ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, float beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - -ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - -ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - -ONEMATH_EXPORT void omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - -ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer &b, - std::int64_t ldb, std::int64_t strideb); - -ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer &b, - std::int64_t ldb, std::int64_t strideb); - -ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t strideb); +ONEMATH_EXPORT void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, + std::int64_t incy); -ONEMATH_EXPORT void omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t strideb); - -ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, - std::int64_t lda, std::int64_t ldb); - -ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, - std::int64_t lda, std::int64_t ldb); - -ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, - std::int64_t ldb); - -ONEMATH_EXPORT void imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, - std::int64_t ldb); - -ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, float beta, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, - std::int64_t ldc); - -ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, double beta, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, - std::int64_t ldc); - -ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); - -ONEMATH_EXPORT void omatadd(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); +ONEMATH_EXPORT void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); -// USM APIs +ONEMATH_EXPORT void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void copy(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void copy(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::half alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, - sycl::half beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const bfloat16 *a, std::int64_t lda, - const bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, oneapi::math::offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::int8_t *a, std::int64_t lda, std::int8_t ao, - const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, oneapi::math::offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::int8_t *a, std::int64_t lda, std::int8_t ao, - const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, oneapi::math::offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, - const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, oneapi::math::offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, - const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hemm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, float beta, - float *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, double beta, - double *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void copy_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *k, float *alpha, const float **a, - std::int64_t *lda, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *k, double *alpha, const double **a, - std::int64_t *lda, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void copy_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - std::complex *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event herk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - float alpha, const std::complex *a, std::int64_t lda, - float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event herk(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - double alpha, const std::complex *a, std::int64_t lda, - double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, float beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, - std::int64_t ldb, double beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, float *b, - std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, double *b, - std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, float *b, - std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, double *b, - std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, - float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, - oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, - double *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm_batch( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm_batch( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, - oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, - oneapi::math::diag *unit_diag, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, - oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, - oneapi::math::diag *unit_diag, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm_batch( - sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsm_batch( - sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, +ONEMATH_EXPORT void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + +ONEMATH_EXPORT void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + +ONEMATH_EXPORT void sdsdot(sycl::queue& queue, std::int64_t n, float sb, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + +ONEMATH_EXPORT void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + +ONEMATH_EXPORT void nrm2(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void nrm2(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + +ONEMATH_EXPORT void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + +ONEMATH_EXPORT void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + +ONEMATH_EXPORT void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, float c, float s); + +ONEMATH_EXPORT void rot(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, + double s); + +ONEMATH_EXPORT void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, float c, + float s); + +ONEMATH_EXPORT void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, double c, + double s); + +ONEMATH_EXPORT void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s); + +ONEMATH_EXPORT void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s); + +ONEMATH_EXPORT void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); + +ONEMATH_EXPORT void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); + +ONEMATH_EXPORT void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param); + +ONEMATH_EXPORT void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param); + +ONEMATH_EXPORT void rotmg(sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, float y1, + sycl::buffer& param); + +ONEMATH_EXPORT void rotmg(sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, double y1, + sycl::buffer& param); + +ONEMATH_EXPORT void scal(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx); + +ONEMATH_EXPORT void scal(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx); + +ONEMATH_EXPORT void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void scal(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void scal(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx); + +ONEMATH_EXPORT void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); + +ONEMATH_EXPORT void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); + +ONEMATH_EXPORT void swap(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + +ONEMATH_EXPORT void swap(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + +ONEMATH_EXPORT void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size); + +ONEMATH_EXPORT void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, double beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size); + +ONEMATH_EXPORT void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::half alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, + sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, float beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, float beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, float beta, sycl::buffer& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void trsm_batch(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size); + +ONEMATH_EXPORT void trsm_batch(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size); + +ONEMATH_EXPORT void trsm_batch(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size); + +ONEMATH_EXPORT void trsm_batch(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size); + +ONEMATH_EXPORT void gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + double beta, sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t n, std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc); + +ONEMATH_EXPORT void gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t n, std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, + std::int64_t ldc); + +ONEMATH_EXPORT void gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, int8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); + +ONEMATH_EXPORT void gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, int8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); + +ONEMATH_EXPORT void gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); + +ONEMATH_EXPORT void gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); + +ONEMATH_EXPORT void omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); + +ONEMATH_EXPORT void omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); + +ONEMATH_EXPORT void omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); -ONEMATH_EXPORT sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, +ONEMATH_EXPORT void omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stridea, - const float *x, std::int64_t incx, std::int64_t stridex, - float beta, float *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stridea, - const double *x, std::int64_t incx, std::int64_t stridex, - double beta, double *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); + +ONEMATH_EXPORT void imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, + std::int64_t stride, std::int64_t batch_size); + +ONEMATH_EXPORT void imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, + std::int64_t stride, std::int64_t batch_size); + +ONEMATH_EXPORT void imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, + std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT sycl::event gemv_batch( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, std::complex beta, - std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, + std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -ONEMATH_EXPORT sycl::event gemv_batch( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, - const float **a, std::int64_t *lda, const float **x, - std::int64_t *incx, float *beta, float **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, - const double **a, std::int64_t *lda, const double **x, - std::int64_t *incx, double *beta, double **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **x, - std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, float beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT sycl::event gemv_batch( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, - std::int64_t m, std::int64_t n, const float *a, - std::int64_t lda, std::int64_t stridea, const float *x, - std::int64_t incx, std::int64_t stridex, float *c, - std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, double beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, - std::int64_t m, std::int64_t n, const double *a, - std::int64_t lda, std::int64_t stridea, const double *x, - std::int64_t incx, std::int64_t stridex, double *c, - std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, - std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, - std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, - std::int64_t *m, std::int64_t *n, const float **a, - std::int64_t *lda, const float **x, std::int64_t *incx, - float **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, - std::int64_t *m, std::int64_t *n, const double **a, - std::int64_t *lda, const double **x, std::int64_t *incx, - double **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, - std::int64_t *m, std::int64_t *n, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, - std::int64_t *m, std::int64_t *n, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event her(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const std::complex *x, - std::int64_t incx, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event her(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const std::complex *x, - std::int64_t incx, std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event her2(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event her2(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *a, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *a, const std::complex *x, - std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const std::complex *x, - std::int64_t incx, std::complex *a, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const std::complex *x, - std::int64_t incx, std::complex *a, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event symv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event symv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syr(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, std::int64_t incx, - float *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syr(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, std::int64_t incx, - double *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *a, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *a, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spr(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, std::int64_t incx, - float *a, const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spr(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, std::int64_t incx, - double *a, const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *a, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *a, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, const double *a, - std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dotc(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dotc(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dotu(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dotu(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event iamax(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event iamin(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event asum(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, float *alpha, - const float **x, std::int64_t *incx, float **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, double *alpha, - const double **x, std::int64_t *incx, double **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, - std::complex *alpha, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, std::int64_t stridex, - float *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, + std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); + +ONEMATH_EXPORT void omatcopy(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb); + +ONEMATH_EXPORT void omatcopy(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb); + +ONEMATH_EXPORT void omatcopy(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); -ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, std::int64_t stridex, - double *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float beta, float *y, - std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double beta, - double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event axpby(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const float **x, - std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const double **x, - std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, - const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t stridex, float *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void omatcopy(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); -ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void omatcopy2(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer& b, + std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void omatcopy2(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer& b, + std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT sycl::event copy_batch(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, - std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - float *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dot(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, - double *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sdsdot(sycl::queue &queue, std::int64_t n, float sb, - const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - double *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event nrm2(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - float c, float s, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - double c, double s, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float c, float s, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rot(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double c, - double s, const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rotg(sycl::queue &queue, float *a, float *b, float *c, float *s, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rotg(sycl::queue &queue, double *a, double *b, double *c, - double *s, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rotg(sycl::queue &queue, std::complex *a, - std::complex *b, float *c, std::complex *s, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rotg(sycl::queue &queue, std::complex *a, - std::complex *b, double *c, std::complex *s, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rotm(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float *param, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rotm(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double *param, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, - float y1, float *param, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, - double y1, double *param, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, float *x, - std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, double *x, - std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, - std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, - const float **a, std::int64_t *lda, const float **b, - std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, double *alpha, - const double **a, std::int64_t *lda, const double **b, - std::int64_t *ldb, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, - std::int64_t *ldb, std::complex *beta, - std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void omatcopy2(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT sycl::event gemm_batch( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, const std::complex **b, - std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, sycl::half *alpha, - const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, sycl::half *beta, - sycl::half **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, - const sycl::half **a, std::int64_t *lda, const sycl::half **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, - const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, - float **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float *alpha, - const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, - std::int32_t **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, - const float *b, std::int64_t ldb, std::int64_t stride_b, - float beta, float *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, - const double *b, std::int64_t ldb, std::int64_t stride_b, - double beta, double *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void omatcopy2(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t strideb); -ONEMATH_EXPORT sycl::event gemm_batch( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void imatcopy(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& ab, + std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT sycl::event gemm_batch( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}); +ONEMATH_EXPORT void imatcopy(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& ab, + std::int64_t lda, std::int64_t ldb); -ONEMATH_EXPORT sycl::event gemm_batch( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, - std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, std::int64_t stride_a, const sycl::half *b, - std::int64_t ldb, std::int64_t stride_b, float beta, float *c, - std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, - std::int64_t ldb, std::int64_t stride_b, float beta, float *c, - std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, - std::int64_t ldb, std::int64_t stride_b, float beta, - std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, - float *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, - std::int64_t ldc, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, - float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void imatcopy(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, + std::int64_t ldb); -ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, - double *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void imatcopy(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, + std::int64_t ldb); + +ONEMATH_EXPORT void omatadd(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, float beta, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& c, + std::int64_t ldc); + +ONEMATH_EXPORT void omatadd(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, double beta, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc); + +ONEMATH_EXPORT void omatadd(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); + +ONEMATH_EXPORT void omatadd(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); -ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}); +// USM APIs -ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::half alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, + sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const sycl::half* a, std::int64_t lda, + const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, const bfloat16* a, std::int64_t lda, + const bfloat16* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const std::int8_t* a, std::int64_t lda, std::int8_t ao, + const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const std::int8_t* a, std::int64_t lda, std::int8_t ao, + const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, + const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, oneapi::math::offset offsetc, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, + const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event symm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event symm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event symm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event symm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hemm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hemm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, float beta, float* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, double beta, + double* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* k, float* alpha, const float** a, + std::int64_t* lda, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* k, double* alpha, const double** a, + std::int64_t* lda, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float beta, float* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double beta, double* c, + std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, + std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event herk(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + float alpha, const std::complex* a, std::int64_t lda, + float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event herk(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + double alpha, const std::complex* a, std::int64_t lda, + double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + float alpha, const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + double alpha, const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event her2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event her2k(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trmm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, float* b, + std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trmm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, double* b, + std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trmm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trmm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, float* b, + std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, double* b, + std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side left_right, + oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side* left_right, + oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, + std::int64_t* m, std::int64_t* n, float* alpha, + const float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side* left_right, + oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, + std::int64_t* m, std::int64_t* n, double* alpha, + const double** a, std::int64_t* lda, double** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side* left_right, + oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side* left_right, + oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stridea, const float* x, + std::int64_t incx, std::int64_t stridex, float beta, float* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stridea, const double* x, + std::int64_t incx, std::int64_t stridex, double beta, + double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, - std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event gemv_batch( + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, std::complex beta, + std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, - std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event gemv_batch( + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, + const float** a, std::int64_t* lda, const float** x, + std::int64_t* incx, float* beta, float** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, + const double** a, std::int64_t* lda, const double** x, + std::int64_t* incx, double* beta, double** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, + std::int64_t m, std::int64_t n, const float* a, + std::int64_t lda, std::int64_t stridea, const float* x, + std::int64_t incx, std::int64_t stridex, float* c, + std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, + std::int64_t m, std::int64_t n, const double* a, + std::int64_t lda, std::int64_t stridea, const double* x, + std::int64_t incx, std::int64_t stridex, double* c, + std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, + std::int64_t m, std::int64_t n, const std::complex* a, + std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, + std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, + std::int64_t m, std::int64_t n, const std::complex* a, + std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, + std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side* left_right, + std::int64_t* m, std::int64_t* n, const float** a, + std::int64_t* lda, const float** x, std::int64_t* incx, + float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side* left_right, + std::int64_t* m, std::int64_t* n, const double** a, + std::int64_t* lda, const double** x, std::int64_t* incx, + double** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side* left_right, + std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side* left_right, + std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, + const float* a, std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, + const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, + const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hemv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hemv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event her(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event her(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event her2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event her2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hpr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hpr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hpr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hpr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::int64_t k, float alpha, const float* a, std::int64_t lda, + const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::int64_t k, double alpha, const double* a, std::int64_t lda, + const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event symv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event symv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, float* a, + std::int64_t lda, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, double* a, + std::int64_t lda, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, const float* y, + std::int64_t incy, float* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, const float* a, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, const double* a, const double* x, std::int64_t incx, + double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, float* a, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, double* a, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, const float* y, + std::int64_t incy, float* a, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event iamax(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event iamax(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event iamin(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event iamin(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event asum(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event asum(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy(sycl::queue& queue, std::int64_t n, float alpha, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy(sycl::queue& queue, std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, float* alpha, + const float** x, std::int64_t* incx, float** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, double* alpha, + const double** x, std::int64_t* incx, double** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, std::int64_t stridex, + float* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, double alpha, + const double* x, std::int64_t incx, std::int64_t stridex, + double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpby(sycl::queue& queue, std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpby(sycl::queue& queue, std::int64_t n, double alpha, const double* x, + std::int64_t incx, const double beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, + std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const float** x, + std::int64_t* incx, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const double** x, + std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, + const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, + const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t stridex, float* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t stridex, double* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event copy_batch(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dot(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, + double* result, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sdsdot(sycl::queue& queue, std::int64_t n, float sb, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, + float* result, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, double* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event nrm2(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event nrm2(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + float c, float s, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + double c, double s, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rot(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, + float* y, std::int64_t incy, float c, float s, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rot(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, + double* y, std::int64_t incy, double c, double s, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rotg(sycl::queue& queue, float* a, float* b, float* c, float* s, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rotg(sycl::queue& queue, double* a, double* b, double* c, double* s, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, + float* c, std::complex* s, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rotg(sycl::queue& queue, std::complex* a, + std::complex* b, double* c, std::complex* s, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rotm(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, + float* y, std::int64_t incy, float* param, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rotm(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, + double* y, std::int64_t incy, double* param, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rotmg(sycl::queue& queue, float* d1, float* d2, float* x1, float y1, + float* param, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event rotmg(sycl::queue& queue, double* d1, double* d2, double* x1, double y1, + double* param, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event scal(sycl::queue& queue, std::int64_t n, float alpha, float* x, + std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event scal(sycl::queue& queue, std::int64_t n, double alpha, double* x, + std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event scal(sycl::queue& queue, std::int64_t n, float alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event scal(sycl::queue& queue, std::int64_t n, double alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event swap(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, + float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event swap(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, + double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, + const float** a, std::int64_t* lda, const float** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, double* alpha, + const double** a, std::int64_t* lda, const double** b, + std::int64_t* ldb, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, sycl::half* alpha, + const sycl::half** a, std::int64_t* lda, const sycl::half** b, + std::int64_t* ldb, sycl::half* beta, sycl::half** c, + std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, + const sycl::half** a, std::int64_t* lda, const sycl::half** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, + const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, + float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float* alpha, + const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, + std::int32_t** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, const float* b, + std::int64_t ldb, std::int64_t stride_b, float beta, float* c, + std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, const double* b, + std::int64_t ldb, std::int64_t stride_b, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, +ONEMATH_EXPORT sycl::event gemm_batch( + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch( + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::half alpha, + const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, + sycl::half beta, sycl::half* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, + const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, + float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, + const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, + float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, + const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, + float beta, std::int32_t* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, + float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, + double* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, + std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, std::int64_t stride_a, + float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, std::int64_t stride_a, + double* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float alpha, float* ab, + std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double alpha, double* ab, + std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + std::complex* ab, std::int64_t lda, + std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, + std::complex alpha, std::complex* ab, + std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float beta, + const float* b, std::int64_t ldb, std::int64_t stride_b, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, +ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, + std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double beta, + const double* b, std::int64_t ldb, std::int64_t stride_b, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event omatadd_batch( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex *c, std::int64_t ldc, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, const std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event omatadd_batch( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, std::complex *c, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - float *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - double *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stridea, float *b, - std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, double *b, - std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, std::int64_t ldb, - std::int64_t strideb, - const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue &queue, oneapi::math::transpose trans, +ONEMATH_EXPORT sycl::event omatcopy(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, std::int64_t ldb, - std::int64_t strideb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, float *ab, std::int64_t lda, - std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, double *ab, std::int64_t lda, - std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event imatcopy(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, float beta, - const float *b, std::int64_t ldb, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stridea, float* b, + std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stridea, double* b, + std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, std::complex* b, std::int64_t ldb, + std::int64_t strideb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatcopy2(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, std::complex* b, + std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float alpha, float* ab, + std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double alpha, double* ab, + std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + std::complex* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, double beta, - const double *b, std::int64_t ldb, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event imatcopy(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + std::complex* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event omatadd(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, float beta, + const float* b, std::int64_t ldb, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event omatadd(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, double beta, + const double* b, std::int64_t ldb, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); -ONEMATH_EXPORT sycl::event omatadd(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event omatadd(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event omatadd(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, - std::int64_t* n, float* alpha, const float** a, - std::int64_t* lda, float** b, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* groupsize, - const std::vector& dependencies = {}); + std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, - std::int64_t* n, double* alpha, const double** a, - std::int64_t* lda, double** b, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* groupsize, - const std::vector& dependencies = {}); + std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, - std::int64_t* n, std::complex* alpha, - const std::complex** a, std::int64_t* lda, - std::complex** b, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* groupsize, - const std::vector& dependencies = {}); + std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, - std::int64_t* n, std::complex* alpha, - const std::complex** a, std::int64_t* lda, - std::complex** b, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* groupsize, - const std::vector& dependencies = {}); + std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, - std::int64_t* n, float* alpha, float** ab, - std::int64_t* lda, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* groupsize, - const std::vector& dependencies = {}); + std::int64_t* n, float* alpha, float** ab, + std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, - std::int64_t* n, double* alpha, double** ab, - std::int64_t* lda, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* groupsize, - const std::vector& dependencies = {}); + std::int64_t* n, double* alpha, double** ab, + std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, - std::int64_t* n, std::complex* alpha, - std::complex** ab, std::int64_t* lda, - std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies = {}); + std::int64_t* n, std::complex* alpha, + std::complex** ab, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, std::int64_t* m, - std::int64_t* n, std::complex* alpha, - std::complex** ab, std::int64_t* lda, - std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies = {}); + std::int64_t* n, std::complex* alpha, + std::complex** ab, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, + const std::vector& dependencies = {}); diff --git a/include/oneapi/math/blas/detail/portblas/blas_ct.hxx b/include/oneapi/math/blas/detail/portblas/blas_ct.hxx index 8f505ef8c..2f3694c6e 100644 --- a/include/oneapi/math/blas/detail/portblas/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/portblas/blas_ct.hxx @@ -20,1894 +20,1894 @@ // Buffer APIs void herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, - std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { + std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1>& a, + std::int64_t lda, float beta, sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, - std::int64_t lda, double beta, sycl::buffer, 1> &c, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1>& a, + std::int64_t lda, double beta, sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx) { + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx) { + sycl::buffer& x, std::int64_t incx) { oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& x, std::int64_t incx) { oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, x, incx); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { oneapi::math::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, - a); + a); } void spr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { oneapi::math::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, - a); + a); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, double beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::half beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::half beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - float beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::portblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, - y, incy, a, lda); + y, incy, a, lda); } void her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { oneapi::math::blas::portblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, - y, incy, a, lda); + y, incy, a, lda); } void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + lda, x, incx, beta, y, incy); } void hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + lda, x, incx, beta, y, incy); } void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, float c, float s) { oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, double s) { oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { +void rot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, float c, float s) { oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { +void rot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, double c, double s) { oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, + stridex, y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, + stridex, y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, + stridex, y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, + stridex, y, incy, stridey, batch_size); } void axpby(backend_selector selector, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, - incy); + incy); } void axpby(backend_selector selector, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, - incy); + incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, - incy); + incy); } void axpby(backend_selector selector, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, - incy); + incy); } void sdsdot(backend_selector selector, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &result) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, - result); + result); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { + oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, + a, lda); } void gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { + oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, + a, lda); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { - oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + alpha, a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + alpha, a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + alpha, a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, + alpha, a, lda, b, ldb, beta, c, ldc); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, float beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, - stridey, batch_size); + oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, + lda, stridea, x, incx, stridex, beta, y, incy, + stridey, batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, double beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, - stridey, batch_size); + oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, + lda, stridea, x, incx, stridex, beta, y, incy, + stridey, batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, - stridey, batch_size); + oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, + lda, stridea, x, incx, stridex, beta, y, incy, + stridey, batch_size); } void gemv_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, - stridey, batch_size); + oneapi::math::blas::portblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, + lda, stridea, x, incx, stridex, beta, y, incy, + stridey, batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + float alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { + oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, + a, lda); } void her(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + double alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { + oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, + a, lda); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { + float alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { oneapi::math::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, - a); + a); } void hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { + double alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { oneapi::math::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, - a); + a); } -void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void iamin(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } -void iamin(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void iamin(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, - incx, beta, y, incy); + incx, beta, y, incy); } void hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, - incx, beta, y, incy); + incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy) { + float alpha, sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, - incx, beta, y, incy); + incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy) { + double alpha, sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { oneapi::math::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, - n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, - co); + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, + ldc, co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { oneapi::math::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, - n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, - co); + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, + ldc, co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { oneapi::math::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, - n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, - co); + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, + ldc, co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { oneapi::math::blas::portblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, - n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, - co); + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, + ldc, co); } -void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void swap(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } -void swap(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void swap(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { + oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, + a, lda); } void geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { - oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { + oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, + a, lda); } void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::half beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::portblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, - y, incy, a, lda); + y, incy, a, lda); } void syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::portblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, - y, incy, a, lda); + y, incy, a, lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void ger(backend_selector selector, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { oneapi::math::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { oneapi::math::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, - y, incy, a); + y, incy, a); } void hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { oneapi::math::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, - y, incy, a); + y, incy, a); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy); + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy); + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy); + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy); + lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, + n, k, a, lda, x, incx); } void symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { - oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, + a, lda); } void syr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { - oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, + a, lda); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, float y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, float y1, + sycl::buffer& param) { oneapi::math::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, double y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, double y1, + sycl::buffer& param) { oneapi::math::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, a, lda, x, incx); } -void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void copy(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } -void copy(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void copy(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, - incy, stridey, batch_size); + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, - incy, stridey, batch_size); + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, - incy, stridey, batch_size); + incy, stridey, batch_size); } void copy_batch(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, - incy, stridey, batch_size); + incy, stridey, batch_size); } void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::portblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::portblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, + x, incx, beta, y, incy); } void hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - oneapi::math::blas::portblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + oneapi::math::blas::portblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, + x, incx, beta, y, incy); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, - k, alpha, a, lda, b, ldb, beta, c, ldc); + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, - k, alpha, a, lda, b, ldb, beta, c, ldc); + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, - k, alpha, a, lda, b, ldb, beta, c, ldc); + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, - k, alpha, a, lda, b, ldb, beta, c, ldc); + k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { - oneapi::math::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + oneapi::math::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + lda, x, incx, beta, y, incy); } void sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { - oneapi::math::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + oneapi::math::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, + lda, x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { - oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { + oneapi::math::blas::portblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, + n, k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a) { + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { oneapi::math::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, - y, incy, a); + y, incy, a); } void spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a) { + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { oneapi::math::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, - y, incy, a); + y, incy, a); } -void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void iamax(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void iamax(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void iamax(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m) { +void rotm(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { oneapi::math::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void rotm(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m) { +void rotm(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param) { oneapi::math::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(backend_selector selector, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, - b, ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, - b, ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, - b, ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, - b, ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + alpha, a, lda, b, ldb, beta, c, ldc); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, double beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, + alpha, a, lda, b, ldb, beta, c, ldc); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::portblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + float alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { + oneapi::math::blas::portblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, + x, incx, beta, y, incy); } void symv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - oneapi::math::blas::portblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { + oneapi::math::blas::portblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, + x, incx, beta, y, incy); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + oneapi::math::blas::portblas::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + oneapi::math::blas::portblas::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + oneapi::math::blas::portblas::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + oneapi::math::blas::portblas::MAJOR::omatcopy_batch( + selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, + ab, lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, + ab, lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, + ab, lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { - oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + oneapi::math::blas::portblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, + ab, lda, ldb, stride, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::portblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, - b, ldb); + b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, - b, ldb); + b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, - b, ldb); + b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, - b, ldb); + b, ldb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::portblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, float beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc); + a, lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc); + a, lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc); + a, lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, - a, lda, beta, b, ldb, c, ldc); + a, lda, beta, b, ldb, c, ldc); } // USM APIs sycl::event syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *a, std::int64_t lda, const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syr2( selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } sycl::event syr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syr2( selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, float alpha, - float *x, std::int64_t incx, const std::vector &dependencies) { + float* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, double alpha, - double *x, std::int64_t incx, const std::vector &dependencies) { + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, float alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, std::int64_t n, double alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, - const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + float alpha, const float* x, std::int64_t incx, float* a, + const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, + alpha, x, incx, a, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, - const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + double alpha, const double* x, std::int64_t incx, double* a, + const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::spr(selector.get_queue(), upper_lower, n, + alpha, x, incx, a, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::hpmv( selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::hpmv( selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; @@ -1915,9 +1915,9 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; @@ -1925,54 +1925,54 @@ sycl::event syrk(backend_selector selector, uplo upper_lower, sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, - const float **a, std::int64_t *lda, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, float* alpha, + const float** a, std::int64_t* lda, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, - const double **a, std::int64_t *lda, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, double* alpha, + const double** a, std::int64_t* lda, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); @@ -1980,10 +1980,10 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float beta, float *c, + transpose trans, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -1992,9 +1992,9 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_ sycl::event syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2003,10 +2003,10 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_ sycl::event syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, std::complex *c, std::int64_t ldc, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2015,10 +2015,10 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_ sycl::event syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, std::complex *c, std::int64_t ldc, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2026,330 +2026,330 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_ } sycl::event her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::her2( selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } sycl::event her2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::her2( selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, float c, float s, const std::vector &dependencies) { + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, float c, float s, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, - c, s, dependencies); + c, s, dependencies); return done; } sycl::event rot(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, + std::complex* x, std::int64_t incx, std::complex* y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, - c, s, dependencies); + c, s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, - c, s, dependencies); + c, s, dependencies); return done; } -sycl::event rot(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, - c, s, dependencies); + c, s, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + const float* x, std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, + y, incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + const double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, + y, incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, + y, incy, dependencies); return done; } sycl::event axpy(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, + y, incy, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, float *alpha, - const float **x, std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, float* alpha, + const float** x, std::int64_t* incx, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, double *alpha, - const double **x, std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, double* alpha, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, + const float* x, std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, + const double* x, std::int64_t incx, std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, std::int64_t incy, + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, std::int64_t incy, + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + const float* x, std::int64_t incx, const float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double beta, double *y, - std::int64_t incy, const std::vector &dependencies) { + const double* x, std::int64_t incx, const double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, + incx, y, incy, a, lda, dependencies); return done; } sycl::event gerc(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, + incx, y, incy, a, lda, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, - n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + n, k, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, - n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + n, k, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, - n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + n, k, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, - n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + n, k, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float beta, float *y, std::int64_t incy, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float beta, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); @@ -2357,11 +2357,11 @@ sycl::event gemv_batch(backend_selector selector, transpose t } sycl::event gemv_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double beta, double *y, std::int64_t incy, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double beta, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); @@ -2370,11 +2370,11 @@ sycl::event gemv_batch(backend_selector selector, transpose t sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); @@ -2383,58 +2383,58 @@ sycl::event gemv_batch(backend_selector selector, transpose t sycl::event gemv_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, - float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, const float** x, std::int64_t* incx, float* beta, + float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, - double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, const double** x, std::int64_t* incx, double* beta, + double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); @@ -2442,10 +2442,10 @@ sycl::event gemv_batch(backend_selector selector, transpose * } sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, const float* a, std::int64_t lda, + std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); @@ -2453,10 +2453,10 @@ sycl::event dgmm_batch(backend_selector selector, side left_r } sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t m, std::int64_t n, const double* a, std::int64_t lda, + std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); @@ -2464,11 +2464,11 @@ sycl::event dgmm_batch(backend_selector selector, side left_r } sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, + std::int64_t m, std::int64_t n, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); @@ -2476,55 +2476,55 @@ sycl::event dgmm_batch(backend_selector selector, side left_r } sycl::event dgmm_batch(backend_selector selector, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, + std::int64_t m, std::int64_t n, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const float** a, std::int64_t* lda, + const float** x, std::int64_t* incx, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const double** a, std::int64_t* lda, + const double** x, std::int64_t* incx, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); @@ -2532,162 +2532,162 @@ sycl::event dgmm_batch(backend_selector selector, side *left_ } sycl::event her(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, + alpha, x, incx, a, lda, dependencies); return done; } sycl::event her(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::her(selector.get_queue(), upper_lower, n, + alpha, x, incx, a, lda, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, + alpha, x, incx, a, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, + alpha, x, incx, a, dependencies); return done; } -sycl::event iamin(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamin(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - sycl::half *alpha, const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, sycl::half *beta, sycl::half **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + sycl::half* alpha, const sycl::half** a, std::int64_t* lda, + const sycl::half** b, std::int64_t* ldb, sycl::half* beta, sycl::half** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const sycl::half** a, std::int64_t* lda, const sycl::half** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, std::int32_t** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const float **a, std::int64_t *lda, const float **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const float** a, std::int64_t* lda, const float** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - double *alpha, const double **a, std::int64_t *lda, const double **b, - std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + double* alpha, const double** a, std::int64_t* lda, const double** b, + std::int64_t* ldb, double* beta, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **b, std::int64_t *ldb, std::complex *beta, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, std::complex* beta, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); @@ -2696,11 +2696,11 @@ sycl::event gemm_batch(backend_selector selector, transpose * sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, const sycl::half *a, std::int64_t lda, - std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, + sycl::half alpha, const sycl::half* a, std::int64_t lda, + std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, sycl::half* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2709,10 +2709,10 @@ sycl::event gemm_batch(backend_selector selector, transpose t sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2721,10 +2721,10 @@ sycl::event gemm_batch(backend_selector selector, transpose t sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2733,10 +2733,10 @@ sycl::event gemm_batch(backend_selector selector, transpose t sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2745,10 +2745,10 @@ sycl::event gemm_batch(backend_selector selector, transpose t sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, - const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, + const float* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2757,10 +2757,10 @@ sycl::event gemm_batch(backend_selector selector, transpose t sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, - const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, + const double* b, std::int64_t ldb, std::int64_t stride_b, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2769,11 +2769,11 @@ sycl::event gemm_batch(backend_selector selector, transpose t sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, const std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::complex beta, std::complex *c, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2782,11 +2782,11 @@ sycl::event gemm_batch(backend_selector selector, transpose t sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, const std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::complex beta, std::complex *c, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2794,180 +2794,180 @@ sycl::event gemm_batch(backend_selector selector, transpose t } sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *a, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, const std::vector &dependencies) { + float alpha, const float* a, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::spmv( selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } sycl::event spmv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *a, const double *x, std::int64_t incx, double beta, - double *y, std::int64_t incy, const std::vector &dependencies) { + double alpha, const double* a, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::spmv( selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event swap(backend_selector selector, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, + incx, y, incy, a, lda, dependencies); return done; } sycl::event geru(backend_selector selector, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, + incx, y, incy, a, lda, dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event nrm2(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *result, const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *result, const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - sycl::half beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, - std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = - oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16* a, + std::int64_t lda, const bfloat16* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -2976,10 +2976,10 @@ sycl::event gemm_bias(backend_selector selector, transpose tr sycl::event gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -2988,10 +2988,10 @@ sycl::event gemm_bias(backend_selector selector, transpose tr sycl::event gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3000,10 +3000,10 @@ sycl::event gemm_bias(backend_selector selector, transpose tr sycl::event gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -3011,87 +3011,87 @@ sycl::event gemm_bias(backend_selector selector, transpose tr } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const std::complex *a, - std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const std::complex* a, + std::int64_t lda, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const std::complex *a, - std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, double alpha, const std::complex* a, + std::int64_t lda, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event ger(backend_selector selector, std::int64_t m, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *a, std::int64_t lda, const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event ger(backend_selector selector, std::int64_t m, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3100,9 +3100,9 @@ sycl::event trsm_batch(backend_selector selector, side left_r sycl::event trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3111,10 +3111,10 @@ sycl::event trsm_batch(backend_selector selector, side left_r sycl::event trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3123,56 +3123,56 @@ sycl::event trsm_batch(backend_selector selector, side left_r sycl::event trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, - double **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); @@ -3180,760 +3180,760 @@ sycl::event trsm_batch(backend_selector selector, side *left_ } sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotu(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, - upper_lower, m, n, alpha, a, lda, b, ldb, - beta, c, ldc, dependencies); + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::hemm(selector.get_queue(), left_right, - upper_lower, m, n, alpha, a, lda, b, ldb, - beta, c, ldc, dependencies); + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, - alpha, x, incx, y, incy, a, dependencies); + alpha, x, incx, y, incy, a, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, - alpha, x, incx, y, incy, a, dependencies); + alpha, x, incx, y, incy, a, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, - a, lda, x, incx, beta, y, incy, dependencies); + a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, const std::vector &dependencies) { + std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, - a, lda, x, incx, beta, y, incy, dependencies); + a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, - a, lda, x, incx, beta, y, incy, dependencies); + a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, - a, lda, x, incx, beta, y, incy, dependencies); + a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, - double *x, std::int64_t incx, const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, - upper_lower, m, n, alpha, a, lda, b, ldb, - beta, c, ldc, dependencies); + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, - upper_lower, m, n, alpha, a, lda, b, ldb, - beta, c, ldc, dependencies); + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, - upper_lower, m, n, alpha, a, lda, b, ldb, - beta, c, ldc, dependencies); + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::symm(selector.get_queue(), left_right, - upper_lower, m, n, alpha, a, lda, b, ldb, - beta, c, ldc, dependencies); + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotc(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + float alpha, const float* x, std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, + alpha, x, incx, a, lda, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + double alpha, const double* x, std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::syr(selector.get_queue(), upper_lower, n, + alpha, x, incx, a, lda, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, + alpha, a, lda, b, ldb, dependencies); return done; } -sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, - float y1, float *param, const std::vector &dependencies) { +sycl::event rotmg(backend_selector selector, float* d1, float* d2, float* x1, + float y1, float* param, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, - param, dependencies); + param, dependencies); return done; } -sycl::event rotmg(backend_selector selector, double *d1, double *d2, double *x1, - double y1, double *param, const std::vector &dependencies) { +sycl::event rotmg(backend_selector selector, double* d1, double* d2, double* x1, + double y1, double* param, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, - param, dependencies); + param, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -sycl::event copy(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const float **x, std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const float** x, std::int64_t* incx, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const double **x, std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t *n, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, +sycl::event copy_batch(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, std::int64_t n, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, + const double* x, std::int64_t incx, std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transpose transb, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, + transa, transb, n, k, alpha, a, lda, b, + ldb, beta, c, ldc, dependencies); return done; } sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transpose transb, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, + transa, transb, n, k, alpha, a, lda, b, + ldb, beta, c, ldc, dependencies); return done; } sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, + transa, transb, n, k, alpha, a, lda, b, + ldb, beta, c, ldc, dependencies); return done; } sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::gemmt(selector.get_queue(), upper_lower, + transa, transb, n, k, alpha, a, lda, b, + ldb, beta, c, ldc, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + std::int64_t k, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event asum(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, float *result, const std::vector &dependencies) { +sycl::event asum(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, double *result, const std::vector &dependencies) { +sycl::event asum(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, - double *x, std::int64_t incx, const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *a, const std::vector &dependencies) { + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, - alpha, x, incx, y, incy, a, dependencies); + alpha, x, incx, y, incy, a, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, const std::vector &dependencies) { + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, - alpha, x, incx, y, incy, a, dependencies); + alpha, x, incx, y, incy, a, dependencies); return done; } -sycl::event iamax(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamax(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event rotm(backend_selector selector, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotm(backend_selector selector, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotg(backend_selector selector, float *a, float *b, float *c, - float *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, float* a, float* b, float* c, + float* s, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, double *a, double *b, double *c, - double *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, double* a, double* b, double* c, + double* s, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, float *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, float* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, double *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, double* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } sycl::event sdsdot(backend_selector selector, std::int64_t n, float sb, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *result, const std::vector &dependencies) { + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, - incy, result, dependencies); + incy, result, dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, - n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + n, k, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, - n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies); + n, k, alpha, a, lda, b, ldb, beta, c, + ldc, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, std::int64_t n, - float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, - float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { + float alpha, const float* a, std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, std::int64_t n, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3941,10 +3941,10 @@ sycl::event omatcopy_batch(backend_selector selector, transpo } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3953,9 +3953,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpo sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3964,9 +3964,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpo sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3974,18 +3974,18 @@ sycl::event omatcopy_batch(backend_selector selector, transpo } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, + std::int64_t m, std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, + std::int64_t m, std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -3993,9 +3993,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpo sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4003,9 +4003,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpo sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -4013,10 +4013,10 @@ sycl::event imatcopy_batch(backend_selector selector, transpo sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, + const float* a, std::int64_t lda, std::int64_t stride_a, float beta, + const float* b, std::int64_t ldb, std::int64_t stride_b, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4025,10 +4025,10 @@ sycl::event omatadd_batch(backend_selector selector, transpos sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + const double* b, std::int64_t ldb, std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4037,11 +4037,11 @@ sycl::event omatadd_batch(backend_selector selector, transpos sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4050,11 +4050,11 @@ sycl::event omatadd_batch(backend_selector selector, transpos sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -4062,233 +4062,233 @@ sycl::event omatadd_batch(backend_selector selector, transpos } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, - std::int64_t ldb, const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, float* b, + std::int64_t ldb, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, dependencies); + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, - std::int64_t ldb, const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, double* b, + std::int64_t ldb, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, dependencies); + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, dependencies); + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, dependencies); + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, double* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, dependencies); + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, dependencies); + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, dependencies); + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, dependencies); + alpha, ab, lda, ldb, dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, - n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + transpose transb, std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, float beta, const float* b, std::int64_t ldb, float* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, + m, n, alpha, a, lda, beta, b, ldb, c, + ldc, dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, - n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + transpose transb, std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double beta, const double* b, std::int64_t ldb, double* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, + m, n, alpha, a, lda, beta, b, ldb, c, + ldc, dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, - n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, + m, n, alpha, a, lda, beta, b, ldb, c, + ldc, dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, - n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + auto done = oneapi::math::blas::portblas::MAJOR::omatadd(selector.get_queue(), transa, transb, + m, n, alpha, a, lda, beta, b, ldb, c, + ldc, dependencies); return done; } -sycl::event omatcopy_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; } -sycl::event omatcopy_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; } -sycl::event omatcopy_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event omatcopy_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; } -sycl::event omatcopy_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event omatcopy_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); return done; } -sycl::event imatcopy_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, float **ab, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event imatcopy_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, float** ab, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; } -sycl::event imatcopy_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, double **ab, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event imatcopy_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, double** ab, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; } -sycl::event imatcopy_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - std::complex **ab, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + std::complex** ab, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); return done; } -sycl::event imatcopy_batch(backend_selector selector, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - std::complex **ab, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(backend_selector selector, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + std::complex** ab, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { auto done = oneapi::math::blas::portblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); diff --git a/include/oneapi/math/blas/detail/rocblas/blas_ct.hxx b/include/oneapi/math/blas/detail/rocblas/blas_ct.hxx index cdfeae51b..42332ff59 100644 --- a/include/oneapi/math/blas/detail/rocblas/blas_ct.hxx +++ b/include/oneapi/math/blas/detail/rocblas/blas_ct.hxx @@ -20,1898 +20,1902 @@ **************************************************************************/ void herk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, - int64_t k, float alpha, sycl::buffer, 1> &a, int64_t lda, float beta, - sycl::buffer, 1> &c, int64_t ldc) { + int64_t k, float alpha, sycl::buffer, 1>& a, int64_t lda, float beta, + sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void herk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, - int64_t k, double alpha, sycl::buffer, 1> &a, int64_t lda, - double beta, sycl::buffer, 1> &c, int64_t ldc) { + int64_t k, double alpha, sycl::buffer, 1>& a, int64_t lda, + double beta, sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::herk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void scal(backend_selector selector, int64_t n, float alpha, - sycl::buffer &x, int64_t incx) { + sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, int64_t n, double alpha, - sycl::buffer &x, int64_t incx) { + sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx) { + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx) { + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, int64_t n, float alpha, - sycl::buffer, 1> &x, int64_t incx) { + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void scal(backend_selector selector, int64_t n, double alpha, - sycl::buffer, 1> &x, int64_t incx) { + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::trmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer &a, sycl::buffer &x, + diag unit_diag, int64_t n, sycl::buffer& a, sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer &a, sycl::buffer &x, + diag unit_diag, int64_t n, sycl::buffer& a, sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void spr(backend_selector selector, uplo upper_lower, int64_t n, float alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a) { - oneapi::math::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + sycl::buffer& x, int64_t incx, sycl::buffer& a) { + oneapi::math::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, + a); } void spr(backend_selector selector, uplo upper_lower, int64_t n, double alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a) { - oneapi::math::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + sycl::buffer& x, int64_t incx, sycl::buffer& a) { + oneapi::math::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, x, incx, + a); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, - int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, + int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, + int64_t stride_b, float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, double alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, - int64_t stride_b, double beta, sycl::buffer &c, int64_t ldc, + int64_t m, int64_t n, int64_t k, double alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, + int64_t stride_b, double beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, sycl::half alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, - int64_t stride_b, sycl::half beta, sycl::buffer &c, int64_t ldc, + int64_t m, int64_t n, int64_t k, sycl::half alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, + int64_t stride_b, sycl::half beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, - int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, + int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, + int64_t stride_b, float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, - int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, + int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, + int64_t stride_b, float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void gemm_batch(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, - int64_t stride_b, float beta, sycl::buffer &c, int64_t ldc, + int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, + int64_t stride_b, float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemm_batch(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, stride_a, b, ldb, stride_b, beta, - c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, b, ldb, stride_b, beta, + c, ldc, stride_c, batch_size); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, float beta, - sycl::buffer &c, int64_t ldc) { + int64_t k, float alpha, sycl::buffer& a, int64_t lda, float beta, + sycl::buffer& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, - int64_t k, double alpha, sycl::buffer &a, int64_t lda, double beta, - sycl::buffer &c, int64_t ldc) { + int64_t k, double alpha, sycl::buffer& a, int64_t lda, double beta, + sycl::buffer& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, std::complex beta, sycl::buffer, 1> &c, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void syrk(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, std::complex beta, sycl::buffer, 1> &c, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::syrk(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, beta, c, ldc); + a, lda, beta, c, ldc); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, float beta, sycl::buffer &c, int64_t ldc, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, double beta, sycl::buffer &c, int64_t ldc, + int64_t n, int64_t k, double alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, double beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::syrk_batch(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, stride_a, beta, c, ldc, stride_c, - batch_size); + alpha, a, lda, stride_a, beta, c, ldc, stride_c, + batch_size); } void her2(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { - oneapi::math::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + std::complex alpha, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { + oneapi::math::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, + y, incy, a, lda); } void her2(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { - oneapi::math::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + std::complex alpha, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { + oneapi::math::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, x, incx, + y, incy, a, lda); } void hbmv(backend_selector selector, uplo upper_lower, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void hbmv(backend_selector selector, uplo upper_lower, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void rot(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, float c, float s) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, float c, float s) { oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void rot(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, double c, double s) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, double c, double s) { oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, float c, float s) { +void rot(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, float c, float s) { oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } -void rot(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, double c, double s) { +void rot(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, double c, double s) { oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, s); } void axpy(backend_selector selector, int64_t n, float alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, int64_t n, double alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy(backend_selector selector, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, incy); } void axpy_batch(backend_selector selector, int64_t n, float alpha, - sycl::buffer &x, int64_t incx, int64_t stridex, sycl::buffer &y, + sycl::buffer& x, int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, int64_t n, double alpha, - sycl::buffer &x, int64_t incx, int64_t stridex, - sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { + sycl::buffer& x, int64_t incx, int64_t stridex, + sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpy_batch(backend_selector selector, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, incx, stridex, - y, incy, stridey, batch_size); + y, incy, stridey, batch_size); } void axpby(backend_selector selector, int64_t n, float alpha, - sycl::buffer &x, int64_t incx, float beta, sycl::buffer &y, + sycl::buffer& x, int64_t incx, float beta, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, - incy); + incy); } void axpby(backend_selector selector, int64_t n, double alpha, - sycl::buffer &x, int64_t incx, double beta, sycl::buffer &y, + sycl::buffer& x, int64_t incx, double beta, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, - incy); + incy); } void axpby(backend_selector selector, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, - incy); + incy); } void axpby(backend_selector selector, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, beta, y, - incy); + incy); } void sdsdot(backend_selector selector, int64_t n, float sb, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &result) { + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, incy, - result); + result); } void gerc(backend_selector selector, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { oneapi::math::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void gerc(backend_selector selector, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { oneapi::math::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, + int64_t n, int64_t k, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, double beta, sycl::buffer& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { + int64_t n, int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void syr2k(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void gemv(backend_selector selector, transpose trans, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, float beta, sycl::buffer &y, int64_t incy) { + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, float beta, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, double beta, sycl::buffer &y, int64_t incy) { + double alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, double beta, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv(backend_selector selector, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemv_batch(backend_selector selector, transpose trans, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, int64_t stridea, - sycl::buffer &x, int64_t incx, int64_t stridex, float beta, - sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { + float alpha, sycl::buffer& a, int64_t lda, int64_t stridea, + sycl::buffer& x, int64_t incx, int64_t stridex, float beta, + sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, + stridey, batch_size); } void gemv_batch(backend_selector selector, transpose trans, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, int64_t stridea, - sycl::buffer &x, int64_t incx, int64_t stridex, double beta, - sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { + double alpha, sycl::buffer& a, int64_t lda, int64_t stridea, + sycl::buffer& x, int64_t incx, int64_t stridex, double beta, + sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, + stridey, batch_size); } void gemv_batch(backend_selector selector, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stridea, sycl::buffer, 1> &x, int64_t incx, - int64_t stridex, std::complex beta, sycl::buffer, 1> &y, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stridea, sycl::buffer, 1>& x, int64_t incx, + int64_t stridex, std::complex beta, sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, + stridey, batch_size); } void gemv_batch(backend_selector selector, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stridea, sycl::buffer, 1> &x, int64_t incx, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stridea, sycl::buffer, 1>& x, int64_t incx, int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::gemv_batch(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, x, incx, stridex, beta, y, incy, stridey, - batch_size); + stridea, x, incx, stridex, beta, y, incy, + stridey, batch_size); } void dgmm_batch(backend_selector selector, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &c, int64_t ldc, + sycl::buffer& a, int64_t lda, int64_t stridea, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stridea, - sycl::buffer &x, int64_t incx, int64_t stridex, - sycl::buffer &c, int64_t ldc, int64_t stridec, int64_t batch_size) { + sycl::buffer& a, int64_t lda, int64_t stridea, + sycl::buffer& x, int64_t incx, int64_t stridex, + sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stridea, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &c, int64_t ldc, int64_t stridec, + sycl::buffer, 1>& a, int64_t lda, int64_t stridea, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& c, int64_t ldc, int64_t stridec, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void dgmm_batch(backend_selector selector, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stridea, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &c, int64_t ldc, int64_t stridec, + sycl::buffer, 1>& a, int64_t lda, int64_t stridea, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& c, int64_t ldc, int64_t stridec, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::dgmm_batch(selector.get_queue(), left_right, m, n, a, lda, - stridea, x, incx, stridex, c, ldc, stridec, - batch_size); + stridea, x, incx, stridex, c, ldc, stridec, + batch_size); } void her(backend_selector selector, uplo upper_lower, int64_t n, float alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &a, int64_t lda) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& a, int64_t lda) { oneapi::math::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void her(backend_selector selector, uplo upper_lower, int64_t n, double alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &a, int64_t lda) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& a, int64_t lda) { oneapi::math::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void hpr(backend_selector selector, uplo upper_lower, int64_t n, float alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &a) { - oneapi::math::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& a) { + oneapi::math::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, + a); } void hpr(backend_selector selector, uplo upper_lower, int64_t n, double alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &a) { - oneapi::math::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, a); + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& a) { + oneapi::math::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, x, incx, + a); } -void iamin(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &result) { +void iamin(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } -void iamin(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &result) { +void iamin(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void iamin(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result); } void hpmv(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { - oneapi::math::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { + oneapi::math::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, + incx, beta, y, incy); } void hpmv(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { - oneapi::math::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { + oneapi::math::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, a, x, + incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, int64_t n, float alpha, - sycl::buffer &a, sycl::buffer &x, int64_t incx, float beta, - sycl::buffer &y, int64_t incy) { - oneapi::math::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + sycl::buffer& a, sycl::buffer& x, int64_t incx, float beta, + sycl::buffer& y, int64_t incy) { + oneapi::math::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, + incx, beta, y, incy); } void spmv(backend_selector selector, uplo upper_lower, int64_t n, double alpha, - sycl::buffer &a, sycl::buffer &x, int64_t incx, double beta, - sycl::buffer &y, int64_t incy) { - oneapi::math::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, incx, - beta, y, incy); + sycl::buffer& a, sycl::buffer& x, int64_t incx, double beta, + sycl::buffer& y, int64_t incy) { + oneapi::math::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, a, x, + incx, beta, y, incy); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, - sycl::buffer &a, int64_t lda, int8_t ao, sycl::buffer &b, - int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, int64_t ldc, - sycl::buffer &co) { + sycl::buffer& a, int64_t lda, int8_t ao, sycl::buffer& b, + int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, int64_t ldc, + sycl::buffer& co) { oneapi::math::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, - n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, - co); + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, - sycl::buffer &a, int64_t lda, int8_t ao, sycl::buffer &b, - int64_t ldb, int8_t bo, float beta, sycl::buffer &c, int64_t ldc, - sycl::buffer &co) { + sycl::buffer& a, int64_t lda, int8_t ao, sycl::buffer& b, + int64_t ldb, int8_t bo, float beta, sycl::buffer& c, int64_t ldc, + sycl::buffer& co) { oneapi::math::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, - n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, - co); + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, - sycl::buffer &a, int64_t lda, uint8_t ao, sycl::buffer &b, - int64_t ldb, int8_t bo, float beta, sycl::buffer &c, int64_t ldc, - sycl::buffer &co) { + sycl::buffer& a, int64_t lda, uint8_t ao, sycl::buffer& b, + int64_t ldb, int8_t bo, float beta, sycl::buffer& c, int64_t ldc, + sycl::buffer& co) { oneapi::math::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, - n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, - co); + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } void gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, - sycl::buffer &a, int64_t lda, uint8_t ao, sycl::buffer &b, - int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, int64_t ldc, - sycl::buffer &co) { + sycl::buffer& a, int64_t lda, uint8_t ao, sycl::buffer& b, + int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, int64_t ldc, + sycl::buffer& co) { oneapi::math::blas::rocblas::MAJOR::gemm_bias(selector.get_queue(), transa, transb, offsetc, m, - n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, - co); + n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co); } -void swap(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy) { +void swap(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } -void swap(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy) { +void swap(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void swap(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy); } void geru(backend_selector selector, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { oneapi::math::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void geru(backend_selector selector, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { + std::complex alpha, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { oneapi::math::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void nrm2(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { + sycl::buffer, 1>& x, int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void nrm2(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { + sycl::buffer, 1>& x, int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } -void nrm2(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &result) { +void nrm2(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result); } void gemm(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, + int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { - oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, + int64_t m, int64_t n, int64_t k, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, double beta, sycl::buffer& c, int64_t ldc) { - oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { - oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { - oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, sycl::half alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &b, int64_t ldb, sycl::half beta, - sycl::buffer &c, int64_t ldc) { - oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + int64_t m, int64_t n, int64_t k, sycl::half alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& b, int64_t ldb, sycl::half beta, + sycl::buffer& c, int64_t ldc) { + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, + int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { - oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc); } void gemm(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, + int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { - oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, a, - lda, b, ldb, beta, c, ldc); + oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc); } void syr2(backend_selector selector, uplo upper_lower, int64_t n, float alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { - oneapi::math::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { + oneapi::math::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + y, incy, a, lda); } void syr2(backend_selector selector, uplo upper_lower, int64_t n, double alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { - oneapi::math::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a, lda); + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { + oneapi::math::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + y, incy, a, lda); } void ger(backend_selector selector, int64_t m, int64_t n, float alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { oneapi::math::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void ger(backend_selector selector, int64_t m, int64_t n, double alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { oneapi::math::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, y, incy, a, - lda); + lda); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void dotu(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void dotu(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, result); } void hemm(backend_selector selector, side left_right, uplo upper_lower, int64_t m, - int64_t n, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { + int64_t n, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hemm(backend_selector selector, side left_right, uplo upper_lower, int64_t m, - int64_t n, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { + int64_t n, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void hpr2(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a) { - oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + std::complex alpha, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a) { + oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + y, incy, a); } void hpr2(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a) { - oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + std::complex alpha, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a) { + oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + y, incy, a); } void gbmv(backend_selector selector, transpose trans, int64_t m, int64_t n, - int64_t kl, int64_t ku, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx, float beta, sycl::buffer &y, + int64_t kl, int64_t ku, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx, float beta, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy); + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, int64_t m, int64_t n, - int64_t kl, int64_t ku, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx, double beta, sycl::buffer &y, + int64_t kl, int64_t ku, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx, double beta, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy); + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy); + lda, x, incx, beta, y, incy); } void gbmv(backend_selector selector, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy); + lda, x, incx, beta, y, incy); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { + diag unit_diag, int64_t n, int64_t k, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { + diag unit_diag, int64_t n, int64_t k, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tbmv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void symm(backend_selector selector, side left_right, uplo upper_lower, int64_t m, - int64_t n, float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { + int64_t n, float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, int64_t m, - int64_t n, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, + int64_t n, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, double beta, sycl::buffer& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, int64_t m, - int64_t n, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { + int64_t n, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void symm(backend_selector selector, side left_right, uplo upper_lower, int64_t m, - int64_t n, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { + int64_t n, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc); + alpha, a, lda, b, ldb, beta, c, ldc); } void dotc(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void dotc(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result) { oneapi::math::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, result); } void syr(backend_selector selector, uplo upper_lower, int64_t n, float alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { + sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { oneapi::math::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void syr(backend_selector selector, uplo upper_lower, int64_t n, double alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { + sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { oneapi::math::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, x, incx, a, - lda); + lda); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } void trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, upper_lower, trans, - unit_diag, m, n, alpha, a, lda, b, ldb); + unit_diag, m, n, alpha, a, lda, b, ldb); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, float y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, float y1, + sycl::buffer& param) { oneapi::math::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } -void rotmg(backend_selector selector, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, double y1, - sycl::buffer ¶m) { +void rotmg(backend_selector selector, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, double y1, + sycl::buffer& param) { oneapi::math::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, param); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer &a, sycl::buffer &x, + diag unit_diag, int64_t n, sycl::buffer& a, sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer &a, sycl::buffer &x, + diag unit_diag, int64_t n, sycl::buffer& a, sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, x, incx); + a, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } void trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::trsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - a, lda, x, incx); + a, lda, x, incx); } -void copy(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy) { +void copy(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } -void copy(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy) { +void copy(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } void copy(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy); } -void copy_batch(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, +void copy_batch(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, - incy, stridey, batch_size); + incy, stridey, batch_size); } -void copy_batch(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, +void copy_batch(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, - incy, stridey, batch_size); + incy, stridey, batch_size); } void copy_batch(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, - incy, stridey, batch_size); + incy, stridey, batch_size); } void copy_batch(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::copy_batch(selector.get_queue(), n, x, incx, stridex, y, - incy, stridey, batch_size); + incy, stridey, batch_size); } void hemv(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void hemv(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::hemv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, - transpose transb, int64_t n, int64_t k, float alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &b, int64_t ldb, float beta, - sycl::buffer &c, int64_t ldc) { + transpose transb, int64_t n, int64_t k, float alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& b, int64_t ldb, float beta, + sycl::buffer& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, - k, alpha, a, lda, b, ldb, beta, c, ldc); + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, - transpose transb, int64_t n, int64_t k, double alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &b, int64_t ldb, double beta, - sycl::buffer &c, int64_t ldc) { + transpose transb, int64_t n, int64_t k, double alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& b, int64_t ldb, double beta, + sycl::buffer& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, - k, alpha, a, lda, b, ldb, beta, c, ldc); + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, - k, alpha, a, lda, b, ldb, beta, c, ldc); + k, alpha, a, lda, b, ldb, beta, c, ldc); } void gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, transb, n, - k, alpha, a, lda, b, ldb, beta, c, ldc); + k, alpha, a, lda, b, ldb, beta, c, ldc); } void asum(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { + sycl::buffer, 1>& x, int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void asum(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, sycl::buffer &result) { + sycl::buffer, 1>& x, int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } -void asum(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &result) { +void asum(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result); } void sbmv(backend_selector selector, uplo upper_lower, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, float beta, sycl::buffer &y, int64_t incy) { + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, float beta, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void sbmv(backend_selector selector, uplo upper_lower, int64_t n, int64_t k, - double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, double beta, sycl::buffer &y, int64_t incy) { + double alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, double beta, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, lda, - x, incx, beta, y, incy); + x, incx, beta, y, incy); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { + diag unit_diag, int64_t n, int64_t k, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { + diag unit_diag, int64_t n, int64_t k, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &x, int64_t incx) { + diag unit_diag, int64_t n, int64_t k, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& x, int64_t incx) { oneapi::math::blas::rocblas::MAJOR::tbsv(selector.get_queue(), upper_lower, trans, unit_diag, n, - k, a, lda, x, incx); + k, a, lda, x, incx); } void spr2(backend_selector selector, uplo upper_lower, int64_t n, float alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a) { - oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a) { + oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + y, incy, a); } void spr2(backend_selector selector, uplo upper_lower, int64_t n, double alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a) { - oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, y, - incy, a); + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a) { + oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, x, incx, + y, incy, a); } -void iamax(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &result) { +void iamax(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void iamax(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &result) { +void iamax(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } void iamax(backend_selector selector, int64_t n, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result); } -void rotm(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { +void rotm(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& param) { oneapi::math::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void rotm(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { +void rotm(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& param) { oneapi::math::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, param); } -void dot(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &result) { +void dot(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &result) { +void dot(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } -void dot(backend_selector selector, int64_t n, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &result) { +void dot(backend_selector selector, int64_t n, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& result) { oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, result); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &b, + sycl::buffer& a, int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, - b, ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, - b, ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, - b, ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::trsm_batch(selector.get_queue(), left_right, upper_lower, - trans, unit_diag, m, n, alpha, a, lda, stride_a, - b, ldb, stride_b, batch_size); + trans, unit_diag, m, n, alpha, a, lda, stride_a, + b, ldb, stride_b, batch_size); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, float beta, - sycl::buffer, 1> &c, int64_t ldc) { + int64_t n, int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, float beta, + sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } void her2k(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, double beta, - sycl::buffer, 1> &c, int64_t ldc) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, double beta, + sycl::buffer, 1>& c, int64_t ldc) { oneapi::math::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, alpha, - a, lda, b, ldb, beta, c, ldc); + a, lda, b, ldb, beta, c, ldc); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { +void rotg(backend_selector selector, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } -void rotg(backend_selector selector, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(backend_selector selector, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s); } void symv(backend_selector selector, uplo upper_lower, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - float beta, sycl::buffer &y, int64_t incy) { + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + float beta, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void symv(backend_selector selector, uplo upper_lower, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - double beta, sycl::buffer &y, int64_t incy) { + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + double beta, sycl::buffer& y, int64_t incy) { oneapi::math::blas::rocblas::MAJOR::symv(selector.get_queue(), upper_lower, n, alpha, a, lda, x, - incx, beta, y, incy); + incx, beta, y, incy); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::omatcopy_batch(selector.get_queue(), trans, m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size); + lda, stride_a, b, ldb, stride_b, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, alpha, ab, - lda, ldb, stride, batch_size); + lda, ldb, stride, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { oneapi::math::blas::rocblas::MAJOR::omatadd_batch(selector.get_queue(), transa, transb, m, n, - alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size); + alpha, a, lda, stride_a, beta, b, ldb, + stride_b, c, ldc, stride_c, batch_size); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { - oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { + oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, + b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { - oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { + oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, + b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { + oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, + b, ldb); } void omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { - oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, b, - ldb); + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { + oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, alpha, a, lda, + b, ldb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { oneapi::math::blas::rocblas::MAJOR::omatcopy2(selector.get_queue(), trans, m, n, alpha, a, lda, - stridea, b, ldb, strideb); + stridea, b, ldb, strideb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, alpha, ab, lda, - ldb); + ldb); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, float beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + a, lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { - oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { + oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + a, lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + a, lda, beta, b, ldb, c, ldc); } void omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { - oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, a, - lda, beta, b, ldb, c, ldc); + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { + oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, n, alpha, + a, lda, beta, b, ldb, c, ldc); } // USM APIs sycl::event syr2(backend_selector selector, uplo upper_lower, int64_t n, - float alpha, const float *x, int64_t incx, const float *y, int64_t incy, float *a, - int64_t lda, const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + float alpha, const float* x, int64_t incx, const float* y, int64_t incy, float* a, + int64_t lda, const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::syr2( + selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } sycl::event syr2(backend_selector selector, uplo upper_lower, int64_t n, - double alpha, const double *x, int64_t incx, const double *y, int64_t incy, - double *a, int64_t lda, const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::syr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + double alpha, const double* x, int64_t incx, const double* y, int64_t incy, + double* a, int64_t lda, const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::syr2( + selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } -sycl::event scal(backend_selector selector, int64_t n, float alpha, float *x, - int64_t incx, const std::vector &dependencies) { +sycl::event scal(backend_selector selector, int64_t n, float alpha, float* x, + int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } -sycl::event scal(backend_selector selector, int64_t n, double alpha, double *x, - int64_t incx, const std::vector &dependencies) { +sycl::event scal(backend_selector selector, int64_t n, double alpha, double* x, + int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, int64_t n, std::complex alpha, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, int64_t n, std::complex alpha, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, int64_t n, float alpha, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event scal(backend_selector selector, int64_t n, double alpha, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::scal(selector.get_queue(), n, alpha, x, incx, - dependencies); + dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const float *a, int64_t lda, float *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const float* a, int64_t lda, float* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const double *a, int64_t lda, double *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const double* a, int64_t lda, double* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trmv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const float *a, float *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const float* a, float* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const double *a, double *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const double* a, double* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, std::complex *x, - int64_t incx, const std::vector &dependencies) { + diag unit_diag, int64_t n, const std::complex* a, std::complex* x, + int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, std::complex *x, - int64_t incx, const std::vector &dependencies) { + diag unit_diag, int64_t n, const std::complex* a, std::complex* x, + int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tpmv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, int64_t n, - float alpha, const float *x, int64_t incx, float *a, - const std::vector &dependencies) { + float alpha, const float* x, int64_t incx, float* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event spr(backend_selector selector, uplo upper_lower, int64_t n, - double alpha, const double *x, int64_t incx, double *a, - const std::vector &dependencies) { + double alpha, const double* x, int64_t incx, double* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::spr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + std::complex alpha, const std::complex* a, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::hpmv( + selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } sycl::event hpmv(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::hpmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + std::complex alpha, const std::complex* a, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::hpmv( + selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, float alpha, const float *a, int64_t lda, float beta, - float *c, int64_t ldc, const std::vector &dependencies) { + int64_t n, int64_t k, float alpha, const float* a, int64_t lda, float beta, + float* c, int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, double alpha, const double *a, int64_t lda, double beta, - double *c, int64_t ldc, const std::vector &dependencies) { + int64_t n, int64_t k, double alpha, const double* a, int64_t lda, double beta, + double* c, int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event syrk(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, int64_t *n, int64_t *k, float *alpha, const float **a, - int64_t *lda, float *beta, float **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, int64_t* n, int64_t* k, float* alpha, const float** a, + int64_t* lda, float* beta, float** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, int64_t *n, int64_t *k, double *alpha, const double **a, - int64_t *lda, double *beta, double **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, int64_t* n, int64_t* k, double* alpha, const double** a, + int64_t* lda, double* beta, double** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, int64_t *n, int64_t *k, std::complex *alpha, - const std::complex **a, int64_t *lda, std::complex *beta, - std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, int64_t* n, int64_t* k, std::complex* alpha, + const std::complex** a, int64_t* lda, std::complex* beta, + std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event syrk_batch(backend_selector selector, uplo *upper_lower, - transpose *trans, int64_t *n, int64_t *k, std::complex *alpha, - const std::complex **a, int64_t *lda, std::complex *beta, - std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event syrk_batch(backend_selector selector, uplo* upper_lower, + transpose* trans, int64_t* n, int64_t* k, std::complex* alpha, + const std::complex** a, int64_t* lda, std::complex* beta, + std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); @@ -1919,10 +1923,10 @@ sycl::event syrk_batch(backend_selector selector, uplo *upper_ } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, int64_t n, int64_t k, float alpha, const float *a, - int64_t lda, int64_t stride_a, float beta, float *c, int64_t ldc, + transpose trans, int64_t n, int64_t k, float alpha, const float* a, + int64_t lda, int64_t stride_a, float beta, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -1930,10 +1934,10 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_l } sycl::event syrk_batch(backend_selector selector, uplo upper_lower, - transpose trans, int64_t n, int64_t k, double alpha, const double *a, - int64_t lda, int64_t stride_a, double beta, double *c, int64_t ldc, + transpose trans, int64_t n, int64_t k, double alpha, const double* a, + int64_t lda, int64_t stride_a, double beta, double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -1942,10 +1946,10 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_l sycl::event syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - std::complex beta, std::complex *c, int64_t ldc, + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -1954,10 +1958,10 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_l sycl::event syrk_batch(backend_selector selector, uplo upper_lower, transpose trans, int64_t n, int64_t k, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - std::complex beta, std::complex *c, int64_t ldc, + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syrk_batch( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); @@ -1965,321 +1969,321 @@ sycl::event syrk_batch(backend_selector selector, uplo upper_l } sycl::event her2(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + std::complex alpha, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::her2( + selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } sycl::event her2(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::her2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, lda, dependencies); + std::complex alpha, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::her2( + selector.get_queue(), upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { + int64_t k, std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hbmv(backend_selector selector, uplo upper_lower, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { + int64_t k, std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::hbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } -sycl::event rot(backend_selector selector, int64_t n, std::complex *x, - int64_t incx, std::complex *y, int64_t incy, float c, float s, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); +sycl::event rot(backend_selector selector, int64_t n, std::complex* x, + int64_t incx, std::complex* y, int64_t incy, float c, float s, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, + c, s, dependencies); return done; } -sycl::event rot(backend_selector selector, int64_t n, std::complex *x, - int64_t incx, std::complex *y, int64_t incy, double c, double s, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); +sycl::event rot(backend_selector selector, int64_t n, std::complex* x, + int64_t incx, std::complex* y, int64_t incy, double c, double s, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, + c, s, dependencies); return done; } -sycl::event rot(backend_selector selector, int64_t n, float *x, int64_t incx, - float *y, int64_t incy, float c, float s, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); +sycl::event rot(backend_selector selector, int64_t n, float* x, int64_t incx, + float* y, int64_t incy, float c, float s, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, + c, s, dependencies); return done; } -sycl::event rot(backend_selector selector, int64_t n, double *x, int64_t incx, - double *y, int64_t incy, double c, double s, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, c, - s, dependencies); +sycl::event rot(backend_selector selector, int64_t n, double* x, int64_t incx, + double* y, int64_t incy, double c, double s, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::rot(selector.get_queue(), n, x, incx, y, incy, + c, s, dependencies); return done; } sycl::event axpy(backend_selector selector, int64_t n, float alpha, - const float *x, int64_t incx, float *y, int64_t incy, - const std::vector &dependencies) { + const float* x, int64_t incx, float* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, int64_t n, double alpha, - const double *x, int64_t incx, double *y, int64_t incy, - const std::vector &dependencies) { + const double* x, int64_t incx, double* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex *y, int64_t incy, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } sycl::event axpy(backend_selector selector, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex *y, int64_t incy, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy(selector.get_queue(), n, alpha, x, incx, y, - incy, dependencies); + incy, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, int64_t *n, float *alpha, - const float **x, int64_t *incx, float **y, int64_t *incy, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, int64_t* n, float* alpha, + const float** x, int64_t* incx, float** y, int64_t* incy, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, int64_t *n, double *alpha, - const double **x, int64_t *incx, double **y, int64_t *incy, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, int64_t* n, double* alpha, + const double** x, int64_t* incx, double** y, int64_t* incy, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, int64_t *n, - std::complex *alpha, const std::complex **x, int64_t *incx, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, int64_t* n, + std::complex* alpha, const std::complex** x, int64_t* incx, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event axpy_batch(backend_selector selector, int64_t *n, - std::complex *alpha, const std::complex **x, int64_t *incx, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event axpy_batch(backend_selector selector, int64_t* n, + std::complex* alpha, const std::complex** x, int64_t* incx, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch( selector.get_queue(), n, alpha, x, incx, y, incy, group_count, group_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, int64_t n, float alpha, - const float *x, int64_t incx, int64_t stridex, float *y, int64_t incy, + const float* x, int64_t incx, int64_t stridex, float* y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, int64_t n, double alpha, - const double *x, int64_t incx, int64_t stridex, double *y, int64_t incy, + const double* x, int64_t incx, int64_t stridex, double* y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, int64_t n, - std::complex alpha, const std::complex *x, int64_t incx, - int64_t stridex, std::complex *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, int64_t incx, + int64_t stridex, std::complex* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpy_batch(backend_selector selector, int64_t n, - std::complex alpha, const std::complex *x, int64_t incx, - int64_t stridex, std::complex *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { + std::complex alpha, const std::complex* x, int64_t incx, + int64_t stridex, std::complex* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpy_batch(selector.get_queue(), n, alpha, x, - incx, stridex, y, incy, stridey, - batch_size, dependencies); + incx, stridex, y, incy, stridey, + batch_size, dependencies); return done; } sycl::event axpby(backend_selector selector, int64_t n, float alpha, - const float *x, int64_t incx, const float beta, float *y, int64_t incy, - const std::vector &dependencies) { + const float* x, int64_t incx, const float beta, float* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, int64_t n, double alpha, - const double *x, int64_t incx, const double beta, double *y, int64_t incy, - const std::vector &dependencies) { + const double* x, int64_t incx, const double beta, double* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, const std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event axpby(backend_selector selector, int64_t n, - std::complex alpha, const std::complex *x, int64_t incx, - const std::complex beta, std::complex *y, int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, int64_t incx, + const std::complex beta, std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::axpby(selector.get_queue(), n, alpha, x, incx, - beta, y, incy, dependencies); + beta, y, incy, dependencies); return done; } sycl::event gerc(backend_selector selector, int64_t m, int64_t n, - std::complex alpha, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event gerc(backend_selector selector, int64_t m, int64_t n, - std::complex alpha, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gerc(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, float alpha, const float *a, int64_t lda, const float *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t n, int64_t k, float alpha, const float* a, int64_t lda, const float* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, double alpha, const double *a, int64_t lda, const double *b, - int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t n, int64_t k, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event syr2k(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::syr2k(selector.get_queue(), upper_lower, trans, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, int64_t m, int64_t n, - float alpha, const float *a, int64_t lda, const float *x, int64_t incx, float beta, - float *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + float alpha, const float* a, int64_t lda, const float* x, int64_t incx, float beta, + float* y, int64_t incy, const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::gemv( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, int64_t m, int64_t n, - double alpha, const double *a, int64_t lda, const double *x, int64_t incx, - double beta, double *y, int64_t incy, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + double alpha, const double* a, int64_t lda, const double* x, int64_t incx, + double beta, double* y, int64_t incy, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::gemv( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::gemv( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv(backend_selector selector, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::gemv(selector.get_queue(), trans, m, n, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::gemv( + selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemv_batch(backend_selector selector, transpose trans, int64_t m, - int64_t n, float alpha, const float *a, int64_t lda, int64_t stridea, - const float *x, int64_t incx, int64_t stridex, float beta, float *y, + int64_t n, float alpha, const float* a, int64_t lda, int64_t stridea, + const float* x, int64_t incx, int64_t stridex, float beta, float* y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); @@ -2287,10 +2291,10 @@ sycl::event gemv_batch(backend_selector selector, transpose tr } sycl::event gemv_batch(backend_selector selector, transpose trans, int64_t m, - int64_t n, double alpha, const double *a, int64_t lda, int64_t stridea, - const double *x, int64_t incx, int64_t stridex, double beta, double *y, + int64_t n, double alpha, const double* a, int64_t lda, int64_t stridea, + const double* x, int64_t incx, int64_t stridex, double beta, double* y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); @@ -2298,11 +2302,11 @@ sycl::event gemv_batch(backend_selector selector, transpose tr } sycl::event gemv_batch(backend_selector selector, transpose trans, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, - int64_t lda, int64_t stridea, const std::complex *x, int64_t incx, - int64_t stridex, std::complex beta, std::complex *y, + int64_t n, std::complex alpha, const std::complex* a, + int64_t lda, int64_t stridea, const std::complex* x, int64_t incx, + int64_t stridex, std::complex beta, std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); @@ -2310,55 +2314,55 @@ sycl::event gemv_batch(backend_selector selector, transpose tr } sycl::event gemv_batch(backend_selector selector, transpose trans, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, - int64_t lda, int64_t stridea, const std::complex *x, int64_t incx, - int64_t stridex, std::complex beta, std::complex *y, + int64_t n, std::complex alpha, const std::complex* a, + int64_t lda, int64_t stridea, const std::complex* x, int64_t incx, + int64_t stridex, std::complex beta, std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, int64_t *m, - int64_t *n, float *alpha, const float **a, int64_t *lda, const float **x, - int64_t *incx, float *beta, float **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose* trans, int64_t* m, + int64_t* n, float* alpha, const float** a, int64_t* lda, const float** x, + int64_t* incx, float* beta, float** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, int64_t *m, - int64_t *n, double *alpha, const double **a, int64_t *lda, const double **x, - int64_t *incx, double *beta, double **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose* trans, int64_t* m, + int64_t* n, double* alpha, const double** a, int64_t* lda, const double** x, + int64_t* incx, double* beta, double** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, int64_t *m, - int64_t *n, std::complex *alpha, const std::complex **a, - int64_t *lda, const std::complex **x, int64_t *incx, - std::complex *beta, std::complex **y, int64_t *incy, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose* trans, int64_t* m, + int64_t* n, std::complex* alpha, const std::complex** a, + int64_t* lda, const std::complex** x, int64_t* incx, + std::complex* beta, std::complex** y, int64_t* incy, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); return done; } -sycl::event gemv_batch(backend_selector selector, transpose *trans, int64_t *m, - int64_t *n, std::complex *alpha, const std::complex **a, - int64_t *lda, const std::complex **x, int64_t *incx, - std::complex *beta, std::complex **y, int64_t *incy, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemv_batch(backend_selector selector, transpose* trans, int64_t* m, + int64_t* n, std::complex* alpha, const std::complex** a, + int64_t* lda, const std::complex** x, int64_t* incx, + std::complex* beta, std::complex** y, int64_t* incy, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemv_batch( selector.get_queue(), trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); @@ -2366,9 +2370,9 @@ sycl::event gemv_batch(backend_selector selector, transpose *t } sycl::event dgmm_batch(backend_selector selector, side left_right, int64_t m, - int64_t n, const float *a, int64_t lda, int64_t stridea, const float *x, - int64_t incx, int64_t stridex, float *c, int64_t ldc, int64_t stridec, - int64_t batch_size, const std::vector &dependencies) { + int64_t n, const float* a, int64_t lda, int64_t stridea, const float* x, + int64_t incx, int64_t stridex, float* c, int64_t ldc, int64_t stridec, + int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); @@ -2376,9 +2380,9 @@ sycl::event dgmm_batch(backend_selector selector, side left_ri } sycl::event dgmm_batch(backend_selector selector, side left_right, int64_t m, - int64_t n, const double *a, int64_t lda, int64_t stridea, const double *x, - int64_t incx, int64_t stridex, double *c, int64_t ldc, int64_t stridec, - int64_t batch_size, const std::vector &dependencies) { + int64_t n, const double* a, int64_t lda, int64_t stridea, const double* x, + int64_t incx, int64_t stridex, double* c, int64_t ldc, int64_t stridec, + int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); @@ -2386,10 +2390,10 @@ sycl::event dgmm_batch(backend_selector selector, side left_ri } sycl::event dgmm_batch(backend_selector selector, side left_right, int64_t m, - int64_t n, const std::complex *a, int64_t lda, int64_t stridea, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *c, int64_t ldc, int64_t stridec, int64_t batch_size, - const std::vector &dependencies) { + int64_t n, const std::complex* a, int64_t lda, int64_t stridea, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* c, int64_t ldc, int64_t stridec, int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); @@ -2397,52 +2401,52 @@ sycl::event dgmm_batch(backend_selector selector, side left_ri } sycl::event dgmm_batch(backend_selector selector, side left_right, int64_t m, - int64_t n, const std::complex *a, int64_t lda, int64_t stridea, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *c, int64_t ldc, int64_t stridec, int64_t batch_size, - const std::vector &dependencies) { + int64_t n, const std::complex* a, int64_t lda, int64_t stridea, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* c, int64_t ldc, int64_t stridec, int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, int64_t *m, - int64_t *n, const float **a, int64_t *lda, const float **x, int64_t *incx, - float **c, int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side* left_right, int64_t* m, + int64_t* n, const float** a, int64_t* lda, const float** x, int64_t* incx, + float** c, int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, int64_t *m, - int64_t *n, const double **a, int64_t *lda, const double **x, int64_t *incx, - double **c, int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side* left_right, int64_t* m, + int64_t* n, const double** a, int64_t* lda, const double** x, int64_t* incx, + double** c, int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, int64_t *m, - int64_t *n, const std::complex **a, int64_t *lda, - const std::complex **x, int64_t *incx, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side* left_right, int64_t* m, + int64_t* n, const std::complex** a, int64_t* lda, + const std::complex** x, int64_t* incx, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event dgmm_batch(backend_selector selector, side *left_right, int64_t *m, - int64_t *n, const std::complex **a, int64_t *lda, - const std::complex **x, int64_t *incx, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(backend_selector selector, side* left_right, int64_t* m, + int64_t* n, const std::complex** a, int64_t* lda, + const std::complex** x, int64_t* incx, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dgmm_batch( selector.get_queue(), left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); @@ -2450,151 +2454,151 @@ sycl::event dgmm_batch(backend_selector selector, side *left_r } sycl::event her(backend_selector selector, uplo upper_lower, int64_t n, - float alpha, const std::complex *x, int64_t incx, std::complex *a, - int64_t lda, const std::vector &dependencies) { + float alpha, const std::complex* x, int64_t incx, std::complex* a, + int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event her(backend_selector selector, uplo upper_lower, int64_t n, - double alpha, const std::complex *x, int64_t incx, std::complex *a, - int64_t lda, const std::vector &dependencies) { + double alpha, const std::complex* x, int64_t incx, std::complex* a, + int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::her(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, int64_t n, - float alpha, const std::complex *x, int64_t incx, std::complex *a, - const std::vector &dependencies) { + float alpha, const std::complex* x, int64_t incx, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } sycl::event hpr(backend_selector selector, uplo upper_lower, int64_t n, - double alpha, const std::complex *x, int64_t incx, std::complex *a, - const std::vector &dependencies) { + double alpha, const std::complex* x, int64_t incx, std::complex* a, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::hpr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, dependencies); + x, incx, a, dependencies); return done; } -sycl::event iamin(backend_selector selector, int64_t n, const float *x, - int64_t incx, int64_t *result, const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, int64_t n, const float* x, + int64_t incx, int64_t* result, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamin(backend_selector selector, int64_t n, const double *x, - int64_t incx, int64_t *result, const std::vector &dependencies) { +sycl::event iamin(backend_selector selector, int64_t n, const double* x, + int64_t incx, int64_t* result, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, int64_t *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamin(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, int64_t *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::iamin(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, int64_t *m, int64_t *n, int64_t *k, float *alpha, - const float **a, int64_t *lda, const float **b, int64_t *ldb, float *beta, - float **c, int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, int64_t* m, int64_t* n, int64_t* k, float* alpha, + const float** a, int64_t* lda, const float** b, int64_t* ldb, float* beta, + float** c, int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, int64_t *m, int64_t *n, int64_t *k, double *alpha, - const double **a, int64_t *lda, const double **b, int64_t *ldb, double *beta, - double **c, int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, int64_t* m, int64_t* n, int64_t* k, double* alpha, + const double** a, int64_t* lda, const double** b, int64_t* ldb, double* beta, + double** c, int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, int64_t *m, int64_t *n, int64_t *k, - std::complex *alpha, const std::complex **a, int64_t *lda, - const std::complex **b, int64_t *ldb, std::complex *beta, - std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, int64_t* m, int64_t* n, int64_t* k, + std::complex* alpha, const std::complex** a, int64_t* lda, + const std::complex** b, int64_t* ldb, std::complex* beta, + std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, int64_t *m, int64_t *n, int64_t *k, - std::complex *alpha, const std::complex **a, int64_t *lda, - const std::complex **b, int64_t *ldb, std::complex *beta, - std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, int64_t* m, int64_t* n, int64_t* k, + std::complex* alpha, const std::complex** a, int64_t* lda, + const std::complex** b, int64_t* ldb, std::complex* beta, + std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, int64_t *m, int64_t *n, int64_t *k, sycl::half *alpha, - const sycl::half **a, int64_t *lda, const sycl::half **b, int64_t *ldb, - sycl::half *beta, sycl::half **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, int64_t* m, int64_t* n, int64_t* k, sycl::half* alpha, + const sycl::half** a, int64_t* lda, const sycl::half** b, int64_t* ldb, + sycl::half* beta, sycl::half** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, int64_t *m, int64_t *n, int64_t *k, float *alpha, - const sycl::half **a, int64_t *lda, const sycl::half **b, int64_t *ldb, - float *beta, float **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, int64_t* m, int64_t* n, int64_t* k, float* alpha, + const sycl::half** a, int64_t* lda, const sycl::half** b, int64_t* ldb, + float* beta, float** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, int64_t *m, int64_t *n, int64_t *k, float *alpha, - const std::int8_t **a, int64_t *lda, const std::int8_t **b, int64_t *ldb, - float *beta, float **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, int64_t* m, int64_t* n, int64_t* k, float* alpha, + const std::int8_t** a, int64_t* lda, const std::int8_t** b, int64_t* ldb, + float* beta, float** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); return done; } -sycl::event gemm_batch(backend_selector selector, transpose *transa, - transpose *transb, int64_t *m, int64_t *n, int64_t *k, float *alpha, - const std::int8_t **a, int64_t *lda, const std::int8_t **b, int64_t *ldb, - float *beta, std::int32_t **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event gemm_batch(backend_selector selector, transpose* transa, + transpose* transb, int64_t* m, int64_t* n, int64_t* k, float* alpha, + const std::int8_t** a, int64_t* lda, const std::int8_t** b, int64_t* ldb, + float* beta, std::int32_t** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); @@ -2603,9 +2607,9 @@ sycl::event gemm_batch(backend_selector selector, transpose *t sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, float alpha, - const float *a, int64_t lda, int64_t stride_a, const float *b, int64_t ldb, - int64_t stride_b, float beta, float *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { + const float* a, int64_t lda, int64_t stride_a, const float* b, int64_t ldb, + int64_t stride_b, float beta, float* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2614,9 +2618,9 @@ sycl::event gemm_batch(backend_selector selector, transpose tr sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, double alpha, - const double *a, int64_t lda, int64_t stride_a, const double *b, int64_t ldb, - int64_t stride_b, double beta, double *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { + const double* a, int64_t lda, int64_t stride_a, const double* b, int64_t ldb, + int64_t stride_b, double beta, double* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2625,11 +2629,11 @@ sycl::event gemm_batch(backend_selector selector, transpose tr sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex beta, std::complex *c, int64_t ldc, + const std::complex* a, int64_t lda, int64_t stride_a, + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2638,11 +2642,11 @@ sycl::event gemm_batch(backend_selector selector, transpose tr sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, const std::complex *b, int64_t ldb, - int64_t stride_b, std::complex beta, std::complex *c, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, const std::complex* b, int64_t ldb, + int64_t stride_b, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2651,10 +2655,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tr sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, sycl::half alpha, - const sycl::half *a, int64_t lda, int64_t stride_a, const sycl::half *b, - int64_t ldb, int64_t stride_b, sycl::half beta, sycl::half *c, int64_t ldc, + const sycl::half* a, int64_t lda, int64_t stride_a, const sycl::half* b, + int64_t ldb, int64_t stride_b, sycl::half beta, sycl::half* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2663,10 +2667,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tr sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, float alpha, - const sycl::half *a, int64_t lda, int64_t stride_a, const sycl::half *b, - int64_t ldb, int64_t stride_b, float beta, float *c, int64_t ldc, + const sycl::half* a, int64_t lda, int64_t stride_a, const sycl::half* b, + int64_t ldb, int64_t stride_b, float beta, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2675,10 +2679,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tr sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, float alpha, - const std::int8_t *a, int64_t lda, int64_t stride_a, const std::int8_t *b, - int64_t ldb, int64_t stride_b, float beta, float *c, int64_t ldc, + const std::int8_t* a, int64_t lda, int64_t stride_a, const std::int8_t* b, + int64_t ldb, int64_t stride_b, float beta, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2687,10 +2691,10 @@ sycl::event gemm_batch(backend_selector selector, transpose tr sycl::event gemm_batch(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, float alpha, - const std::int8_t *a, int64_t lda, int64_t stride_a, const std::int8_t *b, - int64_t ldb, int64_t stride_b, float beta, std::int32_t *c, int64_t ldc, + const std::int8_t* a, int64_t lda, int64_t stride_a, const std::int8_t* b, + int64_t ldb, int64_t stride_b, float beta, std::int32_t* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_batch( selector.get_queue(), transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); @@ -2698,177 +2702,177 @@ sycl::event gemm_batch(backend_selector selector, transpose tr } sycl::event spmv(backend_selector selector, uplo upper_lower, int64_t n, - float alpha, const float *a, const float *x, int64_t incx, float beta, float *y, - int64_t incy, const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + float alpha, const float* a, const float* x, int64_t incx, float beta, float* y, + int64_t incy, const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::spmv( + selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } sycl::event spmv(backend_selector selector, uplo upper_lower, int64_t n, - double alpha, const double *a, const double *x, int64_t incx, double beta, - double *y, int64_t incy, const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::spmv(selector.get_queue(), upper_lower, n, alpha, - a, x, incx, beta, y, incy, dependencies); + double alpha, const double* a, const double* x, int64_t incx, double beta, + double* y, int64_t incy, const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::spmv( + selector.get_queue(), upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); return done; } -sycl::event swap(backend_selector selector, int64_t n, float *x, int64_t incx, - float *y, int64_t incy, const std::vector &dependencies) { +sycl::event swap(backend_selector selector, int64_t n, float* x, int64_t incx, + float* y, int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, int64_t n, double *x, int64_t incx, - double *y, int64_t incy, const std::vector &dependencies) { +sycl::event swap(backend_selector selector, int64_t n, double* x, int64_t incx, + double* y, int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, int64_t n, std::complex *x, - int64_t incx, std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, int64_t n, std::complex* x, + int64_t incx, std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event swap(backend_selector selector, int64_t n, std::complex *x, - int64_t incx, std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event swap(backend_selector selector, int64_t n, std::complex* x, + int64_t incx, std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::swap(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event geru(backend_selector selector, int64_t m, int64_t n, - std::complex alpha, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event geru(backend_selector selector, int64_t m, int64_t n, - std::complex alpha, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { + std::complex alpha, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::geru(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event nrm2(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event nrm2(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, int64_t n, const float *x, - int64_t incx, float *result, const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, int64_t n, const float* x, + int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event nrm2(backend_selector selector, int64_t n, const double *x, - int64_t incx, double *result, const std::vector &dependencies) { +sycl::event nrm2(backend_selector selector, int64_t n, const double* x, + int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::nrm2(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, float alpha, const float *a, int64_t lda, - const float *b, int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { + int64_t m, int64_t n, int64_t k, float alpha, const float* a, int64_t lda, + const float* b, int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, double alpha, const double *a, int64_t lda, - const double *b, int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { + int64_t m, int64_t n, int64_t k, double alpha, const double* a, int64_t lda, + const double* b, int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, std::complex alpha, - const std::complex *a, int64_t lda, const std::complex *b, - int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, const std::complex* b, + int64_t ldb, std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, std::complex alpha, - const std::complex *a, int64_t lda, const std::complex *b, - int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, const std::complex* b, + int64_t ldb, std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, sycl::half alpha, const sycl::half *a, - int64_t lda, const sycl::half *b, int64_t ldb, sycl::half beta, sycl::half *c, - int64_t ldc, const std::vector &dependencies) { + int64_t m, int64_t n, int64_t k, sycl::half alpha, const sycl::half* a, + int64_t lda, const sycl::half* b, int64_t ldb, sycl::half beta, sycl::half* c, + int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, float alpha, const sycl::half *a, int64_t lda, - const sycl::half *b, int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { + int64_t m, int64_t n, int64_t k, float alpha, const sycl::half* a, int64_t lda, + const sycl::half* b, int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm(backend_selector selector, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, float alpha, const bfloat16 *a, int64_t lda, - const bfloat16 *b, int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { + int64_t m, int64_t n, int64_t k, float alpha, const bfloat16* a, int64_t lda, + const bfloat16* b, int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm(selector.get_queue(), transa, transb, m, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + alpha, a, lda, b, ldb, beta, c, ldc, dependencies); return done; } sycl::event gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, - float alpha, const std::int8_t *a, int64_t lda, std::int8_t ao, - const std::uint8_t *b, int64_t ldb, std::uint8_t bo, float beta, - std::int32_t *c, int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + float alpha, const std::int8_t* a, int64_t lda, std::int8_t ao, + const std::uint8_t* b, int64_t ldb, std::uint8_t bo, float beta, + std::int32_t* c, int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -2877,10 +2881,10 @@ sycl::event gemm_bias(backend_selector selector, transpose tra sycl::event gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, - float alpha, const std::int8_t *a, int64_t lda, std::int8_t ao, - const std::int8_t *b, int64_t ldb, std::int8_t bo, float beta, - std::int32_t *c, int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + float alpha, const std::int8_t* a, int64_t lda, std::int8_t ao, + const std::int8_t* b, int64_t ldb, std::int8_t bo, float beta, + std::int32_t* c, int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -2889,10 +2893,10 @@ sycl::event gemm_bias(backend_selector selector, transpose tra sycl::event gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, - float alpha, const std::uint8_t *a, int64_t lda, std::uint8_t ao, - const std::int8_t *b, int64_t ldb, std::int8_t bo, float beta, - std::int32_t *c, int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + float alpha, const std::uint8_t* a, int64_t lda, std::uint8_t ao, + const std::int8_t* b, int64_t ldb, std::int8_t bo, float beta, + std::int32_t* c, int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -2901,10 +2905,10 @@ sycl::event gemm_bias(backend_selector selector, transpose tra sycl::event gemm_bias(backend_selector selector, transpose transa, transpose transb, offset offsetc, int64_t m, int64_t n, int64_t k, - float alpha, const std::uint8_t *a, int64_t lda, std::uint8_t ao, - const std::uint8_t *b, int64_t ldb, std::uint8_t bo, float beta, - std::int32_t *c, int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + float alpha, const std::uint8_t* a, int64_t lda, std::uint8_t ao, + const std::uint8_t* b, int64_t ldb, std::uint8_t bo, float beta, + std::int32_t* c, int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemm_bias( selector.get_queue(), transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); @@ -2912,83 +2916,83 @@ sycl::event gemm_bias(backend_selector selector, transpose tra } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, float alpha, const std::complex *a, int64_t lda, - float beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { + int64_t n, int64_t k, float alpha, const std::complex* a, int64_t lda, + float beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event herk(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, double alpha, const std::complex *a, int64_t lda, - double beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { + int64_t n, int64_t k, double alpha, const std::complex* a, int64_t lda, + double beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::herk( selector.get_queue(), upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); return done; } sycl::event ger(backend_selector selector, int64_t m, int64_t n, float alpha, - const float *x, int64_t incx, const float *y, int64_t incy, float *a, int64_t lda, - const std::vector &dependencies) { + const float* x, int64_t incx, const float* y, int64_t incy, float* a, int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event ger(backend_selector selector, int64_t m, int64_t n, double alpha, - const double *x, int64_t incx, const double *y, int64_t incy, double *a, - int64_t lda, const std::vector &dependencies) { + const double* x, int64_t incx, const double* y, int64_t incy, double* a, + int64_t lda, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::ger(selector.get_queue(), m, n, alpha, x, incx, - y, incy, a, lda, dependencies); + y, incy, a, lda, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, - int64_t lda, float *b, int64_t ldb, const std::vector &dependencies) { + transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, const float* a, + int64_t lda, float* b, int64_t ldb, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, double *b, int64_t ldb, - const std::vector &dependencies) { + const double* a, int64_t lda, double* b, int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, - float alpha, const float *a, int64_t lda, int64_t stride_a, float *b, + float alpha, const float* a, int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -2997,9 +3001,9 @@ sycl::event trsm_batch(backend_selector selector, side left_ri sycl::event trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, - double alpha, const double *a, int64_t lda, int64_t stride_a, double *b, + double alpha, const double* a, int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3008,9 +3012,9 @@ sycl::event trsm_batch(backend_selector selector, side left_ri sycl::event trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3019,53 +3023,53 @@ sycl::event trsm_batch(backend_selector selector, side left_ri sycl::event trsm_batch(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, - float *alpha, const float **a, int64_t *lda, float **b, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, int64_t* m, int64_t* n, + float* alpha, const float** a, int64_t* lda, float** b, int64_t* ldb, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, - double *alpha, const double **a, int64_t *lda, double **b, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, int64_t* m, int64_t* n, + double* alpha, const double** a, int64_t* lda, double** b, int64_t* ldb, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - std::complex **b, int64_t *ldb, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + std::complex** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); return done; } -sycl::event trsm_batch(backend_selector selector, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - std::complex **b, int64_t *ldb, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event trsm_batch(backend_selector selector, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + std::complex** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsm_batch( selector.get_queue(), left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); @@ -3073,747 +3077,747 @@ sycl::event trsm_batch(backend_selector selector, side *left_r } sycl::event dotu(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotu(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dotu(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, - int64_t m, int64_t n, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t m, int64_t n, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event hemm(backend_selector selector, side left_right, uplo upper_lower, - int64_t m, int64_t n, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t m, int64_t n, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::hemm(selector.get_queue(), left_right, + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *a, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + std::complex alpha, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* a, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, + alpha, x, incx, y, incy, a, dependencies); return done; } sycl::event hpr2(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *a, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + std::complex alpha, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* a, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::hpr2(selector.get_queue(), upper_lower, n, + alpha, x, incx, y, incy, a, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, int64_t m, int64_t n, - int64_t kl, int64_t ku, float alpha, const float *a, int64_t lda, const float *x, - int64_t incx, float beta, float *y, int64_t incy, - const std::vector &dependencies) { + int64_t kl, int64_t ku, float alpha, const float* a, int64_t lda, const float* x, + int64_t incx, float beta, float* y, int64_t incy, + const std::vector& dependencies) { auto done = - oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, + a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, int64_t m, int64_t n, - int64_t kl, int64_t ku, double alpha, const double *a, int64_t lda, - const double *x, int64_t incx, double beta, double *y, int64_t incy, - const std::vector &dependencies) { + int64_t kl, int64_t ku, double alpha, const double* a, int64_t lda, + const double* x, int64_t incx, double beta, double* y, int64_t incy, + const std::vector& dependencies) { auto done = - oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, + a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, int64_t m, int64_t n, - int64_t kl, int64_t ku, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { + int64_t kl, int64_t ku, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = - oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, + a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gbmv(backend_selector selector, transpose trans, int64_t m, int64_t n, - int64_t kl, int64_t ku, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *x, int64_t incx, - std::complex beta, std::complex *y, int64_t incy, - const std::vector &dependencies) { + int64_t kl, int64_t ku, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* x, int64_t incx, + std::complex beta, std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = - oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + oneapi::math::blas::rocblas::MAJOR::gbmv(selector.get_queue(), trans, m, n, kl, ku, alpha, + a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const float *a, int64_t lda, float *x, - int64_t incx, const std::vector &dependencies) { + diag unit_diag, int64_t n, int64_t k, const float* a, int64_t lda, float* x, + int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const double *a, int64_t lda, double *x, - int64_t incx, const std::vector &dependencies) { + diag unit_diag, int64_t n, int64_t k, const double* a, int64_t lda, double* x, + int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, int64_t k, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbmv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, int64_t k, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tbmv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - int64_t m, int64_t n, float alpha, const float *a, int64_t lda, const float *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t m, int64_t n, float alpha, const float* a, int64_t lda, const float* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - int64_t m, int64_t n, double alpha, const double *a, int64_t lda, const double *b, - int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t m, int64_t n, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - int64_t m, int64_t n, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t m, int64_t n, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event symm(backend_selector selector, side left_right, uplo upper_lower, - int64_t m, int64_t n, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, upper_lower, m, n, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t m, int64_t n, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::symm(selector.get_queue(), left_right, + upper_lower, m, n, alpha, a, lda, b, ldb, + beta, c, ldc, dependencies); return done; } sycl::event dotc(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event dotc(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dotc(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, int64_t n, - float alpha, const float *x, int64_t incx, float *a, int64_t lda, - const std::vector &dependencies) { + float alpha, const float* x, int64_t incx, float* a, int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event syr(backend_selector selector, uplo upper_lower, int64_t n, - double alpha, const double *x, int64_t incx, double *a, int64_t lda, - const std::vector &dependencies) { + double alpha, const double* x, int64_t incx, double* a, int64_t lda, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::syr(selector.get_queue(), upper_lower, n, alpha, - x, incx, a, lda, dependencies); + x, incx, a, lda, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, - int64_t lda, float *b, int64_t ldb, const std::vector &dependencies) { + transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, const float* a, + int64_t lda, float* b, int64_t ldb, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, double *b, int64_t ldb, - const std::vector &dependencies) { + const double* a, int64_t lda, double* b, int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } sycl::event trmm(backend_selector selector, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trmm(selector.get_queue(), left_right, - upper_lower, trans, unit_diag, m, n, alpha, - a, lda, b, ldb, dependencies); + upper_lower, trans, unit_diag, m, n, alpha, + a, lda, b, ldb, dependencies); return done; } -sycl::event rotmg(backend_selector selector, float *d1, float *d2, float *x1, - float y1, float *param, const std::vector &dependencies) { +sycl::event rotmg(backend_selector selector, float* d1, float* d2, float* x1, + float y1, float* param, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, - param, dependencies); + param, dependencies); return done; } -sycl::event rotmg(backend_selector selector, double *d1, double *d2, double *x1, - double y1, double *param, const std::vector &dependencies) { +sycl::event rotmg(backend_selector selector, double* d1, double* d2, double* x1, + double y1, double* param, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::rotmg(selector.get_queue(), d1, d2, x1, y1, - param, dependencies); + param, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const float *a, float *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const float* a, float* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const double *a, double *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const double* a, double* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, std::complex *x, - int64_t incx, const std::vector &dependencies) { + diag unit_diag, int64_t n, const std::complex* a, std::complex* x, + int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event tpsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, std::complex *x, - int64_t incx, const std::vector &dependencies) { + diag unit_diag, int64_t n, const std::complex* a, std::complex* x, + int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tpsv(selector.get_queue(), upper_lower, trans, - unit_diag, n, a, x, incx, dependencies); + unit_diag, n, a, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const float *a, int64_t lda, float *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const float* a, int64_t lda, float* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const double *a, int64_t lda, double *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const double* a, int64_t lda, double* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } sycl::event trsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::trsv( selector.get_queue(), upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); return done; } -sycl::event copy(backend_selector selector, int64_t n, const float *x, - int64_t incx, float *y, int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, int64_t n, const float* x, + int64_t incx, float* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy(backend_selector selector, int64_t n, const double *x, - int64_t incx, double *y, int64_t incy, - const std::vector &dependencies) { +sycl::event copy(backend_selector selector, int64_t n, const double* x, + int64_t incx, double* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, std::complex *y, int64_t incy, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } sycl::event copy(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, std::complex *y, int64_t incy, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy(selector.get_queue(), n, x, incx, y, incy, - dependencies); + dependencies); return done; } -sycl::event copy_batch(backend_selector selector, int64_t *n, const float **x, - int64_t *incx, float **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, int64_t* n, const float** x, + int64_t* incx, float** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, int64_t *n, const double **x, - int64_t *incx, double **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, int64_t* n, const double** x, + int64_t* incx, double** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, int64_t *n, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, int64_t* n, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, int64_t *n, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, int64_t* n, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, y, incy, group_count, group_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, int64_t n, const float *x, - int64_t incx, int64_t stridex, float *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, int64_t n, const float* x, + int64_t incx, int64_t stridex, float* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } -sycl::event copy_batch(backend_selector selector, int64_t n, const double *x, - int64_t incx, int64_t stridex, double *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(backend_selector selector, int64_t n, const double* x, + int64_t incx, int64_t stridex, double* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event copy_batch(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::copy_batch( selector.get_queue(), n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event hemv(backend_selector selector, uplo upper_lower, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::hemv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, - transpose transb, int64_t n, int64_t k, float alpha, const float *a, int64_t lda, - const float *b, int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { + transpose transb, int64_t n, int64_t k, float alpha, const float* a, int64_t lda, + const float* b, int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, - transpose transb, int64_t n, int64_t k, double alpha, const double *a, - int64_t lda, const double *b, int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { + transpose transb, int64_t n, int64_t k, double alpha, const double* a, + int64_t lda, const double* b, int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, int64_t n, int64_t k, std::complex alpha, - const std::complex *a, int64_t lda, const std::complex *b, - int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, const std::complex* b, + int64_t ldb, std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } sycl::event gemmt(backend_selector selector, uplo upper_lower, transpose transa, transpose transb, int64_t n, int64_t k, std::complex alpha, - const std::complex *a, int64_t lda, const std::complex *b, - int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, const std::complex* b, + int64_t ldb, std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::gemmt(selector.get_queue(), upper_lower, transa, - transb, n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies); + transb, n, k, alpha, a, lda, b, ldb, beta, + c, ldc, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, int64_t n, - int64_t k, float alpha, const float *a, int64_t lda, const float *x, int64_t incx, - float beta, float *y, int64_t incy, const std::vector &dependencies) { + int64_t k, float alpha, const float* a, int64_t lda, const float* x, int64_t incx, + float beta, float* y, int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event sbmv(backend_selector selector, uplo upper_lower, int64_t n, - int64_t k, double alpha, const double *a, int64_t lda, const double *x, - int64_t incx, double beta, double *y, int64_t incy, - const std::vector &dependencies) { + int64_t k, double alpha, const double* a, int64_t lda, const double* x, + int64_t incx, double beta, double* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::sbmv(selector.get_queue(), upper_lower, n, k, alpha, a, - lda, x, incx, beta, y, incy, dependencies); + lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event asum(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, float *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event asum(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, double *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, int64_t n, const float *x, - int64_t incx, float *result, const std::vector &dependencies) { +sycl::event asum(backend_selector selector, int64_t n, const float* x, + int64_t incx, float* result, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event asum(backend_selector selector, int64_t n, const double *x, - int64_t incx, double *result, const std::vector &dependencies) { +sycl::event asum(backend_selector selector, int64_t n, const double* x, + int64_t incx, double* result, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::asum(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const float *a, int64_t lda, float *x, - int64_t incx, const std::vector &dependencies) { + diag unit_diag, int64_t n, int64_t k, const float* a, int64_t lda, float* x, + int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const double *a, int64_t lda, double *x, - int64_t incx, const std::vector &dependencies) { + diag unit_diag, int64_t n, int64_t k, const double* a, int64_t lda, double* x, + int64_t incx, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, int64_t k, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event tbsv(backend_selector selector, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { + diag unit_diag, int64_t n, int64_t k, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::tbsv( selector.get_queue(), upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, int64_t n, - float alpha, const float *x, int64_t incx, const float *y, int64_t incy, float *a, - const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + float alpha, const float* x, int64_t incx, const float* y, int64_t incy, float* a, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, + alpha, x, incx, y, incy, a, dependencies); return done; } sycl::event spr2(backend_selector selector, uplo upper_lower, int64_t n, - double alpha, const double *x, int64_t incx, const double *y, int64_t incy, - double *a, const std::vector &dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, alpha, - x, incx, y, incy, a, dependencies); + double alpha, const double* x, int64_t incx, const double* y, int64_t incy, + double* a, const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::spr2(selector.get_queue(), upper_lower, n, + alpha, x, incx, y, incy, a, dependencies); return done; } -sycl::event iamax(backend_selector selector, int64_t n, const float *x, - int64_t incx, int64_t *result, const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, int64_t n, const float* x, + int64_t incx, int64_t* result, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event iamax(backend_selector selector, int64_t n, const double *x, - int64_t incx, int64_t *result, const std::vector &dependencies) { +sycl::event iamax(backend_selector selector, int64_t n, const double* x, + int64_t incx, int64_t* result, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, int64_t *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } sycl::event iamax(backend_selector selector, int64_t n, - const std::complex *x, int64_t incx, int64_t *result, - const std::vector &dependencies) { + const std::complex* x, int64_t incx, int64_t* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::iamax(selector.get_queue(), n, x, incx, result, - dependencies); + dependencies); return done; } -sycl::event rotm(backend_selector selector, int64_t n, float *x, int64_t incx, - float *y, int64_t incy, float *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, int64_t n, float* x, int64_t incx, + float* y, int64_t incy, float* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotm(backend_selector selector, int64_t n, double *x, int64_t incx, - double *y, int64_t incy, double *param, - const std::vector &dependencies) { +sycl::event rotm(backend_selector selector, int64_t n, double* x, int64_t incx, + double* y, int64_t incy, double* param, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::rotm(selector.get_queue(), n, x, incx, y, incy, - param, dependencies); + param, dependencies); return done; } -sycl::event rotg(backend_selector selector, float *a, float *b, float *c, - float *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, float* a, float* b, float* c, + float* s, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, double *a, double *b, double *c, - double *s, const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, double* a, double* b, double* c, + double* s, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, float *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, float* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event rotg(backend_selector selector, std::complex *a, - std::complex *b, double *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(backend_selector selector, std::complex* a, + std::complex* b, double* c, std::complex* s, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::rotg(selector.get_queue(), a, b, c, s, dependencies); return done; } -sycl::event sdsdot(backend_selector selector, int64_t n, float sb, const float *x, - int64_t incx, const float *y, int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event sdsdot(backend_selector selector, int64_t n, float sb, const float* x, + int64_t incx, const float* y, int64_t incy, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::sdsdot(selector.get_queue(), n, sb, x, incx, y, - incy, result, dependencies); + incy, result, dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, float beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, float beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } sycl::event her2k(backend_selector selector, uplo upper_lower, transpose trans, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, double beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = - oneapi::math::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, n, k, - alpha, a, lda, b, ldb, beta, c, ldc, dependencies); + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, double beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::her2k(selector.get_queue(), upper_lower, trans, + n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies); return done; } -sycl::event dot(backend_selector selector, int64_t n, const float *x, - int64_t incx, const float *y, int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, int64_t n, const float* x, + int64_t incx, const float* y, int64_t incy, float* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, int64_t n, const double *x, - int64_t incx, const double *y, int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, int64_t n, const double* x, + int64_t incx, const double* y, int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } -sycl::event dot(backend_selector selector, int64_t n, const float *x, - int64_t incx, const float *y, int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(backend_selector selector, int64_t n, const float* x, + int64_t incx, const float* y, int64_t incy, double* result, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::dot(selector.get_queue(), n, x, incx, y, incy, - result, dependencies); + result, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, int64_t n, - float alpha, const float *a, int64_t lda, const float *x, int64_t incx, float beta, - float *y, int64_t incy, const std::vector &dependencies) { + float alpha, const float* a, int64_t lda, const float* x, int64_t incx, float beta, + float* y, int64_t incy, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event symv(backend_selector selector, uplo upper_lower, int64_t n, - double alpha, const double *a, int64_t lda, const double *x, int64_t incx, - double beta, double *y, int64_t incy, - const std::vector &dependencies) { + double alpha, const double* a, int64_t lda, const double* x, int64_t incx, + double beta, double* y, int64_t incy, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::symv( selector.get_queue(), upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); return done; } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3821,10 +3825,10 @@ sycl::event omatcopy_batch(backend_selector selector, transpos } sycl::event omatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3833,9 +3837,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpos sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3844,9 +3848,9 @@ sycl::event omatcopy_batch(backend_selector selector, transpos sycl::event omatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); @@ -3854,18 +3858,18 @@ sycl::event omatcopy_batch(backend_selector selector, transpos } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, + std::int64_t m, std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, + std::int64_t m, std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -3873,9 +3877,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpos sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -3883,9 +3887,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpos sycl::event imatcopy_batch(backend_selector selector, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); return done; @@ -3893,10 +3897,10 @@ sycl::event imatcopy_batch(backend_selector selector, transpos sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, + const float* a, std::int64_t lda, std::int64_t stride_a, float beta, + const float* b, std::int64_t ldb, std::int64_t stride_b, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -3905,10 +3909,10 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + const double* b, std::int64_t ldb, std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -3917,11 +3921,11 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -3930,11 +3934,11 @@ sycl::event omatadd_batch(backend_selector selector, transpose sycl::event omatadd_batch(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatadd_batch( selector.get_queue(), transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); @@ -3942,150 +3946,150 @@ sycl::event omatadd_batch(backend_selector selector, transpose } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, float *b, - std::int64_t ldb, const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, float* b, + std::int64_t ldb, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, dependencies); + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, double *b, - std::int64_t ldb, const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, double* b, + std::int64_t ldb, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, dependencies); + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, dependencies); + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy(selector.get_queue(), trans, m, n, - alpha, a, lda, b, ldb, dependencies); + alpha, a, lda, b, ldb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, double* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event omatcopy2(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy2( selector.get_queue(), trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, float alpha, float *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, dependencies); + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, double alpha, double *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, dependencies); + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, dependencies); + alpha, ab, lda, ldb, dependencies); return done; } sycl::event imatcopy(backend_selector selector, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, dependencies); + alpha, ab, lda, ldb, dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - float beta, const float *b, std::int64_t ldb, float *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + float beta, const float* b, std::int64_t ldb, float* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, - n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, - double beta, const double *b, std::int64_t ldb, double *c, std::int64_t ldc, - const std::vector &dependencies) { + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + double beta, const double* b, std::int64_t ldb, double* c, std::int64_t ldc, + const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, - n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, - n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } sycl::event omatadd(backend_selector selector, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatadd(selector.get_queue(), transa, transb, m, - n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies); + n, alpha, a, lda, beta, b, ldb, c, ldc, + dependencies); return done; } @@ -4115,8 +4119,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpos std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); @@ -4127,8 +4130,7 @@ sycl::event omatcopy_batch(backend_selector selector, transpos std::int64_t* m, std::int64_t* n, std::complex* alpha, const std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { + std::int64_t* groupsize, const std::vector& dependencies) { auto done = oneapi::math::blas::rocblas::MAJOR::omatcopy_batch( selector.get_queue(), trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); @@ -4138,22 +4140,20 @@ sycl::event omatcopy_batch(backend_selector selector, transpos sycl::event imatcopy_batch(backend_selector selector, transpose* trans, std::int64_t* m, std::int64_t* n, float* alpha, float** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + std::int64_t* groupsize, const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( + selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, + dependencies); return done; } sycl::event imatcopy_batch(backend_selector selector, transpose* trans, std::int64_t* m, std::int64_t* n, double* alpha, double** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* groupsize, - const std::vector& dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + std::int64_t* groupsize, const std::vector& dependencies) { + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( + selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, + dependencies); return done; } @@ -4162,9 +4162,9 @@ sycl::event imatcopy_batch(backend_selector selector, transpos std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( + selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, + dependencies); return done; } @@ -4173,8 +4173,8 @@ sycl::event imatcopy_batch(backend_selector selector, transpos std::complex** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, const std::vector& dependencies) { - auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch(selector.get_queue(), trans, m, n, - alpha, ab, lda, ldb, group_count, - groupsize, dependencies); + auto done = oneapi::math::blas::rocblas::MAJOR::imatcopy_batch( + selector.get_queue(), trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, + dependencies); return done; } diff --git a/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hxx b/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hxx index 70aabaaf9..f6c3eeee5 100644 --- a/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hxx +++ b/include/oneapi/math/blas/detail/rocblas/onemath_blas_rocblas.hxx @@ -21,2102 +21,2102 @@ // Buffer APIs -void asum(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result); +void asum(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result); -void asum(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result); +void asum(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result); -void asum(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result); +void asum(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result); -void asum(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result); +void asum(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result); -void axpy(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy); +void axpy(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy); -void axpy(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy); +void axpy(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy); -void axpy(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy); +void axpy(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy); -void axpy(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy); +void axpy(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy); -void axpy_batch(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size); -void axpy_batch(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, +void axpy_batch(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size); -void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size); -void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size); -void axpby(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - float beta, sycl::buffer &y, int64_t incy); +void axpby(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + float beta, sycl::buffer& y, int64_t incy); -void axpby(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx, - double beta, sycl::buffer &y, int64_t incy); +void axpby(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, int64_t incx, + double beta, sycl::buffer& y, int64_t incy); -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy); +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy); -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy); +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy); -void copy(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy); +void copy(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy); -void copy(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy); +void copy(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy); -void copy(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy); +void copy(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy); -void copy(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy); +void copy(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy); -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size); -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size); -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, - int64_t incx, int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, + int64_t incx, int64_t stridex, sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size); -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, - int64_t incx, int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, + int64_t incx, int64_t stridex, sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size); -void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result); +void dot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result); -void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result); +void dot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result); -void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result); +void dot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result); -void dotc(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result); +void dotc(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result); -void dotc(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result); +void dotc(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result); -void dotu(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result); +void dotu(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result); -void dotu(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result); +void dotu(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result); -void iamin(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result); +void iamin(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result); -void iamin(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result); +void iamin(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result); -void iamin(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result); +void iamin(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result); -void iamin(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result); +void iamin(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result); -void iamax(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result); +void iamax(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result); -void iamax(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result); +void iamax(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result); -void iamax(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result); +void iamax(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result); -void iamax(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result); +void iamax(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result); -void nrm2(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result); +void nrm2(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result); -void nrm2(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result); +void nrm2(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result); -void nrm2(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result); +void nrm2(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result); -void nrm2(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result); +void nrm2(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result); -void rot(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, float c, float s); +void rot(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, float c, float s); -void rot(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, double c, double s); +void rot(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, double c, double s); -void rot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, float c, float s); +void rot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, float c, float s); -void rot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, double c, double s); +void rot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, double c, double s); -void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s); +void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s); -void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s); +void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s); -void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s); +void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); -void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s); +void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); -void rotm(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer ¶m); +void rotm(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& param); -void rotm(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer ¶m); +void rotm(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& param); -void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, float y1, sycl::buffer ¶m); +void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, float y1, sycl::buffer& param); -void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, double y1, sycl::buffer ¶m); +void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, double y1, sycl::buffer& param); -void scal(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx); +void scal(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx); -void scal(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx); +void scal(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, int64_t incx); -void scal(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx); +void scal(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx); -void scal(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx); +void scal(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx); -void scal(sycl::queue &queue, int64_t n, float alpha, sycl::buffer, 1> &x, +void scal(sycl::queue& queue, int64_t n, float alpha, sycl::buffer, 1>& x, int64_t incx); -void scal(sycl::queue &queue, int64_t n, double alpha, sycl::buffer, 1> &x, +void scal(sycl::queue& queue, int64_t n, double alpha, sycl::buffer, 1>& x, int64_t incx); -void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result); +void sdsdot(sycl::queue& queue, int64_t n, float sb, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result); -void swap(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy); +void swap(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy); -void swap(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy); +void swap(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy); -void swap(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy); +void swap(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy); -void swap(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy); +void swap(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, float beta, sycl::buffer &y, int64_t incy); +void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, float beta, sycl::buffer& y, int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, double beta, sycl::buffer &y, int64_t incy); +void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + double alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, double beta, sycl::buffer& y, int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy); +void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy); +void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy); -void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - float beta, sycl::buffer &y, int64_t incy); +void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + float beta, sycl::buffer& y, int64_t incy); -void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - double beta, sycl::buffer &y, int64_t incy); +void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + double beta, sycl::buffer& y, int64_t incy); -void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy); +void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy); -void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy); +void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy); -void gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &x, - int64_t incx, int64_t stridex, float beta, sycl::buffer &y, int64_t incy, +void gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, int64_t stridea, sycl::buffer& x, + int64_t incx, int64_t stridex, float beta, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size); -void gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stridea, - sycl::buffer &x, int64_t incx, int64_t stridex, double beta, - sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size); +void gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, int64_t stridea, + sycl::buffer& x, int64_t incx, int64_t stridex, double beta, + sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size); -void gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stridea, sycl::buffer, 1> &x, int64_t incx, - int64_t stridex, std::complex beta, sycl::buffer, 1> &y, +void gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stridea, sycl::buffer, 1>& x, int64_t incx, + int64_t stridex, std::complex beta, sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size); -void gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stridea, sycl::buffer, 1> &x, int64_t incx, +void gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stridea, sycl::buffer, 1>& x, int64_t incx, int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size); -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &c, int64_t ldc, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stridea, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size); -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stridea, - sycl::buffer &x, int64_t incx, int64_t stridex, - sycl::buffer &c, int64_t ldc, int64_t stridec, int64_t batch_size); +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stridea, + sycl::buffer& x, int64_t incx, int64_t stridex, + sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size); -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stridea, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &c, int64_t ldc, int64_t stridec, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer, 1>& a, int64_t lda, int64_t stridea, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& c, int64_t ldc, int64_t stridec, int64_t batch_size); -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stridea, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &c, int64_t ldc, int64_t stridec, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer, 1>& a, int64_t lda, int64_t stridea, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& c, int64_t ldc, int64_t stridec, int64_t batch_size); -void ger(sycl::queue &queue, int64_t m, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, +void ger(sycl::queue& queue, int64_t m, int64_t n, float alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, int64_t lda); -void ger(sycl::queue &queue, int64_t m, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, +void ger(sycl::queue& queue, int64_t m, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, int64_t lda); -void gerc(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda); - -void gerc(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda); - -void geru(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda); - -void geru(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda); - -void hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy); - -void hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy); - -void hemv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy); - -void hemv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy); - -void her(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &a, int64_t lda); - -void her(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &a, int64_t lda); - -void her2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda); - -void her2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda); - -void hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, - int64_t incx, std::complex beta, sycl::buffer, 1> &y, +void gerc(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda); + +void gerc(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda); + +void geru(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda); + +void geru(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda); + +void hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy); + +void hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy); + +void hemv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy); + +void hemv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy); + +void her(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& a, int64_t lda); + +void her(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& a, int64_t lda); + +void her2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda); + +void her2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda); + +void hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, + int64_t incx, std::complex beta, sycl::buffer, 1>& y, int64_t incy); -void hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, - int64_t incx, std::complex beta, sycl::buffer, 1> &y, +void hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, + int64_t incx, std::complex beta, sycl::buffer, 1>& y, int64_t incy); -void hpr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &a); +void hpr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& a); -void hpr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &a); +void hpr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& a); -void hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a); +void hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a); -void hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a); +void hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a); -void sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - float beta, sycl::buffer &y, int64_t incy); +void sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + float beta, sycl::buffer& y, int64_t incy); -void sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - double beta, sycl::buffer &y, int64_t incy); +void sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, double alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + double beta, sycl::buffer& y, int64_t incy); -void spmv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &a, - sycl::buffer &x, int64_t incx, float beta, sycl::buffer &y, +void spmv(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& a, + sycl::buffer& x, int64_t incx, float beta, sycl::buffer& y, int64_t incy); -void spmv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer &a, - sycl::buffer &x, int64_t incx, double beta, sycl::buffer &y, +void spmv(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer& a, + sycl::buffer& x, int64_t incx, double beta, sycl::buffer& y, int64_t incy); -void spr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &a); +void spr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& a); -void spr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &a); +void spr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& a); -void spr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a); +void spr2(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a); -void spr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a); +void spr2(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a); -void symv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &x, int64_t incx, float beta, - sycl::buffer &y, int64_t incy); +void symv(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& x, int64_t incx, float beta, + sycl::buffer& y, int64_t incy); -void symv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &x, int64_t incx, double beta, - sycl::buffer &y, int64_t incy); +void symv(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& x, int64_t incx, double beta, + sycl::buffer& y, int64_t incy); -void syr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &a, int64_t lda); +void syr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& a, int64_t lda); -void syr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &a, int64_t lda); +void syr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& a, int64_t lda); -void syr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, +void syr2(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, int64_t lda); -void syr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, +void syr2(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, int64_t lda); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx); +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx); +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx); +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx); +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &a, sycl::buffer &x, int64_t incx); +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& a, sycl::buffer& x, int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &a, sycl::buffer &x, int64_t incx); +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& a, sycl::buffer& x, int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &a, sycl::buffer &x, int64_t incx); +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& a, sycl::buffer& x, int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &a, sycl::buffer &x, int64_t incx); +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& a, sycl::buffer& x, int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, int64_t incx); -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx); - -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx); +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx); + +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx); -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx); - -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx); - -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx); - -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx); - -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx); - -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, float beta, sycl::buffer &c, int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, double beta, sycl::buffer &c, int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - sycl::half alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, sycl::half beta, - sycl::buffer &c, int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, float beta, sycl::buffer &c, int64_t ldc); - -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, float beta, sycl::buffer &c, int64_t ldc); - -void hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc); - -void hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc); - -void herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, - sycl::buffer, 1> &a, int64_t lda, float beta, - sycl::buffer, 1> &c, int64_t ldc); - -void herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, - sycl::buffer, 1> &a, int64_t lda, double beta, - sycl::buffer, 1> &c, int64_t ldc); - -void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, float beta, - sycl::buffer, 1> &c, int64_t ldc); - -void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, double beta, - sycl::buffer, 1> &c, int64_t ldc); - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, - float beta, sycl::buffer &c, int64_t ldc); - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, - double beta, sycl::buffer &c, int64_t ldc); - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc); - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc); - -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, - sycl::buffer &a, int64_t lda, float beta, sycl::buffer &c, +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx); + +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx); + +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx); + +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx); + +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx); + +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx); + +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, float beta, sycl::buffer& c, int64_t ldc); + +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + double alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, double beta, sycl::buffer& c, int64_t ldc); + +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc); + +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc); + +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + sycl::half alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, sycl::half beta, + sycl::buffer& c, int64_t ldc); + +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, float beta, sycl::buffer& c, int64_t ldc); + +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, float beta, sycl::buffer& c, int64_t ldc); + +void hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc); + +void hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc); + +void herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, + sycl::buffer, 1>& a, int64_t lda, float beta, + sycl::buffer, 1>& c, int64_t ldc); + +void herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, + sycl::buffer, 1>& a, int64_t lda, double beta, + sycl::buffer, 1>& c, int64_t ldc); + +void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, float beta, + sycl::buffer, 1>& c, int64_t ldc); + +void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, double beta, + sycl::buffer, 1>& c, int64_t ldc); + +void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb, + float beta, sycl::buffer& c, int64_t ldc); + +void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb, + double beta, sycl::buffer& c, int64_t ldc); + +void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc); + +void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc); + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, + sycl::buffer& a, int64_t lda, float beta, sycl::buffer& c, int64_t ldc); -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, - sycl::buffer &a, int64_t lda, double beta, sycl::buffer &c, +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, + sycl::buffer& a, int64_t lda, double beta, sycl::buffer& c, int64_t ldc); -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc); +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc); -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc); +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc); -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, float beta, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size); +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, float beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, std::complex beta, sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, - float beta, sycl::buffer &c, int64_t ldc); +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb, + float beta, sycl::buffer& c, int64_t ldc); -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, double beta, sycl::buffer &c, int64_t ldc); +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, double beta, sycl::buffer& c, int64_t ldc); -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc); +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc); -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc); +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc); -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb); +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + int64_t m, int64_t n, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb); -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb); +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + int64_t m, int64_t n, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb); -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb); +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb); -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb); + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb); -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb); +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + int64_t m, int64_t n, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb); -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb); +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + int64_t m, int64_t n, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb); -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb); +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb); -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb); - -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size); - -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, double beta, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size); - -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, - int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb); + +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, float beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); + +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, double beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); + +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, + int64_t stride_b, std::complex beta, sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, sycl::half alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::half beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, sycl::half alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::half beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size); -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, float beta, sycl::buffer& c, int64_t ldc); -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, double beta, sycl::buffer& c, int64_t ldc); -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc); - -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc); - -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co); - -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co); - -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co); - -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co); - -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size); - -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size); - -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc); + +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc); + +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co); + +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co); + +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co); + +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co); + +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size); + +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size); + +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size); -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - float beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size); +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + float beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size); +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb); -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb); -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb); -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb); -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb); +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb); -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb); +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb); -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &b, int64_t ldb, std::int64_t strideb); +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& b, int64_t ldb, std::int64_t strideb); -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, int64_t ldb, +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, int64_t ldb, std::int64_t strideb); -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb); -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb); -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb); -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, float beta, - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, sycl::buffer& a, int64_t lda, float beta, + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, double beta, - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, sycl::buffer& a, int64_t lda, double beta, + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &b, int64_t ldb, - sycl::buffer, 1> &c, int64_t ldc); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& b, int64_t ldb, + sycl::buffer, 1>& c, int64_t ldc); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &b, int64_t ldb, - sycl::buffer, 1> &c, int64_t ldc); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& b, int64_t ldb, + sycl::buffer, 1>& c, int64_t ldc); // USM APIs -sycl::event asum(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - float *result, const std::vector &dependencies = {}); +sycl::event asum(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + float* result, const std::vector& dependencies = {}); -sycl::event asum(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - double *result, const std::vector &dependencies = {}); +sycl::event asum(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + double* result, const std::vector& dependencies = {}); -sycl::event asum(sycl::queue &queue, int64_t n, const float *x, int64_t incx, float *result, - const std::vector &dependencies = {}); +sycl::event asum(sycl::queue& queue, int64_t n, const float* x, int64_t incx, float* result, + const std::vector& dependencies = {}); -sycl::event asum(sycl::queue &queue, int64_t n, const double *x, int64_t incx, double *result, - const std::vector &dependencies = {}); +sycl::event asum(sycl::queue& queue, int64_t n, const double* x, int64_t incx, double* result, + const std::vector& dependencies = {}); -sycl::event axpy(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, float *y, - int64_t incy, const std::vector &dependencies = {}); +sycl::event axpy(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, float* y, + int64_t incy, const std::vector& dependencies = {}); -sycl::event axpy(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - double *y, int64_t incy, const std::vector &dependencies = {}); +sycl::event axpy(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + double* y, int64_t incy, const std::vector& dependencies = {}); -sycl::event axpy(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event axpy(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event axpy(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event axpy(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, float *alpha, const float **x, int64_t *incx, - float **y, int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, float* alpha, const float** x, int64_t* incx, + float** y, int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, double *alpha, const double **x, - int64_t *incx, double **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies = {}); +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, double* alpha, const double** x, + int64_t* incx, double** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies = {}); -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, std::complex* alpha, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, std::complex* alpha, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); -sycl::event axpy_batch(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, - int64_t stridex, float *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies = {}); +sycl::event axpy_batch(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, + int64_t stridex, float* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event axpy_batch(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - int64_t stridex, double *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event axpy_batch(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + int64_t stridex, double* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies = {}); +sycl::event axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies = {}); +sycl::event axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event axpby(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, - const float beta, float *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event axpby(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, + const float beta, float* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event axpby(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - const double beta, double *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event axpby(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + const double beta, double* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event copy(sycl::queue &queue, int64_t n, const float *x, int64_t incx, float *y, - int64_t incy, const std::vector &dependencies = {}); +sycl::event copy(sycl::queue& queue, int64_t n, const float* x, int64_t incx, float* y, + int64_t incy, const std::vector& dependencies = {}); -sycl::event copy(sycl::queue &queue, int64_t n, const double *x, int64_t incx, double *y, - int64_t incy, const std::vector &dependencies = {}); +sycl::event copy(sycl::queue& queue, int64_t n, const double* x, int64_t incx, double* y, + int64_t incy, const std::vector& dependencies = {}); -sycl::event copy(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event copy(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event copy(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event copy(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const float **x, int64_t *incx, float **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const float** x, int64_t* incx, float** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const double **x, int64_t *incx, double **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const double** x, int64_t* incx, double** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, int64_t *incx, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const std::complex** x, int64_t* incx, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, - int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const std::complex** x, + int64_t* incx, std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, int64_t n, const float *x, int64_t incx, int64_t stridex, - float *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, int64_t n, const float* x, int64_t incx, int64_t stridex, + float* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, int64_t n, const double *x, int64_t incx, - int64_t stridex, double *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, int64_t n, const double* x, int64_t incx, + int64_t stridex, double* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - int64_t stridex, std::complex *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + int64_t stridex, std::complex* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - int64_t stridex, std::complex *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + int64_t stridex, std::complex* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event dot(sycl::queue &queue, int64_t n, const float *x, int64_t incx, const float *y, - int64_t incy, float *result, const std::vector &dependencies = {}); +sycl::event dot(sycl::queue& queue, int64_t n, const float* x, int64_t incx, const float* y, + int64_t incy, float* result, const std::vector& dependencies = {}); -sycl::event dot(sycl::queue &queue, int64_t n, const double *x, int64_t incx, const double *y, - int64_t incy, double *result, const std::vector &dependencies = {}); +sycl::event dot(sycl::queue& queue, int64_t n, const double* x, int64_t incx, const double* y, + int64_t incy, double* result, const std::vector& dependencies = {}); -sycl::event dot(sycl::queue &queue, int64_t n, const float *x, int64_t incx, const float *y, - int64_t incy, double *result, const std::vector &dependencies = {}); +sycl::event dot(sycl::queue& queue, int64_t n, const float* x, int64_t incx, const float* y, + int64_t incy, double* result, const std::vector& dependencies = {}); -sycl::event dotc(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *result, - const std::vector &dependencies = {}); +sycl::event dotc(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* result, + const std::vector& dependencies = {}); -sycl::event dotc(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *result, - const std::vector &dependencies = {}); +sycl::event dotc(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* result, + const std::vector& dependencies = {}); -sycl::event dotu(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *result, - const std::vector &dependencies = {}); +sycl::event dotu(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* result, + const std::vector& dependencies = {}); -sycl::event dotu(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *result, - const std::vector &dependencies = {}); +sycl::event dotu(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* result, + const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, int64_t n, const float *x, int64_t incx, int64_t *result, - const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, int64_t n, const float* x, int64_t incx, int64_t* result, + const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, int64_t n, const double *x, int64_t incx, int64_t *result, - const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, int64_t n, const double* x, int64_t incx, int64_t* result, + const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - int64_t *result, const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + int64_t* result, const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - int64_t *result, const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + int64_t* result, const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, int64_t n, const float *x, int64_t incx, int64_t *result, - const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, int64_t n, const float* x, int64_t incx, int64_t* result, + const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, int64_t n, const double *x, int64_t incx, int64_t *result, - const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, int64_t n, const double* x, int64_t incx, int64_t* result, + const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - int64_t *result, const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + int64_t* result, const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - int64_t *result, const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + int64_t* result, const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - float *result, const std::vector &dependencies = {}); +sycl::event nrm2(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + float* result, const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - double *result, const std::vector &dependencies = {}); +sycl::event nrm2(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + double* result, const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, int64_t n, const float *x, int64_t incx, float *result, - const std::vector &dependencies = {}); +sycl::event nrm2(sycl::queue& queue, int64_t n, const float* x, int64_t incx, float* result, + const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, int64_t n, const double *x, int64_t incx, double *result, - const std::vector &dependencies = {}); +sycl::event nrm2(sycl::queue& queue, int64_t n, const double* x, int64_t incx, double* result, + const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, int64_t n, std::complex *x, int64_t incx, - std::complex *y, int64_t incy, float c, float s, - const std::vector &dependencies = {}); +sycl::event rot(sycl::queue& queue, int64_t n, std::complex* x, int64_t incx, + std::complex* y, int64_t incy, float c, float s, + const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, int64_t n, std::complex *x, int64_t incx, - std::complex *y, int64_t incy, double c, double s, - const std::vector &dependencies = {}); +sycl::event rot(sycl::queue& queue, int64_t n, std::complex* x, int64_t incx, + std::complex* y, int64_t incy, double c, double s, + const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, int64_t n, float *x, int64_t incx, float *y, int64_t incy, - float c, float s, const std::vector &dependencies = {}); +sycl::event rot(sycl::queue& queue, int64_t n, float* x, int64_t incx, float* y, int64_t incy, + float c, float s, const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, int64_t n, double *x, int64_t incx, double *y, int64_t incy, - double c, double s, const std::vector &dependencies = {}); +sycl::event rot(sycl::queue& queue, int64_t n, double* x, int64_t incx, double* y, int64_t incy, + double c, double s, const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, float *a, float *b, float *c, float *s, - const std::vector &dependencies = {}); +sycl::event rotg(sycl::queue& queue, float* a, float* b, float* c, float* s, + const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, double *a, double *b, double *c, double *s, - const std::vector &dependencies = {}); +sycl::event rotg(sycl::queue& queue, double* a, double* b, double* c, double* s, + const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, float *c, - std::complex *s, const std::vector &dependencies = {}); +sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, float* c, + std::complex* s, const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, double *c, - std::complex *s, const std::vector &dependencies = {}); +sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, double* c, + std::complex* s, const std::vector& dependencies = {}); -sycl::event rotm(sycl::queue &queue, int64_t n, float *x, int64_t incx, float *y, int64_t incy, - float *param, const std::vector &dependencies = {}); +sycl::event rotm(sycl::queue& queue, int64_t n, float* x, int64_t incx, float* y, int64_t incy, + float* param, const std::vector& dependencies = {}); -sycl::event rotm(sycl::queue &queue, int64_t n, double *x, int64_t incx, double *y, int64_t incy, - double *param, const std::vector &dependencies = {}); +sycl::event rotm(sycl::queue& queue, int64_t n, double* x, int64_t incx, double* y, int64_t incy, + double* param, const std::vector& dependencies = {}); -sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, float y1, float *param, - const std::vector &dependencies = {}); +sycl::event rotmg(sycl::queue& queue, float* d1, float* d2, float* x1, float y1, float* param, + const std::vector& dependencies = {}); -sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, double y1, double *param, - const std::vector &dependencies = {}); +sycl::event rotmg(sycl::queue& queue, double* d1, double* d2, double* x1, double y1, double* param, + const std::vector& dependencies = {}); -sycl::event scal(sycl::queue &queue, int64_t n, float alpha, float *x, int64_t incx, - const std::vector &dependencies = {}); +sycl::event scal(sycl::queue& queue, int64_t n, float alpha, float* x, int64_t incx, + const std::vector& dependencies = {}); -sycl::event scal(sycl::queue &queue, int64_t n, double alpha, double *x, int64_t incx, - const std::vector &dependencies = {}); +sycl::event scal(sycl::queue& queue, int64_t n, double alpha, double* x, int64_t incx, + const std::vector& dependencies = {}); -sycl::event scal(sycl::queue &queue, int64_t n, std::complex alpha, std::complex *x, - int64_t incx, const std::vector &dependencies = {}); +sycl::event scal(sycl::queue& queue, int64_t n, std::complex alpha, std::complex* x, + int64_t incx, const std::vector& dependencies = {}); -sycl::event scal(sycl::queue &queue, int64_t n, std::complex alpha, std::complex *x, - int64_t incx, const std::vector &dependencies = {}); +sycl::event scal(sycl::queue& queue, int64_t n, std::complex alpha, std::complex* x, + int64_t incx, const std::vector& dependencies = {}); -sycl::event scal(sycl::queue &queue, int64_t n, float alpha, std::complex *x, int64_t incx, - const std::vector &dependencies = {}); +sycl::event scal(sycl::queue& queue, int64_t n, float alpha, std::complex* x, int64_t incx, + const std::vector& dependencies = {}); -sycl::event scal(sycl::queue &queue, int64_t n, double alpha, std::complex *x, int64_t incx, - const std::vector &dependencies = {}); +sycl::event scal(sycl::queue& queue, int64_t n, double alpha, std::complex* x, int64_t incx, + const std::vector& dependencies = {}); -sycl::event sdsdot(sycl::queue &queue, int64_t n, float sb, const float *x, int64_t incx, - const float *y, int64_t incy, float *result, - const std::vector &dependencies = {}); +sycl::event sdsdot(sycl::queue& queue, int64_t n, float sb, const float* x, int64_t incx, + const float* y, int64_t incy, float* result, + const std::vector& dependencies = {}); -sycl::event swap(sycl::queue &queue, int64_t n, float *x, int64_t incx, float *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event swap(sycl::queue& queue, int64_t n, float* x, int64_t incx, float* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event swap(sycl::queue &queue, int64_t n, double *x, int64_t incx, double *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event swap(sycl::queue& queue, int64_t n, double* x, int64_t incx, double* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event swap(sycl::queue &queue, int64_t n, std::complex *x, int64_t incx, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event swap(sycl::queue& queue, int64_t n, std::complex* x, int64_t incx, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event swap(sycl::queue &queue, int64_t n, std::complex *x, int64_t incx, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event swap(sycl::queue& queue, int64_t n, std::complex* x, int64_t incx, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - float alpha, const float *a, int64_t lda, const float *x, int64_t incx, float beta, - float *y, int64_t incy, const std::vector &dependencies = {}); +sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + float alpha, const float* a, int64_t lda, const float* x, int64_t incx, float beta, + float* y, int64_t incy, const std::vector& dependencies = {}); -sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - double alpha, const double *a, int64_t lda, const double *x, int64_t incx, - double beta, double *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + double alpha, const double* a, int64_t lda, const double* x, int64_t incx, + double beta, double* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, const float *x, int64_t incx, float beta, float *y, - int64_t incy, const std::vector &dependencies = {}); +sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, const float* x, int64_t incx, float beta, float* y, + int64_t incy, const std::vector& dependencies = {}); -sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, const double *x, int64_t incx, double beta, - double *y, int64_t incy, const std::vector &dependencies = {}); +sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, const double* x, int64_t incx, double beta, + double* y, int64_t incy, const std::vector& dependencies = {}); -sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); +sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); -sycl::event gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stridea, const float *x, int64_t incx, - int64_t stridex, float beta, float *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stridea, const float* x, int64_t incx, + int64_t stridex, float beta, float* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stridea, const double *x, int64_t incx, - int64_t stridex, double beta, double *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stridea, const double* x, int64_t incx, + int64_t stridex, double beta, double* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stridea, const std::complex *x, int64_t incx, int64_t stridex, - std::complex beta, std::complex *y, int64_t incy, +sycl::event gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stridea, const std::complex* x, int64_t incx, int64_t stridex, + std::complex beta, std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stridea, const std::complex *x, int64_t incx, - int64_t stridex, std::complex beta, std::complex *y, +sycl::event gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stridea, const std::complex* x, int64_t incx, + int64_t stridex, std::complex beta, std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, float *alpha, - const float **a, int64_t *lda, const float **x, int64_t *incx, float *beta, - float **y, int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, double *alpha, - const double **a, int64_t *lda, const double **x, int64_t *incx, - double *beta, double **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - const std::complex **x, int64_t *incx, std::complex *beta, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - const std::complex **x, int64_t *incx, std::complex *beta, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, const float *a, - int64_t lda, int64_t stridea, const float *x, int64_t incx, int64_t stridex, - float *c, int64_t ldc, int64_t stridec, int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, const double *a, - int64_t lda, int64_t stridea, const double *x, int64_t incx, int64_t stridex, - double *c, int64_t ldc, int64_t stridec, int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const std::complex *a, int64_t lda, int64_t stridea, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *c, int64_t ldc, int64_t stridec, int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const std::complex *a, int64_t lda, int64_t stridea, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *c, int64_t ldc, int64_t stridec, int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const float **a, int64_t *lda, const float **x, int64_t *incx, float **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const double **a, int64_t *lda, const double **x, int64_t *incx, double **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const std::complex **a, int64_t *lda, const std::complex **x, - int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const std::complex **a, int64_t *lda, const std::complex **x, - int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies = {}); - -sycl::event ger(sycl::queue &queue, int64_t m, int64_t n, float alpha, const float *x, int64_t incx, - const float *y, int64_t incy, float *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event ger(sycl::queue &queue, int64_t m, int64_t n, double alpha, const double *x, - int64_t incx, const double *y, int64_t incy, double *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event gerc(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event gerc(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event geru(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event geru(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hemv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, const std::complex *x, - int64_t incx, std::complex beta, std::complex *y, int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hemv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, const std::complex *x, - int64_t incx, std::complex beta, std::complex *y, int64_t incy, - const std::vector &dependencies = {}); - -sycl::event her(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, - const std::complex *x, int64_t incx, std::complex *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event her(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, - const std::complex *x, int64_t incx, std::complex *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event her2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event her2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *a, const std::complex *x, int64_t incx, - std::complex beta, std::complex *y, int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *a, const std::complex *x, int64_t incx, - std::complex beta, std::complex *y, int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hpr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, - const std::complex *x, int64_t incx, std::complex *a, - const std::vector &dependencies = {}); - -sycl::event hpr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, - const std::complex *x, int64_t incx, std::complex *a, - const std::vector &dependencies = {}); - -sycl::event hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, - const std::vector &dependencies = {}); - -sycl::event hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, - const std::vector &dependencies = {}); - -sycl::event sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, float alpha, - const float *a, int64_t lda, const float *x, int64_t incx, float beta, float *y, - int64_t incy, const std::vector &dependencies = {}); - -sycl::event sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, double alpha, - const double *a, int64_t lda, const double *x, int64_t incx, double beta, - double *y, int64_t incy, const std::vector &dependencies = {}); - -sycl::event spmv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *a, - const float *x, int64_t incx, float beta, float *y, int64_t incy, - const std::vector &dependencies = {}); - -sycl::event spmv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *a, - const double *x, int64_t incx, double beta, double *y, int64_t incy, - const std::vector &dependencies = {}); - -sycl::event spr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *x, - int64_t incx, float *a, const std::vector &dependencies = {}); - -sycl::event spr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *x, - int64_t incx, double *a, const std::vector &dependencies = {}); - -sycl::event spr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *x, - int64_t incx, const float *y, int64_t incy, float *a, - const std::vector &dependencies = {}); - -sycl::event spr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *x, - int64_t incx, const double *y, int64_t incy, double *a, - const std::vector &dependencies = {}); - -sycl::event symv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *a, - int64_t lda, const float *x, int64_t incx, float beta, float *y, int64_t incy, - const std::vector &dependencies = {}); - -sycl::event symv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *a, - int64_t lda, const double *x, int64_t incx, double beta, double *y, int64_t incy, - const std::vector &dependencies = {}); - -sycl::event syr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *x, - int64_t incx, float *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event syr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *x, - int64_t incx, double *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event syr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *x, - int64_t incx, const float *y, int64_t incy, float *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event syr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *x, - int64_t incx, const double *y, int64_t incy, double *a, int64_t lda, - const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const float *a, int64_t lda, float *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const double *a, int64_t lda, double *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const std::complex *a, int64_t lda, std::complex *x, - int64_t incx, const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const std::complex *a, int64_t lda, std::complex *x, - int64_t incx, const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const float *a, int64_t lda, float *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const double *a, int64_t lda, double *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const std::complex *a, int64_t lda, std::complex *x, - int64_t incx, const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const std::complex *a, int64_t lda, std::complex *x, - int64_t incx, const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const float *a, float *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const double *a, double *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *a, std::complex *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *a, std::complex *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const float *a, float *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const double *a, double *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *a, std::complex *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *a, std::complex *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const float *a, int64_t lda, float *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const double *a, int64_t lda, double *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *a, int64_t lda, std::complex *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *a, int64_t lda, std::complex *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const float *a, int64_t lda, float *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const double *a, int64_t lda, double *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *a, int64_t lda, std::complex *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *a, int64_t lda, std::complex *x, int64_t incx, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const float *a, int64_t lda, const float *b, int64_t ldb, - float beta, float *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, double alpha, const double *a, int64_t lda, const double *b, - int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, sycl::half alpha, const sycl::half *a, int64_t lda, const sycl::half *b, - int64_t ldb, sycl::half beta, sycl::half *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const sycl::half *a, int64_t lda, const sycl::half *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const bfloat16 *a, int64_t lda, const bfloat16 *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, const std::complex *a, int64_t lda, float beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, const std::complex *a, int64_t lda, double beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, float beta, std::complex *c, - int64_t ldc, const std::vector &dependencies = {}); - -sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, double beta, std::complex *c, - int64_t ldc, const std::vector &dependencies = {}); - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - float alpha, const float *a, int64_t lda, const float *b, int64_t ldb, float beta, - float *c, int64_t ldc, const std::vector &dependencies = {}); - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - double alpha, const double *a, int64_t lda, const double *b, int64_t ldb, - double beta, double *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, const float *a, int64_t lda, float beta, float *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, const double *a, int64_t lda, double beta, double *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, float *alpha, const float **a, int64_t *lda, float *beta, - float **c, int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, double *alpha, const double **a, int64_t *lda, double *beta, - double **c, int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, const float *a, int64_t lda, int64_t stride_a, float beta, - float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, const double *a, int64_t lda, int64_t stride_a, double beta, - double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex beta, std::complex *c, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, float* alpha, + const float** a, int64_t* lda, const float** x, int64_t* incx, float* beta, + float** y, int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, double* alpha, + const double** a, int64_t* lda, const double** x, int64_t* incx, + double* beta, double** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + const std::complex** x, int64_t* incx, std::complex* beta, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + const std::complex** x, int64_t* incx, std::complex* beta, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, const float* a, + int64_t lda, int64_t stridea, const float* x, int64_t incx, int64_t stridex, + float* c, int64_t ldc, int64_t stridec, int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, const double* a, + int64_t lda, int64_t stridea, const double* x, int64_t incx, int64_t stridex, + double* c, int64_t ldc, int64_t stridec, int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + const std::complex* a, int64_t lda, int64_t stridea, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* c, int64_t ldc, int64_t stridec, int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + const std::complex* a, int64_t lda, int64_t stridea, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* c, int64_t ldc, int64_t stridec, int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const float** a, int64_t* lda, const float** x, int64_t* incx, float** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const double** a, int64_t* lda, const double** x, int64_t* incx, double** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const std::complex** a, int64_t* lda, const std::complex** x, + int64_t* incx, std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const std::complex** a, int64_t* lda, const std::complex** x, + int64_t* incx, std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event ger(sycl::queue& queue, int64_t m, int64_t n, float alpha, const float* x, int64_t incx, + const float* y, int64_t incy, float* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event ger(sycl::queue& queue, int64_t m, int64_t n, double alpha, const double* x, + int64_t incx, const double* y, int64_t incy, double* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event gerc(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event gerc(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event geru(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event geru(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); + +sycl::event hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies = {}); + +sycl::event hemv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* a, int64_t lda, const std::complex* x, + int64_t incx, std::complex beta, std::complex* y, int64_t incy, + const std::vector& dependencies = {}); + +sycl::event hemv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* a, int64_t lda, const std::complex* x, + int64_t incx, std::complex beta, std::complex* y, int64_t incy, + const std::vector& dependencies = {}); + +sycl::event her(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, + const std::complex* x, int64_t incx, std::complex* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event her(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, + const std::complex* x, int64_t incx, std::complex* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event her2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event her2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* a, const std::complex* x, int64_t incx, + std::complex beta, std::complex* y, int64_t incy, + const std::vector& dependencies = {}); + +sycl::event hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* a, const std::complex* x, int64_t incx, + std::complex beta, std::complex* y, int64_t incy, + const std::vector& dependencies = {}); + +sycl::event hpr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, + const std::complex* x, int64_t incx, std::complex* a, + const std::vector& dependencies = {}); + +sycl::event hpr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, + const std::complex* x, int64_t incx, std::complex* a, + const std::vector& dependencies = {}); + +sycl::event hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, + const std::vector& dependencies = {}); + +sycl::event hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, + const std::vector& dependencies = {}); + +sycl::event sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, float alpha, + const float* a, int64_t lda, const float* x, int64_t incx, float beta, float* y, + int64_t incy, const std::vector& dependencies = {}); + +sycl::event sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, double alpha, + const double* a, int64_t lda, const double* x, int64_t incx, double beta, + double* y, int64_t incy, const std::vector& dependencies = {}); + +sycl::event spmv(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* a, + const float* x, int64_t incx, float beta, float* y, int64_t incy, + const std::vector& dependencies = {}); + +sycl::event spmv(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* a, + const double* x, int64_t incx, double beta, double* y, int64_t incy, + const std::vector& dependencies = {}); + +sycl::event spr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* x, + int64_t incx, float* a, const std::vector& dependencies = {}); + +sycl::event spr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* x, + int64_t incx, double* a, const std::vector& dependencies = {}); + +sycl::event spr2(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* x, + int64_t incx, const float* y, int64_t incy, float* a, + const std::vector& dependencies = {}); + +sycl::event spr2(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* x, + int64_t incx, const double* y, int64_t incy, double* a, + const std::vector& dependencies = {}); + +sycl::event symv(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* a, + int64_t lda, const float* x, int64_t incx, float beta, float* y, int64_t incy, + const std::vector& dependencies = {}); + +sycl::event symv(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* a, + int64_t lda, const double* x, int64_t incx, double beta, double* y, int64_t incy, + const std::vector& dependencies = {}); + +sycl::event syr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* x, + int64_t incx, float* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event syr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* x, + int64_t incx, double* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event syr2(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* x, + int64_t incx, const float* y, int64_t incy, float* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event syr2(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* x, + int64_t incx, const double* y, int64_t incy, double* a, int64_t lda, + const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const float* a, int64_t lda, float* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const double* a, int64_t lda, double* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const std::complex* a, int64_t lda, std::complex* x, + int64_t incx, const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const std::complex* a, int64_t lda, std::complex* x, + int64_t incx, const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const float* a, int64_t lda, float* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const double* a, int64_t lda, double* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const std::complex* a, int64_t lda, std::complex* x, + int64_t incx, const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const std::complex* a, int64_t lda, std::complex* x, + int64_t incx, const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const float* a, float* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const double* a, double* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* a, std::complex* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* a, std::complex* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const float* a, float* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const double* a, double* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* a, std::complex* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* a, std::complex* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const float* a, int64_t lda, float* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const double* a, int64_t lda, double* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* a, int64_t lda, std::complex* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* a, int64_t lda, std::complex* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const float* a, int64_t lda, float* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const double* a, int64_t lda, double* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* a, int64_t lda, std::complex* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* a, int64_t lda, std::complex* x, int64_t incx, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const float* a, int64_t lda, const float* b, int64_t ldb, + float beta, float* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, sycl::half alpha, const sycl::half* a, int64_t lda, const sycl::half* b, + int64_t ldb, sycl::half beta, sycl::half* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const sycl::half* a, int64_t lda, const sycl::half* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const bfloat16* a, int64_t lda, const bfloat16* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, const std::complex* a, int64_t lda, float beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, const std::complex* a, int64_t lda, double beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, float beta, std::complex* c, + int64_t ldc, const std::vector& dependencies = {}); + +sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, double beta, std::complex* c, + int64_t ldc, const std::vector& dependencies = {}); + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + float alpha, const float* a, int64_t lda, const float* b, int64_t ldb, float beta, + float* c, int64_t ldc, const std::vector& dependencies = {}); + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + double alpha, const double* a, int64_t lda, const double* b, int64_t ldb, + double beta, double* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, const float* a, int64_t lda, float beta, float* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, const double* a, int64_t lda, double beta, double* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, float* alpha, const float** a, int64_t* lda, float* beta, + float** c, int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, double* alpha, const double** a, int64_t* lda, double* beta, + double** c, int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, const float* a, int64_t lda, int64_t stride_a, float beta, + float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, const double* a, int64_t lda, int64_t stride_a, double beta, + double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex beta, std::complex *c, +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, const float *a, int64_t lda, const float *b, int64_t ldb, float beta, - float *c, int64_t ldc, const std::vector &dependencies = {}); - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, const double *a, int64_t lda, const double *b, int64_t ldb, - double beta, double *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, - float *b, int64_t ldb, const std::vector &dependencies = {}); - -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, - double *b, int64_t ldb, const std::vector &dependencies = {}); - -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, + const std::vector& dependencies = {}); + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, const float* a, int64_t lda, const float* b, int64_t ldb, float beta, + float* c, int64_t ldc, const std::vector& dependencies = {}); + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, const double* a, int64_t lda, const double* b, int64_t ldb, + double beta, double* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, const float* a, int64_t lda, + float* b, int64_t ldb, const std::vector& dependencies = {}); + +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, const double* a, int64_t lda, + double* b, int64_t ldb, const std::vector& dependencies = {}); + +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies = {}); + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies = {}); -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies = {}); + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies = {}); -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, - float *b, int64_t ldb, const std::vector &dependencies = {}); +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, const float* a, int64_t lda, + float* b, int64_t ldb, const std::vector& dependencies = {}); -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, - double *b, int64_t ldb, const std::vector &dependencies = {}); +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, const double* a, int64_t lda, + double* b, int64_t ldb, const std::vector& dependencies = {}); -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies = {}); + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies = {}); -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies = {}); + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies = {}); -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, - int64_t lda, int64_t stride_a, float *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, const float* a, + int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, - int64_t lda, int64_t stride_a, double *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, const double* a, + int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, float *alpha, const float **a, - int64_t *lda, float **b, int64_t *ldb, int64_t group_count, - int64_t *group_size, const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, double *alpha, const double **a, - int64_t *lda, double **b, int64_t *ldb, int64_t group_count, - int64_t *group_size, const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, std::complex *alpha, - const std::complex **a, int64_t *lda, std::complex **b, - int64_t *ldb, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, std::complex *alpha, - const std::complex **a, int64_t *lda, std::complex **b, - int64_t *ldb, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const float **a, int64_t *lda, - const float **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, double *alpha, const double **a, int64_t *lda, - const double **b, int64_t *ldb, double *beta, double **c, int64_t *ldc, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, std::complex *alpha, - const std::complex **a, int64_t *lda, const std::complex **b, - int64_t *ldb, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, std::complex *alpha, - const std::complex **a, int64_t *lda, const std::complex **b, - int64_t *ldb, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, sycl::half *alpha, const sycl::half **a, - int64_t *lda, const sycl::half **b, int64_t *ldb, sycl::half *beta, - sycl::half **c, int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const sycl::half **a, int64_t *lda, - const sycl::half **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const std::int8_t **a, int64_t *lda, - const std::int8_t **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const std::int8_t **a, int64_t *lda, - const std::int8_t **b, int64_t *ldb, float *beta, std::int32_t **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const float *a, int64_t lda, int64_t stride_a, - const float *b, int64_t ldb, int64_t stride_b, float beta, float *c, + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, float* alpha, const float** a, + int64_t* lda, float** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, double* alpha, const double** a, + int64_t* lda, double** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, std::complex* alpha, + const std::complex** a, int64_t* lda, std::complex** b, + int64_t* ldb, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, std::complex* alpha, + const std::complex** a, int64_t* lda, std::complex** b, + int64_t* ldb, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const float** a, int64_t* lda, + const float** b, int64_t* ldb, float* beta, float** c, int64_t* ldc, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, double* alpha, const double** a, int64_t* lda, + const double** b, int64_t* ldb, double* beta, double** c, int64_t* ldc, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, std::complex* alpha, + const std::complex** a, int64_t* lda, const std::complex** b, + int64_t* ldb, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, std::complex* alpha, + const std::complex** a, int64_t* lda, const std::complex** b, + int64_t* ldb, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, sycl::half* alpha, const sycl::half** a, + int64_t* lda, const sycl::half** b, int64_t* ldb, sycl::half* beta, + sycl::half** c, int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const sycl::half** a, int64_t* lda, + const sycl::half** b, int64_t* ldb, float* beta, float** c, int64_t* ldc, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const std::int8_t** a, int64_t* lda, + const std::int8_t** b, int64_t* ldb, float* beta, float** c, int64_t* ldc, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const std::int8_t** a, int64_t* lda, + const std::int8_t** b, int64_t* ldb, float* beta, std::int32_t** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const float* a, int64_t lda, int64_t stride_a, + const float* b, int64_t ldb, int64_t stride_b, float beta, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, double alpha, const double *a, int64_t lda, int64_t stride_a, - const double *b, int64_t ldb, int64_t stride_b, double beta, double *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, double alpha, const double* a, int64_t lda, int64_t stride_a, + const double* b, int64_t ldb, int64_t stride_b, double beta, double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, int64_t stride_a, const std::complex *b, int64_t ldb, - int64_t stride_b, std::complex beta, std::complex *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, int64_t stride_a, const std::complex* b, int64_t ldb, + int64_t stride_b, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, int64_t stride_a, const std::complex *b, int64_t ldb, - int64_t stride_b, std::complex beta, std::complex *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, int64_t stride_a, const std::complex* b, int64_t ldb, + int64_t stride_b, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, sycl::half alpha, const sycl::half *a, int64_t lda, - int64_t stride_a, const sycl::half *b, int64_t ldb, int64_t stride_b, - sycl::half beta, sycl::half *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, sycl::half alpha, const sycl::half* a, int64_t lda, + int64_t stride_a, const sycl::half* b, int64_t ldb, int64_t stride_b, + sycl::half beta, sycl::half* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const sycl::half *a, int64_t lda, int64_t stride_a, - const sycl::half *b, int64_t ldb, int64_t stride_b, float beta, float *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const sycl::half* a, int64_t lda, int64_t stride_a, + const sycl::half* b, int64_t ldb, int64_t stride_b, float beta, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const std::int8_t *a, int64_t lda, int64_t stride_a, - const std::int8_t *b, int64_t ldb, int64_t stride_b, float beta, float *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const std::int8_t* a, int64_t lda, int64_t stride_a, + const std::int8_t* b, int64_t ldb, int64_t stride_b, float beta, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const std::int8_t *a, int64_t lda, int64_t stride_a, - const std::int8_t *b, int64_t ldb, int64_t stride_b, float beta, - std::int32_t *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, float alpha, const float *a, int64_t lda, const float *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, double alpha, const double *a, int64_t lda, const double *b, - int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const std::int8_t *a, - int64_t lda, std::int8_t ao, const std::uint8_t *b, int64_t ldb, - std::uint8_t bo, float beta, std::int32_t *c, int64_t ldc, - const std::int32_t *co, const std::vector &dependencies = {}); - -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const std::int8_t *a, - int64_t lda, std::int8_t ao, const std::int8_t *b, int64_t ldb, - std::int8_t bo, float beta, std::int32_t *c, int64_t ldc, - const std::int32_t *co, const std::vector &dependencies = {}); - -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const std::uint8_t *a, - int64_t lda, std::uint8_t ao, const std::int8_t *b, int64_t ldb, - std::int8_t bo, float beta, std::int32_t *c, int64_t ldc, - const std::int32_t *co, const std::vector &dependencies = {}); - -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const std::uint8_t *a, - int64_t lda, std::uint8_t ao, const std::uint8_t *b, int64_t ldb, - std::uint8_t bo, float beta, std::int32_t *c, int64_t ldc, - const std::int32_t *co, const std::vector &dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const std::int8_t* a, int64_t lda, int64_t stride_a, + const std::int8_t* b, int64_t ldb, int64_t stride_b, float beta, + std::int32_t* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, float alpha, const float* a, int64_t lda, const float* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const std::int8_t* a, + int64_t lda, std::int8_t ao, const std::uint8_t* b, int64_t ldb, + std::uint8_t bo, float beta, std::int32_t* c, int64_t ldc, + const std::int32_t* co, const std::vector& dependencies = {}); + +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const std::int8_t* a, + int64_t lda, std::int8_t ao, const std::int8_t* b, int64_t ldb, + std::int8_t bo, float beta, std::int32_t* c, int64_t ldc, + const std::int32_t* co, const std::vector& dependencies = {}); + +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const std::uint8_t* a, + int64_t lda, std::uint8_t ao, const std::int8_t* b, int64_t ldb, + std::int8_t bo, float beta, std::int32_t* c, int64_t ldc, + const std::int32_t* co, const std::vector& dependencies = {}); + +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const std::uint8_t* a, + int64_t lda, std::uint8_t ao, const std::uint8_t* b, int64_t ldb, + std::uint8_t bo, float beta, std::int32_t* c, int64_t ldc, + const std::int32_t* co, const std::vector& dependencies = {}); + +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies = {}); +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies = {}); -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies = {}); +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies = {}); +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, - float beta, const float *b, int64_t ldb, int64_t stride_b, float *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, float alpha, const float* a, int64_t lda, int64_t stride_a, + float beta, const float* b, int64_t ldb, int64_t stride_b, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, double alpha, const double *a, int64_t lda, int64_t stride_a, - double beta, const double *b, int64_t ldb, int64_t stride_b, double *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, double alpha, const double* a, int64_t lda, int64_t stride_a, + double beta, const double* b, int64_t ldb, int64_t stride_b, double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies = {}); + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, float *b, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, double *b, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, std::int64_t stridea, float *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies = {}); - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, std::int64_t stridea, double *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies = {}); - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::int64_t stridea, std::complex *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies = {}); - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::int64_t stridea, std::complex *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies = {}); - -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, const float *a, int64_t lda, float beta, const float *b, - int64_t ldb, float *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, const double *a, int64_t lda, double beta, const double *b, - int64_t ldb, double *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, const std::complex *b, int64_t ldb, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, const std::complex *b, int64_t ldb, - std::complex *c, int64_t ldc, - const std::vector &dependencies = {}); + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, float* b, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, double* b, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex* b, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex* b, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, std::int64_t stridea, float* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies = {}); + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, std::int64_t stridea, double* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies = {}); + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::int64_t stridea, std::complex* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies = {}); + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::int64_t stridea, std::complex* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies = {}); + +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, const float* a, int64_t lda, float beta, const float* b, + int64_t ldb, float* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, const double* a, int64_t lda, double beta, const double* b, + int64_t ldb, double* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, const std::complex* b, int64_t ldb, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, const std::complex* b, int64_t ldb, + std::complex* c, int64_t ldc, + const std::vector& dependencies = {}); sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, float* alpha, const float** a, int64_t* lda, float** b, int64_t* ldb, diff --git a/include/oneapi/math/detail/exceptions.hpp b/include/oneapi/math/detail/exceptions.hpp index 8a82db01a..eb8cc0b12 100644 --- a/include/oneapi/math/detail/exceptions.hpp +++ b/include/oneapi/math/detail/exceptions.hpp @@ -31,14 +31,14 @@ namespace math { class backend_not_found : public oneapi::math::exception { public: - backend_not_found(const std::string &info = "") + backend_not_found(const std::string& info = "") : oneapi::math::exception( "", "", ((info.length() != 0) ? info : "Couldn't load selected backend")) {} }; class function_not_found : public oneapi::math::exception { public: - function_not_found(const std::string &info = "") + function_not_found(const std::string& info = "") : oneapi::math::exception( "", "", ((info.length() != 0) ? info : "Couldn't load functions from selected backend")) { @@ -48,8 +48,8 @@ class function_not_found : public oneapi::math::exception { class library_not_found : public oneapi::math::exception { public: library_not_found(const std::string& message) : exception(message) {} - library_not_found(const std::string &domain, const std::string &function, - const std::string &info = "") + library_not_found(const std::string& domain, const std::string& function, + const std::string& info = "") : oneapi::math::exception( domain, function, "library not found" + ((info.length() != 0) ? (": " + info) : "")) {} @@ -57,10 +57,11 @@ class library_not_found : public oneapi::math::exception { class specification_mismatch : public oneapi::math::exception { public: - specification_mismatch(const std::string &info = "") + specification_mismatch(const std::string& info = "") : oneapi::math::exception( "", "", - ((info.length() != 0) ? info : "Loaded oneMath specification version mismatch")) {} + ((info.length() != 0) ? info : "Loaded oneMath specification version mismatch")) { + } }; } // namespace math diff --git a/include/oneapi/math/detail/get_device_id.hpp b/include/oneapi/math/detail/get_device_id.hpp index c81f96c13..32d493059 100644 --- a/include/oneapi/math/detail/get_device_id.hpp +++ b/include/oneapi/math/detail/get_device_id.hpp @@ -40,7 +40,7 @@ namespace oneapi { namespace math { -inline oneapi::math::device get_device_id(sycl::queue &queue) { +inline oneapi::math::device get_device_id(sycl::queue& queue) { oneapi::math::device device_id; if (queue.get_device().is_cpu()) device_id = device::x86cpu; diff --git a/include/oneapi/math/dft/backward.hpp b/include/oneapi/math/dft/backward.hpp index 50c9f4234..6d7f7f122 100644 --- a/include/oneapi/math/dft/backward.hpp +++ b/include/oneapi/math/dft/backward.hpp @@ -33,7 +33,7 @@ namespace oneapi::math::dft { //In-place transform template -void compute_backward(descriptor_type &desc, sycl::buffer &inout) { +void compute_backward(descriptor_type& desc, sycl::buffer& inout) { static_assert(detail::valid_compute_arg::value, "unexpected type for data_type"); @@ -46,8 +46,8 @@ void compute_backward(descriptor_type &desc, sycl::buffer &inout) //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template , bool> = true> -void compute_backward(descriptor_type &desc, sycl::buffer &inout_re, - sycl::buffer &inout_im) { +void compute_backward(descriptor_type& desc, sycl::buffer& inout_re, + sycl::buffer& inout_im) { static_assert(detail::valid_compute_arg::value, "unexpected type for data_type"); @@ -61,8 +61,8 @@ void compute_backward(descriptor_type &desc, sycl::buffer &inout_r //Out-of-place transform template -void compute_backward(descriptor_type &desc, sycl::buffer &in, - sycl::buffer &out) { +void compute_backward(descriptor_type& desc, sycl::buffer& in, + sycl::buffer& out) { static_assert(detail::valid_compute_arg::value, "unexpected type for input_type"); static_assert(detail::valid_compute_arg::value, @@ -79,9 +79,9 @@ void compute_backward(descriptor_type &desc, sycl::buffer &in, //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -void compute_backward(descriptor_type &desc, sycl::buffer &in_re, - sycl::buffer &in_im, sycl::buffer &out_re, - sycl::buffer &out_im) { +void compute_backward(descriptor_type& desc, sycl::buffer& in_re, + sycl::buffer& in_im, sycl::buffer& out_re, + sycl::buffer& out_im) { static_assert(detail::valid_compute_arg::value, "unexpected type for input_type"); static_assert(detail::valid_compute_arg::value, @@ -104,34 +104,32 @@ void compute_backward(descriptor_type &desc, sycl::buffer &in_re, //In-place transform template -sycl::event compute_backward(descriptor_type &desc, data_type *inout, - const std::vector &dependencies = {}) { +sycl::event compute_backward(descriptor_type& desc, data_type* inout, + const std::vector& dependencies = {}) { static_assert(detail::valid_compute_arg::value, "unexpected type for data_type"); using fwd_type = typename detail::descriptor_info::forward_type; - return get_commit(desc)->backward_ip_cc(desc, reinterpret_cast(inout), - dependencies); + return get_commit(desc)->backward_ip_cc(desc, reinterpret_cast(inout), dependencies); } //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template , bool> = true> -sycl::event compute_backward(descriptor_type &desc, data_type *inout_re, data_type *inout_im, - const std::vector &dependencies = {}) { +sycl::event compute_backward(descriptor_type& desc, data_type* inout_re, data_type* inout_im, + const std::vector& dependencies = {}) { static_assert(detail::valid_compute_arg::value, "unexpected type for data_type"); using scalar_type = typename detail::descriptor_info::scalar_type; - return get_commit(desc)->backward_ip_rr(desc, reinterpret_cast(inout_re), - reinterpret_cast(inout_im), - dependencies); + return get_commit(desc)->backward_ip_rr(desc, reinterpret_cast(inout_re), + reinterpret_cast(inout_im), dependencies); } //Out-of-place transform template -sycl::event compute_backward(descriptor_type &desc, input_type *in, output_type *out, - const std::vector &dependencies = {}) { +sycl::event compute_backward(descriptor_type& desc, input_type* in, output_type* out, + const std::vector& dependencies = {}) { static_assert(detail::valid_compute_arg::value, "unexpected type for input_type"); static_assert(detail::valid_compute_arg::value, @@ -139,25 +137,25 @@ sycl::event compute_backward(descriptor_type &desc, input_type *in, output_type using fwd_type = typename detail::descriptor_info::forward_type; using bwd_type = typename detail::descriptor_info::backward_type; - return get_commit(desc)->backward_op_cc(desc, reinterpret_cast(in), - reinterpret_cast(out), dependencies); + return get_commit(desc)->backward_op_cc(desc, reinterpret_cast(in), + reinterpret_cast(out), dependencies); } //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -sycl::event compute_backward(descriptor_type &desc, input_type *in_re, input_type *in_im, - output_type *out_re, output_type *out_im, - const std::vector &dependencies = {}) { +sycl::event compute_backward(descriptor_type& desc, input_type* in_re, input_type* in_im, + output_type* out_re, output_type* out_im, + const std::vector& dependencies = {}) { static_assert(detail::valid_compute_arg::value, "unexpected type for input_type"); static_assert(detail::valid_compute_arg::value, "unexpected type for output_type"); using scalar_type = typename detail::descriptor_info::scalar_type; - return get_commit(desc)->backward_op_rr(desc, reinterpret_cast(in_re), - reinterpret_cast(in_im), - reinterpret_cast(out_re), - reinterpret_cast(out_im), dependencies); + return get_commit(desc)->backward_op_rr(desc, reinterpret_cast(in_re), + reinterpret_cast(in_im), + reinterpret_cast(out_re), + reinterpret_cast(out_im), dependencies); } } // namespace oneapi::math::dft diff --git a/include/oneapi/math/dft/detail/commit_impl.hpp b/include/oneapi/math/dft/detail/commit_impl.hpp index d3f7d7d8d..6aaa92580 100644 --- a/include/oneapi/math/dft/detail/commit_impl.hpp +++ b/include/oneapi/math/dft/detail/commit_impl.hpp @@ -54,18 +54,18 @@ class commit_impl { public: commit_impl(sycl::queue queue, math::backend backend, - const dft::detail::dft_values &config_values) + const dft::detail::dft_values& config_values) : queue_(queue), backend_(backend), external_workspace_helper_(config_values.workspace_placement == dft::detail::config_value::WORKSPACE_EXTERNAL) {} // rule of three - commit_impl(const commit_impl &other) = delete; - commit_impl &operator=(const commit_impl &other) = delete; + commit_impl(const commit_impl& other) = delete; + commit_impl& operator=(const commit_impl& other) = delete; virtual ~commit_impl() = default; - sycl::queue &get_queue() noexcept { + sycl::queue& get_queue() noexcept { return queue_; } @@ -73,9 +73,9 @@ class commit_impl { return backend_; } - virtual void *get_handle() noexcept = 0; + virtual void* get_handle() noexcept = 0; - virtual void commit(const dft_values &) = 0; + virtual void commit(const dft_values&) = 0; inline std::int64_t get_workspace_external_bytes() { return external_workspace_helper_.get_rqd_workspace_bytes(*this); @@ -87,54 +87,54 @@ class commit_impl { // When not overridden, external workspace support is faked: an external workspace can be set, // and errors will be generated according to the specificiation, // but the required workspace size will always be zero, and any given workspace will not actually be used. - virtual void set_workspace(scalar_type *usm_workspace) { + virtual void set_workspace(scalar_type* usm_workspace) { external_workspace_helper_.set_workspace_throw(*this, usm_workspace); } - virtual void set_workspace(sycl::buffer &buffer_workspace) { + virtual void set_workspace(sycl::buffer& buffer_workspace) { external_workspace_helper_.set_workspace_throw(*this, buffer_workspace); } - virtual void forward_ip_cc(descriptor_type &desc, sycl::buffer &inout) = 0; - virtual void forward_ip_rr(descriptor_type &desc, sycl::buffer &inout_re, - sycl::buffer &inout_im) = 0; - virtual void forward_op_cc(descriptor_type &desc, sycl::buffer &in, - sycl::buffer &out) = 0; - virtual void forward_op_rr(descriptor_type &desc, sycl::buffer &in_re, - sycl::buffer &in_im, - sycl::buffer &out_re, - sycl::buffer &out_im) = 0; - - virtual sycl::event forward_ip_cc(descriptor_type &desc, fwd_type *inout, - const std::vector &dependencies) = 0; - virtual sycl::event forward_ip_rr(descriptor_type &desc, scalar_type *inout_re, - scalar_type *inout_im, - const std::vector &dependencies) = 0; - virtual sycl::event forward_op_cc(descriptor_type &desc, fwd_type *in, bwd_type *out, - const std::vector &dependencies) = 0; - virtual sycl::event forward_op_rr(descriptor_type &desc, scalar_type *in_re, scalar_type *in_im, - scalar_type *out_re, scalar_type *out_im, - const std::vector &dependencies) = 0; - - virtual void backward_ip_cc(descriptor_type &desc, sycl::buffer &inout) = 0; - virtual void backward_ip_rr(descriptor_type &desc, sycl::buffer &inout_re, - sycl::buffer &inout_im) = 0; - virtual void backward_op_cc(descriptor_type &desc, sycl::buffer &in, - sycl::buffer &out) = 0; - virtual void backward_op_rr(descriptor_type &desc, sycl::buffer &in_re, - sycl::buffer &in_im, - sycl::buffer &out_re, - sycl::buffer &out_im) = 0; - - virtual sycl::event backward_ip_cc(descriptor_type &desc, fwd_type *inout, - const std::vector &dependencies) = 0; - virtual sycl::event backward_ip_rr(descriptor_type &desc, scalar_type *inout_re, - scalar_type *inout_im, - const std::vector &dependencies) = 0; - virtual sycl::event backward_op_cc(descriptor_type &desc, bwd_type *in, fwd_type *out, - const std::vector &dependencies) = 0; - virtual sycl::event backward_op_rr(descriptor_type &desc, scalar_type *in_re, - scalar_type *in_im, scalar_type *out_re, scalar_type *out_im, - const std::vector &dependencies) = 0; + virtual void forward_ip_cc(descriptor_type& desc, sycl::buffer& inout) = 0; + virtual void forward_ip_rr(descriptor_type& desc, sycl::buffer& inout_re, + sycl::buffer& inout_im) = 0; + virtual void forward_op_cc(descriptor_type& desc, sycl::buffer& in, + sycl::buffer& out) = 0; + virtual void forward_op_rr(descriptor_type& desc, sycl::buffer& in_re, + sycl::buffer& in_im, + sycl::buffer& out_re, + sycl::buffer& out_im) = 0; + + virtual sycl::event forward_ip_cc(descriptor_type& desc, fwd_type* inout, + const std::vector& dependencies) = 0; + virtual sycl::event forward_ip_rr(descriptor_type& desc, scalar_type* inout_re, + scalar_type* inout_im, + const std::vector& dependencies) = 0; + virtual sycl::event forward_op_cc(descriptor_type& desc, fwd_type* in, bwd_type* out, + const std::vector& dependencies) = 0; + virtual sycl::event forward_op_rr(descriptor_type& desc, scalar_type* in_re, scalar_type* in_im, + scalar_type* out_re, scalar_type* out_im, + const std::vector& dependencies) = 0; + + virtual void backward_ip_cc(descriptor_type& desc, sycl::buffer& inout) = 0; + virtual void backward_ip_rr(descriptor_type& desc, sycl::buffer& inout_re, + sycl::buffer& inout_im) = 0; + virtual void backward_op_cc(descriptor_type& desc, sycl::buffer& in, + sycl::buffer& out) = 0; + virtual void backward_op_rr(descriptor_type& desc, sycl::buffer& in_re, + sycl::buffer& in_im, + sycl::buffer& out_re, + sycl::buffer& out_im) = 0; + + virtual sycl::event backward_ip_cc(descriptor_type& desc, fwd_type* inout, + const std::vector& dependencies) = 0; + virtual sycl::event backward_ip_rr(descriptor_type& desc, scalar_type* inout_re, + scalar_type* inout_im, + const std::vector& dependencies) = 0; + virtual sycl::event backward_op_cc(descriptor_type& desc, bwd_type* in, fwd_type* out, + const std::vector& dependencies) = 0; + virtual sycl::event backward_op_rr(descriptor_type& desc, scalar_type* in_re, + scalar_type* in_im, scalar_type* out_re, scalar_type* out_im, + const std::vector& dependencies) = 0; /** For compute calls, throw errors for the external workspace as required. * @tparam ArgTs The non-descriptor arg(s) for the compute call. First one is used to check @@ -142,7 +142,7 @@ class commit_impl { * @param function_name The function name to user in generated exceptions. */ template - void compute_call_throw(const char *function_name) { + void compute_call_throw(const char* function_name) { external_workspace_helper_.template compute_call_throw(function_name); } @@ -151,14 +151,14 @@ class commit_impl { * @param function_name The function name to user in generated exceptions. * @param cgh The command group handler to associate the accessor with. */ - void add_buffer_workspace_dependency_if_rqd(const char *function_name, sycl::handler &cgh) { + void add_buffer_workspace_dependency_if_rqd(const char* function_name, sycl::handler& cgh) { external_workspace_helper_.add_buffer_dependency_if_rqd(function_name, cgh); } /** If WORKSPACE_EXTERNAL is set, depend on the last USM workspace event added via set_last_usm_workspace_event. * @param cgh The command group handler to associate the accessor with. */ - void depend_on_last_usm_workspace_event_if_rqd(sycl::handler &cgh) { + void depend_on_last_usm_workspace_event_if_rqd(sycl::handler& cgh) { external_workspace_helper_.depend_on_last_usm_workspace_event_if_rqd(cgh); } @@ -166,7 +166,7 @@ class commit_impl { * subsequent calls to depend_on_last_usm_workspace_event. * @param sycl_event The last usage of the USM workspace. */ - void set_last_usm_workspace_event_if_rqd(sycl::event &sycl_event) { + void set_last_usm_workspace_event_if_rqd(sycl::event& sycl_event) { external_workspace_helper_.set_last_usm_workspace_event_if_rqd(sycl_event); } diff --git a/include/oneapi/math/dft/detail/dft_ct.hxx b/include/oneapi/math/dft/detail/dft_ct.hxx index bf479bde1..e03298c5e 100644 --- a/include/oneapi/math/dft/detail/dft_ct.hxx +++ b/include/oneapi/math/dft/detail/dft_ct.hxx @@ -20,8 +20,8 @@ // Commit template -ONEMATH_EXPORT dft::detail::commit_impl *create_commit( - const dft::detail::descriptor &desc, sycl::queue &sycl_queue); +ONEMATH_EXPORT dft::detail::commit_impl* create_commit( + const dft::detail::descriptor& desc, sycl::queue& sycl_queue); // BUFFER version @@ -34,105 +34,107 @@ using bwd = typename detail::descriptor_info::backward_type; //In-place transform template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> &inout); +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& inout); //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> &inout_re, - sycl::buffer, 1> &inout_im); +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& inout_re, + sycl::buffer, 1>& inout_im); //Out-of-place transform template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, - sycl::buffer, 1> &out); +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& in, + sycl::buffer, 1>& out); //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> &in_re, - sycl::buffer, 1> &in_im, - sycl::buffer, 1> &out_re, - sycl::buffer, 1> &out_im); +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& in_re, + sycl::buffer, 1>& in_im, + sycl::buffer, 1>& out_re, + sycl::buffer, 1>& out_im); //USM version //In-place transform template -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, - const std::vector &dependencies); +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, fwd* inout, + const std::vector& dependencies); //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *inout_re, - scalar *inout_im, - const std::vector &dependencies); +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, scalar* inout_re, + scalar* inout_im, + const std::vector& dependencies); //Out-of-place transform template -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, - bwd *out, - const std::vector &dependencies); +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, fwd* in, + bwd* out, + const std::vector& dependencies); //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *in_re, - scalar *in_im, - scalar *out_re, - scalar *out_im, - const std::vector &dependencies); +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, scalar* in_re, + scalar* in_im, + scalar* out_re, + scalar* out_im, + const std::vector& dependencies); // BUFFER version //In-place transform template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &inout); +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& inout); //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &inout_re, - sycl::buffer, 1> &inout_im); +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& inout_re, + sycl::buffer, 1>& inout_im); //Out-of-place transform template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &in, - sycl::buffer, 1> &out); +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& in, + sycl::buffer, 1>& out); //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &in_re, - sycl::buffer, 1> &in_im, - sycl::buffer, 1> &out_re, - sycl::buffer, 1> &out_im); +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& in_re, + sycl::buffer, 1>& in_im, + sycl::buffer, 1>& out_re, + sycl::buffer, 1>& out_im); //USM version //In-place transform template -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, - const std::vector &dependencies); +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, fwd* inout, + const std::vector& dependencies); //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *inout_re, - scalar *inout_im, - const std::vector &dependencies); +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, + scalar* inout_re, + scalar* inout_im, + const std::vector& dependencies); //Out-of-place transform template -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, - fwd *out, - const std::vector &dependencies); +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, bwd* in, + fwd* out, + const std::vector& dependencies); //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *in_re, - scalar *in_im, - scalar *out_re, - scalar *out_im, - const std::vector &dependencies); +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, scalar* in_re, + scalar* in_im, + scalar* out_re, + scalar* out_im, + const std::vector& dependencies); diff --git a/include/oneapi/math/dft/detail/dft_loader.hpp b/include/oneapi/math/dft/detail/dft_loader.hpp index 41def11db..34fb2da56 100644 --- a/include/oneapi/math/dft/detail/dft_loader.hpp +++ b/include/oneapi/math/dft/detail/dft_loader.hpp @@ -42,7 +42,7 @@ class descriptor; template ONEMATH_EXPORT commit_impl* create_commit(const descriptor& desc, - sycl::queue& queue); + sycl::queue& queue); } // namespace detail } // namespace dft diff --git a/include/oneapi/math/dft/detail/external_workspace_helper.hpp b/include/oneapi/math/dft/detail/external_workspace_helper.hpp index 24bdf93c5..31d1d7ffc 100644 --- a/include/oneapi/math/dft/detail/external_workspace_helper.hpp +++ b/include/oneapi/math/dft/detail/external_workspace_helper.hpp @@ -91,7 +91,7 @@ class external_workspace_helper { void set_workspace_throw(commit_impl_t& committed_desc, scalar_t* usm_workspace) { if (get_rqd_workspace_bytes(committed_desc) > 0 && usm_workspace == nullptr) { throw math::invalid_argument("DFT", "set_workspace", - "Backend expected a non-null workspace pointer."); + "Backend expected a non-null workspace pointer."); } m_ext_workspace_rqd = true; m_workspace_type = ext_workspace_type::usm; @@ -110,7 +110,7 @@ class external_workspace_helper { } if (buffer_workspace.is_sub_buffer()) { throw math::invalid_argument("DFT", "set_workspace", - "Cannot use sub-buffers for workspace"); + "Cannot use sub-buffers for workspace"); return; } m_ext_workspace_rqd = true; diff --git a/include/oneapi/math/dft/detail/types_impl.hpp b/include/oneapi/math/dft/detail/types_impl.hpp index da17c762c..233a8bd0b 100644 --- a/include/oneapi/math/dft/detail/types_impl.hpp +++ b/include/oneapi/math/dft/detail/types_impl.hpp @@ -113,7 +113,7 @@ using valid_compute_arg = typename std::bool_constant< template constexpr bool valid_ip_realreal_impl = - is_complex_dft&& std::is_same_v, data_t>; + is_complex_dft && std::is_same_v, data_t>; // compute the range of a reinterpreted buffer template diff --git a/include/oneapi/math/dft/forward.hpp b/include/oneapi/math/dft/forward.hpp index 67f106923..297582423 100644 --- a/include/oneapi/math/dft/forward.hpp +++ b/include/oneapi/math/dft/forward.hpp @@ -34,7 +34,7 @@ namespace oneapi::math::dft { //In-place transform template -void compute_forward(descriptor_type &desc, sycl::buffer &inout) { +void compute_forward(descriptor_type& desc, sycl::buffer& inout) { static_assert(detail::valid_compute_arg::value, "unexpected type for data_type"); @@ -47,8 +47,8 @@ void compute_forward(descriptor_type &desc, sycl::buffer &inout) { //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template , bool> = true> -void compute_forward(descriptor_type &desc, sycl::buffer &inout_re, - sycl::buffer &inout_im) { +void compute_forward(descriptor_type& desc, sycl::buffer& inout_re, + sycl::buffer& inout_im) { static_assert(detail::valid_compute_arg::value, "unexpected type for data_type"); @@ -62,8 +62,8 @@ void compute_forward(descriptor_type &desc, sycl::buffer &inout_re //Out-of-place transform template -void compute_forward(descriptor_type &desc, sycl::buffer &in, - sycl::buffer &out) { +void compute_forward(descriptor_type& desc, sycl::buffer& in, + sycl::buffer& out) { static_assert(detail::valid_compute_arg::value, "unexpected type for input_type"); static_assert(detail::valid_compute_arg::value, @@ -80,9 +80,9 @@ void compute_forward(descriptor_type &desc, sycl::buffer &in, //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -void compute_forward(descriptor_type &desc, sycl::buffer &in_re, - sycl::buffer &in_im, sycl::buffer &out_re, - sycl::buffer &out_im) { +void compute_forward(descriptor_type& desc, sycl::buffer& in_re, + sycl::buffer& in_im, sycl::buffer& out_re, + sycl::buffer& out_im) { static_assert(detail::valid_compute_arg::value, "unexpected type for input_type"); static_assert(detail::valid_compute_arg::value, @@ -105,56 +105,56 @@ void compute_forward(descriptor_type &desc, sycl::buffer &in_re, //In-place transform template -sycl::event compute_forward(descriptor_type &desc, data_type *inout, - const std::vector &dependencies = {}) { +sycl::event compute_forward(descriptor_type& desc, data_type* inout, + const std::vector& dependencies = {}) { static_assert(detail::valid_compute_arg::value, "unexpected type for data_type"); using fwd_type = typename detail::descriptor_info::forward_type; - return get_commit(desc)->forward_ip_cc(desc, reinterpret_cast(inout), dependencies); + return get_commit(desc)->forward_ip_cc(desc, reinterpret_cast(inout), dependencies); } //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template , bool> = true> -sycl::event compute_forward(descriptor_type &desc, data_type *inout_re, data_type *inout_im, - const std::vector &dependencies = {}) { +sycl::event compute_forward(descriptor_type& desc, data_type* inout_re, data_type* inout_im, + const std::vector& dependencies = {}) { static_assert(detail::valid_compute_arg::value, "unexpected type for data_type"); using scalar_type = typename detail::descriptor_info::scalar_type; - return get_commit(desc)->forward_ip_rr(desc, reinterpret_cast(inout_re), - reinterpret_cast(inout_im), dependencies); + return get_commit(desc)->forward_ip_rr(desc, reinterpret_cast(inout_re), + reinterpret_cast(inout_im), dependencies); } //Out-of-place transform template -sycl::event compute_forward(descriptor_type &desc, input_type *in, output_type *out, - const std::vector &dependencies = {}) { +sycl::event compute_forward(descriptor_type& desc, input_type* in, output_type* out, + const std::vector& dependencies = {}) { static_assert(detail::valid_compute_arg::value, "unexpected type for input_type"); static_assert(detail::valid_compute_arg::value, "unexpected type for output_type"); using fwd_type = typename detail::descriptor_info::forward_type; using bwd_type = typename detail::descriptor_info::backward_type; - return get_commit(desc)->forward_op_cc(desc, reinterpret_cast(in), - reinterpret_cast(out), dependencies); + return get_commit(desc)->forward_op_cc(desc, reinterpret_cast(in), + reinterpret_cast(out), dependencies); } //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -sycl::event compute_forward(descriptor_type &desc, input_type *in_re, input_type *in_im, - output_type *out_re, output_type *out_im, - const std::vector &dependencies = {}) { +sycl::event compute_forward(descriptor_type& desc, input_type* in_re, input_type* in_im, + output_type* out_re, output_type* out_im, + const std::vector& dependencies = {}) { static_assert(detail::valid_compute_arg::value, "unexpected type for input_type"); static_assert(detail::valid_compute_arg::value, "unexpected type for output_type"); using scalar_type = typename detail::descriptor_info::scalar_type; - return get_commit(desc)->forward_op_rr(desc, reinterpret_cast(in_re), - reinterpret_cast(in_im), - reinterpret_cast(out_re), - reinterpret_cast(out_im), dependencies); + return get_commit(desc)->forward_op_rr(desc, reinterpret_cast(in_re), + reinterpret_cast(in_im), + reinterpret_cast(out_re), + reinterpret_cast(out_im), dependencies); } } // namespace oneapi::math::dft diff --git a/include/oneapi/math/exceptions.hpp b/include/oneapi/math/exceptions.hpp index ef99a1fb9..76f37b90f 100644 --- a/include/oneapi/math/exceptions.hpp +++ b/include/oneapi/math/exceptions.hpp @@ -39,7 +39,7 @@ class exception : public std::exception { public: exception(const std::string& message) : std::exception(), msg_(message) {} - exception(const std::string &domain, const std::string &function, const std::string &info = "") + exception(const std::string& domain, const std::string& function, const std::string& info = "") : std::exception() { msg_ = std::string("oneMath: ") + domain + ((domain.length() != 0 && function.length() != 0) ? "/" : "") + function + @@ -48,7 +48,7 @@ class exception : public std::exception { : ""); } - const char *what() const noexcept override { + const char* what() const noexcept override { return msg_.c_str(); } }; @@ -56,8 +56,8 @@ class exception : public std::exception { class unsupported_device : public oneapi::math::exception { public: unsupported_device(const std::string& message) : exception(message) {} - unsupported_device(const std::string &domain, const std::string &function, - const sycl::device &device) + unsupported_device(const std::string& domain, const std::string& function, + const sycl::device& device) : oneapi::math::exception( domain, function, device.get_info() + " is not supported") {} @@ -66,15 +66,15 @@ class unsupported_device : public oneapi::math::exception { class host_bad_alloc : public oneapi::math::exception { public: host_bad_alloc(const std::string& message) : exception(message) {} - host_bad_alloc(const std::string &domain, const std::string &function) + host_bad_alloc(const std::string& domain, const std::string& function) : oneapi::math::exception(domain, function, "cannot allocate memory on host") {} }; class device_bad_alloc : public oneapi::math::exception { public: device_bad_alloc(const std::string& message) : exception(message) {} - device_bad_alloc(const std::string &domain, const std::string &function, - const sycl::device &device) + device_bad_alloc(const std::string& domain, const std::string& function, + const sycl::device& device) : oneapi::math::exception( domain, function, "cannot allocate memory on " + device.get_info()) {} @@ -83,33 +83,33 @@ class device_bad_alloc : public oneapi::math::exception { class unimplemented : public oneapi::math::exception { public: unimplemented(const std::string& message) : exception(message) {} - unimplemented(const std::string &domain, const std::string &function, - const std::string &info = "") + unimplemented(const std::string& domain, const std::string& function, + const std::string& info = "") : oneapi::math::exception(domain, function, "function is not implemented " + info) {} }; class invalid_argument : public oneapi::math::exception { public: invalid_argument(const std::string& message) : exception(message) {} - invalid_argument(const std::string &domain, const std::string &function, - const std::string &info = "") + invalid_argument(const std::string& domain, const std::string& function, + const std::string& info = "") : oneapi::math::exception(domain, function, "invalid argument " + info) {} }; class uninitialized : public oneapi::math::exception { public: uninitialized(const std::string& message) : exception(message) {} - uninitialized(const std::string &domain, const std::string &function, - const std::string &info = "") + uninitialized(const std::string& domain, const std::string& function, + const std::string& info = "") : oneapi::math::exception(domain, function, - "handle/descriptor is not initialized " + info) {} + "handle/descriptor is not initialized " + info) {} }; class computation_error : public oneapi::math::exception { public: computation_error(const std::string& message) : exception(message) {} - computation_error(const std::string &domain, const std::string &function, - const std::string &info = "") + computation_error(const std::string& domain, const std::string& function, + const std::string& info = "") : oneapi::math::exception( domain, function, "computation error" + ((info.length() != 0) ? (": " + info) : "")) {} @@ -118,10 +118,10 @@ class computation_error : public oneapi::math::exception { class batch_error : public oneapi::math::exception { public: batch_error(const std::string& message) : exception(message) {} - batch_error(const std::string &domain, const std::string &function, - const std::string &info = "") - : oneapi::math::exception(domain, function, - "batch error" + ((info.length() != 0) ? (": " + info) : "")) {} + batch_error(const std::string& domain, const std::string& function, + const std::string& info = "") + : oneapi::math::exception( + domain, function, "batch error" + ((info.length() != 0) ? (": " + info) : "")) {} }; } // namespace math diff --git a/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hxx b/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hxx index 175e4500e..62998cae2 100644 --- a/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hxx +++ b/include/oneapi/math/lapack/detail/cusolver/lapack_ct.hxx @@ -20,2293 +20,2309 @@ // Buffer APIs static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, - sycl::buffer &taup, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, sycl::buffer& tauq, + sycl::buffer& taup, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); -} -static inline void getrs(backend_selector selector, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, - scratchpad, scratchpad_size); -} -static inline void getrs(backend_selector selector, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, - scratchpad, scratchpad_size); -} -static inline void getrs(backend_selector selector, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, - scratchpad, scratchpad_size); -} -static inline void getrs(backend_selector selector, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, - scratchpad, scratchpad_size); + scratchpad_size); +} +static inline void getrs(backend_selector selector, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, + ldb, scratchpad, scratchpad_size); +} +static inline void getrs(backend_selector selector, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, + ldb, scratchpad, scratchpad_size); +} +static inline void getrs(backend_selector selector, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, + ldb, scratchpad, scratchpad_size); +} +static inline void getrs(backend_selector selector, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, + ldb, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, - std::int64_t ldvt, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& s, + sycl::buffer& u, std::int64_t ldu, sycl::buffer& vt, + std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, - vt, ldvt, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, + ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, - std::int64_t ldvt, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& s, + sycl::buffer& u, std::int64_t ldu, sycl::buffer& vt, + std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, - vt, ldvt, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, + ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, - std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, + std::int64_t ldu, sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, - vt, ldvt, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, + ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, - std::int64_t ldu, sycl::buffer> &vt, - std::int64_t ldvt, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, + std::int64_t ldu, sycl::buffer>& vt, + std::int64_t ldvt, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, - vt, ldvt, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, + ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void heevd(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, - scratchpad_size); + oneapi::math::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + scratchpad, scratchpad_size); } static inline void heevd(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, - scratchpad_size); + oneapi::math::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, - w, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, + ldb, w, scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, - w, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, + ldb, w, scratchpad, scratchpad_size); } static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void orgbr(backend_selector selector, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void orgbr(backend_selector selector, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void ormtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, - c, ldc, scratchpad, scratchpad_size); + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, + tau, c, ldc, scratchpad, scratchpad_size); } static inline void ormtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, - c, ldc, scratchpad, scratchpad_size); + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, + tau, c, ldc, scratchpad, scratchpad_size); } static inline void ormrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void ormrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void ormqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void ormqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void syevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, - scratchpad_size); + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + scratchpad, scratchpad_size); } static inline void syevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, - scratchpad_size); + oneapi::math::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, - w, scratchpad, scratchpad_size); + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, + ldb, w, scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, - w, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, + ldb, w, scratchpad, scratchpad_size); } static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void unmrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void unmrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void unmqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void unmqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void unmtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, - c, ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, + tau, c, ldc, scratchpad, scratchpad_size); } static inline void unmtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, - c, ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, + tau, c, ldc, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, + sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, + sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer> &b, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer> &b, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, + sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, + sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void orgqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void orgqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrf_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, - batch_size, scratchpad, scratchpad_size); + batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrf_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, - batch_size, scratchpad, scratchpad_size); + batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, +static inline void potrf_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, - batch_size, scratchpad, scratchpad_size); + batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, +static inline void potrf_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, - batch_size, scratchpad, scratchpad_size); + batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrs_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size); -} -static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + stride_a, b, ldb, stride_b, batch_size, scratchpad, + scratchpad_size); +} +static inline void potrs_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size); + stride_a, b, ldb, stride_b, batch_size, scratchpad, + scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &b, +static inline void potrs_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size); + stride_a, b, ldb, stride_b, batch_size, scratchpad, + scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &b, +static inline void potrs_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size); + stride_a, b, ldb, stride_b, batch_size, scratchpad, + scratchpad_size); } static inline void ungqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, + std::int64_t n, std::int64_t k, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, + sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void ungqr_batch(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } // USM APIs static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, float *d, - float *e, std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::complex* a, std::int64_t lda, float* d, + float* e, std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, - taup, scratchpad, scratchpad_size, dependencies); + taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *d, double *e, - double *tauq, double *taup, double *scratchpad, + std::int64_t n, double* a, std::int64_t lda, double* d, double* e, + double* tauq, double* taup, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, - taup, scratchpad, scratchpad_size, dependencies); + taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *d, float *e, - float *tauq, float *taup, float *scratchpad, + std::int64_t n, float* a, std::int64_t lda, float* d, float* e, + float* tauq, float* taup, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, - taup, scratchpad, scratchpad_size, dependencies); + taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - double *d, double *e, std::complex *tauq, - std::complex *taup, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + double* d, double* e, std::complex* tauq, + std::complex* taup, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, - taup, scratchpad, scratchpad_size, dependencies); + taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + std::int64_t n, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + std::int64_t n, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::gerqf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + std::int64_t n, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + std::int64_t n, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::geqrf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, double* a, std::int64_t lda, std::int64_t* ipiv, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, float* a, std::int64_t lda, std::int64_t* ipiv, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, + double* a, std::int64_t lda, std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, + float* a, std::int64_t lda, std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size, dependencies); + b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, std::int64_t *ipiv, double *b, - std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* a, std::int64_t lda, std::int64_t* ipiv, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size, dependencies); + b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - float *a, std::int64_t lda, std::int64_t *ipiv, float *b, - std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* a, std::int64_t lda, std::int64_t* ipiv, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size, dependencies); + b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size, dependencies); + b, ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *s, double *u, - std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* s, double* u, std::int64_t ldu, double* vt, + std::int64_t ldvt, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size, - dependencies); + u, ldu, vt, ldvt, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *s, float *u, - std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* s, float* u, std::int64_t ldu, float* vt, std::int64_t ldvt, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size, - dependencies); + u, ldu, vt, ldvt, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, float *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, float* s, std::complex* u, + std::int64_t ldu, std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size, - dependencies); + u, ldu, vt, ldvt, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - double *s, std::complex *u, std::int64_t ldu, - std::complex *vt, std::int64_t ldvt, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, double* s, std::complex* u, + std::int64_t ldu, std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size, - dependencies); + u, ldu, vt, ldvt, scratchpad, scratchpad_size, + dependencies); } -static inline sycl::event heevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, float *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event heevd(backend_selector selector, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, float* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } -static inline sycl::event heevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, double *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event heevd(backend_selector selector, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, double* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, float *w, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, float* w, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size, - dependencies); + b, ldb, w, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, double *w, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, double* w, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size, - dependencies); + b, ldb, w, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, float *d, float *e, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, float* d, float* e, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, double *d, double *e, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, double* d, double* e, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* a, std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ormtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* a, std::int64_t lda, double* tau, double* c, + std::int64_t ldc, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ormrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ormrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, double *tau, double *c, std::int64_t ldc, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, double* tau, double* c, std::int64_t ldc, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ormqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, double *tau, double *c, std::int64_t ldc, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, double* tau, double* c, std::int64_t ldc, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ormqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, float* b, std::int64_t ldb, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, + ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, double *b, std::int64_t ldb, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); + double* a, std::int64_t lda, double* b, std::int64_t ldb, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, + ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, + ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); -} -static inline sycl::event syevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *w, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, + ldb, scratchpad, scratchpad_size, dependencies); +} +static inline sycl::event syevd(backend_selector selector, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* w, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } -static inline sycl::event syevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *w, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event syevd(backend_selector selector, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* w, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* a, std::int64_t lda, double* b, std::int64_t ldb, double* w, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size, - dependencies); + b, ldb, w, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* a, std::int64_t lda, float* b, std::int64_t ldb, float* w, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size, - dependencies); + b, ldb, w, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *d, double *e, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* d, double* e, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *d, float *e, float *tau, float *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* d, float* e, float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, + a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, double *b, std::int64_t ldb, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size, - dependencies); + double* a, std::int64_t lda, double* b, std::int64_t ldb, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, + a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size, - dependencies); + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, float* b, std::int64_t ldb, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, + a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::cusolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, + a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event unmrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event unmqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event unmqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event unmtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event unmtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, + std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size, dependencies); + tau, stride_tau, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, + std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size, dependencies); + tau, stride_tau, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size, dependencies); + tau, stride_tau, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + tau, stride_tau, batch_size, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, float **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, +static inline sycl::event getrf_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, double **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, +static inline sycl::event getrf_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event getrf_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event getrf_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, std::int64_t n, - float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, float *scratchpad, + float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, std::int64_t n, - double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, double *scratchpad, + double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event getri_batch(backend_selector selector, std::int64_t *n, - float **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event getri_batch(backend_selector selector, std::int64_t* n, + float** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event getri_batch(backend_selector selector, std::int64_t *n, - double **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event getri_batch(backend_selector selector, std::int64_t* n, + double** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event getri_batch(backend_selector selector, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event getri_batch(backend_selector selector, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event getri_batch(backend_selector selector, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event getri_batch(backend_selector selector, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, float *b, std::int64_t ldb, + std::int64_t nrhs, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, double *b, std::int64_t ldb, + std::int64_t nrhs, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}) { + std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}) { + std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, - std::int64_t **ipiv, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, float** a, std::int64_t* lda, + std::int64_t** ipiv, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + ipiv, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, - std::int64_t **ipiv, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, double** a, std::int64_t* lda, + std::int64_t** ipiv, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + ipiv, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + ipiv, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( - backend_selector selector, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + backend_selector selector, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + ipiv, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgqr_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float **a, - std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, +static inline sycl::event orgqr_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float** a, + std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event orgqr_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::int64_t *k, double **a, - std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, +static inline sycl::event orgqr_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, std::int64_t* k, double** a, + std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, float *scratchpad, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, + oneapi::math::uplo uplo, std::int64_t n, double* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, double *scratchpad, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - float *a, std::int64_t lda, std::int64_t stride_a, float *b, + float* a, std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, float *scratchpad, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, std::int64_t stride_a, double *b, + double* a, std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, double *scratchpad, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - float **a, std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + float** a, std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + b, ldb, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - double **a, std::int64_t *lda, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + double** a, std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + b, ldb, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + b, ldb, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + b, ldb, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, + std::int64_t n, std::int64_t k, std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, + std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, + std::int64_t n, std::int64_t k, std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, + std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size, dependencies); -} -static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::int64_t *k, std::complex **a, - std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); +} +static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::complex** a, + std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::int64_t *k, std::complex **a, - std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::complex** a, + std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::cusolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } // SCRATCHPAD APIs template std::int64_t gebrd_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::gebrd_scratchpad_size(selector.get_queue(), m, n, - lda); + return oneapi::math::lapack::cusolver::gebrd_scratchpad_size(selector.get_queue(), m, + n, lda); } template std::int64_t gerqf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::gerqf_scratchpad_size(selector.get_queue(), m, n, - lda); + return oneapi::math::lapack::cusolver::gerqf_scratchpad_size(selector.get_queue(), m, + n, lda); } template std::int64_t geqrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::geqrf_scratchpad_size(selector.get_queue(), m, n, - lda); + return oneapi::math::lapack::cusolver::geqrf_scratchpad_size(selector.get_queue(), m, + n, lda); } template std::int64_t gesvd_scratchpad_size(backend_selector selector, @@ -2319,28 +2335,28 @@ std::int64_t gesvd_scratchpad_size(backend_selector selector, template std::int64_t getrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::getrf_scratchpad_size(selector.get_queue(), m, n, - lda); + return oneapi::math::lapack::cusolver::getrf_scratchpad_size(selector.get_queue(), m, + n, lda); } template std::int64_t getri_scratchpad_size(backend_selector selector, std::int64_t n, std::int64_t lda) { return oneapi::math::lapack::cusolver::getri_scratchpad_size(selector.get_queue(), n, - lda); + lda); } template std::int64_t getrs_scratchpad_size(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return oneapi::math::lapack::cusolver::getrs_scratchpad_size(selector.get_queue(), - trans, n, nrhs, lda, ldb); + trans, n, nrhs, lda, ldb); } template std::int64_t heevd_scratchpad_size(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::heevd_scratchpad_size(selector.get_queue(), jobz, - uplo, n, lda); + return oneapi::math::lapack::cusolver::heevd_scratchpad_size(selector.get_queue(), + jobz, uplo, n, lda); } template std::int64_t hegvd_scratchpad_size(backend_selector selector, std::int64_t itype, @@ -2352,49 +2368,49 @@ std::int64_t hegvd_scratchpad_size(backend_selector selector, template std::int64_t hetrd_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::hetrd_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::cusolver::hetrd_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t hetrf_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::hetrf_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::cusolver::hetrf_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t orgbr_scratchpad_size(backend_selector selector, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::cusolver::orgbr_scratchpad_size(selector.get_queue(), vect, - m, n, k, lda); + return oneapi::math::lapack::cusolver::orgbr_scratchpad_size(selector.get_queue(), + vect, m, n, k, lda); } template std::int64_t orgtr_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::orgtr_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::cusolver::orgtr_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t orgqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::cusolver::orgqr_scratchpad_size(selector.get_queue(), m, n, - k, lda); + return oneapi::math::lapack::cusolver::orgqr_scratchpad_size(selector.get_queue(), m, + n, k, lda); } template std::int64_t ormrq_scratchpad_size(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::math::lapack::cusolver::ormrq_scratchpad_size(selector.get_queue(), side, - trans, m, n, k, lda, ldc); + return oneapi::math::lapack::cusolver::ormrq_scratchpad_size( + selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t ormqr_scratchpad_size(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::math::lapack::cusolver::ormqr_scratchpad_size(selector.get_queue(), side, - trans, m, n, k, lda, ldc); + return oneapi::math::lapack::cusolver::ormqr_scratchpad_size( + selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t ormtr_scratchpad_size(backend_selector selector, @@ -2407,34 +2423,34 @@ std::int64_t ormtr_scratchpad_size(backend_selector selector, template std::int64_t potrf_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::potrf_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::cusolver::potrf_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t potrs_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::math::lapack::cusolver::potrs_scratchpad_size(selector.get_queue(), uplo, - n, nrhs, lda, ldb); + return oneapi::math::lapack::cusolver::potrs_scratchpad_size(selector.get_queue(), + uplo, n, nrhs, lda, ldb); } template std::int64_t potri_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::potri_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::cusolver::potri_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t sytrf_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::sytrf_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::cusolver::sytrf_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t syevd_scratchpad_size(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::syevd_scratchpad_size(selector.get_queue(), jobz, - uplo, n, lda); + return oneapi::math::lapack::cusolver::syevd_scratchpad_size(selector.get_queue(), + jobz, uplo, n, lda); } template std::int64_t sygvd_scratchpad_size(backend_selector selector, std::int64_t itype, @@ -2446,8 +2462,8 @@ std::int64_t sygvd_scratchpad_size(backend_selector selector, template std::int64_t sytrd_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::sytrd_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::cusolver::sytrd_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t trtrs_scratchpad_size(backend_selector selector, @@ -2461,36 +2477,36 @@ template std::int64_t ungbr_scratchpad_size(backend_selector selector, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::cusolver::ungbr_scratchpad_size(selector.get_queue(), vect, - m, n, k, lda); + return oneapi::math::lapack::cusolver::ungbr_scratchpad_size(selector.get_queue(), + vect, m, n, k, lda); } template std::int64_t ungqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::cusolver::ungqr_scratchpad_size(selector.get_queue(), m, n, - k, lda); + return oneapi::math::lapack::cusolver::ungqr_scratchpad_size(selector.get_queue(), m, + n, k, lda); } template std::int64_t ungtr_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::cusolver::ungtr_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::cusolver::ungtr_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t unmrq_scratchpad_size(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::math::lapack::cusolver::unmrq_scratchpad_size(selector.get_queue(), side, - trans, m, n, k, lda, ldc); + return oneapi::math::lapack::cusolver::unmrq_scratchpad_size( + selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t unmqr_scratchpad_size(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::math::lapack::cusolver::unmqr_scratchpad_size(selector.get_queue(), side, - trans, m, n, k, lda, ldc); + return oneapi::math::lapack::cusolver::unmqr_scratchpad_size( + selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t unmtr_scratchpad_size(backend_selector selector, @@ -2566,62 +2582,62 @@ std::int64_t ungqr_batch_scratchpad_size(backend_selector sel } template std::int64_t getrf_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::cusolver::getrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } template std::int64_t getri_batch_scratchpad_size(backend_selector selector, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::cusolver::getri_batch_scratchpad_size( selector.get_queue(), n, lda, group_count, group_sizes); } template std::int64_t getrs_batch_scratchpad_size(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes) { + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::cusolver::getrs_batch_scratchpad_size( selector.get_queue(), trans, n, nrhs, lda, ldb, group_count, group_sizes); } template std::int64_t geqrf_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::cusolver::geqrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } template std::int64_t orgqr_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return oneapi::math::lapack::cusolver::orgqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } template std::int64_t potrf_batch_scratchpad_size(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return oneapi::math::lapack::cusolver::potrf_batch_scratchpad_size( selector.get_queue(), uplo, n, lda, group_count, group_sizes); } template std::int64_t potrs_batch_scratchpad_size(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes) { + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::cusolver::potrs_batch_scratchpad_size( selector.get_queue(), uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template std::int64_t ungqr_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return oneapi::math::lapack::cusolver::ungqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } diff --git a/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx index ae8e8a147..a3d4de61c 100644 --- a/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx +++ b/include/oneapi/math/lapack/detail/cusolver/onemath_lapack_cusolver.hxx @@ -19,1812 +19,1836 @@ // Buffer APIs -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tauq, - sycl::buffer &taup, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tauq, + sycl::buffer& taup, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tauq, + sycl::buffer& taup, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); // USM APIs -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *d, double *e, double *tauq, double *taup, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *d, float *e, float *tauq, float *taup, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, - double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, - float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, - double *vt, std::int64_t ldvt, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, - std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, float *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, float *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - double *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - float *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - double *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, - std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, - std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, float *b, - std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, double *b, - std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, double *a, std::int64_t lda, double *w, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, float *a, std::int64_t lda, float *w, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *b, std::int64_t ldb, double *w, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *b, std::int64_t ldb, float *w, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *d, double *e, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *d, float *e, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, - double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, - float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, float *tau, - std::int64_t stride_tau, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, double *tau, - std::int64_t stride_tau, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - float **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - double **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, float **a, - std::int64_t *lda, std::int64_t **ipiv, float **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, double **a, - std::int64_t *lda, std::int64_t **ipiv, double **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - float **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - double **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::complex **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* d, double* e, double* tauq, double* taup, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* d, float* e, float* tauq, float* taup, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, + std::int64_t* ipiv, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t* ipiv, + double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t* ipiv, + float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, double* s, double* u, + std::int64_t ldu, double* vt, std::int64_t ldvt, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + float* a, std::int64_t lda, float* s, float* u, std::int64_t ldu, + float* vt, std::int64_t ldvt, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, float* s, + std::complex* u, std::int64_t ldu, std::complex* vt, + std::int64_t ldvt, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, double* s, + std::complex* u, std::int64_t ldu, + std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, float* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, double* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + float* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + double* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, float* tau, float* c, + std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, float* tau, float* c, + std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* w, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* w, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* b, std::int64_t ldb, double* w, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* b, std::int64_t ldb, float* w, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* d, double* e, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* d, float* e, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, std::int64_t* ipiv, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, double* a, std::int64_t lda, + double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, float* a, std::int64_t lda, + float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, float* tau, + std::int64_t stride_tau, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + float** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + double** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, float** a, + std::int64_t* lda, std::int64_t** ipiv, float** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, double** a, + std::int64_t* lda, std::int64_t** ipiv, double** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, float** a, + std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, double** a, + std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch( + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); // SCRATCHPAD APIs template -ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::int64_t lda, std::int64_t ldu, - std::int64_t ldvt); +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t ldu, std::int64_t ldvt); template -ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, + std::int64_t lda); template -ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template -ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda, - std::int64_t ldb); +ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda, + std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, - std::int64_t m, std::int64_t n, std::int64_t k, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, + std::int64_t m, std::int64_t n, std::int64_t k, + std::int64_t lda); template -ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda); +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template -ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda, - std::int64_t ldb); +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda, + std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, - std::int64_t ldb); +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, std::int64_t lda, + std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, - std::int64_t m, std::int64_t n, std::int64_t k, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, + std::int64_t m, std::int64_t n, std::int64_t k, + std::int64_t lda); template -ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda); +ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t stride_a, - std::int64_t stride_ipiv, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t stride_a, + std::int64_t stride_ipiv, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, - std::int64_t lda, std::int64_t stride_a, - std::int64_t stride_ipiv, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, + std::int64_t lda, std::int64_t stride_a, + std::int64_t stride_ipiv, + std::int64_t batch_size); template ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t stride_a, - std::int64_t stride_tau, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t stride_a, + std::int64_t stride_tau, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda, - std::int64_t stride_a, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda, + std::int64_t stride_a, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t stride_a, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t stride_a, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda, std::int64_t stride_a, - std::int64_t stride_tau, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, + std::int64_t lda, std::int64_t stride_a, + std::int64_t stride_tau, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda, std::int64_t stride_a, - std::int64_t stride_tau, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, + std::int64_t lda, std::int64_t stride_a, + std::int64_t stride_tau, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); diff --git a/include/oneapi/math/lapack/detail/lapack_loader.hpp b/include/oneapi/math/lapack/detail/lapack_loader.hpp index 7e0da894b..be5015257 100644 --- a/include/oneapi/math/lapack/detail/lapack_loader.hpp +++ b/include/oneapi/math/lapack/detail/lapack_loader.hpp @@ -38,2344 +38,2356 @@ namespace math { namespace lapack { namespace detail { -ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, - sycl::buffer &taup, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, - std::int64_t ldvt, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, - std::int64_t ldvt, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, - std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, - std::int64_t ldu, sycl::buffer> &vt, - std::int64_t ldvt, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, float *d, - float *e, std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *d, double *e, - double *tauq, double *taup, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *d, float *e, - float *tauq, float *taup, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - double *d, double *e, std::complex *tauq, - std::complex *taup, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, +ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, sycl::buffer& tauq, + sycl::buffer& taup, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, + std::int64_t ldu, sycl::buffer>& vt, + std::int64_t ldvt, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gesvd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, + std::int64_t ldu, sycl::buffer>& vt, + std::int64_t ldvt, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void heevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void heevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hegvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hegvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hetrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hetrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hetrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hetrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgbr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgbr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void syevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void syevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sygvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void sygvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungbr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungbr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, std::int64_t *ipiv, double *b, - std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - float *a, std::int64_t lda, std::int64_t *ipiv, float *b, - std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *s, double *u, - std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *s, float *u, - std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, float *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - double *s, std::complex *u, std::int64_t ldu, - std::complex *vt, std::int64_t ldvt, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, float *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, double *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, float *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, double *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, float *d, float *e, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, double *d, double *e, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, double *tau, double *c, std::int64_t ldc, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, double *tau, double *c, std::int64_t ldc, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, double *b, std::int64_t ldb, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *w, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *w, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *d, double *e, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *d, float *e, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, double *b, std::int64_t ldb, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, float **a, - std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, double **a, - std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, float **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, double **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, float* d, + float* e, std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, double* a, std::int64_t lda, double* d, double* e, + double* tauq, double* taup, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, float* a, std::int64_t lda, float* d, float* e, + float* tauq, float* taup, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, + double* d, double* e, std::complex* tauq, + std::complex* taup, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, double* a, std::int64_t lda, std::int64_t* ipiv, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, float* a, std::int64_t lda, std::int64_t* ipiv, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double* a, std::int64_t lda, std::int64_t* ipiv, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + float* a, std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + double* a, std::int64_t lda, std::int64_t* ipiv, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, std::int64_t* ipiv, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* s, double* u, std::int64_t ldu, double* vt, + std::int64_t ldvt, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* s, float* u, std::int64_t ldu, float* vt, std::int64_t ldvt, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, float* s, std::complex* u, + std::int64_t ldu, std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gesvd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, double* s, std::complex* u, + std::int64_t ldu, std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event heevd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, float* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event heevd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, double* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hegvd(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + float* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hegvd(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + double* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hetrd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, float* d, float* e, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hetrd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, double* d, double* e, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hetrf(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hetrf(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgbr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgbr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgtr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgtr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormtr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + float* a, std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormtr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, double* tau, double* c, + std::int64_t ldc, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormrq(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormrq(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, double* tau, double* c, std::int64_t ldc, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormqr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, double* tau, double* c, std::int64_t ldc, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormqr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potri(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, float* b, std::int64_t ldb, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + double* a, std::int64_t lda, double* b, std::int64_t ldb, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syevd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* w, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syevd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* w, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sygvd(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* b, std::int64_t ldb, double* w, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sygvd(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* b, std::int64_t ldb, float* w, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* d, double* e, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrd(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* d, float* e, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrf(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + double* a, std::int64_t lda, double* b, std::int64_t ldb, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, float* b, std::int64_t ldb, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trtrs(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungbr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungbr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungtr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungtr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmrq(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmrq(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmqr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmqr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmtr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmtr(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, + std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, + std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event getrs_batch( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, float** a, std::int64_t* lda, + std::int64_t** ipiv, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, double** a, std::int64_t* lda, + std::int64_t** ipiv, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event getrs_batch( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, - std::int64_t **ipiv, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, - std::int64_t **ipiv, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event getrs_batch( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, std::int64_t stride_a, float *tau, - std::int64_t stride_tau, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, std::int64_t stride_a, double *tau, - std::int64_t stride_tau, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, float **a, - std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, double **a, - std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - float *a, std::int64_t lda, std::int64_t stride_a, float *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, std::int64_t stride_a, double *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - float **a, std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - double **a, std::int64_t *lda, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, std::int64_t stride_a, float* tau, + std::int64_t stride_tau, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, std::int64_t stride_a, double* tau, + std::int64_t stride_tau, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, float** a, + std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, std::int64_t stride_a, float* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + double* a, std::int64_t lda, std::int64_t stride_a, + double* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, double** a, std::int64_t* lda, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungqr_batch( + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); ONEMATH_EXPORT sycl::event ungqr_batch( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); template = nullptr> -std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, +std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, +std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, +std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template = nullptr> -std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template = nullptr> -std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, +std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t getri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, +std::int64_t getri_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t heevd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t heevd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hegvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t hegvd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t hetrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t hetrd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hetrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t hetrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, +std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t potri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potri_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t ungbr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ungbr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ungqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, +std::int64_t ungqr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ungtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ungtr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t unmrq_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t unmrq_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t unmqr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t unmtr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template = nullptr> -std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template = nullptr> -std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> -std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template = nullptr> -std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ungqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> -std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template = nullptr> -std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template = nullptr> -std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes); template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template = nullptr> -std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template = nullptr> -std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes); template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +std::int64_t ungqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - std::int64_t m, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, + std::int64_t m, + std::int64_t n, + std::int64_t lda); +template <> ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - std::int64_t m, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, + std::int64_t m, + std::int64_t n, + std::int64_t lda); +template <> ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - std::int64_t m, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, + std::int64_t m, + std::int64_t n, + std::int64_t lda); +template <> ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, - oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t ldu, std::int64_t ldvt); -template <> -ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, + std::int64_t ldvt); +template <> ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - std::int64_t m, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, + std::int64_t m, + std::int64_t n, + std::int64_t lda); +template <> ONEMATH_EXPORT std::int64_t getri_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t n, - std::int64_t lda); + sycl::queue& queue, std::int64_t n, + std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t getri_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t n, - std::int64_t lda); + sycl::queue& queue, std::int64_t n, + std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, - oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, + sycl::queue& queue, + oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t ldb); +template <> ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); +template <> ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, - oneapi::math::generate vect, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda); + sycl::queue& queue, + oneapi::math::generate vect, + std::int64_t m, std::int64_t n, + std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, - oneapi::math::generate vect, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda); + sycl::queue& queue, + oneapi::math::generate vect, + std::int64_t m, std::int64_t n, + std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, + sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc); -template <> -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc); -template <> -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc); -template <> -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t lda, std::int64_t ldc); -template <> -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t lda, std::int64_t ldc); +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); +template <> +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); +template <> +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); +template <> +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); +template <> +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t lda, std::int64_t ldc); +template <> +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); +template <> ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::int64_t lda, + std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::int64_t lda, + std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t potri_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t potri_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> +ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); +template <> ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::job jobz, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, + sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, + sycl::queue& queue, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, + sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, + sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda, std::int64_t ldb); +template <> ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, - std::int64_t ldb); +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, - std::int64_t ldb); +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(oneapi::math::device libkey, + sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); +template <> ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, - std::int64_t ldc); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, - std::int64_t ldc); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes); + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes); } //namespace detail } //namespace lapack } //namespace math diff --git a/include/oneapi/math/lapack/detail/lapack_rt.hpp b/include/oneapi/math/lapack/detail/lapack_rt.hpp index 8f7aa92f6..a84331ceb 100644 --- a/include/oneapi/math/lapack/detail/lapack_rt.hpp +++ b/include/oneapi/math/lapack/detail/lapack_rt.hpp @@ -38,2132 +38,2137 @@ namespace oneapi { namespace math { namespace lapack { -static inline void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, +static inline void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::gebrd(get_device_id(queue), queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -static inline void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tauq, - sycl::buffer &taup, sycl::buffer &scratchpad, +static inline void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tauq, + sycl::buffer& taup, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::gebrd(get_device_id(queue), queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -static inline void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::gebrd(get_device_id(queue), queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -static inline void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, +static inline void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::gebrd(get_device_id(queue), queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -static inline void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::gerqf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::gerqf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::gerqf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::gerqf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::geqrf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::geqrf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::geqrf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::geqrf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getrf(get_device_id(queue), queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, +static inline void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getrf(get_device_id(queue), queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getrf(get_device_id(queue), queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getrf(get_device_id(queue), queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getri(get_device_id(queue), queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getri(get_device_id(queue), queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getri(get_device_id(queue), queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getri(get_device_id(queue), queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, +static inline void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, +static inline void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, +static inline void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -static inline void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, +static inline void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, +static inline void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -static inline void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, +static inline void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::heevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, +static inline void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::heevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, +static inline void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::hegvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, +static inline void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::hegvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::hetrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::hetrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::hetrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::hetrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, +static inline void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::orgbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, +static inline void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::orgbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::orgqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::orgqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::orgtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::orgtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, +static inline void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::ormtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, +static inline void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::ormtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, +static inline void ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::ormrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, +static inline void ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::ormrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, +static inline void ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::ormqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, +static inline void ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::ormqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, +static inline void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, +static inline void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, +static inline void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, +static inline void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, +static inline void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, +static inline void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -static inline void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, +static inline void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, +static inline void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, +static inline void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, +static inline void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::syevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, +static inline void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::syevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -static inline void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, +static inline void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::sygvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, +static inline void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::sygvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -static inline void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::sytrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::sytrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -static inline void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, +static inline void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -static inline void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, +static inline void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, +static inline void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, +static inline void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, +static inline void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -static inline void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::ungbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::ungbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::ungqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::ungqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::ungtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::ungtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -static inline void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, +static inline void unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::unmrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, +static inline void unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::unmrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, +static inline void unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::unmqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, +static inline void unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::unmqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, +static inline void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::unmtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, +static inline void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::unmtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, +static inline void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -static inline void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, +static inline void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -static inline void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, +static inline void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -static inline void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, +static inline void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -static inline void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +static inline void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getri_batch(get_device_id(queue), queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -static inline void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, +static inline void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getri_batch(get_device_id(queue), queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -static inline void getri_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, +static inline void getri_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getri_batch(get_device_id(queue), queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -static inline void getri_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, +static inline void getri_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getri_batch(get_device_id(queue), queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -static inline void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, +static inline void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, +static inline void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, +static inline void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, +static inline void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +static inline void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -static inline void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +static inline void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -static inline void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, +static inline void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -static inline void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, +static inline void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -static inline void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, +static inline void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::orgqr_batch(get_device_id(queue), queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -static inline void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, +static inline void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::orgqr_batch(get_device_id(queue), queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, +static inline void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, +static inline void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, +static inline void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, +static inline void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, +static inline void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, +static inline void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, +static inline void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, +static inline void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -static inline void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, +static inline void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::ungqr_batch(get_device_id(queue), queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -static inline void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, +static inline void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { detail::ungqr_batch(get_device_id(queue), queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -static inline sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::gebrd(get_device_id(queue), queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *d, double *e, double *tauq, double *taup, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* d, double* e, double* tauq, double* taup, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::gebrd(get_device_id(queue), queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *d, float *e, float *tauq, float *taup, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* d, float* e, float* tauq, float* taup, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::gebrd(get_device_id(queue), queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::gebrd(get_device_id(queue), queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, +static inline sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::gerqf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, +static inline sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::gerqf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::gerqf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, +static inline sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::gerqf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::geqrf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, +static inline sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::geqrf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, +static inline sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::geqrf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, +static inline sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::geqrf(get_device_id(queue), queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrf(get_device_id(queue), queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t *ipiv, double *scratchpad, +static inline sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrf(get_device_id(queue), queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t *ipiv, float *scratchpad, +static inline sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrf(get_device_id(queue), queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrf(get_device_id(queue), queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getri(get_device_id(queue), queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, +static inline sycl::event getri(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, + std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getri(get_device_id(queue), queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getri(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getri(get_device_id(queue), queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getri(get_device_id(queue), queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, - double *b, std::int64_t ldb, double *scratchpad, +static inline sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t* ipiv, + double* b, std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, - float *b, std::int64_t ldb, float *scratchpad, +static inline sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t* ipiv, + float* b, std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrs(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, +static inline sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, - double *vt, std::int64_t ldvt, double *scratchpad, + double* a, std::int64_t lda, double* s, double* u, std::int64_t ldu, + double* vt, std::int64_t ldvt, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, - std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + float* a, std::int64_t lda, float* s, float* u, std::int64_t ldu, + float* vt, std::int64_t ldvt, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, +static inline sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, float *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, + std::complex* a, std::int64_t lda, float* s, + std::complex* u, std::int64_t ldu, std::complex* vt, + std::int64_t ldvt, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, +static inline sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, + std::complex* a, std::int64_t lda, double* s, + std::complex* u, std::int64_t ldu, std::complex* vt, + std::int64_t ldvt, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::gesvd(get_device_id(queue), queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, float *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, float* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::heevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - double *w, std::complex *scratchpad, +static inline sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, + double* w, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::heevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - float *w, std::complex *scratchpad, +static inline sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + float* w, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::hegvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - double *w, std::complex *scratchpad, +static inline sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + double* w, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::hegvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tau, std::complex *scratchpad, +static inline sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::hetrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tau, std::complex *scratchpad, +static inline sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::hetrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::hetrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::hetrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::orgbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::orgbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - double *a, std::int64_t lda, double *tau, double *scratchpad, +static inline sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + double* a, std::int64_t lda, double* tau, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::orgqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - float *a, std::int64_t lda, float *tau, float *scratchpad, +static inline sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + float* a, std::int64_t lda, float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::orgqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *scratchpad, +static inline sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::orgtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *scratchpad, +static inline sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* tau, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::orgtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::ormtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::ormtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, +static inline sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, - std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, float* a, std::int64_t lda, float* tau, float* c, + std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::ormrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, +static inline sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, double* a, std::int64_t lda, double* tau, double* c, + std::int64_t ldc, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::ormrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, +static inline sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, double* a, std::int64_t lda, double* tau, double* c, + std::int64_t ldc, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::ormqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, +static inline sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, - std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, float* a, std::int64_t lda, float* tau, float* c, + std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::ormqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *scratchpad, +static inline sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *scratchpad, +static inline sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potrf(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *scratchpad, +static inline sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *scratchpad, +static inline sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potri(get_device_id(queue), queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, float *b, - std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, double *b, - std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potrs(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, double *a, std::int64_t lda, double *w, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, double* a, std::int64_t lda, double* w, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::syevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, float *a, std::int64_t lda, float *w, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, float* a, std::int64_t lda, float* w, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::syevd(get_device_id(queue), queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *b, std::int64_t ldb, double *w, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* b, std::int64_t ldb, double* w, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::sygvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *b, std::int64_t ldb, float *w, float *scratchpad, +static inline sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* b, std::int64_t ldb, float* w, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::sygvd(get_device_id(queue), queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *d, double *e, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* d, double* e, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::sytrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *d, float *e, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* d, float* e, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::sytrd(get_device_id(queue), queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, +static inline sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, +static inline sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::sytrf(get_device_id(queue), queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, +static inline sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, +static inline sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, - double *b, std::int64_t ldb, double *scratchpad, + std::int64_t n, std::int64_t nrhs, double* a, std::int64_t lda, + double* b, std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, +static inline sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, - float *b, std::int64_t ldb, float *scratchpad, + std::int64_t n, std::int64_t nrhs, float* a, std::int64_t lda, + float* b, std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, +static inline sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::trtrs(get_device_id(queue), queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::ungbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::ungbr(get_device_id(queue), queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::ungqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, +static inline sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::ungqr(get_device_id(queue), queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::ungtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, +static inline sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::ungtr(get_device_id(queue), queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, +static inline sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::unmrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, +static inline sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::unmrq(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, +static inline sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::unmqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, +static inline sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::unmqr(get_device_id(queue), queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::unmtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::unmtr(get_device_id(queue), queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, float *tau, +static inline sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, float* tau, std::int64_t stride_tau, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, double *tau, +static inline sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, double* tau, std::int64_t stride_tau, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, +static inline sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, +static inline sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::geqrf_batch(get_device_id(queue), queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, +static inline sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, +static inline sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, +static inline sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, +static inline sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - float **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + float** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - double **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + double** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrf_batch(get_device_id(queue), queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, +static inline sycl::event getri_batch(sycl::queue& queue, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getri_batch(get_device_id(queue), queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, +static inline sycl::event getri_batch(sycl::queue& queue, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getri_batch(get_device_id(queue), queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, +static inline sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getri_batch(get_device_id(queue), queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, +static inline sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getri_batch(get_device_id(queue), queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getri_batch(get_device_id(queue), queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getri_batch(get_device_id(queue), queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getri_batch(get_device_id(queue), queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, +static inline sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getri_batch(get_device_id(queue), queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, float *b, std::int64_t ldb, +static inline sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, double *b, std::int64_t ldb, +static inline sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, +static inline sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, +static inline sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, float **a, - std::int64_t *lda, std::int64_t **ipiv, float **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, +static inline sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, float** a, + std::int64_t* lda, std::int64_t** ipiv, float** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, double **a, - std::int64_t *lda, std::int64_t **ipiv, double **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, +static inline sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, double** a, + std::int64_t* lda, std::int64_t** ipiv, double** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, +static inline sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::getrs_batch(get_device_id(queue), queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, +static inline sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::orgqr_batch(get_device_id(queue), queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, +static inline sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::orgqr_batch(get_device_id(queue), queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::orgqr_batch(get_device_id(queue), queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::orgqr_batch(get_device_id(queue), queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, float *scratchpad, +static inline sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, double *scratchpad, +static inline sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, +static inline sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, +static inline sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - float **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, +static inline sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + float** a, std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - double **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, +static inline sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + double** a, std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, +static inline sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrf_batch(get_device_id(queue), queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, +static inline sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, +static inline sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, +static inline sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, +static inline sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, +static inline sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, +static inline sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, double** a, std::int64_t* lda, double** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::complex **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::potrs_batch(get_device_id(queue), queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, +static inline sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return detail::ungqr_batch(get_device_id(queue), queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, +static inline sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::ungqr_batch(get_device_id(queue), queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::ungqr_batch(get_device_id(queue), queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -static inline sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return detail::ungqr_batch(get_device_id(queue), queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } template = nullptr> -std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gebrd_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return detail::gebrd_scratchpad_size(get_device_id(queue), queue, m, n, lda); } template = nullptr> -std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return detail::gerqf_scratchpad_size(get_device_id(queue), queue, m, n, lda); } template = nullptr> -std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return detail::geqrf_scratchpad_size(get_device_id(queue), queue, m, n, lda); } template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, +std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { return detail::gesvd_scratchpad_size(get_device_id(queue), queue, jobu, jobvt, m, n, lda, ldu, ldvt); } template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, +std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { return detail::gesvd_scratchpad_size(get_device_id(queue), queue, jobu, jobvt, m, n, lda, ldu, ldvt); } template = nullptr> -std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return detail::getrf_scratchpad_size(get_device_id(queue), queue, m, n, lda); } template = nullptr> -std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda) { +std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda) { return detail::getri_scratchpad_size(get_device_id(queue), queue, n, lda); } template = nullptr> -std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { +std::int64_t getrs_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::int64_t lda, + std::int64_t ldb) { return detail::getrs_scratchpad_size(get_device_id(queue), queue, trans, n, nrhs, lda, ldb); } template = nullptr> -std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, +std::int64_t heevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::heevd_scratchpad_size(get_device_id(queue), queue, jobz, uplo, n, lda); } template = nullptr> -std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, +std::int64_t hegvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { return detail::hegvd_scratchpad_size(get_device_id(queue), queue, itype, jobz, uplo, n, lda, ldb); } template = nullptr> -std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t hetrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::hetrd_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t hetrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::hetrf_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, +std::int64_t orgbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return detail::orgbr_scratchpad_size(get_device_id(queue), queue, vect, m, n, k, lda); } template = nullptr> -std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t orgtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::orgtr_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return detail::orgqr_scratchpad_size(get_device_id(queue), queue, m, n, k, lda); } template = nullptr> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return detail::ormrq_scratchpad_size(get_device_id(queue), queue, side, trans, m, n, k, lda, ldc); } template = nullptr> -std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return detail::ormqr_scratchpad_size(get_device_id(queue), queue, side, trans, m, n, k, @@ -2171,7 +2176,7 @@ std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, } template = nullptr> -std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { @@ -2179,45 +2184,45 @@ std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, n, lda, ldc); } template = nullptr> -std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t potrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::potrf_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t potrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return detail::potrs_scratchpad_size(get_device_id(queue), queue, uplo, n, nrhs, lda, ldb); } template = nullptr> -std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t potri_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::potri_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t sytrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::sytrf_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, +std::int64_t syevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::syevd_scratchpad_size(get_device_id(queue), queue, jobz, uplo, n, lda); } template = nullptr> -std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, +std::int64_t sygvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { return detail::sygvd_scratchpad_size(get_device_id(queue), queue, itype, jobz, uplo, n, lda, ldb); } template = nullptr> -std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t sytrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::sytrd_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t trtrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { @@ -2226,31 +2231,31 @@ std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, } template = nullptr> -std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, +std::int64_t ungbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return detail::ungbr_scratchpad_size(get_device_id(queue), queue, vect, m, n, k, lda); } template = nullptr> -std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t ungqr_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return detail::ungqr_scratchpad_size(get_device_id(queue), queue, m, n, k, lda); } template = nullptr> -std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t ungtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return detail::ungtr_scratchpad_size(get_device_id(queue), queue, uplo, n, lda); } template = nullptr> -std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return detail::unmrq_scratchpad_size(get_device_id(queue), queue, side, trans, m, n, k, lda, ldc); } template = nullptr> -std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { return detail::unmqr_scratchpad_size(get_device_id(queue), queue, side, trans, m, n, k, @@ -2258,7 +2263,7 @@ std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, } template = nullptr> -std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { @@ -2266,21 +2271,21 @@ std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, n, lda, ldc); } template = nullptr> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return detail::getrf_batch_scratchpad_size(get_device_id(queue), queue, m, n, lda, stride_a, stride_ipiv, batch_size); } template = nullptr> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return detail::getri_batch_scratchpad_size(get_device_id(queue), queue, n, lda, stride_a, stride_ipiv, batch_size); } template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, @@ -2290,30 +2295,30 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::trans batch_size); } template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return detail::geqrf_batch_scratchpad_size(get_device_id(queue), queue, m, n, lda, stride_a, stride_tau, batch_size); } template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda, std::int64_t stride_a, +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { return detail::potrf_batch_scratchpad_size(get_device_id(queue), queue, uplo, n, lda, stride_a, batch_size); } template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size) { +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, std::int64_t lda, + std::int64_t stride_a, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size) { return detail::potrs_batch_scratchpad_size(get_device_id(queue), queue, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return detail::orgqr_batch_scratchpad_size(get_device_id(queue), queue, m, n, k, lda, @@ -2321,68 +2326,68 @@ std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std } template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return detail::ungqr_batch_scratchpad_size(get_device_id(queue), queue, m, n, k, lda, stride_a, stride_tau, batch_size); } template = nullptr> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return detail::getrf_batch_scratchpad_size(get_device_id(queue), queue, m, n, lda, group_count, group_sizes); } template = nullptr> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return detail::getri_batch_scratchpad_size(get_device_id(queue), queue, n, lda, group_count, group_sizes); } template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes) { return detail::getrs_batch_scratchpad_size(get_device_id(queue), queue, trans, n, nrhs, lda, ldb, group_count, group_sizes); } template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return detail::geqrf_batch_scratchpad_size(get_device_id(queue), queue, m, n, lda, group_count, group_sizes); } template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return detail::orgqr_batch_scratchpad_size(get_device_id(queue), queue, m, n, k, lda, group_count, group_sizes); } template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return detail::potrf_batch_scratchpad_size(get_device_id(queue), queue, uplo, n, lda, group_count, group_sizes); } template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes) { return detail::potrs_batch_scratchpad_size(get_device_id(queue), queue, uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { +std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return detail::ungqr_batch_scratchpad_size(get_device_id(queue), queue, m, n, k, lda, group_count, group_sizes); } diff --git a/include/oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx b/include/oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx index 41e7e8753..1dd86d2ed 100644 --- a/include/oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx +++ b/include/oneapi/math/lapack/detail/mkl_common/lapack_ct.hxx @@ -18,2315 +18,2345 @@ *******************************************************************************/ static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, - sycl::buffer &taup, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, sycl::buffer& tauq, + sycl::buffer& taup, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + taup, scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, + taup, scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, + scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, + scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, + scratchpad, scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, + scratchpad, scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size); + u, ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size); + u, ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, - std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, + std::int64_t ldu, sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size); + u, ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, - std::int64_t ldu, sycl::buffer> &vt, - std::int64_t ldvt, sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, + std::int64_t ldu, sycl::buffer>& vt, + std::int64_t ldvt, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size); + u, ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void heevd(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void heevd(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size); + b, ldb, w, scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size); + b, ldb, w, scratchpad, scratchpad_size); } -static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void hetrd(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void hetrd(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void hetrf(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void hetrf(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size) { +static inline void orgtr(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size) { +static inline void orgtr(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void ormtr(backend_selector selector, oneapi::math::side side, - oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void ormtr(backend_selector selector, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size); + lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormtr(backend_selector selector, oneapi::math::side side, - oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void ormtr(backend_selector selector, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size); + lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(backend_selector selector, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void ormrq(backend_selector selector, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size); + tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormrq(backend_selector selector, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void ormrq(backend_selector selector, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size); + tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(backend_selector selector, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void ormqr(backend_selector selector, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size); + tau, c, ldc, scratchpad, scratchpad_size); } -static inline void ormqr(backend_selector selector, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void ormqr(backend_selector selector, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size); + tau, c, ldc, scratchpad, scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrf(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrf(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, +static inline void potrf(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } -static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, +static inline void potrf(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potri(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potri(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, +static inline void potri(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } -static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, +static inline void potri(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrs(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrs(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, +static inline void potrs(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, +static inline void potrs(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void syevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void syevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size); + b, ldb, w, scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size); + b, ldb, w, scratchpad, scratchpad_size); } -static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tau, sycl::buffer &scratchpad, +static inline void sytrd(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void sytrd(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size) { +static inline void sytrf(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size) { +static inline void sytrf(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void sytrf(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, +static inline void sytrf(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, +static inline void trtrs(backend_selector selector, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size); + lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void trtrs(backend_selector selector, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size); + lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, +static inline void trtrs(backend_selector selector, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size); + lda, b, ldb, scratchpad, scratchpad_size); } -static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, +static inline void trtrs(backend_selector selector, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size); + lda, b, ldb, scratchpad, scratchpad_size); } static inline void ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void ungtr(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, +static inline void ungtr(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } -static inline void unmrq(backend_selector selector, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, +static inline void unmrq(backend_selector selector, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size); + tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmrq(backend_selector selector, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, +static inline void unmrq(backend_selector selector, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size); + tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(backend_selector selector, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, +static inline void unmqr(backend_selector selector, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size); + tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmqr(backend_selector selector, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, +static inline void unmqr(backend_selector selector, + oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size); + tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(backend_selector selector, oneapi::math::side side, - oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, +static inline void unmtr(backend_selector selector, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size); + lda, tau, c, ldc, scratchpad, scratchpad_size); } -static inline void unmtr(backend_selector selector, oneapi::math::side side, - oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, +static inline void unmtr(backend_selector selector, + oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size); + lda, tau, c, ldc, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size); + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size); + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, + sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size); + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, + sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size); + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer> &b, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer> &b, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, + sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, + sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void orgqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size); + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size); } static inline void orgqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size); + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size); } static inline void potrf_batch(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrf_batch( selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } static inline void potrf_batch(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrf_batch( selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrf_batch( selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrf_batch( selector.get_queue(), uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size); } static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size); } static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &b, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size); } static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &b, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size); } static inline void ungqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, + std::int64_t n, std::int64_t k, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, + sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size); + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size); } static inline void ungqr_batch(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size); + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, float *d, - float *e, std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::complex* a, std::int64_t lda, float* d, + float* e, std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, - tauq, taup, scratchpad, scratchpad_size, - dependencies); + tauq, taup, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *d, double *e, - double *tauq, double *taup, double *scratchpad, + std::int64_t n, double* a, std::int64_t lda, double* d, double* e, + double* tauq, double* taup, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, - tauq, taup, scratchpad, scratchpad_size, - dependencies); + tauq, taup, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *d, float *e, - float *tauq, float *taup, float *scratchpad, + std::int64_t n, float* a, std::int64_t lda, float* d, float* e, + float* tauq, float* taup, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, - tauq, taup, scratchpad, scratchpad_size, - dependencies); + tauq, taup, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - double *d, double *e, std::complex *tauq, - std::complex *taup, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + double* d, double* e, std::complex* tauq, + std::complex* taup, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gebrd(selector.get_queue(), m, n, a, lda, d, e, - tauq, taup, scratchpad, scratchpad_size, - dependencies); + tauq, taup, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gerqf(selector.get_queue(), m, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::geqrf(selector.get_queue(), m, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, double* a, std::int64_t lda, std::int64_t* ipiv, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, float* a, std::int64_t lda, std::int64_t* ipiv, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, + double* a, std::int64_t lda, std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, + float* a, std::int64_t lda, std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri(selector.get_queue(), n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, scratchpad, scratchpad_size, - dependencies); + ipiv, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, std::int64_t *ipiv, double *b, - std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* a, std::int64_t lda, std::int64_t* ipiv, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, scratchpad, scratchpad_size, - dependencies); + ipiv, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - float *a, std::int64_t lda, std::int64_t *ipiv, float *b, - std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* a, std::int64_t lda, std::int64_t* ipiv, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, scratchpad, scratchpad_size, - dependencies); + ipiv, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, scratchpad, scratchpad_size, - dependencies); + ipiv, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *s, double *u, - std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* s, double* u, std::int64_t ldu, double* vt, + std::int64_t ldvt, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, - lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size, dependencies); + lda, s, u, ldu, vt, ldvt, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *s, float *u, - std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* s, float* u, std::int64_t ldu, float* vt, std::int64_t ldvt, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, - lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size, dependencies); + lda, s, u, ldu, vt, ldvt, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, float *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, float* s, std::complex* u, + std::int64_t ldu, std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, - lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size, dependencies); + lda, s, u, ldu, vt, ldvt, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - double *s, std::complex *u, std::int64_t ldu, - std::complex *vt, std::int64_t ldvt, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, double* s, std::complex* u, + std::int64_t ldu, std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, - lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size, dependencies); + lda, s, u, ldu, vt, ldvt, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event heevd(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, float *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, - w, scratchpad, scratchpad_size, dependencies); + std::complex* a, std::int64_t lda, float* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::heevd( + selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event heevd(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, double *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::heevd(selector.get_queue(), jobz, uplo, n, a, lda, - w, scratchpad, scratchpad_size, dependencies); + std::complex* a, std::int64_t lda, double* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::heevd( + selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, float *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, - lda, b, ldb, w, scratchpad, scratchpad_size, - dependencies); + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, float* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, + a, lda, b, ldb, w, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, double *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, - lda, b, ldb, w, scratchpad, scratchpad_size, - dependencies); + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, double* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::hegvd(selector.get_queue(), itype, jobz, uplo, n, + a, lda, b, ldb, w, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, float *d, float *e, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, float* d, float* e, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, - tau, scratchpad, scratchpad_size, - dependencies); + tau, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, double *d, double *e, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, double* d, double* e, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, - tau, scratchpad, scratchpad_size, - dependencies); + tau, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::orgbr( selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::orgbr( selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::orgqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::orgtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, - a, lda, tau, c, ldc, scratchpad, - scratchpad_size, dependencies); + float* a, std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, + n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event ormtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, n, - a, lda, tau, c, ldc, scratchpad, - scratchpad_size, dependencies); + double* a, std::int64_t lda, double* tau, double* c, + std::int64_t ldc, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::ormtr(selector.get_queue(), side, uplo, trans, m, + n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event ormrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, + a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event ormrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, double *tau, double *c, std::int64_t ldc, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, double* tau, double* c, std::int64_t ldc, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::ormrq(selector.get_queue(), side, trans, m, n, k, + a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event ormqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, double *tau, double *c, std::int64_t ldc, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, double* tau, double* c, std::int64_t ldc, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, + a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event ormqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::ormqr(selector.get_queue(), side, trans, m, n, k, + a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf(selector.get_queue(), uplo, n, a, lda, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potri(selector.get_queue(), uplo, n, a, lda, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, float* b, std::int64_t ldb, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size, - dependencies); + b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, double *b, std::int64_t ldb, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* a, std::int64_t lda, double* b, std::int64_t ldb, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size, - dependencies); + b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size, - dependencies); + b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size, - dependencies); + b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event syevd(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *w, double *scratchpad, + double* a, std::int64_t lda, double* w, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, - w, scratchpad, scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::syevd( + selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event syevd(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *w, float *scratchpad, + float* a, std::int64_t lda, float* w, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::syevd(selector.get_queue(), jobz, uplo, n, a, lda, - w, scratchpad, scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::syevd( + selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } static inline sycl::event sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, double *a, std::int64_t lda, double *b, - std::int64_t ldb, double *w, double *scratchpad, + std::int64_t n, double* a, std::int64_t lda, double* b, + std::int64_t ldb, double* w, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, - lda, b, ldb, w, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, + a, lda, b, ldb, w, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, float *a, std::int64_t lda, float *b, - std::int64_t ldb, float *w, float *scratchpad, + std::int64_t n, float* a, std::int64_t lda, float* b, + std::int64_t ldb, float* w, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, - lda, b, ldb, w, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::sygvd(selector.get_queue(), itype, jobz, uplo, n, + a, lda, b, ldb, w, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *d, double *e, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* d, double* e, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, - tau, scratchpad, scratchpad_size, - dependencies); + tau, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *d, float *e, float *tau, float *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* d, float* e, float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, - tau, scratchpad, scratchpad_size, - dependencies); + tau, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, - nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies); + nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, double *b, std::int64_t ldb, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* a, std::int64_t lda, double* b, std::int64_t ldb, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, - nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies); + nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, float* b, std::int64_t ldb, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, - nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies); + nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::trtrs(selector.get_queue(), uplo, trans, diag, n, - nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies); + nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::ungbr( selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::ungbr( selector.get_queue(), vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::ungqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::ungtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, + a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event unmrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::unmrq(selector.get_queue(), side, trans, m, n, k, + a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event unmqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, + a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event unmqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::unmqr(selector.get_queue(), side, trans, m, n, k, + a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event unmtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, - a, lda, tau, c, ldc, scratchpad, - scratchpad_size, dependencies); + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, + n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event unmtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, n, - a, lda, tau, c, ldc, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::unmtr(selector.get_queue(), side, uplo, trans, m, + n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, - std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, - std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, - std::int64_t m, std::int64_t n, std::complex *a, + std::int64_t m, std::int64_t n, std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, + std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, - std::int64_t m, std::int64_t n, std::complex *a, + std::int64_t m, std::int64_t n, std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, + std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( selector.get_queue(), m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, float **a, - std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, + std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( + selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, double **a, - std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, + std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( + selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( + selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch( + selector.get_queue(), m, n, a, lda, tau, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, - std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, - std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, - std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, - std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch( selector.get_queue(), m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, - std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri_batch( selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, - std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri_batch( selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri_batch( selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri_batch( selector.get_queue(), n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, - std::int64_t *n, float **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, + std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, - std::int64_t *n, double **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, + std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, float *b, std::int64_t ldb, + std::int64_t nrhs, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, double *b, std::int64_t ldb, + std::int64_t nrhs, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, - std::int64_t **ipiv, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, float** a, std::int64_t* lda, + std::int64_t** ipiv, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, - std::int64_t **ipiv, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, double** a, std::int64_t* lda, + std::int64_t** ipiv, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( - backend_selector selector, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + backend_selector selector, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr_batch(backend_selector selector, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, std::int64_t stride_a, float *tau, + std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, std::int64_t stride_a, float* tau, std::int64_t stride_tau, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch( selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr_batch(backend_selector selector, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, std::int64_t stride_a, double *tau, + std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, std::int64_t stride_a, double* tau, std::int64_t stride_tau, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch( selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *k, float **a, - std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, + std::int64_t* m, std::int64_t* n, std::int64_t* k, float** a, + std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch( selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *k, double **a, - std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, + std::int64_t* m, std::int64_t* n, std::int64_t* k, double** a, + std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch( selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, float *scratchpad, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, + oneapi::math::uplo uplo, std::int64_t n, double* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, double *scratchpad, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - float *a, std::int64_t lda, std::int64_t stride_a, float *b, + float* a, std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, float *scratchpad, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, std::int64_t stride_a, double *b, + double* a, std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, double *scratchpad, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - float **a, std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + float** a, std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - double **a, std::int64_t *lda, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + double** a, std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch( selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch( selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch( selector.get_queue(), m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch( selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr_batch(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch( selector.get_queue(), m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); @@ -2335,20 +2365,20 @@ static inline sycl::event ungqr_batch(backend_selector template = nullptr> std::int64_t gebrd_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::gebrd_scratchpad_size(selector.get_queue(), - m, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::gebrd_scratchpad_size( + selector.get_queue(), m, n, lda); } template = nullptr> std::int64_t gerqf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::gerqf_scratchpad_size(selector.get_queue(), - m, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::gerqf_scratchpad_size( + selector.get_queue(), m, n, lda); } template = nullptr> std::int64_t geqrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::geqrf_scratchpad_size(selector.get_queue(), - m, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::geqrf_scratchpad_size( + selector.get_queue(), m, n, lda); } template = nullptr> std::int64_t gesvd_scratchpad_size(backend_selector selector, @@ -2369,14 +2399,14 @@ std::int64_t gesvd_scratchpad_size(backend_selector sel template = nullptr> std::int64_t getrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::getrf_scratchpad_size(selector.get_queue(), - m, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::getrf_scratchpad_size( + selector.get_queue(), m, n, lda); } template = nullptr> std::int64_t getri_scratchpad_size(backend_selector selector, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::getri_scratchpad_size(selector.get_queue(), - n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::getri_scratchpad_size( + selector.get_queue(), n, lda); } template = nullptr> std::int64_t getrs_scratchpad_size(backend_selector selector, @@ -2389,8 +2419,8 @@ template = nullp std::int64_t heevd_scratchpad_size(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::heevd_scratchpad_size(selector.get_queue(), - jobz, uplo, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::heevd_scratchpad_size( + selector.get_queue(), jobz, uplo, n, lda); } template = nullptr> std::int64_t hegvd_scratchpad_size(backend_selector selector, @@ -2403,37 +2433,37 @@ std::int64_t hegvd_scratchpad_size(backend_selector sel template = nullptr> std::int64_t hetrd_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::hetrd_scratchpad_size(selector.get_queue(), - uplo, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::hetrd_scratchpad_size( + selector.get_queue(), uplo, n, lda); } template = nullptr> std::int64_t hetrf_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::hetrf_scratchpad_size(selector.get_queue(), - uplo, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::hetrf_scratchpad_size( + selector.get_queue(), uplo, n, lda); } template = nullptr> std::int64_t orgbr_scratchpad_size(backend_selector selector, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::orgbr_scratchpad_size(selector.get_queue(), - vect, m, n, k, lda); + return oneapi::math::lapack::LAPACK_BACKEND::orgbr_scratchpad_size( + selector.get_queue(), vect, m, n, k, lda); } template = nullptr> std::int64_t orgtr_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::orgtr_scratchpad_size(selector.get_queue(), - uplo, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::orgtr_scratchpad_size( + selector.get_queue(), uplo, n, lda); } template = nullptr> std::int64_t orgqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::orgqr_scratchpad_size(selector.get_queue(), - m, n, k, lda); + return oneapi::math::lapack::LAPACK_BACKEND::orgqr_scratchpad_size( + selector.get_queue(), m, n, k, lda); } template = nullptr> @@ -2464,8 +2494,8 @@ std::int64_t ormtr_scratchpad_size(backend_selector sel template = nullptr> std::int64_t potrf_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::potrf_scratchpad_size(selector.get_queue(), - uplo, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::potrf_scratchpad_size( + selector.get_queue(), uplo, n, lda); } template = nullptr> std::int64_t potrs_scratchpad_size(backend_selector selector, @@ -2477,21 +2507,21 @@ std::int64_t potrs_scratchpad_size(backend_selector sel template = nullptr> std::int64_t potri_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::potri_scratchpad_size(selector.get_queue(), - uplo, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::potri_scratchpad_size( + selector.get_queue(), uplo, n, lda); } template = nullptr> std::int64_t sytrf_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::sytrf_scratchpad_size(selector.get_queue(), - uplo, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::sytrf_scratchpad_size( + selector.get_queue(), uplo, n, lda); } template = nullptr> std::int64_t syevd_scratchpad_size(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::syevd_scratchpad_size(selector.get_queue(), - jobz, uplo, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::syevd_scratchpad_size( + selector.get_queue(), jobz, uplo, n, lda); } template = nullptr> std::int64_t sygvd_scratchpad_size(backend_selector selector, @@ -2504,8 +2534,8 @@ std::int64_t sygvd_scratchpad_size(backend_selector sel template = nullptr> std::int64_t sytrd_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::sytrd_scratchpad_size(selector.get_queue(), - uplo, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::sytrd_scratchpad_size( + selector.get_queue(), uplo, n, lda); } template = nullptr> std::int64_t trtrs_scratchpad_size(backend_selector selector, @@ -2520,22 +2550,22 @@ template selector, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::ungbr_scratchpad_size(selector.get_queue(), - vect, m, n, k, lda); + return oneapi::math::lapack::LAPACK_BACKEND::ungbr_scratchpad_size( + selector.get_queue(), vect, m, n, k, lda); } template = nullptr> std::int64_t ungqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::ungqr_scratchpad_size(selector.get_queue(), - m, n, k, lda); + return oneapi::math::lapack::LAPACK_BACKEND::ungqr_scratchpad_size( + selector.get_queue(), m, n, k, lda); } template = nullptr> std::int64_t ungtr_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::LAPACK_BACKEND::ungtr_scratchpad_size(selector.get_queue(), - uplo, n, lda); + return oneapi::math::lapack::LAPACK_BACKEND::ungtr_scratchpad_size( + selector.get_queue(), uplo, n, lda); } template = nullptr> @@ -2631,64 +2661,64 @@ std::int64_t ungqr_batch_scratchpad_size(backend_selector = nullptr> std::int64_t getrf_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::LAPACK_BACKEND::getrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } template = nullptr> std::int64_t getri_batch_scratchpad_size(backend_selector selector, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::LAPACK_BACKEND::getri_batch_scratchpad_size( selector.get_queue(), n, lda, group_count, group_sizes); } template = nullptr> std::int64_t getrs_batch_scratchpad_size(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes) { + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::LAPACK_BACKEND::getrs_batch_scratchpad_size( selector.get_queue(), trans, n, nrhs, lda, ldb, group_count, group_sizes); } template = nullptr> std::int64_t geqrf_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::LAPACK_BACKEND::geqrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } template = nullptr> std::int64_t orgqr_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return oneapi::math::lapack::LAPACK_BACKEND::orgqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } template = nullptr> std::int64_t potrf_batch_scratchpad_size(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return oneapi::math::lapack::LAPACK_BACKEND::potrf_batch_scratchpad_size( selector.get_queue(), uplo, n, lda, group_count, group_sizes); } template = nullptr> std::int64_t potrs_batch_scratchpad_size(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes) { + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::LAPACK_BACKEND::potrs_batch_scratchpad_size( selector.get_queue(), uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template = nullptr> std::int64_t ungqr_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return oneapi::math::lapack::LAPACK_BACKEND::ungqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } diff --git a/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx b/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx index 5c1bdb095..15f59a7c7 100644 --- a/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx +++ b/include/oneapi/math/lapack/detail/mkl_common/onemath_lapack_backends.hxx @@ -17,2123 +17,2153 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tauq, - sycl::buffer &taup, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *d, double *e, double *tauq, double *taup, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *d, float *e, float *tauq, float *taup, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, - double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, - float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, - double *vt, std::int64_t ldvt, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, - std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, float *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, float *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - double *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - float *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - double *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, - std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, - std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, float *b, - std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, double *b, - std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, double *a, std::int64_t lda, double *w, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, float *a, std::int64_t lda, float *w, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *b, std::int64_t ldb, double *w, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *b, std::int64_t ldb, float *w, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *d, double *e, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *d, float *e, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, - double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, - float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, float *tau, - std::int64_t stride_tau, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, double *tau, - std::int64_t stride_tau, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - float **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - double **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, float **a, - std::int64_t *lda, std::int64_t **ipiv, float **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, double **a, - std::int64_t *lda, std::int64_t **ipiv, double **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - float **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - double **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::complex **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tauq, + sycl::buffer& taup, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tauq, + sycl::buffer& taup, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* d, double* e, double* tauq, double* taup, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* d, float* e, float* tauq, float* taup, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, + std::int64_t* ipiv, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t* ipiv, + double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t* ipiv, + float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, double* s, double* u, + std::int64_t ldu, double* vt, std::int64_t ldvt, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + float* a, std::int64_t lda, float* s, float* u, std::int64_t ldu, + float* vt, std::int64_t ldvt, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, float* s, + std::complex* u, std::int64_t ldu, std::complex* vt, + std::int64_t ldvt, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, double* s, + std::complex* u, std::int64_t ldu, + std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, float* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, double* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + float* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + double* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, float* tau, float* c, + std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, float* tau, float* c, + std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* w, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* w, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* b, std::int64_t ldb, double* w, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* b, std::int64_t ldb, float* w, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* d, double* e, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* d, float* e, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, std::int64_t* ipiv, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, double* a, std::int64_t lda, + double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, float* a, std::int64_t lda, + float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, float* tau, + std::int64_t stride_tau, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + float** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + double** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, float** a, + std::int64_t* lda, std::int64_t** ipiv, float** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, double** a, + std::int64_t* lda, std::int64_t** ipiv, double** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, float** a, + std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, double** a, + std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event potrs_batch( + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); template = nullptr> -std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gebrd_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, +std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, +std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template = nullptr> -std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda); +std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); +std::int64_t getrs_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::int64_t lda, + std::int64_t ldb); template = nullptr> -std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, +std::int64_t heevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, +std::int64_t hegvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t hetrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t hetrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, +std::int64_t orgbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t orgtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t potrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t potrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t potri_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t sytrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, +std::int64_t syevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, +std::int64_t sygvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t sytrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t trtrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, +std::int64_t ungbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t ungqr_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t ungtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template = nullptr> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda, std::int64_t stride_a, +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size); +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, std::int64_t lda, + std::int64_t stride_a, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size); template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template = nullptr> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes); template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes); template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); -template <> -ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); +std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template <> -ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size>(sycl::queue& queue, + std::int64_t m, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size>(sycl::queue& queue, + std::int64_t m, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, - oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t ldu, std::int64_t ldvt); +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size>(sycl::queue& queue, + std::int64_t m, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size( + sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, + std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); +template <> ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( - sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, + sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size>( - sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, + sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template <> -ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(sycl::queue &queue, - std::int64_t m, - std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size>(sycl::queue& queue, + std::int64_t m, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(sycl::queue &queue, - std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(sycl::queue& queue, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, - oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT std::int64_t getri_scratchpad_size>(sycl::queue& queue, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue& queue, + oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t ldb); +template <> ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t getrs_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>(sycl::queue &queue, - oneapi::math::job jobz, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size>(sycl::queue& queue, + oneapi::math::job jobz, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); +template <> ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( - sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, + sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size>( - sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, + sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, - oneapi::math::generate vect, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, - oneapi::math::generate vect, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue& queue, + oneapi::math::generate vect, + std::int64_t m, std::int64_t n, + std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue& queue, + oneapi::math::generate vect, + std::int64_t m, std::int64_t n, + std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size( + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size( + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size( + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t lda, std::int64_t ldc); +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size( + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue& queue, + oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue& queue, + oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, + std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::int64_t lda, + std::int64_t ldb); +template <> +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::int64_t lda, + std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t ldb); + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t potrs_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t ldb); -template <> -ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t potri_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue& queue, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda, std::int64_t ldb); template <> -ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, - std::int64_t ldb); +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, - std::int64_t ldb); +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size( + sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); +template <> +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size( + sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template <> ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( - sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size>( - sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template <> -ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, - std::int64_t lda); -template <> -ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template <> +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, + std::int64_t lda); +template <> ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( - sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size>( - sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( - sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size>( - sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( - sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size>( - sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template <> -ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t stride_a, - std::int64_t stride_ipiv, - std::int64_t batch_size); -template <> -ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t stride_a, + std::int64_t stride_ipiv, + std::int64_t batch_size); +template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> -ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, - std::int64_t lda, - std::int64_t stride_a, - std::int64_t stride_ipiv, - std::int64_t batch_size); -template <> -ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, + std::int64_t lda, + std::int64_t stride_a, + std::int64_t stride_ipiv, + std::int64_t batch_size); +template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, + sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, + sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template <> -ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t stride_a, - std::int64_t stride_tau, - std::int64_t batch_size); -template <> -ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t stride_a, + std::int64_t stride_tau, + std::int64_t batch_size); +template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, - oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda, - std::int64_t stride_a, - std::int64_t batch_size); -template <> -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda, + std::int64_t stride_a, + std::int64_t batch_size); +template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template <> -ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); -template <> -ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); + std::int64_t* group_sizes); +template <> +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template <> -ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); -template <> -ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); + std::int64_t* group_sizes); +template <> +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* n, std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t *n, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* n, std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); template <> -ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); -template <> -ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); + std::int64_t* group_sizes); +template <> +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); -template <> -ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); -template <> -ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, + sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); +template <> +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); + std::int64_t* group_sizes); template <> -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, - oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template <> -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, - oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); + std::int64_t* group_sizes); +template <> +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template <> ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); diff --git a/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hxx b/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hxx index 40468215c..f9db456a9 100644 --- a/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hxx +++ b/include/oneapi/math/lapack/detail/rocsolver/lapack_ct.hxx @@ -22,2293 +22,2320 @@ // Buffer APIs static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, sycl::buffer &tauq, - sycl::buffer &taup, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, sycl::buffer& tauq, + sycl::buffer& taup, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, taup, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getrf(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void getri(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size); -} -static inline void getrs(backend_selector selector, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, - scratchpad, scratchpad_size); -} -static inline void getrs(backend_selector selector, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, - scratchpad, scratchpad_size); -} -static inline void getrs(backend_selector selector, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, - scratchpad, scratchpad_size); -} -static inline void getrs(backend_selector selector, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, ldb, - scratchpad, scratchpad_size); + scratchpad_size); +} +static inline void getrs(backend_selector selector, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, + ldb, scratchpad, scratchpad_size); +} +static inline void getrs(backend_selector selector, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, + ldb, scratchpad, scratchpad_size); +} +static inline void getrs(backend_selector selector, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, + ldb, scratchpad, scratchpad_size); +} +static inline void getrs(backend_selector selector, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, b, + ldb, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, - std::int64_t ldvt, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& s, + sycl::buffer& u, std::int64_t ldu, sycl::buffer& vt, + std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, - vt, ldvt, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, + ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, - std::int64_t ldvt, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& s, + sycl::buffer& u, std::int64_t ldu, sycl::buffer& vt, + std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, - vt, ldvt, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, + ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, - std::int64_t ldu, sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, + std::int64_t ldu, sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, - vt, ldvt, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, + ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void gesvd(backend_selector selector, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, - std::int64_t ldu, sycl::buffer> &vt, - std::int64_t ldvt, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, + std::int64_t ldu, sycl::buffer>& vt, + std::int64_t ldvt, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, ldu, - vt, ldvt, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, u, + ldu, vt, ldvt, scratchpad, scratchpad_size); } static inline void heevd(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, - scratchpad_size); + oneapi::math::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + scratchpad, scratchpad_size); } static inline void heevd(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, - scratchpad_size); + oneapi::math::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, - w, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, + ldb, w, scratchpad, scratchpad_size); } static inline void hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, - w, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, + ldb, w, scratchpad, scratchpad_size); } static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void hetrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void hetrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void orgbr(backend_selector selector, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void orgbr(backend_selector selector, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void orgtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void ormtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, - c, ldc, scratchpad, scratchpad_size); + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, + tau, c, ldc, scratchpad, scratchpad_size); } static inline void ormtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, - c, ldc, scratchpad, scratchpad_size); + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, + tau, c, ldc, scratchpad, scratchpad_size); } static inline void ormrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void ormrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void ormqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void ormqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potri(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void potrs(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void syevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, - scratchpad_size); + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + scratchpad, scratchpad_size); } static inline void syevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &w, sycl::buffer &scratchpad, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, scratchpad, - scratchpad_size); + oneapi::math::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, + scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, - w, scratchpad, scratchpad_size); + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, + ldb, w, scratchpad, scratchpad_size); } static inline void sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, ldb, - w, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, b, + ldb, w, scratchpad, scratchpad_size); } static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tau, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void sytrd(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void sytrf(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, lda, - b, ldb, scratchpad, scratchpad_size); + b, ldb, scratchpad, scratchpad_size); } static inline void ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size); + scratchpad, scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void ungtr(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, scratchpad, - scratchpad_size); + scratchpad_size); } static inline void unmrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void unmrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void unmqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void unmqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, c, - ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, tau, + c, ldc, scratchpad, scratchpad_size); } static inline void unmtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, - c, ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, + tau, c, ldc, scratchpad, scratchpad_size); } static inline void unmtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, tau, - c, ldc, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, lda, + tau, c, ldc, scratchpad, scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, + sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, + sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getri_batch(backend_selector selector, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer> &b, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer> &b, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - stride_a, ipiv, stride_ipiv, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size); + stride_a, ipiv, stride_ipiv, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, + sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, sycl::buffer> &a, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, + sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, ipiv, - stride_ipiv, batch_size, scratchpad, - scratchpad_size); + stride_ipiv, batch_size, scratchpad, + scratchpad_size); } static inline void orgqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void orgqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrf_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, - batch_size, scratchpad, scratchpad_size); + batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrf_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, - batch_size, scratchpad, scratchpad_size); + batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, +static inline void potrf_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, - batch_size, scratchpad, scratchpad_size); + batch_size, scratchpad, scratchpad_size); } -static inline void potrf_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, +static inline void potrf_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, stride_a, - batch_size, scratchpad, scratchpad_size); + batch_size, scratchpad, scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +static inline void potrs_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size); -} -static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + stride_a, b, ldb, stride_b, batch_size, scratchpad, + scratchpad_size); +} +static inline void potrs_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size); + stride_a, b, ldb, stride_b, batch_size, scratchpad, + scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &b, +static inline void potrs_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size); + stride_a, b, ldb, stride_b, batch_size, scratchpad, + scratchpad_size); } -static inline void potrs_batch(backend_selector selector, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &b, +static inline void potrs_batch(backend_selector selector, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size); + stride_a, b, ldb, stride_b, batch_size, scratchpad, + scratchpad_size); } static inline void ungqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, + std::int64_t n, std::int64_t k, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, + sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } static inline void ungqr_batch(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, tau, - stride_tau, batch_size, scratchpad, scratchpad_size); + oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, scratchpad, + scratchpad_size); } // USM APIs static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, float *d, - float *e, std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::complex* a, std::int64_t lda, float* d, + float* e, std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, - taup, scratchpad, scratchpad_size, dependencies); + taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *d, double *e, - double *tauq, double *taup, double *scratchpad, + std::int64_t n, double* a, std::int64_t lda, double* d, double* e, + double* tauq, double* taup, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, - taup, scratchpad, scratchpad_size, dependencies); + taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *d, float *e, - float *tauq, float *taup, float *scratchpad, + std::int64_t n, float* a, std::int64_t lda, float* d, float* e, + float* tauq, float* taup, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, - taup, scratchpad, scratchpad_size, dependencies); + taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gebrd(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - double *d, double *e, std::complex *tauq, - std::complex *taup, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + double* d, double* e, std::complex* tauq, + std::complex* taup, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::gebrd(selector.get_queue(), m, n, a, lda, d, e, tauq, - taup, scratchpad, scratchpad_size, dependencies); + taup, scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + std::int64_t n, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + std::int64_t n, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event gerqf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::gerqf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + std::int64_t n, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + std::int64_t n, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::geqrf(selector.get_queue(), m, n, a, lda, tau, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, double* a, std::int64_t lda, std::int64_t* ipiv, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, float* a, std::int64_t lda, std::int64_t* ipiv, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrf(selector.get_queue(), m, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, + double* a, std::int64_t lda, std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, + float* a, std::int64_t lda, std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event getri(backend_selector selector, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri(selector.get_queue(), n, a, lda, ipiv, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size, dependencies); + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, std::int64_t *ipiv, double *b, - std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size, dependencies); + double* a, std::int64_t lda, std::int64_t* ipiv, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - float *a, std::int64_t lda, std::int64_t *ipiv, float *b, - std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size, dependencies); + float* a, std::int64_t lda, std::int64_t* ipiv, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event getrs(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, ipiv, - b, ldb, scratchpad, scratchpad_size, dependencies); + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrs(selector.get_queue(), trans, n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *s, double *u, - std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size, - dependencies); + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* s, double* u, std::int64_t ldu, double* vt, + std::int64_t ldvt, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, + s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *s, float *u, - std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size, - dependencies); + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* s, float* u, std::int64_t ldu, float* vt, std::int64_t ldvt, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, + s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, float *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size, - dependencies); + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, float* s, std::complex* u, + std::int64_t ldu, std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, + s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event gesvd(backend_selector selector, - oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - double *s, std::complex *u, std::int64_t ldu, - std::complex *vt, std::int64_t ldvt, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, s, - u, ldu, vt, ldvt, scratchpad, scratchpad_size, - dependencies); -} -static inline sycl::event heevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, float *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, double* s, std::complex* u, + std::int64_t ldu, std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::gesvd(selector.get_queue(), jobu, jobvt, m, n, a, lda, + s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, + dependencies); +} +static inline sycl::event heevd(backend_selector selector, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, float* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } -static inline sycl::event heevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, double *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event heevd(backend_selector selector, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, double* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::heevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, float *w, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, float* w, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, + lda, b, ldb, w, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event hegvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, double *w, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, double* w, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::hegvd(selector.get_queue(), itype, jobz, uplo, n, a, + lda, b, ldb, w, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, float *d, float *e, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, float* d, float* e, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, double *d, double *e, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, double* d, double* e, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::hetrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event hetrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::hetrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::orgbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::orgqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::orgtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ormtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* a, std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ormtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* a, std::int64_t lda, double* tau, double* c, + std::int64_t ldc, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ormtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ormrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ormrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, double *tau, double *c, std::int64_t ldc, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, double* tau, double* c, std::int64_t ldc, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::ormrq(selector.get_queue(), side, trans, m, n, k, a, + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ormqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, double *tau, double *c, std::int64_t ldc, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, double* tau, double* c, std::int64_t ldc, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ormqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::ormqr(selector.get_queue(), side, trans, m, n, k, a, + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potri(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potri(selector.get_queue(), uplo, n, a, lda, scratchpad, - scratchpad_size, dependencies); + scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, float* b, std::int64_t ldb, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, + ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, double *b, std::int64_t ldb, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); + double* a, std::int64_t lda, double* b, std::int64_t ldb, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, + ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, + ldb, scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, + ldb, scratchpad, scratchpad_size, dependencies); +} +static inline sycl::event syevd(backend_selector selector, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* w, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::potrs(selector.get_queue(), uplo, n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies); -} -static inline sycl::event syevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *w, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } -static inline sycl::event syevd(backend_selector selector, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *w, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { +static inline sycl::event syevd(backend_selector selector, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* w, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::syevd(selector.get_queue(), jobz, uplo, n, a, lda, w, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *b, std::int64_t ldb, double *w, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size, - dependencies); + double* a, std::int64_t lda, double* b, std::int64_t ldb, double* w, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, + lda, b, ldb, w, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event sygvd(backend_selector selector, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *b, std::int64_t ldb, float *w, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, lda, - b, ldb, w, scratchpad, scratchpad_size, - dependencies); + float* a, std::int64_t lda, float* b, std::int64_t ldb, float* w, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::sygvd(selector.get_queue(), itype, jobz, uplo, n, a, + lda, b, ldb, w, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *d, double *e, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* d, double* e, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrd(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *d, float *e, float *tau, float *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + float* d, float* e, float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::sytrd(selector.get_queue(), uplo, n, a, lda, d, e, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event sytrf(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::sytrf(selector.get_queue(), uplo, n, a, lda, ipiv, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, + a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, double *b, std::int64_t ldb, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size, - dependencies); + double* a, std::int64_t lda, double* b, std::int64_t ldb, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, + a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size, - dependencies); + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, float* b, std::int64_t ldb, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, + a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event trtrs(backend_selector selector, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, a, - lda, b, ldb, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::trtrs(selector.get_queue(), uplo, trans, diag, n, nrhs, + a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungbr(backend_selector selector, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ungbr(selector.get_queue(), vec, m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ungqr(selector.get_queue(), m, n, k, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungtr(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ungtr(selector.get_queue(), uplo, n, a, lda, tau, - scratchpad, scratchpad_size, dependencies); + scratchpad, scratchpad_size, dependencies); } static inline sycl::event unmrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event unmrq(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::unmrq(selector.get_queue(), side, trans, m, n, k, a, + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event unmqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event unmqr(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, lda, - tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::unmqr(selector.get_queue(), side, trans, m, n, k, a, + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event unmtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event unmtr(backend_selector selector, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::unmtr(selector.get_queue(), side, uplo, trans, m, n, a, - lda, tau, c, ldc, scratchpad, scratchpad_size, - dependencies); + lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, + std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, + std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size, dependencies); + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size, dependencies); + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - tau, stride_tau, batch_size, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); +} +static inline sycl::event geqrf_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event geqrf_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event geqrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event geqrf_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event geqrf_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::geqrf_batch(selector.get_queue(), m, n, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, + stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, + stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, + stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrf_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event getrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, float **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, + stride_a, ipiv, stride_ipiv, batch_size, + scratchpad, scratchpad_size, dependencies); +} +static inline sycl::event getrf_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event getrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, double **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event getrf_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event getrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event getrf_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event getrf_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event getrf_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrf_batch(selector.get_queue(), m, n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, std::int64_t n, - float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, float *scratchpad, + float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, std::int64_t n, - double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, double *scratchpad, + double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getri_batch(backend_selector selector, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, stride_a, - ipiv, stride_ipiv, batch_size, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event getri_batch(backend_selector selector, std::int64_t *n, - float **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + ipiv, stride_ipiv, batch_size, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event getri_batch(backend_selector selector, + std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event getri_batch(backend_selector selector, std::int64_t *n, - double **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event getri_batch(backend_selector selector, + std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event getri_batch(backend_selector selector, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event getri_batch(backend_selector selector, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event getri_batch(backend_selector selector, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +static inline sycl::event getri_batch(backend_selector selector, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getri_batch(selector.get_queue(), n, a, lda, ipiv, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, float *b, std::int64_t ldb, + std::int64_t nrhs, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, double *b, std::int64_t ldb, + std::int64_t nrhs, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}) { + std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( backend_selector selector, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}) { + std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::getrs_batch( selector.get_queue(), trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, - std::int64_t **ipiv, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, float** a, std::int64_t* lda, + std::int64_t** ipiv, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, + lda, ipiv, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, - std::int64_t **ipiv, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, double** a, std::int64_t* lda, + std::int64_t** ipiv, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, + lda, ipiv, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, + lda, ipiv, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event getrs_batch( - backend_selector selector, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { - return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, lda, - ipiv, b, ldb, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies); + backend_selector selector, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { + return oneapi::math::lapack::rocsolver::getrs_batch(selector.get_queue(), trans, n, nrhs, a, + lda, ipiv, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event orgqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); } -static inline sycl::event orgqr_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float **a, - std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, +static inline sycl::event orgqr_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, std::int64_t* k, float** a, + std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } -static inline sycl::event orgqr_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::int64_t *k, double **a, - std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, +static inline sycl::event orgqr_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, std::int64_t* k, double** a, + std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::orgqr_batch(selector.get_queue(), m, n, k, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, float *a, + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, float *scratchpad, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo uplo, std::int64_t n, double *a, + oneapi::math::uplo uplo, std::int64_t n, double* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, double *scratchpad, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - stride_a, batch_size, scratchpad, - scratchpad_size, dependencies); + stride_a, batch_size, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrf_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrf_batch(selector.get_queue(), uplo, n, a, lda, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - float *a, std::int64_t lda, std::int64_t stride_a, float *b, + float* a, std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, float *scratchpad, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, std::int64_t stride_a, double *b, + double* a, std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, double *scratchpad, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - stride_a, b, ldb, stride_b, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, b, ldb, stride_b, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - float **a, std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + float** a, std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - double **a, std::int64_t *lda, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + double** a, std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event potrs_batch(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::potrs_batch(selector.get_queue(), uplo, n, nrhs, a, lda, - b, ldb, group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, + std::int64_t n, std::int64_t k, std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, + std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size, dependencies); + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); } static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, + std::int64_t n, std::int64_t k, std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, + std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, - stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size, dependencies); -} -static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::int64_t *k, std::complex **a, - std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + stride_a, tau, stride_tau, batch_size, + scratchpad, scratchpad_size, dependencies); +} +static inline sycl::event ungqr_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); -} -static inline sycl::event ungqr_batch(backend_selector selector, std::int64_t *m, - std::int64_t *n, std::int64_t *k, std::complex **a, - std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); +} +static inline sycl::event ungqr_batch(backend_selector selector, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return oneapi::math::lapack::rocsolver::ungqr_batch(selector.get_queue(), m, n, k, a, lda, tau, - group_count, group_sizes, scratchpad, - scratchpad_size, dependencies); + group_count, group_sizes, scratchpad, + scratchpad_size, dependencies); } // SCRATCHPAD APIs template std::int64_t gebrd_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::gebrd_scratchpad_size(selector.get_queue(), m, n, - lda); + return oneapi::math::lapack::rocsolver::gebrd_scratchpad_size(selector.get_queue(), m, + n, lda); } template std::int64_t gerqf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::gerqf_scratchpad_size(selector.get_queue(), m, n, - lda); + return oneapi::math::lapack::rocsolver::gerqf_scratchpad_size(selector.get_queue(), m, + n, lda); } template std::int64_t geqrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::geqrf_scratchpad_size(selector.get_queue(), m, n, - lda); + return oneapi::math::lapack::rocsolver::geqrf_scratchpad_size(selector.get_queue(), m, + n, lda); } template std::int64_t gesvd_scratchpad_size(backend_selector selector, @@ -2321,82 +2348,83 @@ std::int64_t gesvd_scratchpad_size(backend_selector selector template std::int64_t getrf_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::getrf_scratchpad_size(selector.get_queue(), m, n, - lda); + return oneapi::math::lapack::rocsolver::getrf_scratchpad_size(selector.get_queue(), m, + n, lda); } template std::int64_t getri_scratchpad_size(backend_selector selector, std::int64_t n, std::int64_t lda) { return oneapi::math::lapack::rocsolver::getri_scratchpad_size(selector.get_queue(), n, - lda); + lda); } template std::int64_t getrs_scratchpad_size(backend_selector selector, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::math::lapack::rocsolver::getrs_scratchpad_size(selector.get_queue(), - trans, n, nrhs, lda, ldb); + return oneapi::math::lapack::rocsolver::getrs_scratchpad_size( + selector.get_queue(), trans, n, nrhs, lda, ldb); } template std::int64_t heevd_scratchpad_size(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::heevd_scratchpad_size(selector.get_queue(), jobz, - uplo, n, lda); + return oneapi::math::lapack::rocsolver::heevd_scratchpad_size(selector.get_queue(), + jobz, uplo, n, lda); } template -std::int64_t hegvd_scratchpad_size(backend_selector selector, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda, std::int64_t ldb) { +std::int64_t hegvd_scratchpad_size(backend_selector selector, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, + std::int64_t ldb) { return oneapi::math::lapack::rocsolver::hegvd_scratchpad_size( selector.get_queue(), itype, jobz, uplo, n, lda, ldb); } template std::int64_t hetrd_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::hetrd_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::rocsolver::hetrd_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t hetrf_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::hetrf_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::rocsolver::hetrf_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t orgbr_scratchpad_size(backend_selector selector, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::orgbr_scratchpad_size(selector.get_queue(), vect, - m, n, k, lda); + return oneapi::math::lapack::rocsolver::orgbr_scratchpad_size(selector.get_queue(), + vect, m, n, k, lda); } template std::int64_t orgtr_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::orgtr_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::rocsolver::orgtr_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t orgqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::orgqr_scratchpad_size(selector.get_queue(), m, n, - k, lda); + return oneapi::math::lapack::rocsolver::orgqr_scratchpad_size(selector.get_queue(), m, + n, k, lda); } template std::int64_t ormrq_scratchpad_size(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::math::lapack::rocsolver::ormrq_scratchpad_size(selector.get_queue(), side, - trans, m, n, k, lda, ldc); + return oneapi::math::lapack::rocsolver::ormrq_scratchpad_size( + selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t ormqr_scratchpad_size(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::math::lapack::rocsolver::ormqr_scratchpad_size(selector.get_queue(), side, - trans, m, n, k, lda, ldc); + return oneapi::math::lapack::rocsolver::ormqr_scratchpad_size( + selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t ormtr_scratchpad_size(backend_selector selector, @@ -2409,47 +2437,48 @@ std::int64_t ormtr_scratchpad_size(backend_selector selector template std::int64_t potrf_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::potrf_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::rocsolver::potrf_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t potrs_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - return oneapi::math::lapack::rocsolver::potrs_scratchpad_size(selector.get_queue(), uplo, - n, nrhs, lda, ldb); + return oneapi::math::lapack::rocsolver::potrs_scratchpad_size(selector.get_queue(), + uplo, n, nrhs, lda, ldb); } template std::int64_t potri_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::potri_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::rocsolver::potri_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t sytrf_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::sytrf_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::rocsolver::sytrf_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t syevd_scratchpad_size(backend_selector selector, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::syevd_scratchpad_size(selector.get_queue(), jobz, - uplo, n, lda); + return oneapi::math::lapack::rocsolver::syevd_scratchpad_size(selector.get_queue(), + jobz, uplo, n, lda); } template -std::int64_t sygvd_scratchpad_size(backend_selector selector, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda, std::int64_t ldb) { +std::int64_t sygvd_scratchpad_size(backend_selector selector, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, + std::int64_t ldb) { return oneapi::math::lapack::rocsolver::sygvd_scratchpad_size( selector.get_queue(), itype, jobz, uplo, n, lda, ldb); } template std::int64_t sytrd_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::sytrd_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::rocsolver::sytrd_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t trtrs_scratchpad_size(backend_selector selector, @@ -2463,36 +2492,36 @@ template std::int64_t ungbr_scratchpad_size(backend_selector selector, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::ungbr_scratchpad_size(selector.get_queue(), vect, - m, n, k, lda); + return oneapi::math::lapack::rocsolver::ungbr_scratchpad_size(selector.get_queue(), + vect, m, n, k, lda); } template std::int64_t ungqr_scratchpad_size(backend_selector selector, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::ungqr_scratchpad_size(selector.get_queue(), m, n, - k, lda); + return oneapi::math::lapack::rocsolver::ungqr_scratchpad_size(selector.get_queue(), m, + n, k, lda); } template std::int64_t ungtr_scratchpad_size(backend_selector selector, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - return oneapi::math::lapack::rocsolver::ungtr_scratchpad_size(selector.get_queue(), uplo, - n, lda); + return oneapi::math::lapack::rocsolver::ungtr_scratchpad_size(selector.get_queue(), + uplo, n, lda); } template std::int64_t unmrq_scratchpad_size(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::math::lapack::rocsolver::unmrq_scratchpad_size(selector.get_queue(), side, - trans, m, n, k, lda, ldc); + return oneapi::math::lapack::rocsolver::unmrq_scratchpad_size( + selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t unmqr_scratchpad_size(backend_selector selector, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - return oneapi::math::lapack::rocsolver::unmqr_scratchpad_size(selector.get_queue(), side, - trans, m, n, k, lda, ldc); + return oneapi::math::lapack::rocsolver::unmqr_scratchpad_size( + selector.get_queue(), side, trans, m, n, k, lda, ldc); } template std::int64_t unmtr_scratchpad_size(backend_selector selector, @@ -2568,62 +2597,62 @@ std::int64_t ungqr_batch_scratchpad_size(backend_selector se } template std::int64_t getrf_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::rocsolver::getrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } template std::int64_t getri_batch_scratchpad_size(backend_selector selector, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::rocsolver::getri_batch_scratchpad_size( selector.get_queue(), n, lda, group_count, group_sizes); } template std::int64_t getrs_batch_scratchpad_size(backend_selector selector, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes) { + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::rocsolver::getrs_batch_scratchpad_size( selector.get_queue(), trans, n, nrhs, lda, ldb, group_count, group_sizes); } template std::int64_t geqrf_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::rocsolver::geqrf_batch_scratchpad_size( selector.get_queue(), m, n, lda, group_count, group_sizes); } template std::int64_t orgqr_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return oneapi::math::lapack::rocsolver::orgqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } template std::int64_t potrf_batch_scratchpad_size(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return oneapi::math::lapack::rocsolver::potrf_batch_scratchpad_size( selector.get_queue(), uplo, n, lda, group_count, group_sizes); } template std::int64_t potrs_batch_scratchpad_size(backend_selector selector, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes) { + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes) { return oneapi::math::lapack::rocsolver::potrs_batch_scratchpad_size( selector.get_queue(), uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template std::int64_t ungqr_batch_scratchpad_size(backend_selector selector, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return oneapi::math::lapack::rocsolver::ungqr_batch_scratchpad_size( selector.get_queue(), m, n, k, lda, group_count, group_sizes); } diff --git a/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx index 11a9fe326..f8ebfd700 100644 --- a/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx +++ b/include/oneapi/math/lapack/detail/rocsolver/onemath_lapack_rocsolver.hxx @@ -21,1815 +21,1836 @@ // Buffer APIs -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tauq, - sycl::buffer &taup, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getri_batch(sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - -ONEMATH_EXPORT void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, - std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tauq, + sycl::buffer& taup, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tauq, + sycl::buffer& taup, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getri_batch(sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + +ONEMATH_EXPORT void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, + std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); // USM APIs -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *d, double *e, double *tauq, double *taup, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *d, float *e, float *tauq, float *taup, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, - double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, - float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *s, double *u, std::int64_t ldu, - double *vt, std::int64_t ldvt, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, - std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, float *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *s, - std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, float *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - double *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - float *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - double *w, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, - std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, - std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, float *b, - std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, double *b, - std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, double *a, std::int64_t lda, double *w, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, float *a, std::int64_t lda, float *w, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, - double *b, std::int64_t ldb, double *w, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, - float *b, std::int64_t ldb, float *w, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *d, double *e, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *d, float *e, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, - double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, - float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, float *tau, - std::int64_t stride_tau, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, double *tau, - std::int64_t stride_tau, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - float **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - double **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, float **a, - std::int64_t *lda, std::int64_t **ipiv, float **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, double **a, - std::int64_t *lda, std::int64_t **ipiv, double **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - float **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - double **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::complex **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* d, double* e, double* tauq, double* taup, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* d, float* e, float* tauq, float* taup, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, + std::int64_t* ipiv, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t* ipiv, + double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t* ipiv, + float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, double* s, double* u, + std::int64_t ldu, double* vt, std::int64_t ldvt, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + float* a, std::int64_t lda, float* s, float* u, std::int64_t ldu, + float* vt, std::int64_t ldvt, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, float* s, + std::complex* u, std::int64_t ldu, std::complex* vt, + std::int64_t ldvt, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, double* s, + std::complex* u, std::int64_t ldu, + std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, float* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, double* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + float* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + double* w, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, float* tau, float* c, + std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, float* tau, float* c, + std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* w, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* w, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* b, std::int64_t ldb, double* w, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* b, std::int64_t ldb, float* w, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* d, double* e, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* d, float* e, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, std::int64_t* ipiv, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, double* a, std::int64_t lda, + double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, float* a, std::int64_t lda, + float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, float* tau, + std::int64_t stride_tau, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + float** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + double** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, float** a, + std::int64_t* lda, std::int64_t** ipiv, float** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, double** a, + std::int64_t* lda, std::int64_t** ipiv, double** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, float** a, + std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, double** a, + std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event potrs_batch( + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, + std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies = {}); // SCRATCHPAD APIs template -ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t gebrd_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t geqrf_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, std::int64_t lda, std::int64_t ldu, - std::int64_t ldvt); +ONEMATH_EXPORT std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t ldu, std::int64_t ldvt); template -ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t getrf_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, + std::int64_t lda); template -ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT std::int64_t getrs_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t heevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template -ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda, - std::int64_t ldb); +ONEMATH_EXPORT std::int64_t hegvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda, + std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t hetrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t hetrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, - std::int64_t m, std::int64_t n, std::int64_t k, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t orgbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, + std::int64_t m, std::int64_t n, std::int64_t k, + std::int64_t lda); template -ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t orgtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda); +ONEMATH_EXPORT std::int64_t orgqr_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t ormqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t ormtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t potrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb); +ONEMATH_EXPORT std::int64_t potrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t potri_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t sytrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t syevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda); template -ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda, - std::int64_t ldb); +ONEMATH_EXPORT std::int64_t sygvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda, + std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t sytrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, - std::int64_t ldb); +ONEMATH_EXPORT std::int64_t trtrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, std::int64_t lda, + std::int64_t ldb); template -ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, - std::int64_t m, std::int64_t n, std::int64_t k, - std::int64_t lda); +ONEMATH_EXPORT std::int64_t ungbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, + std::int64_t m, std::int64_t n, std::int64_t k, + std::int64_t lda); template -ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda); +ONEMATH_EXPORT std::int64_t ungqr_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda); +ONEMATH_EXPORT std::int64_t ungtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda); template -ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t unmrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t unmqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t ldc); +ONEMATH_EXPORT std::int64_t unmtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t ldc); template -ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t stride_a, - std::int64_t stride_ipiv, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t stride_a, + std::int64_t stride_ipiv, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, - std::int64_t lda, std::int64_t stride_a, - std::int64_t stride_ipiv, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, + std::int64_t lda, std::int64_t stride_a, + std::int64_t stride_ipiv, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, - oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t stride_a, - std::int64_t stride_ipiv, std::int64_t ldb, - std::int64_t stride_b, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t lda, - std::int64_t stride_a, - std::int64_t stride_tau, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t lda, + std::int64_t stride_a, + std::int64_t stride_tau, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda, - std::int64_t stride_a, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda, + std::int64_t stride_a, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t stride_a, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t stride_a, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda, std::int64_t stride_a, - std::int64_t stride_tau, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, + std::int64_t lda, std::int64_t stride_a, + std::int64_t stride_tau, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda, std::int64_t stride_a, - std::int64_t stride_tau, - std::int64_t batch_size); +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, + std::int64_t lda, std::int64_t stride_a, + std::int64_t stride_tau, + std::int64_t batch_size); template -ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template ONEMATH_EXPORT std::int64_t getrs_batch_scratchpad_size( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, + std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes); template -ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +ONEMATH_EXPORT std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); diff --git a/include/oneapi/math/lapack/exceptions.hpp b/include/oneapi/math/lapack/exceptions.hpp index 33eddea4d..db9baf9d6 100644 --- a/include/oneapi/math/lapack/exceptions.hpp +++ b/include/oneapi/math/lapack/exceptions.hpp @@ -25,7 +25,7 @@ namespace lapack { class exception { public: - exception(oneapi::math::exception *_ex, std::int64_t info, std::int64_t detail = 0) + exception(oneapi::math::exception* _ex, std::int64_t info, std::int64_t detail = 0) : _info(info), _detail(detail), _ex(_ex) {} @@ -35,20 +35,20 @@ class exception { std::int64_t detail() const { return _detail; } - const char *what() const { + const char* what() const { return _ex->what(); } private: std::int64_t _info; std::int64_t _detail; - math::exception *_ex; + math::exception* _ex; }; class computation_error : public oneapi::math::computation_error, public oneapi::math::lapack::exception { public: - computation_error(const std::string &function, const std::string &info, std::int64_t code) + computation_error(const std::string& function, const std::string& info, std::int64_t code) : oneapi::math::computation_error("LAPACK", function, info), oneapi::math::lapack::exception(this, code) {} using oneapi::math::computation_error::what; @@ -56,17 +56,17 @@ class computation_error : public oneapi::math::computation_error, class batch_error : public oneapi::math::batch_error, public oneapi::math::lapack::exception { public: - batch_error(const std::string &function, const std::string &info, std::int64_t num_errors, + batch_error(const std::string& function, const std::string& info, std::int64_t num_errors, std::vector ids = {}, std::vector exceptions = {}) : oneapi::math::batch_error("LAPACK", function, info), oneapi::math::lapack::exception(this, num_errors), _ids(ids), _exceptions(exceptions) {} using oneapi::math::batch_error::what; - const std::vector &ids() const { + const std::vector& ids() const { return _ids; } - const std::vector &exceptions() const { + const std::vector& exceptions() const { return _exceptions; } @@ -78,7 +78,7 @@ class batch_error : public oneapi::math::batch_error, public oneapi::math::lapac class invalid_argument : public oneapi::math::invalid_argument, public oneapi::math::lapack::exception { public: - invalid_argument(const std::string &function, const std::string &info, + invalid_argument(const std::string& function, const std::string& info, std::int64_t arg_position = 0, std::int64_t detail = 0) : oneapi::math::invalid_argument("LAPACK", function, info), oneapi::math::lapack::exception(this, arg_position, detail) {} diff --git a/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp b/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp index 874432e19..7aa400bf8 100644 --- a/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp +++ b/include/oneapi/math/rng/detail/curand/onemath_rng_curand.hpp @@ -75,13 +75,13 @@ namespace rng { namespace curand { ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, - std::uint64_t seed); + std::uint64_t seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, - std::uint32_t seed); + std::uint32_t seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); diff --git a/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp b/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp index ff875be2f..9848a5cee 100644 --- a/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp +++ b/include/oneapi/math/rng/detail/mklcpu/onemath_rng_mklcpu.hpp @@ -36,13 +36,13 @@ namespace rng { namespace mklcpu { ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, - std::uint64_t seed); + std::uint64_t seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, - std::uint32_t seed); + std::uint32_t seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); diff --git a/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp b/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp index 9fc40dfcd..f36fe3997 100644 --- a/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp +++ b/include/oneapi/math/rng/detail/mklgpu/onemath_rng_mklgpu.hpp @@ -36,13 +36,13 @@ namespace rng { namespace mklgpu { ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, - std::uint64_t seed); + std::uint64_t seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, - std::uint32_t seed); + std::uint32_t seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); diff --git a/include/oneapi/math/rng/detail/rng_loader.hpp b/include/oneapi/math/rng/detail/rng_loader.hpp index 19de23529..1855d641e 100644 --- a/include/oneapi/math/rng/detail/rng_loader.hpp +++ b/include/oneapi/math/rng/detail/rng_loader.hpp @@ -38,16 +38,16 @@ namespace rng { namespace detail { ONEMATH_EXPORT engine_impl* create_philox4x32x10(oneapi::math::device libkey, sycl::queue queue, - std::uint64_t seed); + std::uint64_t seed); ONEMATH_EXPORT engine_impl* create_philox4x32x10(oneapi::math::device libkey, sycl::queue queue, - std::initializer_list seed); + std::initializer_list seed); ONEMATH_EXPORT engine_impl* create_mrg32k3a(oneapi::math::device libkey, sycl::queue queue, - std::uint32_t seed); + std::uint32_t seed); ONEMATH_EXPORT engine_impl* create_mrg32k3a(oneapi::math::device libkey, sycl::queue queue, - std::initializer_list seed); + std::initializer_list seed); } // namespace detail } // namespace rng diff --git a/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp b/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp index 4e2538800..901f618f0 100644 --- a/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp +++ b/include/oneapi/math/rng/detail/rocrand/onemath_rng_rocrand.hpp @@ -77,13 +77,13 @@ namespace rng { namespace rocrand { ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, - std::uint64_t seed); + std::uint64_t seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_philox4x32x10( sycl::queue queue, std::initializer_list seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, - std::uint32_t seed); + std::uint32_t seed); ONEMATH_EXPORT oneapi::math::rng::detail::engine_impl* create_mrg32k3a( sycl::queue queue, std::initializer_list seed); diff --git a/include/oneapi/math/rng/device/detail/exponential_impl.hpp b/include/oneapi/math/rng/device/detail/exponential_impl.hpp index f754fd548..5c89bf824 100644 --- a/include/oneapi/math/rng/device/detail/exponential_impl.hpp +++ b/include/oneapi/math/rng/device/detail/exponential_impl.hpp @@ -126,7 +126,8 @@ class distribution_base oneapi::math::rng::device::poisson>; friend class distribution_base< oneapi::math::rng::device::poisson>; - friend class distribution_base>; + friend class distribution_base< + oneapi::math::rng::device::gamma>; friend class distribution_base< oneapi::math::rng::device::gamma>; friend class distribution_base< diff --git a/include/oneapi/math/rng/device/detail/gamma_impl.hpp b/include/oneapi/math/rng/device/detail/gamma_impl.hpp index f338f2705..816ce0ecb 100644 --- a/include/oneapi/math/rng/device/detail/gamma_impl.hpp +++ b/include/oneapi/math/rng/device/detail/gamma_impl.hpp @@ -250,7 +250,8 @@ class distribution_base> { typename std::conditional>::type { if (algorithm_ == gamma_algorithm::Exponential) { - distribution_base> distr_exp(a_, beta_); + distribution_base> distr_exp(a_, + beta_); return distr_exp.generate(engine); } sycl::vec res{}; @@ -262,7 +263,8 @@ class distribution_base> { template RealType generate_single(EngineType& engine) { if (algorithm_ == gamma_algorithm::Exponential) { - distribution_base> distr_exp(a_, beta_); + distribution_base> distr_exp(a_, + beta_); RealType z = distr_exp.generate_single(engine); return z; } diff --git a/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp b/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp index 6bca38b19..8b4d91967 100644 --- a/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp +++ b/include/oneapi/math/rng/device/detail/mrg32k3a_impl.hpp @@ -182,7 +182,7 @@ static inline void skip_ahead(engine_state(state.s, n, num_to_skip_ptr, skip_ahead_matrix[0]); @@ -190,7 +190,8 @@ static inline void skip_ahead(engine_state -static inline void validate_seed(engine_state>& state) { +static inline void validate_seed( + engine_state>& state) { int i; for (i = 0; i < 3; i++) { if (state.s[i] >= mrg32k3a_params::m1) { diff --git a/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp b/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp index 7ec22d816..ca133d22f 100644 --- a/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp +++ b/include/oneapi/math/rng/device/detail/philox4x32x10_impl.hpp @@ -119,8 +119,9 @@ static inline void round_10(std::uint32_t* cnt, std::uint32_t* k) { } template -static inline void skip_ahead(engine_state>& state, - std::uint64_t num_to_skip) { +static inline void skip_ahead( + engine_state>& state, + std::uint64_t num_to_skip) { std::uint64_t num_to_skip_tmp = num_to_skip; std::uint64_t c_inc; std::uint32_t counter[4]; @@ -157,8 +158,9 @@ static inline void skip_ahead(engine_state -static inline void skip_ahead(engine_state>& state, - std::uint64_t n, const std::uint64_t* num_to_skip_ptr) { +static inline void skip_ahead( + engine_state>& state, std::uint64_t n, + const std::uint64_t* num_to_skip_ptr) { constexpr std::uint64_t uint_max = 0xFFFFFFFFFFFFFFFF; std::uint64_t post_buffer, pre_buffer; std::int32_t num_elements = 0; diff --git a/include/oneapi/math/rng/distributions.hpp b/include/oneapi/math/rng/distributions.hpp index 0af2b80e0..f20e1333f 100644 --- a/include/oneapi/math/rng/distributions.hpp +++ b/include/oneapi/math/rng/distributions.hpp @@ -76,7 +76,7 @@ class uniform { explicit uniform(Type a, Type b) : a_(a), b_(b) { if (a >= b) { throw oneapi::math::invalid_argument("rng", "uniform", - "parameters are incorrect, a >= b"); + "parameters are incorrect, a >= b"); } } @@ -104,7 +104,7 @@ class uniform { explicit uniform(std::int32_t a, std::int32_t b) : a_(a), b_(b) { if (a >= b) { throw oneapi::math::invalid_argument("rng", "uniform", - "parameters are incorrect, a >= b"); + "parameters are incorrect, a >= b"); } } @@ -161,7 +161,7 @@ class gaussian { explicit gaussian(RealType mean, RealType stddev) : mean_(mean), stddev_(stddev) { if (stddev <= static_cast(0.0)) { throw oneapi::math::invalid_argument("rng", "gaussian", - "stddev parameter is incorrect, stddev <= 0.0"); + "stddev parameter is incorrect, stddev <= 0.0"); } } diff --git a/include/oneapi/math/sparse_blas/detail/helper_types.hpp b/include/oneapi/math/sparse_blas/detail/helper_types.hpp index b3082b736..49c462d04 100644 --- a/include/oneapi/math/sparse_blas/detail/helper_types.hpp +++ b/include/oneapi/math/sparse_blas/detail/helper_types.hpp @@ -40,7 +40,7 @@ inline constexpr bool is_int_supported_v = template inline constexpr bool are_fp_int_supported_v = - is_fp_supported_v&& is_int_supported_v; + is_fp_supported_v && is_int_supported_v; } // namespace detail } // namespace sparse diff --git a/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx b/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx index 310a9a4e8..71e22a664 100644 --- a/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx +++ b/include/oneapi/math/sparse_blas/detail/onemath_sparse_blas_backends.hxx @@ -22,196 +22,197 @@ // Dense vector template -ONEMATH_EXPORT void init_dense_vector(sycl::queue &queue, dense_vector_handle_t *p_dvhandle, - std::int64_t size, sycl::buffer val); +ONEMATH_EXPORT void init_dense_vector(sycl::queue& queue, dense_vector_handle_t* p_dvhandle, + std::int64_t size, sycl::buffer val); template -ONEMATH_EXPORT void init_dense_vector(sycl::queue &queue, dense_vector_handle_t *p_dvhandle, - std::int64_t size, dataType *val); +ONEMATH_EXPORT void init_dense_vector(sycl::queue& queue, dense_vector_handle_t* p_dvhandle, + std::int64_t size, dataType* val); template -ONEMATH_EXPORT void set_dense_vector_data(sycl::queue &queue, dense_vector_handle_t dvhandle, - std::int64_t size, sycl::buffer val); +ONEMATH_EXPORT void set_dense_vector_data(sycl::queue& queue, dense_vector_handle_t dvhandle, + std::int64_t size, sycl::buffer val); template -ONEMATH_EXPORT void set_dense_vector_data(sycl::queue &queue, dense_vector_handle_t dvhandle, - std::int64_t size, dataType *val); +ONEMATH_EXPORT void set_dense_vector_data(sycl::queue& queue, dense_vector_handle_t dvhandle, + std::int64_t size, dataType* val); -ONEMATH_EXPORT sycl::event release_dense_vector(sycl::queue &queue, dense_vector_handle_t dvhandle, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event release_dense_vector(sycl::queue& queue, dense_vector_handle_t dvhandle, + const std::vector& dependencies = {}); // Dense matrix template -ONEMATH_EXPORT void init_dense_matrix(sycl::queue &queue, dense_matrix_handle_t *p_dmhandle, - std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, - layout dense_layout, sycl::buffer val); +ONEMATH_EXPORT void init_dense_matrix(sycl::queue& queue, dense_matrix_handle_t* p_dmhandle, + std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, + layout dense_layout, sycl::buffer val); template -ONEMATH_EXPORT void init_dense_matrix(sycl::queue &queue, dense_matrix_handle_t *p_dmhandle, - std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, - layout dense_layout, dataType *val); +ONEMATH_EXPORT void init_dense_matrix(sycl::queue& queue, dense_matrix_handle_t* p_dmhandle, + std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, + layout dense_layout, dataType* val); template -ONEMATH_EXPORT void set_dense_matrix_data(sycl::queue &queue, dense_matrix_handle_t dmhandle, - std::int64_t num_rows, std::int64_t num_cols, - std::int64_t ld, layout dense_layout, - sycl::buffer val); +ONEMATH_EXPORT void set_dense_matrix_data(sycl::queue& queue, dense_matrix_handle_t dmhandle, + std::int64_t num_rows, std::int64_t num_cols, + std::int64_t ld, layout dense_layout, + sycl::buffer val); template -ONEMATH_EXPORT void set_dense_matrix_data(sycl::queue &queue, dense_matrix_handle_t dmhandle, - std::int64_t num_rows, std::int64_t num_cols, - std::int64_t ld, layout dense_layout, dataType *val); +ONEMATH_EXPORT void set_dense_matrix_data(sycl::queue& queue, dense_matrix_handle_t dmhandle, + std::int64_t num_rows, std::int64_t num_cols, + std::int64_t ld, layout dense_layout, dataType* val); -ONEMATH_EXPORT sycl::event release_dense_matrix(sycl::queue &queue, dense_matrix_handle_t dmhandle, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event release_dense_matrix(sycl::queue& queue, dense_matrix_handle_t dmhandle, + const std::vector& dependencies = {}); // COO matrix template -ONEMATH_EXPORT void init_coo_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, - index_base index, sycl::buffer row_ind, - sycl::buffer col_ind, - sycl::buffer val); +ONEMATH_EXPORT void init_coo_matrix(sycl::queue& queue, matrix_handle_t* p_smhandle, + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, + index_base index, sycl::buffer row_ind, + sycl::buffer col_ind, + sycl::buffer val); template -ONEMATH_EXPORT void init_coo_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, - index_base index, indexType *row_ind, indexType *col_ind, - dataType *val); +ONEMATH_EXPORT void init_coo_matrix(sycl::queue& queue, matrix_handle_t* p_smhandle, + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, + index_base index, indexType* row_ind, indexType* col_ind, + dataType* val); template -ONEMATH_EXPORT void set_coo_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, - std::int64_t num_rows, std::int64_t num_cols, - std::int64_t nnz, index_base index, - sycl::buffer row_ind, - sycl::buffer col_ind, - sycl::buffer val); +ONEMATH_EXPORT void set_coo_matrix_data(sycl::queue& queue, matrix_handle_t smhandle, + std::int64_t num_rows, std::int64_t num_cols, + std::int64_t nnz, index_base index, + sycl::buffer row_ind, + sycl::buffer col_ind, + sycl::buffer val); template -ONEMATH_EXPORT void set_coo_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, - std::int64_t num_rows, std::int64_t num_cols, - std::int64_t nnz, index_base index, indexType *row_ind, - indexType *col_ind, dataType *val); +ONEMATH_EXPORT void set_coo_matrix_data(sycl::queue& queue, matrix_handle_t smhandle, + std::int64_t num_rows, std::int64_t num_cols, + std::int64_t nnz, index_base index, indexType* row_ind, + indexType* col_ind, dataType* val); // CSR matrix template -ONEMATH_EXPORT void init_csr_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, - index_base index, sycl::buffer row_ptr, - sycl::buffer col_ind, - sycl::buffer val); +ONEMATH_EXPORT void init_csr_matrix(sycl::queue& queue, matrix_handle_t* p_smhandle, + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, + index_base index, sycl::buffer row_ptr, + sycl::buffer col_ind, + sycl::buffer val); template -ONEMATH_EXPORT void init_csr_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, - index_base index, indexType *row_ptr, indexType *col_ind, - dataType *val); +ONEMATH_EXPORT void init_csr_matrix(sycl::queue& queue, matrix_handle_t* p_smhandle, + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, + index_base index, indexType* row_ptr, indexType* col_ind, + dataType* val); template -ONEMATH_EXPORT void set_csr_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, - std::int64_t num_rows, std::int64_t num_cols, - std::int64_t nnz, index_base index, - sycl::buffer row_ptr, - sycl::buffer col_ind, - sycl::buffer val); +ONEMATH_EXPORT void set_csr_matrix_data(sycl::queue& queue, matrix_handle_t smhandle, + std::int64_t num_rows, std::int64_t num_cols, + std::int64_t nnz, index_base index, + sycl::buffer row_ptr, + sycl::buffer col_ind, + sycl::buffer val); template -ONEMATH_EXPORT void set_csr_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, - std::int64_t num_rows, std::int64_t num_cols, - std::int64_t nnz, index_base index, indexType *row_ptr, - indexType *col_ind, dataType *val); +ONEMATH_EXPORT void set_csr_matrix_data(sycl::queue& queue, matrix_handle_t smhandle, + std::int64_t num_rows, std::int64_t num_cols, + std::int64_t nnz, index_base index, indexType* row_ptr, + indexType* col_ind, dataType* val); // Common sparse matrix functions -ONEMATH_EXPORT sycl::event release_sparse_matrix(sycl::queue &queue, matrix_handle_t smhandle, - const std::vector &dependencies = {}); +ONEMATH_EXPORT sycl::event release_sparse_matrix(sycl::queue& queue, matrix_handle_t smhandle, + const std::vector& dependencies = {}); -bool set_matrix_property(sycl::queue &queue, matrix_handle_t smhandle, matrix_property property); +bool set_matrix_property(sycl::queue& queue, matrix_handle_t smhandle, matrix_property property); // SPMM -ONEMATH_EXPORT void init_spmm_descr(sycl::queue &queue, spmm_descr_t *p_spmm_descr); - -ONEMATH_EXPORT sycl::event release_spmm_descr(sycl::queue &queue, spmm_descr_t spmm_descr, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT void spmm_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, - oneapi::math::transpose opB, const void *alpha, - matrix_view A_view, matrix_handle_t A_handle, - dense_matrix_handle_t B_handle, const void *beta, - dense_matrix_handle_t C_handle, spmm_alg alg, - spmm_descr_t spmm_descr, std::size_t &temp_buffer_size); - -ONEMATH_EXPORT void spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, - oneapi::math::transpose opB, const void *alpha, matrix_view A_view, - matrix_handle_t A_handle, dense_matrix_handle_t B_handle, - const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, - spmm_descr_t spmm_descr, sycl::buffer workspace); - -ONEMATH_EXPORT sycl::event spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, - oneapi::math::transpose opB, const void *alpha, - matrix_view A_view, matrix_handle_t A_handle, - dense_matrix_handle_t B_handle, const void *beta, - dense_matrix_handle_t C_handle, spmm_alg alg, - spmm_descr_t spmm_descr, void *workspace, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spmm(sycl::queue &queue, oneapi::math::transpose opA, - oneapi::math::transpose opB, const void *alpha, matrix_view A_view, - matrix_handle_t A_handle, dense_matrix_handle_t B_handle, - const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, - spmm_descr_t spmm_descr, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void init_spmm_descr(sycl::queue& queue, spmm_descr_t* p_spmm_descr); + +ONEMATH_EXPORT sycl::event release_spmm_descr(sycl::queue& queue, spmm_descr_t spmm_descr, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT void spmm_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void* alpha, + matrix_view A_view, matrix_handle_t A_handle, + dense_matrix_handle_t B_handle, const void* beta, + dense_matrix_handle_t C_handle, spmm_alg alg, + spmm_descr_t spmm_descr, std::size_t& temp_buffer_size); + +ONEMATH_EXPORT void spmm_optimize(sycl::queue& queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void* alpha, + matrix_view A_view, matrix_handle_t A_handle, + dense_matrix_handle_t B_handle, const void* beta, + dense_matrix_handle_t C_handle, spmm_alg alg, + spmm_descr_t spmm_descr, sycl::buffer workspace); + +ONEMATH_EXPORT sycl::event spmm_optimize(sycl::queue& queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void* alpha, + matrix_view A_view, matrix_handle_t A_handle, + dense_matrix_handle_t B_handle, const void* beta, + dense_matrix_handle_t C_handle, spmm_alg alg, + spmm_descr_t spmm_descr, void* workspace, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spmm(sycl::queue& queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void* alpha, matrix_view A_view, + matrix_handle_t A_handle, dense_matrix_handle_t B_handle, + const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, + spmm_descr_t spmm_descr, + const std::vector& dependencies = {}); // SPMV -ONEMATH_EXPORT void init_spmv_descr(sycl::queue &queue, spmv_descr_t *p_spmv_descr); - -ONEMATH_EXPORT sycl::event release_spmv_descr(sycl::queue &queue, spmv_descr_t spmv_descr, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT void spmv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_vector_handle_t x_handle, const void *beta, - dense_vector_handle_t y_handle, spmv_alg alg, - spmv_descr_t spmv_descr, std::size_t &temp_buffer_size); - -ONEMATH_EXPORT void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, - matrix_view A_view, matrix_handle_t A_handle, - dense_vector_handle_t x_handle, const void *beta, - dense_vector_handle_t y_handle, spmv_alg alg, - spmv_descr_t spmv_descr, sycl::buffer workspace); - -ONEMATH_EXPORT sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, - const void *alpha, matrix_view A_view, - matrix_handle_t A_handle, dense_vector_handle_t x_handle, - const void *beta, dense_vector_handle_t y_handle, - spmv_alg alg, spmv_descr_t spmv_descr, void *workspace, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spmv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, - matrix_view A_view, matrix_handle_t A_handle, - dense_vector_handle_t x_handle, const void *beta, - dense_vector_handle_t y_handle, spmv_alg alg, - spmv_descr_t spmv_descr, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void init_spmv_descr(sycl::queue& queue, spmv_descr_t* p_spmv_descr); + +ONEMATH_EXPORT sycl::event release_spmv_descr(sycl::queue& queue, spmv_descr_t spmv_descr, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT void spmv_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, + const void* alpha, matrix_view A_view, + matrix_handle_t A_handle, dense_vector_handle_t x_handle, + const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, + spmv_descr_t spmv_descr, std::size_t& temp_buffer_size); + +ONEMATH_EXPORT void spmv_optimize(sycl::queue& queue, oneapi::math::transpose opA, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_vector_handle_t x_handle, const void* beta, + dense_vector_handle_t y_handle, spmv_alg alg, + spmv_descr_t spmv_descr, sycl::buffer workspace); + +ONEMATH_EXPORT sycl::event spmv_optimize(sycl::queue& queue, oneapi::math::transpose opA, + const void* alpha, matrix_view A_view, + matrix_handle_t A_handle, dense_vector_handle_t x_handle, + const void* beta, dense_vector_handle_t y_handle, + spmv_alg alg, spmv_descr_t spmv_descr, void* workspace, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spmv(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, + matrix_view A_view, matrix_handle_t A_handle, + dense_vector_handle_t x_handle, const void* beta, + dense_vector_handle_t y_handle, spmv_alg alg, + spmv_descr_t spmv_descr, + const std::vector& dependencies = {}); // SPSV -ONEMATH_EXPORT void init_spsv_descr(sycl::queue &queue, spsv_descr_t *p_spsv_descr); - -ONEMATH_EXPORT sycl::event release_spsv_descr(sycl::queue &queue, spsv_descr_t spsv_descr, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT void spsv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, - spsv_alg alg, spsv_descr_t spsv_descr, - std::size_t &temp_buffer_size); - -ONEMATH_EXPORT void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, - matrix_view A_view, matrix_handle_t A_handle, - dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, - spsv_alg alg, spsv_descr_t spsv_descr, - sycl::buffer workspace); - -ONEMATH_EXPORT sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, - const void *alpha, matrix_view A_view, - matrix_handle_t A_handle, dense_vector_handle_t x_handle, - dense_vector_handle_t y_handle, spsv_alg alg, - spsv_descr_t spsv_descr, void *workspace, - const std::vector &dependencies = {}); - -ONEMATH_EXPORT sycl::event spsv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, - matrix_view A_view, matrix_handle_t A_handle, - dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, - spsv_alg alg, spsv_descr_t spsv_descr, - const std::vector &dependencies = {}); +ONEMATH_EXPORT void init_spsv_descr(sycl::queue& queue, spsv_descr_t* p_spsv_descr); + +ONEMATH_EXPORT sycl::event release_spsv_descr(sycl::queue& queue, spsv_descr_t spsv_descr, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT void spsv_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, + const void* alpha, matrix_view A_view, + matrix_handle_t A_handle, dense_vector_handle_t x_handle, + dense_vector_handle_t y_handle, spsv_alg alg, + spsv_descr_t spsv_descr, std::size_t& temp_buffer_size); + +ONEMATH_EXPORT void spsv_optimize(sycl::queue& queue, oneapi::math::transpose opA, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, + spsv_alg alg, spsv_descr_t spsv_descr, + sycl::buffer workspace); + +ONEMATH_EXPORT sycl::event spsv_optimize(sycl::queue& queue, oneapi::math::transpose opA, + const void* alpha, matrix_view A_view, + matrix_handle_t A_handle, dense_vector_handle_t x_handle, + dense_vector_handle_t y_handle, spsv_alg alg, + spsv_descr_t spsv_descr, void* workspace, + const std::vector& dependencies = {}); + +ONEMATH_EXPORT sycl::event spsv(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, + matrix_view A_view, matrix_handle_t A_handle, + dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, + spsv_alg alg, spsv_descr_t spsv_descr, + const std::vector& dependencies = {}); diff --git a/include/oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx b/include/oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx index 5f28b1ecb..6d7b05d3d 100644 --- a/include/oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx +++ b/include/oneapi/math/sparse_blas/detail/sparse_blas_ct.hxx @@ -27,14 +27,14 @@ // Dense vector template std::enable_if_t> init_dense_vector( - backend_selector selector, dense_vector_handle_t *p_dvhandle, + backend_selector selector, dense_vector_handle_t* p_dvhandle, std::int64_t size, sycl::buffer val) { BACKEND::init_dense_vector(selector.get_queue(), p_dvhandle, size, val); } template std::enable_if_t> init_dense_vector( - backend_selector selector, dense_vector_handle_t *p_dvhandle, - std::int64_t size, dataType *val) { + backend_selector selector, dense_vector_handle_t* p_dvhandle, + std::int64_t size, dataType* val) { BACKEND::init_dense_vector(selector.get_queue(), p_dvhandle, size, val); } @@ -47,20 +47,20 @@ std::enable_if_t> set_dense_vector_data( template std::enable_if_t> set_dense_vector_data( backend_selector selector, dense_vector_handle_t dvhandle, std::int64_t size, - dataType *val) { + dataType* val) { BACKEND::set_dense_vector_data(selector.get_queue(), dvhandle, size, val); } inline sycl::event release_dense_vector(backend_selector selector, dense_vector_handle_t dvhandle, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return BACKEND::release_dense_vector(selector.get_queue(), dvhandle, dependencies); } // Dense matrix template std::enable_if_t> init_dense_matrix( - backend_selector selector, dense_matrix_handle_t *p_dmhandle, + backend_selector selector, dense_matrix_handle_t* p_dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, layout dense_layout, sycl::buffer val) { BACKEND::init_dense_matrix(selector.get_queue(), p_dmhandle, num_rows, num_cols, ld, @@ -68,9 +68,9 @@ std::enable_if_t> init_dense_matrix( } template std::enable_if_t> init_dense_matrix( - backend_selector selector, dense_matrix_handle_t *p_dmhandle, + backend_selector selector, dense_matrix_handle_t* p_dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, layout dense_layout, - dataType *val) { + dataType* val) { BACKEND::init_dense_matrix(selector.get_queue(), p_dmhandle, num_rows, num_cols, ld, dense_layout, val); } @@ -87,21 +87,21 @@ template std::enable_if_t> set_dense_matrix_data( backend_selector selector, dense_matrix_handle_t dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, layout dense_layout, - dataType *val) { + dataType* val) { BACKEND::set_dense_matrix_data(selector.get_queue(), dmhandle, num_rows, num_cols, ld, dense_layout, val); } inline sycl::event release_dense_matrix(backend_selector selector, dense_matrix_handle_t dmhandle, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return BACKEND::release_dense_matrix(selector.get_queue(), dmhandle, dependencies); } // COO matrix template std::enable_if_t> init_coo_matrix( - backend_selector selector, matrix_handle_t *p_smhandle, std::int64_t num_rows, + backend_selector selector, matrix_handle_t* p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, sycl::buffer row_ind, sycl::buffer col_ind, sycl::buffer val) { BACKEND::init_coo_matrix(selector.get_queue(), p_smhandle, num_rows, num_cols, nnz, index, @@ -109,9 +109,9 @@ std::enable_if_t> init_coo_m } template std::enable_if_t> init_coo_matrix( - backend_selector selector, matrix_handle_t *p_smhandle, std::int64_t num_rows, - std::int64_t num_cols, std::int64_t nnz, index_base index, indexType *row_ind, - indexType *col_ind, dataType *val) { + backend_selector selector, matrix_handle_t* p_smhandle, std::int64_t num_rows, + std::int64_t num_cols, std::int64_t nnz, index_base index, indexType* row_ind, + indexType* col_ind, dataType* val) { BACKEND::init_coo_matrix(selector.get_queue(), p_smhandle, num_rows, num_cols, nnz, index, row_ind, col_ind, val); } @@ -127,8 +127,8 @@ std::enable_if_t> set_coo_ma template std::enable_if_t> set_coo_matrix_data( backend_selector selector, matrix_handle_t smhandle, std::int64_t num_rows, - std::int64_t num_cols, std::int64_t nnz, index_base index, indexType *row_ind, - indexType *col_ind, dataType *val) { + std::int64_t num_cols, std::int64_t nnz, index_base index, indexType* row_ind, + indexType* col_ind, dataType* val) { BACKEND::set_coo_matrix_data(selector.get_queue(), smhandle, num_rows, num_cols, nnz, index, row_ind, col_ind, val); } @@ -136,7 +136,7 @@ std::enable_if_t> set_coo_ma // CSR matrix template std::enable_if_t> init_csr_matrix( - backend_selector selector, matrix_handle_t *p_smhandle, std::int64_t num_rows, + backend_selector selector, matrix_handle_t* p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, sycl::buffer row_ptr, sycl::buffer col_ind, sycl::buffer val) { BACKEND::init_csr_matrix(selector.get_queue(), p_smhandle, num_rows, num_cols, nnz, index, @@ -144,9 +144,9 @@ std::enable_if_t> init_csr_m } template std::enable_if_t> init_csr_matrix( - backend_selector selector, matrix_handle_t *p_smhandle, std::int64_t num_rows, - std::int64_t num_cols, std::int64_t nnz, index_base index, indexType *row_ptr, - indexType *col_ind, dataType *val) { + backend_selector selector, matrix_handle_t* p_smhandle, std::int64_t num_rows, + std::int64_t num_cols, std::int64_t nnz, index_base index, indexType* row_ptr, + indexType* col_ind, dataType* val) { BACKEND::init_csr_matrix(selector.get_queue(), p_smhandle, num_rows, num_cols, nnz, index, row_ptr, col_ind, val); } @@ -162,8 +162,8 @@ std::enable_if_t> set_csr_ma template std::enable_if_t> set_csr_matrix_data( backend_selector selector, matrix_handle_t smhandle, std::int64_t num_rows, - std::int64_t num_cols, std::int64_t nnz, index_base index, indexType *row_ptr, - indexType *col_ind, dataType *val) { + std::int64_t num_cols, std::int64_t nnz, index_base index, indexType* row_ptr, + indexType* col_ind, dataType* val) { BACKEND::set_csr_matrix_data(selector.get_queue(), smhandle, num_rows, num_cols, nnz, index, row_ptr, col_ind, val); } @@ -171,7 +171,7 @@ std::enable_if_t> set_csr_ma // Common sparse matrix functions inline sycl::event release_sparse_matrix(backend_selector selector, matrix_handle_t smhandle, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return BACKEND::release_sparse_matrix(selector.get_queue(), smhandle, dependencies); } @@ -182,30 +182,30 @@ inline bool set_matrix_property(backend_selector selector, // SPMM inline void init_spmm_descr(backend_selector selector, - spmm_descr_t *p_spmm_descr) { + spmm_descr_t* p_spmm_descr) { BACKEND::init_spmm_descr(selector.get_queue(), p_spmm_descr); } inline sycl::event release_spmm_descr(backend_selector selector, spmm_descr_t spmm_descr, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return BACKEND::release_spmm_descr(selector.get_queue(), spmm_descr, dependencies); } inline void spmm_buffer_size(backend_selector selector, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_matrix_handle_t B_handle, const void *beta, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_matrix_handle_t B_handle, const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, - std::size_t &temp_buffer_size) { + std::size_t& temp_buffer_size) { BACKEND::spmm_buffer_size(selector.get_queue(), opA, opB, alpha, A_view, A_handle, B_handle, beta, C_handle, alg, spmm_descr, temp_buffer_size); } inline void spmm_optimize(backend_selector selector, oneapi::math::transpose opA, - oneapi::math::transpose opB, const void *alpha, matrix_view A_view, + oneapi::math::transpose opB, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, - const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, + const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, sycl::buffer workspace) { BACKEND::spmm_optimize(selector.get_queue(), opA, opB, alpha, A_view, A_handle, B_handle, beta, C_handle, alg, spmm_descr, workspace); @@ -213,48 +213,48 @@ inline void spmm_optimize(backend_selector selector, oneapi::m inline sycl::event spmm_optimize(backend_selector selector, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_matrix_handle_t B_handle, const void *beta, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_matrix_handle_t B_handle, const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, - spmm_descr_t spmm_descr, void *workspace, - const std::vector &dependencies = {}) { + spmm_descr_t spmm_descr, void* workspace, + const std::vector& dependencies = {}) { return BACKEND::spmm_optimize(selector.get_queue(), opA, opB, alpha, A_view, A_handle, B_handle, beta, C_handle, alg, spmm_descr, workspace, dependencies); } inline sycl::event spmm(backend_selector selector, oneapi::math::transpose opA, - oneapi::math::transpose opB, const void *alpha, matrix_view A_view, - matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::transpose opB, const void* alpha, matrix_view A_view, + matrix_handle_t A_handle, dense_matrix_handle_t B_handle, const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return BACKEND::spmm(selector.get_queue(), opA, opB, alpha, A_view, A_handle, B_handle, beta, C_handle, alg, spmm_descr, dependencies); } // SPMV inline void init_spmv_descr(backend_selector selector, - spmv_descr_t *p_spmv_descr) { + spmv_descr_t* p_spmv_descr) { BACKEND::init_spmv_descr(selector.get_queue(), p_spmv_descr); } inline sycl::event release_spmv_descr(backend_selector selector, spmv_descr_t spmv_descr, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return BACKEND::release_spmv_descr(selector.get_queue(), spmv_descr, dependencies); } inline void spmv_buffer_size(backend_selector selector, - oneapi::math::transpose opA, const void *alpha, matrix_view A_view, + oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, - const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, - spmv_descr_t spmv_descr, std::size_t &temp_buffer_size) { + const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, + spmv_descr_t spmv_descr, std::size_t& temp_buffer_size) { BACKEND::spmv_buffer_size(selector.get_queue(), opA, alpha, A_view, A_handle, x_handle, beta, y_handle, alg, spmv_descr, temp_buffer_size); } inline void spmv_optimize(backend_selector selector, oneapi::math::transpose opA, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_vector_handle_t x_handle, const void *beta, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_vector_handle_t x_handle, const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, sycl::buffer workspace) { BACKEND::spmv_optimize(selector.get_queue(), opA, alpha, A_view, A_handle, x_handle, beta, @@ -262,47 +262,47 @@ inline void spmv_optimize(backend_selector selector, oneapi::m } inline sycl::event spmv_optimize(backend_selector selector, - oneapi::math::transpose opA, const void *alpha, matrix_view A_view, + oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, - const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, - spmv_descr_t spmv_descr, void *workspace, - const std::vector &dependencies = {}) { + const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, + spmv_descr_t spmv_descr, void* workspace, + const std::vector& dependencies = {}) { return BACKEND::spmv_optimize(selector.get_queue(), opA, alpha, A_view, A_handle, x_handle, beta, y_handle, alg, spmv_descr, workspace, dependencies); } inline sycl::event spmv(backend_selector selector, oneapi::math::transpose opA, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_vector_handle_t x_handle, const void *beta, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_vector_handle_t x_handle, const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return BACKEND::spmv(selector.get_queue(), opA, alpha, A_view, A_handle, x_handle, beta, y_handle, alg, spmv_descr, dependencies); } // SPSV inline void init_spsv_descr(backend_selector selector, - spsv_descr_t *p_spsv_descr) { + spsv_descr_t* p_spsv_descr) { BACKEND::init_spsv_descr(selector.get_queue(), p_spsv_descr); } inline sycl::event release_spsv_descr(backend_selector selector, spsv_descr_t spsv_descr, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return BACKEND::release_spsv_descr(selector.get_queue(), spsv_descr, dependencies); } inline void spsv_buffer_size(backend_selector selector, - oneapi::math::transpose opA, const void *alpha, matrix_view A_view, + oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, - std::size_t &temp_buffer_size) { + std::size_t& temp_buffer_size) { BACKEND::spsv_buffer_size(selector.get_queue(), opA, alpha, A_view, A_handle, x_handle, y_handle, alg, spsv_descr, temp_buffer_size); } inline void spsv_optimize(backend_selector selector, oneapi::math::transpose opA, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, sycl::buffer workspace) { @@ -311,20 +311,20 @@ inline void spsv_optimize(backend_selector selector, oneapi::m } inline sycl::event spsv_optimize(backend_selector selector, - oneapi::math::transpose opA, const void *alpha, matrix_view A_view, + oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, - spsv_descr_t spsv_descr, void *workspace, - const std::vector &dependencies = {}) { + spsv_descr_t spsv_descr, void* workspace, + const std::vector& dependencies = {}) { return BACKEND::spsv_optimize(selector.get_queue(), opA, alpha, A_view, A_handle, x_handle, y_handle, alg, spsv_descr, workspace, dependencies); } inline sycl::event spsv(backend_selector selector, oneapi::math::transpose opA, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, - const std::vector &dependencies = {}) { + const std::vector& dependencies = {}) { return BACKEND::spsv(selector.get_queue(), opA, alpha, A_view, A_handle, x_handle, y_handle, alg, spsv_descr, dependencies); } diff --git a/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp b/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp index df3341814..8066691e5 100644 --- a/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp +++ b/include/oneapi/math/sparse_blas/detail/sparse_blas_rt.hpp @@ -30,176 +30,176 @@ namespace sparse { // Dense vector template std::enable_if_t> init_dense_vector( - sycl::queue &queue, dense_vector_handle_t *p_dvhandle, std::int64_t size, + sycl::queue& queue, dense_vector_handle_t* p_dvhandle, std::int64_t size, sycl::buffer val); template std::enable_if_t> init_dense_vector( - sycl::queue &queue, dense_vector_handle_t *p_dvhandle, std::int64_t size, dataType *val); + sycl::queue& queue, dense_vector_handle_t* p_dvhandle, std::int64_t size, dataType* val); template std::enable_if_t> set_dense_vector_data( - sycl::queue &queue, dense_vector_handle_t dvhandle, std::int64_t size, + sycl::queue& queue, dense_vector_handle_t dvhandle, std::int64_t size, sycl::buffer val); template std::enable_if_t> set_dense_vector_data( - sycl::queue &queue, dense_vector_handle_t dvhandle, std::int64_t size, dataType *val); + sycl::queue& queue, dense_vector_handle_t dvhandle, std::int64_t size, dataType* val); -sycl::event release_dense_vector(sycl::queue &queue, dense_vector_handle_t dvhandle, - const std::vector &dependencies = {}); +sycl::event release_dense_vector(sycl::queue& queue, dense_vector_handle_t dvhandle, + const std::vector& dependencies = {}); // Dense matrix template std::enable_if_t> init_dense_matrix( - sycl::queue &queue, dense_matrix_handle_t *p_dmhandle, std::int64_t num_rows, + sycl::queue& queue, dense_matrix_handle_t* p_dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, layout dense_layout, sycl::buffer val); template std::enable_if_t> init_dense_matrix( - sycl::queue &queue, dense_matrix_handle_t *p_dmhandle, std::int64_t num_rows, - std::int64_t num_cols, std::int64_t ld, layout dense_layout, dataType *val); + sycl::queue& queue, dense_matrix_handle_t* p_dmhandle, std::int64_t num_rows, + std::int64_t num_cols, std::int64_t ld, layout dense_layout, dataType* val); template std::enable_if_t> set_dense_matrix_data( - sycl::queue &queue, dense_matrix_handle_t dmhandle, std::int64_t num_rows, + sycl::queue& queue, dense_matrix_handle_t dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, layout dense_layout, sycl::buffer val); template std::enable_if_t> set_dense_matrix_data( - sycl::queue &queue, dense_matrix_handle_t dmhandle, std::int64_t num_rows, - std::int64_t num_cols, std::int64_t ld, layout dense_layout, dataType *val); + sycl::queue& queue, dense_matrix_handle_t dmhandle, std::int64_t num_rows, + std::int64_t num_cols, std::int64_t ld, layout dense_layout, dataType* val); -sycl::event release_dense_matrix(sycl::queue &queue, dense_matrix_handle_t dmhandle, - const std::vector &dependencies = {}); +sycl::event release_dense_matrix(sycl::queue& queue, dense_matrix_handle_t dmhandle, + const std::vector& dependencies = {}); // COO matrix template std::enable_if_t> init_coo_matrix( - sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, + sycl::queue& queue, matrix_handle_t* p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, sycl::buffer row_ind, sycl::buffer col_ind, sycl::buffer val); template std::enable_if_t> init_coo_matrix( - sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, - std::int64_t nnz, index_base index, indexType *row_ind, indexType *col_ind, dataType *val); + sycl::queue& queue, matrix_handle_t* p_smhandle, std::int64_t num_rows, std::int64_t num_cols, + std::int64_t nnz, index_base index, indexType* row_ind, indexType* col_ind, dataType* val); template std::enable_if_t> set_coo_matrix_data( - sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, + sycl::queue& queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, sycl::buffer row_ind, sycl::buffer col_ind, sycl::buffer val); template std::enable_if_t> set_coo_matrix_data( - sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, - std::int64_t nnz, index_base index, indexType *row_ind, indexType *col_ind, dataType *val); + sycl::queue& queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, + std::int64_t nnz, index_base index, indexType* row_ind, indexType* col_ind, dataType* val); // CSR matrix template std::enable_if_t> init_csr_matrix( - sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, + sycl::queue& queue, matrix_handle_t* p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, sycl::buffer row_ptr, sycl::buffer col_ind, sycl::buffer val); template std::enable_if_t> init_csr_matrix( - sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, std::int64_t num_cols, - std::int64_t nnz, index_base index, indexType *row_ptr, indexType *col_ind, dataType *val); + sycl::queue& queue, matrix_handle_t* p_smhandle, std::int64_t num_rows, std::int64_t num_cols, + std::int64_t nnz, index_base index, indexType* row_ptr, indexType* col_ind, dataType* val); template std::enable_if_t> set_csr_matrix_data( - sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, + sycl::queue& queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, index_base index, sycl::buffer row_ptr, sycl::buffer col_ind, sycl::buffer val); template std::enable_if_t> set_csr_matrix_data( - sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, - std::int64_t nnz, index_base index, indexType *row_ptr, indexType *col_ind, dataType *val); + sycl::queue& queue, matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, + std::int64_t nnz, index_base index, indexType* row_ptr, indexType* col_ind, dataType* val); // Common sparse matrix functions -sycl::event release_sparse_matrix(sycl::queue &queue, matrix_handle_t smhandle, - const std::vector &dependencies = {}); +sycl::event release_sparse_matrix(sycl::queue& queue, matrix_handle_t smhandle, + const std::vector& dependencies = {}); -bool set_matrix_property(sycl::queue &queue, matrix_handle_t smhandle, matrix_property property); +bool set_matrix_property(sycl::queue& queue, matrix_handle_t smhandle, matrix_property property); // SPMM -void init_spmm_descr(sycl::queue &queue, spmm_descr_t *p_spmm_descr); +void init_spmm_descr(sycl::queue& queue, spmm_descr_t* p_spmm_descr); -sycl::event release_spmm_descr(sycl::queue &queue, spmm_descr_t spmm_descr, - const std::vector &dependencies = {}); +sycl::event release_spmm_descr(sycl::queue& queue, spmm_descr_t spmm_descr, + const std::vector& dependencies = {}); -void spmm_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_matrix_handle_t B_handle, const void *beta, +void spmm_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_matrix_handle_t B_handle, const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, - std::size_t &temp_buffer_size); + std::size_t& temp_buffer_size); -void spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, +void spmm_optimize(sycl::queue& queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_matrix_handle_t B_handle, const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, sycl::buffer workspace); -sycl::event spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, - oneapi::math::transpose opB, const void *alpha, matrix_view A_view, +sycl::event spmm_optimize(sycl::queue& queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, - const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, - spmm_descr_t spmm_descr, void *workspace, - const std::vector &dependencies = {}); + const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, + spmm_descr_t spmm_descr, void* workspace, + const std::vector& dependencies = {}); -sycl::event spmm(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, +sycl::event spmm(sycl::queue& queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_matrix_handle_t B_handle, const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); // SPMV -void init_spmv_descr(sycl::queue &queue, spmv_descr_t *p_spmv_descr); +void init_spmv_descr(sycl::queue& queue, spmv_descr_t* p_spmv_descr); -sycl::event release_spmv_descr(sycl::queue &queue, spmv_descr_t spmv_descr, - const std::vector &dependencies = {}); +sycl::event release_spmv_descr(sycl::queue& queue, spmv_descr_t spmv_descr, + const std::vector& dependencies = {}); -void spmv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spmv_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, - const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, - spmv_descr_t spmv_descr, std::size_t &temp_buffer_size); + const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, + spmv_descr_t spmv_descr, std::size_t& temp_buffer_size); -void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spmv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, - const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, + const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, sycl::buffer workspace); -sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spmv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_vector_handle_t x_handle, const void *beta, + dense_vector_handle_t x_handle, const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, - void *workspace, const std::vector &dependencies = {}); + void* workspace, const std::vector& dependencies = {}); -sycl::event spmv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spmv(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, - const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, - spmv_descr_t spmv_descr, const std::vector &dependencies = {}); + const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, + spmv_descr_t spmv_descr, const std::vector& dependencies = {}); // SPSV -void init_spsv_descr(sycl::queue &queue, spsv_descr_t *p_spsv_descr); +void init_spsv_descr(sycl::queue& queue, spsv_descr_t* p_spsv_descr); -sycl::event release_spsv_descr(sycl::queue &queue, spsv_descr_t spsv_descr, - const std::vector &dependencies = {}); +sycl::event release_spsv_descr(sycl::queue& queue, spsv_descr_t spsv_descr, + const std::vector& dependencies = {}); -void spsv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spsv_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, - std::size_t &temp_buffer_size); + std::size_t& temp_buffer_size); -void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spsv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, sycl::buffer workspace); -sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spsv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, - spsv_alg alg, spsv_descr_t spsv_descr, void *workspace, - const std::vector &dependencies = {}); + spsv_alg alg, spsv_descr_t spsv_descr, void* workspace, + const std::vector& dependencies = {}); -sycl::event spsv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spsv(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); } // namespace sparse } // namespace math diff --git a/include/oneapi/mkl/blas.hpp b/include/oneapi/mkl/blas.hpp index a8befa71e..71b2fed1e 100644 --- a/include/oneapi/mkl/blas.hpp +++ b/include/oneapi/mkl/blas.hpp @@ -21,7 +21,8 @@ #define ONEMATH_MKL_BLAS_HPP // Deprecated header is planned to be removed late 2025. -#pragma message("Header `oneapi/mkl/blas.hpp` is deprecated, please use `oneapi/math/blas.hpp` instead") +#pragma message( \ + "Header `oneapi/mkl/blas.hpp` is deprecated, please use `oneapi/math/blas.hpp` instead") #include "oneapi/math/blas.hpp" diff --git a/include/oneapi/mkl/dft.hpp b/include/oneapi/mkl/dft.hpp index 09ec86cc7..d86952974 100644 --- a/include/oneapi/mkl/dft.hpp +++ b/include/oneapi/mkl/dft.hpp @@ -21,7 +21,8 @@ #define ONEMATH_MKL_DFT_HPP // Deprecated header is planned to be removed late 2025. -#pragma message("Header `oneapi/mkl/dft.hpp` is deprecated, please use `oneapi/math/dft.hpp` instead") +#pragma message( \ + "Header `oneapi/mkl/dft.hpp` is deprecated, please use `oneapi/math/dft.hpp` instead") #include "oneapi/math/dft.hpp" diff --git a/include/oneapi/mkl/lapack.hpp b/include/oneapi/mkl/lapack.hpp index 990511249..0a358f564 100644 --- a/include/oneapi/mkl/lapack.hpp +++ b/include/oneapi/mkl/lapack.hpp @@ -21,7 +21,8 @@ #define ONEMATH_MKL_LAPACK_HPP // Deprecated header is planned to be removed late 2025. -#pragma message("Header `oneapi/mkl/lapack.hpp` is deprecated, please use `oneapi/math/lapack.hpp` instead") +#pragma message( \ + "Header `oneapi/mkl/lapack.hpp` is deprecated, please use `oneapi/math/lapack.hpp` instead") #include "oneapi/math/lapack.hpp" diff --git a/include/oneapi/mkl/namespace_alias.hpp b/include/oneapi/mkl/namespace_alias.hpp index fa52bad2c..3b09b5867 100644 --- a/include/oneapi/mkl/namespace_alias.hpp +++ b/include/oneapi/mkl/namespace_alias.hpp @@ -18,8 +18,9 @@ **************************************************************************/ namespace oneapi { - // Deprecated namespace is planned to be removed late 2025. - namespace [[deprecated("Namespace `oneapi::mkl` is deprecated, please use `oneapi::math` instead")]] mkl { - using namespace math; - } +// Deprecated namespace is planned to be removed late 2025. +namespace + [[deprecated("Namespace `oneapi::mkl` is deprecated, please use `oneapi::math` instead")]] mkl { +using namespace math; } +} // namespace oneapi diff --git a/include/oneapi/mkl/rng.hpp b/include/oneapi/mkl/rng.hpp index 094a446a1..be4760dad 100644 --- a/include/oneapi/mkl/rng.hpp +++ b/include/oneapi/mkl/rng.hpp @@ -21,7 +21,8 @@ #define ONEMATH_MKL_RNG_HPP // Deprecated header is planned to be removed late 2025. -#pragma message("Header `oneapi/mkl/rng.hpp` is deprecated, please use `oneapi/math/rng.hpp` instead") +#pragma message( \ + "Header `oneapi/mkl/rng.hpp` is deprecated, please use `oneapi/math/rng.hpp` instead") #include "oneapi/math/rng.hpp" diff --git a/include/oneapi/mkl/rng/device.hpp b/include/oneapi/mkl/rng/device.hpp index 012649884..9024e066b 100644 --- a/include/oneapi/mkl/rng/device.hpp +++ b/include/oneapi/mkl/rng/device.hpp @@ -21,7 +21,8 @@ #define ONEMATH_MKL_RNG_DEVICE_HPP // Deprecated header is planned to be removed late 2025. -#pragma message("Header `oneapi/mkl/rng/device.hpp` is deprecated, please use `oneapi/math/rng/math.hpp` instead") +#pragma message( \ + "Header `oneapi/mkl/rng/device.hpp` is deprecated, please use `oneapi/math/rng/math.hpp` instead") #include "oneapi/math/rng/device.hpp" diff --git a/include/oneapi/mkl/sparse_blas.hpp b/include/oneapi/mkl/sparse_blas.hpp index 805f55a0a..1ec890462 100644 --- a/include/oneapi/mkl/sparse_blas.hpp +++ b/include/oneapi/mkl/sparse_blas.hpp @@ -21,7 +21,8 @@ #define ONEMATH_MKL_SPARSE_BLAS_HPP // Deprecated header is planned to be removed late 2025. -#pragma message("Header `oneapi/mkl/sparse_blas.hpp` is deprecated, please use `oneapi/math/sparse_blas.hpp` instead") +#pragma message( \ + "Header `oneapi/mkl/sparse_blas.hpp` is deprecated, please use `oneapi/math/sparse_blas.hpp` instead") #include "oneapi/math/sparse_blas.hpp" diff --git a/src/blas/backends/cublas/cublas_batch.cpp b/src/blas/backends/cublas/cublas_batch.cpp index dc53e0679..822817d6f 100644 --- a/src/blas/backends/cublas/cublas_batch.cpp +++ b/src/blas/backends/cublas/cublas_batch.cpp @@ -29,122 +29,122 @@ namespace column_major { // Buffer APIs -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, - int64_t incx, int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, + int64_t incx, int64_t stridex, sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, - int64_t incx, int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, + int64_t incx, int64_t stridex, sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -void axpy_batch(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -void axpy_batch(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, +void axpy_batch(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, - int64_t incx, int64_t stride_x, float beta, sycl::buffer &y, int64_t incy, +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, sycl::buffer& x, + int64_t incx, int64_t stride_x, float beta, sycl::buffer& y, int64_t incy, int64_t stride_y, int64_t batch_size) { throw unimplemented("blas", "gemv_batch", "for column_major layout"); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &x, int64_t incx, int64_t stride_x, double beta, - sycl::buffer &y, int64_t incy, int64_t stride_y, int64_t batch_size) { +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& x, int64_t incx, int64_t stride_x, double beta, + sycl::buffer& y, int64_t incy, int64_t stride_y, int64_t batch_size) { throw unimplemented("blas", "gemv_batch", "for column_major layout"); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &x, int64_t incx, - int64_t stride_x, std::complex beta, sycl::buffer, 1> &y, +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& x, int64_t incx, + int64_t stride_x, std::complex beta, sycl::buffer, 1>& y, int64_t incy, int64_t stride_y, int64_t batch_size) { throw unimplemented("blas", "gemv_batch", "for column_major layout"); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &x, int64_t incx, +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& x, int64_t incx, int64_t stride_x, std::complex beta, - sycl::buffer, 1> &y, int64_t incy, int64_t stride_y, + sycl::buffer, 1>& y, int64_t incy, int64_t stride_y, int64_t batch_size) { throw unimplemented("blas", "gemv_batch", "for column_major layout"); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, - int64_t incx, int64_t stride_x, sycl::buffer &c, int64_t ldc, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stride_a, sycl::buffer& x, + int64_t incx, int64_t stride_x, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &x, int64_t incx, int64_t stride_x, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& x, int64_t incx, int64_t stride_x, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& x, int64_t incx, int64_t stride_x, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& x, int64_t incx, int64_t stride_x, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } template -inline void gemm_batch_impl(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, Ts alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - Ts beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +inline void gemm_batch_impl(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, int64_t k, Ts alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + Ts beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { using cuTypeA = typename CudaEquivalentType::Type; using cuTypeB = typename CudaEquivalentType::Type; @@ -153,7 +153,7 @@ inline void gemm_batch_impl(sycl::queue &queue, transpose transa, transpose tran overflow_check(m, n, k, lda, ldb, ldc, stride_a, stride_b, stride_c, batch_size); cublasGemmAlgo_t cublas_gemm_algo = CUBLAS_GEMM_DEFAULT; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { if (!verify_support(queue, sycl::aspect::fp16)) { throw oneapi::math::unimplemented( "blas", "sycl::half", "half is not supported by the device or the sycl compiler"); @@ -161,19 +161,19 @@ inline void gemm_batch_impl(sycl::queue &queue, transpose transa, transpose tran auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); cublasStatus_t err; #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - CUBLAS_ERROR_FUNC_T( - "cublasGemmStridedBatchedEx", cublasGemmStridedBatchedEx, err, handle, - get_cublas_operation(transa), get_cublas_operation(transb), m, n, k, &alpha, a_, - get_cublas_datatype(), lda, stride_a, b_, get_cublas_datatype(), - ldb, stride_b, &beta, c_, get_cublas_datatype(), ldc, stride_c, batch_size, - get_cublas_datatype(), cublas_gemm_algo); + CUBLAS_ERROR_FUNC_T("cublasGemmStridedBatchedEx", cublasGemmStridedBatchedEx, err, + handle, get_cublas_operation(transa), get_cublas_operation(transb), + m, n, k, &alpha, a_, get_cublas_datatype(), lda, stride_a, + b_, get_cublas_datatype(), ldb, stride_b, &beta, c_, + get_cublas_datatype(), ldc, stride_c, batch_size, + get_cublas_datatype(), cublas_gemm_algo); #else CUBLAS_ERROR_FUNC_T_SYNC( "cublasGemmStridedBatchedEx", cublasGemmStridedBatchedEx, err, handle, @@ -187,10 +187,10 @@ inline void gemm_batch_impl(sycl::queue &queue, transpose transa, transpose tran } #define GEMM_STRIDED_BATCH_LAUNCHER(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_S alpha, sycl::buffer &a, int64_t lda, \ - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, \ - TYPE_S beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, \ + void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_S alpha, sycl::buffer& a, int64_t lda, \ + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, \ + TYPE_S beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, \ int64_t batch_size) { \ gemm_batch_impl(queue, transa, transb, m, n, k, alpha, a, \ lda, stride_a, b, ldb, stride_b, beta, c, \ @@ -209,10 +209,10 @@ GEMM_STRIDED_BATCH_LAUNCHER(std::complex, std::complex, std::com #undef GEMM_STRIDED_BATCH_LAUNCHER #define GEMM_STRIDED_BATCH_LAUNCHER(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_S alpha, sycl::buffer &a, int64_t lda, \ - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, \ - TYPE_S beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, \ + void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_S alpha, sycl::buffer& a, int64_t lda, \ + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, \ + TYPE_S beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, \ int64_t batch_size) { \ throw unimplemented("blas", "gemm_batch", \ std::string("for dtype unimplemented dtype combination <") + \ @@ -225,307 +225,308 @@ GEMM_STRIDED_BATCH_LAUNCHER(std::int8_t, std::int8_t, std::int32_t, float) #undef GEMM_STRIDED_BATCH_LAUNCHER -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "trsm_batch", "for column_major layout"); } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "trsm_batch", "for column_major layout"); } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "trsm_batch", "for column_major layout"); } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "trsm_batch", "for column_major layout"); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, float beta, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, float beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, std::complex beta, sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - float beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + float beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "omatadd_batch", "for column_major layout"); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "omatadd_batch", "for column_major layout"); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "omatadd_batch", "for column_major layout"); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "omatadd_batch", "for column_major layout"); } // USM APIs -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const float **x, int64_t *incx, float **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const float** x, int64_t* incx, float** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const double **x, int64_t *incx, double **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const double** x, int64_t* incx, double** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, int64_t *incx, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const std::complex** x, int64_t* incx, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, - int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const std::complex** x, + int64_t* incx, std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const float *x, int64_t incx, - std::int64_t stridex, float *y, int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t n, const float* x, int64_t incx, + std::int64_t stridex, float* y, int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const double *x, int64_t incx, - std::int64_t stridex, double *y, int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t n, const double* x, int64_t incx, + std::int64_t stridex, double* y, int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - std::int64_t stridex, std::complex *y, int64_t incy, +sycl::event copy_batch(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::int64_t stridex, std::complex* y, int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - std::int64_t stridex, std::complex *y, int64_t incy, +sycl::event copy_batch(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::int64_t stridex, std::complex* y, int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for column_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, float *alpha, const float **x, int64_t *incx, - float **y, int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, float* alpha, const float** x, int64_t* incx, + float** y, int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, double *alpha, const double **x, - int64_t *incx, double **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, double* alpha, const double** x, + int64_t* incx, double** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, std::complex* alpha, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, std::complex* alpha, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, - int64_t stridex, float *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, + int64_t stridex, float* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - int64_t stridex, double *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + int64_t stridex, double* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for column_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, const float *x, int64_t incx, - int64_t stride_x, float beta, float *y, int64_t incy, int64_t stride_y, - int64_t batch_size, const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stride_a, const float* x, int64_t incx, + int64_t stride_x, float beta, float* y, int64_t incy, int64_t stride_y, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for column_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, const double *x, - int64_t incx, int64_t stride_x, double beta, double *y, int64_t incy, +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stride_a, const double* x, + int64_t incx, int64_t stride_x, double beta, double* y, int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for column_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, const std::complex *x, int64_t incx, - int64_t stride_x, std::complex beta, std::complex *y, +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, const std::complex* x, int64_t incx, + int64_t stride_x, std::complex beta, std::complex* y, int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for column_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, const std::complex *x, int64_t incx, - int64_t stride_x, std::complex beta, std::complex *y, +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, const std::complex* x, int64_t incx, + int64_t stride_x, std::complex beta, std::complex* y, int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for column_major layout"); } template -inline sycl::event gemv_batch(const char *func_name, Func func, sycl::queue &queue, transpose *trans, int64_t *m, - int64_t *n, T *alpha, const T **a, int64_t *lda, const T **x, - int64_t *incx, T *beta, T **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +inline sycl::event gemv_batch(const char* func_name, Func func, sycl::queue& queue, + transpose* trans, int64_t* m, int64_t* n, T* alpha, const T** a, + int64_t* lda, const T** x, int64_t* incx, T* beta, T** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; for (int64_t i = 0; i < group_count; i++) { overflow_check(m[i], n[i], lda[i], incx[i], incy[i], group_size[i]); } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; cublasStatus_t err; - auto **a_ = reinterpret_cast(a); - auto **x_ = reinterpret_cast(x); - auto **y_ = reinterpret_cast(y); + auto** a_ = reinterpret_cast(a); + auto** x_ = reinterpret_cast(x); + auto** y_ = reinterpret_cast(y); for (int64_t i = 0; i < group_count; i++) { - cublas_native_named_func( - func_name, func, err, handle, get_cublas_operation(trans[i]), - (int)m[i], (int)n[i], - (cuDataType *)&alpha[i], a_ + offset, (int)lda[i], x_ + offset, (int)incx[i], - (cuDataType *)&beta[i], y_ + offset, (int)incy[i], (int)group_size[i]); + cublas_native_named_func(func_name, func, err, handle, + get_cublas_operation(trans[i]), (int)m[i], (int)n[i], + (cuDataType*)&alpha[i], a_ + offset, (int)lda[i], + x_ + offset, (int)incx[i], (cuDataType*)&beta[i], + y_ + offset, (int)incy[i], (int)group_size[i]); offset += group_size[i]; } }); @@ -533,14 +534,13 @@ inline sycl::event gemv_batch(const char *func_name, Func func, sycl::queue &que return done; } -#define GEMV_BATCH_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event gemv_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, \ - TYPE *alpha, const TYPE **a, int64_t *lda, const TYPE **x, \ - int64_t *incx, TYPE *beta, TYPE **y, int64_t *incy, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ - return gemv_batch(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, \ - x, incx, beta, y, incy, group_count, group_size, dependencies); \ +#define GEMV_BATCH_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ + sycl::event gemv_batch( \ + sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, TYPE* alpha, const TYPE** a, \ + int64_t* lda, const TYPE** x, int64_t* incx, TYPE* beta, TYPE** y, int64_t* incy, \ + int64_t group_count, int64_t* group_size, const std::vector& dependencies) { \ + return gemv_batch(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, x, \ + incx, beta, y, incy, group_count, group_size, dependencies); \ } GEMV_BATCH_LAUNCHER_USM(float, cublasSgemvBatched) @@ -550,72 +550,72 @@ GEMV_BATCH_LAUNCHER_USM(std::complex, cublasZgemvBatched) #undef GEMV_BATCH_LAUNCHER_USM -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, const float *a, - int64_t lda, int64_t stride_a, const float *x, int64_t incx, - int64_t stride_x, float *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, const float* a, + int64_t lda, int64_t stride_a, const float* x, int64_t incx, + int64_t stride_x, float* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, const double *a, - int64_t lda, int64_t stride_a, const double *x, int64_t incx, - int64_t stride_x, double *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, const double* a, + int64_t lda, int64_t stride_a, const double* x, int64_t incx, + int64_t stride_x, double* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const std::complex *a, int64_t lda, int64_t stride_a, - const std::complex *x, int64_t incx, int64_t stride_x, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + const std::complex* a, int64_t lda, int64_t stride_a, + const std::complex* x, int64_t incx, int64_t stride_x, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const std::complex *a, int64_t lda, int64_t stride_a, - const std::complex *x, int64_t incx, int64_t stride_x, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + const std::complex* a, int64_t lda, int64_t stride_a, + const std::complex* x, int64_t incx, int64_t stride_x, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const float **a, int64_t *lda, const float **x, int64_t *incx, float **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const float** a, int64_t* lda, const float** x, int64_t* incx, float** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const double **a, int64_t *lda, const double **x, int64_t *incx, double **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const double** a, int64_t* lda, const double** x, int64_t* incx, double** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const std::complex **a, int64_t *lda, const std::complex **x, - int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const std::complex** a, int64_t* lda, const std::complex** x, + int64_t* incx, std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const std::complex **a, int64_t *lda, const std::complex **x, - int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const std::complex** a, int64_t* lda, const std::complex** x, + int64_t* incx, std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for column_major layout"); } template -inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose transa, +inline sycl::event gemm_batch_strided_usm_impl(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - Ts alpha, const Ta *a, int64_t lda, int64_t stride_a, - const Tb *b, int64_t ldb, int64_t stride_b, Ts beta, - Tc *c, int64_t ldc, int64_t stride_c, + Ts alpha, const Ta* a, int64_t lda, int64_t stride_a, + const Tb* b, int64_t ldb, int64_t stride_b, Ts beta, + Tc* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuTypeA = typename CudaEquivalentType::Type; using cuTypeB = typename CudaEquivalentType::Type; using cuTypeC = typename CudaEquivalentType::Type; @@ -623,7 +623,7 @@ inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose tra overflow_check(m, n, k, lda, ldb, ldc, stride_a, stride_b, stride_c, batch_size); cublasGemmAlgo_t cublas_gemm_algo = CUBLAS_GEMM_DEFAULT; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { if (!verify_support(queue, sycl::aspect::fp16)) { throw oneapi::math::unimplemented( "blas", "sycl::half", "half is not supported by the device or the sycl compiler"); @@ -632,16 +632,16 @@ inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose tra for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); cublasStatus_t err; #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - CUBLAS_ERROR_FUNC_T( - "cublasGemmStridedBatchedEx", cublasGemmStridedBatchedEx, err, handle, - get_cublas_operation(transa), get_cublas_operation(transb), m, n, k, &alpha, a, - get_cublas_datatype(), lda, stride_a, b, get_cublas_datatype(), - ldb, stride_b, &beta, c, get_cublas_datatype(), ldc, stride_c, batch_size, - get_cublas_datatype(), cublas_gemm_algo); + CUBLAS_ERROR_FUNC_T("cublasGemmStridedBatchedEx", cublasGemmStridedBatchedEx, err, + handle, get_cublas_operation(transa), get_cublas_operation(transb), + m, n, k, &alpha, a, get_cublas_datatype(), lda, stride_a, + b, get_cublas_datatype(), ldb, stride_b, &beta, c, + get_cublas_datatype(), ldc, stride_c, batch_size, + get_cublas_datatype(), cublas_gemm_algo); #else CUBLAS_ERROR_FUNC_T_SYNC( "cublasGemmStridedBatchedEx", cublasGemmStridedBatchedEx, err, handle, @@ -656,11 +656,11 @@ inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose tra } #define GEMM_STRIDED_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, int64_t k, TYPE_S alpha, const TYPE_A *a, int64_t lda, \ - int64_t stride_a, const TYPE_B *b, int64_t ldb, int64_t stride_b, \ - TYPE_S beta, TYPE_C *c, int64_t ldc, int64_t stride_c, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, int64_t k, TYPE_S alpha, const TYPE_A* a, int64_t lda, \ + int64_t stride_a, const TYPE_B* b, int64_t ldb, int64_t stride_b, \ + TYPE_S beta, TYPE_C* c, int64_t ldc, int64_t stride_c, \ + int64_t batch_size, const std::vector& dependencies) { \ return gemm_batch_strided_usm_impl(queue, transa, transb, m, n, k, alpha, a, lda, \ stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, \ batch_size, dependencies); \ @@ -678,11 +678,11 @@ GEMM_STRIDED_BATCH_LAUNCHER_USM(std::complex, std::complex, std: #undef GEMM_STRIDED_BATCH_LAUNCHER_USM #define GEMM_STRIDED_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, int64_t k, TYPE_S alpha, const TYPE_A *a, int64_t lda, \ - int64_t stride_a, const TYPE_B *b, int64_t ldb, int64_t stride_b, \ - TYPE_S beta, TYPE_C *c, int64_t ldc, int64_t stride_c, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, int64_t k, TYPE_S alpha, const TYPE_A* a, int64_t lda, \ + int64_t stride_a, const TYPE_B* b, int64_t ldb, int64_t stride_b, \ + TYPE_S beta, TYPE_C* c, int64_t ldc, int64_t stride_c, \ + int64_t batch_size, const std::vector& dependencies) { \ throw unimplemented("blas", "gemm_batch", \ std::string("for dtype unimplemented dtype combination <") + \ dtype_string() + "," + dtype_string() + "," + \ @@ -695,11 +695,11 @@ GEMM_STRIDED_BATCH_LAUNCHER_USM(std::int8_t, std::int8_t, std::int32_t, float) #undef GEMM_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, transpose *transb, - int64_t *m, int64_t *n, int64_t *k, Ts *alpha, const Ta **a, - int64_t *lda, const Tb **b, int64_t *ldb, Ts *beta, Tc **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event gemm_batch_usm_impl(sycl::queue& queue, transpose* transa, transpose* transb, + int64_t* m, int64_t* n, int64_t* k, Ts* alpha, const Ta** a, + int64_t* lda, const Tb** b, int64_t* ldb, Ts* beta, Tc** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { using cuTypeA = typename CudaEquivalentType::Type; using cuTypeB = typename CudaEquivalentType::Type; using cuTypeC = typename CudaEquivalentType::Type; @@ -709,7 +709,7 @@ inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, tr } cublasGemmAlgo_t cublas_gemm_algo = CUBLAS_GEMM_DEFAULT; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { if (!verify_support(queue, sycl::aspect::fp16)) { throw oneapi::math::unimplemented( "blas", "sycl::half", "half is not supported by the device or the sycl compiler"); @@ -718,7 +718,7 @@ inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, tr for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; cublasStatus_t err; @@ -727,10 +727,10 @@ inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, tr CUBLAS_ERROR_FUNC_T( "cublasGemmBatchedEx", cublasGemmBatchedEx, err, handle, get_cublas_operation(transa[i]), get_cublas_operation(transb[i]), (int)m[i], - (int)n[i], (int)k[i], &alpha[i], (const void *const *)(a + offset), - get_cublas_datatype(), (int)lda[i], (const void *const *)(b + offset), + (int)n[i], (int)k[i], &alpha[i], (const void* const*)(a + offset), + get_cublas_datatype(), (int)lda[i], (const void* const*)(b + offset), get_cublas_datatype(), (int)ldb[i], &beta[i], - (void *const *)(c + offset), get_cublas_datatype(), (int)ldc[i], + (void* const*)(c + offset), get_cublas_datatype(), (int)ldc[i], (int)group_size[i], get_cublas_datatype(), cublas_gemm_algo); #else CUBLAS_ERROR_FUNC_T_SYNC( @@ -750,11 +750,11 @@ inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, tr } #define GEMM_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, \ - int64_t *n, int64_t *k, TYPE_S *alpha, const TYPE_A **a, int64_t *lda, \ - const TYPE_B **b, int64_t *ldb, TYPE_S *beta, TYPE_C **c, int64_t *ldc, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, \ + int64_t* n, int64_t* k, TYPE_S* alpha, const TYPE_A** a, int64_t* lda, \ + const TYPE_B** b, int64_t* ldb, TYPE_S* beta, TYPE_C** c, int64_t* ldc, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return gemm_batch_usm_impl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, \ ldc, group_count, group_size, dependencies); \ } @@ -771,11 +771,11 @@ GEMM_BATCH_LAUNCHER_USM(std::complex, std::complex, std::complex #undef GEMM_BATCH_LAUNCHER_USM #define GEMM_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, \ - int64_t *n, int64_t *k, TYPE_S *alpha, const TYPE_A **a, int64_t *lda, \ - const TYPE_B **b, int64_t *ldb, TYPE_S *beta, TYPE_C **c, int64_t *ldc, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, \ + int64_t* n, int64_t* k, TYPE_S* alpha, const TYPE_A** a, int64_t* lda, \ + const TYPE_B** b, int64_t* ldb, TYPE_S* beta, TYPE_C** c, int64_t* ldc, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ throw unimplemented("blas", "gemm_batch", \ std::string("for dtype unimplemented dtype combination <") + \ dtype_string() + "," + dtype_string() + "," + \ @@ -787,63 +787,63 @@ GEMM_BATCH_LAUNCHER_USM(std::int8_t, std::int8_t, std::int32_t, float) #undef GEMM_BATCH_LAUNCHER_USM -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, - int64_t lda, int64_t stride_a, float *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, const float* a, + int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", "for column_major layout"); } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, - int64_t lda, int64_t stride_a, double *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, const double* a, + int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", "for column_major layout"); } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", "for column_major layout"); } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", "for column_major layout"); } template -inline sycl::event trsm_batch(const char *func_name, Func func, sycl::queue &queue, - side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, T *alpha, const T **a, - int64_t *lda, T **b, int64_t *ldb, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +inline sycl::event trsm_batch(const char* func_name, Func func, sycl::queue& queue, + side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, T* alpha, const T** a, + int64_t* lda, T** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; for (int64_t i = 0; i < group_count; i++) { overflow_check(m[i], n[i], lda[i], ldb[i], group_size[i]); } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; cublasStatus_t err; for (int64_t i = 0; i < group_count; i++) { - auto **a_ = reinterpret_cast(a); - auto **b_ = reinterpret_cast(b); + auto** a_ = reinterpret_cast(a); + auto** b_ = reinterpret_cast(b); cublas_native_named_func( func_name, func, err, handle, get_cublas_side_mode(left_right[i]), get_cublas_fill_mode(upper_lower[i]), get_cublas_operation(trans[i]), get_cublas_diag_type(unit_diag[i]), (int)m[i], (int)n[i], - (cuDataType *)&alpha[i], a_ + offset, (int)lda[i], b_ + offset, (int)ldb[i], + (cuDataType*)&alpha[i], a_ + offset, (int)lda[i], b_ + offset, (int)ldb[i], (int)group_size[i]); offset += group_size[i]; @@ -854,11 +854,11 @@ inline sycl::event trsm_batch(const char *func_name, Func func, sycl::queue &que } #define TRSM_BATCH_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, \ - transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, TYPE *alpha, \ - const TYPE **a, int64_t *lda, TYPE **b, int64_t *ldb, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, \ + transpose* trans, diag* unit_diag, int64_t* m, int64_t* n, TYPE* alpha, \ + const TYPE** a, int64_t* lda, TYPE** b, int64_t* ldb, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return trsm_batch(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, trans, \ unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, \ dependencies); \ @@ -871,208 +871,208 @@ TRSM_BATCH_LAUNCHER_USM(std::complex, cublasZtrsmBatched) #undef TRSM_BATCH_LAUNCHER_USM -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, float *alpha, const float **a, int64_t *lda, float *beta, - float **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, float* alpha, const float** a, int64_t* lda, float* beta, + float** c, int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, double *alpha, const double **a, int64_t *lda, double *beta, - double **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, double* alpha, const double** a, int64_t* lda, double* beta, + double** c, int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, const float *a, int64_t lda, int64_t stride_a, float beta, - float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, const float* a, int64_t lda, int64_t stride_a, float beta, + float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, const double *a, int64_t lda, int64_t stride_a, double beta, - double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, const double* a, int64_t lda, int64_t stride_a, double beta, + double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex beta, std::complex *c, +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex beta, std::complex *c, +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for column_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, - float beta, const float *b, int64_t ldb, int64_t stride_b, float *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, float alpha, const float* a, int64_t lda, int64_t stride_a, + float beta, const float* b, int64_t ldb, int64_t stride_b, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "omatadd_batch", "for column_major layout"); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, double alpha, const double *a, int64_t lda, int64_t stride_a, - double beta, const double *b, int64_t ldb, int64_t stride_b, double *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, double alpha, const double* a, int64_t lda, int64_t stride_a, + double beta, const double* b, int64_t ldb, int64_t stride_b, double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "omatadd_batch", "for column_major layout"); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "omatadd_batch", "for column_major layout"); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "omatadd_batch", "for column_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - float *alpha, const float **a, int64_t *lda, float **b, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + float* alpha, const float** a, int64_t* lda, float** b, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - double *alpha, const double **a, int64_t *lda, double **b, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + double* alpha, const double** a, int64_t* lda, double** b, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - std::complex **b, int64_t *ldb, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + std::complex** b, int64_t* ldb, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex **b, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex** b, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - float *alpha, float **ab, int64_t *lda, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + float* alpha, float** ab, int64_t* lda, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - double *alpha, double **ab, int64_t *lda, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + double* alpha, double** ab, int64_t* lda, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, std::complex **ab, int64_t *lda, - int64_t *ldb, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, std::complex** ab, int64_t* lda, + int64_t* ldb, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, std::complex **ab, int64_t *lda, - int64_t *ldb, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, std::complex** ab, int64_t* lda, + int64_t* ldb, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } @@ -1081,122 +1081,122 @@ namespace row_major { // Buffer APIs -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, - int64_t incx, int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, + int64_t incx, int64_t stridex, sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, - int64_t incx, int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, + int64_t incx, int64_t stridex, sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -void axpy_batch(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -void axpy_batch(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, +void axpy_batch(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, - int64_t incx, int64_t stride_x, float beta, sycl::buffer &y, int64_t incy, +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, sycl::buffer& x, + int64_t incx, int64_t stride_x, float beta, sycl::buffer& y, int64_t incy, int64_t stride_y, int64_t batch_size) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &x, int64_t incx, int64_t stride_x, double beta, - sycl::buffer &y, int64_t incy, int64_t stride_y, int64_t batch_size) { +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& x, int64_t incx, int64_t stride_x, double beta, + sycl::buffer& y, int64_t incy, int64_t stride_y, int64_t batch_size) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &x, int64_t incx, - int64_t stride_x, std::complex beta, sycl::buffer, 1> &y, +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& x, int64_t incx, + int64_t stride_x, std::complex beta, sycl::buffer, 1>& y, int64_t incy, int64_t stride_y, int64_t batch_size) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &x, int64_t incx, +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& x, int64_t incx, int64_t stride_x, std::complex beta, - sycl::buffer, 1> &y, int64_t incy, int64_t stride_y, + sycl::buffer, 1>& y, int64_t incy, int64_t stride_y, int64_t batch_size) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, - int64_t incx, int64_t stride_x, sycl::buffer &c, int64_t ldc, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stride_a, sycl::buffer& x, + int64_t incx, int64_t stride_x, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &x, int64_t incx, int64_t stride_x, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& x, int64_t incx, int64_t stride_x, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& x, int64_t incx, int64_t stride_x, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& x, int64_t incx, int64_t stride_x, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } #define GEMM_STRIDED_BATCH_LAUNCHER(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_S alpha, sycl::buffer &a, int64_t lda, \ - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, \ - TYPE_S beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, \ + void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_S alpha, sycl::buffer& a, int64_t lda, \ + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, \ + TYPE_S beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, \ int64_t batch_size) { \ throw unimplemented("blas", "gemm_batch", "for row_major layout"); \ } @@ -1214,377 +1214,377 @@ GEMM_STRIDED_BATCH_LAUNCHER(std::complex, std::complex, std::com #undef GEMM_STRIDED_BATCH_LAUNCHER -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "trsm_batch", "for row_major layout"); } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "trsm_batch", "for row_major layout"); } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "trsm_batch", "for row_major layout"); } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "trsm_batch", "for row_major layout"); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, float beta, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, float beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, std::complex beta, sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - float beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + float beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "omatadd_batch", "for row_major layout"); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "omatadd_batch", "for row_major layout"); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "omatadd_batch", "for row_major layout"); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "omatadd_batch", "for row_major layout"); } // USM APIs -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const float **x, int64_t *incx, float **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const float** x, int64_t* incx, float** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const double **x, int64_t *incx, double **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const double** x, int64_t* incx, double** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, int64_t *incx, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const std::complex** x, int64_t* incx, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, - int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const std::complex** x, + int64_t* incx, std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const float *x, int64_t incx, - std::int64_t stridex, float *y, int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t n, const float* x, int64_t incx, + std::int64_t stridex, float* y, int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const double *x, int64_t incx, - std::int64_t stridex, double *y, int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t n, const double* x, int64_t incx, + std::int64_t stridex, double* y, int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - std::int64_t stridex, std::complex *y, int64_t incy, +sycl::event copy_batch(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::int64_t stridex, std::complex* y, int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - std::int64_t stridex, std::complex *y, int64_t incy, +sycl::event copy_batch(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::int64_t stridex, std::complex* y, int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", "for row_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, float *alpha, const float **x, int64_t *incx, - float **y, int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, float* alpha, const float** x, int64_t* incx, + float** y, int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, double *alpha, const double **x, - int64_t *incx, double **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, double* alpha, const double** x, + int64_t* incx, double** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, std::complex* alpha, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, std::complex* alpha, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, - int64_t stridex, float *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, + int64_t stridex, float* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - int64_t stridex, double *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + int64_t stridex, double* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", "for row_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, const float *x, int64_t incx, - int64_t stride_x, float beta, float *y, int64_t incy, int64_t stride_y, - int64_t batch_size, const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stride_a, const float* x, int64_t incx, + int64_t stride_x, float beta, float* y, int64_t incy, int64_t stride_y, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, const double *x, - int64_t incx, int64_t stride_x, double beta, double *y, int64_t incy, +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stride_a, const double* x, + int64_t incx, int64_t stride_x, double beta, double* y, int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, const std::complex *x, int64_t incx, - int64_t stride_x, std::complex beta, std::complex *y, +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, const std::complex* x, int64_t incx, + int64_t stride_x, std::complex beta, std::complex* y, int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, const std::complex *x, int64_t incx, - int64_t stride_x, std::complex beta, std::complex *y, +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, const std::complex* x, int64_t incx, + int64_t stride_x, std::complex beta, std::complex* y, int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, float *alpha, - const float **a, int64_t *lda, const float **x, int64_t *incx, float *beta, - float **y, int64_t *incy, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, float* alpha, + const float** a, int64_t* lda, const float** x, int64_t* incx, float* beta, + float** y, int64_t* incy, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, double *alpha, - const double **a, int64_t *lda, const double **x, int64_t *incx, - double *beta, double **y, int64_t *incy, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, double* alpha, + const double** a, int64_t* lda, const double** x, int64_t* incx, + double* beta, double** y, int64_t* incy, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - const std::complex **x, int64_t *incx, std::complex *beta, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + const std::complex** x, int64_t* incx, std::complex* beta, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - const std::complex **x, int64_t *incx, std::complex *beta, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + const std::complex** x, int64_t* incx, std::complex* beta, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", "for row_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, const float *a, - int64_t lda, int64_t stride_a, const float *x, int64_t incx, - int64_t stride_x, float *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, const float* a, + int64_t lda, int64_t stride_a, const float* x, int64_t incx, + int64_t stride_x, float* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, const double *a, - int64_t lda, int64_t stride_a, const double *x, int64_t incx, - int64_t stride_x, double *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, const double* a, + int64_t lda, int64_t stride_a, const double* x, int64_t incx, + int64_t stride_x, double* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const std::complex *a, int64_t lda, int64_t stride_a, - const std::complex *x, int64_t incx, int64_t stride_x, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + const std::complex* a, int64_t lda, int64_t stride_a, + const std::complex* x, int64_t incx, int64_t stride_x, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const std::complex *a, int64_t lda, int64_t stride_a, - const std::complex *x, int64_t incx, int64_t stride_x, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + const std::complex* a, int64_t lda, int64_t stride_a, + const std::complex* x, int64_t incx, int64_t stride_x, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const float **a, int64_t *lda, const float **x, int64_t *incx, float **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const float** a, int64_t* lda, const float** x, int64_t* incx, float** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const double **a, int64_t *lda, const double **x, int64_t *incx, double **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const double** a, int64_t* lda, const double** x, int64_t* incx, double** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const std::complex **a, int64_t *lda, const std::complex **x, - int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const std::complex** a, int64_t* lda, const std::complex** x, + int64_t* incx, std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const std::complex **a, int64_t *lda, const std::complex **x, - int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const std::complex** a, int64_t* lda, const std::complex** x, + int64_t* incx, std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", "for row_major layout"); } #define GEMM_STRIDED_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, int64_t k, TYPE_S alpha, const TYPE_A *a, int64_t lda, \ - int64_t stride_a, const TYPE_B *b, int64_t ldb, int64_t stride_b, \ - TYPE_S beta, TYPE_C *c, int64_t ldc, int64_t stride_c, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, int64_t k, TYPE_S alpha, const TYPE_A* a, int64_t lda, \ + int64_t stride_a, const TYPE_B* b, int64_t ldb, int64_t stride_b, \ + TYPE_S beta, TYPE_C* c, int64_t ldc, int64_t stride_c, \ + int64_t batch_size, const std::vector& dependencies) { \ throw unimplemented("blas", "gemm_batch", "for row_major layout"); \ } @@ -1602,11 +1602,11 @@ GEMM_STRIDED_BATCH_LAUNCHER_USM(std::complex, std::complex, std: #undef GEMM_STRIDED_BATCH_LAUNCHER_USM #define GEMM_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, \ - int64_t *n, int64_t *k, TYPE_S *alpha, const TYPE_A **a, int64_t *lda, \ - const TYPE_B **b, int64_t *ldb, TYPE_S *beta, TYPE_C **c, int64_t *ldc, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, \ + int64_t* n, int64_t* k, TYPE_S* alpha, const TYPE_A** a, int64_t* lda, \ + const TYPE_B** b, int64_t* ldb, TYPE_S* beta, TYPE_C** c, int64_t* ldc, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ throw unimplemented("blas", "gemm_batch", "for row_major layout"); \ } @@ -1623,51 +1623,51 @@ GEMM_BATCH_LAUNCHER_USM(std::complex, std::complex, std::complex #undef GEMM_BATCH_LAUNCHER_USM -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, - int64_t lda, int64_t stride_a, float *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, const float* a, + int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", "for row_major layout"); } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, - int64_t lda, int64_t stride_a, double *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, const double* a, + int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", "for row_major layout"); } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", "for row_major layout"); } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", "for row_major layout"); } template -inline sycl::event trsm_batch(const char *func_name, Func func, sycl::queue &queue, - side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, T *alpha, const T **a, - int64_t *lda, T **b, int64_t *ldb, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +inline sycl::event trsm_batch(const char* func_name, Func func, sycl::queue& queue, + side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, T* alpha, const T** a, + int64_t* lda, T** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", "for row_major layout"); } #define TRSM_BATCH_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, \ - transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, TYPE *alpha, \ - const TYPE **a, int64_t *lda, TYPE **b, int64_t *ldb, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, \ + transpose* trans, diag* unit_diag, int64_t* m, int64_t* n, TYPE* alpha, \ + const TYPE** a, int64_t* lda, TYPE** b, int64_t* ldb, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return trsm_batch(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, trans, \ unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, \ dependencies); \ @@ -1680,208 +1680,208 @@ TRSM_BATCH_LAUNCHER_USM(std::complex, cublasZtrsmBatched) #undef TRSM_BATCH_LAUNCHER_USM -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, float *alpha, const float **a, int64_t *lda, float *beta, - float **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, float* alpha, const float** a, int64_t* lda, float* beta, + float** c, int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, double *alpha, const double **a, int64_t *lda, double *beta, - double **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, double* alpha, const double** a, int64_t* lda, double* beta, + double** c, int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, const float *a, int64_t lda, int64_t stride_a, float beta, - float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, const float* a, int64_t lda, int64_t stride_a, float beta, + float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, const double *a, int64_t lda, int64_t stride_a, double beta, - double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, const double* a, int64_t lda, int64_t stride_a, double beta, + double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex beta, std::complex *c, +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex beta, std::complex *c, +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", "for row_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, - float beta, const float *b, int64_t ldb, int64_t stride_b, float *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, float alpha, const float* a, int64_t lda, int64_t stride_a, + float beta, const float* b, int64_t ldb, int64_t stride_b, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "omatadd_batch", "for row_major layout"); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, double alpha, const double *a, int64_t lda, int64_t stride_a, - double beta, const double *b, int64_t ldb, int64_t stride_b, double *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, double alpha, const double* a, int64_t lda, int64_t stride_a, + double beta, const double* b, int64_t ldb, int64_t stride_b, double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "omatadd_batch", "for row_major layout"); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "omatadd_batch", "for row_major layout"); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "omatadd_batch", "for row_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - float *alpha, const float **a, int64_t *lda, float **b, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + float* alpha, const float** a, int64_t* lda, float** b, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - double *alpha, const double **a, int64_t *lda, double **b, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + double* alpha, const double** a, int64_t* lda, double** b, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - std::complex **b, int64_t *ldb, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + std::complex** b, int64_t* ldb, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex **b, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex** b, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - float *alpha, float **ab, int64_t *lda, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + float* alpha, float** ab, int64_t* lda, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - double *alpha, double **ab, int64_t *lda, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + double* alpha, double** ab, int64_t* lda, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, std::complex **ab, int64_t *lda, - int64_t *ldb, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, std::complex** ab, int64_t* lda, + int64_t* ldb, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, std::complex **ab, int64_t *lda, - int64_t *ldb, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, std::complex** ab, int64_t* lda, + int64_t* ldb, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } diff --git a/src/blas/backends/cublas/cublas_extensions.cpp b/src/blas/backends/cublas/cublas_extensions.cpp index db6c0aa05..ab1157135 100644 --- a/src/blas/backends/cublas/cublas_extensions.cpp +++ b/src/blas/backends/cublas/cublas_extensions.cpp @@ -29,88 +29,88 @@ namespace column_major { // Buffer APIs -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, double beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for column_major layout"); } template -void omatcopy(const char *func_name, Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, +void omatcopy(const char* func_name, Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); const int64_t logical_m = (trans == oneapi::math::transpose::nontrans ? m : n); const int64_t logical_n = (trans == oneapi::math::transpose::nontrans ? n : m); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); cublasStatus_t err; CUBLAS_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_operation(trans), get_cublas_operation(trans), logical_m, logical_n, - (cuDataType *)&alpha, a_, lda, nullptr, nullptr, lda, b_, ldb); + (cuDataType*)&alpha, a_, lda, nullptr, nullptr, lda, b_, ldb); }); }); } #define OMATCOPY_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { \ omatcopy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, b, ldb); \ } @@ -122,16 +122,16 @@ OMATCOPY_LAUNCHER(std::complex, cublasZgeam) #undef OMATCOPY_LAUNCHER template -void omatcopy2(const char *func_name, Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb) { +void omatcopy2(const char* func_name, Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb) { throw unimplemented("blas", "omatcopy2", ""); } #define OMATCOPY2_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, int64_t stridea, \ - sycl::buffer &b, int64_t ldb, int64_t strideb) { \ + void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, int64_t stridea, \ + sycl::buffer& b, int64_t ldb, int64_t strideb) { \ omatcopy2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, stridea, lda, b, \ ldb, strideb); \ } @@ -142,53 +142,53 @@ OMATCOPY2_LAUNCHER(std::complex, "unimplemented") OMATCOPY2_LAUNCHER(std::complex, "unimplemented") #undef OMATCOPY2_LAUNCHER -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } template -void omatadd(const char *func_name, Func func, sycl::queue &queue, transpose transa, - transpose transb, int64_t m, int64_t n, T alpha, sycl::buffer &a, int64_t lda, - T beta, sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { +void omatadd(const char* func_name, Func func, sycl::queue& queue, transpose transa, + transpose transb, int64_t m, int64_t n, T alpha, sycl::buffer& a, int64_t lda, + T beta, sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); cublasStatus_t err; CUBLAS_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_operation(transa), - get_cublas_operation(transb), m, n, (cuDataType *)&alpha, a_, - lda, (cuDataType *)&beta, b_, ldb, c_, ldc); + get_cublas_operation(transb), m, n, (cuDataType*)&alpha, a_, + lda, (cuDataType*)&beta, b_, ldb, c_, ldc); }); }); } #define OMATADD_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - TYPE alpha, sycl::buffer &a, int64_t lda, TYPE beta, \ - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { \ + void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + TYPE alpha, sycl::buffer& a, int64_t lda, TYPE beta, \ + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { \ omatadd(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, lda, beta, \ b, ldb, c, ldc); \ } @@ -202,95 +202,95 @@ OMATADD_LAUNCHER(std::complex, cublasZgeam) // USM APIs -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const int8_t *a, int64_t lda, - int8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const int8_t *a, int64_t lda, - int8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const uint8_t *a, int64_t lda, - uint8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const uint8_t *a, int64_t lda, - uint8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, float alpha, const float *a, int64_t lda, const float *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, float alpha, const float* a, int64_t lda, const float* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, double alpha, const double *a, int64_t lda, const double *b, - int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for column_major layout"); } template -sycl::event omatcopy(const char *func_name, Func func, sycl::queue &queue, transpose trans, - int64_t m, int64_t n, T alpha, const T *a, int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { +sycl::event omatcopy(const char* func_name, Func func, sycl::queue& queue, transpose trans, + int64_t m, int64_t n, T alpha, const T* a, int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); const int64_t logical_m = (trans == oneapi::math::transpose::nontrans ? m : n); const int64_t logical_n = (trans == oneapi::math::transpose::nontrans ? n : m); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); cublasStatus_t err; CUBLAS_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_operation(trans), get_cublas_operation(trans), logical_m, logical_n, - (cuDataType *)&alpha, a_, lda, nullptr, nullptr, lda, b_, ldb); + (cuDataType*)&alpha, a_, lda, nullptr, nullptr, lda, b_, ldb); }); }); return done; } #define OMATCOPY_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, TYPE *b, int64_t ldb, \ - const std::vector &dependencies) { \ + sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, TYPE* b, int64_t ldb, \ + const std::vector& dependencies) { \ return omatcopy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, b, \ ldb, dependencies); \ } @@ -303,16 +303,16 @@ OMATCOPY_LAUNCHER_USM(std::complex, cublasZgeam) #undef OMATCOPY_LAUNCHER_USM template -sycl::event omatcopy2(const char *func_name, Func func, sycl::queue &queue, transpose trans, - int64_t m, int64_t n, T alpha, const T *a, int64_t lda, int64_t stridea, T *b, - int64_t ldb, int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(const char* func_name, Func func, sycl::queue& queue, transpose trans, + int64_t m, int64_t n, T alpha, const T* a, int64_t lda, int64_t stridea, T* b, + int64_t ldb, int64_t strideb, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy2", ""); } #define OMATCOPY2_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, int64_t stridea, TYPE *b, int64_t ldb, \ - int64_t strideb, const std::vector &dependencies) { \ + sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, int64_t stridea, TYPE* b, int64_t ldb, \ + int64_t strideb, const std::vector& dependencies) { \ return omatcopy2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, stridea, \ lda, b, ldb, strideb, dependencies); \ } @@ -323,58 +323,58 @@ OMATCOPY2_LAUNCHER_USM(std::complex, "unimplemented") OMATCOPY2_LAUNCHER_USM(std::complex, "unimplemented") #undef OMATCOPY2_LAUNCHER_USM -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } template -inline sycl::event omatadd(const char *func_name, Func func, sycl::queue &queue, transpose transa, - transpose transb, int64_t m, int64_t n, T alpha, const T *a, int64_t lda, - T beta, const T *b, int64_t ldb, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event omatadd(const char* func_name, Func func, sycl::queue& queue, transpose transa, + transpose transb, int64_t m, int64_t n, T alpha, const T* a, int64_t lda, + T beta, const T* b, int64_t ldb, T* c, int64_t ldc, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); cublasStatus_t err; CUBLAS_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_operation(transa), - get_cublas_operation(transb), m, n, (cuDataType *)&alpha, a_, - lda, (cuDataType *)&beta, b_, ldb, c_, ldc); + get_cublas_operation(transb), m, n, (cuDataType*)&alpha, a_, + lda, (cuDataType*)&beta, b_, ldb, c_, ldc); }); }); return done; } #define OMATADD_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, TYPE alpha, const TYPE *a, int64_t lda, TYPE beta, \ - const TYPE *b, int64_t ldb, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, TYPE alpha, const TYPE* a, int64_t lda, TYPE beta, \ + const TYPE* b, int64_t ldb, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return omatadd(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, \ lda, beta, b, ldb, c, ldc, dependencies); \ } @@ -392,88 +392,88 @@ namespace row_major { // Buffer APIs -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, double beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for row_major layout"); } template -void omatcopy(const char *func_name, Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, +void omatcopy(const char* func_name, Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); const int64_t logical_m = (trans == oneapi::math::transpose::nontrans ? n : m); const int64_t logical_n = (trans == oneapi::math::transpose::nontrans ? m : n); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); cublasStatus_t err; CUBLAS_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_operation(trans), get_cublas_operation(trans), logical_m, logical_n, - (cuDataType *)&alpha, a_, lda, nullptr, nullptr, lda, b_, ldb); + (cuDataType*)&alpha, a_, lda, nullptr, nullptr, lda, b_, ldb); }); }); } #define OMATCOPY_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { \ omatcopy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, b, ldb); \ } @@ -485,16 +485,16 @@ OMATCOPY_LAUNCHER(std::complex, cublasZgeam) #undef OMATCOPY_LAUNCHER template -void omatcopy2(const char *func_name, Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb) { +void omatcopy2(const char* func_name, Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb) { throw unimplemented("blas", "omatcopy2", ""); } #define OMATCOPY2_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, int64_t stridea, \ - sycl::buffer &b, int64_t ldb, int64_t strideb) { \ + void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, int64_t stridea, \ + sycl::buffer& b, int64_t ldb, int64_t strideb) { \ omatcopy2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, stridea, lda, b, \ ldb, strideb); \ } @@ -505,53 +505,53 @@ OMATCOPY2_LAUNCHER(std::complex, "unimplemented") OMATCOPY2_LAUNCHER(std::complex, "unimplemented") #undef OMATCOPY2_LAUNCHER -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } template -void omatadd(const char *func_name, Func func, sycl::queue &queue, transpose transa, - transpose transb, int64_t m, int64_t n, T alpha, sycl::buffer &a, int64_t lda, - T beta, sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { +void omatadd(const char* func_name, Func func, sycl::queue& queue, transpose transa, + transpose transb, int64_t m, int64_t n, T alpha, sycl::buffer& a, int64_t lda, + T beta, sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); cublasStatus_t err; CUBLAS_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_operation(transa), - get_cublas_operation(transb), n, m, (cuDataType *)&alpha, a_, - lda, (cuDataType *)&beta, b_, ldb, c_, ldc); + get_cublas_operation(transb), n, m, (cuDataType*)&alpha, a_, + lda, (cuDataType*)&beta, b_, ldb, c_, ldc); }); }); } #define OMATADD_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - TYPE alpha, sycl::buffer &a, int64_t lda, TYPE beta, \ - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { \ + void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + TYPE alpha, sycl::buffer& a, int64_t lda, TYPE beta, \ + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { \ omatadd(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, lda, beta, \ b, ldb, c, ldc); \ } @@ -565,95 +565,95 @@ OMATADD_LAUNCHER(std::complex, cublasZgeam) // USM APIs -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const int8_t *a, int64_t lda, - int8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const int8_t *a, int64_t lda, - int8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const uint8_t *a, int64_t lda, - uint8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const uint8_t *a, int64_t lda, - uint8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, float alpha, const float *a, int64_t lda, const float *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, float alpha, const float* a, int64_t lda, const float* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, double alpha, const double *a, int64_t lda, const double *b, - int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for row_major layout"); } template -sycl::event omatcopy(const char *func_name, Func func, sycl::queue &queue, transpose trans, - int64_t m, int64_t n, T alpha, const T *a, int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { +sycl::event omatcopy(const char* func_name, Func func, sycl::queue& queue, transpose trans, + int64_t m, int64_t n, T alpha, const T* a, int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); const int64_t logical_m = (trans == oneapi::math::transpose::nontrans ? n : m); const int64_t logical_n = (trans == oneapi::math::transpose::nontrans ? m : n); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); cublasStatus_t err; CUBLAS_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_operation(trans), get_cublas_operation(trans), logical_m, logical_n, - (cuDataType *)&alpha, a_, lda, nullptr, nullptr, ldb, b_, ldb); + (cuDataType*)&alpha, a_, lda, nullptr, nullptr, ldb, b_, ldb); }); }); return done; } #define OMATCOPY_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, TYPE *b, int64_t ldb, \ - const std::vector &dependencies) { \ + sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, TYPE* b, int64_t ldb, \ + const std::vector& dependencies) { \ return omatcopy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, b, \ ldb, dependencies); \ } @@ -666,16 +666,16 @@ OMATCOPY_LAUNCHER_USM(std::complex, cublasZgeam) #undef OMATCOPY_LAUNCHER_USM template -sycl::event omatcopy2(const char *func_name, Func func, sycl::queue &queue, transpose trans, - int64_t m, int64_t n, T alpha, const T *a, int64_t lda, int64_t stridea, T *b, - int64_t ldb, int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(const char* func_name, Func func, sycl::queue& queue, transpose trans, + int64_t m, int64_t n, T alpha, const T* a, int64_t lda, int64_t stridea, T* b, + int64_t ldb, int64_t strideb, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy2", ""); } #define OMATCOPY2_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, int64_t stridea, TYPE *b, int64_t ldb, \ - int64_t strideb, const std::vector &dependencies) { \ + sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, int64_t stridea, TYPE* b, int64_t ldb, \ + int64_t strideb, const std::vector& dependencies) { \ return omatcopy2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, stridea, \ lda, b, ldb, strideb, dependencies); \ } @@ -686,58 +686,58 @@ OMATCOPY2_LAUNCHER_USM(std::complex, "unimplemented") OMATCOPY2_LAUNCHER_USM(std::complex, "unimplemented") #undef OMATCOPY2_LAUNCHER_USM -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } template -inline sycl::event omatadd(const char *func_name, Func func, sycl::queue &queue, transpose transa, - transpose transb, int64_t m, int64_t n, T alpha, const T *a, int64_t lda, - T beta, const T *b, int64_t ldb, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event omatadd(const char* func_name, Func func, sycl::queue& queue, transpose transa, + transpose transb, int64_t m, int64_t n, T alpha, const T* a, int64_t lda, + T beta, const T* b, int64_t ldb, T* c, int64_t ldc, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); cublasStatus_t err; CUBLAS_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_operation(transa), - get_cublas_operation(transb), n, m, (cuDataType *)&alpha, a_, - lda, (cuDataType *)&beta, b_, ldb, c_, ldc); + get_cublas_operation(transb), n, m, (cuDataType*)&alpha, a_, + lda, (cuDataType*)&beta, b_, ldb, c_, ldc); }); }); return done; } #define OMATADD_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, TYPE alpha, const TYPE *a, int64_t lda, TYPE beta, \ - const TYPE *b, int64_t ldb, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, TYPE alpha, const TYPE* a, int64_t lda, TYPE beta, \ + const TYPE* b, int64_t ldb, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return omatadd(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, \ lda, beta, b, ldb, c, ldc, dependencies); \ } diff --git a/src/blas/backends/cublas/cublas_handle.hpp b/src/blas/backends/cublas/cublas_handle.hpp index 254c1c160..34a71b185 100644 --- a/src/blas/backends/cublas/cublas_handle.hpp +++ b/src/blas/backends/cublas/cublas_handle.hpp @@ -28,10 +28,10 @@ namespace cublas { template struct cublas_handle { - using handle_container_t = std::unordered_map *>; + using handle_container_t = std::unordered_map*>; handle_container_t cublas_handle_mapper_{}; ~cublas_handle() noexcept(false) { - for (auto &handle_pair : cublas_handle_mapper_) { + for (auto& handle_pair : cublas_handle_mapper_) { cublasStatus_t err; if (handle_pair.second != nullptr) { auto handle = handle_pair.second->exchange(nullptr); diff --git a/src/blas/backends/cublas/cublas_helper.hpp b/src/blas/backends/cublas/cublas_helper.hpp index 27b4a6bf9..a67d89126 100644 --- a/src/blas/backends/cublas/cublas_helper.hpp +++ b/src/blas/backends/cublas/cublas_helper.hpp @@ -81,7 +81,7 @@ void overflow_check(Index index, Next... indices) { class cublas_error : virtual public std::runtime_error { protected: - inline const char *cublas_error_map(cublasStatus_t error) { + inline const char* cublas_error_map(cublasStatus_t error) { switch (error) { case CUBLAS_STATUS_SUCCESS: return "CUBLAS_STATUS_SUCCESS"; @@ -133,7 +133,7 @@ class cublas_error : virtual public std::runtime_error { class cuda_error : virtual public std::runtime_error { protected: - inline const char *cuda_error_map(CUresult result) { + inline const char* cuda_error_map(CUresult result) { switch (result) { case CUDA_SUCCESS: return "CUDA_SUCCESS"; case CUDA_ERROR_NOT_PERMITTED: return "CUDA_ERROR_NOT_PERMITTED"; @@ -206,23 +206,22 @@ class cuda_error : virtual public std::runtime_error { cuStreamSynchronize(currentStreamId); template -inline void cublas_native_func(Func func, cublasStatus_t err, - cublasHandle_t handle, Types... args) { +inline void cublas_native_func(Func func, cublasStatus_t err, cublasHandle_t handle, + Types... args) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - CUBLAS_ERROR_FUNC(func, err, handle, args...) + CUBLAS_ERROR_FUNC(func, err, handle, args...) #else - CUBLAS_ERROR_FUNC_SYNC(func, err, handle, args...) + CUBLAS_ERROR_FUNC_SYNC(func, err, handle, args...) #endif }; template -inline void cublas_native_named_func(const char *func_name, Func func, - cublasStatus_t err, cublasHandle_t handle, - Types... args) { +inline void cublas_native_named_func(const char* func_name, Func func, cublasStatus_t err, + cublasHandle_t handle, Types... args) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - CUBLAS_ERROR_FUNC_T(func_name, func, err, handle, args...) + CUBLAS_ERROR_FUNC_T(func_name, func, err, handle, args...) #else - CUBLAS_ERROR_FUNC_T_SYNC(func_name, func, err, handle, args...) + CUBLAS_ERROR_FUNC_T_SYNC(func_name, func, err, handle, args...) #endif }; diff --git a/src/blas/backends/cublas/cublas_level1.cpp b/src/blas/backends/cublas/cublas_level1.cpp index 5b0e39978..830a1b1e0 100644 --- a/src/blas/backends/cublas/cublas_level1.cpp +++ b/src/blas/backends/cublas/cublas_level1.cpp @@ -32,16 +32,16 @@ namespace column_major { // Level 1 template -inline void asum(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &result) { +inline void asum(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& result) { using cuDataType1 = typename CudaEquivalentType::Type; using cuDataType2 = typename CudaEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -49,8 +49,8 @@ inline void asum(const char *func_name, Func func, sycl::queue &queue, int64_t n // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); - auto x_ = sc.get_mem(x_acc); - auto res_ = sc.get_mem(res_acc); + auto x_ = sc.get_mem(x_acc); + auto res_ = sc.get_mem(res_acc); cublasStatus_t err; // ASUM does not support negative index cublas_native_named_func(func_name, func, err, handle, n, x_, std::abs(incx), res_); @@ -63,8 +63,8 @@ inline void asum(const char *func_name, Func func, sycl::queue &queue, int64_t n } #define ASUM_LAUNCHER(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - void asum(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void asum(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ asum(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result); \ } ASUM_LAUNCHER(float, float, cublasSasum) @@ -74,26 +74,26 @@ ASUM_LAUNCHER(std::complex, double, cublasDzasum) #undef ASUM_LAUNCHER template -inline void scal(const char *func_name, Func func, sycl::queue &queue, int64_t n, T1 a, - sycl::buffer &x, int64_t incx) { +inline void scal(const char* func_name, Func func, sycl::queue& queue, int64_t n, T1 a, + sycl::buffer& x, int64_t incx) { using cuDataType1 = typename CudaEquivalentType::Type; using cuDataType2 = typename CudaEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = sc.get_mem(x_acc); + auto x_ = sc.get_mem(x_acc); cublasStatus_t err; // SCAL does not support negative incx - cublas_native_named_func(func_name, func, err, handle, n, (cuDataType1 *)&a, x_, + cublas_native_named_func(func_name, func, err, handle, n, (cuDataType1*)&a, x_, std::abs(incx)); }); }); } #define SCAL_LAUNCHER(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - void scal(sycl::queue &queue, int64_t n, TYPE1 a, sycl::buffer &x, int64_t incx) { \ + void scal(sycl::queue& queue, int64_t n, TYPE1 a, sycl::buffer& x, int64_t incx) { \ scal(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, a, x, incx); \ } SCAL_LAUNCHER(float, float, cublasSscal) @@ -105,27 +105,27 @@ SCAL_LAUNCHER(double, std::complex, cublasZdscal) #undef SCAL_LAUNCHER template -inline void axpy(const char *func_name, Func func, sycl::queue &queue, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { +inline void axpy(const char* func_name, Func func, sycl::queue& queue, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; - cublas_native_named_func(func_name, func, err, handle, n, (cuDataType *)&alpha, x_, - incx, y_, incy); + cublas_native_named_func(func_name, func, err, handle, n, (cuDataType*)&alpha, x_, incx, + y_, incy); }); }); } #define AXPY_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void axpy(sycl::queue &queue, int64_t n, TYPE alpha, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void axpy(sycl::queue& queue, int64_t n, TYPE alpha, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ axpy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, alpha, x, incx, y, incy); \ } @@ -135,39 +135,39 @@ AXPY_LAUNCHER(std::complex, cublasCaxpy) AXPY_LAUNCHER(std::complex, cublasZaxpy) #undef AXPY_LAUNCHER -void axpby(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - float beta, sycl::buffer &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + float beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "axpby", "for column_major layout"); } -void axpby(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx, - double beta, sycl::buffer &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, int64_t incx, + double beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "axpby", "for column_major layout"); } -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { throw unimplemented("blas", "axpby", "for column_major layout"); } -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { throw unimplemented("blas", "axpby", "for column_major layout"); } template -inline void rotg(const char *func_name, Func func, sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { +inline void rotg(const char* func_name, Func func, sycl::queue& queue, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { using cuDataType1 = typename CudaEquivalentType::Type; using cuDataType2 = typename CudaEquivalentType::Type; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto s_acc = s.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -175,10 +175,10 @@ inline void rotg(const char *func_name, Func func, sycl::queue &queue, sycl::buf // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); - auto s_ = sc.get_mem(s_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); + auto s_ = sc.get_mem(s_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, a_, b_, c_, s_); // Higher level BLAS functions expect CUBLAS_POINTER_MODE_HOST @@ -190,8 +190,8 @@ inline void rotg(const char *func_name, Func func, sycl::queue &queue, sycl::buf } #define ROTG_LAUNCHER(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, \ - sycl::buffer &c, sycl::buffer &s) { \ + void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, \ + sycl::buffer& c, sycl::buffer& s) { \ rotg(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, a, b, c, s); \ } @@ -202,16 +202,16 @@ ROTG_LAUNCHER(std::complex, double, cublasZrotg) #undef ROTG_LAUNCHER template -inline void rotm(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer ¶m) { +inline void rotm(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& param) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); auto param_acc = param.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -219,9 +219,9 @@ inline void rotm(const char *func_name, Func func, sycl::queue &queue, int64_t n // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); - auto param_ = sc.get_mem(param_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); + auto param_ = sc.get_mem(param_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, n, x_, incx, y_, incy, param_); // Higher level BLAS functions expect CUBLAS_POINTER_MODE_HOST @@ -233,8 +233,8 @@ inline void rotm(const char *func_name, Func func, sycl::queue &queue, int64_t n } #define ROTM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void rotm(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { \ + void rotm(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy, sycl::buffer& param) { \ rotm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, param); \ } @@ -243,17 +243,17 @@ ROTM_LAUNCHER(double, cublasDrotm) #undef ROTM_LAUNCHER template -inline void copy(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { +inline void copy(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, n, x_, incx, y_, incy); }); @@ -261,8 +261,8 @@ inline void copy(const char *func_name, Func func, sycl::queue &queue, int64_t n } #define COPY_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void copy(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void copy(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ copy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy); \ } @@ -273,16 +273,16 @@ COPY_LAUNCHER(std::complex, cublasZcopy) #undef COPY_LAUNCHER template -inline void dot(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &result) { +inline void dot(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& result) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -290,9 +290,9 @@ inline void dot(const char *func_name, Func func, sycl::queue &queue, int64_t n, // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); - auto res_ = sc.get_mem(res_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); + auto res_ = sc.get_mem(res_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, n, x_, incx, y_, incy, res_); // Higher level BLAS functions expect CUBLAS_POINTER_MODE_HOST @@ -304,8 +304,8 @@ inline void dot(const char *func_name, Func func, sycl::queue &queue, int64_t n, } #define DOT_LAUNCHER(EXT, TYPE, CUBLAS_ROUTINE) \ - void dot##EXT(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &y, const int64_t incy, sycl::buffer &result) { \ + void dot##EXT(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& y, const int64_t incy, sycl::buffer& result) { \ dot(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, result); \ } DOT_LAUNCHER(, float, cublasSdot) @@ -317,17 +317,17 @@ DOT_LAUNCHER(u, std::complex, cublasZdotu) #undef DOT_LAUNCHER template -inline void rot(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &y, int64_t incy, +inline void rot(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& y, int64_t incy, T2 c, T3 s) { using cuDataType1 = typename CudaEquivalentType::Type; using cuDataType2 = typename CudaEquivalentType::Type; using cuDataType3 = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -335,18 +335,18 @@ inline void rot(const char *func_name, Func func, sycl::queue &queue, int64_t n, // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. // cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, n, x_, incx, y_, incy, - (cuDataType2 *)&c, (cuDataType3 *)&s); + (cuDataType2*)&c, (cuDataType3*)&s); }); }); } #define ROT_LAUNCHER(TYPE1, TYPE2, TYPE3, CUBLAS_ROUTINE) \ - void rot(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &y, int64_t incy, TYPE2 c, TYPE3 s) { \ + void rot(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& y, int64_t incy, TYPE2 c, TYPE3 s) { \ rot(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, c, s); \ } @@ -356,15 +356,15 @@ ROT_LAUNCHER(std::complex, float, float, cublasCsrot) ROT_LAUNCHER(std::complex, double, double, cublasZdrot) #undef ROT_LAUNCHER -void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { +void sdsdot(sycl::queue& queue, int64_t n, float sb, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { overflow_check(n, incx, incy); // cuBLAS does not support sdot so we need to mimic sdot. - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.get_access(cgh); auto y_acc = y.get_access(cgh); auto res_acc = result.get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -372,9 +372,9 @@ void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); - auto res_ = sc.get_mem(res_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); + auto res_ = sc.get_mem(res_acc); cublasStatus_t err; cublas_native_func(cublasSdot, err, handle, n, x_, incx, y_, incy, res_); // Higher level BLAS functions expect CUBLAS_POINTER_MODE_HOST @@ -388,23 +388,23 @@ void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, result.get_host_access(sycl::read_write)[0] += sb; } -void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { +void dot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { throw unimplemented("blas", "dot", "for column_major layout"); } template -inline void rotmg(const char *func_name, Func func, sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, T y1, sycl::buffer ¶m) { +inline void rotmg(const char* func_name, Func func, sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, T y1, sycl::buffer& param) { using cuDataType = typename CudaEquivalentType::Type; sycl::buffer y1_buff(&y1, sycl::range<1>(1)); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto d1_acc = d1.template get_access(cgh); auto d2_acc = d2.template get_access(cgh); auto x1_acc = x1.template get_access(cgh); auto y1_acc = y1_buff.template get_access(cgh); auto param_acc = param.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -412,11 +412,11 @@ inline void rotmg(const char *func_name, Func func, sycl::queue &queue, sycl::bu // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); - auto d1_ = sc.get_mem(d1_acc); - auto d2_ = sc.get_mem(d2_acc); - auto x1_ = sc.get_mem(x1_acc); - auto y1_ = sc.get_mem(y1_acc); - auto param_ = sc.get_mem(param_acc); + auto d1_ = sc.get_mem(d1_acc); + auto d2_ = sc.get_mem(d2_acc); + auto x1_ = sc.get_mem(x1_acc); + auto y1_ = sc.get_mem(y1_acc); + auto param_ = sc.get_mem(param_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, d1_, d2_, x1_, y1_, param_); // Higher level BLAS functions expect CUBLAS_POINTER_MODE_HOST @@ -428,8 +428,8 @@ inline void rotmg(const char *func_name, Func func, sycl::queue &queue, sycl::bu } #define ROTMG_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, \ - sycl::buffer &x1, TYPE y1, sycl::buffer ¶m) { \ + void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, \ + sycl::buffer& x1, TYPE y1, sycl::buffer& param) { \ rotmg(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, d1, d2, x1, y1, param); \ } @@ -438,8 +438,8 @@ ROTMG_LAUNCHER(double, cublasDrotmg) #undef ROTMG_LAUNCHER template -inline void iamax(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &result) { +inline void iamax(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& result) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); // cuBLAS does not support int64_t as return type for the data. So we need to @@ -450,10 +450,10 @@ inline void iamax(const char *func_name, Func func, sycl::queue &queue, int64_t // to elementwise copy the data between two buffer, or allow reinterpret cast // to convert to different type with different typesize size. sycl::buffer int_res_buff{ sycl::range<1>(1) }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto int_res_acc = int_res_buff.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -461,8 +461,8 @@ inline void iamax(const char *func_name, Func func, sycl::queue &queue, int64_t // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); - auto x_ = sc.get_mem(x_acc); - auto int_res_ = sc.get_mem(int_res_acc); + auto x_ = sc.get_mem(x_acc); + auto int_res_ = sc.get_mem(int_res_acc); cublasStatus_t err; // For negative incx, iamax returns 0. This behaviour is similar to that of // reference netlib BLAS. @@ -474,7 +474,7 @@ inline void iamax(const char *func_name, Func func, sycl::queue &queue, int64_t }); }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto int_res_acc = int_res_buff.template get_access(cgh); auto result_acc = result.template get_access(cgh); cgh.single_task( @@ -483,8 +483,8 @@ inline void iamax(const char *func_name, Func func, sycl::queue &queue, int64_t } #define IAMAX_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void iamax(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void iamax(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ iamax(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result); \ } IAMAX_LAUNCHER(float, cublasIsamax) @@ -494,17 +494,17 @@ IAMAX_LAUNCHER(std::complex, cublasIzamax) #undef IAMAX_LAUNCHER template -inline void swap(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { +inline void swap(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, n, x_, incx, y_, incy); }); @@ -512,8 +512,8 @@ inline void swap(const char *func_name, Func func, sycl::queue &queue, int64_t n } #define SWAP_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void swap(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void swap(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ swap(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy); \ } @@ -524,8 +524,8 @@ SWAP_LAUNCHER(std::complex, cublasZswap) #undef SWAP_LAUNCHER template -inline void iamin(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &result) { +inline void iamin(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& result) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); // cuBLAS does not support int64_t as return type for the data. So we need to @@ -536,10 +536,10 @@ inline void iamin(const char *func_name, Func func, sycl::queue &queue, int64_t // to elementwise copy the data between two buffer, or allow reinterpret cast // to convert to different type with different typesize size. sycl::buffer int_res_buff{ sycl::range<1>(1) }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto int_res_acc = int_res_buff.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -547,8 +547,8 @@ inline void iamin(const char *func_name, Func func, sycl::queue &queue, int64_t // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); - auto x_ = sc.get_mem(x_acc); - auto int_res_ = sc.get_mem(int_res_acc); + auto x_ = sc.get_mem(x_acc); + auto int_res_ = sc.get_mem(int_res_acc); cublasStatus_t err; // For negative incx, iamin returns 0. This behaviour is similar to that of // implemented as a reference IAMIN. @@ -560,7 +560,7 @@ inline void iamin(const char *func_name, Func func, sycl::queue &queue, int64_t }); }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto int_res_acc = int_res_buff.template get_access(cgh); auto result_acc = result.template get_access(cgh); cgh.single_task( @@ -569,8 +569,8 @@ inline void iamin(const char *func_name, Func func, sycl::queue &queue, int64_t } #define IAMIN_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void iamin(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void iamin(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ iamin(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result); \ } IAMIN_LAUNCHER(float, cublasIsamin) @@ -580,16 +580,16 @@ IAMIN_LAUNCHER(std::complex, cublasIzamin) #undef IAMIN_LAUNCHER template -inline void nrm2(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &result) { +inline void nrm2(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& result) { using cuDataType1 = typename CudaEquivalentType::Type; using cuDataType2 = typename CudaEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the CUBLAS_POINTER_MODE_HOST // when the data is on buffer, it must be set to @@ -597,8 +597,8 @@ inline void nrm2(const char *func_name, Func func, sycl::queue &queue, int64_t n // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); - auto x_ = sc.get_mem(x_acc); - auto res_ = sc.get_mem(res_acc); + auto x_ = sc.get_mem(x_acc); + auto res_ = sc.get_mem(res_acc); cublasStatus_t err; // NRM2 does not support negative index cublas_native_named_func(func_name, func, err, handle, n, x_, std::abs(incx), res_); @@ -611,8 +611,8 @@ inline void nrm2(const char *func_name, Func func, sycl::queue &queue, int64_t n } #define NRM2_LAUNCHER(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - void nrm2(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void nrm2(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ nrm2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result); \ } NRM2_LAUNCHER(float, float, cublasSnrm2) @@ -625,24 +625,24 @@ NRM2_LAUNCHER(std::complex, double, cublasDznrm2) // Level 1 template -inline sycl::event asum(const char *func_name, Func func, sycl::queue &queue, int64_t n, - const T1 *x, const int64_t incx, T2 *result, - const std::vector &dependencies) { +inline sycl::event asum(const char* func_name, Func func, sycl::queue& queue, int64_t n, + const T1* x, const int64_t incx, T2* result, + const std::vector& dependencies) { using cuDataType1 = typename CudaEquivalentType::Type; using cuDataType2 = typename CudaEquivalentType::Type; overflow_check(n, incx); bool result_on_device = sycl::get_pointer_type(result, queue.get_context()) == sycl::usm::alloc::device; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto res_ = reinterpret_cast(result); + auto x_ = reinterpret_cast(x); + auto res_ = reinterpret_cast(result); if (result_on_device) { cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); } @@ -658,8 +658,8 @@ inline sycl::event asum(const char *func_name, Func func, sycl::queue &queue, in } #define ASUM_LAUNCHER_USM(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - sycl::event asum(sycl::queue &queue, int64_t n, const TYPE1 *x, const int64_t incx, \ - TYPE2 *result, const std::vector &dependencies) { \ + sycl::event asum(sycl::queue& queue, int64_t n, const TYPE1* x, const int64_t incx, \ + TYPE2* result, const std::vector& dependencies) { \ return asum(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } ASUM_LAUNCHER_USM(float, float, cublasSasum) @@ -669,22 +669,22 @@ ASUM_LAUNCHER_USM(std::complex, double, cublasDzasum) #undef ASUM_LAUNCHER_USM template -inline sycl::event scal(const char *func_name, Func func, sycl::queue &queue, int64_t n, T1 a, - T2 *x, int64_t incx, const std::vector &dependencies) { +inline sycl::event scal(const char* func_name, Func func, sycl::queue& queue, int64_t n, T1 a, + T2* x, int64_t incx, const std::vector& dependencies) { using cuDataType1 = typename CudaEquivalentType::Type; using cuDataType2 = typename CudaEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); + auto x_ = reinterpret_cast(x); cublasStatus_t err; // SCAL does not support negative incx - cublas_native_named_func(func_name, func, err, handle, n, (cuDataType1 *)&a, x_, + cublas_native_named_func(func_name, func, err, handle, n, (cuDataType1*)&a, x_, std::abs(incx)); }); }); @@ -692,8 +692,8 @@ inline sycl::event scal(const char *func_name, Func func, sycl::queue &queue, in } #define SCAL_LAUNCHER_USM(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - sycl::event scal(sycl::queue &queue, int64_t n, TYPE1 a, TYPE2 *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event scal(sycl::queue& queue, int64_t n, TYPE1 a, TYPE2* x, int64_t incx, \ + const std::vector& dependencies) { \ return scal(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, a, x, incx, dependencies); \ } SCAL_LAUNCHER_USM(float, float, cublasSscal) @@ -705,31 +705,31 @@ SCAL_LAUNCHER_USM(double, std::complex, cublasZdscal) #undef SCAL_LAUNCHER_USM template -inline sycl::event axpy(const char *func_name, Func func, sycl::queue &queue, int64_t n, T alpha, - const T *x, int64_t incx, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event axpy(const char* func_name, Func func, sycl::queue& queue, int64_t n, T alpha, + const T* x, int64_t incx, T* y, int64_t incy, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; - cublas_native_named_func(func_name, func, err, handle, n, (cuDataType *)&alpha, x_, - incx, y_, incy); + cublas_native_named_func(func_name, func, err, handle, n, (cuDataType*)&alpha, x_, incx, + y_, incy); }); }); return done; } #define AXPY_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event axpy(sycl::queue &queue, int64_t n, TYPE alpha, const TYPE *x, int64_t incx, \ - TYPE *y, int64_t incy, const std::vector &dependencies) { \ + sycl::event axpy(sycl::queue& queue, int64_t n, TYPE alpha, const TYPE* x, int64_t incx, \ + TYPE* y, int64_t incy, const std::vector& dependencies) { \ return axpy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, alpha, x, incx, y, incy, \ dependencies); \ } @@ -740,32 +740,32 @@ AXPY_LAUNCHER_USM(std::complex, cublasCaxpy) AXPY_LAUNCHER_USM(std::complex, cublasZaxpy) #undef AXPY_LAUNCHER_USM -sycl::event axpby(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, - float beta, float *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, + float beta, float* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for column_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - double beta, double *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + double beta, double* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for column_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for column_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for column_major layout"); } template -inline sycl::event rotg(const char *func_name, Func func, sycl::queue &queue, T1 *a, T1 *b, T2 *c, - T1 *s, const std::vector &dependencies) { +inline sycl::event rotg(const char* func_name, Func func, sycl::queue& queue, T1* a, T1* b, T2* c, + T1* s, const std::vector& dependencies) { using cuDataType1 = typename CudaEquivalentType::Type; using cuDataType2 = typename CudaEquivalentType::Type; auto ctx = queue.get_context(); @@ -783,17 +783,17 @@ inline sycl::event rotg(const char *func_name, Func func, sycl::queue &queue, T1 "If any pointer is only device accessible, all must be device accessible"); } } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); - auto s_ = reinterpret_cast(s); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); + auto s_ = reinterpret_cast(s); if (results_on_device) { cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); } @@ -808,8 +808,8 @@ inline sycl::event rotg(const char *func_name, Func func, sycl::queue &queue, T1 } #define ROTG_LAUNCHER_USM(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - sycl::event rotg(sycl::queue &queue, TYPE1 *a, TYPE1 *b, TYPE2 *c, TYPE1 *s, \ - const std::vector &dependencies) { \ + sycl::event rotg(sycl::queue& queue, TYPE1* a, TYPE1* b, TYPE2* c, TYPE1* s, \ + const std::vector& dependencies) { \ return rotg(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, a, b, c, s, dependencies); \ } @@ -820,21 +820,21 @@ ROTG_LAUNCHER_USM(std::complex, double, cublasZrotg) #undef ROTG_LAUNCHER_USM template -inline sycl::event rotm(const char *func_name, Func func, sycl::queue &queue, int64_t n, T *x, - int64_t incx, T *y, int64_t incy, T *param, - const std::vector &dependencies) { +inline sycl::event rotm(const char* func_name, Func func, sycl::queue& queue, int64_t n, T* x, + int64_t incx, T* y, int64_t incy, T* param, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); - auto param_ = reinterpret_cast(param); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); + auto param_ = reinterpret_cast(param); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, n, x_, incx, y_, incy, param_); }); @@ -843,8 +843,8 @@ inline sycl::event rotm(const char *func_name, Func func, sycl::queue &queue, in } #define ROTM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event rotm(sycl::queue &queue, int64_t n, TYPE *x, int64_t incx, TYPE *y, int64_t incy, \ - TYPE *param, const std::vector &dependencies) { \ + sycl::event rotm(sycl::queue& queue, int64_t n, TYPE* x, int64_t incx, TYPE* y, int64_t incy, \ + TYPE* param, const std::vector& dependencies) { \ return rotm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, param, \ dependencies); \ } @@ -854,20 +854,20 @@ ROTM_LAUNCHER_USM(double, cublasDrotm) #undef ROTM_LAUNCHER_USM template -inline sycl::event copy(const char *func_name, Func func, sycl::queue &queue, int64_t n, const T *x, - int64_t incx, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event copy(const char* func_name, Func func, sycl::queue& queue, int64_t n, const T* x, + int64_t incx, T* y, int64_t incy, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, n, x_, incx, y_, incy); }); @@ -876,8 +876,8 @@ inline sycl::event copy(const char *func_name, Func func, sycl::queue &queue, in } #define COPY_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event copy(sycl::queue &queue, int64_t n, const TYPE *x, int64_t incx, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event copy(sycl::queue& queue, int64_t n, const TYPE* x, int64_t incx, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return copy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, dependencies); \ } @@ -888,23 +888,23 @@ COPY_LAUNCHER_USM(std::complex, cublasZcopy) #undef COPY_LAUNCHER_USM template -inline sycl::event dot(const char *func_name, Func func, sycl::queue &queue, int64_t n, const T *x, - const int64_t incx, const T *y, int64_t incy, T *result, - const std::vector &dependencies) { +inline sycl::event dot(const char* func_name, Func func, sycl::queue& queue, int64_t n, const T* x, + const int64_t incx, const T* y, int64_t incy, T* result, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); bool result_on_device = sycl::get_pointer_type(result, queue.get_context()) == sycl::usm::alloc::device; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); - auto res_ = reinterpret_cast(result); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); + auto res_ = reinterpret_cast(result); if (result_on_device) { cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); } @@ -919,9 +919,9 @@ inline sycl::event dot(const char *func_name, Func func, sycl::queue &queue, int } #define DOT_LAUNCHER_USM(EXT, TYPE, CUBLAS_ROUTINE) \ - sycl::event dot##EXT(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - const TYPE *y, const int64_t incy, TYPE *result, \ - const std::vector &dependencies) { \ + sycl::event dot##EXT(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + const TYPE* y, const int64_t incy, TYPE* result, \ + const std::vector& dependencies) { \ return dot(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, result, \ dependencies); \ } @@ -934,34 +934,34 @@ DOT_LAUNCHER_USM(u, std::complex, cublasZdotu) #undef DOT_LAUNCHER_USM template -inline sycl::event rot(const char *func_name, Func func, sycl::queue &queue, int64_t n, T1 *x, - const int64_t incx, T1 *y, int64_t incy, T2 c, T3 s, - const std::vector &dependencies) { +inline sycl::event rot(const char* func_name, Func func, sycl::queue& queue, int64_t n, T1* x, + const int64_t incx, T1* y, int64_t incy, T2 c, T3 s, + const std::vector& dependencies) { using cuDataType1 = typename CudaEquivalentType::Type; using cuDataType2 = typename CudaEquivalentType::Type; using cuDataType3 = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, n, x_, incx, y_, incy, - (cuDataType2 *)&c, (cuDataType3 *)&s); + (cuDataType2*)&c, (cuDataType3*)&s); }); }); return done; } #define ROT_LAUNCHER_USM(TYPE1, TYPE2, TYPE3, CUBLAS_ROUTINE) \ - sycl::event rot(sycl::queue &queue, int64_t n, TYPE1 *x, const int64_t incx, TYPE1 *y, \ + sycl::event rot(sycl::queue& queue, int64_t n, TYPE1* x, const int64_t incx, TYPE1* y, \ int64_t incy, TYPE2 c, TYPE3 s, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return rot(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, c, s, \ dependencies); \ } @@ -972,23 +972,23 @@ ROT_LAUNCHER_USM(std::complex, float, float, cublasCsrot) ROT_LAUNCHER_USM(std::complex, double, double, cublasZdrot) #undef ROT_LAUNCHER_USM -sycl::event sdsdot(sycl::queue &queue, int64_t n, float sb, const float *x, int64_t incx, - const float *y, int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event sdsdot(sycl::queue& queue, int64_t n, float sb, const float* x, int64_t incx, + const float* y, int64_t incy, float* result, + const std::vector& dependencies) { overflow_check(n, incx, incy); bool result_on_device = sycl::get_pointer_type(result, queue.get_context()) == sycl::usm::alloc::device; // cuBLAS does not support sdsdot so we need to mimic sdot. - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); - auto res_ = reinterpret_cast(result); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); + auto res_ = reinterpret_cast(result); if (result_on_device) { cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); } @@ -1017,14 +1017,14 @@ sycl::event sdsdot(sycl::queue &queue, int64_t n, float sb, const float *x, int6 } } -sycl::event dot(sycl::queue &queue, int64_t n, const float *x, int64_t incx, const float *y, - int64_t incy, double *result, const std::vector &dependencies) { +sycl::event dot(sycl::queue& queue, int64_t n, const float* x, int64_t incx, const float* y, + int64_t incy, double* result, const std::vector& dependencies) { throw unimplemented("blas", "dot", "for column_major layout"); } template -inline sycl::event rotmg(const char *func_name, Func func, sycl::queue &queue, T *d1, T *d2, T *x1, - T y1, T *param, const std::vector &dependencies) { +inline sycl::event rotmg(const char* func_name, Func func, sycl::queue& queue, T* d1, T* d2, T* x1, + T y1, T* param, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; auto ctx = queue.get_context(); bool results_on_device = (sycl::get_pointer_type(d1, ctx) == sycl::usm::alloc::device || @@ -1039,22 +1039,22 @@ inline sycl::event rotmg(const char *func_name, Func func, sycl::queue &queue, T "If any pointer is only device accessible, all must be device accessible"); } } - cuDataType *y1_; + cuDataType* y1_; if (results_on_device) { y1_ = sycl::malloc_device(1, queue); queue.memcpy(y1_, &y1, sizeof(cuDataType)).wait(); } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto d1_ = reinterpret_cast(d1); - auto d2_ = reinterpret_cast(d2); - auto x1_ = reinterpret_cast(x1); - auto param_ = reinterpret_cast(param); + auto d1_ = reinterpret_cast(d1); + auto d2_ = reinterpret_cast(d2); + auto x1_ = reinterpret_cast(x1); + auto param_ = reinterpret_cast(param); cublasStatus_t err; if (results_on_device) { cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); @@ -1062,7 +1062,7 @@ inline sycl::event rotmg(const char *func_name, Func func, sycl::queue &queue, T cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_HOST); } else { - auto y1_c = reinterpret_cast(&y1); + auto y1_c = reinterpret_cast(&y1); cublas_native_named_func(func_name, func, err, handle, d1_, d2_, x1_, y1_c, param_); } }); @@ -1076,8 +1076,8 @@ inline sycl::event rotmg(const char *func_name, Func func, sycl::queue &queue, T } #define ROTMG_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event rotmg(sycl::queue &queue, TYPE *d1, TYPE *d2, TYPE *x1, TYPE y1, TYPE *param, \ - const std::vector &dependencies) { \ + sycl::event rotmg(sycl::queue& queue, TYPE* d1, TYPE* d2, TYPE* x1, TYPE y1, TYPE* param, \ + const std::vector& dependencies) { \ return rotmg(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, d1, d2, x1, y1, param, dependencies); \ } @@ -1086,9 +1086,9 @@ ROTMG_LAUNCHER_USM(double, cublasDrotmg) #undef ROTMG_LAUNCHER_USM template -inline sycl::event iamax(const char *func_name, Func func, sycl::queue &queue, int64_t n, - const T *x, const int64_t incx, int64_t *result, - const std::vector &dependencies) { +inline sycl::event iamax(const char* func_name, Func func, sycl::queue& queue, int64_t n, + const T* x, const int64_t incx, int64_t* result, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); // cuBLAS does not support int64_t as return type for the data. So we need to @@ -1097,7 +1097,7 @@ inline sycl::event iamax(const char *func_name, Func func, sycl::queue &queue, i // This change may cause failure as the result of integer overflow // based on the size. int int_res = 0; - int *int_res_p = nullptr; + int* int_res_p = nullptr; bool result_on_device = sycl::get_pointer_type(result, queue.get_context()) == sycl::usm::alloc::device; if (result_on_device) { @@ -1106,14 +1106,14 @@ inline sycl::event iamax(const char *func_name, Func func, sycl::queue &queue, i else { int_res_p = &int_res; } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); + auto x_ = reinterpret_cast(x); if (result_on_device) { cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); } @@ -1128,7 +1128,7 @@ inline sycl::event iamax(const char *func_name, Func func, sycl::queue &queue, i }); done.wait(); if (result_on_device) { - auto last_ev = queue.submit([&](sycl::handler &cgh) { + auto last_ev = queue.submit([&](sycl::handler& cgh) { cgh.single_task([=]() { *result = std::max((int64_t)*int_res_p - 1, (int64_t)0); }); }); last_ev.wait(); @@ -1142,8 +1142,8 @@ inline sycl::event iamax(const char *func_name, Func func, sycl::queue &queue, i } #define IAMAX_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event iamax(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - int64_t *result, const std::vector &dependencies) { \ + sycl::event iamax(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + int64_t* result, const std::vector& dependencies) { \ return iamax(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } IAMAX_LAUNCHER_USM(float, cublasIsamax) @@ -1153,20 +1153,20 @@ IAMAX_LAUNCHER_USM(std::complex, cublasIzamax) #undef IAMAX_LAUNCHER_USM template -inline sycl::event swap(const char *func_name, Func func, sycl::queue &queue, int64_t n, T *x, - int64_t incx, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event swap(const char* func_name, Func func, sycl::queue& queue, int64_t n, T* x, + int64_t incx, T* y, int64_t incy, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, n, x_, incx, y_, incy); }); @@ -1175,8 +1175,8 @@ inline sycl::event swap(const char *func_name, Func func, sycl::queue &queue, in } #define SWAP_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event swap(sycl::queue &queue, int64_t n, TYPE *x, int64_t incx, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event swap(sycl::queue& queue, int64_t n, TYPE* x, int64_t incx, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return swap(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, dependencies); \ } @@ -1187,9 +1187,9 @@ SWAP_LAUNCHER_USM(std::complex, cublasZswap) #undef SWAP_LAUNCHER_USM template -inline sycl::event iamin(const char *func_name, Func func, sycl::queue &queue, int64_t n, - const T *x, const int64_t incx, int64_t *result, - const std::vector &dependencies) { +inline sycl::event iamin(const char* func_name, Func func, sycl::queue& queue, int64_t n, + const T* x, const int64_t incx, int64_t* result, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); // cuBLAS does not support int64_t as return type for the data. So we need to @@ -1198,7 +1198,7 @@ inline sycl::event iamin(const char *func_name, Func func, sycl::queue &queue, i // This change may cause failure as the result of integer overflow // based on the size. int int_res = 0; - int *int_res_p = nullptr; + int* int_res_p = nullptr; bool result_on_device = sycl::get_pointer_type(result, queue.get_context()) == sycl::usm::alloc::device; if (result_on_device) { @@ -1207,14 +1207,14 @@ inline sycl::event iamin(const char *func_name, Func func, sycl::queue &queue, i else { int_res_p = &int_res; } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); + auto x_ = reinterpret_cast(x); if (result_on_device) { cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); } @@ -1229,7 +1229,7 @@ inline sycl::event iamin(const char *func_name, Func func, sycl::queue &queue, i }); done.wait(); if (result_on_device) { - auto last_ev = queue.submit([&](sycl::handler &cgh) { + auto last_ev = queue.submit([&](sycl::handler& cgh) { cgh.single_task([=]() { *result = std::max((int64_t)*int_res_p - 1, (int64_t)0); }); }); last_ev.wait(); @@ -1243,8 +1243,8 @@ inline sycl::event iamin(const char *func_name, Func func, sycl::queue &queue, i } #define IAMIN_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event iamin(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - int64_t *result, const std::vector &dependencies) { \ + sycl::event iamin(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + int64_t* result, const std::vector& dependencies) { \ return iamin(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } IAMIN_LAUNCHER_USM(float, cublasIsamin) @@ -1254,24 +1254,24 @@ IAMIN_LAUNCHER_USM(std::complex, cublasIzamin) #undef IAMIN_LAUNCHER_USM template -inline sycl::event nrm2(const char *func_name, Func func, sycl::queue &queue, int64_t n, - const T1 *x, const int64_t incx, T2 *result, - const std::vector &dependencies) { +inline sycl::event nrm2(const char* func_name, Func func, sycl::queue& queue, int64_t n, + const T1* x, const int64_t incx, T2* result, + const std::vector& dependencies) { using cuDataType1 = typename CudaEquivalentType::Type; using cuDataType2 = typename CudaEquivalentType::Type; overflow_check(n, incx); bool result_on_device = sycl::get_pointer_type(result, queue.get_context()) == sycl::usm::alloc::device; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto res_ = reinterpret_cast(result); + auto x_ = reinterpret_cast(x); + auto res_ = reinterpret_cast(result); if (result_on_device) { cublasSetPointerMode(handle, CUBLAS_POINTER_MODE_DEVICE); } @@ -1287,8 +1287,8 @@ inline sycl::event nrm2(const char *func_name, Func func, sycl::queue &queue, in } #define NRM2_LAUNCHER_USM(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - sycl::event nrm2(sycl::queue &queue, int64_t n, const TYPE1 *x, const int64_t incx, \ - TYPE2 *result, const std::vector &dependencies) { \ + sycl::event nrm2(sycl::queue& queue, int64_t n, const TYPE1* x, const int64_t incx, \ + TYPE2* result, const std::vector& dependencies) { \ return nrm2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } NRM2_LAUNCHER_USM(float, float, cublasSnrm2) @@ -1304,14 +1304,14 @@ namespace row_major { // Level 1 template -inline void asum(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &result) { +inline void asum(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& result) { throw unimplemented("blas", "asum", "for row_major layout"); } #define ASUM_LAUNCHER(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - void asum(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void asum(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ asum(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result); \ } ASUM_LAUNCHER(float, float, cublasSasum) @@ -1321,13 +1321,13 @@ ASUM_LAUNCHER(std::complex, double, cublasDzasum) #undef ASUM_LAUNCHER template -inline void scal(const char *func_name, Func func, sycl::queue &queue, int64_t n, T1 a, - sycl::buffer &x, int64_t incx) { +inline void scal(const char* func_name, Func func, sycl::queue& queue, int64_t n, T1 a, + sycl::buffer& x, int64_t incx) { throw unimplemented("blas", "scal", "for row_major layout"); } #define SCAL_LAUNCHER(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - void scal(sycl::queue &queue, int64_t n, TYPE1 a, sycl::buffer &x, int64_t incx) { \ + void scal(sycl::queue& queue, int64_t n, TYPE1 a, sycl::buffer& x, int64_t incx) { \ scal(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, a, x, incx); \ } SCAL_LAUNCHER(float, float, cublasSscal) @@ -1339,14 +1339,14 @@ SCAL_LAUNCHER(double, std::complex, cublasZdscal) #undef SCAL_LAUNCHER template -inline void axpy(const char *func_name, Func func, sycl::queue &queue, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { +inline void axpy(const char* func_name, Func func, sycl::queue& queue, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "axpy", "for row_major layout"); } #define AXPY_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void axpy(sycl::queue &queue, int64_t n, TYPE alpha, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void axpy(sycl::queue& queue, int64_t n, TYPE alpha, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ axpy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, alpha, x, incx, y, incy); \ } @@ -1356,37 +1356,37 @@ AXPY_LAUNCHER(std::complex, cublasCaxpy) AXPY_LAUNCHER(std::complex, cublasZaxpy) #undef AXPY_LAUNCHER -void axpby(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - float beta, sycl::buffer &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + float beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "axpby", "for row_major layout"); } -void axpby(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx, - double beta, sycl::buffer &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, int64_t incx, + double beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "axpby", "for row_major layout"); } -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { throw unimplemented("blas", "axpby", "for row_major layout"); } -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { throw unimplemented("blas", "axpby", "for row_major layout"); } template -inline void rotg(const char *func_name, Func func, sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { +inline void rotg(const char* func_name, Func func, sycl::queue& queue, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { throw unimplemented("blas", "rotg", "for row_major layout"); } #define ROTG_LAUNCHER(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, \ - sycl::buffer &c, sycl::buffer &s) { \ + void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, \ + sycl::buffer& c, sycl::buffer& s) { \ rotg(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, a, b, c, s); \ } @@ -1397,15 +1397,15 @@ ROTG_LAUNCHER(std::complex, double, cublasZrotg) #undef ROTG_LAUNCHER template -inline void rotm(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer ¶m) { +inline void rotm(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& param) { throw unimplemented("blas", "rotm", "for row_major layout"); } #define ROTM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void rotm(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { \ + void rotm(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy, sycl::buffer& param) { \ rotm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, param); \ } @@ -1414,14 +1414,14 @@ ROTM_LAUNCHER(double, cublasDrotm) #undef ROTM_LAUNCHER template -inline void copy(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { +inline void copy(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "copy", "for row_major layout"); } #define COPY_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void copy(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void copy(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ copy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy); \ } @@ -1432,15 +1432,15 @@ COPY_LAUNCHER(std::complex, cublasZcopy) #undef COPY_LAUNCHER template -inline void dot(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &result) { +inline void dot(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& result) { throw unimplemented("blas", "dot", "for row_major layout"); } #define DOT_LAUNCHER(EXT, TYPE, CUBLAS_ROUTINE) \ - void dot##EXT(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &y, const int64_t incy, sycl::buffer &result) { \ + void dot##EXT(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& y, const int64_t incy, sycl::buffer& result) { \ dot(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, result); \ } DOT_LAUNCHER(, float, cublasSdot) @@ -1452,15 +1452,15 @@ DOT_LAUNCHER(u, std::complex, cublasZdotu) #undef DOT_LAUNCHER template -inline void rot(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &y, int64_t incy, +inline void rot(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& y, int64_t incy, T2 c, T3 s) { throw unimplemented("blas", "rot", "for row_major layout"); } #define ROT_LAUNCHER(TYPE1, TYPE2, TYPE3, CUBLAS_ROUTINE) \ - void rot(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &y, int64_t incy, TYPE2 c, TYPE3 s) { \ + void rot(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& y, int64_t incy, TYPE2 c, TYPE3 s) { \ rot(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, c, s); \ } @@ -1470,25 +1470,25 @@ ROT_LAUNCHER(std::complex, float, float, cublasCsrot) ROT_LAUNCHER(std::complex, double, double, cublasZdrot) #undef ROT_LAUNCHER -void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { +void sdsdot(sycl::queue& queue, int64_t n, float sb, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { throw unimplemented("blas", "sdsdot", "for row_major layout"); } -void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { +void dot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { throw unimplemented("blas", "dot", "for row_major layout"); } template -inline void rotmg(const char *func_name, Func func, sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, T y1, sycl::buffer ¶m) { +inline void rotmg(const char* func_name, Func func, sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, T y1, sycl::buffer& param) { throw unimplemented("blas", "rotmg", "for row_major layout"); } #define ROTMG_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, \ - sycl::buffer &x1, TYPE y1, sycl::buffer ¶m) { \ + void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, \ + sycl::buffer& x1, TYPE y1, sycl::buffer& param) { \ rotmg(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, d1, d2, x1, y1, param); \ } @@ -1497,14 +1497,14 @@ ROTMG_LAUNCHER(double, cublasDrotmg) #undef ROTMG_LAUNCHER template -inline void iamax(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &result) { +inline void iamax(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& result) { throw unimplemented("blas", "iamax", "for row_major layout"); } #define IAMAX_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void iamax(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void iamax(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ iamax(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result); \ } IAMAX_LAUNCHER(float, cublasIsamax) @@ -1514,14 +1514,14 @@ IAMAX_LAUNCHER(std::complex, cublasIzamax) #undef IAMAX_LAUNCHER template -inline void swap(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy) { +inline void swap(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "swap", "for row_major layout"); } #define SWAP_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void swap(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void swap(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ swap(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy); \ } @@ -1532,14 +1532,14 @@ SWAP_LAUNCHER(std::complex, cublasZswap) #undef SWAP_LAUNCHER template -inline void iamin(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &result) { +inline void iamin(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& result) { throw unimplemented("blas", "iamin", "for row_major layout"); } #define IAMIN_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void iamin(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void iamin(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ iamin(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result); \ } IAMIN_LAUNCHER(float, cublasIsamin) @@ -1549,14 +1549,14 @@ IAMIN_LAUNCHER(std::complex, cublasIzamin) #undef IAMIN_LAUNCHER template -inline void nrm2(const char *func_name, Func func, sycl::queue &queue, int64_t n, - sycl::buffer &x, const int64_t incx, sycl::buffer &result) { +inline void nrm2(const char* func_name, Func func, sycl::queue& queue, int64_t n, + sycl::buffer& x, const int64_t incx, sycl::buffer& result) { throw unimplemented("blas", "nrm2", "for row_major layout"); } #define NRM2_LAUNCHER(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - void nrm2(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void nrm2(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ nrm2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result); \ } NRM2_LAUNCHER(float, float, cublasSnrm2) @@ -1569,15 +1569,15 @@ NRM2_LAUNCHER(std::complex, double, cublasDznrm2) // Level 1 template -inline sycl::event asum(const char *func_name, Func func, sycl::queue &queue, int64_t n, - const T1 *x, const int64_t incx, T2 *result, - const std::vector &dependencies) { +inline sycl::event asum(const char* func_name, Func func, sycl::queue& queue, int64_t n, + const T1* x, const int64_t incx, T2* result, + const std::vector& dependencies) { throw unimplemented("blas", "asum", "for row_major layout"); } #define ASUM_LAUNCHER_USM(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - sycl::event asum(sycl::queue &queue, int64_t n, const TYPE1 *x, const int64_t incx, \ - TYPE2 *result, const std::vector &dependencies) { \ + sycl::event asum(sycl::queue& queue, int64_t n, const TYPE1* x, const int64_t incx, \ + TYPE2* result, const std::vector& dependencies) { \ return asum(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } ASUM_LAUNCHER_USM(float, float, cublasSasum) @@ -1587,14 +1587,14 @@ ASUM_LAUNCHER_USM(std::complex, double, cublasDzasum) #undef ASUM_LAUNCHER_USM template -inline sycl::event scal(const char *func_name, Func func, sycl::queue &queue, int64_t n, T1 a, - T2 *x, int64_t incx, const std::vector &dependencies) { +inline sycl::event scal(const char* func_name, Func func, sycl::queue& queue, int64_t n, T1 a, + T2* x, int64_t incx, const std::vector& dependencies) { throw unimplemented("blas", "scal", "for row_major layout"); } #define SCAL_LAUNCHER_USM(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - sycl::event scal(sycl::queue &queue, int64_t n, TYPE1 a, TYPE2 *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event scal(sycl::queue& queue, int64_t n, TYPE1 a, TYPE2* x, int64_t incx, \ + const std::vector& dependencies) { \ return scal(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, a, x, incx, dependencies); \ } SCAL_LAUNCHER_USM(float, float, cublasSscal) @@ -1606,15 +1606,15 @@ SCAL_LAUNCHER_USM(double, std::complex, cublasZdscal) #undef SCAL_LAUNCHER_USM template -inline sycl::event axpy(const char *func_name, Func func, sycl::queue &queue, int64_t n, T alpha, - const T *x, int64_t incx, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event axpy(const char* func_name, Func func, sycl::queue& queue, int64_t n, T alpha, + const T* x, int64_t incx, T* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpy", "for row_major layout"); } #define AXPY_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event axpy(sycl::queue &queue, int64_t n, TYPE alpha, const TYPE *x, int64_t incx, \ - TYPE *y, int64_t incy, const std::vector &dependencies) { \ + sycl::event axpy(sycl::queue& queue, int64_t n, TYPE alpha, const TYPE* x, int64_t incx, \ + TYPE* y, int64_t incy, const std::vector& dependencies) { \ return axpy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, alpha, x, incx, y, incy, \ dependencies); \ } @@ -1625,38 +1625,38 @@ AXPY_LAUNCHER_USM(std::complex, cublasCaxpy) AXPY_LAUNCHER_USM(std::complex, cublasZaxpy) #undef AXPY_LAUNCHER_USM -sycl::event axpby(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, - float beta, float *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, + float beta, float* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for row_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - double beta, double *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + double beta, double* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for row_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for row_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for row_major layout"); } template -inline sycl::event rotg(const char *func_name, Func func, sycl::queue &queue, T1 *a, T1 *b, T2 *c, - T1 *s, const std::vector &dependencies) { +inline sycl::event rotg(const char* func_name, Func func, sycl::queue& queue, T1* a, T1* b, T2* c, + T1* s, const std::vector& dependencies) { throw unimplemented("blas", "rotg", "for row_major layout"); } #define ROTG_LAUNCHER_USM(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - sycl::event rotg(sycl::queue &queue, TYPE1 *a, TYPE1 *b, TYPE2 *c, TYPE1 *s, \ - const std::vector &dependencies) { \ + sycl::event rotg(sycl::queue& queue, TYPE1* a, TYPE1* b, TYPE2* c, TYPE1* s, \ + const std::vector& dependencies) { \ return rotg(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, a, b, c, s, dependencies); \ } @@ -1667,15 +1667,15 @@ ROTG_LAUNCHER_USM(std::complex, double, cublasZrotg) #undef ROTG_LAUNCHER_USM template -inline sycl::event rotm(const char *func_name, Func func, sycl::queue &queue, int64_t n, T *x, - int64_t incx, T *y, int64_t incy, T *param, - const std::vector &dependencies) { +inline sycl::event rotm(const char* func_name, Func func, sycl::queue& queue, int64_t n, T* x, + int64_t incx, T* y, int64_t incy, T* param, + const std::vector& dependencies) { throw unimplemented("blas", "rotm", "for row_major layout"); } #define ROTM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event rotm(sycl::queue &queue, int64_t n, TYPE *x, int64_t incx, TYPE *y, int64_t incy, \ - TYPE *param, const std::vector &dependencies) { \ + sycl::event rotm(sycl::queue& queue, int64_t n, TYPE* x, int64_t incx, TYPE* y, int64_t incy, \ + TYPE* param, const std::vector& dependencies) { \ return rotm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, param, \ dependencies); \ } @@ -1685,15 +1685,15 @@ ROTM_LAUNCHER_USM(double, cublasDrotm) #undef ROTM_LAUNCHER_USM template -inline sycl::event copy(const char *func_name, Func func, sycl::queue &queue, int64_t n, const T *x, - int64_t incx, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event copy(const char* func_name, Func func, sycl::queue& queue, int64_t n, const T* x, + int64_t incx, T* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "copy", "for row_major layout"); } #define COPY_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event copy(sycl::queue &queue, int64_t n, const TYPE *x, int64_t incx, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event copy(sycl::queue& queue, int64_t n, const TYPE* x, int64_t incx, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return copy(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, dependencies); \ } @@ -1704,16 +1704,16 @@ COPY_LAUNCHER_USM(std::complex, cublasZcopy) #undef COPY_LAUNCHER_USM template -inline sycl::event dot(const char *func_name, Func func, sycl::queue &queue, int64_t n, const T *x, - const int64_t incx, const T *y, int64_t incy, T *result, - const std::vector &dependencies) { +inline sycl::event dot(const char* func_name, Func func, sycl::queue& queue, int64_t n, const T* x, + const int64_t incx, const T* y, int64_t incy, T* result, + const std::vector& dependencies) { throw unimplemented("blas", "dot", "for row_major layout"); } #define DOT_LAUNCHER_USM(EXT, TYPE, CUBLAS_ROUTINE) \ - sycl::event dot##EXT(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - const TYPE *y, const int64_t incy, TYPE *result, \ - const std::vector &dependencies) { \ + sycl::event dot##EXT(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + const TYPE* y, const int64_t incy, TYPE* result, \ + const std::vector& dependencies) { \ return dot(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, result, \ dependencies); \ } @@ -1726,16 +1726,16 @@ DOT_LAUNCHER_USM(u, std::complex, cublasZdotu) #undef DOT_LAUNCHER_USM template -inline sycl::event rot(const char *func_name, Func func, sycl::queue &queue, int64_t n, T1 *x, - const int64_t incx, T1 *y, int64_t incy, T2 c, T3 s, - const std::vector &dependencies) { +inline sycl::event rot(const char* func_name, Func func, sycl::queue& queue, int64_t n, T1* x, + const int64_t incx, T1* y, int64_t incy, T2 c, T3 s, + const std::vector& dependencies) { throw unimplemented("blas", "rot", "for row_major layout"); } #define ROT_LAUNCHER_USM(TYPE1, TYPE2, TYPE3, CUBLAS_ROUTINE) \ - sycl::event rot(sycl::queue &queue, int64_t n, TYPE1 *x, const int64_t incx, TYPE1 *y, \ + sycl::event rot(sycl::queue& queue, int64_t n, TYPE1* x, const int64_t incx, TYPE1* y, \ int64_t incy, TYPE2 c, TYPE3 s, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return rot(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, c, s, \ dependencies); \ } @@ -1746,26 +1746,26 @@ ROT_LAUNCHER_USM(std::complex, float, float, cublasCsrot) ROT_LAUNCHER_USM(std::complex, double, double, cublasZdrot) #undef ROT_LAUNCHER_USM -sycl::event sdsdot(sycl::queue &queue, int64_t n, float sb, const float *x, int64_t incx, - const float *y, int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event sdsdot(sycl::queue& queue, int64_t n, float sb, const float* x, int64_t incx, + const float* y, int64_t incy, float* result, + const std::vector& dependencies) { throw unimplemented("blas", "sdsdot", "for row_major layout"); } -sycl::event dot(sycl::queue &queue, int64_t n, const float *x, int64_t incx, const float *y, - int64_t incy, double *result, const std::vector &dependencies) { +sycl::event dot(sycl::queue& queue, int64_t n, const float* x, int64_t incx, const float* y, + int64_t incy, double* result, const std::vector& dependencies) { throw unimplemented("blas", "dot", "for row_major layout"); } template -inline sycl::event rotmg(const char *func_name, Func func, sycl::queue &queue, T *d1, T *d2, T *x1, - T y1, T *param, const std::vector &dependencies) { +inline sycl::event rotmg(const char* func_name, Func func, sycl::queue& queue, T* d1, T* d2, T* x1, + T y1, T* param, const std::vector& dependencies) { throw unimplemented("blas", "rotmg", "for row_major layout"); } #define ROTMG_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event rotmg(sycl::queue &queue, TYPE *d1, TYPE *d2, TYPE *x1, TYPE y1, TYPE *param, \ - const std::vector &dependencies) { \ + sycl::event rotmg(sycl::queue& queue, TYPE* d1, TYPE* d2, TYPE* x1, TYPE y1, TYPE* param, \ + const std::vector& dependencies) { \ return rotmg(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, d1, d2, x1, y1, param, dependencies); \ } @@ -1774,15 +1774,15 @@ ROTMG_LAUNCHER_USM(double, cublasDrotmg) #undef ROTMG_LAUNCHER_USM template -inline sycl::event iamax(const char *func_name, Func func, sycl::queue &queue, int64_t n, - const T *x, const int64_t incx, int64_t *result, - const std::vector &dependencies) { +inline sycl::event iamax(const char* func_name, Func func, sycl::queue& queue, int64_t n, + const T* x, const int64_t incx, int64_t* result, + const std::vector& dependencies) { throw unimplemented("blas", "iamax", "for row_major layout"); } #define IAMAX_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event iamax(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - int64_t *result, const std::vector &dependencies) { \ + sycl::event iamax(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + int64_t* result, const std::vector& dependencies) { \ return iamax(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } IAMAX_LAUNCHER_USM(float, cublasIsamax) @@ -1792,15 +1792,15 @@ IAMAX_LAUNCHER_USM(std::complex, cublasIzamax) #undef IAMAX_LAUNCHER_USM template -inline sycl::event swap(const char *func_name, Func func, sycl::queue &queue, int64_t n, T *x, - int64_t incx, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event swap(const char* func_name, Func func, sycl::queue& queue, int64_t n, T* x, + int64_t incx, T* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "swap", "for row_major layout"); } #define SWAP_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event swap(sycl::queue &queue, int64_t n, TYPE *x, int64_t incx, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event swap(sycl::queue& queue, int64_t n, TYPE* x, int64_t incx, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return swap(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, y, incy, dependencies); \ } @@ -1811,15 +1811,15 @@ SWAP_LAUNCHER_USM(std::complex, cublasZswap) #undef SWAP_LAUNCHER_USM template -inline sycl::event iamin(const char *func_name, Func func, sycl::queue &queue, int64_t n, - const T *x, const int64_t incx, int64_t *result, - const std::vector &dependencies) { +inline sycl::event iamin(const char* func_name, Func func, sycl::queue& queue, int64_t n, + const T* x, const int64_t incx, int64_t* result, + const std::vector& dependencies) { throw unimplemented("blas", "iamin", "for row_major layout"); } #define IAMIN_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event iamin(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - int64_t *result, const std::vector &dependencies) { \ + sycl::event iamin(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + int64_t* result, const std::vector& dependencies) { \ return iamin(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } IAMIN_LAUNCHER_USM(float, cublasIsamin) @@ -1829,15 +1829,15 @@ IAMIN_LAUNCHER_USM(std::complex, cublasIzamin) #undef IAMIN_LAUNCHER_USM template -inline sycl::event nrm2(const char *func_name, Func func, sycl::queue &queue, int64_t n, - const T1 *x, const int64_t incx, T2 *result, - const std::vector &dependencies) { +inline sycl::event nrm2(const char* func_name, Func func, sycl::queue& queue, int64_t n, + const T1* x, const int64_t incx, T2* result, + const std::vector& dependencies) { throw unimplemented("blas", "nrm2", "for row_major layout"); } #define NRM2_LAUNCHER_USM(TYPE1, TYPE2, CUBLAS_ROUTINE) \ - sycl::event nrm2(sycl::queue &queue, int64_t n, const TYPE1 *x, const int64_t incx, \ - TYPE2 *result, const std::vector &dependencies) { \ + sycl::event nrm2(sycl::queue& queue, int64_t n, const TYPE1* x, const int64_t incx, \ + TYPE2* result, const std::vector& dependencies) { \ return nrm2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } NRM2_LAUNCHER_USM(float, float, cublasSnrm2) diff --git a/src/blas/backends/cublas/cublas_level2.cpp b/src/blas/backends/cublas/cublas_level2.cpp index 1bbfbc60f..87adc7fab 100644 --- a/src/blas/backends/cublas/cublas_level2.cpp +++ b/src/blas/backends/cublas/cublas_level2.cpp @@ -31,32 +31,32 @@ namespace column_major { // Buffer APIs template -inline void gemv(const char *func_name, Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, T beta, sycl::buffer &y, int64_t incy) { +inline void gemv(const char* func_name, Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, T beta, sycl::buffer& y, int64_t incy) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, m, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_operation(trans), m, - n, (cuDataType *)&alpha, a_, lda, x_, incx, - (cuDataType *)&beta, y_, incy); + n, (cuDataType*)&alpha, a_, lda, x_, incx, (cuDataType*)&beta, + y_, incy); }); }); } #define GEMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ gemv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, x, incx, beta, y, \ incy); \ } @@ -68,32 +68,32 @@ GEMV_LAUNCHER(std::complex, cublasZgemv) #undef GEMV_LAUNCHER template -inline void gbmv(const char *func_name, Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, int64_t kl, int64_t ku, T alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx, T beta, sycl::buffer &y, int64_t incy) { +inline void gbmv(const char* func_name, Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, int64_t kl, int64_t ku, T alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx, T beta, sycl::buffer& y, int64_t incy) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, m, lda, kl, ku, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_operation(trans), m, - n, kl, ku, (cuDataType *)&alpha, a_, lda, x_, incx, - (cuDataType *)&beta, y_, incy); + n, kl, ku, (cuDataType*)&alpha, a_, lda, x_, incx, + (cuDataType*)&beta, y_, incy); }); }); } #define GBMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ - int64_t incx, TYPE beta, sycl::buffer &y, int64_t incy) { \ + void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ + int64_t incx, TYPE beta, sycl::buffer& y, int64_t incy) { \ gbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, \ beta, y, incy); \ } @@ -105,30 +105,30 @@ GBMV_LAUNCHER(std::complex, cublasZgbmv) #undef GBMV_LAUNCHER template -inline void ger(const char *func_name, Func func, sycl::queue &queue, int64_t m, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { +inline void ger(const char* func_name, Func func, sycl::queue& queue, int64_t m, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, m, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; - cublas_native_named_func(func_name, func, err, handle, m, n, (cuDataType *)&alpha, x_, + cublas_native_named_func(func_name, func, err, handle, m, n, (cuDataType*)&alpha, x_, incx, y_, incy, a_, lda); }); }); } #define GER_LAUNCHER(EXT, TYPE, CUBLAS_ROUTINE) \ - void ger##EXT(sycl::queue &queue, int64_t m, int64_t n, TYPE alpha, sycl::buffer &x, \ - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, \ + void ger##EXT(sycl::queue& queue, int64_t m, int64_t n, TYPE alpha, sycl::buffer& x, \ + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, \ int64_t lda) { \ ger(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, m, n, alpha, x, incx, y, incy, a, lda); \ } @@ -142,32 +142,32 @@ GER_LAUNCHER(c, std::complex, cublasZgerc) #undef GER_LAUNCHER template -inline void hbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - int64_t k, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, T beta, sycl::buffer &y, int64_t incy) { +inline void hbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + int64_t k, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, T beta, sycl::buffer& y, int64_t incy) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, k, (cuDataType *)&alpha, - a_, lda, x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, k, (cuDataType*)&alpha, + a_, lda, x_, incx, (cuDataType*)&beta, y_, incy); }); }); } #define HBMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ hbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, \ beta, y, incy); \ } @@ -177,32 +177,32 @@ HBMV_LAUNCHER(std::complex, cublasZhbmv) #undef HBMV_LAUNCHER template -inline void hemv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - T beta, sycl::buffer &y, int64_t incy) { +inline void hemv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + T beta, sycl::buffer& y, int64_t incy) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, a_, - lda, x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, a_, + lda, x_, incx, (cuDataType*)&beta, y_, incy); }); }); } #define HEMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void hemv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void hemv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ hemv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, \ y, incy); \ } @@ -212,31 +212,31 @@ HEMV_LAUNCHER(std::complex, cublasZhemv) #undef HEMV_LAUNCHER template -inline void her(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - ScalarType alpha, sycl::buffer &x, int64_t incx, - sycl::buffer &a, int64_t lda) { +inline void her(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + ScalarType alpha, sycl::buffer& x, int64_t incx, + sycl::buffer& a, int64_t lda) { using cuScalarType = typename CudaEquivalentType::Type; using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuScalarType *)&alpha, + get_cublas_fill_mode(upper_lower), n, (cuScalarType*)&alpha, x_, incx, a_, lda); }); }); } #define HER_LAUNCHER(SCALAR_TYPE, DATA_TYPE, CUBLAS_ROUTINE) \ - void her(sycl::queue &queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a, \ + void her(sycl::queue& queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a, \ int64_t lda) { \ her(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda); \ } @@ -247,32 +247,32 @@ HER_LAUNCHER(double, std::complex, cublasZher) #undef HER_LAUNCHER template -inline void her2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { +inline void her2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, y_, incy, a_, lda); }); }); } #define HER2_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void her2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a, int64_t lda) { \ + void her2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a, int64_t lda) { \ her2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, \ lda); \ } @@ -283,32 +283,32 @@ HER2_LAUNCHER(std::complex, cublasZher2) #undef HER2_LAUNCHER template -inline void hpmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &a, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void hpmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& a, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, a_, - x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, a_, + x_, incx, (cuDataType*)&beta, y_, incy); }); }); } #define HPMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx, TYPE beta, \ - sycl::buffer &y, int64_t incy) { \ + void hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx, TYPE beta, \ + sycl::buffer& y, int64_t incy) { \ hpmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, \ incy); \ } @@ -319,30 +319,30 @@ HPMV_LAUNCHER(std::complex, cublasZhpmv) #undef HPMV_LAUNCHER template -inline void hpr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - ScalarType alpha, sycl::buffer &x, int64_t incx, - sycl::buffer &a) { +inline void hpr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + ScalarType alpha, sycl::buffer& x, int64_t incx, + sycl::buffer& a) { using cuScalarType = typename CudaEquivalentType::Type; using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuScalarType *)&alpha, + get_cublas_fill_mode(upper_lower), n, (cuScalarType*)&alpha, x_, incx, a_); }); }); } #define HPR_LAUNCHER(SCALAR_TYPE, DATA_TYPE, CUBLAS_ROUTINE) \ - void hpr(sycl::queue &queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a) { \ + void hpr(sycl::queue& queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a) { \ hpr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a); \ } @@ -352,32 +352,32 @@ HPR_LAUNCHER(double, std::complex, cublasZhpr) #undef HPR_LAUNCHER template -inline void hpr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a) { +inline void hpr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, y_, incy, a_); }); }); } #define HPR2_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a) { \ + void hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a) { \ hpr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a); \ } @@ -387,32 +387,32 @@ HPR2_LAUNCHER(std::complex, cublasZhpr2) #undef HPR2_LAUNCHER template -inline void sbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - int64_t k, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, T beta, sycl::buffer &y, int64_t incy) { +inline void sbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + int64_t k, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, T beta, sycl::buffer& y, int64_t incy) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, k, (cuDataType *)&alpha, - a_, lda, x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, k, (cuDataType*)&alpha, + a_, lda, x_, incx, (cuDataType*)&beta, y_, incy); }); }); } #define SBMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ sbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, \ beta, y, incy); \ } @@ -423,32 +423,32 @@ SBMV_LAUNCHER(double, cublasDsbmv) #undef SBMV_LAUNCHER template -inline void symv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - T beta, sycl::buffer &y, int64_t incy) { +inline void symv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + T beta, sycl::buffer& y, int64_t incy) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, a_, - lda, x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, a_, + lda, x_, incx, (cuDataType*)&beta, y_, incy); }); }); } #define SYMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void symv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void symv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ symv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, \ y, incy); \ } @@ -459,28 +459,28 @@ SYMV_LAUNCHER(double, cublasDsymv) #undef SYMV_LAUNCHER template -inline void syr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { +inline void syr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, a_, lda); }); }); } #define SYR_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void syr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { \ + void syr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { \ syr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda); \ } @@ -492,32 +492,32 @@ SYR_LAUNCHER(std::complex, cublasZsyr) #undef SYR_LAUNCHER template -inline void syr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { +inline void syr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, y_, incy, a_, lda); }); }); } #define SYR2_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void syr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a, int64_t lda) { \ + void syr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a, int64_t lda) { \ syr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, \ lda); \ } @@ -531,32 +531,32 @@ SYR2_LAUNCHER(std::complex, cublasZsyr2) #undef SYR2_LAUNCHER template -inline void spmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &a, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void spmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& a, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, a_, - x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, a_, + x_, incx, (cuDataType*)&beta, y_, incy); }); }); } #define SPMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void spmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx, TYPE beta, \ - sycl::buffer &y, int64_t incy) { \ + void spmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx, TYPE beta, \ + sycl::buffer& y, int64_t incy) { \ spmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, \ incy); \ } @@ -567,28 +567,28 @@ SPMV_LAUNCHER(double, cublasDspmv) #undef SPMV_LAUNCHER template -inline void spr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a) { +inline void spr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& a) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, a_); }); }); } #define SPR_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void spr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a) { \ + void spr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a) { \ spr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a); \ } @@ -598,32 +598,32 @@ SPR_LAUNCHER(double, cublasDspr) #undef SPR_LAUNCHER template -inline void spr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a) { +inline void spr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, y_, incy, a_); }); }); } #define SPR2_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void spr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a) { \ + void spr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a) { \ spr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a); \ } @@ -633,18 +633,18 @@ SPR2_LAUNCHER(double, cublasDspr2) #undef SPR2_LAUNCHER template -inline void tbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, - int64_t lda, sycl::buffer &x, int64_t incx) { +inline void tbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer& a, + int64_t lda, sycl::buffer& x, int64_t incx) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -654,8 +654,8 @@ inline void tbmv(const char *func_name, Func func, sycl::queue &queue, uplo uppe } #define TBMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ + void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ int64_t incx) { \ tbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, \ x, incx); \ @@ -669,18 +669,18 @@ TBMV_LAUNCHER(std::complex, cublasZtbmv) #undef TBMV_LAUNCHER template -inline void tbsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, - int64_t lda, sycl::buffer &x, int64_t incx) { +inline void tbsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer& a, + int64_t lda, sycl::buffer& x, int64_t incx) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -690,8 +690,8 @@ inline void tbsv(const char *func_name, Func func, sycl::queue &queue, uplo uppe } #define TBSV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ + void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ int64_t incx) { \ tbsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, \ x, incx); \ @@ -705,18 +705,18 @@ TBSV_LAUNCHER(std::complex, cublasZtbsv) #undef TBSV_LAUNCHER template -inline void tpmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, - sycl::buffer &x, int64_t incx) { +inline void tpmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, sycl::buffer& a, + sycl::buffer& x, int64_t incx) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -726,8 +726,8 @@ inline void tpmv(const char *func_name, Func func, sycl::queue &queue, uplo uppe } #define TPMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx) { \ + void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx) { \ tpmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, \ incx); \ } @@ -740,18 +740,18 @@ TPMV_LAUNCHER(std::complex, cublasZtpmv) #undef TPMV_LAUNCHER template -inline void tpsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, - sycl::buffer &x, int64_t incx) { +inline void tpsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, sycl::buffer& a, + sycl::buffer& x, int64_t incx) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -761,8 +761,8 @@ inline void tpsv(const char *func_name, Func func, sycl::queue &queue, uplo uppe } #define TPSV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx) { \ + void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx) { \ tpsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, \ incx); \ } @@ -775,18 +775,18 @@ TPSV_LAUNCHER(std::complex, cublasZtpsv) #undef TPSV_LAUNCHER template -inline void trmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { +inline void trmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -796,8 +796,8 @@ inline void trmv(const char *func_name, Func func, sycl::queue &queue, uplo uppe } #define TRMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { \ + void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { \ trmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, \ incx); \ } @@ -810,18 +810,18 @@ TRMV_LAUNCHER(std::complex, cublasZtrmv) #undef TRMV_LAUNCHER template -inline void trsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { +inline void trsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -831,8 +831,8 @@ inline void trsv(const char *func_name, Func func, sycl::queue &queue, uplo uppe } #define TRSV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { \ + void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { \ trsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, \ incx); \ } @@ -847,35 +847,35 @@ TRSV_LAUNCHER(std::complex, cublasZtrsv) // USM APIs template -inline sycl::event gemv(const char *func_name, Func func, sycl::queue &queue, transpose trans, - int64_t m, int64_t n, T alpha, const T *a, int64_t lda, const T *x, - int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event gemv(const char* func_name, Func func, sycl::queue& queue, transpose trans, + int64_t m, int64_t n, T alpha, const T* a, int64_t lda, const T* x, + int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, m, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_operation(trans), m, - n, (cuDataType *)&alpha, a_, lda, x_, incx, - (cuDataType *)&beta, y_, incy); + n, (cuDataType*)&alpha, a_, lda, x_, incx, (cuDataType*)&beta, + y_, incy); }); }); return done; } #define GEMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return gemv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, x, incx, \ beta, y, incy, dependencies); \ } @@ -887,36 +887,36 @@ GEMV_LAUNCHER_USM(std::complex, cublasZgemv) #undef GEMV_LAUNCHER_USM template -inline sycl::event gbmv(const char *func_name, Func func, sycl::queue &queue, transpose trans, - int64_t m, int64_t n, int64_t kl, int64_t ku, T alpha, const T *a, - int64_t lda, const T *x, int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event gbmv(const char* func_name, Func func, sycl::queue& queue, transpose trans, + int64_t m, int64_t n, int64_t kl, int64_t ku, T alpha, const T* a, + int64_t lda, const T* x, int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, m, lda, kl, ku, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_operation(trans), m, - n, kl, ku, (cuDataType *)&alpha, a_, lda, x_, incx, - (cuDataType *)&beta, y_, incy); + n, kl, ku, (cuDataType*)&alpha, a_, lda, x_, incx, + (cuDataType*)&beta, y_, incy); }); }); return done; } #define GBMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, \ - int64_t ku, TYPE alpha, const TYPE *a, int64_t lda, const TYPE *x, \ - int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, \ + int64_t ku, TYPE alpha, const TYPE* a, int64_t lda, const TYPE* x, \ + int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return gbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, kl, ku, alpha, a, lda, x, \ incx, beta, y, incy, dependencies); \ } @@ -928,23 +928,23 @@ GBMV_LAUNCHER_USM(std::complex, cublasZgbmv) #undef GBMV_LAUNCHER_USM template -inline sycl::event ger(const char *func_name, Func func, sycl::queue &queue, int64_t m, int64_t n, - T alpha, const T *x, int64_t incx, const T *y, int64_t incy, T *a, - int64_t lda, const std::vector &dependencies) { +inline sycl::event ger(const char* func_name, Func func, sycl::queue& queue, int64_t m, int64_t n, + T alpha, const T* x, int64_t incx, const T* y, int64_t incy, T* a, + int64_t lda, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, m, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; - cublas_native_named_func(func_name, func, err, handle, m, n, (cuDataType *)&alpha, x_, + cublas_native_named_func(func_name, func, err, handle, m, n, (cuDataType*)&alpha, x_, incx, y_, incy, a_, lda); }); }); @@ -952,9 +952,9 @@ inline sycl::event ger(const char *func_name, Func func, sycl::queue &queue, int } #define GER_LAUNCHER_USM(EXT, TYPE, CUBLAS_ROUTINE) \ - sycl::event ger##EXT(sycl::queue &queue, int64_t m, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event ger##EXT(sycl::queue& queue, int64_t m, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return ger(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, m, n, alpha, x, incx, y, incy, a, lda, \ dependencies); \ } @@ -968,35 +968,35 @@ GER_LAUNCHER_USM(c, std::complex, cublasZgerc) #undef GER_LAUNCHER_USM template -inline sycl::event hbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, int64_t k, T alpha, const T *a, int64_t lda, const T *x, - int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event hbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, int64_t k, T alpha, const T* a, int64_t lda, const T* x, + int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, k, (cuDataType *)&alpha, - a_, lda, x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, k, (cuDataType*)&alpha, + a_, lda, x_, incx, (cuDataType*)&beta, y_, incy); }); }); return done; } #define HBMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return hbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, \ incx, beta, y, incy, dependencies); \ } @@ -1006,34 +1006,34 @@ HBMV_LAUNCHER_USM(std::complex, cublasZhbmv) #undef HBMV_LAUNCHER_USM template -inline sycl::event hemv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *a, int64_t lda, const T *x, int64_t incx, - T beta, T *y, int64_t incy, const std::vector &dependencies) { +inline sycl::event hemv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* a, int64_t lda, const T* x, int64_t incx, + T beta, T* y, int64_t incy, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, a_, - lda, x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, a_, + lda, x_, incx, (cuDataType*)&beta, y_, incy); }); }); return done; } #define HEMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event hemv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event hemv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return hemv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, \ incx, beta, y, incy, dependencies); \ } @@ -1043,25 +1043,25 @@ HEMV_LAUNCHER_USM(std::complex, cublasZhemv) #undef HEMV_LAUNCHER_USM template -inline sycl::event her(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, const ScalarType alpha, const DataType *x, int64_t incx, - DataType *a, int64_t lda, const std::vector &dependencies) { +inline sycl::event her(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, const ScalarType alpha, const DataType* x, int64_t incx, + DataType* a, int64_t lda, const std::vector& dependencies) { using cuScalarType = typename CudaEquivalentType::Type; using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuScalarType *)&alpha, + get_cublas_fill_mode(upper_lower), n, (cuScalarType*)&alpha, x_, incx, a_, lda); }); }); @@ -1069,9 +1069,9 @@ inline sycl::event her(const char *func_name, Func func, sycl::queue &queue, upl } #define HER_LAUNCHER_USM(SCALAR_TYPE, DATA_TYPE, CUBLAS_ROUTINE) \ - sycl::event her(sycl::queue &queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ - const DATA_TYPE *x, int64_t incx, DATA_TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event her(sycl::queue& queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ + const DATA_TYPE* x, int64_t incx, DATA_TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return her(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda, \ dependencies); \ } @@ -1082,24 +1082,24 @@ HER_LAUNCHER_USM(double, std::complex, cublasZher) #undef HER_LAUNCHER_USM template -inline sycl::event her2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, - T *a, int64_t lda, const std::vector &dependencies) { +inline sycl::event her2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, const T* y, int64_t incy, + T* a, int64_t lda, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, y_, incy, a_, lda); }); }); @@ -1107,9 +1107,9 @@ inline sycl::event her2(const char *func_name, Func func, sycl::queue &queue, up } #define HER2_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event her2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event her2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return her2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, \ incy, a, lda, dependencies); \ } @@ -1120,34 +1120,34 @@ HER2_LAUNCHER_USM(std::complex, cublasZher2) #undef HER2_LAUNCHER_USM template -inline sycl::event hpmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *a, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event hpmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* a, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, a_, - x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, a_, + x_, incx, (cuDataType*)&beta, y_, incy); }); }); return done; } #define HPMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return hpmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, \ beta, y, incy, dependencies); \ } @@ -1158,24 +1158,24 @@ HPMV_LAUNCHER_USM(std::complex, cublasZhpmv) #undef HPMV_LAUNCHER_USM template -inline sycl::event hpr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, const ScalarType alpha, const DataType *x, int64_t incx, - DataType *a, const std::vector &dependencies) { +inline sycl::event hpr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, const ScalarType alpha, const DataType* x, int64_t incx, + DataType* a, const std::vector& dependencies) { using cuScalarType = typename CudaEquivalentType::Type; using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuScalarType *)&alpha, + get_cublas_fill_mode(upper_lower), n, (cuScalarType*)&alpha, x_, incx, a_); }); }); @@ -1183,9 +1183,9 @@ inline sycl::event hpr(const char *func_name, Func func, sycl::queue &queue, upl } #define HPR_LAUNCHER_USM(SCALAR_TYPE, DATA_TYPE, CUBLAS_ROUTINE) \ - sycl::event hpr(sycl::queue &queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ - const DATA_TYPE *x, int64_t incx, DATA_TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event hpr(sycl::queue& queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ + const DATA_TYPE* x, int64_t incx, DATA_TYPE* a, \ + const std::vector& dependencies) { \ return hpr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, \ dependencies); \ } @@ -1196,24 +1196,24 @@ HPR_LAUNCHER_USM(double, std::complex, cublasZhpr) #undef HPR_LAUNCHER_USM template -inline sycl::event hpr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, - T *a, const std::vector &dependencies) { +inline sycl::event hpr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, const T* y, int64_t incy, + T* a, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, y_, incy, a_); }); }); @@ -1221,9 +1221,9 @@ inline sycl::event hpr2(const char *func_name, Func func, sycl::queue &queue, up } #define HPR2_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, \ + const std::vector& dependencies) { \ return hpr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, \ incy, a, dependencies); \ } @@ -1234,35 +1234,35 @@ HPR2_LAUNCHER_USM(std::complex, cublasZhpr2) #undef HPR2_LAUNCHER_USM template -inline sycl::event sbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, int64_t k, T alpha, const T *a, int64_t lda, const T *x, - int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event sbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, int64_t k, T alpha, const T* a, int64_t lda, const T* x, + int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, k, (cuDataType *)&alpha, - a_, lda, x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, k, (cuDataType*)&alpha, + a_, lda, x_, incx, (cuDataType*)&beta, y_, incy); }); }); return done; } #define SBMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return sbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, \ incx, beta, y, incy, dependencies); \ } @@ -1273,34 +1273,34 @@ SBMV_LAUNCHER_USM(double, cublasDsbmv) #undef SBMV_LAUNCHER_USM template -inline sycl::event symv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *a, int64_t lda, const T *x, int64_t incx, - T beta, T *y, int64_t incy, const std::vector &dependencies) { +inline sycl::event symv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* a, int64_t lda, const T* x, int64_t incx, + T beta, T* y, int64_t incy, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, a_, - lda, x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, a_, + lda, x_, incx, (cuDataType*)&beta, y_, incy); }); }); return done; } #define SYMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event symv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event symv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return symv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, \ incx, beta, y, incy, dependencies); \ } @@ -1311,23 +1311,23 @@ SYMV_LAUNCHER_USM(double, cublasDsymv) #undef SYMV_LAUNCHER_USM template -inline sycl::event syr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, T *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event syr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, T* a, int64_t lda, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, a_, lda); }); }); @@ -1335,9 +1335,9 @@ inline sycl::event syr(const char *func_name, Func func, sycl::queue &queue, upl } #define SYR_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event syr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event syr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return syr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda, \ dependencies); \ } @@ -1350,24 +1350,24 @@ SYR_LAUNCHER_USM(std::complex, cublasZsyr) #undef SYR_LAUNCHER_USM template -inline sycl::event syr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, - T *a, int64_t lda, const std::vector &dependencies) { +inline sycl::event syr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, const T* y, int64_t incy, + T* a, int64_t lda, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, y_, incy, a_, lda); }); }); @@ -1375,9 +1375,9 @@ inline sycl::event syr2(const char *func_name, Func func, sycl::queue &queue, up } #define SYR2_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event syr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event syr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return syr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, \ incy, a, lda, dependencies); \ } @@ -1391,34 +1391,34 @@ SYR2_LAUNCHER_USM(std::complex, cublasZsyr2) #undef SYR2_LAUNCHER_USM template -inline sycl::event spmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *a, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event spmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* a, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, a_, - x_, incx, (cuDataType *)&beta, y_, incy); + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, a_, + x_, incx, (cuDataType*)&beta, y_, incy); }); }); return done; } #define SPMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event spmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event spmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return spmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, \ beta, y, incy, dependencies); \ } @@ -1429,23 +1429,23 @@ SPMV_LAUNCHER_USM(double, cublasDspmv) #undef SPMV_LAUNCHER_USM template -inline sycl::event spr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, T *a, - const std::vector &dependencies) { +inline sycl::event spr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, T* a, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, a_); }); }); @@ -1453,8 +1453,8 @@ inline sycl::event spr(const char *func_name, Func func, sycl::queue &queue, upl } #define SPR_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event spr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, TYPE *a, const std::vector &dependencies) { \ + sycl::event spr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, TYPE* a, const std::vector& dependencies) { \ return spr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, \ dependencies); \ } @@ -1465,24 +1465,24 @@ SPR_LAUNCHER_USM(double, cublasDspr) #undef SPR_LAUNCHER_USM template -inline sycl::event spr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, - T *a, const std::vector &dependencies) { +inline sycl::event spr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, const T* y, int64_t incy, + T* a, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, - get_cublas_fill_mode(upper_lower), n, (cuDataType *)&alpha, x_, + get_cublas_fill_mode(upper_lower), n, (cuDataType*)&alpha, x_, incx, y_, incy, a_); }); }); @@ -1490,9 +1490,9 @@ inline sycl::event spr2(const char *func_name, Func func, sycl::queue &queue, up } #define SPR2_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event spr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event spr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, \ + const std::vector& dependencies) { \ return spr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, \ incy, a, dependencies); \ } @@ -1503,21 +1503,21 @@ SPR2_LAUNCHER_USM(double, cublasDspr2) #undef SPR2_LAUNCHER_USM template -inline sycl::event tbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, int64_t k, const T *a, - int64_t lda, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, int64_t k, const T* a, + int64_t lda, T* x, int64_t incx, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -1528,9 +1528,9 @@ inline sycl::event tbmv(const char *func_name, Func func, sycl::queue &queue, up } #define TBMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, int64_t k, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, int64_t k, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, \ a, lda, x, incx, dependencies); \ } @@ -1543,21 +1543,21 @@ TBMV_LAUNCHER_USM(std::complex, cublasZtbmv) #undef TBMV_LAUNCHER_USM template -inline sycl::event tbsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, int64_t k, const T *a, - int64_t lda, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tbsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, int64_t k, const T* a, + int64_t lda, T* x, int64_t incx, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -1568,9 +1568,9 @@ inline sycl::event tbsv(const char *func_name, Func func, sycl::queue &queue, up } #define TBSV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, int64_t k, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, int64_t k, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tbsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, \ a, lda, x, incx, dependencies); \ } @@ -1583,20 +1583,20 @@ TBSV_LAUNCHER_USM(std::complex, cublasZtbsv) #undef TBSV_LAUNCHER_USM template -inline sycl::event tpmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, const T *a, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tpmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, const T* a, T* x, int64_t incx, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -1607,9 +1607,9 @@ inline sycl::event tpmv(const char *func_name, Func func, sycl::queue &queue, up } #define TPMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tpmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, \ x, incx, dependencies); \ } @@ -1622,20 +1622,20 @@ TPMV_LAUNCHER_USM(std::complex, cublasZtpmv) #undef TPMV_LAUNCHER_USM template -inline sycl::event tpsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, const T *a, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tpsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, const T* a, T* x, int64_t incx, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -1646,9 +1646,9 @@ inline sycl::event tpsv(const char *func_name, Func func, sycl::queue &queue, up } #define TPSV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tpsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, \ x, incx, dependencies); \ } @@ -1661,20 +1661,20 @@ TPSV_LAUNCHER_USM(std::complex, cublasZtpsv) #undef TPSV_LAUNCHER_USM template -inline sycl::event trmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, const T *a, int64_t lda, T *x, - int64_t incx, const std::vector &dependencies) { +inline sycl::event trmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, const T* a, int64_t lda, T* x, + int64_t incx, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -1685,9 +1685,9 @@ inline sycl::event trmv(const char *func_name, Func func, sycl::queue &queue, up } #define TRMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return trmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, \ lda, x, incx, dependencies); \ } @@ -1700,20 +1700,20 @@ TRMV_LAUNCHER_USM(std::complex, cublasZtrmv) #undef TRMV_LAUNCHER_USM template -inline sycl::event trsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, const T *a, int64_t lda, T *x, - int64_t incx, const std::vector &dependencies) { +inline sycl::event trsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, const T* a, int64_t lda, T* x, + int64_t incx, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), @@ -1724,9 +1724,9 @@ inline sycl::event trsv(const char *func_name, Func func, sycl::queue &queue, up } #define TRSV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return trsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, \ lda, x, incx, dependencies); \ } @@ -1744,16 +1744,16 @@ namespace row_major { // Buffer APIs template -inline void gemv(const char *func_name, Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, T beta, sycl::buffer &y, int64_t incy) { +inline void gemv(const char* func_name, Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, T beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "gemv", "for row_major layout"); } #define GEMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ gemv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, x, incx, beta, y, \ incy); \ } @@ -1765,16 +1765,16 @@ GEMV_LAUNCHER(std::complex, cublasZgemv) #undef GEMV_LAUNCHER template -inline void gbmv(const char *func_name, Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, int64_t kl, int64_t ku, T alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx, T beta, sycl::buffer &y, int64_t incy) { +inline void gbmv(const char* func_name, Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, int64_t kl, int64_t ku, T alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx, T beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "gbmv", "for row_major layout"); } #define GBMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ - int64_t incx, TYPE beta, sycl::buffer &y, int64_t incy) { \ + void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ + int64_t incx, TYPE beta, sycl::buffer& y, int64_t incy) { \ gbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, \ beta, y, incy); \ } @@ -1786,15 +1786,15 @@ GBMV_LAUNCHER(std::complex, cublasZgbmv) #undef GBMV_LAUNCHER template -inline void ger(const char *func_name, Func func, sycl::queue &queue, int64_t m, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { +inline void ger(const char* func_name, Func func, sycl::queue& queue, int64_t m, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { throw unimplemented("blas", "ger", "for row_major layout"); } #define GER_LAUNCHER(EXT, TYPE, CUBLAS_ROUTINE) \ - void ger##EXT(sycl::queue &queue, int64_t m, int64_t n, TYPE alpha, sycl::buffer &x, \ - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, \ + void ger##EXT(sycl::queue& queue, int64_t m, int64_t n, TYPE alpha, sycl::buffer& x, \ + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, \ int64_t lda) { \ ger(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, m, n, alpha, x, incx, y, incy, a, lda); \ } @@ -1808,16 +1808,16 @@ GER_LAUNCHER(c, std::complex, cublasZgerc) #undef GER_LAUNCHER template -inline void hbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - int64_t k, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, T beta, sycl::buffer &y, int64_t incy) { +inline void hbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + int64_t k, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, T beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "hbmv", "for row_major layout"); } #define HBMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ hbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, \ beta, y, incy); \ } @@ -1827,16 +1827,16 @@ HBMV_LAUNCHER(std::complex, cublasZhbmv) #undef HBMV_LAUNCHER template -inline void hemv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - T beta, sycl::buffer &y, int64_t incy) { +inline void hemv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + T beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "hemv", "for row_major layout"); } #define HEMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void hemv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void hemv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ hemv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, \ y, incy); \ } @@ -1846,15 +1846,15 @@ HEMV_LAUNCHER(std::complex, cublasZhemv) #undef HEMV_LAUNCHER template -inline void her(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - ScalarType alpha, sycl::buffer &x, int64_t incx, - sycl::buffer &a, int64_t lda) { +inline void her(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + ScalarType alpha, sycl::buffer& x, int64_t incx, + sycl::buffer& a, int64_t lda) { throw unimplemented("blas", "her", "for row_major layout"); } #define HER_LAUNCHER(SCALAR_TYPE, DATA_TYPE, CUBLAS_ROUTINE) \ - void her(sycl::queue &queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a, \ + void her(sycl::queue& queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a, \ int64_t lda) { \ her(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda); \ } @@ -1865,16 +1865,16 @@ HER_LAUNCHER(double, std::complex, cublasZher) #undef HER_LAUNCHER template -inline void her2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { +inline void her2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { throw unimplemented("blas", "her2", "for row_major layout"); } #define HER2_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void her2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a, int64_t lda) { \ + void her2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a, int64_t lda) { \ her2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, \ lda); \ } @@ -1885,16 +1885,16 @@ HER2_LAUNCHER(std::complex, cublasZher2) #undef HER2_LAUNCHER template -inline void hpmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &a, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void hpmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& a, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "hpmv", "for row_major layout"); } #define HPMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx, TYPE beta, \ - sycl::buffer &y, int64_t incy) { \ + void hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx, TYPE beta, \ + sycl::buffer& y, int64_t incy) { \ hpmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, \ incy); \ } @@ -1905,15 +1905,15 @@ HPMV_LAUNCHER(std::complex, cublasZhpmv) #undef HPMV_LAUNCHER template -inline void hpr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - ScalarType alpha, sycl::buffer &x, int64_t incx, - sycl::buffer &a) { +inline void hpr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + ScalarType alpha, sycl::buffer& x, int64_t incx, + sycl::buffer& a) { throw unimplemented("blas", "hpr", "for row_major layout"); } #define HPR_LAUNCHER(SCALAR_TYPE, DATA_TYPE, CUBLAS_ROUTINE) \ - void hpr(sycl::queue &queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a) { \ + void hpr(sycl::queue& queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a) { \ hpr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a); \ } @@ -1923,16 +1923,16 @@ HPR_LAUNCHER(double, std::complex, cublasZhpr) #undef HPR_LAUNCHER template -inline void hpr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a) { +inline void hpr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a) { throw unimplemented("blas", "hpr2", "for row_major layout"); } #define HPR2_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a) { \ + void hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a) { \ hpr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a); \ } @@ -1942,16 +1942,16 @@ HPR2_LAUNCHER(std::complex, cublasZhpr2) #undef HPR2_LAUNCHER template -inline void sbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - int64_t k, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, T beta, sycl::buffer &y, int64_t incy) { +inline void sbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + int64_t k, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, T beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "sbmv", "for row_major layout"); } #define SBMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ sbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, \ beta, y, incy); \ } @@ -1962,16 +1962,16 @@ SBMV_LAUNCHER(double, cublasDsbmv) #undef SBMV_LAUNCHER template -inline void symv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - T beta, sycl::buffer &y, int64_t incy) { +inline void symv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + T beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "symv", "for row_major layout"); } #define SYMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void symv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void symv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ symv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, \ y, incy); \ } @@ -1982,14 +1982,14 @@ SYMV_LAUNCHER(double, cublasDsymv) #undef SYMV_LAUNCHER template -inline void syr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { +inline void syr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { throw unimplemented("blas", "syr", "for row_major layout"); } #define SYR_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void syr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { \ + void syr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { \ syr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda); \ } @@ -2001,16 +2001,16 @@ SYR_LAUNCHER(std::complex, cublasZsyr) #undef SYR_LAUNCHER template -inline void syr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { +inline void syr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { throw unimplemented("blas", "syr2", "for row_major layout"); } #define SYR2_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void syr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a, int64_t lda) { \ + void syr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a, int64_t lda) { \ syr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, \ lda); \ } @@ -2024,16 +2024,16 @@ SYR2_LAUNCHER(std::complex, cublasZsyr2) #undef SYR2_LAUNCHER template -inline void spmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &a, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void spmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& a, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "spmv", "for row_major layout"); } #define SPMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void spmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx, TYPE beta, \ - sycl::buffer &y, int64_t incy) { \ + void spmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx, TYPE beta, \ + sycl::buffer& y, int64_t incy) { \ spmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, \ incy); \ } @@ -2044,14 +2044,14 @@ SPMV_LAUNCHER(double, cublasDspmv) #undef SPMV_LAUNCHER template -inline void spr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &a) { +inline void spr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& a) { throw unimplemented("blas", "spr", "for row_major layout"); } #define SPR_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void spr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a) { \ + void spr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a) { \ spr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a); \ } @@ -2061,16 +2061,16 @@ SPR_LAUNCHER(double, cublasDspr) #undef SPR_LAUNCHER template -inline void spr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - T alpha, sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a) { +inline void spr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + T alpha, sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a) { throw unimplemented("blas", "spr2", "for row_major layout"); } #define SPR2_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void spr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a) { \ + void spr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a) { \ spr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a); \ } @@ -2080,15 +2080,15 @@ SPR2_LAUNCHER(double, cublasDspr2) #undef SPR2_LAUNCHER template -inline void tbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, - int64_t lda, sycl::buffer &x, int64_t incx) { +inline void tbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer& a, + int64_t lda, sycl::buffer& x, int64_t incx) { throw unimplemented("blas", "tbmv", "for row_major layout"); } #define TBMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ + void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ int64_t incx) { \ tbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, \ x, incx); \ @@ -2102,15 +2102,15 @@ TBMV_LAUNCHER(std::complex, cublasZtbmv) #undef TBMV_LAUNCHER template -inline void tbsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer &a, - int64_t lda, sycl::buffer &x, int64_t incx) { +inline void tbsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, int64_t k, sycl::buffer& a, + int64_t lda, sycl::buffer& x, int64_t incx) { throw unimplemented("blas", "tbsv", "for row_major layout"); } #define TBSV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ + void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ int64_t incx) { \ tbsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, \ x, incx); \ @@ -2124,15 +2124,15 @@ TBSV_LAUNCHER(std::complex, cublasZtbsv) #undef TBSV_LAUNCHER template -inline void tpmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, - sycl::buffer &x, int64_t incx) { +inline void tpmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, sycl::buffer& a, + sycl::buffer& x, int64_t incx) { throw unimplemented("blas", "tpmv", "for row_major layout"); } #define TPMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx) { \ + void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx) { \ tpmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, \ incx); \ } @@ -2145,15 +2145,15 @@ TPMV_LAUNCHER(std::complex, cublasZtpmv) #undef TPMV_LAUNCHER template -inline void tpsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, - sycl::buffer &x, int64_t incx) { +inline void tpsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, sycl::buffer& a, + sycl::buffer& x, int64_t incx) { throw unimplemented("blas", "tpsv", "for row_major layout"); } #define TPSV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx) { \ + void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx) { \ tpsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, \ incx); \ } @@ -2166,15 +2166,15 @@ TPSV_LAUNCHER(std::complex, cublasZtpsv) #undef TPSV_LAUNCHER template -inline void trmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { +inline void trmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { throw unimplemented("blas", "trmv", "for row_major layout"); } #define TRMV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { \ + void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { \ trmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, \ incx); \ } @@ -2187,15 +2187,15 @@ TRMV_LAUNCHER(std::complex, cublasZtrmv) #undef TRMV_LAUNCHER template -inline void trsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, sycl::buffer &a, int64_t lda, - sycl::buffer &x, int64_t incx) { +inline void trsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, sycl::buffer& a, int64_t lda, + sycl::buffer& x, int64_t incx) { throw unimplemented("blas", "trsv", "for row_major layout"); } #define TRSV_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { \ + void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { \ trsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, \ incx); \ } @@ -2210,17 +2210,17 @@ TRSV_LAUNCHER(std::complex, cublasZtrsv) // USM APIs template -inline sycl::event gemv(const char *func_name, Func func, sycl::queue &queue, transpose trans, - int64_t m, int64_t n, T alpha, const T *a, int64_t lda, const T *x, - int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event gemv(const char* func_name, Func func, sycl::queue& queue, transpose trans, + int64_t m, int64_t n, T alpha, const T* a, int64_t lda, const T* x, + int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "gemv", "for row_major layout"); } #define GEMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return gemv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, x, incx, \ beta, y, incy, dependencies); \ } @@ -2232,18 +2232,18 @@ GEMV_LAUNCHER_USM(std::complex, cublasZgemv) #undef GEMV_LAUNCHER_USM template -inline sycl::event gbmv(const char *func_name, Func func, sycl::queue &queue, transpose trans, - int64_t m, int64_t n, int64_t kl, int64_t ku, T alpha, const T *a, - int64_t lda, const T *x, int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event gbmv(const char* func_name, Func func, sycl::queue& queue, transpose trans, + int64_t m, int64_t n, int64_t kl, int64_t ku, T alpha, const T* a, + int64_t lda, const T* x, int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "gbmv", "for row_major layout"); } #define GBMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, \ - int64_t ku, TYPE alpha, const TYPE *a, int64_t lda, const TYPE *x, \ - int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, \ + int64_t ku, TYPE alpha, const TYPE* a, int64_t lda, const TYPE* x, \ + int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return gbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, trans, m, n, kl, ku, alpha, a, lda, x, \ incx, beta, y, incy, dependencies); \ } @@ -2255,16 +2255,16 @@ GBMV_LAUNCHER_USM(std::complex, cublasZgbmv) #undef GBMV_LAUNCHER_USM template -inline sycl::event ger(const char *func_name, Func func, sycl::queue &queue, int64_t m, int64_t n, - T alpha, const T *x, int64_t incx, const T *y, int64_t incy, T *a, - int64_t lda, const std::vector &dependencies) { +inline sycl::event ger(const char* func_name, Func func, sycl::queue& queue, int64_t m, int64_t n, + T alpha, const T* x, int64_t incx, const T* y, int64_t incy, T* a, + int64_t lda, const std::vector& dependencies) { throw unimplemented("blas", "ger", "for row_major layout"); } #define GER_LAUNCHER_USM(EXT, TYPE, CUBLAS_ROUTINE) \ - sycl::event ger##EXT(sycl::queue &queue, int64_t m, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event ger##EXT(sycl::queue& queue, int64_t m, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return ger(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, m, n, alpha, x, incx, y, incy, a, lda, \ dependencies); \ } @@ -2278,17 +2278,17 @@ GER_LAUNCHER_USM(c, std::complex, cublasZgerc) #undef GER_LAUNCHER_USM template -inline sycl::event hbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, int64_t k, T alpha, const T *a, int64_t lda, const T *x, - int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event hbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, int64_t k, T alpha, const T* a, int64_t lda, const T* x, + int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "hbmv", "for row_major layout"); } #define HBMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return hbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, \ incx, beta, y, incy, dependencies); \ } @@ -2298,16 +2298,16 @@ HBMV_LAUNCHER_USM(std::complex, cublasZhbmv) #undef HBMV_LAUNCHER_USM template -inline sycl::event hemv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *a, int64_t lda, const T *x, int64_t incx, - T beta, T *y, int64_t incy, const std::vector &dependencies) { +inline sycl::event hemv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* a, int64_t lda, const T* x, int64_t incx, + T beta, T* y, int64_t incy, const std::vector& dependencies) { throw unimplemented("blas", "hemv", "for row_major layout"); } #define HEMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event hemv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event hemv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return hemv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, \ incx, beta, y, incy, dependencies); \ } @@ -2317,16 +2317,16 @@ HEMV_LAUNCHER_USM(std::complex, cublasZhemv) #undef HEMV_LAUNCHER_USM template -inline sycl::event her(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, const ScalarType alpha, const DataType *x, int64_t incx, - DataType *a, int64_t lda, const std::vector &dependencies) { +inline sycl::event her(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, const ScalarType alpha, const DataType* x, int64_t incx, + DataType* a, int64_t lda, const std::vector& dependencies) { throw unimplemented("blas", "her", "for row_major layout"); } #define HER_LAUNCHER_USM(SCALAR_TYPE, DATA_TYPE, CUBLAS_ROUTINE) \ - sycl::event her(sycl::queue &queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ - const DATA_TYPE *x, int64_t incx, DATA_TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event her(sycl::queue& queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ + const DATA_TYPE* x, int64_t incx, DATA_TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return her(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda, \ dependencies); \ } @@ -2337,16 +2337,16 @@ HER_LAUNCHER_USM(double, std::complex, cublasZher) #undef HER_LAUNCHER_USM template -inline sycl::event her2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, - T *a, int64_t lda, const std::vector &dependencies) { +inline sycl::event her2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, const T* y, int64_t incy, + T* a, int64_t lda, const std::vector& dependencies) { throw unimplemented("blas", "her2", "for row_major layout"); } #define HER2_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event her2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event her2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return her2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, \ incy, a, lda, dependencies); \ } @@ -2357,16 +2357,16 @@ HER2_LAUNCHER_USM(std::complex, cublasZher2) #undef HER2_LAUNCHER_USM template -inline sycl::event hpmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *a, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event hpmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* a, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { throw unimplemented("blas", "hpmv", "for row_major layout"); } #define HPMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return hpmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, \ beta, y, incy, dependencies); \ } @@ -2377,16 +2377,16 @@ HPMV_LAUNCHER_USM(std::complex, cublasZhpmv) #undef HPMV_LAUNCHER_USM template -inline sycl::event hpr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, const ScalarType alpha, const DataType *x, int64_t incx, - DataType *a, const std::vector &dependencies) { +inline sycl::event hpr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, const ScalarType alpha, const DataType* x, int64_t incx, + DataType* a, const std::vector& dependencies) { throw unimplemented("blas", "hpr", "for row_major layout"); } #define HPR_LAUNCHER_USM(SCALAR_TYPE, DATA_TYPE, CUBLAS_ROUTINE) \ - sycl::event hpr(sycl::queue &queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ - const DATA_TYPE *x, int64_t incx, DATA_TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event hpr(sycl::queue& queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ + const DATA_TYPE* x, int64_t incx, DATA_TYPE* a, \ + const std::vector& dependencies) { \ return hpr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, \ dependencies); \ } @@ -2397,16 +2397,16 @@ HPR_LAUNCHER_USM(double, std::complex, cublasZhpr) #undef HPR_LAUNCHER_USM template -inline sycl::event hpr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, - T *a, const std::vector &dependencies) { +inline sycl::event hpr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, const T* y, int64_t incy, + T* a, const std::vector& dependencies) { throw unimplemented("blas", "hpr2", "for row_major layout"); } #define HPR2_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, \ + const std::vector& dependencies) { \ return hpr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, \ incy, a, dependencies); \ } @@ -2417,17 +2417,17 @@ HPR2_LAUNCHER_USM(std::complex, cublasZhpr2) #undef HPR2_LAUNCHER_USM template -inline sycl::event sbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, int64_t k, T alpha, const T *a, int64_t lda, const T *x, - int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event sbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, int64_t k, T alpha, const T* a, int64_t lda, const T* x, + int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "sbmv", "for row_major layout"); } #define SBMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return sbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, \ incx, beta, y, incy, dependencies); \ } @@ -2438,16 +2438,16 @@ SBMV_LAUNCHER_USM(double, cublasDsbmv) #undef SBMV_LAUNCHER_USM template -inline sycl::event symv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *a, int64_t lda, const T *x, int64_t incx, - T beta, T *y, int64_t incy, const std::vector &dependencies) { +inline sycl::event symv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* a, int64_t lda, const T* x, int64_t incx, + T beta, T* y, int64_t incy, const std::vector& dependencies) { throw unimplemented("blas", "symv", "for row_major layout"); } #define SYMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event symv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event symv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return symv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, \ incx, beta, y, incy, dependencies); \ } @@ -2458,16 +2458,16 @@ SYMV_LAUNCHER_USM(double, cublasDsymv) #undef SYMV_LAUNCHER_USM template -inline sycl::event syr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, T *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event syr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, T* a, int64_t lda, + const std::vector& dependencies) { throw unimplemented("blas", "syr", "for row_major layout"); } #define SYR_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event syr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event syr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return syr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda, \ dependencies); \ } @@ -2480,16 +2480,16 @@ SYR_LAUNCHER_USM(std::complex, cublasZsyr) #undef SYR_LAUNCHER_USM template -inline sycl::event syr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, - T *a, int64_t lda, const std::vector &dependencies) { +inline sycl::event syr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, const T* y, int64_t incy, + T* a, int64_t lda, const std::vector& dependencies) { throw unimplemented("blas", "syr2", "for row_major layout"); } #define SYR2_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event syr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event syr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return syr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, \ incy, a, lda, dependencies); \ } @@ -2503,16 +2503,16 @@ SYR2_LAUNCHER_USM(std::complex, cublasZsyr2) #undef SYR2_LAUNCHER_USM template -inline sycl::event spmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *a, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event spmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* a, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { throw unimplemented("blas", "spmv", "for row_major layout"); } #define SPMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event spmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event spmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return spmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, \ beta, y, incy, dependencies); \ } @@ -2523,15 +2523,15 @@ SPMV_LAUNCHER_USM(double, cublasDspmv) #undef SPMV_LAUNCHER_USM template -inline sycl::event spr(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, T *a, - const std::vector &dependencies) { +inline sycl::event spr(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, T* a, + const std::vector& dependencies) { throw unimplemented("blas", "spr", "for row_major layout"); } #define SPR_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event spr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, TYPE *a, const std::vector &dependencies) { \ + sycl::event spr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, TYPE* a, const std::vector& dependencies) { \ return spr(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, \ dependencies); \ } @@ -2542,16 +2542,16 @@ SPR_LAUNCHER_USM(double, cublasDspr) #undef SPR_LAUNCHER_USM template -inline sycl::event spr2(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - int64_t n, T alpha, const T *x, int64_t incx, const T *y, int64_t incy, - T *a, const std::vector &dependencies) { +inline sycl::event spr2(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + int64_t n, T alpha, const T* x, int64_t incx, const T* y, int64_t incy, + T* a, const std::vector& dependencies) { throw unimplemented("blas", "spr2", "for row_major layout"); } #define SPR2_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event spr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event spr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, \ + const std::vector& dependencies) { \ return spr2(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, \ incy, a, dependencies); \ } @@ -2562,17 +2562,17 @@ SPR2_LAUNCHER_USM(double, cublasDspr2) #undef SPR2_LAUNCHER_USM template -inline sycl::event tbmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, int64_t k, const T *a, - int64_t lda, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tbmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, int64_t k, const T* a, + int64_t lda, T* x, int64_t incx, + const std::vector& dependencies) { throw unimplemented("blas", "tbmv", "for row_major layout"); } #define TBMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, int64_t k, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, int64_t k, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tbmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, \ a, lda, x, incx, dependencies); \ } @@ -2585,17 +2585,17 @@ TBMV_LAUNCHER_USM(std::complex, cublasZtbmv) #undef TBMV_LAUNCHER_USM template -inline sycl::event tbsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, int64_t k, const T *a, - int64_t lda, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tbsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, int64_t k, const T* a, + int64_t lda, T* x, int64_t incx, + const std::vector& dependencies) { throw unimplemented("blas", "tbsv", "for row_major layout"); } #define TBSV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, int64_t k, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, int64_t k, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tbsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, \ a, lda, x, incx, dependencies); \ } @@ -2608,16 +2608,16 @@ TBSV_LAUNCHER_USM(std::complex, cublasZtbsv) #undef TBSV_LAUNCHER_USM template -inline sycl::event tpmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, const T *a, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tpmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, const T* a, T* x, int64_t incx, + const std::vector& dependencies) { throw unimplemented("blas", "tpmv", "for row_major layout"); } #define TPMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tpmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, \ x, incx, dependencies); \ } @@ -2630,16 +2630,16 @@ TPMV_LAUNCHER_USM(std::complex, cublasZtpmv) #undef TPMV_LAUNCHER_USM template -inline sycl::event tpsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, const T *a, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tpsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, const T* a, T* x, int64_t incx, + const std::vector& dependencies) { throw unimplemented("blas", "tpsv", "for row_major layout"); } #define TPSV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tpsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, \ x, incx, dependencies); \ } @@ -2652,16 +2652,16 @@ TPSV_LAUNCHER_USM(std::complex, cublasZtpsv) #undef TPSV_LAUNCHER_USM template -inline sycl::event trmv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, const T *a, int64_t lda, T *x, - int64_t incx, const std::vector &dependencies) { +inline sycl::event trmv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, const T* a, int64_t lda, T* x, + int64_t incx, const std::vector& dependencies) { throw unimplemented("blas", "trmv", "for row_major layout"); } #define TRMV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return trmv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, \ lda, x, incx, dependencies); \ } @@ -2674,16 +2674,16 @@ TRMV_LAUNCHER_USM(std::complex, cublasZtrmv) #undef TRMV_LAUNCHER_USM template -inline sycl::event trsv(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, diag unit_diag, int64_t n, const T *a, int64_t lda, T *x, - int64_t incx, const std::vector &dependencies) { +inline sycl::event trsv(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, diag unit_diag, int64_t n, const T* a, int64_t lda, T* x, + int64_t incx, const std::vector& dependencies) { throw unimplemented("blas", "trsv", "for row_major layout"); } #define TRSV_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return trsv(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, \ lda, x, incx, dependencies); \ } diff --git a/src/blas/backends/cublas/cublas_level3.cpp b/src/blas/backends/cublas/cublas_level3.cpp index ec6dc5192..58d1f4273 100644 --- a/src/blas/backends/cublas/cublas_level3.cpp +++ b/src/blas/backends/cublas/cublas_level3.cpp @@ -31,33 +31,33 @@ namespace column_major { // Buffer APIs template -inline void gemm(const char *func_name, Func func, sycl::queue &queue, transpose transa, - transpose transb, int64_t m, int64_t n, int64_t k, T alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, +inline void gemm(const char* func_name, Func func, sycl::queue& queue, transpose transa, + transpose transb, int64_t m, int64_t n, int64_t k, T alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& b, int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_operation(transa), - get_cublas_operation(transb), m, n, k, (cuDataType *)&alpha, - a_, lda, b_, ldb, (cuDataType *)&beta, c_, ldc); + get_cublas_operation(transb), m, n, k, (cuDataType*)&alpha, a_, + lda, b_, ldb, (cuDataType*)&beta, c_, ldc); }); }); } #define GEMM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, TYPE beta, sycl::buffer &c, \ + void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, TYPE beta, sycl::buffer& c, \ int64_t ldc) { \ gemm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, transa, transb, m, n, k, alpha, a, lda, b, \ ldb, beta, c, ldc); \ @@ -72,15 +72,15 @@ GEMM_LAUNCHER(std::complex, cublasZgemm) template -inline void gemm_ex(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, sycl::queue &queue, +inline void gemm_ex(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, T_C alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, - T_C beta, sycl::buffer &c, int64_t ldc) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb, + T_C beta, sycl::buffer& c, int64_t ldc) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; using cuDataType_C = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { if (!verify_support(queue, sycl::aspect::fp16)) { throw oneapi::math::unimplemented( "blas", "sycl::half", "half is not supported by the device or the sycl compiler"); @@ -88,17 +88,17 @@ inline void gemm_ex(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, sycl::que auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); cublasStatus_t err; #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND CUBLAS_ERROR_FUNC_SYNC(cublasGemmEx, err, handle, get_cublas_operation(transa), - get_cublas_operation(transb), m, n, k, (cuDataType_C *)&alpha, - a_, DT_A, lda, b_, DT_B, ldb, (cuDataType_C *)&beta, c_, DT_C, - ldc, DT_C, CUBLAS_GEMM_DEFAULT); + get_cublas_operation(transb), m, n, k, (cuDataType_C*)&alpha, a_, + DT_A, lda, b_, DT_B, ldb, (cuDataType_C*)&beta, c_, DT_C, ldc, + DT_C, CUBLAS_GEMM_DEFAULT); #else CUBLAS_ERROR_FUNC(cublasGemmEx, err, handle, get_cublas_operation(transa), get_cublas_operation(transb), m, n, k, (cuDataType_C *)&alpha, @@ -110,9 +110,9 @@ inline void gemm_ex(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, sycl::que } #define GEMM_EX_LAUNCHER(TYPE_A, TYPE_B, TYPE_C, CUDADATATYPE_A, CUDADATATYPE_B, CUDADATATYPE_C) \ - void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_C alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, TYPE_C beta, sycl::buffer &c, \ + void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_C alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, TYPE_C beta, sycl::buffer& c, \ int64_t ldc) { \ gemm_ex(CUDADATATYPE_A, CUDADATATYPE_B, CUDADATATYPE_C, queue, transa, transb, m, n, k, \ alpha, a, lda, b, ldb, beta, c, ldc); \ @@ -123,40 +123,40 @@ GEMM_EX_LAUNCHER(sycl::half, sycl::half, sycl::half, CUDA_R_16F, CUDA_R_16F, CUD #undef GEMM_EX_LAUNCHER -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemm", "for column_major layout"); } template -inline void symm(const char *func_name, Func func, sycl::queue &queue, side left_right, - uplo upper_lower, int64_t m, int64_t n, T alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, +inline void symm(const char* func_name, Func func, sycl::queue& queue, side left_right, + uplo upper_lower, int64_t m, int64_t n, T alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& b, int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_side_mode(left_right), - get_cublas_fill_mode(upper_lower), m, n, (cuDataType *)&alpha, - a_, lda, b_, ldb, (cuDataType *)&beta, c_, ldc); + get_cublas_fill_mode(upper_lower), m, n, (cuDataType*)&alpha, + a_, lda, b_, ldb, (cuDataType*)&beta, c_, ldc); }); }); } #define SYMM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ symm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, \ b, ldb, beta, c, ldc); \ } @@ -169,33 +169,33 @@ SYMM_LAUNCHER(std::complex, cublasZsymm) #undef SYMM_LAUNCHER template -inline void hemm(const char *func_name, Func func, sycl::queue &queue, side left_right, - uplo upper_lower, int64_t m, int64_t n, T alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, +inline void hemm(const char* func_name, Func func, sycl::queue& queue, side left_right, + uplo upper_lower, int64_t m, int64_t n, T alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& b, int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_side_mode(left_right), - get_cublas_fill_mode(upper_lower), m, n, (cuDataType *)&alpha, - a_, lda, b_, ldb, (cuDataType *)&beta, c_, ldc); + get_cublas_fill_mode(upper_lower), m, n, (cuDataType*)&alpha, + a_, lda, b_, ldb, (cuDataType*)&beta, c_, ldc); }); }); } #define HEMM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ hemm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, \ b, ldb, beta, c, ldc); \ } @@ -205,31 +205,31 @@ HEMM_LAUNCHER(std::complex, cublasZhemm) #undef HEMM_LAUNCHER template -inline void syrk(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, int64_t n, int64_t k, T alpha, sycl::buffer &a, int64_t lda, - T beta, sycl::buffer &c, int64_t ldc) { +inline void syrk(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, int64_t n, int64_t k, T alpha, sycl::buffer& a, int64_t lda, + T beta, sycl::buffer& c, int64_t ldc) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto c_ = sc.get_mem(c_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - n, k, (cuDataType *)&alpha, a_, lda, (cuDataType *)&beta, c_, + n, k, (cuDataType*)&alpha, a_, lda, (cuDataType*)&beta, c_, ldc); }); }); } #define SYRK_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, sycl::buffer &a, int64_t lda, TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, sycl::buffer& a, int64_t lda, TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ syrk(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, \ beta, c, ldc); \ } @@ -242,33 +242,33 @@ SYRK_LAUNCHER(std::complex, cublasZsyrk) #undef SYRK_LAUNCHER template -inline void herk(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, +inline void herk(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, ScalarType alpha, - sycl::buffer &a, int64_t lda, ScalarType beta, - sycl::buffer &c, int64_t ldc) { + sycl::buffer& a, int64_t lda, ScalarType beta, + sycl::buffer& c, int64_t ldc) { using cuDataType = typename CudaEquivalentType::Type; using cuScalarType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto c_ = sc.get_mem(c_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - n, k, (cuScalarType *)&alpha, a_, lda, (cuScalarType *)&beta, - c_, ldc); + n, k, (cuScalarType*)&alpha, a_, lda, (cuScalarType*)&beta, c_, + ldc); }); }); } #define HERK_LAUNCHER(DATA_TYPE, SCALAR_TYPE, CUBLAS_ROUTINE) \ - void herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - SCALAR_TYPE alpha, sycl::buffer &a, int64_t lda, SCALAR_TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + SCALAR_TYPE alpha, sycl::buffer& a, int64_t lda, SCALAR_TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ herk(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, \ beta, c, ldc); \ } @@ -279,34 +279,34 @@ HERK_LAUNCHER(std::complex, double, cublasZherk) #undef HERK_LAUNCHER template -inline void syr2k(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, int64_t n, int64_t k, T alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, +inline void syr2k(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, int64_t n, int64_t k, T alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& b, int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - n, k, (cuDataType *)&alpha, a_, lda, b_, ldb, - (cuDataType *)&beta, c_, ldc); + n, k, (cuDataType*)&alpha, a_, lda, b_, ldb, + (cuDataType*)&beta, c_, ldc); }); }); } #define SYR2K_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ syr2k(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, \ ldb, beta, c, ldc); \ } @@ -318,36 +318,36 @@ SYR2K_LAUNCHER(std::complex, cublasZsyr2k) #undef SYR2K_LAUNCHER template -inline void her2k(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, +inline void her2k(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, DataType alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, ScalarType beta, sycl::buffer &c, int64_t ldc) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, ScalarType beta, sycl::buffer& c, int64_t ldc) { using cuDataType = typename CudaEquivalentType::Type; using cuScalarType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - n, k, (cuDataType *)&alpha, a_, lda, b_, ldb, - (cuScalarType *)&beta, c_, ldc); + n, k, (cuDataType*)&alpha, a_, lda, b_, ldb, + (cuScalarType*)&beta, c_, ldc); }); }); } #define HER2K_LAUNCHER(DATA_TYPE, SCALAR_TYPE, CUBLAS_ROUTINE) \ - void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - DATA_TYPE alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, SCALAR_TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + DATA_TYPE alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, SCALAR_TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ her2k(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, \ ldb, beta, c, ldc); \ } @@ -362,31 +362,31 @@ HER2K_LAUNCHER(std::complex, double, cublasZher2k) // separated from the B matrix. It is possible to use B instead of C, but this // will slow-down the code. template -inline void trmm(const char *func_name, Func func, sycl::queue &queue, side left_right, +inline void trmm(const char* func_name, Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_side_mode(left_right), get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - get_cublas_diag_type(unit_diag), m, n, (cuDataType *)&alpha, - a_, lda, b_, ldb, b_, ldb); + get_cublas_diag_type(unit_diag), m, n, (cuDataType*)&alpha, a_, + lda, b_, ldb, b_, ldb); }); }); } #define TRMM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer &a, \ - int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer& a, \ + int64_t lda, sycl::buffer& b, int64_t ldb) { \ trmm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, \ n, alpha, a, lda, b, ldb); \ } @@ -398,31 +398,31 @@ TRMM_LAUNCHER(std::complex, cublasZtrmm) #undef TRMM_LAUNCHER template -inline void trsm(const char *func_name, Func func, sycl::queue &queue, side left_right, +inline void trsm(const char* func_name, Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_side_mode(left_right), get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - get_cublas_diag_type(unit_diag), m, n, (cuDataType *)&alpha, - a_, lda, b_, ldb); + get_cublas_diag_type(unit_diag), m, n, (cuDataType*)&alpha, a_, + lda, b_, ldb); }); }); } #define TRSM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer &a, \ - int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer& a, \ + int64_t lda, sycl::buffer& b, int64_t ldb) { \ trsm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, \ n, alpha, a, lda, b, ldb); \ } @@ -436,36 +436,36 @@ TRSM_LAUNCHER(std::complex, cublasZtrsm) // USM APIs template -inline sycl::event gemm(const char *func_name, Func func, sycl::queue &queue, transpose transa, - transpose transb, int64_t m, int64_t n, int64_t k, T alpha, const T *a, - int64_t lda, const T *b, int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event gemm(const char* func_name, Func func, sycl::queue& queue, transpose transa, + transpose transb, int64_t m, int64_t n, int64_t k, T alpha, const T* a, + int64_t lda, const T* b, int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_operation(transa), - get_cublas_operation(transb), m, n, k, (cuDataType *)&alpha, - a_, lda, b_, ldb, (cuDataType *)&beta, c_, ldc); + get_cublas_operation(transb), m, n, k, (cuDataType*)&alpha, a_, + lda, b_, ldb, (cuDataType*)&beta, c_, ldc); }); }); return done; } #define GEMM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, \ - int64_t ldb, TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, \ + int64_t ldb, TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return gemm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, transa, transb, m, n, k, alpha, a, \ lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -480,30 +480,30 @@ GEMM_LAUNCHER_USM(std::complex, cublasZgemm) template inline sycl::event gemm_ex_usm(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, - sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, T_C alpha, const T_A *a, int64_t lda, - const T_B *b, int64_t ldb, T_C beta, T_C *c, int64_t ldc, - const std::vector &dependencies) { + sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, int64_t k, T_C alpha, const T_A* a, int64_t lda, + const T_B* b, int64_t ldb, T_C beta, T_C* c, int64_t ldc, + const std::vector& dependencies) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; using cuDataType_C = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); cublasStatus_t err; #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND CUBLAS_ERROR_FUNC_SYNC(cublasGemmEx, err, handle, get_cublas_operation(transa), - get_cublas_operation(transb), m, n, k, (cuDataType_C *)&alpha, - a_, DT_A, lda, b_, DT_B, ldb, (cuDataType_C *)&beta, c_, DT_C, - ldc, DT_C, CUBLAS_GEMM_DEFAULT); + get_cublas_operation(transb), m, n, k, (cuDataType_C*)&alpha, a_, + DT_A, lda, b_, DT_B, ldb, (cuDataType_C*)&beta, c_, DT_C, ldc, + DT_C, CUBLAS_GEMM_DEFAULT); #else CUBLAS_ERROR_FUNC(cublasGemmEx, err, handle, get_cublas_operation(transa), get_cublas_operation(transb), m, n, k, (cuDataType_C *)&alpha, @@ -517,10 +517,10 @@ inline sycl::event gemm_ex_usm(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C #define GEMM_EX_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, CUDADATATYPE_A, CUDADATATYPE_B, \ CUDADATATYPE_C) \ - sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_C alpha, const TYPE_A *a, int64_t lda, const TYPE_B *b, \ - int64_t ldb, TYPE_C beta, TYPE_C *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_C alpha, const TYPE_A* a, int64_t lda, const TYPE_B* b, \ + int64_t ldb, TYPE_C beta, TYPE_C* c, int64_t ldc, \ + const std::vector& dependencies) { \ return gemm_ex_usm(CUDADATATYPE_A, CUDADATATYPE_B, CUDADATATYPE_C, queue, transa, transb, \ m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -530,44 +530,44 @@ GEMM_EX_LAUNCHER_USM(sycl::half, sycl::half, sycl::half, CUDA_R_16F, CUDA_R_16F, #undef GEMM_EX_LAUNCHER_USM -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const bfloat16 *a, int64_t lda, const bfloat16 *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const bfloat16* a, int64_t lda, const bfloat16* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemm", "for column_major layout"); } template -inline sycl::event symm(const char *func_name, Func func, sycl::queue &queue, side left_right, - uplo upper_lower, int64_t m, int64_t n, T alpha, const T *a, int64_t lda, - const T *b, int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event symm(const char* func_name, Func func, sycl::queue& queue, side left_right, + uplo upper_lower, int64_t m, int64_t n, T alpha, const T* a, int64_t lda, + const T* b, int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_side_mode(left_right), - get_cublas_fill_mode(upper_lower), m, n, (cuDataType *)&alpha, - a_, lda, b_, ldb, (cuDataType *)&beta, c_, ldc); + get_cublas_fill_mode(upper_lower), m, n, (cuDataType*)&alpha, + a_, lda, b_, ldb, (cuDataType*)&beta, c_, ldc); }); }); return done; } #define SYMM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return symm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, \ a, lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -580,36 +580,36 @@ SYMM_LAUNCHER_USM(std::complex, cublasZsymm) #undef SYMM_LAUNCHER_USM template -inline sycl::event hemm(const char *func_name, Func func, sycl::queue &queue, side left_right, - uplo upper_lower, int64_t m, int64_t n, T alpha, const T *a, int64_t lda, - const T *b, int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event hemm(const char* func_name, Func func, sycl::queue& queue, side left_right, + uplo upper_lower, int64_t m, int64_t n, T alpha, const T* a, int64_t lda, + const T* b, int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_side_mode(left_right), - get_cublas_fill_mode(upper_lower), m, n, (cuDataType *)&alpha, - a_, lda, b_, ldb, (cuDataType *)&beta, c_, ldc); + get_cublas_fill_mode(upper_lower), m, n, (cuDataType*)&alpha, + a_, lda, b_, ldb, (cuDataType*)&beta, c_, ldc); }); }); return done; } #define HEMM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return hemm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, \ a, lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -619,24 +619,24 @@ HEMM_LAUNCHER_USM(std::complex, cublasZhemm) #undef HEMM_LAUNCHER_USM template -inline sycl::event syrk(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, int64_t n, int64_t k, T alpha, const T *a, int64_t lda, - T beta, T *c, int64_t ldc, const std::vector &dependencies) { +inline sycl::event syrk(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, int64_t n, int64_t k, T alpha, const T* a, int64_t lda, + T beta, T* c, int64_t ldc, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto c_ = reinterpret_cast(c); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - n, k, (cuDataType *)&alpha, a_, lda, (cuDataType *)&beta, c_, + n, k, (cuDataType*)&alpha, a_, lda, (cuDataType*)&beta, c_, ldc); }); }); @@ -644,9 +644,9 @@ inline sycl::event syrk(const char *func_name, Func func, sycl::queue &queue, up } #define SYRK_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, const TYPE *a, int64_t lda, TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, const TYPE* a, int64_t lda, TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return syrk(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, \ lda, beta, c, ldc, dependencies); \ } @@ -659,37 +659,37 @@ SYRK_LAUNCHER_USM(std::complex, cublasZsyrk) #undef SYRK_LAUNCHER_USM template -inline sycl::event herk(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, +inline sycl::event herk(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, const ScalarType alpha, - const DataType *a, int64_t lda, const ScalarType beta, DataType *c, - int64_t ldc, const std::vector &dependencies) { + const DataType* a, int64_t lda, const ScalarType beta, DataType* c, + int64_t ldc, const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; using cuScalarType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto c_ = reinterpret_cast(c); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - n, k, (cuScalarType *)&alpha, a_, lda, (cuScalarType *)&beta, - c_, ldc); + n, k, (cuScalarType*)&alpha, a_, lda, (cuScalarType*)&beta, c_, + ldc); }); }); return done; } #define HERK_LAUNCHER_USM(DATA_TYPE, SCALAR_TYPE, CUBLAS_ROUTINE) \ - sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - const SCALAR_TYPE alpha, const DATA_TYPE *a, int64_t lda, \ - const SCALAR_TYPE beta, DATA_TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + const SCALAR_TYPE alpha, const DATA_TYPE* a, int64_t lda, \ + const SCALAR_TYPE beta, DATA_TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return herk(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, \ lda, beta, c, ldc, dependencies); \ } @@ -700,37 +700,37 @@ HERK_LAUNCHER_USM(std::complex, double, cublasZherk) #undef HERK_LAUNCHER_USM template -inline sycl::event syr2k(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, int64_t n, int64_t k, T alpha, const T *a, int64_t lda, - const T *b, int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event syr2k(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, int64_t n, int64_t k, T alpha, const T* a, int64_t lda, + const T* b, int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - n, k, (cuDataType *)&alpha, a_, lda, b_, ldb, - (cuDataType *)&beta, c_, ldc); + n, k, (cuDataType*)&alpha, a_, lda, b_, ldb, + (cuDataType*)&beta, c_, ldc); }); }); return done; } #define SYR2K_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return syr2k(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, \ lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -742,39 +742,39 @@ SYR2K_LAUNCHER_USM(std::complex, cublasZsyr2k) #undef SYR2K_LAUNCHER_USM template -inline sycl::event her2k(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, +inline sycl::event her2k(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, const DataType alpha, - const DataType *a, int64_t lda, const DataType *b, int64_t ldb, - const ScalarType beta, DataType *c, int64_t ldc, - const std::vector &dependencies) { + const DataType* a, int64_t lda, const DataType* b, int64_t ldb, + const ScalarType beta, DataType* c, int64_t ldc, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; using cuScalarType = typename CudaEquivalentType::Type; overflow_check(n, k, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - n, k, (cuDataType *)&alpha, a_, lda, b_, ldb, - (cuScalarType *)&beta, c_, ldc); + n, k, (cuDataType*)&alpha, a_, lda, b_, ldb, + (cuScalarType*)&beta, c_, ldc); }); }); return done; } #define HER2K_LAUNCHER_USM(DATA_TYPE, SCALAR_TYPE, CUBLAS_ROUTINE) \ - sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - const DATA_TYPE alpha, const DATA_TYPE *a, int64_t lda, const DATA_TYPE *b, \ - int64_t ldb, const SCALAR_TYPE beta, DATA_TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + const DATA_TYPE alpha, const DATA_TYPE* a, int64_t lda, const DATA_TYPE* b, \ + int64_t ldb, const SCALAR_TYPE beta, DATA_TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return her2k(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, \ lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -789,35 +789,35 @@ HER2K_LAUNCHER_USM(std::complex, double, cublasZher2k) // separated from the B matrix. It is possible to use B instead of C, but this // will slow-down the code. template -inline sycl::event trmm(const char *func_name, Func func, sycl::queue &queue, side left_right, +inline sycl::event trmm(const char* func_name, Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, - T alpha, const T *a, int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { + T alpha, const T* a, int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_side_mode(left_right), get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - get_cublas_diag_type(unit_diag), m, n, (cuDataType *)&alpha, - a_, lda, b_, ldb, b_, ldb); + get_cublas_diag_type(unit_diag), m, n, (cuDataType*)&alpha, a_, + lda, b_, ldb, b_, ldb); }); }); return done; } #define TRMM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE *a, int64_t lda, \ - TYPE *b, int64_t ldb, const std::vector &dependencies) { \ + sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE* a, int64_t lda, \ + TYPE* b, int64_t ldb, const std::vector& dependencies) { \ return trmm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, trans, \ unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); \ } @@ -829,35 +829,35 @@ TRMM_LAUNCHER_USM(std::complex, cublasZtrmm) #undef TRMM_LAUNCHER_USM template -inline sycl::event trsm(const char *func_name, Func func, sycl::queue &queue, side left_right, +inline sycl::event trsm(const char* func_name, Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, - T alpha, const T *a, int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { + T alpha, const T* a, int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldb); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler &sc) { + onemath_cublas_host_task(cgh, queue, [=](CublasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); cublasStatus_t err; cublas_native_named_func(func_name, func, err, handle, get_cublas_side_mode(left_right), get_cublas_fill_mode(upper_lower), get_cublas_operation(trans), - get_cublas_diag_type(unit_diag), m, n, (cuDataType *)&alpha, - a_, lda, b_, ldb); + get_cublas_diag_type(unit_diag), m, n, (cuDataType*)&alpha, a_, + lda, b_, ldb); }); }); return done; } #define TRSM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE *a, int64_t lda, \ - TYPE *b, int64_t ldb, const std::vector &dependencies) { \ + sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE* a, int64_t lda, \ + TYPE* b, int64_t ldb, const std::vector& dependencies) { \ return trsm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, trans, \ unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); \ } @@ -874,17 +874,17 @@ namespace row_major { // Buffer APIs template -inline void gemm(const char *func_name, Func func, sycl::queue &queue, transpose transa, - transpose transb, int64_t m, int64_t n, int64_t k, T alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, +inline void gemm(const char* func_name, Func func, sycl::queue& queue, transpose transa, + transpose transb, int64_t m, int64_t n, int64_t k, T alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& b, int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemm", "for row_major layout"); } #define GEMM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, TYPE beta, sycl::buffer &c, \ + void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, TYPE beta, sycl::buffer& c, \ int64_t ldc) { \ gemm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, transa, transb, m, n, k, alpha, a, lda, b, \ ldb, beta, c, ldc); \ @@ -899,17 +899,17 @@ GEMM_LAUNCHER(std::complex, cublasZgemm) template -inline void gemm_ex(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, sycl::queue &queue, +inline void gemm_ex(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, T_C alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, - T_C beta, sycl::buffer &c, int64_t ldc) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb, + T_C beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemm", "for row_major layout"); } #define GEMM_EX_LAUNCHER(TYPE_A, TYPE_B, TYPE_C, CUDADATATYPE_A, CUDADATATYPE_B, CUDADATATYPE_C) \ - void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_C alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, TYPE_C beta, sycl::buffer &c, \ + void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_C alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, TYPE_C beta, sycl::buffer& c, \ int64_t ldc) { \ gemm_ex(CUDADATATYPE_A, CUDADATATYPE_B, CUDADATATYPE_C, queue, transa, transb, m, n, k, \ alpha, a, lda, b, ldb, beta, c, ldc); \ @@ -920,24 +920,24 @@ GEMM_EX_LAUNCHER(sycl::half, sycl::half, sycl::half, CUDA_R_16F, CUDA_R_16F, CUD #undef GEMM_EX_LAUNCHER -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemm", "for row_major layout"); } template -inline void symm(const char *func_name, Func func, sycl::queue &queue, side left_right, - uplo upper_lower, int64_t m, int64_t n, T alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, +inline void symm(const char* func_name, Func func, sycl::queue& queue, side left_right, + uplo upper_lower, int64_t m, int64_t n, T alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& b, int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "symm", "for row_major layout"); } #define SYMM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ symm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, \ b, ldb, beta, c, ldc); \ } @@ -950,17 +950,17 @@ SYMM_LAUNCHER(std::complex, cublasZsymm) #undef SYMM_LAUNCHER template -inline void hemm(const char *func_name, Func func, sycl::queue &queue, side left_right, - uplo upper_lower, int64_t m, int64_t n, T alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, +inline void hemm(const char* func_name, Func func, sycl::queue& queue, side left_right, + uplo upper_lower, int64_t m, int64_t n, T alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& b, int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "hemm", "for row_major layout"); } #define HEMM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ hemm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, \ b, ldb, beta, c, ldc); \ } @@ -970,16 +970,16 @@ HEMM_LAUNCHER(std::complex, cublasZhemm) #undef HEMM_LAUNCHER template -inline void syrk(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, int64_t n, int64_t k, T alpha, sycl::buffer &a, int64_t lda, - T beta, sycl::buffer &c, int64_t ldc) { +inline void syrk(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, int64_t n, int64_t k, T alpha, sycl::buffer& a, int64_t lda, + T beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "syrk", "for row_major layout"); } #define SYRK_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, sycl::buffer &a, int64_t lda, TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, sycl::buffer& a, int64_t lda, TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ syrk(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, \ beta, c, ldc); \ } @@ -992,17 +992,17 @@ SYRK_LAUNCHER(std::complex, cublasZsyrk) #undef SYRK_LAUNCHER template -inline void herk(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, +inline void herk(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, ScalarType alpha, - sycl::buffer &a, int64_t lda, ScalarType beta, - sycl::buffer &c, int64_t ldc) { + sycl::buffer& a, int64_t lda, ScalarType beta, + sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "herk", "for row_major layout"); } #define HERK_LAUNCHER(DATA_TYPE, SCALAR_TYPE, CUBLAS_ROUTINE) \ - void herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - SCALAR_TYPE alpha, sycl::buffer &a, int64_t lda, SCALAR_TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + SCALAR_TYPE alpha, sycl::buffer& a, int64_t lda, SCALAR_TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ herk(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, \ beta, c, ldc); \ } @@ -1013,17 +1013,17 @@ HERK_LAUNCHER(std::complex, double, cublasZherk) #undef HERK_LAUNCHER template -inline void syr2k(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, int64_t n, int64_t k, T alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, +inline void syr2k(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, int64_t n, int64_t k, T alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& b, int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "syr2k", "for row_major layout"); } #define SYR2K_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ syr2k(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, \ ldb, beta, c, ldc); \ } @@ -1035,18 +1035,18 @@ SYR2K_LAUNCHER(std::complex, cublasZsyr2k) #undef SYR2K_LAUNCHER template -inline void her2k(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, +inline void her2k(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, DataType alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, ScalarType beta, sycl::buffer &c, int64_t ldc) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, ScalarType beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "her2k", "for row_major layout"); } #define HER2K_LAUNCHER(DATA_TYPE, SCALAR_TYPE, CUBLAS_ROUTINE) \ - void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - DATA_TYPE alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, SCALAR_TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + DATA_TYPE alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, SCALAR_TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ her2k(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, \ ldb, beta, c, ldc); \ } @@ -1061,16 +1061,16 @@ HER2K_LAUNCHER(std::complex, double, cublasZher2k) // separated from the B matrix. It is possible to use B instead of C, but this // will slow-down the code. template -inline void trmm(const char *func_name, Func func, sycl::queue &queue, side left_right, +inline void trmm(const char* func_name, Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { throw unimplemented("blas", "trmm", "for row_major layout"); } #define TRMM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer &a, \ - int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer& a, \ + int64_t lda, sycl::buffer& b, int64_t ldb) { \ trmm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, \ n, alpha, a, lda, b, ldb); \ } @@ -1082,16 +1082,16 @@ TRMM_LAUNCHER(std::complex, cublasZtrmm) #undef TRMM_LAUNCHER template -inline void trsm(const char *func_name, Func func, sycl::queue &queue, side left_right, +inline void trsm(const char* func_name, Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { throw unimplemented("blas", "trsm", "for row_major layout"); } #define TRSM_LAUNCHER(TYPE, CUBLAS_ROUTINE) \ - void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer &a, \ - int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer& a, \ + int64_t lda, sycl::buffer& b, int64_t ldb) { \ trsm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, \ n, alpha, a, lda, b, ldb); \ } @@ -1105,18 +1105,18 @@ TRSM_LAUNCHER(std::complex, cublasZtrsm) // USM APIs template -inline sycl::event gemm(const char *func_name, Func func, sycl::queue &queue, transpose transa, - transpose transb, int64_t m, int64_t n, int64_t k, T alpha, const T *a, - int64_t lda, const T *b, int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event gemm(const char* func_name, Func func, sycl::queue& queue, transpose transa, + transpose transb, int64_t m, int64_t n, int64_t k, T alpha, const T* a, + int64_t lda, const T* b, int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemm", "for row_major layout"); } #define GEMM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, \ - int64_t ldb, TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, \ + int64_t ldb, TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return gemm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, transa, transb, m, n, k, alpha, a, \ lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -1131,19 +1131,19 @@ GEMM_LAUNCHER_USM(std::complex, cublasZgemm) template inline sycl::event gemm_ex_usm(DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, - sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, T_C alpha, const T_A *a, int64_t lda, - const T_B *b, int64_t ldb, T_C beta, T_C *c, int64_t ldc, - const std::vector &dependencies) { + sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, int64_t k, T_C alpha, const T_A* a, int64_t lda, + const T_B* b, int64_t ldb, T_C beta, T_C* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemm", "for row_major layout"); } #define GEMM_EX_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, CUDADATATYPE_A, CUDADATATYPE_B, \ CUDADATATYPE_C) \ - sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_C alpha, const TYPE_A *a, int64_t lda, const TYPE_B *b, \ - int64_t ldb, TYPE_C beta, TYPE_C *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_C alpha, const TYPE_A* a, int64_t lda, const TYPE_B* b, \ + int64_t ldb, TYPE_C beta, TYPE_C* c, int64_t ldc, \ + const std::vector& dependencies) { \ return gemm_ex_usm(CUDADATATYPE_A, CUDADATATYPE_B, CUDADATATYPE_C, queue, transa, transb, \ m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -1153,26 +1153,26 @@ GEMM_EX_LAUNCHER_USM(sycl::half, sycl::half, sycl::half, CUDA_R_16F, CUDA_R_16F, #undef GEMM_EX_LAUNCHER_USM -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const bfloat16 *a, int64_t lda, const bfloat16 *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const bfloat16* a, int64_t lda, const bfloat16* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemm", "for row_major layout"); } template -inline sycl::event symm(const char *func_name, Func func, sycl::queue &queue, side left_right, - uplo upper_lower, int64_t m, int64_t n, T alpha, const T *a, int64_t lda, - const T *b, int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event symm(const char* func_name, Func func, sycl::queue& queue, side left_right, + uplo upper_lower, int64_t m, int64_t n, T alpha, const T* a, int64_t lda, + const T* b, int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "symm", "for row_major layout"); } #define SYMM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return symm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, \ a, lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -1185,18 +1185,18 @@ SYMM_LAUNCHER_USM(std::complex, cublasZsymm) #undef SYMM_LAUNCHER_USM template -inline sycl::event hemm(const char *func_name, Func func, sycl::queue &queue, side left_right, - uplo upper_lower, int64_t m, int64_t n, T alpha, const T *a, int64_t lda, - const T *b, int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event hemm(const char* func_name, Func func, sycl::queue& queue, side left_right, + uplo upper_lower, int64_t m, int64_t n, T alpha, const T* a, int64_t lda, + const T* b, int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "hemm", "for row_major layout"); } #define HEMM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return hemm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, \ a, lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -1206,16 +1206,16 @@ HEMM_LAUNCHER_USM(std::complex, cublasZhemm) #undef HEMM_LAUNCHER_USM template -inline sycl::event syrk(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, int64_t n, int64_t k, T alpha, const T *a, int64_t lda, - T beta, T *c, int64_t ldc, const std::vector &dependencies) { +inline sycl::event syrk(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, int64_t n, int64_t k, T alpha, const T* a, int64_t lda, + T beta, T* c, int64_t ldc, const std::vector& dependencies) { throw unimplemented("blas", "syrk", "for row_major layout"); } #define SYRK_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, const TYPE *a, int64_t lda, TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, const TYPE* a, int64_t lda, TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return syrk(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, \ lda, beta, c, ldc, dependencies); \ } @@ -1228,18 +1228,18 @@ SYRK_LAUNCHER_USM(std::complex, cublasZsyrk) #undef SYRK_LAUNCHER_USM template -inline sycl::event herk(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, +inline sycl::event herk(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, const ScalarType alpha, - const DataType *a, int64_t lda, const ScalarType beta, DataType *c, - int64_t ldc, const std::vector &dependencies) { + const DataType* a, int64_t lda, const ScalarType beta, DataType* c, + int64_t ldc, const std::vector& dependencies) { throw unimplemented("blas", "herk", "for row_major layout"); } #define HERK_LAUNCHER_USM(DATA_TYPE, SCALAR_TYPE, CUBLAS_ROUTINE) \ - sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - const SCALAR_TYPE alpha, const DATA_TYPE *a, int64_t lda, \ - const SCALAR_TYPE beta, DATA_TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + const SCALAR_TYPE alpha, const DATA_TYPE* a, int64_t lda, \ + const SCALAR_TYPE beta, DATA_TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return herk(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, \ lda, beta, c, ldc, dependencies); \ } @@ -1250,18 +1250,18 @@ HERK_LAUNCHER_USM(std::complex, double, cublasZherk) #undef HERK_LAUNCHER_USM template -inline sycl::event syr2k(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, - transpose trans, int64_t n, int64_t k, T alpha, const T *a, int64_t lda, - const T *b, int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event syr2k(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, + transpose trans, int64_t n, int64_t k, T alpha, const T* a, int64_t lda, + const T* b, int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "syr2k", "for row_major layout"); } #define SYR2K_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return syr2k(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, \ lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -1273,19 +1273,19 @@ SYR2K_LAUNCHER_USM(std::complex, cublasZsyr2k) #undef SYR2K_LAUNCHER_USM template -inline sycl::event her2k(const char *func_name, Func func, sycl::queue &queue, uplo upper_lower, +inline sycl::event her2k(const char* func_name, Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, const DataType alpha, - const DataType *a, int64_t lda, const DataType *b, int64_t ldb, - const ScalarType beta, DataType *c, int64_t ldc, - const std::vector &dependencies) { + const DataType* a, int64_t lda, const DataType* b, int64_t ldb, + const ScalarType beta, DataType* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "her2k", "for row_major layout"); } #define HER2K_LAUNCHER_USM(DATA_TYPE, SCALAR_TYPE, CUBLAS_ROUTINE) \ - sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - const DATA_TYPE alpha, const DATA_TYPE *a, int64_t lda, const DATA_TYPE *b, \ - int64_t ldb, const SCALAR_TYPE beta, DATA_TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + const DATA_TYPE alpha, const DATA_TYPE* a, int64_t lda, const DATA_TYPE* b, \ + int64_t ldb, const SCALAR_TYPE beta, DATA_TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return her2k(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, \ lda, b, ldb, beta, c, ldc, dependencies); \ } @@ -1300,17 +1300,17 @@ HER2K_LAUNCHER_USM(std::complex, double, cublasZher2k) // separated from the B matrix. It is possible to use B instead of C, but this // will slow-down the code. template -inline sycl::event trmm(const char *func_name, Func func, sycl::queue &queue, side left_right, +inline sycl::event trmm(const char* func_name, Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, - T alpha, const T *a, int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { + T alpha, const T* a, int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "trmm", "for row_major layout"); } #define TRMM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE *a, int64_t lda, \ - TYPE *b, int64_t ldb, const std::vector &dependencies) { \ + sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE* a, int64_t lda, \ + TYPE* b, int64_t ldb, const std::vector& dependencies) { \ return trmm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, trans, \ unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); \ } @@ -1322,17 +1322,17 @@ TRMM_LAUNCHER_USM(std::complex, cublasZtrmm) #undef TRMM_LAUNCHER_USM template -inline sycl::event trsm(const char *func_name, Func func, sycl::queue &queue, side left_right, +inline sycl::event trsm(const char* func_name, Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, - T alpha, const T *a, int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { + T alpha, const T* a, int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "trsm", "for row_major layout"); } #define TRSM_LAUNCHER_USM(TYPE, CUBLAS_ROUTINE) \ - sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE *a, int64_t lda, \ - TYPE *b, int64_t ldb, const std::vector &dependencies) { \ + sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE* a, int64_t lda, \ + TYPE* b, int64_t ldb, const std::vector& dependencies) { \ return trsm(#CUBLAS_ROUTINE, CUBLAS_ROUTINE, queue, left_right, upper_lower, trans, \ unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); \ } diff --git a/src/blas/backends/cublas/cublas_scope_handle.cpp b/src/blas/backends/cublas/cublas_scope_handle.cpp index b83185545..013d24d92 100644 --- a/src/blas/backends/cublas/cublas_scope_handle.cpp +++ b/src/blas/backends/cublas/cublas_scope_handle.cpp @@ -43,7 +43,7 @@ thread_local cublas_handle CublasScopedContextHandler::handle_helper cublas_handle{}; #endif -CublasScopedContextHandler::CublasScopedContextHandler(sycl::queue queue, sycl::interop_handle &ih) +CublasScopedContextHandler::CublasScopedContextHandler(sycl::queue queue, sycl::interop_handle& ih) : ih(ih), needToRecover_(false) { placedContext_ = new sycl::context(queue.get_context()); @@ -73,8 +73,8 @@ CublasScopedContextHandler::~CublasScopedContextHandler() noexcept(false) { delete placedContext_; } -void ContextCallback(void *userData) { - auto *ptr = static_cast *>(userData); +void ContextCallback(void* userData) { + auto* ptr = static_cast*>(userData); if (!ptr) { return; } @@ -92,7 +92,7 @@ void ContextCallback(void *userData) { } } -cublasHandle_t CublasScopedContextHandler::get_handle(const sycl::queue &queue) { +cublasHandle_t CublasScopedContextHandler::get_handle(const sycl::queue& queue) { auto cudaDevice = ih.get_native_device(); CUresult cuErr; CUcontext desired; @@ -139,10 +139,10 @@ cublasHandle_t CublasScopedContextHandler::get_handle(const sycl::queue &queue) return handle; } -CUstream CublasScopedContextHandler::get_stream(const sycl::queue &queue) { +CUstream CublasScopedContextHandler::get_stream(const sycl::queue& queue) { return sycl::get_native(queue); } -sycl::context CublasScopedContextHandler::get_context(const sycl::queue &queue) { +sycl::context CublasScopedContextHandler::get_context(const sycl::queue& queue) { return queue.get_context(); } diff --git a/src/blas/backends/cublas/cublas_scope_handle.hpp b/src/blas/backends/cublas/cublas_scope_handle.hpp index 3e0b4a8af..968286ed6 100644 --- a/src/blas/backends/cublas/cublas_scope_handle.hpp +++ b/src/blas/backends/cublas/cublas_scope_handle.hpp @@ -85,19 +85,19 @@ the handle must be destroyed when the context goes out of scope. This will bind class CublasScopedContextHandler { CUcontext original_; - sycl::context *placedContext_; + sycl::context* placedContext_; bool needToRecover_; - sycl::interop_handle &ih; + sycl::interop_handle& ih; #ifdef ONEMATH_PI_INTERFACE_REMOVED static thread_local cublas_handle handle_helper; #else static thread_local cublas_handle handle_helper; #endif - CUstream get_stream(const sycl::queue &queue); - sycl::context get_context(const sycl::queue &queue); + CUstream get_stream(const sycl::queue& queue); + sycl::context get_context(const sycl::queue& queue); public: - CublasScopedContextHandler(sycl::queue queue, sycl::interop_handle &ih); + CublasScopedContextHandler(sycl::queue queue, sycl::interop_handle& ih); ~CublasScopedContextHandler() noexcept(false); /** @@ -107,7 +107,7 @@ class CublasScopedContextHandler { * @param queue sycl queue. * @return cublasHandle_t a handle to construct cublas routines */ - cublasHandle_t get_handle(const sycl::queue &queue); + cublasHandle_t get_handle(const sycl::queue& queue); // This is a work-around function for reinterpret_casting the memory. This // will be fixed when SYCL-2020 has been implemented for Pi backend. template @@ -116,7 +116,7 @@ class CublasScopedContextHandler { return reinterpret_cast(cudaPtr); } - void wait_stream(const sycl::queue &queue) { + void wait_stream(const sycl::queue& queue) { cuStreamSynchronize(get_stream(queue)); } }; diff --git a/src/blas/backends/cublas/cublas_scope_handle_hipsycl.cpp b/src/blas/backends/cublas/cublas_scope_handle_hipsycl.cpp index 82870f0ae..de1870d28 100644 --- a/src/blas/backends/cublas/cublas_scope_handle_hipsycl.cpp +++ b/src/blas/backends/cublas/cublas_scope_handle_hipsycl.cpp @@ -26,10 +26,10 @@ namespace cublas { thread_local cublas_handle CublasScopedContextHandler::handle_helper = cublas_handle{}; -CublasScopedContextHandler::CublasScopedContextHandler(sycl::queue queue, sycl::interop_handle &ih) +CublasScopedContextHandler::CublasScopedContextHandler(sycl::queue queue, sycl::interop_handle& ih) : interop_h(ih) {} -cublasHandle_t CublasScopedContextHandler::get_handle(const sycl::queue &queue) { +cublasHandle_t CublasScopedContextHandler::get_handle(const sycl::queue& queue) { sycl::device device = queue.get_device(); int current_device = interop_h.get_native_device(); CUstream streamId = get_stream(queue); @@ -64,7 +64,7 @@ cublasHandle_t CublasScopedContextHandler::get_handle(const sycl::queue &queue) return handle; } -CUstream CublasScopedContextHandler::get_stream(const sycl::queue &queue) { +CUstream CublasScopedContextHandler::get_stream(const sycl::queue& queue) { return interop_h.get_native_queue(); } diff --git a/src/blas/backends/cublas/cublas_scope_handle_hipsycl.hpp b/src/blas/backends/cublas/cublas_scope_handle_hipsycl.hpp index bc358d9ca..28354a265 100644 --- a/src/blas/backends/cublas/cublas_scope_handle_hipsycl.hpp +++ b/src/blas/backends/cublas/cublas_scope_handle_hipsycl.hpp @@ -61,13 +61,13 @@ the handle must be destroyed when the context goes out of scope. This will bind class CublasScopedContextHandler { sycl::interop_handle interop_h; static thread_local cublas_handle handle_helper; - sycl::context get_context(const sycl::queue &queue); - CUstream get_stream(const sycl::queue &queue); + sycl::context get_context(const sycl::queue& queue); + CUstream get_stream(const sycl::queue& queue); public: - CublasScopedContextHandler(sycl::queue queue, sycl::interop_handle &ih); + CublasScopedContextHandler(sycl::queue queue, sycl::interop_handle& ih); - cublasHandle_t get_handle(const sycl::queue &queue); + cublasHandle_t get_handle(const sycl::queue& queue); // This is a work-around function for reinterpret_casting the memory. This // will be fixed when SYCL-2020 has been implemented for Pi backend. diff --git a/src/blas/backends/cublas/cublas_task.hpp b/src/blas/backends/cublas/cublas_task.hpp index 58d5306ff..1b5173052 100644 --- a/src/blas/backends/cublas/cublas_task.hpp +++ b/src/blas/backends/cublas/cublas_task.hpp @@ -58,7 +58,7 @@ namespace cublas { #ifdef __HIPSYCL__ template -static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { +static inline void host_task_internal(H& cgh, sycl::queue queue, F f) { cgh.hipSYCL_enqueue_custom_operation([f, queue](sycl::interop_handle ih) { auto sc = CublasScopedContextHandler(queue, ih); f(sc); @@ -66,9 +66,9 @@ static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { } #else template -static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { +static inline void host_task_internal(H& cgh, sycl::queue queue, F f) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - cgh.ext_codeplay_enqueue_native_command([f, queue](sycl::interop_handle ih){ + cgh.ext_codeplay_enqueue_native_command([f, queue](sycl::interop_handle ih) { #else cgh.host_task([f, queue](sycl::interop_handle ih) { #endif @@ -78,7 +78,7 @@ static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { } #endif template -static inline void onemath_cublas_host_task(H &cgh, sycl::queue queue, F f) { +static inline void onemath_cublas_host_task(H& cgh, sycl::queue queue, F f) { (void)host_task_internal(cgh, queue, f); } diff --git a/src/blas/backends/mkl_common/mkl_batch.cxx b/src/blas/backends/mkl_common/mkl_batch.cxx index 93bba84a1..cb97b9693 100644 --- a/src/blas/backends/mkl_common/mkl_batch.cxx +++ b/src/blas/backends/mkl_common/mkl_batch.cxx @@ -19,1015 +19,1142 @@ // Buffer APIs -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - std::int64_t stridex, sycl::buffer &y, int64_t incy, std::int64_t stridey, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + std::int64_t stridex, sycl::buffer& y, int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - std::int64_t stridex, sycl::buffer &y, int64_t incy, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + std::int64_t stridex, sycl::buffer& y, int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, - int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, + int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, - int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, + int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, int64_t incy, std::int64_t stridey, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size)); } -void axpy_batch(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, +void axpy_batch(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); } -void axpy_batch(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); } -void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); } -void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size)); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, - int64_t incx, int64_t stride_x, float beta, sycl::buffer &y, int64_t incy, +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, sycl::buffer& x, + int64_t incx, int64_t stride_x, float beta, sycl::buffer& y, int64_t incy, int64_t stride_y, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, - incy, stride_y, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, + n, alpha, a, lda, stride_a, x, incx, stride_x, + beta, y, incy, stride_y, batch_size)); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &x, int64_t incx, int64_t stride_x, double beta, - sycl::buffer &y, int64_t incy, int64_t stride_y, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, - incy, stride_y, batch_size)); +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& x, int64_t incx, int64_t stride_x, double beta, + sycl::buffer& y, int64_t incy, int64_t stride_y, int64_t batch_size) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, + n, alpha, a, lda, stride_a, x, incx, stride_x, + beta, y, incy, stride_y, batch_size)); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &x, int64_t incx, - int64_t stride_x, std::complex beta, sycl::buffer, 1> &y, +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& x, int64_t incx, + int64_t stride_x, std::complex beta, sycl::buffer, 1>& y, int64_t incy, int64_t stride_y, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, - incy, stride_y, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, + n, alpha, a, lda, stride_a, x, incx, stride_x, + beta, y, incy, stride_y, batch_size)); } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &x, int64_t incx, +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& x, int64_t incx, int64_t stride_x, std::complex beta, - sycl::buffer, 1> &y, int64_t incy, int64_t stride_y, + sycl::buffer, 1>& y, int64_t incy, int64_t stride_y, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, beta, y, - incy, stride_y, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, + n, alpha, a, lda, stride_a, x, incx, stride_x, + beta, y, incy, stride_y, batch_size)); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stride_a, sycl::buffer &x, - int64_t incx, int64_t stride_x, sycl::buffer &c, int64_t ldc, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stride_a, sycl::buffer& x, + int64_t incx, int64_t stride_x, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, - stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, + n, a, lda, stride_a, x, incx, stride_x, c, ldc, + stride_c, batch_size)); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &x, int64_t incx, int64_t stride_x, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, - stride_c, batch_size)); +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& x, int64_t incx, int64_t stride_x, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, + n, a, lda, stride_a, x, incx, stride_x, c, ldc, + stride_c, batch_size)); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& x, int64_t incx, int64_t stride_x, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, - stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, + n, a, lda, stride_a, x, incx, stride_x, c, ldc, + stride_c, batch_size)); } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& x, int64_t incx, int64_t stride_x, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, ldc, - stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, + n, a, lda, stride_a, x, incx, stride_x, c, ldc, + stride_c, batch_size)); } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, float beta, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size)); +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, float beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size)); } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, double beta, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size)); +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, double beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size)); } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, - int64_t stride_b, std::complex beta, sycl::buffer, 1> &c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, + int64_t stride_b, std::complex beta, sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size)); } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size)); } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, sycl::half alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::half beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, sycl::half alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::half beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size)); } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size)); } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { throw unimplemented("blas", "gemm_batch", "unsupported dtype combination: int8_t, int8_t, float, float"); } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size)); } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - stride_a, b, ldb, stride_b, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size)); } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - stride_a, b, ldb, stride_b, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size)); } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - stride_a, b, ldb, stride_b, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size)); } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - stride_a, b, ldb, stride_b, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size)); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, float beta, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, - stride_c, batch_size)); +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, float beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size)); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, - stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size)); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, std::complex beta, sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, - stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size)); } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, ldc, - stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size)); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size)); +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, a, lda, stride_a, b, ldb, + stride_b, batch_size)); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size)); +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, a, lda, stride_a, b, ldb, + stride_b, batch_size)); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, a, lda, stride_a, b, ldb, + stride_b, batch_size)); } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, a, lda, stride_a, b, ldb, + stride_b, batch_size)); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch( + queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch( + queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch( + queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy_batch( + queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size)); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - float beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size)); +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + float beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size)); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size)); +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size)); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size)); } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, ldb, - stride_b, c, ldc, stride_c, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size)); } // USM APIs -sycl::event copy_batch(sycl::queue &queue, int64_t n, const float *x, int64_t incx, - std::int64_t stridex, float *y, int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, - dependencies)); +sycl::event copy_batch(sycl::queue& queue, int64_t n, const float* x, int64_t incx, + std::int64_t stridex, float* y, int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, + stridey, batch_size, dependencies)); } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const double *x, int64_t incx, - std::int64_t stridex, double *y, int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, - dependencies)); +sycl::event copy_batch(sycl::queue& queue, int64_t n, const double* x, int64_t incx, + std::int64_t stridex, double* y, int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, + stridey, batch_size, dependencies)); } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - std::int64_t stridex, std::complex *y, int64_t incy, +sycl::event copy_batch(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::int64_t stridex, std::complex* y, int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, - dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, + stridey, batch_size, dependencies)); } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - std::int64_t stridex, std::complex *y, int64_t incy, +sycl::event copy_batch(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::int64_t stridex, std::complex* y, int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, stridey, batch_size, - dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, stridex, y, incy, + stridey, batch_size, dependencies)); } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const float **x, int64_t *incx, float **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, - dependencies)); +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const float** x, int64_t* incx, float** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, dependencies)); } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const double **x, int64_t *incx, double **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, - dependencies)); +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const double** x, int64_t* incx, double** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, dependencies)); } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, int64_t *incx, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, - dependencies)); +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const std::complex** x, int64_t* incx, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, dependencies)); } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, - int64_t *incx, std::complex **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, - dependencies)); +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const std::complex** x, + int64_t* incx, std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::copy_batch(queue, n, x, incx, y, incy, group_count, group_size, dependencies)); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, - int64_t stridex, float *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, - dependencies)); +sycl::event axpy_batch(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, + int64_t stridex, float* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, + stridey, batch_size, dependencies)); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - int64_t stridex, double *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, - dependencies)); +sycl::event axpy_batch(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + int64_t stridex, double* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, + stridey, batch_size, dependencies)); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, - dependencies)); +sycl::event axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, + stridey, batch_size, dependencies)); } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, - dependencies)); +sycl::event axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, stridex, y, incy, + stridey, batch_size, dependencies)); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, float *alpha, const float **x, int64_t *incx, - float **y, int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, - dependencies)); +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, float* alpha, const float** x, int64_t* incx, + float** y, int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, + group_count, group_size, dependencies)); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, double *alpha, const double **x, - int64_t *incx, double **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, - dependencies)); +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, double* alpha, const double** x, + int64_t* incx, double** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, + group_count, group_size, dependencies)); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, - dependencies)); +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, std::complex* alpha, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, + group_count, group_size, dependencies)); } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, group_count, group_size, - dependencies)); +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, std::complex* alpha, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy_batch(queue, n, alpha, x, incx, y, incy, + group_count, group_size, dependencies)); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, const float *x, int64_t incx, - int64_t stride_x, float beta, float *y, int64_t incy, int64_t stride_y, - int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, - beta, y, incy, stride_y, batch_size, dependencies)); +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stride_a, const float* x, int64_t incx, + int64_t stride_x, float beta, float* y, int64_t incy, int64_t stride_y, + int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch( + queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, + stride_x, beta, y, incy, stride_y, batch_size, dependencies)); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, const double *x, - int64_t incx, int64_t stride_x, double beta, double *y, int64_t incy, +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stride_a, const double* x, + int64_t incx, int64_t stride_x, double beta, double* y, int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, - beta, y, incy, stride_y, batch_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch( + queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, + stride_x, beta, y, incy, stride_y, batch_size, dependencies)); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, const std::complex *x, int64_t incx, - int64_t stride_x, std::complex beta, std::complex *y, +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, const std::complex* x, int64_t incx, + int64_t stride_x, std::complex beta, std::complex* y, int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, - beta, y, incy, stride_y, batch_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch( + queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, + stride_x, beta, y, incy, stride_y, batch_size, dependencies)); } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, const std::complex *x, int64_t incx, - int64_t stride_x, std::complex beta, std::complex *y, +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, const std::complex* x, int64_t incx, + int64_t stride_x, std::complex beta, std::complex* y, int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, stride_x, - beta, y, incy, stride_y, batch_size, dependencies)); -} - -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, float *alpha, - const float **a, int64_t *lda, const float **x, int64_t *incx, float *beta, - float **y, int64_t *incy, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, - group_count, groupsize, dependencies)); -} - -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, double *alpha, - const double **a, int64_t *lda, const double **x, int64_t *incx, - double *beta, double **y, int64_t *incy, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, - group_count, groupsize, dependencies)); -} - -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - const std::complex **x, int64_t *incx, std::complex *beta, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, - group_count, groupsize, dependencies)); -} - -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - const std::complex **x, int64_t *incx, std::complex *beta, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, incx, beta, y, incy, - group_count, groupsize, dependencies)); -} - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, const float *a, - int64_t lda, int64_t stride_a, const float *x, int64_t incx, - int64_t stride_x, float *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, - ldc, stride_c, batch_size, dependencies)); -} - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, const double *a, - int64_t lda, int64_t stride_a, const double *x, int64_t incx, - int64_t stride_x, double *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, - ldc, stride_c, batch_size, dependencies)); -} - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const std::complex *a, int64_t lda, int64_t stride_a, - const std::complex *x, int64_t incx, int64_t stride_x, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, - ldc, stride_c, batch_size, dependencies)); -} - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const std::complex *a, int64_t lda, int64_t stride_a, - const std::complex *x, int64_t incx, int64_t stride_x, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, x, incx, stride_x, c, - ldc, stride_c, batch_size, dependencies)); -} - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const float **a, int64_t *lda, const float **x, int64_t *incx, float **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, - groupsize, dependencies)); -} - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const double **a, int64_t *lda, const double **x, int64_t *incx, double **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, - groupsize, dependencies)); -} - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const std::complex **a, int64_t *lda, const std::complex **x, - int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, - groupsize, dependencies)); -} - -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const std::complex **a, int64_t *lda, const std::complex **x, - int64_t *incx, std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, x, incx, c, ldc, group_count, - groupsize, dependencies)); -} - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const float *a, int64_t lda, int64_t stride_a, - const float *b, int64_t ldb, int64_t stride_b, float beta, float *c, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv_batch( + queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, stride_a, x, incx, + stride_x, beta, y, incy, stride_y, batch_size, dependencies)); +} + +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, float* alpha, + const float** a, int64_t* lda, const float** x, int64_t* incx, float* beta, + float** y, int64_t* incy, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, + incx, beta, y, incy, group_count, groupsize, dependencies)); +} + +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, double* alpha, + const double** a, int64_t* lda, const double** x, int64_t* incx, + double* beta, double** y, int64_t* incy, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, + incx, beta, y, incy, group_count, groupsize, dependencies)); +} + +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + const std::complex** x, int64_t* incx, std::complex* beta, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, + incx, beta, y, incy, group_count, groupsize, dependencies)); +} + +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + const std::complex** x, int64_t* incx, std::complex* beta, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::gemv_batch(queue, detail::get_onemkl_transpose(transa), m, n, alpha, a, lda, x, + incx, beta, y, incy, group_count, groupsize, dependencies)); +} + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, const float* a, + int64_t lda, int64_t stride_a, const float* x, int64_t incx, + int64_t stride_x, float* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, + x, incx, stride_x, c, ldc, stride_c, batch_size, dependencies)); +} + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, const double* a, + int64_t lda, int64_t stride_a, const double* x, int64_t incx, + int64_t stride_x, double* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, + x, incx, stride_x, c, ldc, stride_c, batch_size, dependencies)); +} + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + const std::complex* a, int64_t lda, int64_t stride_a, + const std::complex* x, int64_t incx, int64_t stride_x, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, + x, incx, stride_x, c, ldc, stride_c, batch_size, dependencies)); +} + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + const std::complex* a, int64_t lda, int64_t stride_a, + const std::complex* x, int64_t incx, int64_t stride_x, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), m, n, a, lda, stride_a, + x, incx, stride_x, c, ldc, stride_c, batch_size, dependencies)); +} + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const float** a, int64_t* lda, const float** x, int64_t* incx, float** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), + m, n, a, lda, x, incx, c, ldc, group_count, + groupsize, dependencies)); +} + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const double** a, int64_t* lda, const double** x, int64_t* incx, double** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), + m, n, a, lda, x, incx, c, ldc, group_count, + groupsize, dependencies)); +} + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const std::complex** a, int64_t* lda, const std::complex** x, + int64_t* incx, std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), + m, n, a, lda, x, incx, c, ldc, group_count, + groupsize, dependencies)); +} + +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const std::complex** a, int64_t* lda, const std::complex** x, + int64_t* incx, std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dgmm_batch(queue, detail::get_onemkl_side(left_right), + m, n, a, lda, x, incx, c, ldc, group_count, + groupsize, dependencies)); +} + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const float* a, int64_t lda, int64_t stride_a, + const float* b, int64_t ldb, int64_t stride_b, float beta, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, + dependencies)); } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, double alpha, const double *a, int64_t lda, int64_t stride_a, - const double *b, int64_t ldb, int64_t stride_b, double beta, double *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, double alpha, const double* a, int64_t lda, int64_t stride_a, + const double* b, int64_t ldb, int64_t stride_b, double beta, double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, + dependencies)); } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, int64_t stride_a, const std::complex *b, int64_t ldb, - int64_t stride_b, std::complex beta, std::complex *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, int64_t stride_a, const std::complex* b, int64_t ldb, + int64_t stride_b, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, + dependencies)); } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, int64_t stride_a, const std::complex *b, int64_t ldb, - int64_t stride_b, std::complex beta, std::complex *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, int64_t stride_a, const std::complex* b, int64_t ldb, + int64_t stride_b, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); -} - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, sycl::half alpha, const sycl::half *a, int64_t lda, - int64_t stride_a, const sycl::half *b, int64_t ldb, int64_t stride_b, - sycl::half beta, sycl::half *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); -} - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const sycl::half *a, int64_t lda, int64_t stride_a, - const sycl::half *b, int64_t ldb, int64_t stride_b, float beta, float *c, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, + dependencies)); +} + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, sycl::half alpha, const sycl::half* a, int64_t lda, + int64_t stride_a, const sycl::half* b, int64_t ldb, int64_t stride_b, + sycl::half beta, sycl::half* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, + dependencies)); +} + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const sycl::half* a, int64_t lda, int64_t stride_a, + const sycl::half* b, int64_t ldb, int64_t stride_b, float beta, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, + dependencies)); } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const std::int8_t *a, int64_t lda, int64_t stride_a, - const std::int8_t *b, int64_t ldb, int64_t stride_b, float beta, float *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const std::int8_t* a, int64_t lda, int64_t stride_a, + const std::int8_t* b, int64_t ldb, int64_t stride_b, float beta, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", "unsupported dtype combination: int8_t, int8_t, float, float"); } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const std::int8_t *a, int64_t lda, int64_t stride_a, - const std::int8_t *b, int64_t ldb, int64_t stride_b, float beta, - std::int32_t *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, stride_a, b, ldb, - stride_b, beta, c, ldc, stride_c, batch_size, dependencies)); -} - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const float **a, int64_t *lda, - const float **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, group_size, dependencies)); -} - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, double *alpha, const double **a, int64_t *lda, - const double **b, int64_t *ldb, double *beta, double **c, int64_t *ldc, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, group_size, dependencies)); -} - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, std::complex *alpha, - const std::complex **a, int64_t *lda, const std::complex **b, - int64_t *ldb, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, group_size, dependencies)); -} - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, std::complex *alpha, - const std::complex **a, int64_t *lda, const std::complex **b, - int64_t *ldb, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, group_size, dependencies)); -} - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, sycl::half *alpha, const sycl::half **a, - int64_t *lda, const sycl::half **b, int64_t *ldb, sycl::half *beta, - sycl::half **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, groupsize, dependencies)); -} - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const sycl::half **a, int64_t *lda, - const sycl::half **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, groupsize, dependencies)); -} - -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const std::int8_t **a, int64_t *lda, - const std::int8_t **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const std::int8_t* a, int64_t lda, int64_t stride_a, + const std::int8_t* b, int64_t ldb, int64_t stride_b, float beta, + std::int32_t* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, + dependencies)); +} + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const float** a, int64_t* lda, + const float** b, int64_t* ldb, float* beta, float** c, int64_t* ldc, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies)); +} + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, double* alpha, const double** a, int64_t* lda, + const double** b, int64_t* ldb, double* beta, double** c, int64_t* ldc, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies)); +} + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, std::complex* alpha, + const std::complex** a, int64_t* lda, const std::complex** b, + int64_t* ldb, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies)); +} + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, std::complex* alpha, + const std::complex** a, int64_t* lda, const std::complex** b, + int64_t* ldb, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies)); +} + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, sycl::half* alpha, const sycl::half** a, + int64_t* lda, const sycl::half** b, int64_t* ldb, sycl::half* beta, + sycl::half** c, int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, group_count, groupsize, dependencies)); +} + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const sycl::half** a, int64_t* lda, + const sycl::half** b, int64_t* ldb, float* beta, float** c, int64_t* ldc, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, group_count, groupsize, dependencies)); +} + +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const std::int8_t** a, int64_t* lda, + const std::int8_t** b, int64_t* ldb, float* beta, float** c, int64_t* ldc, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", "unsupported dtype combination: int8_t, int8_t, float, float"); } -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const std::int8_t **a, int64_t *lda, - const std::int8_t **b, int64_t *ldb, float *beta, std::int32_t **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, - ldc, group_count, groupsize, dependencies)); +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const std::int8_t** a, int64_t* lda, + const std::int8_t** b, int64_t* ldb, float* beta, std::int32_t** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, group_count, groupsize, dependencies)); } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, - int64_t lda, int64_t stride_a, float *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, const float* a, + int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, - int64_t lda, int64_t stride_a, double *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, const double* a, + int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - std::complex *b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); -} - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, float *alpha, const float **a, - int64_t *lda, float **b, int64_t *ldb, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, b, ldb, group_count, groupsize, dependencies)); -} - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, double *alpha, const double **a, - int64_t *lda, double **b, int64_t *ldb, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, b, ldb, group_count, groupsize, dependencies)); -} - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, std::complex *alpha, - const std::complex **a, int64_t *lda, std::complex **b, - int64_t *ldb, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, b, ldb, group_count, groupsize, dependencies)); -} - -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, transpose *trans, - diag *unit_diag, int64_t *m, int64_t *n, std::complex *alpha, - const std::complex **a, int64_t *lda, std::complex **b, - int64_t *ldb, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, - lda, b, ldb, group_count, groupsize, dependencies)); -} - -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, const float *a, int64_t lda, int64_t stride_a, float beta, - float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, - ldc, stride_c, batch_size, dependencies)); -} - -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, const double *a, int64_t lda, int64_t stride_a, double beta, - double *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, - ldc, stride_c, batch_size, dependencies)); -} - -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex beta, std::complex *c, + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, stride_a, b, ldb, stride_b, batch_size, dependencies)); +} + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, float* alpha, const float** a, + int64_t* lda, float** b, int64_t* ldb, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, group_count, groupsize, dependencies)); +} + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, double* alpha, const double** a, + int64_t* lda, double** b, int64_t* ldb, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, group_count, groupsize, dependencies)); +} + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, std::complex* alpha, + const std::complex** a, int64_t* lda, std::complex** b, + int64_t* ldb, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, group_count, groupsize, dependencies)); +} + +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, std::complex* alpha, + const std::complex** a, int64_t* lda, std::complex** b, + int64_t* ldb, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm_batch( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, group_count, groupsize, dependencies)); +} + +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, const float* a, int64_t lda, int64_t stride_a, float beta, + float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies)); +} + +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, const double* a, int64_t lda, int64_t stride_a, double beta, + double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies)); +} + +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, - ldc, stride_c, batch_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies)); } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex beta, std::complex *c, +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex beta, std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, stride_a, beta, c, - ldc, stride_c, batch_size, dependencies)); -} - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, float *alpha, const float **a, int64_t *lda, float *beta, - float **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - group_count, groupsize, dependencies)); -} - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, double *alpha, const double **a, int64_t *lda, double *beta, - double **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - group_count, groupsize, dependencies)); -} - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - group_count, groupsize, dependencies)); -} - -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - group_count, groupsize, dependencies)); -} - -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies)); +} + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, float* alpha, const float** a, int64_t* lda, float* beta, + float** c, int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, beta, c, ldc, group_count, groupsize, dependencies)); +} + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, double* alpha, const double** a, int64_t* lda, double* beta, + double** c, int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, beta, c, ldc, group_count, groupsize, dependencies)); +} + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, beta, c, ldc, group_count, groupsize, dependencies)); +} + +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk_batch( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, beta, c, ldc, group_count, groupsize, dependencies)); +} + +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, + stride_a, b, ldb, stride_b, batch_size, dependencies)); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, + stride_a, b, ldb, stride_b, batch_size, dependencies)); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size, dependencies)); +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, + stride_a, b, ldb, stride_b, batch_size, dependencies)); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, stride_a, b, ldb, stride_b, - batch_size, dependencies)); +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, + stride_a, b, ldb, stride_b, batch_size, dependencies)); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, - dependencies)); +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, + ldb, stride, batch_size, dependencies)); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, - dependencies)); +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, + ldb, stride, batch_size, dependencies)); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, - dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, + ldb, stride, batch_size, dependencies)); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, stride, batch_size, - dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, + ldb, stride, batch_size, dependencies)); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, - float beta, const float *b, int64_t ldb, int64_t stride_b, float *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, float alpha, const float* a, int64_t lda, int64_t stride_a, + float beta, const float* b, int64_t ldb, int64_t stride_b, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, - ldb, stride_b, c, ldc, stride_c, batch_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, + dependencies)); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, double alpha, const double *a, int64_t lda, int64_t stride_a, - double beta, const double *b, int64_t ldb, int64_t stride_b, double *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, double alpha, const double* a, int64_t lda, int64_t stride_a, + double beta, const double* b, int64_t ldb, int64_t stride_b, double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, - ldb, stride_b, c, ldc, stride_c, batch_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, + dependencies)); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, - ldb, stride_b, c, ldc, stride_c, batch_size, dependencies)); + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, + dependencies)); } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, stride_a, beta, b, - ldb, stride_b, c, ldc, stride_c, batch_size, dependencies)); + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd_batch( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, + dependencies)); } sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, float* alpha, const float** a, int64_t* lda, float** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, - groupsize, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, + b, ldb, group_count, groupsize, dependencies)); } sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, double* alpha, const double** a, int64_t* lda, double** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, - groupsize, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, + b, ldb, group_count, groupsize, dependencies)); } sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, std::complex* alpha, const std::complex** a, int64_t* lda, std::complex** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, - groupsize, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, + b, ldb, group_count, groupsize, dependencies)); } sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, @@ -1035,38 +1162,43 @@ sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int int64_t* lda, std::complex** b, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, group_count, - groupsize, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::omatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, + b, ldb, group_count, groupsize, dependencies)); } sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, float* alpha, float** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, - groupsize, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, + ldb, group_count, groupsize, dependencies)); } sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, double* alpha, double** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, - groupsize, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, + ldb, group_count, groupsize, dependencies)); } sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, std::complex* alpha, std::complex** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, - groupsize, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, + ldb, group_count, groupsize, dependencies)); } sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, std::complex* alpha, std::complex** ab, int64_t* lda, int64_t* ldb, int64_t group_count, int64_t* groupsize, const std::vector& dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, group_count, - groupsize, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::imatcopy_batch(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, + ldb, group_count, groupsize, dependencies)); } diff --git a/src/blas/backends/mkl_common/mkl_blas_backend.hxx b/src/blas/backends/mkl_common/mkl_blas_backend.hxx index 82588a66a..51937dade 100644 --- a/src/blas/backends/mkl_common/mkl_blas_backend.hxx +++ b/src/blas/backends/mkl_common/mkl_blas_backend.hxx @@ -19,1351 +19,1365 @@ /// level3, buffer -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc); -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::half beta, - sycl::buffer &c, std::int64_t ldc); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::half alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::half beta, + sycl::buffer& c, std::int64_t ldc); -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc); -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, + std::int64_t ldc); -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, float beta, - sycl::buffer &c, std::int64_t ldc); - -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, double beta, - sycl::buffer &c, std::int64_t ldc); - -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); - -void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc); - -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc); - -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc); - -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); - -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); + +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc); + +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, float beta, + sycl::buffer& c, std::int64_t ldc); + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, double beta, + sycl::buffer& c, std::int64_t ldc); + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); + +void herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer, 1>& a, std::int64_t lda, float beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer, 1>& a, std::int64_t lda, double beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); + +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc); + +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, double beta, + sycl::buffer, 1>& c, std::int64_t ldc); + +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb); + +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb); +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb); -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); // level 3, USM -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, value_or_pointer beta, - float *c, std::int64_t ldc, const std::vector &dependencies = {}); +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, value_or_pointer alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, value_or_pointer beta, + float* c, std::int64_t ldc, const std::vector& dependencies = {}); -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, value_or_pointer beta, - double *c, std::int64_t ldc, const std::vector &dependencies = {}); +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, value_or_pointer alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, value_or_pointer beta, + double* c, std::int64_t ldc, const std::vector& dependencies = {}); -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, value_or_pointer> beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, value_or_pointer> beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, + const std::complex* a, std::int64_t lda, const std::complex* b, std::int64_t ldb, value_or_pointer> beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - value_or_pointer beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, value_or_pointer beta, - float *c, std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + value_or_pointer beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, value_or_pointer alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, + value_or_pointer beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, value_or_pointer alpha, const bfloat16* a, + std::int64_t lda, const bfloat16* b, std::int64_t ldb, + value_or_pointer beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, value_or_pointer alpha, const bfloat16* a, + std::int64_t lda, const bfloat16* b, std::int64_t ldb, + value_or_pointer beta, bfloat16* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const bfloat16 *a, std::int64_t lda, const bfloat16 *b, - std::int64_t ldb, value_or_pointer beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); + const std::int8_t* a, std::int64_t lda, const std::int8_t* b, std::int64_t ldb, + value_or_pointer beta, std::int32_t* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const bfloat16 *a, std::int64_t lda, const bfloat16 *b, - std::int64_t ldb, value_or_pointer beta, bfloat16 *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const std::int8_t *a, - std::int64_t lda, const std::int8_t *b, std::int64_t ldb, - value_or_pointer beta, std::int32_t *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const std::int8_t *a, - std::int64_t lda, const std::int8_t *b, std::int64_t ldb, - value_or_pointer beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, value_or_pointer beta, float *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, value_or_pointer beta, double *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, + const std::int8_t* a, std::int64_t lda, const std::int8_t* b, std::int64_t ldb, + value_or_pointer beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, value_or_pointer alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, value_or_pointer beta, float* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, value_or_pointer alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, value_or_pointer beta, double* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, value_or_pointer> beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, value_or_pointer> beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, + const std::complex* a, std::int64_t lda, const std::complex* b, std::int64_t ldb, value_or_pointer> beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, +sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, value_or_pointer> beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, value_or_pointer> beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, +sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, + const std::complex* a, std::int64_t lda, const std::complex* b, std::int64_t ldb, value_or_pointer> beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const float *a, std::int64_t lda, - value_or_pointer beta, float *c, std::int64_t ldc, - const std::vector &dependencies = {}); +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, value_or_pointer alpha, const float* a, std::int64_t lda, + value_or_pointer beta, float* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const double *a, std::int64_t lda, - value_or_pointer beta, double *c, std::int64_t ldc, - const std::vector &dependencies = {}); +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, value_or_pointer alpha, const double* a, std::int64_t lda, + value_or_pointer beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, - value_or_pointer> beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, + value_or_pointer> beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, - value_or_pointer> beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const std::complex *a, - std::int64_t lda, value_or_pointer beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const std::complex *a, - std::int64_t lda, value_or_pointer beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, value_or_pointer beta, float *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, value_or_pointer beta, double *c, - std::int64_t ldc, const std::vector &dependencies = {}); - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, + const std::complex* a, std::int64_t lda, + value_or_pointer> beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, value_or_pointer alpha, const std::complex* a, + std::int64_t lda, value_or_pointer beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, value_or_pointer alpha, const std::complex* a, + std::int64_t lda, value_or_pointer beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, value_or_pointer alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, value_or_pointer beta, float* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, value_or_pointer alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, value_or_pointer beta, double* c, + std::int64_t ldc, const std::vector& dependencies = {}); + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, value_or_pointer> beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, value_or_pointer> beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, + const std::complex* a, std::int64_t lda, const std::complex* b, std::int64_t ldb, value_or_pointer> beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, +sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, value_or_pointer beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, value_or_pointer beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); -sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, +sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, value_or_pointer beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, value_or_pointer beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies = {}); -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies = {}); + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies = {}); -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies = {}); + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies = {}); -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies = {}); + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies = {}); -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies = {}); + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies = {}); -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); // level 2, buffer -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy); -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy); +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); -void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, +void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, sycl::buffer& a, std::int64_t lda); -void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda); - -void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -void hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -void hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -void hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -void hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - -void her(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda); - -void her(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda); - -void her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -void her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - -void hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, +void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda); + +void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +void hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +void hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +void hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); + +void her(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda); + +void her(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda); + +void her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + +void hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); -void hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, +void hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); -void hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a); +void hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a); -void hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a); +void hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a); -void hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a); +void hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a); -void hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a); +void hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a); -void sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); +void sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy); -void sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); +void sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); -void symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); +void symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy); -void symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); +void symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); -void syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, std::int64_t lda); +void syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda); -void syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, +void syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda); -void syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda); +void syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda); -void syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda); +void syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda); -void spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); +void spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); -void spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); +void spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); -void spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a); +void spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a); -void spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a); +void spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a); -void spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a); +void spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a); -void spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a); +void spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx); +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx); +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, std::int64_t incx); -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, std::int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx); +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx); +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, std::int64_t incx); -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, std::int64_t incx); -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx); -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); // level 2, USM -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, value_or_pointer beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, value_or_pointer beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - value_or_pointer> beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - value_or_pointer> beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, value_or_pointer alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, value_or_pointer beta, - float *y, std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, value_or_pointer alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, value_or_pointer beta, - double *y, std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, value_or_pointer beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, value_or_pointer beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + value_or_pointer> beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + value_or_pointer> beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, value_or_pointer alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, value_or_pointer beta, + float* y, std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, value_or_pointer alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, + value_or_pointer beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, value_or_pointer> beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, value_or_pointer> beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, + const std::complex* a, std::int64_t lda, const std::complex* x, std::int64_t incx, value_or_pointer> beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - value_or_pointer> beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - value_or_pointer> beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - value_or_pointer> beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hemv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - value_or_pointer> beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event her(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event her(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event her2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, - value_or_pointer> beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event hpmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, - value_or_pointer> beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - const std::vector &dependencies = {}); - -sycl::event hpr(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - const std::vector &dependencies = {}); - -sycl::event hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *a, const std::vector &dependencies = {}); - -sycl::event hpr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *a, const std::vector &dependencies = {}); - -sycl::event sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - value_or_pointer alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, value_or_pointer beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event sbmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, std::int64_t k, - value_or_pointer alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, value_or_pointer beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const float *a, std::int64_t lda, const float *x, std::int64_t incx, - value_or_pointer beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event symv(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const double *a, std::int64_t lda, const double *x, std::int64_t incx, - value_or_pointer beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); - -sycl::event syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const float *x, std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event syr(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const double *x, std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies = {}); - -sycl::event syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event syr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, - std::int64_t lda, const std::vector &dependencies = {}); - -sycl::event spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const float *a, const float *x, std::int64_t incx, value_or_pointer beta, - float *y, std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event spmv(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const double *a, const double *x, std::int64_t incx, value_or_pointer beta, - double *y, std::int64_t incy, const std::vector &dependencies = {}); - -sycl::event spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const float *x, std::int64_t incx, float *a, - const std::vector &dependencies = {}); - -sycl::event spr(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const double *x, std::int64_t incx, double *a, - const std::vector &dependencies = {}); - -sycl::event spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, float *a, - const std::vector &dependencies = {}); - -sycl::event spr2(sycl::queue &queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const double *x, std::int64_t incx, const double *y, std::int64_t incy, double *a, - const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float *a, std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double *a, std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); - -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies = {}); + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, value_or_pointer alpha, + const float* x, std::int64_t incx, const float* y, std::int64_t incy, float* a, + std::int64_t lda, const std::vector& dependencies = {}); + +sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, value_or_pointer alpha, + const double* x, std::int64_t incx, const double* y, std::int64_t incy, double* a, + std::int64_t lda, const std::vector& dependencies = {}); + +sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + value_or_pointer> beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + value_or_pointer> beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + value_or_pointer> beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + value_or_pointer> beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event her(sycl::queue& queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + std::int64_t lda, const std::vector& dependencies = {}); + +sycl::event her(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer> alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer> alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, + value_or_pointer> beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, + value_or_pointer> beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); + +sycl::event hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + const std::vector& dependencies = {}); + +sycl::event hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies = {}); + +sycl::event hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer> alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, const std::vector& dependencies = {}); + +sycl::event hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer> alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* a, const std::vector& dependencies = {}); + +sycl::event sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + value_or_pointer alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, value_or_pointer beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, + value_or_pointer alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, value_or_pointer beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, value_or_pointer beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, value_or_pointer beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, + const float* x, std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const double* x, std::int64_t incx, double* a, + std::int64_t lda, const std::vector& dependencies = {}); + +sycl::event syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const float* x, std::int64_t incx, const float* y, + std::int64_t incy, float* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const float* a, const float* x, std::int64_t incx, + value_or_pointer beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const double* a, const double* x, + std::int64_t incx, value_or_pointer beta, double* y, std::int64_t incy, + const std::vector& dependencies = {}); + +sycl::event spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, + const float* x, std::int64_t incx, float* a, + const std::vector& dependencies = {}); + +sycl::event spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const double* x, std::int64_t incx, double* a, + const std::vector& dependencies = {}); + +sycl::event spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const float* x, std::int64_t incx, const float* y, + std::int64_t incy, float* a, const std::vector& dependencies = {}); + +sycl::event spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, + value_or_pointer alpha, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* a, + const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); + +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies = {}); // level 1, buffer -void dotc(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); +void dotc(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); -void dotc(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); +void dotc(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); -void dotu(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); +void dotu(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); -void dotu(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); +void dotu(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result, index_base base=index_base::zero); +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result, index_base base = index_base::zero); -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result, index_base base=index_base::zero); +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result, index_base base = index_base::zero); -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result, - index_base base=index_base::zero); +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result, + index_base base = index_base::zero); -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result, - index_base base=index_base::zero); +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result, + index_base base = index_base::zero); -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result, index_base base=index_base::zero); +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result, index_base base = index_base::zero); -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result, index_base base=index_base::zero); +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result, index_base base = index_base::zero); -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result, - index_base base=index_base::zero); +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result, + index_base base = index_base::zero); -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result, - index_base base=index_base::zero); +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result, + index_base base = index_base::zero); -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void axpy(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); +void axpy(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); -void axpy(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); +void axpy(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); -void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); +void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); -void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); +void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); -void axpby(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy); +void axpby(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy); -void axpby(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy); +void axpby(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); -void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); +void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); -void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); +void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy); -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); -void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &result); +void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& result); -void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &result); +void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& result); -void sdsdot(sycl::queue &queue, std::int64_t n, float sb, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); +void sdsdot(sycl::queue& queue, std::int64_t n, float sb, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); -void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &result); +void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& result); -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result); +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result); -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result); +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result); -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, float c, float s); -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, double c, double s); -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, float c, float s); +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, float c, float s); -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, double c, double s); +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, double c, double s); -void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s); +void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s); -void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s); +void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s); -void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s); +void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); -void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s); +void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); -void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m); +void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& param); -void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m); +void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& param); -void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, float y1, sycl::buffer ¶m); +void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, float y1, sycl::buffer& param); -void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, double y1, sycl::buffer ¶m); +void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, double y1, sycl::buffer& param); -void scal(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, +void scal(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx); -void scal(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, +void scal(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, std::int64_t incx); -void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx); +void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx); -void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx); +void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx); -void scal(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer, 1> &x, +void scal(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer, 1>& x, std::int64_t incx); -void scal(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx); +void scal(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx); -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy); +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); // level 1, USM -sycl::event dotc(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, const std::vector &dependencies = {}); +sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies = {}); -sycl::event dotc(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, const std::vector &dependencies = {}); +sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies = {}); -sycl::event dotu(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, const std::vector &dependencies = {}); +sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies = {}); -sycl::event dotu(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, const std::vector &dependencies = {}); +sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); -sycl::event iamax(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); +sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); -sycl::event iamin(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies = {}); +sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies = {}); -sycl::event asum(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, float *result, - const std::vector &dependencies = {}); +sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}); -sycl::event asum(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, double *result, - const std::vector &dependencies = {}); +sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}); -sycl::event asum(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - float *result, const std::vector &dependencies = {}); +sycl::event asum(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + float* result, const std::vector& dependencies = {}); -sycl::event asum(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - double *result, const std::vector &dependencies = {}); +sycl::event asum(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + double* result, const std::vector& dependencies = {}); -sycl::event axpy(sycl::queue &queue, std::int64_t n, value_or_pointer alpha, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies = {}); +sycl::event axpy(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies = {}); -sycl::event axpy(sycl::queue &queue, std::int64_t n, value_or_pointer alpha, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies = {}); +sycl::event axpy(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, + const double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies = {}); -sycl::event axpy(sycl::queue &queue, std::int64_t n, value_or_pointer> alpha, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); +sycl::event axpy(sycl::queue& queue, std::int64_t n, value_or_pointer> alpha, + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); -sycl::event axpy(sycl::queue &queue, std::int64_t n, value_or_pointer> alpha, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); +sycl::event axpy(sycl::queue& queue, std::int64_t n, value_or_pointer> alpha, + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); -sycl::event axpby(sycl::queue &queue, std::int64_t n, value_or_pointer alpha, const float *x, - std::int64_t incx, value_or_pointer beta, float *y, std::int64_t incy, - const std::vector &dependencies = {}); +sycl::event axpby(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, const float* x, + std::int64_t incx, value_or_pointer beta, float* y, std::int64_t incy, + const std::vector& dependencies = {}); -sycl::event axpby(sycl::queue &queue, std::int64_t n, value_or_pointer alpha, const double *x, - std::int64_t incx, value_or_pointer beta, double *y, std::int64_t incy, - const std::vector &dependencies = {}); +sycl::event axpby(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, + const double* x, std::int64_t incx, value_or_pointer beta, double* y, + std::int64_t incy, const std::vector& dependencies = {}); -sycl::event axpby(sycl::queue &queue, std::int64_t n, value_or_pointer> alpha, - const std::complex *x, std::int64_t incx, - value_or_pointer> beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); +sycl::event axpby(sycl::queue& queue, std::int64_t n, value_or_pointer> alpha, + const std::complex* x, std::int64_t incx, + value_or_pointer> beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); -sycl::event axpby(sycl::queue &queue, std::int64_t n, value_or_pointer> alpha, - const std::complex *x, std::int64_t incx, - value_or_pointer> beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies = {}); +sycl::event axpby(sycl::queue& queue, std::int64_t n, value_or_pointer> alpha, + const std::complex* x, std::int64_t incx, + value_or_pointer> beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies = {}); -sycl::event copy(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *y, - std::int64_t incy, const std::vector &dependencies = {}); +sycl::event copy(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, float* y, + std::int64_t incy, const std::vector& dependencies = {}); -sycl::event copy(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *y, - std::int64_t incy, const std::vector &dependencies = {}); +sycl::event copy(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, double* y, + std::int64_t incy, const std::vector& dependencies = {}); -sycl::event copy(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); +sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); -sycl::event copy(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); +sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); -sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *result, - const std::vector &dependencies = {}); +sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* result, + const std::vector& dependencies = {}); -sycl::event dot(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *result, - const std::vector &dependencies = {}); +sycl::event dot(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* result, + const std::vector& dependencies = {}); -sycl::event sdsdot(sycl::queue &queue, std::int64_t n, float sb, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *result, - const std::vector &dependencies = {}); +sycl::event sdsdot(sycl::queue& queue, std::int64_t n, float sb, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* result, + const std::vector& dependencies = {}); -sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, double *result, - const std::vector &dependencies = {}); +sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, double* result, + const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, float *result, - const std::vector &dependencies = {}); +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, float* result, + const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, double *result, - const std::vector &dependencies = {}); +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, double* result, + const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - float *result, const std::vector &dependencies = {}); +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + float* result, const std::vector& dependencies = {}); -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - double *result, const std::vector &dependencies = {}); +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + double* result, const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, value_or_pointer c, - value_or_pointer s, const std::vector &dependencies = {}); +sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, value_or_pointer c, + value_or_pointer s, const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, value_or_pointer c, - value_or_pointer s, const std::vector &dependencies = {}); +sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, value_or_pointer c, + value_or_pointer s, const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, +sycl::event rot(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, std::int64_t incy, value_or_pointer c, value_or_pointer s, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event rot(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, +sycl::event rot(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, std::int64_t incy, value_or_pointer c, value_or_pointer s, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, float *a, float *b, float *c, float *s, - const std::vector &dependencies = {}); +sycl::event rotg(sycl::queue& queue, float* a, float* b, float* c, float* s, + const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, double *a, double *b, double *c, double *s, - const std::vector &dependencies = {}); +sycl::event rotg(sycl::queue& queue, double* a, double* b, double* c, double* s, + const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, float *c, - std::complex *s, const std::vector &dependencies = {}); +sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, float* c, + std::complex* s, const std::vector& dependencies = {}); -sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, double *c, - std::complex *s, const std::vector &dependencies = {}); +sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, double* c, + std::complex* s, const std::vector& dependencies = {}); -sycl::event rotm(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, - std::int64_t incy, const float *param, - const std::vector &dependencies = {}); +sycl::event rotm(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, + std::int64_t incy, const float* param, + const std::vector& dependencies = {}); -sycl::event rotm(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, - std::int64_t incy, const double *param, - const std::vector &dependencies = {}); +sycl::event rotm(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, + std::int64_t incy, const double* param, + const std::vector& dependencies = {}); -sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, value_or_pointer y1, - float *param, const std::vector &dependencies = {}); +sycl::event rotmg(sycl::queue& queue, float* d1, float* d2, float* x1, value_or_pointer y1, + float* param, const std::vector& dependencies = {}); -sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, value_or_pointer y1, - double *param, const std::vector &dependencies = {}); +sycl::event rotmg(sycl::queue& queue, double* d1, double* d2, double* x1, + value_or_pointer y1, double* param, + const std::vector& dependencies = {}); -#define ONEMATH_DECLARE_SCAL(T, Ts) \ - sycl::event scal(sycl::queue &queue, std::int64_t n, value_or_pointer alpha, T *x, \ - std::int64_t incx, const std::vector &dependencies = {}); +#define ONEMATH_DECLARE_SCAL(T, Ts) \ + sycl::event scal(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, T* x, \ + std::int64_t incx, const std::vector& dependencies = {}); ONEMATH_DECLARE_SCAL(float, float) ONEMATH_DECLARE_SCAL(double, double) @@ -1371,1124 +1385,1134 @@ ONEMATH_DECLARE_SCAL(std::complex, std::complex) ONEMATH_DECLARE_SCAL(std::complex, std::complex) ONEMATH_DECLARE_SCAL(std::complex, float) ONEMATH_DECLARE_SCAL(std::complex, double) -sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); -sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, std::complex *x, - std::int64_t incx, const std::vector &dependencies = {}); +sycl::event scal(sycl::queue& queue, std::int64_t n, float alpha, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); +sycl::event scal(sycl::queue& queue, std::int64_t n, double alpha, std::complex* x, + std::int64_t incx, const std::vector& dependencies = {}); #undef ONEMATH_DECLARE_SCAL -sycl::event swap(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, - std::int64_t incy, const std::vector &dependencies = {}); +sycl::event swap(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, + std::int64_t incy, const std::vector& dependencies = {}); -sycl::event swap(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, - std::int64_t incy, const std::vector &dependencies = {}); +sycl::event swap(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, + std::int64_t incy, const std::vector& dependencies = {}); -sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); +sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); -sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies = {}); +sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies = {}); // extensions, buffer -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc); -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc); +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int8_t ao, - sycl::buffer &b, std::int64_t ldb, std::uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, std::int8_t ao, + sycl::buffer& b, std::int64_t ldb, std::uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int8_t ao, - sycl::buffer &b, std::int64_t ldb, std::int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, std::int8_t ao, + sycl::buffer& b, std::int64_t ldb, std::int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::uint8_t ao, - sycl::buffer &b, std::int64_t ldb, std::int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, std::uint8_t ao, + sycl::buffer& b, std::int64_t ldb, std::int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::uint8_t ao, - sycl::buffer &b, std::int64_t ldb, std::uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, std::uint8_t ao, + sycl::buffer& b, std::int64_t ldb, std::uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); // extensions, USM -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, value_or_pointer beta, - float *c, std::int64_t ldc, const std::vector &dependencies = {}); +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + std::int64_t n, std::int64_t k, value_or_pointer alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, value_or_pointer beta, + float* c, std::int64_t ldc, const std::vector& dependencies = {}); -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, value_or_pointer beta, - double *c, std::int64_t ldc, const std::vector &dependencies = {}); +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + std::int64_t n, std::int64_t k, value_or_pointer alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, + value_or_pointer beta, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, value_or_pointer> beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, value_or_pointer> beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, + const std::complex* a, std::int64_t lda, const std::complex* b, std::int64_t ldb, value_or_pointer> beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::uint8_t *b, - std::int64_t ldb, std::uint8_t bo, value_or_pointer beta, std::int32_t *c, - std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies = {}); + const std::int8_t* a, std::int64_t lda, std::int8_t ao, const std::uint8_t* b, + std::int64_t ldb, std::uint8_t bo, value_or_pointer beta, + std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies = {}); -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::int8_t *b, - std::int64_t ldb, std::int8_t bo, value_or_pointer beta, std::int32_t *c, - std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies = {}); + const std::int8_t* a, std::int64_t lda, std::int8_t ao, const std::int8_t* b, + std::int64_t ldb, std::int8_t bo, value_or_pointer beta, + std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies = {}); -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, - const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - value_or_pointer beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies = {}); + const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, + const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + value_or_pointer beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies = {}); -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, - const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - value_or_pointer beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies = {}); + const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, + const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + value_or_pointer beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies = {}); // batch, buffer -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, float beta, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, float beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, std::complex beta, sycl::buffer, 1> &c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, std::complex beta, sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, std::int64_t incy, +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, std::int64_t incy, +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, float beta, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, double beta, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, +void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, +void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -void dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -void axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer &y, +void axpy_batch(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer &y, +void axpy_batch(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, +void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, +void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer &c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer &c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, +void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -void omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, +void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -void omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, +void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -void omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, +void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, +void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &ab, std::int64_t lda, std::int64_t ldb, +void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -void imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, float beta, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, double beta, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); -void omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, +void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb); -void omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, +void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb); -void omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); +void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); -void omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb); +void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb); -void omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); +void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); -void omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &b, std::int64_t ldb, std::int64_t strideb); +void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); -void omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, std::int64_t ldb, +void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb); -void omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, +void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb); -void imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); +void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); -void imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &ab, std::int64_t lda, std::int64_t ldb); +void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); -void imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, std::int64_t lda, +void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb); -void imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, float beta, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, +void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, float beta, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& c, std::int64_t ldc); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, double beta, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, +void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, double beta, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& c, std::int64_t ldc); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); -void omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); // batch, usm -sycl::event syrk_batch(sycl::queue &queue, const uplo *upper_lower, const transpose *trans, - const std::int64_t *n, const std::int64_t *k, const float *alpha, - const float **a, const std::int64_t *lda, const float *beta, float **c, - const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); -sycl::event syrk_batch(sycl::queue &queue, const uplo *upper_lower, const transpose *trans, - const std::int64_t *n, const std::int64_t *k, const double *alpha, - const double **a, const std::int64_t *lda, const double *beta, double **c, - const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); -sycl::event syrk_batch(sycl::queue &queue, const uplo *upper_lower, const transpose *trans, - const std::int64_t *n, const std::int64_t *k, - const std::complex *alpha, const std::complex **a, - const std::int64_t *lda, const std::complex *beta, - std::complex **c, const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); -sycl::event syrk_batch(sycl::queue &queue, const uplo *upper_lower, const transpose *trans, - const std::int64_t *n, const std::int64_t *k, - const std::complex *alpha, const std::complex **a, - const std::int64_t *lda, const std::complex *beta, - std::complex **c, const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, value_or_pointer beta, float *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const double *a, +sycl::event syrk_batch(sycl::queue& queue, const uplo* upper_lower, const transpose* trans, + const std::int64_t* n, const std::int64_t* k, const float* alpha, + const float** a, const std::int64_t* lda, const float* beta, float** c, + const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); +sycl::event syrk_batch(sycl::queue& queue, const uplo* upper_lower, const transpose* trans, + const std::int64_t* n, const std::int64_t* k, const double* alpha, + const double** a, const std::int64_t* lda, const double* beta, double** c, + const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); +sycl::event syrk_batch(sycl::queue& queue, const uplo* upper_lower, const transpose* trans, + const std::int64_t* n, const std::int64_t* k, + const std::complex* alpha, const std::complex** a, + const std::int64_t* lda, const std::complex* beta, + std::complex** c, const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); +sycl::event syrk_batch(sycl::queue& queue, const uplo* upper_lower, const transpose* trans, + const std::int64_t* n, const std::int64_t* k, + const std::complex* alpha, const std::complex** a, + const std::int64_t* lda, const std::complex* beta, + std::complex** c, const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, value_or_pointer alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, value_or_pointer beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, value_or_pointer alpha, const double* a, std::int64_t lda, std::int64_t stride_a, value_or_pointer beta, - double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - value_or_pointer> beta, std::complex *c, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + value_or_pointer> beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, + const std::vector& dependencies = {}); +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - value_or_pointer> beta, std::complex *c, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + value_or_pointer> beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies = {}); +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, const std::int64_t *n, const float **x, - const std::int64_t *incx, float **y, const std::int64_t *incy, - std::int64_t group_count, const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, const std::int64_t *n, const double **x, - const std::int64_t *incx, double **y, const std::int64_t *incy, - std::int64_t group_count, const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, const std::int64_t *n, const std::complex **x, - const std::int64_t *incx, std::complex **y, const std::int64_t *incy, - std::int64_t group_count, const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event copy_batch(sycl::queue &queue, const std::int64_t *n, const std::complex **x, - const std::int64_t *incx, std::complex **y, const std::int64_t *incy, - std::int64_t group_count, const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - const float *a, std::int64_t lda, std::int64_t stridea, const float *x, - std::int64_t incx, std::int64_t stridex, float *c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event copy_batch(sycl::queue& queue, const std::int64_t* n, const float** x, + const std::int64_t* incx, float** y, const std::int64_t* incy, + std::int64_t group_count, const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event copy_batch(sycl::queue& queue, const std::int64_t* n, const double** x, + const std::int64_t* incx, double** y, const std::int64_t* incy, + std::int64_t group_count, const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event copy_batch(sycl::queue& queue, const std::int64_t* n, const std::complex** x, + const std::int64_t* incx, std::complex** y, const std::int64_t* incy, + std::int64_t group_count, const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event copy_batch(sycl::queue& queue, const std::int64_t* n, const std::complex** x, + const std::int64_t* incx, std::complex** y, const std::int64_t* incy, + std::int64_t group_count, const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + const float* a, std::int64_t lda, std::int64_t stridea, const float* x, + std::int64_t incx, std::int64_t stridex, float* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - const double *a, std::int64_t lda, std::int64_t stridea, const double *x, - std::int64_t incx, std::int64_t stridex, double *c, std::int64_t ldc, +sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + const double* a, std::int64_t lda, std::int64_t stridea, const double* x, + std::int64_t incx, std::int64_t stridex, double* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, const side *left_right, const std::int64_t *m, - const std::int64_t *n, const float **a, const std::int64_t *lda, - const float **x, const std::int64_t *incx, float **c, - const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, const side *left_right, const std::int64_t *m, - const std::int64_t *n, const double **a, const std::int64_t *lda, - const double **x, const std::int64_t *incx, double **c, - const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, const side *left_right, const std::int64_t *m, - const std::int64_t *n, const std::complex **a, - const std::int64_t *lda, const std::complex **x, - const std::int64_t *incx, std::complex **c, const std::int64_t *ldc, - std::int64_t group_count, const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event dgmm_batch(sycl::queue &queue, const side *left_right, const std::int64_t *m, - const std::int64_t *n, const std::complex **a, - const std::int64_t *lda, const std::complex **x, - const std::int64_t *incx, std::complex **c, const std::int64_t *ldc, - std::int64_t group_count, const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, value_or_pointer beta, float *y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, value_or_pointer beta, double *y, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, const side* left_right, const std::int64_t* m, + const std::int64_t* n, const float** a, const std::int64_t* lda, + const float** x, const std::int64_t* incx, float** c, + const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, const side* left_right, const std::int64_t* m, + const std::int64_t* n, const double** a, const std::int64_t* lda, + const double** x, const std::int64_t* incx, double** c, + const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, const side* left_right, const std::int64_t* m, + const std::int64_t* n, const std::complex** a, + const std::int64_t* lda, const std::complex** x, + const std::int64_t* incx, std::complex** c, const std::int64_t* ldc, + std::int64_t group_count, const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event dgmm_batch(sycl::queue& queue, const side* left_right, const std::int64_t* m, + const std::int64_t* n, const std::complex** a, + const std::int64_t* lda, const std::complex** x, + const std::int64_t* incx, std::complex** c, const std::int64_t* ldc, + std::int64_t group_count, const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, const float* a, std::int64_t lda, + std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, value_or_pointer beta, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, +sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, const double* a, std::int64_t lda, + std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, value_or_pointer beta, double* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, std::int64_t incx, std::int64_t stridex, - value_or_pointer> beta, std::complex *y, + value_or_pointer> beta, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemv_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, +sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, std::int64_t incx, std::int64_t stridex, - value_or_pointer> beta, std::complex *y, + value_or_pointer> beta, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const float *alpha, const float **a, - const std::int64_t *lda, const float **x, const std::int64_t *incx, - const float *beta, float **y, const std::int64_t *incy, - std::int64_t group_count, const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const double *alpha, const double **a, - const std::int64_t *lda, const double **x, const std::int64_t *incx, - const double *beta, double **y, const std::int64_t *incy, - std::int64_t group_count, const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const std::complex *alpha, - const std::complex **a, const std::int64_t *lda, - const std::complex **x, const std::int64_t *incx, - const std::complex *beta, std::complex **y, - const std::int64_t *incy, std::int64_t group_count, - const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event gemv_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const std::complex *alpha, - const std::complex **a, const std::int64_t *lda, - const std::complex **x, const std::int64_t *incx, - const std::complex *beta, std::complex **y, - const std::int64_t *incy, std::int64_t group_count, - const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, const std::int64_t *n, const double *alpha, - const double **x, const std::int64_t *incx, double **y, - const std::int64_t *incy, std::int64_t group_count, - const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, const std::int64_t *n, const float *alpha, - const float **x, const std::int64_t *incx, float **y, - const std::int64_t *incy, std::int64_t group_count, - const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, const std::int64_t *n, const std::complex *alpha, - const std::complex **x, const std::int64_t *incx, - std::complex **y, const std::int64_t *incy, std::int64_t group_count, - const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, const std::int64_t *n, const std::complex *alpha, - const std::complex **x, const std::int64_t *incx, - std::complex **y, const std::int64_t *incy, std::int64_t group_count, - const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, value_or_pointer alpha, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const float* alpha, const float** a, + const std::int64_t* lda, const float** x, const std::int64_t* incx, + const float* beta, float** y, const std::int64_t* incy, + std::int64_t group_count, const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const double* alpha, const double** a, + const std::int64_t* lda, const double** x, const std::int64_t* incx, + const double* beta, double** y, const std::int64_t* incy, + std::int64_t group_count, const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const std::complex* alpha, + const std::complex** a, const std::int64_t* lda, + const std::complex** x, const std::int64_t* incx, + const std::complex* beta, std::complex** y, + const std::int64_t* incy, std::int64_t group_count, + const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event gemv_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const std::complex* alpha, + const std::complex** a, const std::int64_t* lda, + const std::complex** x, const std::int64_t* incx, + const std::complex* beta, std::complex** y, + const std::int64_t* incy, std::int64_t group_count, + const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, const std::int64_t* n, const double* alpha, + const double** x, const std::int64_t* incx, double** y, + const std::int64_t* incy, std::int64_t group_count, + const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, const std::int64_t* n, const float* alpha, + const float** x, const std::int64_t* incx, float** y, + const std::int64_t* incy, std::int64_t group_count, + const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, const std::int64_t* n, const std::complex* alpha, + const std::complex** x, const std::int64_t* incx, + std::complex** y, const std::int64_t* incy, std::int64_t group_count, + const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, const std::int64_t* n, const std::complex* alpha, + const std::complex** x, const std::int64_t* incx, + std::complex** y, const std::int64_t* incy, std::int64_t group_count, + const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, + const float* x, std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, value_or_pointer alpha, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, + const double* x, std::int64_t incx, std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, value_or_pointer> alpha, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies = {}); - -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, - value_or_pointer> alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, + value_or_pointer> alpha, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, const transpose *transa, const transpose *transb, - const std::int64_t *m, const std::int64_t *n, const std::int64_t *k, - const float *alpha, const float **a, const std::int64_t *lda, - const float **b, const std::int64_t *ldb, const float *beta, float **c, - const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, const transpose *transa, const transpose *transb, - const std::int64_t *m, const std::int64_t *n, const std::int64_t *k, - const double *alpha, const double **a, const std::int64_t *lda, - const double **b, const std::int64_t *ldb, const double *beta, double **c, - const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, const transpose *transa, const transpose *transb, - const std::int64_t *m, const std::int64_t *n, const std::int64_t *k, - const std::complex *alpha, const std::complex **a, - const std::int64_t *lda, const std::complex **b, - const std::int64_t *ldb, const std::complex *beta, - std::complex **c, const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, const transpose *transa, const transpose *transb, - const std::int64_t *m, const std::int64_t *n, const std::int64_t *k, - const std::complex *alpha, const std::complex **a, - const std::int64_t *lda, const std::complex **b, - const std::int64_t *ldb, const std::complex *beta, - std::complex **c, const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, const transpose *transa, const transpose *transb, - const std::int64_t *m, const std::int64_t *n, const std::int64_t *k, - const sycl::half *alpha, const sycl::half **a, const std::int64_t *lda, - const sycl::half **b, const std::int64_t *ldb, const sycl::half *beta, - sycl::half **c, const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, const transpose *transa, const transpose *transb, - const std::int64_t *m, const std::int64_t *n, const std::int64_t *k, - const float *alpha, const sycl::half **a, const std::int64_t *lda, - const sycl::half **b, const std::int64_t *ldb, const float *beta, float **c, - const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, const transpose *transa, const transpose *transb, - const std::int64_t *m, const std::int64_t *n, const std::int64_t *k, - const float *alpha, const bfloat16 **a, const std::int64_t *lda, - const bfloat16 **b, const std::int64_t *ldb, const float *beta, - bfloat16 **c, const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, const transpose *transa, const transpose *transb, - const std::int64_t *m, const std::int64_t *n, const std::int64_t *k, - const float *alpha, const bfloat16 **a, const std::int64_t *lda, - const bfloat16 **b, const std::int64_t *ldb, const float *beta, - float **c, const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, const transpose *transa, const transpose *transb, - const std::int64_t *m, const std::int64_t *n, const std::int64_t *k, - const float *alpha, const std::int8_t **a, const std::int64_t *lda, - const std::int8_t **b, const std::int64_t *ldb, const float *beta, - std::int32_t **c, const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, const transpose *transa, const transpose *transb, - const std::int64_t *m, const std::int64_t *n, const std::int64_t *k, - const float *alpha, const std::int8_t **a, const std::int64_t *lda, - const std::int8_t **b, const std::int64_t *ldb, const float *beta, float **c, - const std::int64_t *ldc, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, const float *b, std::int64_t ldb, - std::int64_t stride_b, value_or_pointer beta, float *c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, const double *b, std::int64_t ldb, - std::int64_t stride_b, value_or_pointer beta, double *c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, + value_or_pointer> alpha, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* y, + std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, + const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, + const float* alpha, const float** a, const std::int64_t* lda, + const float** b, const std::int64_t* ldb, const float* beta, float** c, + const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, + const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, + const double* alpha, const double** a, const std::int64_t* lda, + const double** b, const std::int64_t* ldb, const double* beta, double** c, + const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, + const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, + const std::complex* alpha, const std::complex** a, + const std::int64_t* lda, const std::complex** b, + const std::int64_t* ldb, const std::complex* beta, + std::complex** c, const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, + const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, + const std::complex* alpha, const std::complex** a, + const std::int64_t* lda, const std::complex** b, + const std::int64_t* ldb, const std::complex* beta, + std::complex** c, const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, + const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, + const sycl::half* alpha, const sycl::half** a, const std::int64_t* lda, + const sycl::half** b, const std::int64_t* ldb, const sycl::half* beta, + sycl::half** c, const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, + const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, + const float* alpha, const sycl::half** a, const std::int64_t* lda, + const sycl::half** b, const std::int64_t* ldb, const float* beta, float** c, + const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, + const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, + const float* alpha, const bfloat16** a, const std::int64_t* lda, + const bfloat16** b, const std::int64_t* ldb, const float* beta, bfloat16** c, + const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, + const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, + const float* alpha, const bfloat16** a, const std::int64_t* lda, + const bfloat16** b, const std::int64_t* ldb, const float* beta, float** c, + const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, + const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, + const float* alpha, const std::int8_t** a, const std::int64_t* lda, + const std::int8_t** b, const std::int64_t* ldb, const float* beta, + std::int32_t** c, const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, + const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, + const float* alpha, const std::int8_t** a, const std::int64_t* lda, + const std::int8_t** b, const std::int64_t* ldb, const float* beta, float** c, + const std::int64_t* ldc, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, value_or_pointer alpha, + const float* a, std::int64_t lda, std::int64_t stride_a, const float* b, + std::int64_t ldb, std::int64_t stride_b, value_or_pointer beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, value_or_pointer alpha, + const double* a, std::int64_t lda, std::int64_t stride_a, const double* b, + std::int64_t ldb, std::int64_t stride_b, value_or_pointer beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer> beta, std::complex *c, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + value_or_pointer> beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer> beta, std::complex *c, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + value_or_pointer> beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, sycl::half *c, std::int64_t ldc, + const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, + value_or_pointer beta, sycl::half* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}); + const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, + value_or_pointer beta, float* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const bfloat16 *a, std::int64_t lda, std::int64_t stride_a, - const bfloat16 *b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, bfloat16 *c, std::int64_t ldc, + const bfloat16* a, std::int64_t lda, std::int64_t stride_a, + const bfloat16* b, std::int64_t ldb, std::int64_t stride_b, + value_or_pointer beta, bfloat16* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const bfloat16 *a, std::int64_t lda, std::int64_t stride_a, - const bfloat16 *b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}); + const bfloat16* a, std::int64_t lda, std::int64_t stride_a, + const bfloat16* b, std::int64_t ldb, std::int64_t stride_b, + value_or_pointer beta, float* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, std::int32_t *c, std::int64_t ldc, + const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, + value_or_pointer beta, std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float *b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, + value_or_pointer beta, float* c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies = {}); -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double *b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, + value_or_pointer alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies = {}); -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, + value_or_pointer alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, const side *left_right, const uplo *upper_lower, - const transpose *trans, const diag *unit_diag, const std::int64_t *m, - const std::int64_t *n, const float *alpha, const float **a, - const std::int64_t *lda, float **b, const std::int64_t *ldb, - std::int64_t group_count, const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, const side *left_right, const uplo *upper_lower, - const transpose *trans, const diag *unit_diag, const std::int64_t *m, - const std::int64_t *n, const double *alpha, const double **a, - const std::int64_t *lda, double **b, const std::int64_t *ldb, - std::int64_t group_count, const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, const side *left_right, const uplo *upper_lower, - const transpose *trans, const diag *unit_diag, const std::int64_t *m, - const std::int64_t *n, const std::complex *alpha, - const std::complex **a, const std::int64_t *lda, - std::complex **b, const std::int64_t *ldb, std::int64_t group_count, - const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event trsm_batch(sycl::queue &queue, const side *left_right, const uplo *upper_lower, - const transpose *trans, const diag *unit_diag, const std::int64_t *m, - const std::int64_t *n, const std::complex *alpha, - const std::complex **a, const std::int64_t *lda, - std::complex **b, const std::int64_t *ldb, std::int64_t group_count, - const std::int64_t *group_size, - const std::vector &dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, + const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, const side* left_right, const uplo* upper_lower, + const transpose* trans, const diag* unit_diag, const std::int64_t* m, + const std::int64_t* n, const float* alpha, const float** a, + const std::int64_t* lda, float** b, const std::int64_t* ldb, + std::int64_t group_count, const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, const side* left_right, const uplo* upper_lower, + const transpose* trans, const diag* unit_diag, const std::int64_t* m, + const std::int64_t* n, const double* alpha, const double** a, + const std::int64_t* lda, double** b, const std::int64_t* ldb, + std::int64_t group_count, const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, const side* left_right, const uplo* upper_lower, + const transpose* trans, const diag* unit_diag, const std::int64_t* m, + const std::int64_t* n, const std::complex* alpha, + const std::complex** a, const std::int64_t* lda, + std::complex** b, const std::int64_t* ldb, std::int64_t group_count, + const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event trsm_batch(sycl::queue& queue, const side* left_right, const uplo* upper_lower, + const transpose* trans, const diag* unit_diag, const std::int64_t* m, + const std::int64_t* n, const std::complex* alpha, + const std::complex** a, const std::int64_t* lda, + std::complex** b, const std::int64_t* ldb, std::int64_t group_count, + const std::int64_t* group_size, + const std::vector& dependencies = {}); + +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, float *ab, std::int64_t lda, + const std::vector& dependencies = {}); + +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, double *ab, std::int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, std::complex *ab, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, std::complex *ab, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const float *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, value_or_pointer alpha, const float* a, std::int64_t lda, std::int64_t stride_a, value_or_pointer beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, + const float* b, std::int64_t ldb, std::int64_t stride_b, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const double *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, value_or_pointer alpha, const double* a, std::int64_t lda, std::int64_t stride_a, value_or_pointer beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, + const double* b, std::int64_t ldb, std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - value_or_pointer> beta, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex *c, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + value_or_pointer> beta, const std::complex* b, + std::int64_t ldb, std::int64_t stride_b, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - value_or_pointer> beta, const std::complex *b, - std::int64_t ldb, std::int64_t stride_b, std::complex *c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const float *a, std::int64_t lda, float *b, - std::int64_t ldb, const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const double *a, std::int64_t lda, double *b, - std::int64_t ldb, const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies = {}); - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const float *a, std::int64_t lda, - std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const double *a, std::int64_t lda, - std::int64_t stridea, double *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + value_or_pointer> beta, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, + const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, const float* a, std::int64_t lda, float* b, + std::int64_t ldb, const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, const double* a, std::int64_t lda, double* b, + std::int64_t ldb, const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies = {}); + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, const float* a, std::int64_t lda, + std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies = {}); + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, const double* a, std::int64_t lda, + std::int64_t stridea, double* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies = {}); + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, std::complex *b, +sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, float *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); +sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, float* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); -sycl::event imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, double *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); +sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer alpha, double* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies = {}); -sycl::event imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, std::complex *ab, +sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, std::complex *ab, +sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + value_or_pointer> alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const float *a, std::int64_t lda, - value_or_pointer beta, const float *b, std::int64_t ldb, float *c, - std::int64_t ldc, const std::vector &dependencies = {}); +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, value_or_pointer alpha, const float* a, std::int64_t lda, + value_or_pointer beta, const float* b, std::int64_t ldb, float* c, + std::int64_t ldc, const std::vector& dependencies = {}); -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const double *a, std::int64_t lda, - value_or_pointer beta, const double *b, std::int64_t ldb, double *c, - std::int64_t ldc, const std::vector &dependencies = {}); +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, value_or_pointer alpha, const double* a, + std::int64_t lda, value_or_pointer beta, const double* b, + std::int64_t ldb, double* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, - value_or_pointer> beta, const std::complex *b, - std::int64_t ldb, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, + value_or_pointer> beta, const std::complex* b, + std::int64_t ldb, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, value_or_pointer> alpha, - const std::complex *a, std::int64_t lda, - value_or_pointer> beta, const std::complex *b, - std::int64_t ldb, std::complex *c, std::int64_t ldc, - const std::vector &dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const float *alpha, const float **a, - const std::int64_t *lda, float **b, const std::int64_t *ldb, - std::int64_t group_count, const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const double *alpha, const double **a, - const std::int64_t *lda, double **b, const std::int64_t *ldb, - std::int64_t group_count, const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const std::complex *alpha, - const std::complex **a, const std::int64_t *lda, - std::complex **b, const std::int64_t *ldb, - std::int64_t group_count, const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const std::complex *alpha, - const std::complex **a, const std::int64_t *lda, - std::complex **b, const std::int64_t *ldb, - std::int64_t group_count, const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const float *alpha, float **ab, - const std::int64_t *lda, const std::int64_t *ldb, - std::int64_t group_count, const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const double *alpha, double **ab, - const std::int64_t *lda, const std::int64_t *ldb, - std::int64_t group_count, const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const std::complex *alpha, - std::complex **ab, const std::int64_t *lda, - const std::int64_t *ldb, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue &queue, const transpose *trans, const std::int64_t *m, - const std::int64_t *n, const std::complex *alpha, - std::complex **ab, const std::int64_t *lda, - const std::int64_t *ldb, std::int64_t group_count, - const std::int64_t *groupsize, - const std::vector &dependencies = {}); + const std::complex* a, std::int64_t lda, + value_or_pointer> beta, const std::complex* b, + std::int64_t ldb, std::complex* c, std::int64_t ldc, + const std::vector& dependencies = {}); + +sycl::event omatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const float* alpha, const float** a, + const std::int64_t* lda, float** b, const std::int64_t* ldb, + std::int64_t group_count, const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event omatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const double* alpha, const double** a, + const std::int64_t* lda, double** b, const std::int64_t* ldb, + std::int64_t group_count, const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event omatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const std::complex* alpha, + const std::complex** a, const std::int64_t* lda, + std::complex** b, const std::int64_t* ldb, + std::int64_t group_count, const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event omatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const std::complex* alpha, + const std::complex** a, const std::int64_t* lda, + std::complex** b, const std::int64_t* ldb, + std::int64_t group_count, const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event imatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const float* alpha, float** ab, + const std::int64_t* lda, const std::int64_t* ldb, + std::int64_t group_count, const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event imatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const double* alpha, double** ab, + const std::int64_t* lda, const std::int64_t* ldb, + std::int64_t group_count, const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event imatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const std::complex* alpha, + std::complex** ab, const std::int64_t* lda, + const std::int64_t* ldb, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); + +sycl::event imatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, + const std::int64_t* n, const std::complex* alpha, + std::complex** ab, const std::int64_t* lda, + const std::int64_t* ldb, std::int64_t group_count, + const std::int64_t* groupsize, + const std::vector& dependencies = {}); diff --git a/src/blas/backends/mkl_common/mkl_extensions.cxx b/src/blas/backends/mkl_common/mkl_extensions.cxx index 28500b63c..bca426982 100644 --- a/src/blas/backends/mkl_common/mkl_extensions.cxx +++ b/src/blas/backends/mkl_common/mkl_extensions.cxx @@ -19,341 +19,397 @@ // Buffer APIs -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, - beta, c, ldc, co)); -} - -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, - beta, c, ldc, co)); -} - -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, - beta, c, ldc, co)); -} - -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, - beta, c, ldc, co)); -} - -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), + detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co)); +} + +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), + detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co)); +} + +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), + detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co)); +} + +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm_bias( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), + detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co)); +} + +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, - ldc)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, double beta, sycl::buffer& c, int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, - ldc)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, - ldc)); +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, - ldc)); +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemmt( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); } -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb)); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, b, ldb)); } -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb)); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, b, ldb)); } -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb)); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, b, ldb)); } -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb)); +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, b, ldb)); } -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb) { +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb) { throw unimplemented("blas", "omatcopy2", ""); } -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb) { +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb) { throw unimplemented("blas", "omatcopy2", ""); } -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &b, int64_t ldb, std::int64_t strideb) { +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& b, int64_t ldb, std::int64_t strideb) { throw unimplemented("blas", "omatcopy2", ""); } -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, int64_t ldb, +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, int64_t ldb, std::int64_t strideb) { throw unimplemented("blas", "omatcopy2", ""); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb)); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, ab, lda, ldb)); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb)); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, ab, lda, ldb)); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb)); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, ab, lda, ldb)); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb)); +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, ab, lda, ldb)); } -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, float beta, - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc)); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, sycl::buffer& a, int64_t lda, float beta, + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), m, n, alpha, + a, lda, beta, b, ldb, c, ldc)); } -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, double beta, - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc)); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, sycl::buffer& a, int64_t lda, double beta, + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), m, n, alpha, + a, lda, beta, b, ldb, c, ldc)); } -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &b, int64_t ldb, - sycl::buffer, 1> &c, int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc)); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& b, int64_t ldb, + sycl::buffer, 1>& c, int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), m, n, alpha, + a, lda, beta, b, ldb, c, ldc)); } -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &b, int64_t ldb, - sycl::buffer, 1> &c, int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc)); +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& b, int64_t ldb, + sycl::buffer, 1>& c, int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), m, n, alpha, + a, lda, beta, b, ldb, c, ldc)); } // USM APIs -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const int8_t *a, int64_t lda, - int8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, - bo, beta, c, ldc, co, dependencies)); -} - -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const int8_t *a, int64_t lda, - int8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, - bo, beta, c, ldc, co, dependencies)); -} - -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const uint8_t *a, int64_t lda, - uint8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, - bo, beta, c, ldc, co, dependencies)); -} - -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const uint8_t *a, int64_t lda, - uint8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, - bo, beta, c, ldc, co, dependencies)); -} - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, float alpha, const float *a, int64_t lda, const float *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies)); -} - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, double alpha, const double *a, int64_t lda, const double *b, - int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies)); -} - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies)); -} - -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, - c, ldc, dependencies)); -} - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, float *b, int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies)); -} - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, double *b, int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies)); -} - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies)); -} - -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, b, ldb, dependencies)); -} - -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, std::int64_t stridea, float *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), + detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co, dependencies)); +} + +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), + detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co, dependencies)); +} + +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), + detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co, dependencies)); +} + +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm_bias( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), + detail::get_onemkl_offset(offsetc), m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, + co, dependencies)); +} + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, float alpha, const float* a, int64_t lda, const float* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); +} + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); +} + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); +} + +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemmt( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), n, k, alpha, a, lda, b, ldb, beta, c, ldc, + dependencies)); +} + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, float* b, int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, a, lda, b, ldb, dependencies)); +} + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, double* b, int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, a, lda, b, ldb, dependencies)); +} + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex* b, int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, a, lda, b, ldb, dependencies)); +} + +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex* b, int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatcopy(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, a, lda, b, ldb, dependencies)); +} + +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, std::int64_t stridea, float* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy2", ""); } -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, std::int64_t stridea, double *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, std::int64_t stridea, double* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy2", ""); } -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::int64_t stridea, std::complex *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::int64_t stridea, std::complex* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy2", ""); } -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::int64_t stridea, std::complex *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::int64_t stridea, std::complex* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy2", ""); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies)); +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, ab, lda, ldb, dependencies)); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies)); +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, ab, lda, ldb, dependencies)); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies)); +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, ab, lda, ldb, dependencies)); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), m, n, alpha, ab, lda, ldb, dependencies)); +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::imatcopy(queue, detail::get_onemkl_transpose(trans), + m, n, alpha, ab, lda, ldb, dependencies)); } -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, const float *a, int64_t lda, float beta, const float *b, - int64_t ldb, float *c, int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies)); +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, const float* a, int64_t lda, float beta, const float* b, + int64_t ldb, float* c, int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, beta, b, ldb, c, ldc, dependencies)); } -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, const double *a, int64_t lda, double beta, const double *b, - int64_t ldb, double *c, int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies)); +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, const double* a, int64_t lda, double beta, const double* b, + int64_t ldb, double* c, int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, beta, b, ldb, c, ldc, dependencies)); } -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, const std::complex *b, int64_t ldb, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies)); +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, const std::complex* b, int64_t ldb, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, beta, b, ldb, c, ldc, dependencies)); } -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, const std::complex *b, int64_t ldb, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, alpha, a, lda, beta, b, ldb, c, ldc, - dependencies)); +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, const std::complex* b, int64_t ldb, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::omatadd( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, + alpha, a, lda, beta, b, ldb, c, ldc, dependencies)); } diff --git a/src/blas/backends/mkl_common/mkl_level1.cxx b/src/blas/backends/mkl_common/mkl_level1.cxx index b6a985826..fac39ad36 100644 --- a/src/blas/backends/mkl_common/mkl_level1.cxx +++ b/src/blas/backends/mkl_common/mkl_level1.cxx @@ -19,627 +19,645 @@ // Buffer APIs -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::asum(queue, n, x, incx, result)); } -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::asum(queue, n, x, incx, result)); } -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::asum(queue, n, x, incx, result)); } -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::asum(queue, n, x, incx, result)); } -void axpy(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void axpy(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy(queue, n, alpha, x, incx, y, incy)); } -void axpy(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void axpy(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy(queue, n, alpha, x, incx, y, incy)); } -void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy(queue, n, alpha, x, incx, y, incy)); } -void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpy(queue, n, alpha, x, incx, y, incy)); } -void axpby(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { +void axpby(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy)); } -void axpby(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { +void axpby(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy)); } -void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy)); } -void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy)); } -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy) { +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy(queue, n, x, incx, y, incy)); } -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy) { +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy(queue, n, x, incx, y, incy)); } -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy(queue, n, x, incx, y, incy)); } -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::copy(queue, n, x, incx, y, incy)); } -void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { +void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::dot(queue, n, x, incx, y, incy, result)); } -void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { +void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::dot(queue, n, x, incx, y, incy, result)); } -void sdsdot(sycl::queue &queue, std::int64_t n, float sb, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void sdsdot(sycl::queue& queue, std::int64_t n, float sb, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::sdsdot(queue, n, sb, x, incx, y, incy, result)); } -void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { +void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::dot(queue, n, x, incx, y, incy, result)); } -void dotc(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotc(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::dotc(queue, n, x, incx, y, incy, result)); } -void dotc(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotc(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::dotc(queue, n, x, incx, y, incy, result)); } -void dotu(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotu(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::dotu(queue, n, x, incx, y, incy, result)); } -void dotu(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotu(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::dotu(queue, n, x, incx, y, incy, result)); } -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::nrm2(queue, n, x, incx, result)); } -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::nrm2(queue, n, x, incx, result)); } -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::nrm2(queue, n, x, incx, result)); } -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::nrm2(queue, n, x, incx, result)); } -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, float c, +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, float c, float s) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rot(queue, n, x, incx, y, incy, c, s)); } -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, double c, +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, double c, double s) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rot(queue, n, x, incx, y, incy, c, s)); } -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, float c, float s) { +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, float c, float s) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rot(queue, n, x, incx, y, incy, c, s)); } -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, double c, double s) { +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, double c, double s) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rot(queue, n, x, incx, y, incy, c, s)); } -void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s) { +void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotg(queue, a, b, c, s)); } -void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s) { +void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotg(queue, a, b, c, s)); } -void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotg(queue, a, b, c, s)); } -void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotg(queue, a, b, c, s)); } -void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { +void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& param) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotm(queue, n, x, incx, y, incy, param)); } -void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { +void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& param) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotm(queue, n, x, incx, y, incy, param)); } -void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, float y1, sycl::buffer ¶m) { +void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, float y1, sycl::buffer& param) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotmg(queue, d1, d2, x1, y1, param)); } -void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, double y1, sycl::buffer ¶m) { +void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, double y1, sycl::buffer& param) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::rotmg(queue, d1, d2, x1, y1, param)); } -void scal(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, +void scal(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, std::int64_t incx) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } -void scal(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, +void scal(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, std::int64_t incx) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } -void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } -void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } -void scal(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer, 1> &x, +void scal(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer, 1>& x, std::int64_t incx) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } -void scal(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +void scal(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::scal(queue, n, alpha, x, incx)); } -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy) { +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::swap(queue, n, x, incx, y, incy)); } -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy) { +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::swap(queue, n, x, incx, y, incy)); } -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::swap(queue, n, x, incx, y, incy)); } -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::swap(queue, n, x, incx, y, incy)); } -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamax(queue, n, x, incx, result)); } -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamax(queue, n, x, incx, result)); } -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamax(queue, n, x, incx, result)); } -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamax(queue, n, x, incx, result)); } -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamin(queue, n, x, incx, result)); } -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamin(queue, n, x, incx, result)); } -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamin(queue, n, x, incx, result)); } -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::iamin(queue, n, x, incx, result)); } // USM APIs -sycl::event asum(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, float *result, const std::vector &dependencies) { +sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, float* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::asum(queue, n, x, incx, result, dependencies)); } -sycl::event asum(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, double *result, const std::vector &dependencies) { +sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, double* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::asum(queue, n, x, incx, result, dependencies)); } -sycl::event asum(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - float *result, const std::vector &dependencies) { +sycl::event asum(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + float* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::asum(queue, n, x, incx, result, dependencies)); } -sycl::event asum(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - double *result, const std::vector &dependencies) { +sycl::event asum(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + double* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::asum(queue, n, x, incx, result, dependencies)); } -sycl::event axpy(sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, - float *y, std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); +sycl::event axpy(sycl::queue& queue, std::int64_t n, float alpha, const float* x, std::int64_t incx, + float* y, std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); } -sycl::event axpy(sycl::queue &queue, std::int64_t n, double alpha, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); +sycl::event axpy(sycl::queue& queue, std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); } -sycl::event axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); +sycl::event axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); } -sycl::event axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); +sycl::event axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::axpy(queue, n, alpha, x, incx, y, incy, dependencies)); } -sycl::event axpby(sycl::queue &queue, std::int64_t n, float alpha, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); +sycl::event axpby(sycl::queue& queue, std::int64_t n, float alpha, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); } -sycl::event axpby(sycl::queue &queue, std::int64_t n, double alpha, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); +sycl::event axpby(sycl::queue& queue, std::int64_t n, double alpha, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); } -sycl::event axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); +sycl::event axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); } -sycl::event axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); +sycl::event axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::axpby(queue, n, alpha, x, incx, beta, y, incy, dependencies)); } -sycl::event copy(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, float *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event copy(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, float* y, + std::int64_t incy, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy(queue, n, x, incx, y, incy, dependencies)); } -sycl::event copy(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, double *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event copy(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, double* y, + std::int64_t incy, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy(queue, n, x, incx, y, incy, dependencies)); } -sycl::event copy(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy(queue, n, x, incx, y, incy, dependencies)); } -sycl::event copy(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::copy(queue, n, x, incx, y, incy, dependencies)); } -sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *result, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dot(queue, n, x, incx, y, incy, result, dependencies)); +sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* result, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::dot(queue, n, x, incx, y, incy, result, dependencies)); } -sycl::event dot(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *result, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dot(queue, n, x, incx, y, incy, result, dependencies)); +sycl::event dot(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* result, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::dot(queue, n, x, incx, y, incy, result, dependencies)); } -sycl::event sdsdot(sycl::queue &queue, std::int64_t n, float sb, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *result, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::sdsdot(queue, n, sb, x, incx, y, incy, result, dependencies)); +sycl::event sdsdot(sycl::queue& queue, std::int64_t n, float sb, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* result, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::sdsdot(queue, n, sb, x, incx, y, incy, result, dependencies)); } -sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, double *result, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dot(queue, n, x, incx, y, incy, result, dependencies)); +sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, double* result, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::dot(queue, n, x, incx, y, incy, result, dependencies)); } -sycl::event dotc(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dotc(queue, n, x, incx, y, incy, result, dependencies)); +sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::dotc(queue, n, x, incx, y, incy, result, dependencies)); } -sycl::event dotc(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dotc(queue, n, x, incx, y, incy, result, dependencies)); +sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::dotc(queue, n, x, incx, y, incy, result, dependencies)); } -sycl::event dotu(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dotu(queue, n, x, incx, y, incy, result, dependencies)); +sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::dotu(queue, n, x, incx, y, incy, result, dependencies)); } -sycl::event dotu(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::dotu(queue, n, x, incx, y, incy, result, dependencies)); +sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::dotu(queue, n, x, incx, y, incy, result, dependencies)); } -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, float *result, const std::vector &dependencies) { +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, float* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::nrm2(queue, n, x, incx, result, dependencies)); } -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, double *result, const std::vector &dependencies) { +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, double* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::nrm2(queue, n, x, incx, result, dependencies)); } -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - float *result, const std::vector &dependencies) { +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + float* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::nrm2(queue, n, x, incx, result, dependencies)); } -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - double *result, const std::vector &dependencies) { +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + double* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::nrm2(queue, n, x, incx, result, dependencies)); } -sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies)); } -sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies)); } -sycl::event rot(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, - std::int64_t incy, float c, float s, const std::vector &dependencies) { +sycl::event rot(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, + std::int64_t incy, float c, float s, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies)); } -sycl::event rot(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, +sycl::event rot(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rot(queue, n, x, incx, y, incy, c, s, dependencies)); } -sycl::event rotg(sycl::queue &queue, float *a, float *b, float *c, float *s, - const std::vector &dependencies) { +sycl::event rotg(sycl::queue& queue, float* a, float* b, float* c, float* s, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotg(queue, a, b, c, s, dependencies)); } -sycl::event rotg(sycl::queue &queue, double *a, double *b, double *c, double *s, - const std::vector &dependencies) { +sycl::event rotg(sycl::queue& queue, double* a, double* b, double* c, double* s, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotg(queue, a, b, c, s, dependencies)); } -sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, float *c, - std::complex *s, const std::vector &dependencies) { +sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, float* c, + std::complex* s, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotg(queue, a, b, c, s, dependencies)); } -sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, double *c, - std::complex *s, const std::vector &dependencies) { +sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, double* c, + std::complex* s, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotg(queue, a, b, c, s, dependencies)); } -sycl::event rotm(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, - std::int64_t incy, float *param, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotm(queue, n, x, incx, y, incy, param, dependencies)); +sycl::event rotm(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, + std::int64_t incy, float* param, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::rotm(queue, n, x, incx, y, incy, param, dependencies)); } -sycl::event rotm(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, - std::int64_t incy, double *param, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotm(queue, n, x, incx, y, incy, param, dependencies)); +sycl::event rotm(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, + std::int64_t incy, double* param, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::rotm(queue, n, x, incx, y, incy, param, dependencies)); } -sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, float y1, float *param, - const std::vector &dependencies) { +sycl::event rotmg(sycl::queue& queue, float* d1, float* d2, float* x1, float y1, float* param, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotmg(queue, d1, d2, x1, y1, param, dependencies)); } -sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, double y1, double *param, - const std::vector &dependencies) { +sycl::event rotmg(sycl::queue& queue, double* d1, double* d2, double* x1, double y1, double* param, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::rotmg(queue, d1, d2, x1, y1, param, dependencies)); } -sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, float *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(sycl::queue& queue, std::int64_t n, float alpha, float* x, std::int64_t incx, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } -sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, double *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(sycl::queue& queue, std::int64_t n, double alpha, double* x, std::int64_t incx, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } -sycl::event scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } -sycl::event scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } -sycl::event scal(sycl::queue &queue, std::int64_t n, float alpha, std::complex *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event scal(sycl::queue& queue, std::int64_t n, float alpha, std::complex* x, + std::int64_t incx, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } -sycl::event scal(sycl::queue &queue, std::int64_t n, double alpha, std::complex *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event scal(sycl::queue& queue, std::int64_t n, double alpha, std::complex* x, + std::int64_t incx, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::scal(queue, n, alpha, x, incx, dependencies)); } -sycl::event swap(sycl::queue &queue, std::int64_t n, float *x, std::int64_t incx, float *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event swap(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, + std::int64_t incy, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::swap(queue, n, x, incx, y, incy, dependencies)); } -sycl::event swap(sycl::queue &queue, std::int64_t n, double *x, std::int64_t incx, double *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event swap(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, + std::int64_t incy, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::swap(queue, n, x, incx, y, incy, dependencies)); } -sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::swap(queue, n, x, incx, y, incy, dependencies)); } -sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::swap(queue, n, x, incx, y, incy, dependencies)); } -sycl::event iamax(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies) { +sycl::event iamax(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamax(queue, n, x, incx, result, dependencies)); } -sycl::event iamax(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies) { +sycl::event iamax(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamax(queue, n, x, incx, result, dependencies)); } -sycl::event iamax(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamax(queue, n, x, incx, result, dependencies)); } -sycl::event iamax(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamax(queue, n, x, incx, result, dependencies)); } -sycl::event iamin(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies) { +sycl::event iamin(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamin(queue, n, x, incx, result, dependencies)); } -sycl::event iamin(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies) { +sycl::event iamin(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamin(queue, n, x, incx, result, dependencies)); } -sycl::event iamin(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamin(queue, n, x, incx, result, dependencies)); } -sycl::event iamin(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::iamin(queue, n, x, incx, result, dependencies)); } diff --git a/src/blas/backends/mkl_common/mkl_level2.cxx b/src/blas/backends/mkl_common/mkl_level2.cxx index f2d0a5561..0b0f40c6c 100644 --- a/src/blas/backends/mkl_common/mkl_level2.cxx +++ b/src/blas/backends/mkl_common/mkl_level2.cxx @@ -19,844 +19,1022 @@ // Buffer APIs -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy)); +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, x, incx, beta, y, incy)); } -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy)); +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, x, incx, beta, y, incy)); } -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy)); +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, x, incx, beta, y, incy)); } -void gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy)); +void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, x, incx, beta, y, incy)); } -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, + ku, alpha, a, lda, x, incx, beta, y, incy)); } -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, + ku, alpha, a, lda, x, incx, beta, y, incy)); } -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy)); +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, + ku, alpha, a, lda, x, incx, beta, y, incy)); } -void gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy)); +void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, + std::int64_t ku, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, + ku, alpha, a, lda, x, incx, beta, y, incy)); } -void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, sycl::buffer &a, +void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, sycl::buffer& a, std::int64_t lda) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda)); } -void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { +void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda)); } -void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda)); } -void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda)); } -void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda)); } -void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { RETHROW_ONEMKL_EXCEPTIONS(blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda)); } -void hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy)); +void hbmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, + lda, x, incx, beta, y, incy)); } -void hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy)); +void hbmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::int64_t k, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, + lda, x, incx, beta, y, incy)); } -void hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy)); +void hemv(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, + lda, x, incx, beta, y, incy)); } -void hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy)); +void hemv(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, + lda, x, incx, beta, y, incy)); } -void her(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); +void her(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); } -void her(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); +void her(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); } -void her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); +void her2(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); } -void her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); +void her2(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); } -void hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, +void hpmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, + incx, beta, y, incy)); } -void hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, - std::int64_t incx, std::complex beta, sycl::buffer, 1> &y, +void hpmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, + std::int64_t incx, std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, + incx, beta, y, incy)); } -void hpr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); +void hpr(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); } -void hpr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); +void hpr(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); } -void hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); +void hpr2(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); } -void hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); +void hpr2(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); } -void sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy)); +void sbmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::int64_t k, float alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, + lda, x, incx, beta, y, incy)); } -void sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy)); +void sbmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::int64_t k, double alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, + lda, x, incx, beta, y, incy)); } -void spmv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, +void spmv(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, + incx, beta, y, incy)); } -void spmv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, +void spmv(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, + incx, beta, y, incy)); } -void spr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); +void spr(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& a) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); } -void spr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); +void spr(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& a) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a)); } -void spr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); +void spr2(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); } -void spr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); +void spr2(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a)); } -void symv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy)); +void symv(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, + lda, x, incx, beta, y, incy)); } -void symv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy)); +void symv(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, + lda, x, incx, beta, y, incy)); } -void syr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); +void syr(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& a, std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); } -void syr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); +void syr(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& a, std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda)); } -void syr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); +void syr2(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); } -void syr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); +void syr2(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda)); } -void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void tbmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } -void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void tbmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } -void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); +void tbmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } -void tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); +void tbmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } -void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void tbsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } -void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void tbsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } -void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); +void tbsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } -void tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); +void tbsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, std::int64_t k, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx)); } -void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); +void tpmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx)); } -void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); +void tpmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx)); } -void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx)); } -void tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx)); } -void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); +void tpsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx)); } -void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); +void tpsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx)); } -void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx)); } -void tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer, 1> &a, sycl::buffer, 1> &x, +void tpsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer, 1>& a, sycl::buffer, 1>& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx)); } -void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } -void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } -void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); +void trmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } -void trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); +void trmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } -void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } -void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, +void trsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } -void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); +void trsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } -void trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx)); +void trsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx)); } // USM APIs -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, - dependencies)); -} - -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, - double *y, std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, - dependencies)); -} - -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, - dependencies)); -} - -sycl::event gemv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, alpha, a, lda, x, incx, beta, y, incy, - dependencies)); -} - -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, float alpha, const float *a, std::int64_t lda, - const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, - dependencies)); -} - -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, double alpha, const double *a, std::int64_t lda, - const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, - dependencies)); -} - -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, + const float* a, std::int64_t lda, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, x, incx, beta, y, incy, + dependencies)); +} + +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, + const double* a, std::int64_t lda, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, x, incx, beta, y, incy, + dependencies)); +} + +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, x, incx, beta, y, incy, + dependencies)); +} + +sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemv(queue, detail::get_onemkl_transpose(trans), m, n, + alpha, a, lda, x, incx, beta, y, incy, + dependencies)); +} + +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, float alpha, const float* a, std::int64_t lda, + const float* x, std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, + kl, ku, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); +} + +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, double alpha, const double* a, std::int64_t lda, + const double* x, std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, + kl, ku, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); +} + +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, - dependencies)); + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, + kl, ku, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } -sycl::event gbmv(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, +sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, - dependencies)); + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gbmv(queue, detail::get_onemkl_transpose(trans), m, n, + kl, ku, alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } - -sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); + +sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } -sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); +sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::ger(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } -sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); +sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } -sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); +sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::gerc(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } -sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); +sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } -sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); +sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::geru(queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies)); } -sycl::event hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies)); +sycl::event hbmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, + alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } - -sycl::event hbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies)); + +sycl::event hbmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hbmv(queue, detail::get_onemkl_uplo(uplo), n, k, + alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } -sycl::event hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies)); +sycl::event hemv(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, + a, lda, x, incx, beta, y, incy, dependencies)); } -sycl::event hemv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies)); +sycl::event hemv(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemv(queue, detail::get_onemkl_uplo(uplo), n, alpha, + a, lda, x, incx, beta, y, incy, dependencies)); } -sycl::event her(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies)); +sycl::event her(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, + incx, a, lda, dependencies)); } -sycl::event her(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies)); +sycl::event her(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, + incx, a, lda, dependencies)); } -sycl::event her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies)); +sycl::event her2(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, + x, incx, y, incy, a, lda, dependencies)); } -sycl::event her2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies)); +sycl::event her2(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2(queue, detail::get_onemkl_uplo(uplo), n, alpha, + x, incx, y, incy, a, lda, dependencies)); } -sycl::event hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - const std::complex *a, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies)); +sycl::event hpmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + const std::complex* a, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, + a, x, incx, beta, y, incy, dependencies)); } -sycl::event hpmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - const std::complex *a, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies)); +sycl::event hpmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + const std::complex* a, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, + a, x, incx, beta, y, incy, dependencies)); } -sycl::event hpr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); +sycl::event hpr(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); } -sycl::event hpr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); +sycl::event hpr(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::hpr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); } -sycl::event hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies)); +sycl::event hpr2(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, + x, incx, y, incy, a, dependencies)); } -sycl::event hpr2(sycl::queue &queue, uplo uplo, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies)); +sycl::event hpr2(sycl::queue& queue, uplo uplo, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hpr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, + x, incx, y, incy, a, dependencies)); } -sycl::event sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies)); +sycl::event sbmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, + alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } -sycl::event sbmv(sycl::queue &queue, uplo uplo, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, const double *x, std::int64_t incx, double beta, - double *y, std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies)); +sycl::event sbmv(sycl::queue& queue, uplo uplo, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::sbmv(queue, detail::get_onemkl_uplo(uplo), n, k, + alpha, a, lda, x, incx, beta, y, incy, + dependencies)); } -sycl::event spmv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *a, - const float *x, std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies)); +sycl::event spmv(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, const float* a, + const float* x, std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, + a, x, incx, beta, y, incy, dependencies)); } -sycl::event spmv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *a, - const double *x, std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, x, incx, beta, y, incy, dependencies)); +sycl::event spmv(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, const double* a, + const double* x, std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spmv(queue, detail::get_onemkl_uplo(uplo), n, alpha, + a, x, incx, beta, y, incy, dependencies)); } -sycl::event spr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, - std::int64_t incx, float *a, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); +sycl::event spr(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, const float* x, + std::int64_t incx, float* a, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); } -sycl::event spr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, - std::int64_t incx, double *a, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); +sycl::event spr(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* a, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::spr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, dependencies)); } -sycl::event spr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *a, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies)); +sycl::event spr2(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* a, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, + x, incx, y, incy, a, dependencies)); } -sycl::event spr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *a, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, dependencies)); +sycl::event spr2(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* a, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::spr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, + x, incx, y, incy, a, dependencies)); } -sycl::event symv(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies)); +sycl::event symv(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, + a, lda, x, incx, beta, y, incy, dependencies)); } -sycl::event symv(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, a, lda, x, incx, beta, y, incy, dependencies)); +sycl::event symv(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symv(queue, detail::get_onemkl_uplo(uplo), n, alpha, + a, lda, x, incx, beta, y, incy, dependencies)); } -sycl::event syr(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, - std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies)); +sycl::event syr(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, const float* x, + std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, + incx, a, lda, dependencies)); } -sycl::event syr(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, - std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, a, lda, dependencies)); +sycl::event syr(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, + incx, a, lda, dependencies)); } -sycl::event syr2(sycl::queue &queue, uplo uplo, std::int64_t n, float alpha, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies)); +sycl::event syr2(sycl::queue& queue, uplo uplo, std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, + x, incx, y, incy, a, lda, dependencies)); } -sycl::event syr2(sycl::queue &queue, uplo uplo, std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, x, incx, y, incy, a, lda, dependencies)); +sycl::event syr2(sycl::queue& queue, uplo uplo, std::int64_t n, double alpha, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2(queue, detail::get_onemkl_uplo(uplo), n, alpha, + x, incx, y, incy, a, lda, dependencies)); } -sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); +sycl::event tbmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + std::int64_t k, const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } -sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); +sycl::event tbmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + std::int64_t k, const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } -sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - std::int64_t k, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); +sycl::event tbmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } -sycl::event tbmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - std::int64_t k, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); +sycl::event tbmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tbmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } -sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - std::int64_t k, const float *a, std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); +sycl::event tbsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + std::int64_t k, const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } -sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - std::int64_t k, const double *a, std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); +sycl::event tbsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + std::int64_t k, const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } -sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - std::int64_t k, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); +sycl::event tbsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } -sycl::event tbsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - std::int64_t k, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); +sycl::event tbsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + std::int64_t k, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tbsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, k, a, lda, x, incx, dependencies)); } -sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); +sycl::event tpmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); } -sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); +sycl::event tpmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); } -sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const std::complex *a, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); +sycl::event tpmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const std::complex* a, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); } -sycl::event tpmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const std::complex *a, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); -} +sycl::event tpmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const std::complex* a, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tpmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); +} + +sycl::event tpsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); +} + +sycl::event tpsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); +} + +sycl::event tpsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const std::complex* a, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); +} -sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); -} - -sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); -} - -sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const std::complex *a, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); -} - -sycl::event tpsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const std::complex *a, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); -} +sycl::event tpsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const std::complex* a, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::tpsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, x, incx, dependencies)); +} -sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const float *a, std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); -} +sycl::event trmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); +} -sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const double *a, std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); -} +sycl::event trmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); +} -sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const std::complex *a, std::int64_t lda, std::complex *x, - std::int64_t incx, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); -} +sycl::event trmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const std::complex* a, std::int64_t lda, std::complex* x, + std::int64_t incx, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); +} -sycl::event trmv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const std::complex *a, std::int64_t lda, std::complex *x, - std::int64_t incx, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); -} +sycl::event trmv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const std::complex* a, std::int64_t lda, std::complex* x, + std::int64_t incx, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::trmv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); +} -sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const float *a, std::int64_t lda, float *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); -} +sycl::event trsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const float* a, std::int64_t lda, float* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); +} -sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const double *a, std::int64_t lda, double *x, std::int64_t incx, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); -} +sycl::event trsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const double* a, std::int64_t lda, double* x, std::int64_t incx, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); +} -sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const std::complex *a, std::int64_t lda, std::complex *x, - std::int64_t incx, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); -} +sycl::event trsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const std::complex* a, std::int64_t lda, std::complex* x, + std::int64_t incx, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); +} -sycl::event trsv(sycl::queue &queue, uplo uplo, transpose trans, diag diag, std::int64_t n, - const std::complex *a, std::int64_t lda, std::complex *x, - std::int64_t incx, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); +sycl::event trsv(sycl::queue& queue, uplo uplo, transpose trans, diag diag, std::int64_t n, + const std::complex* a, std::int64_t lda, std::complex* x, + std::int64_t incx, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + blas_major::trsv(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, a, lda, x, incx, dependencies)); } diff --git a/src/blas/backends/mkl_common/mkl_level3.cxx b/src/blas/backends/mkl_common/mkl_level3.cxx index 62d26128e..e76dac543 100644 --- a/src/blas/backends/mkl_common/mkl_level3.cxx +++ b/src/blas/backends/mkl_common/mkl_level3.cxx @@ -19,501 +19,608 @@ // Buffer APIs -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc)); } -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, sycl::buffer &c, +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc)); } -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc)); } -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc)); } -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::half alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::half beta, - sycl::buffer &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::half alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::half beta, + sycl::buffer& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc)); } -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc)); } -void gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, +void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc)); -} - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); -} - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); -} - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); -} - -void symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); -} - -void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); -} + RETHROW_ONEMKL_EXCEPTIONS(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), + detail::get_onemkl_transpose(transb), m, n, k, alpha, + a, lda, b, ldb, beta, c, ldc)); +} -void hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, ldc)); -} +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, + lda, b, ldb, beta, c, ldc)); +} -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, float beta, - sycl::buffer &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); -} - -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, double beta, - sycl::buffer &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); -} +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, + lda, b, ldb, beta, c, ldc)); +} -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); -} +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, + lda, b, ldb, beta, c, ldc)); +} -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); -} +void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::symm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, + lda, b, ldb, beta, c, ldc)); +} -void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer, 1> &a, std::int64_t lda, float beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); +void hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, + lda, b, ldb, beta, c, ldc)); +} + +void hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::hemm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, + lda, b, ldb, beta, c, ldc)); +} + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, float beta, + sycl::buffer& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, beta, c, ldc)); +} + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, double beta, + sycl::buffer& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, beta, c, ldc)); +} + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, beta, c, ldc)); +} + +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, beta, c, ldc)); +} + +void herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer, 1>& a, std::int64_t lda, float beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, beta, c, ldc)); +} + +void herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer, 1>& a, std::int64_t lda, double beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, beta, c, ldc)); +} + +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + float alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, b, ldb, beta, c, ldc)); +} + +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, b, ldb, beta, c, ldc)); +} + +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, b, ldb, beta, c, ldc)); +} + +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, b, ldb, beta, c, ldc)); } - -void herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer, 1> &a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc)); + +void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, b, ldb, beta, c, ldc)); } - -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); + +void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, double beta, + sycl::buffer, 1>& c, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS(blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, a, + lda, b, ldb, beta, c, ldc)); } - -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); -} - -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); -} - -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); -} - -void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); -} - -void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc)); -} - -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb)); -} - -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb)); -} - -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, + +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::trmm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb)); +} + +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::trmm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb)); +} + +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb)); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::trmm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb)); } -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb)); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::trmm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb)); } -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb)); +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::trsm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb)); } -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb)); +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::trsm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb)); } -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb)); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::trsm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb)); } -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, - ldb)); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS( + blas_major::trsm(queue, detail::get_onemkl_side(left_right), + detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), + detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, b, ldb)); } // USM APIs -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); } -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); } -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); } -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); -} - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, sycl::half beta, - sycl::half *c, std::int64_t ldc, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); -} - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, - const sycl::half *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); -} - -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, std::int64_t lda, - const bfloat16 *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm(queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); -} - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies)); -} - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies)); -} - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies)); -} - -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies)); -} - -sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies)); -} - -sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, alpha, a, lda, b, ldb, beta, c, - ldc, dependencies)); -} - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, float beta, - float *c, std::int64_t ldc, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies)); -} - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, double beta, - double *c, std::int64_t ldc, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies)); -} - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies)); -} - -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies)); -} - -sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, const std::complex *a, std::int64_t lda, - float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies)); -} - -sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, const std::complex *a, std::int64_t lda, - double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, beta, c, ldc, - dependencies)); -} - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); -} - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); -} - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); -} - -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); -} - -sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); -} - -sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, double beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2k(queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, alpha, a, lda, b, ldb, beta, c, ldc, - dependencies)); -} - -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, - diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies)); -} - -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, - diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies)); -} - -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, sycl::half beta, + sycl::half* c, std::int64_t ldc, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, std::int64_t lda, + const sycl::half* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::int64_t k, float alpha, const bfloat16* a, std::int64_t lda, + const bfloat16* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::gemm( + queue, detail::get_onemkl_transpose(transa), detail::get_onemkl_transpose(transb), m, n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::symm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::hemm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), m, n, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, const float* a, std::int64_t lda, float beta, + float* c, std::int64_t ldc, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, + a, lda, beta, c, ldc, dependencies)); +} + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, const double* a, std::int64_t lda, double beta, + double* c, std::int64_t ldc, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, + a, lda, beta, c, ldc, dependencies)); +} + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, + a, lda, beta, c, ldc, dependencies)); +} + +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syrk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, + a, lda, beta, c, ldc, dependencies)); +} + +sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, const std::complex* a, std::int64_t lda, + float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, + a, lda, beta, c, ldc, dependencies)); +} + +sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, const std::complex* a, std::int64_t lda, + double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::herk(queue, detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(trans), n, k, alpha, + a, lda, beta, c, ldc, dependencies)); +} + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, float alpha, const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, double alpha, const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::syr2k( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, float beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2k( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, double beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::her2k( + queue, detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(trans), n, k, + alpha, a, lda, b, ldb, beta, c, ldc, dependencies)); +} + +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, + diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, dependencies)); +} + +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, + diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, dependencies)); +} + +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies)); + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, dependencies)); } -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies)); -} - -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, - diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies)); -} - -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, - diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies)); -} - -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trmm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, dependencies)); +} + +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, + diag unit_diag, std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, dependencies)); +} + +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, + diag unit_diag, std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, dependencies)); +} + +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies)); + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, dependencies)); } -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm(queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, lda, - b, ldb, dependencies)); + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(blas_major::trsm( + queue, detail::get_onemkl_side(left_right), detail::get_onemkl_uplo(upper_lower), + detail::get_onemkl_transpose(transa), detail::get_onemkl_diag(unit_diag), m, n, alpha, a, + lda, b, ldb, dependencies)); } diff --git a/src/blas/backends/netlib/netlib_batch.cxx b/src/blas/backends/netlib/netlib_batch.cxx index 7a2839dd4..5af30b80f 100644 --- a/src/blas/backends/netlib/netlib_batch.cxx +++ b/src/blas/backends/netlib/netlib_batch.cxx @@ -19,8 +19,8 @@ // Buffer APIs -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); @@ -30,8 +30,8 @@ void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_ #endif } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); @@ -41,8 +41,8 @@ void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64 #endif } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, - int64_t incx, int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, + int64_t incx, int64_t stridex, sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); @@ -52,8 +52,8 @@ void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, #endif } -void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, - int64_t incx, int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, + int64_t incx, int64_t stridex, sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); @@ -63,9 +63,9 @@ void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer #endif } -void axpy_batch(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, - int64_t stridey, int64_t batch_size) { +void axpy_batch(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, + int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); #endif @@ -74,8 +74,8 @@ void axpy_batch(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, +void axpy_batch(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); @@ -85,9 +85,9 @@ void axpy_batch(sycl::queue &queue, int64_t n, double alpha, sycl::buffer alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); @@ -97,9 +97,9 @@ void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, #endif } -void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, int64_t stridex, - sycl::buffer, 1> &y, int64_t incy, int64_t stridey, +void axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, int64_t stridex, + sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); @@ -109,10 +109,10 @@ void axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, #endif } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &x, int64_t incx, int64_t stride_x, float beta, - sycl::buffer &y, int64_t incy, int64_t stride_y, int64_t batch_size) { +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, sycl::buffer& x, + int64_t incx, int64_t stride_x, float beta, sycl::buffer& y, int64_t incy, + int64_t stride_y, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); #endif @@ -121,11 +121,10 @@ void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, floa #endif } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &x, int64_t incx, int64_t stride_x, double beta, - sycl::buffer &y, int64_t incy, int64_t stride_y, - int64_t batch_size) { +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& x, int64_t incx, int64_t stride_x, double beta, + sycl::buffer& y, int64_t incy, int64_t stride_y, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); #endif @@ -134,12 +133,11 @@ void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, doub #endif } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &x, int64_t incx, - int64_t stride_x, std::complex beta, - sycl::buffer, 1> &y, int64_t incy, int64_t stride_y, - int64_t batch_size) { +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& x, int64_t incx, + int64_t stride_x, std::complex beta, sycl::buffer, 1>& y, + int64_t incy, int64_t stride_y, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); #endif @@ -148,11 +146,11 @@ void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, #endif } -void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &x, - int64_t incx, int64_t stride_x, std::complex beta, - sycl::buffer, 1> &y, int64_t incy, int64_t stride_y, +void gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& x, int64_t incx, + int64_t stride_x, std::complex beta, + sycl::buffer, 1>& y, int64_t incy, int64_t stride_y, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); @@ -162,10 +160,10 @@ void gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, #endif } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &x, int64_t incx, int64_t stride_x, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stride_a, sycl::buffer& x, + int64_t incx, int64_t stride_x, sycl::buffer& c, int64_t ldc, + int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); #endif @@ -174,10 +172,10 @@ void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, #endif } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &x, int64_t incx, int64_t stride_x, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& x, int64_t incx, int64_t stride_x, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); #endif @@ -186,10 +184,10 @@ void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, #endif } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& x, int64_t incx, int64_t stride_x, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); @@ -199,10 +197,10 @@ void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, #endif } -void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &x, int64_t incx, int64_t stride_x, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, +void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& x, int64_t incx, int64_t stride_x, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); @@ -212,11 +210,10 @@ void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, #endif } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, - int64_t batch_size) { +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, float beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -225,11 +222,10 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t #endif } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, double alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, - int64_t batch_size) { +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, double beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -238,12 +234,11 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t #endif } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, - int64_t ldb, int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, - int64_t batch_size) { +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, + int64_t stride_b, std::complex beta, sycl::buffer, 1>& c, + int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -252,11 +247,11 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t #endif } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); @@ -266,10 +261,10 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t #endif } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, sycl::half alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::half beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, sycl::half alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::half beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); @@ -279,10 +274,10 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t #endif } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); @@ -292,10 +287,10 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t #endif } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); @@ -305,10 +300,10 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t #endif } -void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int64_t stride_a, sycl::buffer &b, int64_t ldb, int64_t stride_b, - float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, + float beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); @@ -318,9 +313,9 @@ void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t #endif } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); @@ -330,9 +325,9 @@ void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose #endif } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer &a, - int64_t lda, int64_t stride_a, sycl::buffer &b, int64_t ldb, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, sycl::buffer& a, + int64_t lda, int64_t stride_a, sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); @@ -342,10 +337,10 @@ void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose #endif } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); @@ -355,10 +350,10 @@ void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose #endif } -void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, +void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, int64_t stride_a, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); @@ -368,10 +363,9 @@ void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose #endif } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - float beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, - int64_t batch_size) { +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, float beta, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); #endif @@ -380,9 +374,9 @@ void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n #endif } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &c, int64_t ldc, int64_t stride_c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); @@ -392,11 +386,10 @@ void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n #endif } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, - int64_t batch_size) { +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, std::complex beta, sycl::buffer, 1>& c, + int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); #endif @@ -405,10 +398,10 @@ void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n #endif } -void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, +void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); @@ -418,9 +411,9 @@ void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n #endif } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); #endif @@ -429,9 +422,9 @@ void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, f #endif } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, int64_t stride_a, - sycl::buffer &b, int64_t ldb, int64_t stride_b, int64_t batch_size) { +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, int64_t stride_a, + sycl::buffer& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); #endif @@ -440,9 +433,9 @@ void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, d #endif } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stride_a, sycl::buffer, 1> &b, int64_t ldb, +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); @@ -452,9 +445,9 @@ void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, int64_t stride_a, sycl::buffer, 1> &b, +void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, int64_t stride_a, sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); @@ -464,8 +457,8 @@ void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); @@ -475,8 +468,8 @@ void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, f #endif } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); @@ -486,8 +479,8 @@ void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, d #endif } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); @@ -497,8 +490,8 @@ void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); @@ -508,10 +501,10 @@ void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - float beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + float beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd_batch", "for column_major layout"); #endif @@ -520,10 +513,10 @@ void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64 #endif } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer &b, int64_t ldb, int64_t stride_b, - sycl::buffer &c, int64_t ldc, int64_t stride_c, int64_t batch_size) { +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, + double beta, sycl::buffer& b, int64_t ldb, int64_t stride_b, + sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd_batch", "for column_major layout"); #endif @@ -532,11 +525,11 @@ void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64 #endif } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd_batch", "for column_major layout"); @@ -546,11 +539,11 @@ void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64 #endif } -void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, +void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, int64_t ldb, int64_t stride_b, - sycl::buffer, 1> &c, int64_t ldc, int64_t stride_c, + sycl::buffer, 1>& b, int64_t ldb, int64_t stride_b, + sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, int64_t batch_size) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd_batch", "for column_major layout"); @@ -562,9 +555,9 @@ void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64 // USM APIs -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const float **x, int64_t *incx, - float **y, int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const float** x, int64_t* incx, float** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); #endif @@ -573,9 +566,9 @@ sycl::event copy_batch(sycl::queue &queue, int64_t *n, const float **x, int64_t #endif } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const double **x, int64_t *incx, - double **y, int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const double** x, int64_t* incx, double** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); #endif @@ -584,10 +577,9 @@ sycl::event copy_batch(sycl::queue &queue, int64_t *n, const double **x, int64_t #endif } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, - int64_t *incx, std::complex **y, int64_t *incy, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const std::complex** x, int64_t* incx, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); #endif @@ -596,10 +588,9 @@ sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex #endif } -sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex **x, - int64_t *incx, std::complex **y, int64_t *incy, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t* n, const std::complex** x, + int64_t* incx, std::complex** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); #endif @@ -608,10 +599,9 @@ sycl::event copy_batch(sycl::queue &queue, int64_t *n, const std::complex &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t n, const float* x, int64_t incx, + std::int64_t stridex, float* y, int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); #endif @@ -620,10 +610,9 @@ sycl::event copy_batch(sycl::queue &queue, int64_t n, const float *x, int64_t in #endif } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const double *x, int64_t incx, - std::int64_t stridex, double *y, int64_t incy, std::int64_t stridey, - std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t n, const double* x, int64_t incx, + std::int64_t stridex, double* y, int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); #endif @@ -632,10 +621,10 @@ sycl::event copy_batch(sycl::queue &queue, int64_t n, const double *x, int64_t i #endif } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, - int64_t incx, std::int64_t stridex, std::complex *y, int64_t incy, - std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::int64_t stridex, std::complex* y, int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); #endif @@ -644,10 +633,10 @@ sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex #endif } -sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex *x, - int64_t incx, std::int64_t stridex, std::complex *y, - int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::int64_t stridex, std::complex* y, int64_t incy, + std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "copy_batch", "for column_major layout"); #endif @@ -656,9 +645,9 @@ sycl::event copy_batch(sycl::queue &queue, int64_t n, const std::complex #endif } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, float *alpha, const float **x, - int64_t *incx, float **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, float* alpha, const float** x, int64_t* incx, + float** y, int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); #endif @@ -667,9 +656,9 @@ sycl::event axpy_batch(sycl::queue &queue, int64_t *n, float *alpha, const float #endif } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, double *alpha, const double **x, - int64_t *incx, double **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, double* alpha, const double** x, + int64_t* incx, double** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); #endif @@ -678,10 +667,10 @@ sycl::event axpy_batch(sycl::queue &queue, int64_t *n, double *alpha, const doub #endif } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, std::complex* alpha, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); #endif @@ -690,10 +679,10 @@ sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alph #endif } -sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alpha, - const std::complex **x, int64_t *incx, std::complex **y, - int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t* n, std::complex* alpha, + const std::complex** x, int64_t* incx, std::complex** y, + int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); #endif @@ -702,9 +691,9 @@ sycl::event axpy_batch(sycl::queue &queue, int64_t *n, std::complex *alp #endif } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, float alpha, const float *x, - int64_t incx, int64_t stridex, float *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, + int64_t stridex, float* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); #endif @@ -713,9 +702,9 @@ sycl::event axpy_batch(sycl::queue &queue, int64_t n, float alpha, const float * #endif } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, double alpha, const double *x, - int64_t incx, int64_t stridex, double *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + int64_t stridex, double* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); #endif @@ -724,10 +713,10 @@ sycl::event axpy_batch(sycl::queue &queue, int64_t n, double alpha, const double #endif } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); #endif @@ -736,10 +725,10 @@ sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, #endif } -sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, int64_t stridex, - std::complex *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, int64_t stridex, + std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpy_batch", "for column_major layout"); #endif @@ -748,11 +737,10 @@ sycl::event axpy_batch(sycl::queue &queue, int64_t n, std::complex alpha #endif } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - float alpha, const float *a, int64_t lda, int64_t stride_a, - const float *x, int64_t incx, int64_t stride_x, float beta, float *y, - int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stride_a, const float* x, int64_t incx, + int64_t stride_x, float beta, float* y, int64_t incy, int64_t stride_y, + int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); #endif @@ -761,11 +749,11 @@ sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t #endif } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - double alpha, const double *a, int64_t lda, int64_t stride_a, - const double *x, int64_t incx, int64_t stride_x, double beta, double *y, - int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stride_a, const double* x, + int64_t incx, int64_t stride_x, double beta, double* y, int64_t incy, + int64_t stride_y, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); #endif @@ -774,12 +762,12 @@ sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t #endif } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, const std::complex *x, int64_t incx, - int64_t stride_x, std::complex beta, std::complex *y, - int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, const std::complex* x, int64_t incx, + int64_t stride_x, std::complex beta, std::complex* y, + int64_t incy, int64_t stride_y, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); #endif @@ -788,12 +776,12 @@ sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t #endif } -sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, const std::complex *x, int64_t incx, - int64_t stride_x, std::complex beta, std::complex *y, - int64_t incy, int64_t stride_y, int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose transa, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, const std::complex* x, int64_t incx, + int64_t stride_x, std::complex beta, std::complex* y, + int64_t incy, int64_t stride_y, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); #endif @@ -802,11 +790,10 @@ sycl::event gemv_batch(sycl::queue &queue, transpose transa, int64_t m, int64_t #endif } -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, - float *alpha, const float **a, int64_t *lda, const float **x, - int64_t *incx, float *beta, float **y, int64_t *incy, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, float* alpha, + const float** a, int64_t* lda, const float** x, int64_t* incx, float* beta, + float** y, int64_t* incy, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); #endif @@ -815,11 +802,10 @@ sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_ #endif } -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, - double *alpha, const double **a, int64_t *lda, const double **x, - int64_t *incx, double *beta, double **y, int64_t *incy, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, double* alpha, + const double** a, int64_t* lda, const double** x, int64_t* incx, + double* beta, double** y, int64_t* incy, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); #endif @@ -828,11 +814,11 @@ sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_ #endif } -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - const std::complex **x, int64_t *incx, std::complex *beta, - std::complex **y, int64_t *incy, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + const std::complex** x, int64_t* incx, std::complex* beta, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); #endif @@ -841,12 +827,11 @@ sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_ #endif } -sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, - int64_t *lda, const std::complex **x, int64_t *incx, - std::complex *beta, std::complex **y, int64_t *incy, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, transpose* transa, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, int64_t* lda, + const std::complex** x, int64_t* incx, std::complex* beta, + std::complex** y, int64_t* incy, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemv_batch", "for column_major layout"); #endif @@ -855,10 +840,10 @@ sycl::event gemv_batch(sycl::queue &queue, transpose *transa, int64_t *m, int64_ #endif } -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const float *a, int64_t lda, int64_t stride_a, const float *x, - int64_t incx, int64_t stride_x, float *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, const float* a, + int64_t lda, int64_t stride_a, const float* x, int64_t incx, + int64_t stride_x, float* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); #endif @@ -867,10 +852,10 @@ sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n #endif } -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const double *a, int64_t lda, int64_t stride_a, const double *x, - int64_t incx, int64_t stride_x, double *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, const double* a, + int64_t lda, int64_t stride_a, const double* x, int64_t incx, + int64_t stride_x, double* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); #endif @@ -879,11 +864,11 @@ sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n #endif } -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const std::complex *a, int64_t lda, int64_t stride_a, - const std::complex *x, int64_t incx, int64_t stride_x, - std::complex *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + const std::complex* a, int64_t lda, int64_t stride_a, + const std::complex* x, int64_t incx, int64_t stride_x, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); #endif @@ -892,11 +877,11 @@ sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n #endif } -sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, - const std::complex *a, int64_t lda, int64_t stride_a, - const std::complex *x, int64_t incx, int64_t stride_x, - std::complex *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, + const std::complex* a, int64_t lda, int64_t stride_a, + const std::complex* x, int64_t incx, int64_t stride_x, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); #endif @@ -905,10 +890,10 @@ sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n #endif } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const float **a, int64_t *lda, const float **x, int64_t *incx, float **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const float** a, int64_t* lda, const float** x, int64_t* incx, float** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); #endif @@ -917,10 +902,10 @@ sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t #endif } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const double **a, int64_t *lda, const double **x, int64_t *incx, - double **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const double** a, int64_t* lda, const double** x, int64_t* incx, double** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); #endif @@ -929,11 +914,10 @@ sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t #endif } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const std::complex **a, int64_t *lda, - const std::complex **x, int64_t *incx, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const std::complex** a, int64_t* lda, const std::complex** x, + int64_t* incx, std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); #endif @@ -942,11 +926,10 @@ sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t #endif } -sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, - const std::complex **a, int64_t *lda, - const std::complex **x, int64_t *incx, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, + const std::complex** a, int64_t* lda, const std::complex** x, + int64_t* incx, std::complex** c, int64_t* ldc, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "dgmm_batch", "for column_major layout"); #endif @@ -955,11 +938,11 @@ sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const float **a, int64_t *lda, - const float **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const float** a, int64_t* lda, + const float** b, int64_t* ldb, float* beta, float** c, int64_t* ldc, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -968,11 +951,11 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, double *alpha, const double **a, int64_t *lda, - const double **b, int64_t *ldb, double *beta, double **c, int64_t *ldc, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, double* alpha, const double** a, int64_t* lda, + const double** b, int64_t* ldb, double* beta, double** c, int64_t* ldc, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -981,12 +964,12 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, std::complex *alpha, - const std::complex **a, int64_t *lda, - const std::complex **b, int64_t *ldb, std::complex *beta, - std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, std::complex* alpha, + const std::complex** a, int64_t* lda, const std::complex** b, + int64_t* ldb, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -995,12 +978,12 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, std::complex *alpha, - const std::complex **a, int64_t *lda, - const std::complex **b, int64_t *ldb, std::complex *beta, - std::complex **c, int64_t *ldc, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, std::complex* alpha, + const std::complex** a, int64_t* lda, const std::complex** b, + int64_t* ldb, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1009,11 +992,11 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, sycl::half *alpha, const sycl::half **a, - int64_t *lda, const sycl::half **b, int64_t *ldb, sycl::half *beta, - sycl::half **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, sycl::half* alpha, const sycl::half** a, + int64_t* lda, const sycl::half** b, int64_t* ldb, sycl::half* beta, + sycl::half** c, int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1022,11 +1005,11 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const sycl::half **a, int64_t *lda, - const sycl::half **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const sycl::half** a, int64_t* lda, + const sycl::half** b, int64_t* ldb, float* beta, float** c, int64_t* ldc, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1035,11 +1018,11 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const std::int8_t **a, int64_t *lda, - const std::int8_t **b, int64_t *ldb, float *beta, float **c, int64_t *ldc, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const std::int8_t** a, int64_t* lda, + const std::int8_t** b, int64_t* ldb, float* beta, float** c, int64_t* ldc, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1048,11 +1031,11 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, - int64_t *n, int64_t *k, float *alpha, const std::int8_t **a, int64_t *lda, - const std::int8_t **b, int64_t *ldb, float *beta, std::int32_t **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, + int64_t* n, int64_t* k, float* alpha, const std::int8_t** a, int64_t* lda, + const std::int8_t** b, int64_t* ldb, float* beta, std::int32_t** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1061,11 +1044,11 @@ sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, float alpha, const float *a, int64_t lda, - int64_t stride_a, const float *b, int64_t ldb, int64_t stride_b, - float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const float* a, int64_t lda, int64_t stride_a, + const float* b, int64_t ldb, int64_t stride_b, float beta, float* c, + int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1074,11 +1057,11 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, double alpha, const double *a, int64_t lda, - int64_t stride_a, const double *b, int64_t ldb, int64_t stride_b, - double beta, double *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, double alpha, const double* a, int64_t lda, int64_t stride_a, + const double* b, int64_t ldb, int64_t stride_b, double beta, double* c, + int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1087,13 +1070,12 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex beta, std::complex *c, int64_t ldc, - int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, int64_t stride_a, const std::complex* b, int64_t ldb, + int64_t stride_b, std::complex beta, std::complex* c, + int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1102,13 +1084,12 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, std::complex alpha, - const std::complex *a, int64_t lda, int64_t stride_a, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex beta, std::complex *c, int64_t ldc, - int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, int64_t stride_a, const std::complex* b, int64_t ldb, + int64_t stride_b, std::complex beta, std::complex* c, + int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1117,11 +1098,11 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, sycl::half alpha, const sycl::half *a, int64_t lda, - int64_t stride_a, const sycl::half *b, int64_t ldb, int64_t stride_b, - sycl::half beta, sycl::half *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, sycl::half alpha, const sycl::half* a, int64_t lda, + int64_t stride_a, const sycl::half* b, int64_t ldb, int64_t stride_b, + sycl::half beta, sycl::half* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1130,11 +1111,11 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const sycl::half *a, int64_t lda, int64_t stride_a, - const sycl::half *b, int64_t ldb, int64_t stride_b, float beta, float *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const sycl::half* a, int64_t lda, int64_t stride_a, + const sycl::half* b, int64_t ldb, int64_t stride_b, float beta, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1143,11 +1124,11 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const std::int8_t *a, int64_t lda, int64_t stride_a, - const std::int8_t *b, int64_t ldb, int64_t stride_b, float beta, float *c, +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const std::int8_t* a, int64_t lda, int64_t stride_a, + const std::int8_t* b, int64_t ldb, int64_t stride_b, float beta, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1156,11 +1137,11 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i #endif } -sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const std::int8_t *a, int64_t lda, int64_t stride_a, - const std::int8_t *b, int64_t ldb, int64_t stride_b, float beta, - std::int32_t *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const std::int8_t* a, int64_t lda, int64_t stride_a, + const std::int8_t* b, int64_t ldb, int64_t stride_b, float beta, + std::int32_t* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_batch", "for column_major layout"); #endif @@ -1169,11 +1150,10 @@ sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, i #endif } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, - int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, float alpha, const float* a, + int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); #endif @@ -1182,11 +1162,10 @@ sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, #endif } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, - int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, double alpha, const double* a, + int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); #endif @@ -1195,11 +1174,11 @@ sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, #endif } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, std::complex alpha, + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); #endif @@ -1208,11 +1187,11 @@ sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, #endif } -sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, std::complex alpha, + const std::complex* a, int64_t lda, int64_t stride_a, + std::complex* b, int64_t ldb, int64_t stride_b, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); #endif @@ -1221,11 +1200,10 @@ sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, #endif } -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, - transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, float *alpha, - const float **a, int64_t *lda, float **b, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, float* alpha, const float** a, + int64_t* lda, float** b, int64_t* ldb, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); #endif @@ -1234,11 +1212,10 @@ sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, #endif } -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, - transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, double *alpha, - const double **a, int64_t *lda, double **b, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, double* alpha, const double** a, + int64_t* lda, double** b, int64_t* ldb, int64_t group_count, + int64_t* groupsize, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); #endif @@ -1247,11 +1224,11 @@ sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, #endif } -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, - transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, int64_t *lda, - std::complex **b, int64_t *ldb, int64_t group_count, - int64_t *groupsize, const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, std::complex* alpha, + const std::complex** a, int64_t* lda, std::complex** b, + int64_t* ldb, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); #endif @@ -1260,12 +1237,11 @@ sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, #endif } -sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, - transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex **b, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, transpose* trans, + diag* unit_diag, int64_t* m, int64_t* n, std::complex* alpha, + const std::complex** a, int64_t* lda, std::complex** b, + int64_t* ldb, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "trsm_batch", "for column_major layout"); #endif @@ -1274,10 +1250,10 @@ sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, #endif } -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, float *alpha, const float **a, int64_t *lda, float *beta, - float **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, float* alpha, const float** a, int64_t* lda, float* beta, + float** c, int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); #endif @@ -1286,10 +1262,10 @@ sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, #endif } -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, double *alpha, const double **a, int64_t *lda, double *beta, - double **c, int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, double* alpha, const double** a, int64_t* lda, double* beta, + double** c, int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); #endif @@ -1298,11 +1274,11 @@ sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, #endif } -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); #endif @@ -1311,11 +1287,11 @@ sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, #endif } -sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, - int64_t *k, std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex *beta, std::complex **c, - int64_t *ldc, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, + int64_t* k, std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex* beta, std::complex** c, + int64_t* ldc, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); #endif @@ -1324,10 +1300,10 @@ sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, #endif } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, float alpha, const float *a, int64_t lda, int64_t stride_a, - float beta, float *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, const float* a, int64_t lda, int64_t stride_a, float beta, + float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); #endif @@ -1336,10 +1312,10 @@ sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, in #endif } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, double alpha, const double *a, int64_t lda, int64_t stride_a, - double beta, double *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, const double* a, int64_t lda, int64_t stride_a, double beta, + double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); #endif @@ -1348,11 +1324,11 @@ sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, in #endif } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, int64_t stride_a, std::complex beta, - std::complex *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex beta, std::complex* c, + int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); #endif @@ -1361,11 +1337,11 @@ sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, in #endif } -sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, int64_t stride_a, std::complex beta, - std::complex *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex beta, std::complex* c, + int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "syrk_batch", "for column_major layout"); #endif @@ -1374,10 +1350,10 @@ sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, in #endif } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, int64_t stride_a, float *b, int64_t ldb, +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, int64_t stride_a, float* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); #endif @@ -1386,10 +1362,10 @@ sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 #endif } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, int64_t stride_a, double *b, int64_t ldb, +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, int64_t stride_a, double* b, int64_t ldb, int64_t stride_b, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); #endif @@ -1398,10 +1374,10 @@ sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 #endif } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); #endif @@ -1410,10 +1386,10 @@ sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 #endif } -sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stride_a, std::complex *b, int64_t ldb, int64_t stride_b, - int64_t batch_size, const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stride_a, std::complex* b, int64_t ldb, int64_t stride_b, + int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy_batch", "for column_major layout"); #endif @@ -1422,9 +1398,9 @@ sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 #endif } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); #endif @@ -1433,9 +1409,9 @@ sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 #endif } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); #endif @@ -1444,10 +1420,10 @@ sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 #endif } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); #endif @@ -1456,10 +1432,10 @@ sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 #endif } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); #endif @@ -1468,11 +1444,11 @@ sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64 #endif } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, float alpha, const float *a, int64_t lda, int64_t stride_a, - float beta, const float *b, int64_t ldb, int64_t stride_b, float *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, float alpha, const float* a, int64_t lda, int64_t stride_a, + float beta, const float* b, int64_t ldb, int64_t stride_b, float* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd_batch", "for column_major layout"); #endif @@ -1481,11 +1457,11 @@ sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb #endif } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, double alpha, const double *a, int64_t lda, int64_t stride_a, - double beta, const double *b, int64_t ldb, int64_t stride_b, double *c, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, double alpha, const double* a, int64_t lda, int64_t stride_a, + double beta, const double* b, int64_t ldb, int64_t stride_b, double* c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd_batch", "for column_major layout"); #endif @@ -1494,12 +1470,12 @@ sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb #endif } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, int64_t batch_size, - const std::vector &dependencies) { + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, int64_t batch_size, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd_batch", "for column_major layout"); #endif @@ -1508,12 +1484,12 @@ sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb #endif } -sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, std::complex alpha, const std::complex *a, +sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, std::complex alpha, const std::complex* a, int64_t lda, int64_t stride_a, std::complex beta, - const std::complex *b, int64_t ldb, int64_t stride_b, - std::complex *c, int64_t ldc, int64_t stride_c, - int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, int64_t ldb, int64_t stride_b, + std::complex* c, int64_t ldc, int64_t stride_c, + int64_t batch_size, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd_batch", "for column_major layout"); #endif diff --git a/src/blas/backends/netlib/netlib_common.hpp b/src/blas/backends/netlib/netlib_common.hpp index 51552f2ce..b1848d277 100644 --- a/src/blas/backends/netlib/netlib_common.hpp +++ b/src/blas/backends/netlib/netlib_common.hpp @@ -79,19 +79,19 @@ inline CBLAS_OFFSET convert_to_cblas_offset(offset offsetc) { // host_task automatically uses run_on_host_intel if it is supported by the // compiler. Otherwise, it falls back to single_task. template -static inline auto host_task_internal(H &cgh, F f, int) -> decltype(cgh.host_task(f)) { +static inline auto host_task_internal(H& cgh, F f, int) -> decltype(cgh.host_task(f)) { return cgh.host_task(f); } template -static inline void host_task_internal(H &cgh, F f, long) { +static inline void host_task_internal(H& cgh, F f, long) { #ifndef __SYCL_DEVICE_ONLY__ cgh.template single_task(f); #endif } template -static inline void host_task(H &cgh, F f) { +static inline void host_task(H& cgh, F f) { (void)host_task_internal(cgh, f, 0); } diff --git a/src/blas/backends/netlib/netlib_extensions.cxx b/src/blas/backends/netlib/netlib_extensions.cxx index 8e94cb880..d0c13ebbd 100644 --- a/src/blas/backends/netlib/netlib_extensions.cxx +++ b/src/blas/backends/netlib/netlib_extensions.cxx @@ -19,11 +19,10 @@ // Buffer APIs -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, - int64_t lda, int8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, - float beta, sycl::buffer &c, int64_t ldc, - sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_bias", "for column_major layout"); #endif @@ -32,11 +31,10 @@ void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset of #endif } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, - int64_t lda, int8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, - float beta, sycl::buffer &c, int64_t ldc, - sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_bias", "for column_major layout"); #endif @@ -45,11 +43,10 @@ void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset of #endif } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, - int64_t lda, uint8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, - float beta, sycl::buffer &c, int64_t ldc, - sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_bias", "for column_major layout"); #endif @@ -58,11 +55,10 @@ void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset of #endif } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, sycl::buffer &a, - int64_t lda, uint8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, - float beta, sycl::buffer &c, int64_t ldc, - sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_bias", "for column_major layout"); #endif @@ -71,9 +67,9 @@ void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset of #endif } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemmt", "for column_major layout"); @@ -83,9 +79,9 @@ void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose tra #endif } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, double beta, sycl::buffer& c, int64_t ldc) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemmt", "for column_major layout"); @@ -95,10 +91,10 @@ void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose tra #endif } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemmt", "for column_major layout"); #endif @@ -107,10 +103,10 @@ void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose tra #endif } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemmt", "for column_major layout"); #endif @@ -119,8 +115,8 @@ void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose tra #endif } -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy", "for column_major layout"); #endif @@ -129,8 +125,8 @@ void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float a #endif } -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy", "for column_major layout"); #endif @@ -139,9 +135,9 @@ void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double #endif } -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb) { +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy", "for column_major layout"); #endif @@ -150,9 +146,9 @@ void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::co #endif } -void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb) { +void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy", "for column_major layout"); #endif @@ -161,9 +157,9 @@ void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::co #endif } -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb) { +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy2", "for column_major layout"); #endif @@ -172,9 +168,9 @@ void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float #endif } -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb) { +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy2", "for column_major layout"); #endif @@ -183,9 +179,9 @@ void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double #endif } -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &b, int64_t ldb, std::int64_t strideb) { +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& b, int64_t ldb, std::int64_t strideb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy2", "for column_major layout"); #endif @@ -194,9 +190,9 @@ void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::c #endif } -void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, int64_t ldb, +void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, int64_t ldb, std::int64_t strideb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy2", "for column_major layout"); @@ -206,8 +202,8 @@ void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy", "for column_major layout"); #endif @@ -216,8 +212,8 @@ void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float a #endif } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy", "for column_major layout"); #endif @@ -226,8 +222,8 @@ void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double #endif } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy", "for column_major layout"); #endif @@ -236,8 +232,8 @@ void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::co #endif } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy", "for column_major layout"); #endif @@ -246,9 +242,9 @@ void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::co #endif } -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, sycl::buffer &a, int64_t lda, float beta, - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, sycl::buffer& a, int64_t lda, float beta, + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd", "for column_major layout"); #endif @@ -257,9 +253,9 @@ void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, #endif } -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, sycl::buffer &a, int64_t lda, double beta, - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, sycl::buffer& a, int64_t lda, double beta, + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd", "for column_major layout"); #endif @@ -268,10 +264,10 @@ void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, #endif } -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &b, int64_t ldb, - sycl::buffer, 1> &c, int64_t ldc) { +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& b, int64_t ldb, + sycl::buffer, 1>& c, int64_t ldc) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd", "for column_major layout"); #endif @@ -280,10 +276,10 @@ void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, #endif } -void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &b, int64_t ldb, - sycl::buffer, 1> &c, int64_t ldc) { +void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& b, int64_t ldb, + sycl::buffer, 1>& c, int64_t ldc) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd", "for column_major layout"); #endif @@ -294,11 +290,11 @@ void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, // USM APIs -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, - const int8_t *a, int64_t lda, int8_t ao, const int8_t *b, int64_t ldb, - int8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_bias", "for column_major layout"); #endif @@ -307,11 +303,11 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, #endif } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, - const int8_t *a, int64_t lda, int8_t ao, const uint8_t *b, int64_t ldb, - uint8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_bias", "for column_major layout"); #endif @@ -320,11 +316,11 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, #endif } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, - const uint8_t *a, int64_t lda, uint8_t ao, const int8_t *b, int64_t ldb, - int8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_bias", "for column_major layout"); #endif @@ -333,11 +329,11 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, #endif } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, - offset offsetc, int64_t m, int64_t n, int64_t k, float alpha, - const uint8_t *a, int64_t lda, uint8_t ao, const uint8_t *b, int64_t ldb, - uint8_t bo, float beta, int32_t *c, int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm_bias", "for column_major layout"); #endif @@ -346,10 +342,10 @@ sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, #endif } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, float alpha, const float *a, int64_t lda, - const float *b, int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, float alpha, const float* a, int64_t lda, const float* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemmt", "for column_major layout"); #endif @@ -358,10 +354,10 @@ sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transp #endif } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, double alpha, const double *a, int64_t lda, - const double *b, int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemmt", "for column_major layout"); #endif @@ -370,11 +366,11 @@ sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transp #endif } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemmt", "for column_major layout"); #endif @@ -383,11 +379,11 @@ sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transp #endif } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, - const std::complex *a, int64_t lda, const std::complex *b, - int64_t ldb, std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemmt", "for column_major layout"); #endif @@ -396,9 +392,9 @@ sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transp #endif } -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, float *b, int64_t ldb, - const std::vector &dependencies) { +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, float* b, int64_t ldb, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy", "for column_major layout"); #endif @@ -407,9 +403,9 @@ sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, double *b, int64_t ldb, - const std::vector &dependencies) { +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, double* b, int64_t ldb, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy", "for column_major layout"); #endif @@ -418,10 +414,10 @@ sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb, - const std::vector &dependencies) { +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex* b, int64_t ldb, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy", "for column_major layout"); #endif @@ -430,10 +426,10 @@ sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb, - const std::vector &dependencies) { +sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex* b, int64_t ldb, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy", "for column_major layout"); #endif @@ -442,9 +438,9 @@ sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, std::int64_t stridea, float *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, std::int64_t stridea, float* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy2", "for column_major layout"); #endif @@ -453,9 +449,9 @@ sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, std::int64_t stridea, double *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, std::int64_t stridea, double* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy2", "for column_major layout"); #endif @@ -464,10 +460,10 @@ sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::int64_t stridea, std::complex *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::int64_t stridea, std::complex* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy2", "for column_major layout"); #endif @@ -476,10 +472,10 @@ sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::int64_t stridea, std::complex *b, int64_t ldb, - std::int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::int64_t stridea, std::complex* b, int64_t ldb, + std::int64_t strideb, const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatcopy2", "for column_major layout"); #endif @@ -488,9 +484,9 @@ sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy", "for column_major layout"); #endif @@ -499,9 +495,9 @@ sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy", "for column_major layout"); #endif @@ -510,9 +506,9 @@ sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy", "for column_major layout"); #endif @@ -521,9 +517,9 @@ sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "imatcopy", "for column_major layout"); #endif @@ -532,10 +528,10 @@ sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, #endif } -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - float alpha, const float *a, int64_t lda, float beta, const float *b, - int64_t ldb, float *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + float alpha, const float* a, int64_t lda, float beta, const float* b, + int64_t ldb, float* c, int64_t ldc, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd", "for column_major layout"); #endif @@ -544,10 +540,10 @@ sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int6 #endif } -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - double alpha, const double *a, int64_t lda, double beta, const double *b, - int64_t ldb, double *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + double alpha, const double* a, int64_t lda, double beta, const double* b, + int64_t ldb, double* c, int64_t ldc, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd", "for column_major layout"); #endif @@ -556,11 +552,11 @@ sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int6 #endif } -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, const std::complex *b, int64_t ldb, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, const std::complex* b, int64_t ldb, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd", "for column_major layout"); #endif @@ -569,11 +565,11 @@ sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int6 #endif } -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, const std::complex *b, int64_t ldb, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, const std::complex* b, int64_t ldb, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "omatadd", "for column_major layout"); #endif @@ -581,5 +577,3 @@ sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int6 throw unimplemented("blas", "omatadd", "for row_major layout"); #endif } - - diff --git a/src/blas/backends/netlib/netlib_level1.cpp b/src/blas/backends/netlib/netlib_level1.cpp index ec6b1eb09..aa55ec419 100644 --- a/src/blas/backends/netlib/netlib_level1.cpp +++ b/src/blas/backends/netlib/netlib_level1.cpp @@ -43,7 +43,7 @@ inline double abs_val(std::complex val) { return std::abs(val.real()) + std::abs(val.imag()); } -int cblas_isamin(int n, const float *x, int incx) { +int cblas_isamin(int n, const float* x, int incx) { if (n < 1 || incx < 1) { return 0; } @@ -65,7 +65,7 @@ int cblas_isamin(int n, const float *x, int incx) { return min_idx; } -int cblas_idamin(int n, const double *x, int incx) { +int cblas_idamin(int n, const double* x, int incx) { if (n < 1 || incx < 1) { return 0; } @@ -87,7 +87,7 @@ int cblas_idamin(int n, const double *x, int incx) { return min_idx; } -int cblas_icamin(int n, const std::complex *x, int incx) { +int cblas_icamin(int n, const std::complex* x, int incx) { if (n < 1 || incx < 1) { return 0; } @@ -109,7 +109,7 @@ int cblas_icamin(int n, const std::complex *x, int incx) { return min_idx; } -int cblas_izamin(int n, const std::complex *x, int incx) { +int cblas_izamin(int n, const std::complex* x, int incx) { if (n < 1 || incx < 1) { return 0; } @@ -131,7 +131,7 @@ int cblas_izamin(int n, const std::complex *x, int incx) { return min_idx; } -void cblas_csrot(const int n, std::complex *cx, const int incx, std::complex *cy, +void cblas_csrot(const int n, std::complex* cx, const int incx, std::complex* cy, const int incy, const float c, const float s) { if (n < 1) return; @@ -158,7 +158,7 @@ void cblas_csrot(const int n, std::complex *cx, const int incx, std::comp } } -void cblas_zdrot(const int n, std::complex *zx, const int incx, std::complex *zy, +void cblas_zdrot(const int n, std::complex* zx, const int incx, std::complex* zy, const int incy, const double c, const double s) { if (n < 1) return; @@ -185,8 +185,8 @@ void cblas_zdrot(const int n, std::complex *zx, const int incx, std::com } } -void cblas_crotg(std::complex *ca, const std::complex *cb, float *c, - std::complex *s) { +void cblas_crotg(std::complex* ca, const std::complex* cb, float* c, + std::complex* s) { if (std::abs(ca[0]) == 0) { c[0] = 0.0; s[0] = std::complex(1.0, 0.0); @@ -203,8 +203,8 @@ void cblas_crotg(std::complex *ca, const std::complex *cb, float * } } -void cblas_zrotg(std::complex *ca, const std::complex *cb, double *c, - std::complex *s) { +void cblas_zrotg(std::complex* ca, const std::complex* cb, double* c, + std::complex* s) { if (std::abs(ca[0]) == 0) { c[0] = 0.0; s[0] = std::complex(1.0, 0.0); diff --git a/src/blas/backends/netlib/netlib_level1.cxx b/src/blas/backends/netlib/netlib_level1.cxx index 9f953dc5b..5514a86c1 100644 --- a/src/blas/backends/netlib/netlib_level1.cxx +++ b/src/blas/backends/netlib/netlib_level1.cxx @@ -19,9 +19,9 @@ // Buffer APIs -void asum(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void asum(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -31,9 +31,9 @@ void asum(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx }); } -void asum(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void asum(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -43,9 +43,9 @@ void asum(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t inc }); } -void asum(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void asum(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -55,9 +55,9 @@ void asum(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x }); } -void asum(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void asum(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -67,9 +67,9 @@ void asum(sycl::queue &queue, int64_t n, sycl::buffer, 1> & }); } -void axpy(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void axpy(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -79,9 +79,9 @@ void axpy(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, }); } -void axpy(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void axpy(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -91,34 +91,34 @@ void axpy(sycl::queue &queue, int64_t n, double alpha, sycl::buffer & }); } -void axpy(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void axpy(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { - ::cblas_caxpy((const int)n, (const void *)&alpha, accessor_x.GET_MULTI_PTR, + ::cblas_caxpy((const int)n, (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void axpy(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void axpy(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { - ::cblas_zaxpy((const int)n, (const void *)&alpha, accessor_x.GET_MULTI_PTR, + ::cblas_zaxpy((const int)n, (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void axpby(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - float beta, sycl::buffer &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + float beta, sycl::buffer& y, int64_t incy) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpby", "for column_major layout"); #endif @@ -127,8 +127,8 @@ void axpby(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x #endif } -void axpby(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx, - double beta, sycl::buffer &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, int64_t incx, + double beta, sycl::buffer& y, int64_t incy) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpby", "for column_major layout"); #endif @@ -137,9 +137,9 @@ void axpby(sycl::queue &queue, int64_t n, double alpha, sycl::buffer #endif } -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpby", "for column_major layout"); #endif @@ -148,9 +148,9 @@ void axpby(sycl::queue &queue, int64_t n, std::complex alpha, #endif } -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpby", "for column_major layout"); #endif @@ -159,9 +159,9 @@ void axpby(sycl::queue &queue, int64_t n, std::complex alpha, #endif } -void copy(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void copy(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -171,9 +171,9 @@ void copy(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx }); } -void copy(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void copy(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -183,9 +183,9 @@ void copy(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t inc }); } -void copy(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void copy(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -195,9 +195,9 @@ void copy(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x }); } -void copy(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void copy(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -207,9 +207,9 @@ void copy(sycl::queue &queue, int64_t n, sycl::buffer, 1> & }); } -void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void dot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_result = result.get_access(cgh); @@ -221,9 +221,9 @@ void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, }); } -void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void dot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_result = result.get_access(cgh); @@ -235,9 +235,9 @@ void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx }); } -void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void dot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_result = result.get_access(cgh); @@ -249,10 +249,10 @@ void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, }); } -void dotc(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result) { - queue.submit([&](sycl::handler &cgh) { +void dotc(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_result = result.get_access(cgh); @@ -264,10 +264,10 @@ void dotc(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x }); } -void dotc(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result) { - queue.submit([&](sycl::handler &cgh) { +void dotc(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_result = result.get_access(cgh); @@ -279,10 +279,10 @@ void dotc(sycl::queue &queue, int64_t n, sycl::buffer, 1> & }); } -void dotu(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result) { - queue.submit([&](sycl::handler &cgh) { +void dotu(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_result = result.get_access(cgh); @@ -294,10 +294,10 @@ void dotu(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x }); } -void dotu(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &result) { - queue.submit([&](sycl::handler &cgh) { +void dotu(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_result = result.get_access(cgh); @@ -309,9 +309,9 @@ void dotu(sycl::queue &queue, int64_t n, sycl::buffer, 1> & }); } -void iamin(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void iamin(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -320,9 +320,9 @@ void iamin(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t inc }); } -void iamin(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void iamin(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.template get_access(cgh); auto accessor_result = result.template get_access(cgh); host_task(cgh, [=]() { @@ -331,9 +331,9 @@ void iamin(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t in }); } -void iamin(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void iamin(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -342,9 +342,9 @@ void iamin(sycl::queue &queue, int64_t n, sycl::buffer, 1> & }); } -void iamin(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void iamin(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -353,9 +353,9 @@ void iamin(sycl::queue &queue, int64_t n, sycl::buffer, 1> }); } -void iamax(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void iamax(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -364,9 +364,9 @@ void iamax(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t inc }); } -void iamax(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void iamax(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -375,9 +375,9 @@ void iamax(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t in }); } -void iamax(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void iamax(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -386,9 +386,9 @@ void iamax(sycl::queue &queue, int64_t n, sycl::buffer, 1> & }); } -void iamax(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void iamax(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -397,9 +397,9 @@ void iamax(sycl::queue &queue, int64_t n, sycl::buffer, 1> }); } -void nrm2(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void nrm2(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.template get_access(cgh); auto accessor_result = result.template get_access(cgh); host_task(cgh, [=]() { @@ -409,9 +409,9 @@ void nrm2(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx }); } -void nrm2(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void nrm2(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -421,9 +421,9 @@ void nrm2(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t inc }); } -void nrm2(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void nrm2(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -433,9 +433,9 @@ void nrm2(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x }); } -void nrm2(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void nrm2(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_result = result.get_access(cgh); host_task(cgh, [=]() { @@ -445,9 +445,9 @@ void nrm2(sycl::queue &queue, int64_t n, sycl::buffer, 1> & }); } -void rot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, float c, float s) { - queue.submit([&](sycl::handler &cgh) { +void rot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, float c, float s) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -457,9 +457,9 @@ void rot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, }); } -void rot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, double c, double s) { - queue.submit([&](sycl::handler &cgh) { +void rot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, double c, double s) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -469,9 +469,9 @@ void rot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx }); } -void rot(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, float c, float s) { - queue.submit([&](sycl::handler &cgh) { +void rot(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, float c, float s) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -482,9 +482,9 @@ void rot(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, }); } -void rot(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, double c, double s) { - queue.submit([&](sycl::handler &cgh) { +void rot(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, double c, double s) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -495,9 +495,9 @@ void rot(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x }); } -void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s) { - queue.submit([&](sycl::handler &cgh) { +void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); @@ -509,9 +509,9 @@ void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer }); } -void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s) { - queue.submit([&](sycl::handler &cgh) { +void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); @@ -523,10 +523,10 @@ void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { - queue.submit([&](sycl::handler &cgh) { +void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); @@ -538,10 +538,10 @@ void rotg(sycl::queue &queue, sycl::buffer, 1> &a, }); } -void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { - queue.submit([&](sycl::handler &cgh) { +void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); @@ -553,9 +553,9 @@ void rotg(sycl::queue &queue, sycl::buffer, 1> &a, }); } -void rotm(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { - queue.submit([&](sycl::handler &cgh) { +void rotm(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& param) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_param = param.get_access(cgh); @@ -566,9 +566,9 @@ void rotm(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx }); } -void rotm(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { - queue.submit([&](sycl::handler &cgh) { +void rotm(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& param) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_param = param.get_access(cgh); @@ -579,9 +579,9 @@ void rotm(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t inc }); } -void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, float y1, sycl::buffer ¶m) { - queue.submit([&](sycl::handler &cgh) { +void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, float y1, sycl::buffer& param) { + queue.submit([&](sycl::handler& cgh) { auto accessor_d1 = d1.get_access(cgh); auto accessor_d2 = d2.get_access(cgh); auto accessor_x1 = x1.get_access(cgh); @@ -593,9 +593,9 @@ void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, double y1, sycl::buffer ¶m) { - queue.submit([&](sycl::handler &cgh) { +void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, double y1, sycl::buffer& param) { + queue.submit([&](sycl::handler& cgh) { auto accessor_d1 = d1.get_access(cgh); auto accessor_d2 = d2.get_access(cgh); auto accessor_x1 = x1.get_access(cgh); @@ -607,8 +607,8 @@ void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void scal(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { ::cblas_sscal((const int)n, (const float)alpha, accessor_x.GET_MULTI_PTR, @@ -617,8 +617,8 @@ void scal(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, }); } -void scal(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void scal(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { ::cblas_dscal((const int)n, (const double)alpha, accessor_x.GET_MULTI_PTR, @@ -627,20 +627,20 @@ void scal(sycl::queue &queue, int64_t n, double alpha, sycl::buffer & }); } -void scal(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void scal(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { - ::cblas_cscal((const int)n, (const void *)&alpha, accessor_x.GET_MULTI_PTR, + ::cblas_cscal((const int)n, (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)std::abs(incx)); }); }); } -void scal(sycl::queue &queue, int64_t n, float alpha, sycl::buffer, 1> &x, +void scal(sycl::queue& queue, int64_t n, float alpha, sycl::buffer, 1>& x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { ::cblas_csscal((const int)n, (const float)alpha, accessor_x.GET_MULTI_PTR, @@ -649,20 +649,20 @@ void scal(sycl::queue &queue, int64_t n, float alpha, sycl::buffer alpha, - sycl::buffer, 1> &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void scal(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { - ::cblas_zscal((const int)n, (const void *)&alpha, accessor_x.GET_MULTI_PTR, + ::cblas_zscal((const int)n, (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)std::abs(incx)); }); }); } -void scal(sycl::queue &queue, int64_t n, double alpha, sycl::buffer, 1> &x, +void scal(sycl::queue& queue, int64_t n, double alpha, sycl::buffer, 1>& x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { ::cblas_zdscal((const int)n, (const double)alpha, accessor_x.GET_MULTI_PTR, @@ -671,9 +671,9 @@ void scal(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { - queue.submit([&](sycl::handler &cgh) { +void sdsdot(sycl::queue& queue, int64_t n, float sb, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_result = result.get_access(cgh); @@ -685,9 +685,9 @@ void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, }); } -void swap(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void swap(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -697,9 +697,9 @@ void swap(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx }); } -void swap(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void swap(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -709,9 +709,9 @@ void swap(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t inc }); } -void swap(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void swap(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -721,9 +721,9 @@ void swap(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x }); } -void swap(sycl::queue &queue, int64_t n, sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void swap(sycl::queue& queue, int64_t n, sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { @@ -735,9 +735,9 @@ void swap(sycl::queue &queue, int64_t n, sycl::buffer, 1> & // USM APIs -sycl::event asum(sycl::queue &queue, int64_t n, const float *x, int64_t incx, float *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event asum(sycl::queue& queue, int64_t n, const float* x, int64_t incx, float* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -748,9 +748,9 @@ sycl::event asum(sycl::queue &queue, int64_t n, const float *x, int64_t incx, fl return done; } -sycl::event asum(sycl::queue &queue, int64_t n, const double *x, int64_t incx, double *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event asum(sycl::queue& queue, int64_t n, const double* x, int64_t incx, double* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -761,9 +761,9 @@ sycl::event asum(sycl::queue &queue, int64_t n, const double *x, int64_t incx, d return done; } -sycl::event asum(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - float *result, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event asum(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + float* result, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -774,9 +774,9 @@ sycl::event asum(sycl::queue &queue, int64_t n, const std::complex *x, in return done; } -sycl::event asum(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - double *result, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event asum(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + double* result, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -787,9 +787,9 @@ sycl::event asum(sycl::queue &queue, int64_t n, const std::complex *x, i return done; } -sycl::event axpy(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, float *y, - int64_t incy, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event axpy(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, float* y, + int64_t incy, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -801,9 +801,9 @@ sycl::event axpy(sycl::queue &queue, int64_t n, float alpha, const float *x, int return done; } -sycl::event axpy(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - double *y, int64_t incy, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event axpy(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + double* y, int64_t incy, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -816,41 +816,41 @@ sycl::event axpy(sycl::queue &queue, int64_t n, double alpha, const double *x, i return done; } -sycl::event axpy(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event axpy(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { - ::cblas_caxpy((const int)n, (const void *)&alpha, x, (const int)incx, y, + ::cblas_caxpy((const int)n, (const void*)&alpha, x, (const int)incx, y, (const int)incy); }); }); return done; } -sycl::event axpy(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event axpy(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { - ::cblas_zaxpy((const int)n, (const void *)&alpha, x, (const int)incx, y, + ::cblas_zaxpy((const int)n, (const void*)&alpha, x, (const int)incx, y, (const int)incy); }); }); return done; } -sycl::event axpby(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, - float beta, float *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, + float beta, float* y, int64_t incy, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpby", "for column_major layout"); #endif @@ -859,9 +859,9 @@ sycl::event axpby(sycl::queue &queue, int64_t n, float alpha, const float *x, in #endif } -sycl::event axpby(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - double beta, double *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + double beta, double* y, int64_t incy, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpby", "for column_major layout"); #endif @@ -870,10 +870,10 @@ sycl::event axpby(sycl::queue &queue, int64_t n, double alpha, const double *x, #endif } -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpby", "for column_major layout"); #endif @@ -882,10 +882,10 @@ sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, #endif } -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "axpby", "for column_major layout"); #endif @@ -894,9 +894,9 @@ sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, #endif } -sycl::event copy(sycl::queue &queue, int64_t n, const float *x, int64_t incx, float *y, - int64_t incy, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event copy(sycl::queue& queue, int64_t n, const float* x, int64_t incx, float* y, + int64_t incy, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -907,9 +907,9 @@ sycl::event copy(sycl::queue &queue, int64_t n, const float *x, int64_t incx, fl return done; } -sycl::event copy(sycl::queue &queue, int64_t n, const double *x, int64_t incx, double *y, - int64_t incy, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event copy(sycl::queue& queue, int64_t n, const double* x, int64_t incx, double* y, + int64_t incy, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -920,10 +920,10 @@ sycl::event copy(sycl::queue &queue, int64_t n, const double *x, int64_t incx, d return done; } -sycl::event copy(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event copy(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -934,10 +934,10 @@ sycl::event copy(sycl::queue &queue, int64_t n, const std::complex *x, in return done; } -sycl::event copy(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event copy(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -948,9 +948,9 @@ sycl::event copy(sycl::queue &queue, int64_t n, const std::complex *x, i return done; } -sycl::event dot(sycl::queue &queue, int64_t n, const float *x, int64_t incx, const float *y, - int64_t incy, float *result, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event dot(sycl::queue& queue, int64_t n, const float* x, int64_t incx, const float* y, + int64_t incy, float* result, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -962,9 +962,9 @@ sycl::event dot(sycl::queue &queue, int64_t n, const float *x, int64_t incx, con return done; } -sycl::event dot(sycl::queue &queue, int64_t n, const double *x, int64_t incx, const double *y, - int64_t incy, double *result, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event dot(sycl::queue& queue, int64_t n, const double* x, int64_t incx, const double* y, + int64_t incy, double* result, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -976,9 +976,9 @@ sycl::event dot(sycl::queue &queue, int64_t n, const double *x, int64_t incx, co return done; } -sycl::event dot(sycl::queue &queue, int64_t n, const float *x, int64_t incx, const float *y, - int64_t incy, double *result, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event dot(sycl::queue& queue, int64_t n, const float* x, int64_t incx, const float* y, + int64_t incy, double* result, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -990,10 +990,10 @@ sycl::event dot(sycl::queue &queue, int64_t n, const float *x, int64_t incx, con return done; } -sycl::event dotc(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event dotc(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1005,10 +1005,10 @@ sycl::event dotc(sycl::queue &queue, int64_t n, const std::complex *x, in return done; } -sycl::event dotc(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event dotc(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1020,10 +1020,10 @@ sycl::event dotc(sycl::queue &queue, int64_t n, const std::complex *x, i return done; } -sycl::event dotu(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event dotu(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1035,10 +1035,10 @@ sycl::event dotu(sycl::queue &queue, int64_t n, const std::complex *x, in return done; } -sycl::event dotu(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - const std::complex *y, int64_t incy, std::complex *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event dotu(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + const std::complex* y, int64_t incy, std::complex* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1050,9 +1050,9 @@ sycl::event dotu(sycl::queue &queue, int64_t n, const std::complex *x, i return done; } -sycl::event iamin(sycl::queue &queue, int64_t n, const float *x, int64_t incx, int64_t *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event iamin(sycl::queue& queue, int64_t n, const float* x, int64_t incx, int64_t* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1063,9 +1063,9 @@ sycl::event iamin(sycl::queue &queue, int64_t n, const float *x, int64_t incx, i return done; } -sycl::event iamin(sycl::queue &queue, int64_t n, const double *x, int64_t incx, int64_t *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event iamin(sycl::queue& queue, int64_t n, const double* x, int64_t incx, int64_t* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1076,9 +1076,9 @@ sycl::event iamin(sycl::queue &queue, int64_t n, const double *x, int64_t incx, return done; } -sycl::event iamin(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - int64_t *result, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event iamin(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + int64_t* result, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1089,9 +1089,9 @@ sycl::event iamin(sycl::queue &queue, int64_t n, const std::complex *x, i return done; } -sycl::event iamin(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - int64_t *result, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event iamin(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + int64_t* result, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1102,9 +1102,9 @@ sycl::event iamin(sycl::queue &queue, int64_t n, const std::complex *x, return done; } -sycl::event iamax(sycl::queue &queue, int64_t n, const float *x, int64_t incx, int64_t *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event iamax(sycl::queue& queue, int64_t n, const float* x, int64_t incx, int64_t* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1115,9 +1115,9 @@ sycl::event iamax(sycl::queue &queue, int64_t n, const float *x, int64_t incx, i return done; } -sycl::event iamax(sycl::queue &queue, int64_t n, const double *x, int64_t incx, int64_t *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event iamax(sycl::queue& queue, int64_t n, const double* x, int64_t incx, int64_t* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1128,9 +1128,9 @@ sycl::event iamax(sycl::queue &queue, int64_t n, const double *x, int64_t incx, return done; } -sycl::event iamax(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - int64_t *result, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event iamax(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + int64_t* result, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1141,9 +1141,9 @@ sycl::event iamax(sycl::queue &queue, int64_t n, const std::complex *x, i return done; } -sycl::event iamax(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - int64_t *result, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event iamax(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + int64_t* result, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1154,9 +1154,9 @@ sycl::event iamax(sycl::queue &queue, int64_t n, const std::complex *x, return done; } -sycl::event nrm2(sycl::queue &queue, int64_t n, const float *x, int64_t incx, float *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event nrm2(sycl::queue& queue, int64_t n, const float* x, int64_t incx, float* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1167,9 +1167,9 @@ sycl::event nrm2(sycl::queue &queue, int64_t n, const float *x, int64_t incx, fl return done; } -sycl::event nrm2(sycl::queue &queue, int64_t n, const double *x, int64_t incx, double *result, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event nrm2(sycl::queue& queue, int64_t n, const double* x, int64_t incx, double* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1180,9 +1180,9 @@ sycl::event nrm2(sycl::queue &queue, int64_t n, const double *x, int64_t incx, d return done; } -sycl::event nrm2(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - float *result, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event nrm2(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + float* result, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1193,9 +1193,9 @@ sycl::event nrm2(sycl::queue &queue, int64_t n, const std::complex *x, in return done; } -sycl::event nrm2(sycl::queue &queue, int64_t n, const std::complex *x, int64_t incx, - double *result, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event nrm2(sycl::queue& queue, int64_t n, const std::complex* x, int64_t incx, + double* result, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1206,9 +1206,9 @@ sycl::event nrm2(sycl::queue &queue, int64_t n, const std::complex *x, i return done; } -sycl::event rot(sycl::queue &queue, int64_t n, float *x, int64_t incx, float *y, int64_t incy, - float c, float s, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rot(sycl::queue& queue, int64_t n, float* x, int64_t incx, float* y, int64_t incy, + float c, float s, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1221,9 +1221,9 @@ sycl::event rot(sycl::queue &queue, int64_t n, float *x, int64_t incx, float *y, return done; } -sycl::event rot(sycl::queue &queue, int64_t n, double *x, int64_t incx, double *y, int64_t incy, - double c, double s, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rot(sycl::queue& queue, int64_t n, double* x, int64_t incx, double* y, int64_t incy, + double c, double s, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1236,10 +1236,10 @@ sycl::event rot(sycl::queue &queue, int64_t n, double *x, int64_t incx, double * return done; } -sycl::event rot(sycl::queue &queue, int64_t n, std::complex *x, int64_t incx, - std::complex *y, int64_t incy, float c, float s, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rot(sycl::queue& queue, int64_t n, std::complex* x, int64_t incx, + std::complex* y, int64_t incy, float c, float s, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1252,10 +1252,10 @@ sycl::event rot(sycl::queue &queue, int64_t n, std::complex *x, int64_t i return done; } -sycl::event rot(sycl::queue &queue, int64_t n, std::complex *x, int64_t incx, - std::complex *y, int64_t incy, double c, double s, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rot(sycl::queue& queue, int64_t n, std::complex* x, int64_t incx, + std::complex* y, int64_t incy, double c, double s, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1268,9 +1268,9 @@ sycl::event rot(sycl::queue &queue, int64_t n, std::complex *x, int64_t return done; } -sycl::event rotg(sycl::queue &queue, float *a, float *b, float *c, float *s, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rotg(sycl::queue& queue, float* a, float* b, float* c, float* s, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1280,9 +1280,9 @@ sycl::event rotg(sycl::queue &queue, float *a, float *b, float *c, float *s, return done; } -sycl::event rotg(sycl::queue &queue, double *a, double *b, double *c, double *s, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rotg(sycl::queue& queue, double* a, double* b, double* c, double* s, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1292,9 +1292,9 @@ sycl::event rotg(sycl::queue &queue, double *a, double *b, double *c, double *s, return done; } -sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, float *c, - std::complex *s, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, float* c, + std::complex* s, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1304,9 +1304,9 @@ sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex return done; } -sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, double *c, - std::complex *s, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, double* c, + std::complex* s, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1316,9 +1316,9 @@ sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rotm(sycl::queue& queue, int64_t n, float* x, int64_t incx, float* y, int64_t incy, + float* param, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1330,9 +1330,9 @@ sycl::event rotm(sycl::queue &queue, int64_t n, float *x, int64_t incx, float *y return done; } -sycl::event rotm(sycl::queue &queue, int64_t n, double *x, int64_t incx, double *y, int64_t incy, - double *param, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rotm(sycl::queue& queue, int64_t n, double* x, int64_t incx, double* y, int64_t incy, + double* param, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1344,9 +1344,9 @@ sycl::event rotm(sycl::queue &queue, int64_t n, double *x, int64_t incx, double return done; } -sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, float y1, float *param, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rotmg(sycl::queue& queue, float* d1, float* d2, float* x1, float y1, float* param, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1357,9 +1357,9 @@ sycl::event rotmg(sycl::queue &queue, float *d1, float *d2, float *x1, float y1, return done; } -sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, double y1, double *param, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event rotmg(sycl::queue& queue, double* d1, double* d2, double* x1, double y1, double* param, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1370,9 +1370,9 @@ sycl::event rotmg(sycl::queue &queue, double *d1, double *d2, double *x1, double return done; } -sycl::event scal(sycl::queue &queue, int64_t n, float alpha, float *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event scal(sycl::queue& queue, int64_t n, float alpha, float* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1384,9 +1384,9 @@ sycl::event scal(sycl::queue &queue, int64_t n, float alpha, float *x, int64_t i return done; } -sycl::event scal(sycl::queue &queue, int64_t n, double alpha, double *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event scal(sycl::queue& queue, int64_t n, double alpha, double* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1398,23 +1398,23 @@ sycl::event scal(sycl::queue &queue, int64_t n, double alpha, double *x, int64_t return done; } -sycl::event scal(sycl::queue &queue, int64_t n, std::complex alpha, std::complex *x, - int64_t incx, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event scal(sycl::queue& queue, int64_t n, std::complex alpha, std::complex* x, + int64_t incx, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { - ::cblas_cscal((const int)n, (const void *)&alpha, x, (const int)std::abs(incx)); + ::cblas_cscal((const int)n, (const void*)&alpha, x, (const int)std::abs(incx)); }); }); return done; } -sycl::event scal(sycl::queue &queue, int64_t n, float alpha, std::complex *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event scal(sycl::queue& queue, int64_t n, float alpha, std::complex* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1426,23 +1426,23 @@ sycl::event scal(sycl::queue &queue, int64_t n, float alpha, std::complex return done; } -sycl::event scal(sycl::queue &queue, int64_t n, std::complex alpha, std::complex *x, - int64_t incx, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event scal(sycl::queue& queue, int64_t n, std::complex alpha, std::complex* x, + int64_t incx, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { - ::cblas_zscal((const int)n, (const void *)&alpha, x, (const int)std::abs(incx)); + ::cblas_zscal((const int)n, (const void*)&alpha, x, (const int)std::abs(incx)); }); }); return done; } -sycl::event scal(sycl::queue &queue, int64_t n, double alpha, std::complex *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event scal(sycl::queue& queue, int64_t n, double alpha, std::complex* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1454,10 +1454,10 @@ sycl::event scal(sycl::queue &queue, int64_t n, double alpha, std::complex &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event sdsdot(sycl::queue& queue, int64_t n, float sb, const float* x, int64_t incx, + const float* y, int64_t incy, float* result, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1470,9 +1470,9 @@ sycl::event sdsdot(sycl::queue &queue, int64_t n, float sb, const float *x, int6 return done; } -sycl::event swap(sycl::queue &queue, int64_t n, float *x, int64_t incx, float *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event swap(sycl::queue& queue, int64_t n, float* x, int64_t incx, float* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1483,9 +1483,9 @@ sycl::event swap(sycl::queue &queue, int64_t n, float *x, int64_t incx, float *y return done; } -sycl::event swap(sycl::queue &queue, int64_t n, double *x, int64_t incx, double *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event swap(sycl::queue& queue, int64_t n, double* x, int64_t incx, double* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1496,10 +1496,10 @@ sycl::event swap(sycl::queue &queue, int64_t n, double *x, int64_t incx, double return done; } -sycl::event swap(sycl::queue &queue, int64_t n, std::complex *x, int64_t incx, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event swap(sycl::queue& queue, int64_t n, std::complex* x, int64_t incx, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1510,10 +1510,10 @@ sycl::event swap(sycl::queue &queue, int64_t n, std::complex *x, int64_t return done; } -sycl::event swap(sycl::queue &queue, int64_t n, std::complex *x, int64_t incx, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event swap(sycl::queue& queue, int64_t n, std::complex* x, int64_t incx, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); diff --git a/src/blas/backends/netlib/netlib_level2.cxx b/src/blas/backends/netlib/netlib_level2.cxx index 156ed133b..8e8d74446 100644 --- a/src/blas/backends/netlib/netlib_level2.cxx +++ b/src/blas/backends/netlib/netlib_level2.cxx @@ -19,10 +19,10 @@ // Buffer APIs -void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, float beta, sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, float beta, sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); @@ -36,10 +36,10 @@ void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, }); } -void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, double beta, sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + double alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, double beta, sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); @@ -53,46 +53,46 @@ void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, }); } -void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { ::cblas_cgbmv(MAJOR, convert_to_cblas_trans(trans), (const int)m, (const int)n, - (const int)kl, (const int)ku, (const void *)&alpha, + (const int)kl, (const int)ku, (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_x.GET_MULTI_PTR, - (const int)incx, (const void *)&beta, accessor_y.GET_MULTI_PTR, + (const int)incx, (const void*)&beta, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { ::cblas_zgbmv(MAJOR, convert_to_cblas_trans(trans), (const int)m, (const int)n, - (const int)kl, (const int)ku, (const void *)&alpha, + (const int)kl, (const int)ku, (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_x.GET_MULTI_PTR, - (const int)incx, (const void *)&beta, accessor_y.GET_MULTI_PTR, + (const int)incx, (const void*)&beta, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - float beta, sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + float beta, sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); @@ -105,10 +105,10 @@ void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha }); } -void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - double beta, sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + double beta, sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); @@ -121,44 +121,44 @@ void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alph }); } -void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { ::cblas_cgemv(MAJOR, convert_to_cblas_trans(trans), (const int)m, (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, - accessor_x.GET_MULTI_PTR, (const int)incx, (const void *)&beta, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + accessor_x.GET_MULTI_PTR, (const int)incx, (const void*)&beta, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { ::cblas_zgemv(MAJOR, convert_to_cblas_trans(trans), (const int)m, (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, - accessor_x.GET_MULTI_PTR, (const int)incx, (const void *)&beta, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + accessor_x.GET_MULTI_PTR, (const int)incx, (const void*)&beta, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void ger(sycl::queue &queue, int64_t m, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, +void ger(sycl::queue& queue, int64_t m, int64_t n, float alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_a = a.get_access(cgh); @@ -170,10 +170,10 @@ void ger(sycl::queue &queue, int64_t m, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, +void ger(sycl::queue& queue, int64_t m, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_a = a.get_access(cgh); @@ -185,142 +185,142 @@ void ger(sycl::queue &queue, int64_t m, int64_t n, double alpha, sycl::buffer alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { +void gerc(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_a = a.get_access(cgh); host_task(cgh, [=]() { - ::cblas_cgerc(MAJOR, (const int)m, (const int)n, (const void *)&alpha, + ::cblas_cgerc(MAJOR, (const int)m, (const int)n, (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, accessor_y.GET_MULTI_PTR, (const int)incy, accessor_a.GET_MULTI_PTR, (const int)lda); }); }); } -void gerc(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { +void gerc(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_a = a.get_access(cgh); host_task(cgh, [=]() { - ::cblas_zgerc(MAJOR, (const int)m, (const int)n, (const void *)&alpha, + ::cblas_zgerc(MAJOR, (const int)m, (const int)n, (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, accessor_y.GET_MULTI_PTR, (const int)incy, accessor_a.GET_MULTI_PTR, (const int)lda); }); }); } -void geru(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { +void geru(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_a = a.get_access(cgh); host_task(cgh, [=]() { - ::cblas_cgeru(MAJOR, (const int)m, (const int)n, (const void *)&alpha, + ::cblas_cgeru(MAJOR, (const int)m, (const int)n, (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, accessor_y.GET_MULTI_PTR, (const int)incy, accessor_a.GET_MULTI_PTR, (const int)lda); }); }); } -void geru(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { +void geru(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_a = a.get_access(cgh); host_task(cgh, [=]() { - ::cblas_zgeru(MAJOR, (const int)m, (const int)n, (const void *)&alpha, + ::cblas_zgeru(MAJOR, (const int)m, (const int)n, (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, accessor_y.GET_MULTI_PTR, (const int)incy, accessor_a.GET_MULTI_PTR, (const int)lda); }); }); } -void hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { ::cblas_chbmv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, (const int)k, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, - accessor_x.GET_MULTI_PTR, (const int)incx, (const void *)&beta, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + accessor_x.GET_MULTI_PTR, (const int)incx, (const void*)&beta, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { ::cblas_zhbmv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, (const int)k, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, - accessor_x.GET_MULTI_PTR, (const int)incx, (const void *)&beta, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + accessor_x.GET_MULTI_PTR, (const int)incx, (const void*)&beta, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void hemv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void hemv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { ::cblas_chemv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, - accessor_x.GET_MULTI_PTR, (const int)incx, (const void *)&beta, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + accessor_x.GET_MULTI_PTR, (const int)incx, (const void*)&beta, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void hemv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void hemv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { ::cblas_zhemv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, - accessor_x.GET_MULTI_PTR, (const int)incx, (const void *)&beta, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + accessor_x.GET_MULTI_PTR, (const int)incx, (const void*)&beta, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void her(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { +void her(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& a, int64_t lda) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_a = a.get_access(cgh); host_task(cgh, [=]() { @@ -331,10 +331,10 @@ void her(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, }); } -void her(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { +void her(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& a, int64_t lda) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_a = a.get_access(cgh); host_task(cgh, [=]() { @@ -345,78 +345,78 @@ void her(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, }); } -void her2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { +void her2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_a = a.get_access(cgh); host_task(cgh, [=]() { ::cblas_cher2(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, + (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, accessor_y.GET_MULTI_PTR, (const int)incy, accessor_a.GET_MULTI_PTR, (const int)lda); }); }); } -void her2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { +void her2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_a = a.get_access(cgh); host_task(cgh, [=]() { ::cblas_zher2(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, + (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, accessor_y.GET_MULTI_PTR, (const int)incy, accessor_a.GET_MULTI_PTR, (const int)lda); }); }); } -void hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &ap, sycl::buffer, 1> &x, - int64_t incx, std::complex beta, sycl::buffer, 1> &y, +void hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& ap, sycl::buffer, 1>& x, + int64_t incx, std::complex beta, sycl::buffer, 1>& y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { ::cblas_chpmv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, accessor_ap.GET_MULTI_PTR, accessor_x.GET_MULTI_PTR, - (const int)incx, (const void *)&beta, accessor_y.GET_MULTI_PTR, + (const void*)&alpha, accessor_ap.GET_MULTI_PTR, accessor_x.GET_MULTI_PTR, + (const int)incx, (const void*)&beta, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &ap, sycl::buffer, 1> &x, - int64_t incx, std::complex beta, sycl::buffer, 1> &y, +void hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& ap, sycl::buffer, 1>& x, + int64_t incx, std::complex beta, sycl::buffer, 1>& y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); host_task(cgh, [=]() { ::cblas_zhpmv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, accessor_ap.GET_MULTI_PTR, accessor_x.GET_MULTI_PTR, - (const int)incx, (const void *)&beta, accessor_y.GET_MULTI_PTR, + (const void*)&alpha, accessor_ap.GET_MULTI_PTR, accessor_x.GET_MULTI_PTR, + (const int)incx, (const void*)&beta, accessor_y.GET_MULTI_PTR, (const int)incy); }); }); } -void hpr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &ap) { - queue.submit([&](sycl::handler &cgh) { +void hpr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& ap) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_ap = ap.get_access(cgh); host_task(cgh, [=]() { @@ -427,10 +427,10 @@ void hpr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, }); } -void hpr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &ap) { - queue.submit([&](sycl::handler &cgh) { +void hpr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& ap) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_ap = ap.get_access(cgh); host_task(cgh, [=]() { @@ -441,42 +441,42 @@ void hpr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, }); } -void hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &ap) { - queue.submit([&](sycl::handler &cgh) { +void hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& ap) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_ap = ap.get_access(cgh); host_task(cgh, [=]() { ::cblas_chpr2(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, + (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, accessor_y.GET_MULTI_PTR, (const int)incy, accessor_ap.GET_MULTI_PTR); }); }); } -void hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &ap) { - queue.submit([&](sycl::handler &cgh) { +void hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& ap) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_ap = ap.get_access(cgh); host_task(cgh, [=]() { ::cblas_zhpr2(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, + (const void*)&alpha, accessor_x.GET_MULTI_PTR, (const int)incx, accessor_y.GET_MULTI_PTR, (const int)incy, accessor_ap.GET_MULTI_PTR); }); }); } -void sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - float beta, sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + float beta, sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); @@ -489,10 +489,10 @@ void sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, float alph }); } -void sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, - double beta, sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, double alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, + double beta, sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); @@ -505,10 +505,10 @@ void sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, double alp }); } -void spmv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &ap, - sycl::buffer &x, int64_t incx, float beta, sycl::buffer &y, +void spmv(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& ap, + sycl::buffer& x, int64_t incx, float beta, sycl::buffer& y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); @@ -521,10 +521,10 @@ void spmv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::bu }); } -void spmv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, - sycl::buffer &ap, sycl::buffer &x, int64_t incx, double beta, - sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void spmv(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, + sycl::buffer& ap, sycl::buffer& x, int64_t incx, double beta, + sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); @@ -537,9 +537,9 @@ void spmv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, }); } -void spr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &ap) { - queue.submit([&](sycl::handler &cgh) { +void spr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& ap) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_ap = ap.get_access(cgh); host_task(cgh, [=]() { @@ -550,9 +550,9 @@ void spr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buf }); } -void spr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &ap) { - queue.submit([&](sycl::handler &cgh) { +void spr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& ap) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_ap = ap.get_access(cgh); host_task(cgh, [=]() { @@ -563,9 +563,9 @@ void spr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::bu }); } -void spr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &ap) { - queue.submit([&](sycl::handler &cgh) { +void spr2(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& ap) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_ap = ap.get_access(cgh); @@ -577,9 +577,9 @@ void spr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::bu }); } -void spr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &ap) { - queue.submit([&](sycl::handler &cgh) { +void spr2(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& ap) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_ap = ap.get_access(cgh); @@ -591,10 +591,10 @@ void spr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::b }); } -void symv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &x, int64_t incx, float beta, - sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void symv(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& x, int64_t incx, float beta, + sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); @@ -607,10 +607,10 @@ void symv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::bu }); } -void symv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer &a, - int64_t lda, sycl::buffer &x, int64_t incx, double beta, - sycl::buffer &y, int64_t incy) { - queue.submit([&](sycl::handler &cgh) { +void symv(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer& a, + int64_t lda, sycl::buffer& x, int64_t incx, double beta, + sycl::buffer& y, int64_t incy) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); @@ -623,9 +623,9 @@ void symv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::b }); } -void syr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { +void syr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& a, int64_t lda) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_a = a.get_access(cgh); host_task(cgh, [=]() { @@ -636,9 +636,9 @@ void syr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buf }); } -void syr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { +void syr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& a, int64_t lda) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_a = a.get_access(cgh); host_task(cgh, [=]() { @@ -649,10 +649,10 @@ void syr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::bu }); } -void syr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, +void syr2(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_a = a.get_access(cgh); @@ -665,10 +665,10 @@ void syr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, sycl::bu }); } -void syr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, +void syr2(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, int64_t lda) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_x = x.get_access(cgh); auto accessor_y = y.get_access(cgh); auto accessor_a = a.get_access(cgh); @@ -681,10 +681,10 @@ void syr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, sycl::b }); } -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -696,10 +696,10 @@ void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -711,10 +711,10 @@ void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -726,10 +726,10 @@ void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -741,10 +741,10 @@ void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -756,10 +756,10 @@ void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -771,10 +771,10 @@ void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -786,10 +786,10 @@ void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -801,9 +801,9 @@ void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &ap, sycl::buffer &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& ap, sycl::buffer& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -814,9 +814,9 @@ void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &ap, sycl::buffer &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& ap, sycl::buffer& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -827,10 +827,10 @@ void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &ap, sycl::buffer, 1> &x, +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& ap, sycl::buffer, 1>& x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -841,10 +841,10 @@ void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &ap, sycl::buffer, 1> &x, +void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& ap, sycl::buffer, 1>& x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -855,9 +855,9 @@ void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &ap, sycl::buffer &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& ap, sycl::buffer& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -868,9 +868,9 @@ void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &ap, sycl::buffer &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& ap, sycl::buffer& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -881,10 +881,10 @@ void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &ap, sycl::buffer, 1> &x, +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& ap, sycl::buffer, 1>& x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -895,10 +895,10 @@ void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &ap, sycl::buffer, 1> &x, +void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& ap, sycl::buffer, 1>& x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_ap = ap.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -909,9 +909,9 @@ void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void trmv(sycl::queue& queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { @@ -922,9 +922,9 @@ void trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag }); } -void trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void trmv(sycl::queue& queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { @@ -935,10 +935,10 @@ void trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag }); } -void trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void trmv(sycl::queue& queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { @@ -949,10 +949,10 @@ void trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag }); } -void trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void trmv(sycl::queue& queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { @@ -963,9 +963,9 @@ void trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag }); } -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -976,9 +976,9 @@ void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -989,10 +989,10 @@ void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -1003,10 +1003,10 @@ void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, }); } -void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { - queue.submit([&](sycl::handler &cgh) { +void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_x = x.get_access(cgh); host_task(cgh, [=]() { @@ -1019,10 +1019,10 @@ void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, // USM APIs -sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - float alpha, const float *a, int64_t lda, const float *x, int64_t incx, float beta, - float *y, int64_t incy, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + float alpha, const float* a, int64_t lda, const float* x, int64_t incx, float beta, + float* y, int64_t incy, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1036,11 +1036,11 @@ sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int6 return done; } -sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - double alpha, const double *a, int64_t lda, const double *x, int64_t incx, - double beta, double *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + double alpha, const double* a, int64_t lda, const double* x, int64_t incx, + double beta, double* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1054,48 +1054,48 @@ sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int6 return done; } -sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_cgbmv(MAJOR, convert_to_cblas_trans(trans), (const int)m, (const int)n, - (const int)kl, (const int)ku, (const void *)&alpha, a, (const int)lda, x, - (const int)incx, (const void *)&beta, y, (const int)incy); + (const int)kl, (const int)ku, (const void*)&alpha, a, (const int)lda, x, + (const int)incx, (const void*)&beta, y, (const int)incy); }); }); return done; } -sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_zgbmv(MAJOR, convert_to_cblas_trans(trans), (const int)m, (const int)n, - (const int)kl, (const int)ku, (const void *)&alpha, a, (const int)lda, x, - (const int)incx, (const void *)&beta, y, (const int)incy); + (const int)kl, (const int)ku, (const void*)&alpha, a, (const int)lda, x, + (const int)incx, (const void*)&beta, y, (const int)incy); }); }); return done; } -sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - const float *a, int64_t lda, const float *x, int64_t incx, float beta, float *y, - int64_t incy, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + const float* a, int64_t lda, const float* x, int64_t incx, float beta, float* y, + int64_t incy, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1109,10 +1109,10 @@ sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, floa return done; } -sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - const double *a, int64_t lda, const double *x, int64_t incx, double beta, - double *y, int64_t incy, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + const double* a, int64_t lda, const double* x, int64_t incx, double beta, + double* y, int64_t incy, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1126,48 +1126,48 @@ sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, doub return done; } -sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_cgemv(MAJOR, convert_to_cblas_trans(trans), (const int)m, (const int)n, - (const void *)&alpha, a, (const int)lda, x, (const int)incx, - (const void *)&beta, y, (const int)incy); + (const void*)&alpha, a, (const int)lda, x, (const int)incx, + (const void*)&beta, y, (const int)incy); }); }); return done; } -sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_zgemv(MAJOR, convert_to_cblas_trans(trans), (const int)m, (const int)n, - (const void *)&alpha, a, (const int)lda, x, (const int)incx, - (const void *)&beta, y, (const int)incy); + (const void*)&alpha, a, (const int)lda, x, (const int)incx, + (const void*)&beta, y, (const int)incy); }); }); return done; } -sycl::event ger(sycl::queue &queue, int64_t m, int64_t n, float alpha, const float *x, int64_t incx, - const float *y, int64_t incy, float *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event ger(sycl::queue& queue, int64_t m, int64_t n, float alpha, const float* x, int64_t incx, + const float* y, int64_t incy, float* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1180,10 +1180,10 @@ sycl::event ger(sycl::queue &queue, int64_t m, int64_t n, float alpha, const flo return done; } -sycl::event ger(sycl::queue &queue, int64_t m, int64_t n, double alpha, const double *x, - int64_t incx, const double *y, int64_t incy, double *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event ger(sycl::queue& queue, int64_t m, int64_t n, double alpha, const double* x, + int64_t incx, const double* y, int64_t incy, double* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1196,152 +1196,152 @@ sycl::event ger(sycl::queue &queue, int64_t m, int64_t n, double alpha, const do return done; } -sycl::event gerc(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gerc(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { - ::cblas_cgerc(MAJOR, (const int)m, (const int)n, (const void *)&alpha, x, + ::cblas_cgerc(MAJOR, (const int)m, (const int)n, (const void*)&alpha, x, (const int)incx, y, (const int)incy, a, (const int)lda); }); }); return done; } -sycl::event gerc(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gerc(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { - ::cblas_zgerc(MAJOR, (const int)m, (const int)n, (const void *)&alpha, x, + ::cblas_zgerc(MAJOR, (const int)m, (const int)n, (const void*)&alpha, x, (const int)incx, y, (const int)incy, a, (const int)lda); }); }); return done; } -sycl::event geru(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event geru(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { - ::cblas_cgeru(MAJOR, (const int)m, (const int)n, (const void *)&alpha, x, + ::cblas_cgeru(MAJOR, (const int)m, (const int)n, (const void*)&alpha, x, (const int)incx, y, (const int)incy, a, (const int)lda); }); }); return done; } -sycl::event geru(sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event geru(sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { - ::cblas_zgeru(MAJOR, (const int)m, (const int)n, (const void *)&alpha, x, + ::cblas_zgeru(MAJOR, (const int)m, (const int)n, (const void*)&alpha, x, (const int)incx, y, (const int)incy, a, (const int)lda); }); }); return done; } -sycl::event hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_chbmv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, (const int)k, - (const void *)&alpha, a, (const int)lda, x, (const int)incx, - (const void *)&beta, y, (const int)incy); + (const void*)&alpha, a, (const int)lda, x, (const int)incx, + (const void*)&beta, y, (const int)incy); }); }); return done; } -sycl::event hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_zhbmv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, (const int)k, - (const void *)&alpha, a, (const int)lda, x, (const int)incx, - (const void *)&beta, y, (const int)incy); + (const void*)&alpha, a, (const int)lda, x, (const int)incx, + (const void*)&beta, y, (const int)incy); }); }); return done; } -sycl::event hemv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, const std::complex *x, - int64_t incx, std::complex beta, std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hemv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* a, int64_t lda, const std::complex* x, + int64_t incx, std::complex beta, std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_chemv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, a, (const int)lda, x, (const int)incx, - (const void *)&beta, y, (const int)incy); + (const void*)&alpha, a, (const int)lda, x, (const int)incx, + (const void*)&beta, y, (const int)incy); }); }); return done; } -sycl::event hemv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, const std::complex *x, - int64_t incx, std::complex beta, std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hemv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* a, int64_t lda, const std::complex* x, + int64_t incx, std::complex beta, std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_zhemv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, a, (const int)lda, x, (const int)incx, - (const void *)&beta, y, (const int)incy); + (const void*)&alpha, a, (const int)lda, x, (const int)incx, + (const void*)&beta, y, (const int)incy); }); }); return done; } -sycl::event her(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, - const std::complex *x, int64_t incx, std::complex *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event her(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, + const std::complex* x, int64_t incx, std::complex* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1354,10 +1354,10 @@ sycl::event her(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, return done; } -sycl::event her(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, - const std::complex *x, int64_t incx, std::complex *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event her(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, + const std::complex* x, int64_t incx, std::complex* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1370,82 +1370,82 @@ sycl::event her(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, return done; } -sycl::event her2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event her2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_cher2(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, x, (const int)incx, y, (const int)incy, a, + (const void*)&alpha, x, (const int)incx, y, (const int)incy, a, (const int)lda); }); }); return done; } -sycl::event her2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event her2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_zher2(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, x, (const int)incx, y, (const int)incy, a, + (const void*)&alpha, x, (const int)incx, y, (const int)incy, a, (const int)lda); }); }); return done; } -sycl::event hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *ap, const std::complex *x, int64_t incx, - std::complex beta, std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* ap, const std::complex* x, int64_t incx, + std::complex beta, std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_chpmv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, ap, x, (const int)incx, (const void *)&beta, y, + (const void*)&alpha, ap, x, (const int)incx, (const void*)&beta, y, (const int)incy); }); }); return done; } -sycl::event hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *ap, const std::complex *x, int64_t incx, - std::complex beta, std::complex *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* ap, const std::complex* x, int64_t incx, + std::complex beta, std::complex* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_zhpmv(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, ap, x, (const int)incx, (const void *)&beta, y, + (const void*)&alpha, ap, x, (const int)incx, (const void*)&beta, y, (const int)incy); }); }); return done; } -sycl::event hpr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, - const std::complex *x, int64_t incx, std::complex *ap, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hpr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, + const std::complex* x, int64_t incx, std::complex* ap, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1458,10 +1458,10 @@ sycl::event hpr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, return done; } -sycl::event hpr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, - const std::complex *x, int64_t incx, std::complex *ap, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hpr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, + const std::complex* x, int64_t incx, std::complex* ap, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1474,44 +1474,44 @@ sycl::event hpr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, return done; } -sycl::event hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *ap, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* ap, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_chpr2(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, x, (const int)incx, y, (const int)incy, ap); + (const void*)&alpha, x, (const int)incx, y, (const int)incy, ap); }); }); return done; } -sycl::event hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *ap, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* ap, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_zhpr2(MAJOR, convert_to_cblas_uplo(upper_lower), (const int)n, - (const void *)&alpha, x, (const int)incx, y, (const int)incy, ap); + (const void*)&alpha, x, (const int)incx, y, (const int)incy, ap); }); }); return done; } -sycl::event sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, float alpha, - const float *a, int64_t lda, const float *x, int64_t incx, float beta, float *y, - int64_t incy, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, float alpha, + const float* a, int64_t lda, const float* x, int64_t incx, float beta, float* y, + int64_t incy, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1525,10 +1525,10 @@ sycl::event sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, flo return done; } -sycl::event sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, double alpha, - const double *a, int64_t lda, const double *x, int64_t incx, double beta, - double *y, int64_t incy, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, double alpha, + const double* a, int64_t lda, const double* x, int64_t incx, double beta, + double* y, int64_t incy, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1542,10 +1542,10 @@ sycl::event sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, dou return done; } -sycl::event spmv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *ap, - const float *x, int64_t incx, float beta, float *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event spmv(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* ap, + const float* x, int64_t incx, float beta, float* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1559,10 +1559,10 @@ sycl::event spmv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, c return done; } -sycl::event spmv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *ap, - const double *x, int64_t incx, double beta, double *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event spmv(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* ap, + const double* x, int64_t incx, double beta, double* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1576,9 +1576,9 @@ sycl::event spmv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, return done; } -sycl::event spr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *x, - int64_t incx, float *ap, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event spr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* x, + int64_t incx, float* ap, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1591,9 +1591,9 @@ sycl::event spr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, co return done; } -sycl::event spr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *x, - int64_t incx, double *ap, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event spr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* x, + int64_t incx, double* ap, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1606,10 +1606,10 @@ sycl::event spr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, c return done; } -sycl::event spr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *x, - int64_t incx, const float *y, int64_t incy, float *ap, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event spr2(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* x, + int64_t incx, const float* y, int64_t incy, float* ap, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1622,10 +1622,10 @@ sycl::event spr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, c return done; } -sycl::event spr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *x, - int64_t incx, const double *y, int64_t incy, double *ap, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event spr2(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* x, + int64_t incx, const double* y, int64_t incy, double* ap, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1638,10 +1638,10 @@ sycl::event spr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, return done; } -sycl::event symv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *a, - int64_t lda, const float *x, int64_t incx, float beta, float *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event symv(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* a, + int64_t lda, const float* x, int64_t incx, float beta, float* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1655,10 +1655,10 @@ sycl::event symv(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, c return done; } -sycl::event symv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *a, - int64_t lda, const double *x, int64_t incx, double beta, double *y, int64_t incy, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event symv(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* a, + int64_t lda, const double* x, int64_t incx, double beta, double* y, int64_t incy, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1672,9 +1672,9 @@ sycl::event symv(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, return done; } -sycl::event syr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *x, - int64_t incx, float *a, int64_t lda, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syr(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* x, + int64_t incx, float* a, int64_t lda, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1687,10 +1687,10 @@ sycl::event syr(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, co return done; } -sycl::event syr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *x, - int64_t incx, double *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syr(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* x, + int64_t incx, double* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1703,10 +1703,10 @@ sycl::event syr(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, c return done; } -sycl::event syr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, const float *x, - int64_t incx, const float *y, int64_t incy, float *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syr2(sycl::queue& queue, uplo upper_lower, int64_t n, float alpha, const float* x, + int64_t incx, const float* y, int64_t incy, float* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1720,10 +1720,10 @@ sycl::event syr2(sycl::queue &queue, uplo upper_lower, int64_t n, float alpha, c return done; } -sycl::event syr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, const double *x, - int64_t incx, const double *y, int64_t incy, double *a, int64_t lda, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syr2(sycl::queue& queue, uplo upper_lower, int64_t n, double alpha, const double* x, + int64_t incx, const double* y, int64_t incy, double* a, int64_t lda, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1737,10 +1737,10 @@ sycl::event syr2(sycl::queue &queue, uplo upper_lower, int64_t n, double alpha, return done; } -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const float *a, int64_t lda, float *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const float* a, int64_t lda, float* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1754,10 +1754,10 @@ sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const double *a, int64_t lda, double *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const double* a, int64_t lda, double* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1771,10 +1771,10 @@ sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const std::complex *a, int64_t lda, std::complex *x, - int64_t incx, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const std::complex* a, int64_t lda, std::complex* x, + int64_t incx, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1788,10 +1788,10 @@ sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const std::complex *a, int64_t lda, std::complex *x, - int64_t incx, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const std::complex* a, int64_t lda, std::complex* x, + int64_t incx, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1805,10 +1805,10 @@ sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const float *a, int64_t lda, float *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const float* a, int64_t lda, float* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1822,10 +1822,10 @@ sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const double *a, int64_t lda, double *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const double* a, int64_t lda, double* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1839,10 +1839,10 @@ sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const std::complex *a, int64_t lda, std::complex *x, - int64_t incx, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const std::complex* a, int64_t lda, std::complex* x, + int64_t incx, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1856,10 +1856,10 @@ sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - int64_t k, const std::complex *a, int64_t lda, std::complex *x, - int64_t incx, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + int64_t k, const std::complex* a, int64_t lda, std::complex* x, + int64_t incx, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1873,10 +1873,10 @@ sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const float *ap, float *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const float* ap, float* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1889,10 +1889,10 @@ sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const double *ap, double *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const double* ap, double* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1905,10 +1905,10 @@ sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *ap, std::complex *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* ap, std::complex* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1921,10 +1921,10 @@ sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *ap, std::complex *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* ap, std::complex* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1937,10 +1937,10 @@ sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const float *ap, float *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const float* ap, float* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1953,10 +1953,10 @@ sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const double *ap, double *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const double* ap, double* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1969,10 +1969,10 @@ sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *ap, std::complex *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* ap, std::complex* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1985,10 +1985,10 @@ sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *ap, std::complex *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* ap, std::complex* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -2001,10 +2001,10 @@ sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, - const float *a, int64_t lda, float *b, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, + const float* a, int64_t lda, float* b, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -2018,10 +2018,10 @@ sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag un return done; } -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, - const double *a, int64_t lda, double *b, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, + const double* a, int64_t lda, double* b, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -2035,10 +2035,10 @@ sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag un return done; } -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, - const std::complex *a, int64_t lda, std::complex *b, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, + const std::complex* a, int64_t lda, std::complex* b, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -2052,10 +2052,10 @@ sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag un return done; } -sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, - const std::complex *a, int64_t lda, std::complex *b, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose transa, diag unit_diag, int64_t n, + const std::complex* a, int64_t lda, std::complex* b, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -2069,10 +2069,10 @@ sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose transa, diag un return done; } -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const float *a, int64_t lda, float *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const float* a, int64_t lda, float* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -2086,10 +2086,10 @@ sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const double *a, int64_t lda, double *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const double* a, int64_t lda, double* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -2103,10 +2103,10 @@ sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *a, int64_t lda, std::complex *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* a, int64_t lda, std::complex* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -2120,10 +2120,10 @@ sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag uni return done; } -sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, - const std::complex *a, int64_t lda, std::complex *x, int64_t incx, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, + const std::complex* a, int64_t lda, std::complex* x, int64_t incx, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); diff --git a/src/blas/backends/netlib/netlib_level3.cxx b/src/blas/backends/netlib/netlib_level3.cxx index 8bb6a04ae..2579e66e1 100644 --- a/src/blas/backends/netlib/netlib_level3.cxx +++ b/src/blas/backends/netlib/netlib_level3.cxx @@ -19,10 +19,10 @@ // Buffer APIs -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); @@ -36,10 +36,10 @@ void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int }); } -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, double beta, sycl::buffer &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + double alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, double beta, sycl::buffer& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); @@ -53,46 +53,46 @@ void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int }); } -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_cgemm(MAJOR, convert_to_cblas_trans(transa), convert_to_cblas_trans(transb), - (const int)m, (const int)n, (const int)k, (const void *)&alpha, + (const int)m, (const int)n, (const int)k, (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_b.GET_MULTI_PTR, - (const int)ldb, (const void *)&beta, accessor_c.GET_MULTI_PTR, + (const int)ldb, (const void*)&beta, accessor_c.GET_MULTI_PTR, (const int)ldc); }); }); } -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_zgemm(MAJOR, convert_to_cblas_trans(transa), convert_to_cblas_trans(transb), - (const int)m, (const int)n, (const int)k, (const void *)&alpha, + (const int)m, (const int)n, (const int)k, (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_b.GET_MULTI_PTR, - (const int)ldb, (const void *)&beta, accessor_c.GET_MULTI_PTR, + (const int)ldb, (const void*)&beta, accessor_c.GET_MULTI_PTR, (const int)ldc); }); }); } -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - sycl::half alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, sycl::half beta, - sycl::buffer &c, int64_t ldc) { +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + sycl::half alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, sycl::half beta, + sycl::buffer& c, int64_t ldc) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm", "for column_major layout"); #endif @@ -101,9 +101,9 @@ void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int #endif } -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm", "for column_major layout"); #endif @@ -112,9 +112,9 @@ void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int #endif } -void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - float alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, float beta, sycl::buffer &c, int64_t ldc) { +void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, + float alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm", "for column_major layout"); #endif @@ -123,46 +123,46 @@ void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int #endif } -void hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_chemm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), (const int)m, (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, - accessor_b.GET_MULTI_PTR, (const int)ldb, (const void *)&beta, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + accessor_b.GET_MULTI_PTR, (const int)ldb, (const void*)&beta, accessor_c.GET_MULTI_PTR, (const int)ldc); }); }); } -void hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_zhemm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), (const int)m, (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, - accessor_b.GET_MULTI_PTR, (const int)ldb, (const void *)&beta, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + accessor_b.GET_MULTI_PTR, (const int)ldb, (const void*)&beta, accessor_c.GET_MULTI_PTR, (const int)ldc); }); }); } -void herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, - sycl::buffer, 1> &a, int64_t lda, float beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, + sycl::buffer, 1>& a, int64_t lda, float beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { @@ -174,10 +174,10 @@ void herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int6 }); } -void herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, - sycl::buffer, 1> &a, int64_t lda, double beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, + sycl::buffer, 1>& a, int64_t lda, double beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { @@ -189,17 +189,17 @@ void herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int6 }); } -void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, float beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, float beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_cher2k(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, + (const int)n, (const int)k, (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_b.GET_MULTI_PTR, (const int)ldb, (const float)beta, accessor_c.GET_MULTI_PTR, (const int)ldc); @@ -207,17 +207,17 @@ void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int }); } -void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, double beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, double beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_zher2k(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, + (const int)n, (const int)k, (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_b.GET_MULTI_PTR, (const int)ldb, (const double)beta, accessor_c.GET_MULTI_PTR, (const int)ldc); @@ -225,10 +225,10 @@ void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int }); } -void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, - float beta, sycl::buffer &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb, + float beta, sycl::buffer& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); @@ -242,10 +242,10 @@ void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int6 }); } -void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, double alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, - double beta, sycl::buffer &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, double alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb, + double beta, sycl::buffer& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); @@ -259,46 +259,46 @@ void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int6 }); } -void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_csymm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), (const int)m, (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, - accessor_b.GET_MULTI_PTR, (const int)ldb, (const void *)&beta, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + accessor_b.GET_MULTI_PTR, (const int)ldb, (const void*)&beta, accessor_c.GET_MULTI_PTR, (const int)ldc); }); }); } -void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_zsymm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), (const int)m, (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, - accessor_b.GET_MULTI_PTR, (const int)ldb, (const void *)&beta, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + accessor_b.GET_MULTI_PTR, (const int)ldb, (const void*)&beta, accessor_c.GET_MULTI_PTR, (const int)ldc); }); }); } -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, - sycl::buffer &a, int64_t lda, float beta, sycl::buffer &c, +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, + sycl::buffer& a, int64_t lda, float beta, sycl::buffer& c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { @@ -310,10 +310,10 @@ void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int6 }); } -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, - sycl::buffer &a, int64_t lda, double beta, sycl::buffer &c, +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, double alpha, + sycl::buffer& a, int64_t lda, double beta, sycl::buffer& c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { @@ -325,40 +325,40 @@ void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int6 }); } -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_csyrk(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, - accessor_a.GET_MULTI_PTR, (const int)lda, (const void *)&beta, - accessor_c.GET_MULTI_PTR, (const int)ldc); + (const int)n, (const int)k, (const void*)&alpha, accessor_a.GET_MULTI_PTR, + (const int)lda, (const void*)&beta, accessor_c.GET_MULTI_PTR, + (const int)ldc); }); }); } -void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_zsyrk(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, - accessor_a.GET_MULTI_PTR, (const int)lda, (const void *)&beta, - accessor_c.GET_MULTI_PTR, (const int)ldc); + (const int)n, (const int)k, (const void*)&alpha, accessor_a.GET_MULTI_PTR, + (const int)lda, (const void*)&beta, accessor_c.GET_MULTI_PTR, + (const int)ldc); }); }); } -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb, - float beta, sycl::buffer &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, float alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb, + float beta, sycl::buffer& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); @@ -371,10 +371,10 @@ void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int }); } -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, double beta, sycl::buffer &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, double beta, sycl::buffer& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); @@ -388,46 +388,46 @@ void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int }); } -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_csyr2k(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, + (const int)n, (const int)k, (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_b.GET_MULTI_PTR, - (const int)ldb, (const void *)&beta, accessor_c.GET_MULTI_PTR, + (const int)ldb, (const void*)&beta, accessor_c.GET_MULTI_PTR, (const int)ldc); }); }); } -void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, int64_t ldc) { - queue.submit([&](sycl::handler &cgh) { +void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, int64_t ldc) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); auto accessor_c = c.get_access(cgh); host_task(cgh, [=]() { ::cblas_zsyr2k(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, + (const int)n, (const int)k, (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_b.GET_MULTI_PTR, - (const int)ldb, (const void *)&beta, accessor_c.GET_MULTI_PTR, + (const int)ldb, (const void*)&beta, accessor_c.GET_MULTI_PTR, (const int)ldc); }); }); } -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, - int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb) { - queue.submit([&](sycl::handler &cgh) { +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, + int64_t m, int64_t n, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { @@ -440,10 +440,10 @@ void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans }); } -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, - int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb) { - queue.submit([&](sycl::handler &cgh) { +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, + int64_t m, int64_t n, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { @@ -456,43 +456,43 @@ void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans }); } -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, - int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb) { - queue.submit([&](sycl::handler &cgh) { +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, + int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { ::cblas_ctrmm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(transa), convert_to_cblas_diag(unit_diag), (const int)m, (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_b.GET_MULTI_PTR, (const int)ldb); }); }); } -void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, +void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb) { - queue.submit([&](sycl::handler &cgh) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { ::cblas_ztrmm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(transa), convert_to_cblas_diag(unit_diag), (const int)m, (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_b.GET_MULTI_PTR, (const int)ldb); }); }); } -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, - int64_t m, int64_t n, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb) { - queue.submit([&](sycl::handler &cgh) { +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, + int64_t m, int64_t n, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { @@ -505,10 +505,10 @@ void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans }); } -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, - int64_t m, int64_t n, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb) { - queue.submit([&](sycl::handler &cgh) { +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, + int64_t m, int64_t n, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { @@ -521,34 +521,34 @@ void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans }); } -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, - int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb) { - queue.submit([&](sycl::handler &cgh) { +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, + int64_t m, int64_t n, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { ::cblas_ctrsm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(transa), convert_to_cblas_diag(unit_diag), (const int)m, (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_b.GET_MULTI_PTR, (const int)ldb); }); }); } -void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, +void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &b, int64_t ldb) { - queue.submit([&](sycl::handler &cgh) { + sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& b, int64_t ldb) { + queue.submit([&](sycl::handler& cgh) { auto accessor_a = a.get_access(cgh); auto accessor_b = b.get_access(cgh); host_task(cgh, [=]() { ::cblas_ztrsm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(transa), convert_to_cblas_diag(unit_diag), (const int)m, (const int)n, - (const void *)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, + (const void*)&alpha, accessor_a.GET_MULTI_PTR, (const int)lda, accessor_b.GET_MULTI_PTR, (const int)ldb); }); }); @@ -556,10 +556,10 @@ void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans // USM APIs -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const float *a, int64_t lda, const float *b, int64_t ldb, - float beta, float *c, int64_t ldc, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const float* a, int64_t lda, const float* b, int64_t ldb, + float beta, float* c, int64_t ldc, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -573,11 +573,11 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t return done; } -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, double alpha, const double *a, int64_t lda, const double *b, - int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -591,50 +591,48 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t return done; } -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_cgemm(MAJOR, convert_to_cblas_trans(transa), convert_to_cblas_trans(transb), - (const int)m, (const int)n, (const int)k, (const void *)&alpha, a, - (const int)lda, b, (const int)ldb, (const void *)&beta, c, - (const int)ldc); + (const int)m, (const int)n, (const int)k, (const void*)&alpha, a, + (const int)lda, b, (const int)ldb, (const void*)&beta, c, (const int)ldc); }); }); return done; } -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_zgemm(MAJOR, convert_to_cblas_trans(transa), convert_to_cblas_trans(transb), - (const int)m, (const int)n, (const int)k, (const void *)&alpha, a, - (const int)lda, b, (const int)ldb, (const void *)&beta, c, - (const int)ldc); + (const int)m, (const int)n, (const int)k, (const void*)&alpha, a, + (const int)lda, b, (const int)ldb, (const void*)&beta, c, (const int)ldc); }); }); return done; } -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, sycl::half alpha, const sycl::half *a, int64_t lda, const sycl::half *b, - int64_t ldb, sycl::half beta, sycl::half *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, sycl::half alpha, const sycl::half* a, int64_t lda, const sycl::half* b, + int64_t ldb, sycl::half beta, sycl::half* c, int64_t ldc, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm", "for column_major layout"); #endif @@ -643,10 +641,10 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t #endif } -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const sycl::half *a, int64_t lda, const sycl::half *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const sycl::half* a, int64_t lda, const sycl::half* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm", "for column_major layout"); #endif @@ -655,10 +653,10 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t #endif } -sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const bfloat16 *a, int64_t lda, const bfloat16 *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, float alpha, const bfloat16* a, int64_t lda, const bfloat16* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { #ifdef COLUMN_MAJOR throw unimplemented("blas", "gemm", "for column_major layout"); #endif @@ -667,12 +665,12 @@ sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t #endif } -sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -680,19 +678,19 @@ sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t host_task(cgh, [=]() { ::cblas_chemm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), (const int)m, (const int)n, - (const void *)&alpha, a, (const int)lda, b, (const int)ldb, - (const void *)&beta, c, (const int)ldc); + (const void*)&alpha, a, (const int)lda, b, (const int)ldb, + (const void*)&beta, c, (const int)ldc); }); }); return done; } -sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -700,18 +698,18 @@ sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t host_task(cgh, [=]() { ::cblas_zhemm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), (const int)m, (const int)n, - (const void *)&alpha, a, (const int)lda, b, (const int)ldb, - (const void *)&beta, c, (const int)ldc); + (const void*)&alpha, a, (const int)lda, b, (const int)ldb, + (const void*)&beta, c, (const int)ldc); }); }); return done; } -sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, const std::complex *a, int64_t lda, float beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, const std::complex* a, int64_t lda, float beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -725,11 +723,11 @@ sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t return done; } -sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, const std::complex *a, int64_t lda, double beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, const std::complex* a, int64_t lda, double beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -743,46 +741,46 @@ sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t return done; } -sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, float beta, std::complex *c, - int64_t ldc, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, float beta, std::complex* c, + int64_t ldc, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_cher2k(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, a, (const int)lda, b, + (const int)n, (const int)k, (const void*)&alpha, a, (const int)lda, b, (const int)ldb, (const float)beta, c, (const int)ldc); }); }); return done; } -sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, double beta, std::complex *c, - int64_t ldc, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, double beta, std::complex* c, + int64_t ldc, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_zher2k(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, a, (const int)lda, b, + (const int)n, (const int)k, (const void*)&alpha, a, (const int)lda, b, (const int)ldb, (const double)beta, c, (const int)ldc); }); }); return done; } -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - float alpha, const float *a, int64_t lda, const float *b, int64_t ldb, float beta, - float *c, int64_t ldc, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + float alpha, const float* a, int64_t lda, const float* b, int64_t ldb, float beta, + float* c, int64_t ldc, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -797,11 +795,11 @@ sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t return done; } -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - double alpha, const double *a, int64_t lda, const double *b, int64_t ldb, - double beta, double *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + double alpha, const double* a, int64_t lda, const double* b, int64_t ldb, + double beta, double* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -816,12 +814,12 @@ sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t return done; } -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -829,19 +827,19 @@ sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t host_task(cgh, [=]() { ::cblas_csymm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), (const int)m, (const int)n, - (const void *)&alpha, a, (const int)lda, b, (const int)ldb, - (const void *)&beta, c, (const int)ldc); + (const void*)&alpha, a, (const int)lda, b, (const int)ldb, + (const void*)&beta, c, (const int)ldc); }); }); return done; } -sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -849,17 +847,17 @@ sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t host_task(cgh, [=]() { ::cblas_zsymm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), (const int)m, (const int)n, - (const void *)&alpha, a, (const int)lda, b, (const int)ldb, - (const void *)&beta, c, (const int)ldc); + (const void*)&alpha, a, (const int)lda, b, (const int)ldb, + (const void*)&beta, c, (const int)ldc); }); }); return done; } -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, const float *a, int64_t lda, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, const float* a, int64_t lda, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -873,10 +871,10 @@ sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t return done; } -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, const double *a, int64_t lda, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, const double* a, int64_t lda, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -890,46 +888,46 @@ sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t return done; } -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_csyrk(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, a, (const int)lda, - (const void *)&beta, c, (const int)ldc); + (const int)n, (const int)k, (const void*)&alpha, a, (const int)lda, + (const void*)&beta, c, (const int)ldc); }); }); return done; } -sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_zsyrk(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, a, (const int)lda, - (const void *)&beta, c, (const int)ldc); + (const int)n, (const int)k, (const void*)&alpha, a, (const int)lda, + (const void*)&beta, c, (const int)ldc); }); }); return done; } -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, const float *a, int64_t lda, const float *b, int64_t ldb, float beta, - float *c, int64_t ldc, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + float alpha, const float* a, int64_t lda, const float* b, int64_t ldb, float beta, + float* c, int64_t ldc, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -943,11 +941,11 @@ sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t return done; } -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, const double *a, int64_t lda, const double *b, int64_t ldb, - double beta, double *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + double alpha, const double* a, int64_t lda, const double* b, int64_t ldb, + double beta, double* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -961,48 +959,48 @@ sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t return done; } -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_csyr2k(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, a, (const int)lda, b, - (const int)ldb, (const void *)&beta, c, (const int)ldc); + (const int)n, (const int)k, (const void*)&alpha, a, (const int)lda, b, + (const int)ldb, (const void*)&beta, c, (const int)ldc); }); }); return done; } -sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } host_task(cgh, [=]() { ::cblas_zsyr2k(MAJOR, convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - (const int)n, (const int)k, (const void *)&alpha, a, (const int)lda, b, - (const int)ldb, (const void *)&beta, c, (const int)ldc); + (const int)n, (const int)k, (const void*)&alpha, a, (const int)lda, b, + (const int)ldb, (const void*)&beta, c, (const int)ldc); }); }); return done; } -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, - diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, - float *b, int64_t ldb, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, + diag unit_diag, int64_t m, int64_t n, float alpha, const float* a, int64_t lda, + float* b, int64_t ldb, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1017,10 +1015,10 @@ sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpos return done; } -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, - diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, - double *b, int64_t ldb, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, + diag unit_diag, int64_t m, int64_t n, double alpha, const double* a, int64_t lda, + double* b, int64_t ldb, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1035,11 +1033,11 @@ sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpos return done; } -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1048,17 +1046,17 @@ sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpos ::cblas_ctrmm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(transa), convert_to_cblas_diag(unit_diag), (const int)m, (const int)n, - (const void *)&alpha, a, (const int)lda, b, (const int)ldb); + (const void*)&alpha, a, (const int)lda, b, (const int)ldb); }); }); return done; } -sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, +sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1067,16 +1065,16 @@ sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpos ::cblas_ztrmm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(transa), convert_to_cblas_diag(unit_diag), (const int)m, (const int)n, - (const void *)&alpha, a, (const int)lda, b, (const int)ldb); + (const void*)&alpha, a, (const int)lda, b, (const int)ldb); }); }); return done; } -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, - diag unit_diag, int64_t m, int64_t n, float alpha, const float *a, int64_t lda, - float *b, int64_t ldb, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, + diag unit_diag, int64_t m, int64_t n, float alpha, const float* a, int64_t lda, + float* b, int64_t ldb, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1091,10 +1089,10 @@ sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpos return done; } -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, - diag unit_diag, int64_t m, int64_t n, double alpha, const double *a, int64_t lda, - double *b, int64_t ldb, const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, + diag unit_diag, int64_t m, int64_t n, double alpha, const double* a, int64_t lda, + double* b, int64_t ldb, const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1109,11 +1107,11 @@ sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpos return done; } -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1122,17 +1120,17 @@ sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpos ::cblas_ctrsm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(transa), convert_to_cblas_diag(unit_diag), (const int)m, (const int)n, - (const void *)&alpha, a, (const int)lda, b, (const int)ldb); + (const void*)&alpha, a, (const int)lda, b, (const int)ldb); }); }); return done; } -sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose transa, +sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose transa, diag unit_diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - const std::vector &dependencies) { - auto done = queue.submit([&](sycl::handler &cgh) { + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + const std::vector& dependencies) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); @@ -1141,7 +1139,7 @@ sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpos ::cblas_ztrsm(MAJOR, convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(transa), convert_to_cblas_diag(unit_diag), (const int)m, (const int)n, - (const void *)&alpha, a, (const int)lda, b, (const int)ldb); + (const void*)&alpha, a, (const int)lda, b, (const int)ldb); }); }); return done; diff --git a/src/blas/backends/portblas/portblas_batch.cxx b/src/blas/backends/portblas/portblas_batch.cxx index e48e60890..75b1a115b 100644 --- a/src/blas/backends/portblas/portblas_batch.cxx +++ b/src/blas/backends/portblas/portblas_batch.cxx @@ -19,999 +19,1002 @@ // Buffer APIs -void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, +void syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "syrk_batch", ""); } -void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, +void syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "syrk_batch", ""); } -void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +void syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "syrk_batch", ""); } -void syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +void syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "syrk_batch", ""); } -void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, float beta, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, float beta, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { throw unimplemented("blas", "gemv_batch", ""); } -void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, double beta, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, double beta, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { throw unimplemented("blas", "gemv_batch", ""); } -void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, +void gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { throw unimplemented("blas", "gemv_batch", ""); } -void gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, +void gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { throw unimplemented("blas", "gemv_batch", ""); } -void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", ""); } -void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", ""); } -void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", ""); } -void dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { throw unimplemented("blas", "dgmm_batch", ""); } -void axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer &y, +void axpy_batch(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { CALL_PORTBLAS_FN(::blas::_axpy_batch, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer &y, +void axpy_batch(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { CALL_PORTBLAS_FN(::blas::_axpy_batch, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, +void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { throw unimplemented("blas", "axpy_batch", ""); } -void axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, +void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { throw unimplemented("blas", "axpy_batch", ""); } -void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, std::int64_t incy, +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { throw unimplemented("blas", "copy_batch", ""); } -void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, std::int64_t incy, +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { throw unimplemented("blas", "copy_batch", ""); } -void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { throw unimplemented("blas", "copy_batch", ""); } -void copy_batch(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, +void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { throw unimplemented("blas", "copy_batch", ""); } -void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { CALL_PORTBLAS_FN(::blas::_gemm_strided_batched, queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, double beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { CALL_PORTBLAS_FN(::blas::_gemm_strided_batched, queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "gemm_batch", " for complex"); } -void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "gemm_batch", " for complex"); } -void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::half beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::half beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "gemm_batch", " for complex"); } -void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "gemm_batch", " for unsupported dtype"); } -void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "gemm_batch", " for unsupported dtype"); } -void gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "gemm_batch", " for unsupported dtype"); } -void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +void trsm_batch(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("blas", "trsm_batch", ""); } -void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +void trsm_batch(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("blas", "trsm_batch", ""); } -void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +void trsm_batch(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("blas", "trsm_batch", ""); } -void trsm_batch(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +void trsm_batch(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("blas", "trsm_batch", ""); } -void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, +void omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { CALL_PORTBLAS_FN(::blas::_omatcopy_batch, queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, +void omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { CALL_PORTBLAS_FN(::blas::_omatcopy_batch, queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, +void omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("blas", "omatcopy_batch", ""); } -void omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, +void omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("blas", "omatcopy_batch", ""); } -void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, +void imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", ""); } -void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, +void imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", ""); } -void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, +void imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", ""); } -void imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, +void imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", ""); } -void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { +void omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, float beta, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size) { CALL_PORTBLAS_FN(::blas::_omatadd_batch, queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { +void omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, double beta, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size) { CALL_PORTBLAS_FN(::blas::_omatadd_batch, queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { +void omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size) { throw unimplemented("blas", "omatadd_batch", ""); } -void omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, +void omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, + oneapi::math::transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { throw unimplemented("blas", "omatadd_batch", ""); } // USM APIs -sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, - float *alpha, const float **a, std::int64_t *lda, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* k, + float* alpha, const float** a, std::int64_t* lda, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, - double *alpha, const double **a, std::int64_t *lda, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* k, + double* alpha, const double** a, std::int64_t* lda, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, +sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const float* a, std::int64_t lda, std::int64_t stride_a, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, +sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, +sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, std::complex *c, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event syrk_batch(sycl::queue &queue, oneapi::math::uplo upper_lower, +sycl::event syrk_batch(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, std::complex *c, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "syrk_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float beta, float *y, std::int64_t incy, +sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float beta, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double beta, double *y, std::int64_t incy, +sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double beta, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, +sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, std::int64_t incx, std::int64_t stridex, std::complex beta, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, +sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, std::int64_t incx, std::int64_t stridex, std::complex beta, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float *beta, float **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, + const float** x, std::int64_t* incx, float* beta, float** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double *beta, double **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, + const double** x, std::int64_t* incx, double* beta, double** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event gemv_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemv_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemv_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, - std::int64_t n, const float *a, std::int64_t lda, std::int64_t stridea, - const float *x, std::int64_t incx, std::int64_t stridex, float *c, +sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, + std::int64_t n, const float* a, std::int64_t lda, std::int64_t stridea, + const float* x, std::int64_t incx, std::int64_t stridex, float* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, - std::int64_t n, const double *a, std::int64_t lda, std::int64_t stridea, - const double *x, std::int64_t incx, std::int64_t stridex, double *c, +sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, + std::int64_t n, const double* a, std::int64_t lda, std::int64_t stridea, + const double* x, std::int64_t incx, std::int64_t stridex, double* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *c, std::int64_t ldc, +sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, - std::int64_t n, const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *c, std::int64_t ldc, +sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, + std::int64_t n, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, - std::int64_t *n, const float **a, std::int64_t *lda, const float **x, - std::int64_t *incx, float **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, + std::int64_t* n, const float** a, std::int64_t* lda, const float** x, + std::int64_t* incx, float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, - std::int64_t *n, const double **a, std::int64_t *lda, const double **x, - std::int64_t *incx, double **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, + std::int64_t* n, const double** a, std::int64_t* lda, const double** x, + std::int64_t* incx, double** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, - std::int64_t *n, const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, + std::int64_t* n, const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event dgmm_batch(sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, - std::int64_t *n, const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, + std::int64_t* n, const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "dgmm_batch", " for USM"); } -sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, float *alpha, const float **x, - std::int64_t *incx, float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, float* alpha, const float** x, + std::int64_t* incx, float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", " for USM"); } -sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, double *alpha, const double **x, - std::int64_t *incx, double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, double* alpha, const double** x, + std::int64_t* incx, double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", " for USM"); } -sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, std::complex* alpha, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", " for USM"); } -sycl::event axpy_batch(sycl::queue &queue, std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, std::int64_t* n, std::complex* alpha, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", " for USM"); } -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, float alpha, const float *x, - std::int64_t incx, std::int64_t stridex, float *y, std::int64_t incy, +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, float alpha, const float* x, + std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_axpy_batch, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, double alpha, const double *x, - std::int64_t incx, std::int64_t stridex, double *y, std::int64_t incy, +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, double alpha, const double* x, + std::int64_t incx, std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_axpy_batch, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", " for USM"); } -sycl::event axpy_batch(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "axpy_batch", " for USM"); } -sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const float **x, std::int64_t *incx, - float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const float** x, std::int64_t* incx, + float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", " for USM"); } -sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const double **x, std::int64_t *incx, - double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const double** x, std::int64_t* incx, + double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", " for USM"); } -sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", " for USM"); } -sycl::event copy_batch(sycl::queue &queue, std::int64_t *n, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, std::int64_t* n, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", " for USM"); } -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", " for USM"); } -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", " for USM"); } -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", " for USM"); } -sycl::event copy_batch(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, +sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "copy_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, - const float **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float* alpha, const float** a, std::int64_t* lda, + const float** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, - const double **b, std::int64_t *ldb, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, double* alpha, const double** a, std::int64_t* lda, + const double** b, std::int64_t* ldb, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, sycl::half *alpha, const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, sycl::half *beta, sycl::half **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, sycl::half* alpha, const sycl::half** a, std::int64_t* lda, + const sycl::half** b, std::int64_t* ldb, sycl::half* beta, sycl::half** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float *alpha, const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float* alpha, const sycl::half** a, std::int64_t* lda, + const sycl::half** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose *transa, - oneapi::math::transpose *transb, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose* transa, + oneapi::math::transpose* transb, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, std::int32_t** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, const float *b, std::int64_t ldb, - std::int64_t stride_b, float beta, float *c, std::int64_t ldc, + std::int64_t k, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, const float* b, std::int64_t ldb, + std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_gemm_strided_batched, queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, const double *b, std::int64_t ldb, - std::int64_t stride_b, double beta, double *c, std::int64_t ldc, + std::int64_t k, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, const double* b, std::int64_t ldb, + std::int64_t stride_b, double beta, double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_gemm_strided_batched, queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, const std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, const std::complex *b, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, const std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::complex beta, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::half alpha, const sycl::half *a, std::int64_t lda, - std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, + std::int64_t k, sycl::half alpha, const sycl::half* a, std::int64_t lda, + std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, sycl::half* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const sycl::half *a, std::int64_t lda, - std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, float beta, float *c, std::int64_t ldc, + std::int64_t k, float alpha, const sycl::half* a, std::int64_t lda, + std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, - std::int64_t stride_b, float beta, float *c, std::int64_t ldc, + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int64_t stride_a, const std::int8_t* b, std::int64_t ldb, + std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event gemm_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, - std::int64_t stride_b, float beta, std::int32_t *c, std::int64_t ldc, + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int64_t stride_a, const std::int8_t* b, std::int64_t ldb, + std::int64_t stride_b, float beta, std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "gemm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, +sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float *b, + const float* a, std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, +sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double *b, + const double* a, std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, +sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side left_right, +sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, - oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, float *alpha, - const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side* left_right, + oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + oneapi::math::diag* unit_diag, std::int64_t* m, std::int64_t* n, + float* alpha, const float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, - oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, - double *alpha, const double **a, std::int64_t *lda, double **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side* left_right, + oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + oneapi::math::diag* unit_diag, std::int64_t* m, std::int64_t* n, + double* alpha, const double** a, std::int64_t* lda, double** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, - oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side* left_right, + oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + oneapi::math::diag* unit_diag, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event trsm_batch(sycl::queue &queue, oneapi::math::side *left_right, - oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - oneapi::math::diag *unit_diag, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(sycl::queue& queue, oneapi::math::side* left_right, + oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + oneapi::math::diag* unit_diag, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "trsm_batch", " for USM"); } -sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_omatcopy_batch, queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, +sycl::event omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_omatcopy_batch, queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, +sycl::event omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", " for USM"); } -sycl::event omatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, +sycl::event omatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", " for USM"); } -sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, float *ab, std::int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", " for USM"); } -sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, double *ab, std::int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", " for USM"); } -sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, +sycl::event imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", " for USM"); } -sycl::event imatcopy_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, std::complex *ab, +sycl::event imatcopy_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", " for USM"); } -sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, - float beta, const float *b, std::int64_t ldb, std::int64_t stride_b, - float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, + float beta, const float* b, std::int64_t ldb, std::int64_t stride_b, + float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_omatadd_batch, queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, - double beta, const double *b, std::int64_t ldb, std::int64_t stride_b, - double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, + double beta, const double* b, std::int64_t ldb, std::int64_t stride_b, + double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_omatadd_batch, queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "omatadd_batch", " for USM"); } -sycl::event omatadd_batch(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event omatadd_batch(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { throw unimplemented("blas", "omatadd_batch", " for USM"); } diff --git a/src/blas/backends/portblas/portblas_common.hpp b/src/blas/backends/portblas/portblas_common.hpp index 27d68b3d4..c2713c2f5 100644 --- a/src/blas/backends/portblas/portblas_common.hpp +++ b/src/blas/backends/portblas/portblas_common.hpp @@ -54,9 +54,9 @@ template struct portblas_type; #define DEF_PORTBLAS_TYPE(onemath_t, portblas_t) \ - template <> \ + template <> \ struct portblas_type { \ - using type = portblas_t; \ + using type = portblas_t; \ }; DEF_PORTBLAS_TYPE(sycl::queue, handle_t) @@ -181,7 +181,7 @@ struct throw_if_unsupported_by_device { if (checkTypeInPack) { if (!q.get_info().has(AspectVal)) { throw math::unsupported_device("blas", message, - q.get_info()); + q.get_info()); } } } diff --git a/src/blas/backends/portblas/portblas_gemm_bias.cxx b/src/blas/backends/portblas/portblas_gemm_bias.cxx index 2af4d0bd2..83e39b2d9 100644 --- a/src/blas/backends/portblas/portblas_gemm_bias.cxx +++ b/src/blas/backends/portblas/portblas_gemm_bias.cxx @@ -19,72 +19,72 @@ // Buffer APIs -void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { + float alpha, sycl::buffer& a, std::int64_t lda, int8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", ""); } -void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, int8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { + float alpha, sycl::buffer& a, std::int64_t lda, int8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", ""); } -void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { + float alpha, sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", ""); } -void gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { + float alpha, sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", ""); } // USM APIs -sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, - std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const std::int8_t* a, + std::int64_t lda, std::int8_t ao, const std::uint8_t* b, std::int64_t ldb, + std::uint8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", " for USM"); } -sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int8_t ao, const std::int8_t *b, std::int64_t ldb, - std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const std::int8_t* a, + std::int64_t lda, std::int8_t ao, const std::int8_t* b, std::int64_t ldb, + std::int8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", " for USM"); } -sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, - std::int64_t lda, std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, - std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const std::uint8_t* a, + std::int64_t lda, std::uint8_t ao, const std::int8_t* b, std::int64_t ldb, + std::int8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", " for USM"); } -sycl::event gemm_bias(sycl::queue &queue, oneapi::math::transpose transa, +sycl::event gemm_bias(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, const std::uint8_t *a, - std::int64_t lda, std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, - std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies) { + std::int64_t n, std::int64_t k, float alpha, const std::uint8_t* a, + std::int64_t lda, std::uint8_t ao, const std::uint8_t* b, std::int64_t ldb, + std::uint8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", " for USM"); } diff --git a/src/blas/backends/portblas/portblas_level1.cxx b/src/blas/backends/portblas/portblas_level1.cxx index 0a0af855c..6d1f39463 100644 --- a/src/blas/backends/portblas/portblas_level1.cxx +++ b/src/blas/backends/portblas/portblas_level1.cxx @@ -19,91 +19,91 @@ // Buffer APIs -void dotc(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotc(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { throw unimplemented("blas", "dotc", ""); } -void dotu(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotu(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { throw unimplemented("blas", "dotu", ""); } -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { CALL_PORTBLAS_FN(::blas::_iamax, queue, n, x, incx, result); } -void iamax(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { throw unimplemented("blas", "iamax", ""); } -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { CALL_PORTBLAS_FN(::blas::_iamin, queue, n, x, incx, result); } -void iamin(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { throw unimplemented("blas", "iamin", ""); } -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { throw unimplemented("blas", "asum", ""); } -void asum(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { // portBLAS asum implementation requires that result is initialized to zero // before performing the computation. - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto result_acc = result.template get_access(cgh); cgh.single_task([=]() { result_acc[0] = real_t(0); }); }); CALL_PORTBLAS_FN(::blas::_asum, queue, n, x, incx, result); } -void axpy(sycl::queue &queue, std::int64_t n, real_t alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void axpy(sycl::queue& queue, std::int64_t n, real_t alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_axpy, queue, n, alpha, x, incx, y, incy); } -void axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { throw unimplemented("blas", "axpy", "for complex"); } -void axpby(sycl::queue &queue, std::int64_t n, real_t alpha, sycl::buffer &x, - std::int64_t incx, real_t beta, sycl::buffer &y, std::int64_t incy) { +void axpby(sycl::queue& queue, std::int64_t n, real_t alpha, sycl::buffer& x, + std::int64_t incx, real_t beta, sycl::buffer& y, std::int64_t incy) { throw unimplemented("blas", "axpby", ""); } -void axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { throw unimplemented("blas", "axpby", ""); } -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy) { +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_copy, queue, n, x, incx, y, incy); } -void copy(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { +void copy(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy) { throw unimplemented("blas", "copy", " for complex."); } -void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { +void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& result) { // portBLAS dot implementation requires that result is initialized to zero // before performing the computation. - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto result_acc = result.template get_access(cgh); cgh.single_task([=]() { result_acc[0] = real_t(0); }); }); @@ -111,288 +111,288 @@ void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::in } #ifdef ENABLE_MIXED_PRECISION_WITH_DOUBLE -void dot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer &result) { +void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& result) { throw unimplemented("blas", "dot", " for unmatched return type"); } #endif -void sdsdot(sycl::queue &queue, std::int64_t n, real_t sb, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void sdsdot(sycl::queue& queue, std::int64_t n, real_t sb, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { // portBLAS sdsdot implementation requires that result is initialized to zero // before performing the computation. - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto result_acc = result.template get_access(cgh); cgh.single_task([=]() { result_acc[0] = real_t(0); }); }); CALL_PORTBLAS_FN(::blas::_sdsdot, queue, n, sb, x, incx, y, incy, result); } -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer &result) { +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer& result) { throw unimplemented("blas", "nrm2", " for complex"); } -void nrm2(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &result) { +void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& result) { // portBLAS nrm2 implementation requires that result is initialized to zero // before performing the computation. - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto result_acc = result.template get_access(cgh); cgh.single_task([=]() { result_acc[0] = real_t(0); }); }); CALL_PORTBLAS_FN(::blas::_nrm2, queue, n, x, incx, result); } -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy, real_t c, +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, real_t c, real_t s) { throw unimplemented("blas", "rot", " for complex"); } -void rot(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, real_t c, real_t s) { +void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, real_t c, real_t s) { CALL_PORTBLAS_FN(::blas::_rot, queue, n, x, incx, y, incy, c, s); } -void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s) { +void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s) { CALL_PORTBLAS_FN(::blas::_rotg, queue, a, b, c, s); } -void rotg(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { throw unimplemented("blas", "rotg", " for complex"); } -void rotm(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, sycl::buffer ¶m) { +void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, sycl::buffer& param) { CALL_PORTBLAS_FN(::blas::_rotm, queue, n, x, incx, y, incy, param); } -void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, real_t y1, sycl::buffer ¶m) { +void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, real_t y1, sycl::buffer& param) { CALL_PORTBLAS_FN(::blas::_rotmg, queue, d1, d2, x1, y1, param); } -void scal(sycl::queue &queue, std::int64_t n, real_t alpha, sycl::buffer &x, +void scal(sycl::queue& queue, std::int64_t n, real_t alpha, sycl::buffer& x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_scal, queue, n, alpha, x, incx); } -void scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx) { throw unimplemented("blas", "scal", " for complex"); } -void scal(sycl::queue &queue, std::int64_t n, real_t alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +void scal(sycl::queue& queue, std::int64_t n, real_t alpha, + sycl::buffer, 1>& x, std::int64_t incx) { throw unimplemented("blas", "scal", " for complex"); } -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy) { +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_swap, queue, n, x, incx, y, incy); } -void swap(sycl::queue &queue, std::int64_t n, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, std::int64_t incy) { +void swap(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy) { throw unimplemented("blas", "swap", " for complex"); } // USM APIs -sycl::event dotc(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, const std::vector &dependencies) { +sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies) { throw unimplemented("blas", "dotc", " for USM"); } -sycl::event dotu(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, const std::complex *y, std::int64_t incy, - std::complex *result, const std::vector &dependencies) { +sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, const std::complex* y, std::int64_t incy, + std::complex* result, const std::vector& dependencies) { throw unimplemented("blas", "dotu", " for USM"); } -sycl::event iamax(sycl::queue &queue, std::int64_t n, const real_t *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies) { +sycl::event iamax(sycl::queue& queue, std::int64_t n, const real_t* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_iamax, queue, n, x, incx, result, dependencies); } -sycl::event iamax(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { throw unimplemented("blas", "iamax", " for USM"); } -sycl::event iamin(sycl::queue &queue, std::int64_t n, const real_t *x, std::int64_t incx, - std::int64_t *result, const std::vector &dependencies) { +sycl::event iamin(sycl::queue& queue, std::int64_t n, const real_t* x, std::int64_t incx, + std::int64_t* result, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_iamin, queue, n, x, incx, result, dependencies); } -sycl::event iamin(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { throw unimplemented("blas", "iamin", " for USM"); } -sycl::event asum(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, real_t *result, const std::vector &dependencies) { +sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, real_t* result, const std::vector& dependencies) { throw unimplemented("blas", "asum", " for USM"); } -sycl::event asum(sycl::queue &queue, std::int64_t n, const real_t *x, std::int64_t incx, - real_t *result, const std::vector &dependencies) { +sycl::event asum(sycl::queue& queue, std::int64_t n, const real_t* x, std::int64_t incx, + real_t* result, const std::vector& dependencies) { // portBLAS asum implementation requires result to be initializes to zero // before starting the computation. auto init_res_val = queue.submit( - [&](sycl::handler &cgh) { cgh.single_task([=]() { result[0] = real_t(0); }); }); + [&](sycl::handler& cgh) { cgh.single_task([=]() { result[0] = real_t(0); }); }); std::vector new_dependencies = dependencies; new_dependencies.push_back(init_res_val); CALL_PORTBLAS_USM_FN(::blas::_asum, queue, n, x, incx, result, new_dependencies); } -sycl::event axpy(sycl::queue &queue, std::int64_t n, real_t alpha, const real_t *x, - std::int64_t incx, real_t *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpy(sycl::queue& queue, std::int64_t n, real_t alpha, const real_t* x, + std::int64_t incx, real_t* y, std::int64_t incy, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_axpy, queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { throw unimplemented("blas", "axpy", " for USM"); } -sycl::event axpby(sycl::queue &queue, std::int64_t n, real_t alpha, const real_t *x, - std::int64_t incx, const real_t beta, real_t *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, std::int64_t n, real_t alpha, const real_t* x, + std::int64_t incx, const real_t beta, real_t* y, std::int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", " for USM"); } -sycl::event axpby(sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, const std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", " for USM"); } -sycl::event copy(sycl::queue &queue, std::int64_t n, const real_t *x, std::int64_t incx, real_t *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event copy(sycl::queue& queue, std::int64_t n, const real_t* x, std::int64_t incx, real_t* y, + std::int64_t incy, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_copy, queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "copy", " for USM"); } -sycl::event dot(sycl::queue &queue, std::int64_t n, const real_t *x, std::int64_t incx, - const real_t *y, std::int64_t incy, real_t *result, - const std::vector &dependencies) { +sycl::event dot(sycl::queue& queue, std::int64_t n, const real_t* x, std::int64_t incx, + const real_t* y, std::int64_t incy, real_t* result, + const std::vector& dependencies) { // portBLAS dot implementation requires result to be initializes to zero // before starting the computation. auto init_res_val = queue.submit( - [&](sycl::handler &cgh) { cgh.single_task([=]() { result[0] = real_t(0); }); }); + [&](sycl::handler& cgh) { cgh.single_task([=]() { result[0] = real_t(0); }); }); std::vector new_dependencies = dependencies; new_dependencies.emplace_back(init_res_val); CALL_PORTBLAS_USM_FN(::blas::_dot, queue, n, x, incx, y, incy, result, new_dependencies); } #ifdef ENABLE_MIXED_PRECISION_WITH_DOUBLE -sycl::event dot(sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, double* result, + const std::vector& dependencies) { throw unimplemented("blas", "dot", " for USM"); } #endif -sycl::event sdsdot(sycl::queue &queue, std::int64_t n, real_t sb, const real_t *x, - std::int64_t incx, const real_t *y, std::int64_t incy, real_t *result, - const std::vector &dependencies) { +sycl::event sdsdot(sycl::queue& queue, std::int64_t n, real_t sb, const real_t* x, + std::int64_t incx, const real_t* y, std::int64_t incy, real_t* result, + const std::vector& dependencies) { // portBLAS sdsdot implementation requires result to be initializes to zero // before starting the computation. auto init_res_val = queue.submit( - [&](sycl::handler &cgh) { cgh.single_task([=]() { result[0] = real_t(0); }); }); + [&](sycl::handler& cgh) { cgh.single_task([=]() { result[0] = real_t(0); }); }); std::vector new_dependencies = dependencies; new_dependencies.emplace_back(init_res_val); CALL_PORTBLAS_USM_FN(::blas::_sdsdot, queue, n, sb, x, incx, y, incy, result, new_dependencies); } -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const std::complex *x, - std::int64_t incx, real_t *result, const std::vector &dependencies) { +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, + std::int64_t incx, real_t* result, const std::vector& dependencies) { throw unimplemented("blas", "nrm2", " for USM"); } -sycl::event nrm2(sycl::queue &queue, std::int64_t n, const real_t *x, std::int64_t incx, - real_t *result, const std::vector &dependencies) { +sycl::event nrm2(sycl::queue& queue, std::int64_t n, const real_t* x, std::int64_t incx, + real_t* result, const std::vector& dependencies) { // portBLAS nrm2 implementation requires result to be initializes to zero // before starting the computation. auto init_res_val = queue.submit( - [&](sycl::handler &cgh) { cgh.single_task([=]() { result[0] = real_t(0); }); }); + [&](sycl::handler& cgh) { cgh.single_task([=]() { result[0] = real_t(0); }); }); std::vector new_dependencies = dependencies; new_dependencies.push_back(init_res_val); CALL_PORTBLAS_USM_FN(::blas::_nrm2, queue, n, x, incx, result, new_dependencies); } -sycl::event rot(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, real_t c, real_t s, - const std::vector &dependencies) { +sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, real_t c, real_t s, + const std::vector& dependencies) { throw unimplemented("blas", "rot", " for USM"); } -sycl::event rot(sycl::queue &queue, std::int64_t n, real_t *x, std::int64_t incx, real_t *y, +sycl::event rot(sycl::queue& queue, std::int64_t n, real_t* x, std::int64_t incx, real_t* y, std::int64_t incy, real_t c, real_t s, - const std::vector &dependencies) { + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_rot, queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rotg(sycl::queue &queue, real_t *a, real_t *b, real_t *c, real_t *s, - const std::vector &dependencies) { +sycl::event rotg(sycl::queue& queue, real_t* a, real_t* b, real_t* c, real_t* s, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_rotg, queue, a, b, c, s, dependencies); } -sycl::event rotg(sycl::queue &queue, std::complex *a, std::complex *b, real_t *c, - std::complex *s, const std::vector &dependencies) { +sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, real_t* c, + std::complex* s, const std::vector& dependencies) { throw unimplemented("blas", "rotg", " for USM"); } -sycl::event rotm(sycl::queue &queue, std::int64_t n, real_t *x, std::int64_t incx, real_t *y, - std::int64_t incy, real_t *param, const std::vector &dependencies) { +sycl::event rotm(sycl::queue& queue, std::int64_t n, real_t* x, std::int64_t incx, real_t* y, + std::int64_t incy, real_t* param, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_rotm, queue, n, x, incx, y, incy, param, dependencies); } -sycl::event rotmg(sycl::queue &queue, real_t *d1, real_t *d2, real_t *x1, real_t y1, real_t *param, - const std::vector &dependencies) { +sycl::event rotmg(sycl::queue& queue, real_t* d1, real_t* d2, real_t* x1, real_t y1, real_t* param, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_rotmg, queue, d1, d2, x1, y1, param, dependencies); } -sycl::event scal(sycl::queue &queue, std::int64_t n, real_t alpha, real_t *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(sycl::queue& queue, std::int64_t n, real_t alpha, real_t* x, std::int64_t incx, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_scal, queue, n, alpha, x, incx, dependencies); } -sycl::event scal(sycl::queue &queue, std::int64_t n, std::complex alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(sycl::queue& queue, std::int64_t n, std::complex alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { throw unimplemented("blas", "scal", " for USM"); } -sycl::event scal(sycl::queue &queue, std::int64_t n, real_t alpha, std::complex *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event scal(sycl::queue& queue, std::int64_t n, real_t alpha, std::complex* x, + std::int64_t incx, const std::vector& dependencies) { throw unimplemented("blas", "scal", " for USM"); } -sycl::event swap(sycl::queue &queue, std::int64_t n, real_t *x, std::int64_t incx, real_t *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event swap(sycl::queue& queue, std::int64_t n, real_t* x, std::int64_t incx, real_t* y, + std::int64_t incy, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_swap, queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(sycl::queue &queue, std::int64_t n, std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "swap", " for USM"); } diff --git a/src/blas/backends/portblas/portblas_level2.cxx b/src/blas/backends/portblas/portblas_level2.cxx index 8d01deec2..e1bc4766a 100644 --- a/src/blas/backends/portblas/portblas_level2.cxx +++ b/src/blas/backends/portblas/portblas_level2.cxx @@ -19,452 +19,452 @@ // Buffer APIs -void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, real_t beta, sycl::buffer &y, std::int64_t incy) { +void gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + real_t alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, real_t beta, sycl::buffer& y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_gemv, queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -void gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { throw unimplemented("blas", "gemv", " for complex"); } -void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, real_t alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, real_t beta, - sycl::buffer &y, std::int64_t incy) { +void gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, real_t alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, real_t beta, + sycl::buffer& y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_gbmv, queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -void gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, +void gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { throw unimplemented("blas", "gbmv", " for complex"); } -void ger(sycl::queue &queue, std::int64_t m, std::int64_t n, real_t alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { +void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, real_t alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { CALL_PORTBLAS_FN(::blas::_ger, queue, m, n, alpha, x, incx, y, incy, a, lda); } -void gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { throw unimplemented("blas", "gerc", ""); } -void geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { throw unimplemented("blas", "geru", ""); } -void hbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void hbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { throw unimplemented("blas", "hbmv", ""); } -void hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void hemv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { throw unimplemented("blas", "hemv", ""); } -void her(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { +void her(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { throw unimplemented("blas", "her", ""); } -void her2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void her2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { throw unimplemented("blas", "her2", ""); } -void hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void hpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { throw unimplemented("blas", "hpmv", ""); } -void hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { +void hpr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { throw unimplemented("blas", "hpr", ""); } -void hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { +void hpr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { throw unimplemented("blas", "hpr2", ""); } -void sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, - real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, real_t beta, sycl::buffer &y, std::int64_t incy) { +void sbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, + real_t alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, real_t beta, sycl::buffer& y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_sbmv, queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -void symv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, real_t beta, sycl::buffer &y, std::int64_t incy) { +void symv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, real_t beta, sycl::buffer& y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_symv, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void syr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a, +void syr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { CALL_PORTBLAS_FN(::blas::_syr, queue, upper_lower, n, alpha, x, incx, a, lda); } -void syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { +void syr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { CALL_PORTBLAS_FN(::blas::_syr2, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - sycl::buffer &a, sycl::buffer &x, std::int64_t incx, real_t beta, - sycl::buffer &y, std::int64_t incy) { +void spmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + sycl::buffer& a, sycl::buffer& x, std::int64_t incx, real_t beta, + sycl::buffer& y, std::int64_t incy) { CALL_PORTBLAS_FN(::blas::_spmv, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void spr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { +void spr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { CALL_PORTBLAS_FN(::blas::_spr, queue, upper_lower, n, alpha, x, incx, a); } -void spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a) { +void spr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { CALL_PORTBLAS_FN(::blas::_spr2, queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +void tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_tbmv, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +void tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { throw unimplemented("blas", "tbmv", ""); } -void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +void tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_tbsv, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +void tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { throw unimplemented("blas", "tbsv", ""); } -void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { +void tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_tpmv, queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +void tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { throw unimplemented("blas", "tpmv", ""); } -void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx) { +void tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_tpsv, queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +void tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { throw unimplemented("blas", "tpsv", ""); } -void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +void trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_trmv, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { throw unimplemented("blas", "trmv", " for complex"); } -void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +void trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { CALL_PORTBLAS_FN(::blas::_trsv, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { throw unimplemented("blas", "trsv", ""); } // USM APIs -sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - real_t alpha, const real_t *a, std::int64_t lda, const real_t *x, - std::int64_t incx, real_t beta, real_t *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + real_t alpha, const real_t* a, std::int64_t lda, const real_t* x, + std::int64_t incx, real_t beta, real_t* y, std::int64_t incy, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_gemv, queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event gemv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "gemv", " for USM"); } -sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, real_t alpha, const real_t *a, std::int64_t lda, - const real_t *x, std::int64_t incx, real_t beta, real_t *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, real_t alpha, const real_t* a, std::int64_t lda, + const real_t* x, std::int64_t incx, real_t beta, real_t* y, std::int64_t incy, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_gbmv, queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, +sycl::event gbmv(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { throw unimplemented("blas", "gbmv", " for USM"); } -sycl::event ger(sycl::queue &queue, std::int64_t m, std::int64_t n, real_t alpha, const real_t *x, - std::int64_t incx, const real_t *y, std::int64_t incy, real_t *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, real_t alpha, const real_t* x, + std::int64_t incx, const real_t* y, std::int64_t incy, real_t* a, std::int64_t lda, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_ger, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event gerc(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies) { throw unimplemented("blas", "gerc", " for USM"); } -sycl::event geru(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, + const std::vector& dependencies) { throw unimplemented("blas", "geru", " for USM"); } -sycl::event hbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "hbmv", " for USM"); } -sycl::event hemv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hemv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "hemv", " for USM"); } -sycl::event her(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event her(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { throw unimplemented("blas", "her", " for USM"); } -sycl::event her2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event her2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { throw unimplemented("blas", "her2", " for USM"); } -sycl::event hpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "hpmv", " for USM"); } -sycl::event hpr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - const std::complex *x, std::int64_t incx, std::complex *a, - const std::vector &dependencies) { +sycl::event hpr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + const std::complex* x, std::int64_t incx, std::complex* a, + const std::vector& dependencies) { throw unimplemented("blas", "hpr", " for USM"); } -sycl::event hpr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { +sycl::event hpr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { throw unimplemented("blas", "hpr2", " for USM"); } -sycl::event sbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, - real_t alpha, const real_t *a, std::int64_t lda, const real_t *x, - std::int64_t incx, real_t beta, real_t *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event sbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, + real_t alpha, const real_t* a, std::int64_t lda, const real_t* x, + std::int64_t incx, real_t beta, real_t* y, std::int64_t incy, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_sbmv, queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event symv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - const real_t *a, std::int64_t lda, const real_t *x, std::int64_t incx, real_t beta, - real_t *y, std::int64_t incy, const std::vector &dependencies) { +sycl::event symv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + const real_t* a, std::int64_t lda, const real_t* x, std::int64_t incx, real_t beta, + real_t* y, std::int64_t incy, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_symv, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event syr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - const real_t *x, std::int64_t incx, real_t *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event syr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + const real_t* x, std::int64_t incx, real_t* a, std::int64_t lda, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_syr, queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event syr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - const real_t *x, std::int64_t incx, const real_t *y, std::int64_t incy, real_t *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event syr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + const real_t* x, std::int64_t incx, const real_t* y, std::int64_t incy, real_t* a, + std::int64_t lda, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_syr2, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event spmv(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - const real_t *a, const real_t *x, std::int64_t incx, real_t beta, real_t *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event spmv(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + const real_t* a, const real_t* x, std::int64_t incx, real_t beta, real_t* y, + std::int64_t incy, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_spmv, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event spr(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - const real_t *x, std::int64_t incx, real_t *a, - const std::vector &dependencies) { +sycl::event spr(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + const real_t* x, std::int64_t incx, real_t* a, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_spr, queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event spr2(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, - const real_t *x, std::int64_t incx, const real_t *y, std::int64_t incy, real_t *a, - const std::vector &dependencies) { +sycl::event spr2(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, real_t alpha, + const real_t* x, std::int64_t incx, const real_t* y, std::int64_t incy, real_t* a, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_spr2, queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const real_t *a, - std::int64_t lda, real_t *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const real_t* a, + std::int64_t lda, real_t* x, std::int64_t incx, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_tbmv, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +sycl::event tbmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - const std::complex *a, std::int64_t lda, std::complex *x, - std::int64_t incx, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex* x, + std::int64_t incx, const std::vector& dependencies) { throw unimplemented("blas", "tbmv", " for USM"); } -sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const real_t *a, - std::int64_t lda, real_t *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, const real_t* a, + std::int64_t lda, real_t* x, std::int64_t incx, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_tbsv, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +sycl::event tbsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - const std::complex *a, std::int64_t lda, std::complex *x, - std::int64_t incx, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex* x, + std::int64_t incx, const std::vector& dependencies) { throw unimplemented("blas", "tbsv", " for USM"); } -sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, const real_t *a, real_t *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const real_t* a, real_t* x, + std::int64_t incx, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_tpmv, queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { throw unimplemented("blas", "tpmv", " for USM"); } -sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, const real_t *a, real_t *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const real_t* a, real_t* x, + std::int64_t incx, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_tpsv, queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { throw unimplemented("blas", "tpsv", " for USM"); } -sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, const real_t *a, std::int64_t lda, - real_t *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const real_t* a, std::int64_t lda, + real_t* x, std::int64_t incx, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_trmv, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event trmv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { throw unimplemented("blas", "trmv", " for USM"); } -sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, const real_t *a, std::int64_t lda, - real_t *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const real_t* a, std::int64_t lda, + real_t* x, std::int64_t incx, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_trsv, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - oneapi::math::diag unit_diag, std::int64_t n, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event trsv(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + oneapi::math::diag unit_diag, std::int64_t n, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { throw unimplemented("blas", "trsv", " for USM"); } diff --git a/src/blas/backends/portblas/portblas_level3.cxx b/src/blas/backends/portblas/portblas_level3.cxx index 9887d270d..d1aa32652 100644 --- a/src/blas/backends/portblas/portblas_level3.cxx +++ b/src/blas/backends/portblas/portblas_level3.cxx @@ -19,19 +19,19 @@ // Buffer APIs -void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, real_t beta, - sycl::buffer &c, std::int64_t ldc) { +void gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, real_t beta, + sycl::buffer& c, std::int64_t ldc) { CALL_PORTBLAS_FN(::blas::_gemm, queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { using sycl_complex_real_t = sycl::ext::oneapi::experimental::complex; if (transa == oneapi::math::transpose::conjtrans || transb == oneapi::math::transpose::conjtrans) { @@ -63,184 +63,184 @@ void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::tran queue.copy(out_pb_acc, out_acc); } -void symm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - std::int64_t m, std::int64_t n, real_t alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, real_t beta, - sycl::buffer &c, std::int64_t ldc) { +void symm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + std::int64_t m, std::int64_t n, real_t alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, real_t beta, + sycl::buffer& c, std::int64_t ldc) { CALL_PORTBLAS_FN(::blas::_symm, queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +void symm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { throw unimplemented("blas", "symm", ""); } -void hemm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +void hemm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { throw unimplemented("blas", "hemm", ""); } -void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer &a, - std::int64_t lda, real_t beta, sycl::buffer &c, std::int64_t ldc) { +void syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer& a, + std::int64_t lda, real_t beta, sycl::buffer& c, std::int64_t ldc) { throw unimplemented("blas", "syrk", ""); } -void syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +void syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { throw unimplemented("blas", "syrk", ""); } -void herk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer, 1> &a, - std::int64_t lda, real_t beta, sycl::buffer, 1> &c, +void herk(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer, 1>& a, + std::int64_t lda, real_t beta, sycl::buffer, 1>& c, std::int64_t ldc) { throw unimplemented("blas", "herk", ""); } -void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, real_t beta, - sycl::buffer &c, std::int64_t ldc) { +void syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, real_t alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, real_t beta, + sycl::buffer& c, std::int64_t ldc) { throw unimplemented("blas", "syr2k", ""); } -void syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +void syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { throw unimplemented("blas", "syr2k", ""); } -void her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +void her2k(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, real_t beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, real_t beta, + sycl::buffer, 1>& c, std::int64_t ldc) { throw unimplemented("blas", "her2k", ""); } -void trmm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb) { +void trmm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, + std::int64_t n, real_t alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { throw unimplemented("blas", "trmm", ""); } -void trmm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { +void trmm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { throw unimplemented("blas", "trmm", ""); } -void trsm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - real_t alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb) { +void trsm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, + std::int64_t n, real_t alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { CALL_PORTBLAS_FN(::blas::_trsm, queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { +void trsm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { throw unimplemented("blas", "trsm", " for complex"); } -void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, +void gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, real_t alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, real_t beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, real_t beta, sycl::buffer& c, std::int64_t ldc) { throw unimplemented("blas", "gemmt", ""); } -void gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, +void gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { throw unimplemented("blas", "gemmt", ""); } -void omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, real_t alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, +void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, real_t alpha, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { CALL_PORTBLAS_FN(::blas::_omatcopy, queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { +void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { throw unimplemented("blas", "omatcopy", ""); } -void omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, real_t alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &b, std::int64_t ldb, std::int64_t strideb) { +void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, real_t alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { CALL_PORTBLAS_FN(::blas::_omatcopy2, queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -void omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, +void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { throw unimplemented("blas", "omatcopy2", ""); } -void imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, real_t alpha, - sycl::buffer &ab, std::int64_t lda, std::int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, real_t alpha, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { throw unimplemented("blas", "imatcopy", ""); } -void imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { throw unimplemented("blas", "imatcopy", ""); } -void omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - real_t alpha, sycl::buffer &a, std::int64_t lda, real_t beta, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &c, +void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + real_t alpha, sycl::buffer& a, std::int64_t lda, real_t beta, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& c, std::int64_t ldc) { CALL_PORTBLAS_FN(::blas::_omatadd, queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -void omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { +void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::complex beta, sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { throw unimplemented("blas", "omatadd", ""); } // USM APIs -sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, real_t alpha, const real_t *a, - std::int64_t lda, const real_t *b, std::int64_t ldb, real_t beta, real_t *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, real_t alpha, const real_t* a, + std::int64_t lda, const real_t* b, std::int64_t ldb, real_t beta, real_t* c, + std::int64_t ldc, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_gemm, queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { if (transa == oneapi::math::transpose::conjtrans || transb == oneapi::math::transpose::conjtrans) { throw unimplemented("blas", "gemm", "Conjugate Transpose unsupported yet on portBLAS"); @@ -249,203 +249,204 @@ sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::mat c, ldc, dependencies); } -sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - std::int64_t m, std::int64_t n, real_t alpha, const real_t *a, std::int64_t lda, - const real_t *b, std::int64_t ldb, real_t beta, real_t *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event symm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + std::int64_t m, std::int64_t n, real_t alpha, const real_t* a, std::int64_t lda, + const real_t* b, std::int64_t ldb, real_t beta, real_t* c, std::int64_t ldc, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_symm, queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +sycl::event symm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { throw unimplemented("blas", "symm", " for USM"); } -sycl::event hemm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +sycl::event hemm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { throw unimplemented("blas", "hemm", " for USM"); } -sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, real_t alpha, const real_t *a, std::int64_t lda, - real_t beta, real_t *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, real_t alpha, const real_t* a, std::int64_t lda, + real_t beta, real_t* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "syrk", " for USM"); } -sycl::event syrk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +sycl::event syrk(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "syrk", " for USM"); } -sycl::event herk(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, real_t alpha, const std::complex *a, - std::int64_t lda, real_t beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event herk(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, real_t alpha, const std::complex* a, + std::int64_t lda, real_t beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "herk", " for USM"); } -sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, real_t alpha, const real_t *a, std::int64_t lda, - const real_t *b, std::int64_t ldb, real_t beta, real_t *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, real_t alpha, const real_t* a, std::int64_t lda, + const real_t* b, std::int64_t ldb, real_t beta, real_t* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "syr2k", " for USM"); } -sycl::event syr2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +sycl::event syr2k(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { throw unimplemented("blas", "syr2k", " for USM"); } -sycl::event her2k(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, +sycl::event her2k(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, real_t beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, real_t beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "her2k", " for USM"); } -sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +sycl::event trmm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, real_t alpha, const real_t *a, std::int64_t lda, real_t *b, - std::int64_t ldb, const std::vector &dependencies) { + std::int64_t n, real_t alpha, const real_t* a, std::int64_t lda, real_t* b, + std::int64_t ldb, const std::vector& dependencies) { throw unimplemented("blas", "trmm", " for USM"); } -sycl::event trmm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +sycl::event trmm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "trmm", " for USM"); } -sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +sycl::event trsm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, real_t alpha, const real_t *a, std::int64_t lda, real_t *b, - std::int64_t ldb, const std::vector &dependencies) { + std::int64_t n, real_t alpha, const real_t* a, std::int64_t lda, real_t* b, + std::int64_t ldb, const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_trsm, queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm(sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, +sycl::event trsm(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "trsm", " for USM"); } -sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t n, std::int64_t k, real_t alpha, - const real_t *a, std::int64_t lda, const real_t *b, std::int64_t ldb, real_t beta, - real_t *c, std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, + std::int64_t k, real_t alpha, const real_t* a, std::int64_t lda, const real_t* b, + std::int64_t ldb, real_t beta, real_t* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", " for USM"); } -sycl::event gemmt(sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, oneapi::math::uplo upper_lower, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", " for USM"); } -sycl::event omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - real_t alpha, const real_t *a, std::int64_t lda, real_t *b, std::int64_t ldb, - const std::vector &dependencies) { +sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + real_t alpha, const real_t* a, std::int64_t lda, real_t* b, std::int64_t ldb, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_omatcopy, queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { +sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy", "for USM"); } -sycl::event omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - real_t alpha, const real_t *a, std::int64_t lda, std::int64_t stridea, - real_t *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { +sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + real_t alpha, const real_t* a, std::int64_t lda, std::int64_t stridea, + real_t* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_omatcopy2, queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, std::int64_t ldb, - std::int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stridea, std::complex* b, std::int64_t ldb, + std::int64_t strideb, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy2", "for USM"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - real_t alpha, real_t *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + real_t alpha, real_t* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", ""); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, std::complex *ab, std::int64_t lda, - std::int64_t ldb, const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, std::complex* ab, std::int64_t lda, + std::int64_t ldb, const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", ""); } -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, real_t alpha, const real_t *a, std::int64_t lda, real_t beta, - const real_t *b, std::int64_t ldb, real_t *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, real_t alpha, const real_t* a, std::int64_t lda, real_t beta, + const real_t* b, std::int64_t ldb, real_t* c, std::int64_t ldc, + const std::vector& dependencies) { CALL_PORTBLAS_USM_FN(::blas::_omatadd, queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, const std::complex *b, - std::int64_t ldb, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, const std::complex* b, + std::int64_t ldb, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "omatadd", ""); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - real_t *alpha, const real_t **a, int64_t *lda, real_t **b, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + real_t* alpha, const real_t** a, int64_t* lda, real_t** b, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", ""); } -sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, const std::complex **a, - int64_t *lda, std::complex **b, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, const std::complex** a, + int64_t* lda, std::complex** b, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "omatcopy_batch", ""); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - real_t *alpha, real_t **ab, int64_t *lda, int64_t *ldb, - int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + real_t* alpha, real_t** ab, int64_t* lda, int64_t* ldb, + int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", ""); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, std::complex **ab, int64_t *lda, - int64_t *ldb, int64_t group_count, int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, std::complex** ab, int64_t* lda, + int64_t* ldb, int64_t group_count, int64_t* groupsize, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", ""); } diff --git a/src/blas/backends/portblas/portblas_level3_bfloat16.cpp b/src/blas/backends/portblas/portblas_level3_bfloat16.cpp index d432647d6..7721f58c9 100644 --- a/src/blas/backends/portblas/portblas_level3_bfloat16.cpp +++ b/src/blas/backends/portblas/portblas_level3_bfloat16.cpp @@ -33,20 +33,20 @@ namespace portblas { namespace column_major { // BUFFER -void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { throw unimplemented("blas", "gemm", " for bfloat16"); } // USM -sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const oneapi::math::bfloat16 *a, std::int64_t lda, const oneapi::math::bfloat16 *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + const oneapi::math::bfloat16* a, std::int64_t lda, const oneapi::math::bfloat16* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemm", " for USM"); } @@ -54,20 +54,20 @@ sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::mat namespace row_major { // BUFFER -void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { throw unimplemented("blas", "gemm", " for bfloat16"); } // USM -sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const oneapi::math::bfloat16 *a, std::int64_t lda, const oneapi::math::bfloat16 *b, - std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { + const oneapi::math::bfloat16* a, std::int64_t lda, const oneapi::math::bfloat16* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemm", " for USM"); } diff --git a/src/blas/backends/portblas/portblas_level3_half.cpp b/src/blas/backends/portblas/portblas_level3_half.cpp index 65570e5d9..dbd71ab4b 100644 --- a/src/blas/backends/portblas/portblas_level3_half.cpp +++ b/src/blas/backends/portblas/portblas_level3_half.cpp @@ -33,66 +33,66 @@ namespace portblas { namespace column_major { // BUFFER -void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::half beta, sycl::buffer& c, std::int64_t ldc) { throw unimplemented("blas", "gemm", " half"); } -void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { throw unimplemented("blas", "gemm", " for different argument data types"); } // USM -sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - sycl::half beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies) { + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemm", " for USM"); } -sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { throw unimplemented("blas", "gemm", " for USM"); } } // namespace column_major namespace row_major { // BUFFER -void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::half beta, sycl::buffer& c, std::int64_t ldc) { throw unimplemented("blas", "gemm", " half"); } -void gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +void gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { throw unimplemented("blas", "gemm", " for different argument data types"); } // USM -sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, +sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - sycl::half beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies) { + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemm", " for USM"); } -sycl::event gemm(sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { throw unimplemented("blas", "gemm", " for USM"); } diff --git a/src/blas/backends/rocblas/rocblas_batch.cpp b/src/blas/backends/rocblas/rocblas_batch.cpp index 64338b2a3..b6e550724 100644 --- a/src/blas/backends/rocblas/rocblas_batch.cpp +++ b/src/blas/backends/rocblas/rocblas_batch.cpp @@ -28,7 +28,7 @@ // Helper Functions template -static inline void conj_vector(sycl::handler &cgh, sycl::buffer &buf, const int64_t len, +static inline void conj_vector(sycl::handler& cgh, sycl::buffer& buf, const int64_t len, const int64_t inc, const int64_t stride, const int64_t batch_size) { const auto abs_inc = std::abs(inc); const auto abs_stride = std::abs(stride); @@ -40,7 +40,7 @@ static inline void conj_vector(sycl::handler &cgh, sycl::buffer &buf, const i }); } template -static inline void conj_vector(sycl::handler &cgh, T *ptr, const int64_t len, const int64_t inc, +static inline void conj_vector(sycl::handler& cgh, T* ptr, const int64_t len, const int64_t inc, const int64_t stride, const int64_t batch_size) { const auto abs_inc = std::abs(inc); const auto abs_stride = std::abs(stride); @@ -52,7 +52,7 @@ static inline void conj_vector(sycl::handler &cgh, T *ptr, const int64_t len, co } template -static inline void conj_vector(sycl::handler &cgh, T **ptr, const int64_t len, const int64_t inc, +static inline void conj_vector(sycl::handler& cgh, T** ptr, const int64_t len, const int64_t inc, const int64_t stride, const int64_t group_size) { const auto abs_inc = std::abs(inc); cgh.parallel_for(sycl::range{ (std::size_t)group_size, (std::size_t)len }, @@ -72,30 +72,30 @@ namespace column_major { // Buffer APIs template -inline void copy_batch(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, +inline void copy_batch(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy, stridex, stridey, batch_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, n, x_, incx, stridex, y_, incy, stridey, - batch_size); + batch_size); }); }); } #define COPY_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, \ + void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, \ int64_t batch_size) { \ copy_batch(ROCBLAS_ROUTINE, queue, n, x, incx, stridex, y, incy, stridey, batch_size); \ } @@ -108,30 +108,30 @@ COPY_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zcopy_strided_batched) #undef COPY_STRIDED_BATCH_LAUNCHER template -inline void axpy_batch(Func func, sycl::queue &queue, int64_t n, T alpha, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, +inline void axpy_batch(Func func, sycl::queue& queue, int64_t n, T alpha, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy, stridex, stridey, batch_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; - rocblas_native_func(func, err, handle, n, (rocDataType *)&alpha, x_, incx, stridex, - y_, incy, stridey, batch_size); + rocblas_native_func(func, err, handle, n, (rocDataType*)&alpha, x_, incx, stridex, y_, + incy, stridey, batch_size); }); }); } #define AXPY_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void axpy_batch(sycl::queue &queue, int64_t n, TYPE alpha, sycl::buffer &x, \ - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, \ + void axpy_batch(sycl::queue& queue, int64_t n, TYPE alpha, sycl::buffer& x, \ + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, \ int64_t stridey, int64_t batch_size) { \ axpy_batch(ROCBLAS_ROUTINE, queue, n, alpha, x, incx, stridex, y, incy, stridey, \ batch_size); \ @@ -145,36 +145,36 @@ AXPY_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zaxpy_strided_batched) #undef AXPY_BATCH_LAUNCHER template -inline void gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - T alpha, sycl::buffer &a, int64_t lda, int64_t stridea, - sycl::buffer &x, int64_t incx, int64_t stridex, T beta, - sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { +inline void gemv_batch(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + T alpha, sycl::buffer& a, int64_t lda, int64_t stridea, + sycl::buffer& x, int64_t incx, int64_t stridex, T beta, + sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, incx, incy, stridea, stridex, stridey, batch_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(trans), m, n, - (rocDataType *)&alpha, a_, lda, stridea, x_, incx, stridex, - (rocDataType *)&beta, y_, incy, stridey, batch_size); + (rocDataType*)&alpha, a_, lda, stridea, x_, incx, stridex, + (rocDataType*)&beta, y_, incy, stridey, batch_size); }); }); } #define GEMV_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, int64_t stridea, \ - sycl::buffer &x, int64_t incx, int64_t stridex, TYPE beta, \ - sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { \ + void gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, int64_t stridea, \ + sycl::buffer& x, int64_t incx, int64_t stridex, TYPE beta, \ + sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { \ gemv_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, \ beta, y, incy, stridey, batch_size); \ } @@ -187,35 +187,35 @@ GEMV_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zgemv_strided_batched) #undef GEMV_STRIDED_BATCH_LAUNCHER template -inline void dgmm_batch(Func func, sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &c, int64_t ldc, +inline void dgmm_batch(Func func, sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stridea, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldc, incx, stridea, stridex, stridec, batch_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; - rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), m, n, a_, - lda, stridea, x_, incx, stridex, c_, ldc, stridec, batch_size); + rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), m, n, a_, lda, + stridea, x_, incx, stridex, c_, ldc, stridec, batch_size); }); }); } #define DGMM_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, \ - sycl::buffer &a, int64_t lda, int64_t stridea, \ - sycl::buffer &x, int64_t incx, int64_t stridex, \ - sycl::buffer &c, int64_t ldc, int64_t stridec, int64_t batch_size) { \ + void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, \ + sycl::buffer& a, int64_t lda, int64_t stridea, \ + sycl::buffer& x, int64_t incx, int64_t stridex, \ + sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { \ dgmm_batch(ROCBLAS_ROUTINE, queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, \ ldc, stridec, batch_size); \ } @@ -228,10 +228,10 @@ DGMM_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zdgmm_strided_batched) #undef DGMM_STRIDED_BATCH_LAUNCHER template -inline void gemm_batch_impl(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, Ts alpha, sycl::buffer &a, int64_t lda, - int64_t stridea, sycl::buffer &b, int64_t ldb, int64_t strideb, - Ts beta, sycl::buffer &c, int64_t ldc, int64_t stridec, +inline void gemm_batch_impl(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, int64_t k, Ts alpha, sycl::buffer& a, int64_t lda, + int64_t stridea, sycl::buffer& b, int64_t ldb, int64_t strideb, + Ts beta, sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { using rocTypeA = typename RocEquivalentType::Type; using rocTypeB = typename RocEquivalentType::Type; @@ -241,35 +241,35 @@ inline void gemm_batch_impl(sycl::queue &queue, transpose transa, transpose tran int32_t solution_index = 0; rocblas_gemm_flags flags = rocblas_gemm_flags_none; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocblas_native_func(rocblas_gemm_strided_batched_ex, err, handle, - get_rocblas_operation(transa), get_rocblas_operation(transb), m, - n, k, &alpha, a_, get_rocblas_datatype(), lda, - stridea, b_, get_rocblas_datatype(), ldb, strideb, - &beta, c_, get_rocblas_datatype(), ldc, stridec, c_, - get_rocblas_datatype(), ldc, stridec, batch_size, - get_rocblas_datatype(), rocblas_gemm_algo_standard, - solution_index, flags); + get_rocblas_operation(transa), get_rocblas_operation(transb), m, n, + k, &alpha, a_, get_rocblas_datatype(), lda, stridea, b_, + get_rocblas_datatype(), ldb, strideb, &beta, c_, + get_rocblas_datatype(), ldc, stridec, c_, + get_rocblas_datatype(), ldc, stridec, batch_size, + get_rocblas_datatype(), rocblas_gemm_algo_standard, + solution_index, flags); }); }); } #define GEMM_STRIDED_BATCH_LAUNCHER(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_S alpha, sycl::buffer &a, int64_t lda, \ - int64_t stridea, sycl::buffer &b, int64_t ldb, int64_t strideb, \ - TYPE_S beta, sycl::buffer &c, int64_t ldc, int64_t stridec, \ + void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_S alpha, sycl::buffer& a, int64_t lda, \ + int64_t stridea, sycl::buffer& b, int64_t ldb, int64_t strideb, \ + TYPE_S beta, sycl::buffer& c, int64_t ldc, int64_t stridec, \ int64_t batch_size) { \ gemm_batch_impl(queue, transa, transb, m, n, k, alpha, a, lda, stridea, b, ldb, strideb, \ beta, c, ldc, stridec, batch_size); \ @@ -287,10 +287,10 @@ GEMM_STRIDED_BATCH_LAUNCHER(sycl::half, sycl::half, float, float) #undef GEMM_STRIDED_BATCH_LAUNCHER #define GEMM_STRIDED_BATCH_LAUNCHER(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_S alpha, sycl::buffer &a, int64_t lda, \ - int64_t stridea, sycl::buffer &b, int64_t ldb, int64_t strideb, \ - TYPE_S beta, sycl::buffer &c, int64_t ldc, int64_t stridec, \ + void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_S alpha, sycl::buffer& a, int64_t lda, \ + int64_t stridea, sycl::buffer& b, int64_t ldb, int64_t strideb, \ + TYPE_S beta, sycl::buffer& c, int64_t ldc, int64_t stridec, \ int64_t batch_size) { \ throw unimplemented("blas", "gemm_batch", \ std::string("for dtype unimplemented dtype combination <") + \ @@ -304,35 +304,34 @@ GEMM_STRIDED_BATCH_LAUNCHER(std::int8_t, std::int8_t, std::int32_t, float) #undef GEMM_STRIDED_BATCH_LAUNCHER template -inline void trsm_batch(Func func, sycl::queue &queue, side left_right, uplo upper_lower, +inline void trsm_batch(Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &b, + sycl::buffer& a, int64_t lda, int64_t stridea, sycl::buffer& b, int64_t ldb, int64_t strideb, int64_t batch_size) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, stridea, strideb, batch_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), - get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - m, n, (rocDataType *)&alpha, a_, lda, stridea, b_, ldb, strideb, - batch_size); + get_rocblas_fill_mode(upper_lower), get_rocblas_operation(trans), + get_rocblas_diag_type(unit_diag), m, n, (rocDataType*)&alpha, a_, + lda, stridea, b_, ldb, strideb, batch_size); }); }); } #define TRSM_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer &a, \ - int64_t lda, int64_t stridea, sycl::buffer &b, int64_t ldb, \ + void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer& a, \ + int64_t lda, int64_t stridea, sycl::buffer& b, int64_t ldb, \ int64_t strideb, int64_t batch_size) { \ trsm_batch(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, \ a, lda, stridea, b, ldb, strideb, batch_size); \ @@ -346,34 +345,33 @@ TRSM_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_ztrsm_strided_batched) #undef TRSM_STRIDED_BATCH_LAUNCHER template -inline void syrk_batch(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, T alpha, sycl::buffer &a, int64_t lda, int64_t stridea, - T beta, sycl::buffer &c, int64_t ldc, int64_t stridec, +inline void syrk_batch(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, T alpha, sycl::buffer& a, int64_t lda, int64_t stridea, + T beta, sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, ldc, stridea, stridec, batch_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), n, k, (rocDataType *)&alpha, a_, - lda, stridea, (rocDataType *)&beta, c_, ldc, stridec, - batch_size); + get_rocblas_operation(trans), n, k, (rocDataType*)&alpha, a_, lda, + stridea, (rocDataType*)&beta, c_, ldc, stridec, batch_size); }); }); } #define SYRK_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, sycl::buffer &a, int64_t lda, int64_t stridea, TYPE beta, \ - sycl::buffer &c, int64_t ldc, int64_t stridec, int64_t batch_size) { \ + void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, sycl::buffer& a, int64_t lda, int64_t stridea, TYPE beta, \ + sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { \ syrk_batch(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, stridea, beta, \ c, ldc, stridec, batch_size); \ } @@ -386,9 +384,9 @@ SYRK_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zsyrk_strided_batched) #undef SYRK_STRIDED_BATCH_LAUNCHER template -inline void omatcopy_batch(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - const T alpha, sycl::buffer &a, int64_t lda, int64_t stridea, - sycl::buffer &b, int64_t ldb, int64_t strideb, +inline void omatcopy_batch(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + const T alpha, sycl::buffer& a, int64_t lda, int64_t stridea, + sycl::buffer& b, int64_t ldb, int64_t strideb, int64_t batch_size) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, stridea, strideb, batch_size); @@ -397,27 +395,27 @@ inline void omatcopy_batch(Func func, sycl::queue &queue, transpose trans, int64 const int64_t new_m = trans == oneapi::math::transpose::nontrans ? m : n; const int64_t new_n = trans == oneapi::math::transpose::nontrans ? n : m; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(trans), - get_rocblas_operation(trans), new_m, new_n, - (rocDataType *)&alpha, a_, lda, stridea, (rocDataType *)&beta, - nullptr, lda, stridea, b_, ldb, strideb, batch_size); + get_rocblas_operation(trans), new_m, new_n, (rocDataType*)&alpha, + a_, lda, stridea, (rocDataType*)&beta, nullptr, lda, stridea, b_, + ldb, strideb, batch_size); }); }); } #define OMATCOPY_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, \ - const TYPE alpha, sycl::buffer &a, int64_t lda, int64_t stridea, \ - sycl::buffer &b, int64_t ldb, int64_t strideb, \ + void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, \ + const TYPE alpha, sycl::buffer& a, int64_t lda, int64_t stridea, \ + sycl::buffer& b, int64_t ldb, int64_t strideb, \ int64_t batch_size) { \ omatcopy_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, stridea, b, ldb, \ strideb, batch_size); \ @@ -430,63 +428,63 @@ OMATCOPY_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zgeam_strided_batc #undef OMATCOPY_STRIDED_BATCH_LAUNCHER -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } template -inline void omatadd_batch(Func func, sycl::queue &queue, transpose transa, transpose transb, - int64_t m, int64_t n, const T alpha, sycl::buffer &a, int64_t lda, - int64_t stridea, const T beta, sycl::buffer &b, int64_t ldb, - int64_t strideb, sycl::buffer &c, int64_t ldc, int64_t stridec, +inline void omatadd_batch(Func func, sycl::queue& queue, transpose transa, transpose transb, + int64_t m, int64_t n, const T alpha, sycl::buffer& a, int64_t lda, + int64_t stridea, const T beta, sycl::buffer& b, int64_t ldb, + int64_t strideb, sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc, stridea, strideb, stridec, batch_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(transa), - get_rocblas_operation(transb), m, n, (rocDataType *)&alpha, a_, - lda, stridea, (rocDataType *)&beta, b_, ldb, strideb, c_, ldc, - stridec, batch_size); + get_rocblas_operation(transb), m, n, (rocDataType*)&alpha, a_, lda, + stridea, (rocDataType*)&beta, b_, ldb, strideb, c_, ldc, stridec, + batch_size); }); }); } #define OMATADD_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, const TYPE alpha, sycl::buffer &a, int64_t lda, \ - int64_t stridea, const TYPE beta, sycl::buffer &b, int64_t ldb, \ - int64_t strideb, sycl::buffer &c, int64_t ldc, int64_t stridec, \ + void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, const TYPE alpha, sycl::buffer& a, int64_t lda, \ + int64_t stridea, const TYPE beta, sycl::buffer& b, int64_t ldb, \ + int64_t strideb, sycl::buffer& c, int64_t ldc, int64_t stridec, \ int64_t batch_size) { \ omatadd_batch(ROCBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, lda, stridea, beta, \ b, ldb, strideb, c, ldc, stridec, batch_size); \ @@ -502,26 +500,26 @@ OMATADD_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zgeam_strided_batch // USM APIs template -inline sycl::event copy_batch(Func func, sycl::queue &queue, int64_t *n, const T **x, int64_t *incx, - T **y, int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event copy_batch(Func func, sycl::queue& queue, int64_t* n, const T** x, int64_t* incx, + T** y, int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; for (int64_t i = 0; i < group_count; i++) { overflow_check(n[i], incx[i], incy[i], group_size[i]); } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; for (int64_t i = 0; i < group_count; i++) { - auto **x_ = reinterpret_cast(x); - auto **y_ = reinterpret_cast(y); + auto** x_ = reinterpret_cast(x); + auto** y_ = reinterpret_cast(y); rocblas_native_func(func, err, handle, (int)n[i], x_ + offset, (int)incx[i], - y_ + offset, (int)incy[i], (int)group_size[i]); + y_ + offset, (int)incy[i], (int)group_size[i]); offset += group_size[i]; } }); @@ -531,9 +529,9 @@ inline sycl::event copy_batch(Func func, sycl::queue &queue, int64_t *n, const T } #define COPY_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event copy_batch(sycl::queue &queue, int64_t *n, const TYPE **x, int64_t *incx, \ - TYPE **y, int64_t *incy, int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event copy_batch(sycl::queue& queue, int64_t* n, const TYPE** x, int64_t* incx, \ + TYPE** y, int64_t* incy, int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return copy_batch(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, group_count, group_size, \ dependencies); \ } @@ -546,22 +544,22 @@ COPY_BATCH_LAUNCHER_USM(std::complex, rocblas_zcopy_batched) #undef COPY_BATCH_LAUNCHER_USM template -inline sycl::event copy_batch(Func func, sycl::queue &queue, int64_t n, const T *x, int64_t incx, - int64_t stridex, T *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +inline sycl::event copy_batch(Func func, sycl::queue& queue, int64_t n, const T* x, int64_t incx, + int64_t stridex, T* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy, stridex, stridey, batch_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, n, x_, incx, stridex, y_, incy, stridey, - batch_size); + batch_size); }); }); @@ -569,9 +567,9 @@ inline sycl::event copy_batch(Func func, sycl::queue &queue, int64_t n, const T } #define COPY_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event copy_batch(sycl::queue &queue, int64_t n, const TYPE *x, int64_t incx, \ - int64_t stridex, TYPE *y, int64_t incy, int64_t stridey, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event copy_batch(sycl::queue& queue, int64_t n, const TYPE* x, int64_t incx, \ + int64_t stridex, TYPE* y, int64_t incy, int64_t stridey, \ + int64_t batch_size, const std::vector& dependencies) { \ return copy_batch(ROCBLAS_ROUTINE, queue, n, x, incx, stridex, y, incy, stridey, \ batch_size, dependencies); \ } @@ -584,27 +582,27 @@ COPY_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zcopy_strided_batc #undef COPY_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event axpy_batch(Func func, sycl::queue &queue, int64_t *n, T *alpha, const T **x, - int64_t *incx, T **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +inline sycl::event axpy_batch(Func func, sycl::queue& queue, int64_t* n, T* alpha, const T** x, + int64_t* incx, T** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; for (int64_t i = 0; i < group_count; i++) { overflow_check(n[i], incx[i], incy[i], group_size[i]); } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; for (int64_t i = 0; i < group_count; i++) { - auto **x_ = reinterpret_cast(x); - auto **y_ = reinterpret_cast(y); - rocblas_native_func(func, err, handle, (int)n[i], (rocDataType *)&alpha[i], - x_ + offset, (int)incx[i], y_ + offset, (int)incy[i], - (int)group_size[i]); + auto** x_ = reinterpret_cast(x); + auto** y_ = reinterpret_cast(y); + rocblas_native_func(func, err, handle, (int)n[i], (rocDataType*)&alpha[i], + x_ + offset, (int)incx[i], y_ + offset, (int)incy[i], + (int)group_size[i]); offset += group_size[i]; } }); @@ -614,9 +612,9 @@ inline sycl::event axpy_batch(Func func, sycl::queue &queue, int64_t *n, T *alph } #define AXPY_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event axpy_batch(sycl::queue &queue, int64_t *n, TYPE *alpha, const TYPE **x, \ - int64_t *incx, TYPE **y, int64_t *incy, int64_t group_count, \ - int64_t *group_size, const std::vector &dependencies) { \ + sycl::event axpy_batch(sycl::queue& queue, int64_t* n, TYPE* alpha, const TYPE** x, \ + int64_t* incx, TYPE** y, int64_t* incy, int64_t group_count, \ + int64_t* group_size, const std::vector& dependencies) { \ return axpy_batch(ROCBLAS_ROUTINE, queue, n, alpha, x, incx, y, incy, group_count, \ group_size, dependencies); \ } @@ -629,22 +627,22 @@ AXPY_BATCH_LAUNCHER_USM(std::complex, rocblas_zaxpy_batched) #undef AXPY_BATCH_LAUNCHER_USM template -inline sycl::event axpy_batch(Func func, sycl::queue &queue, int64_t n, T alpha, const T *x, - int64_t incx, int64_t stridex, T *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +inline sycl::event axpy_batch(Func func, sycl::queue& queue, int64_t n, T alpha, const T* x, + int64_t incx, int64_t stridex, T* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy, stridex, stridey, batch_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; - rocblas_native_func(func, err, handle, n, (rocDataType *)&alpha, x_, incx, stridex, - y_, incy, stridey, batch_size); + rocblas_native_func(func, err, handle, n, (rocDataType*)&alpha, x_, incx, stridex, y_, + incy, stridey, batch_size); }); }); @@ -652,9 +650,9 @@ inline sycl::event axpy_batch(Func func, sycl::queue &queue, int64_t n, T alpha, } #define AXPY_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event axpy_batch(sycl::queue &queue, int64_t n, TYPE alpha, const TYPE *x, int64_t incx, \ - int64_t stridex, TYPE *y, int64_t incy, int64_t stridey, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event axpy_batch(sycl::queue& queue, int64_t n, TYPE alpha, const TYPE* x, int64_t incx, \ + int64_t stridex, TYPE* y, int64_t incy, int64_t stridey, \ + int64_t batch_size, const std::vector& dependencies) { \ return axpy_batch(ROCBLAS_ROUTINE, queue, n, alpha, x, incx, stridex, y, incy, stridey, \ batch_size, dependencies); \ } @@ -667,26 +665,26 @@ AXPY_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zaxpy_strided_batc #undef AXPY_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - T alpha, const T *a, int64_t lda, int64_t stridea, const T *x, - int64_t incx, int64_t stridex, T beta, T *y, int64_t incy, +inline sycl::event gemv_batch(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + T alpha, const T* a, int64_t lda, int64_t stridea, const T* x, + int64_t incx, int64_t stridex, T beta, T* y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, m, lda, incx, incy, stridea, stridex, stridey, batch_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(trans), m, n, - (rocDataType *)&alpha, a_, lda, stridea, x_, incx, stridex, - (rocDataType *)&beta, y_, incy, stridey, batch_size); + (rocDataType*)&alpha, a_, lda, stridea, x_, incx, stridex, + (rocDataType*)&beta, y_, incy, stridey, batch_size); }); }); @@ -694,11 +692,11 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose trans, in } #define GEMV_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, int64_t stridea, const TYPE *x, \ - int64_t incx, int64_t stridex, TYPE beta, TYPE *y, int64_t incy, \ + sycl::event gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, int64_t stridea, const TYPE* x, \ + int64_t incx, int64_t stridex, TYPE beta, TYPE* y, int64_t incy, \ int64_t stridey, int64_t batch_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return gemv_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, stridea, x, incx, \ stridex, beta, y, incy, stridey, batch_size, dependencies); \ } @@ -711,30 +709,30 @@ GEMV_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zgemv_strided_batc #undef GEMV_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, int64_t *m, - int64_t *n, T *alpha, const T **a, int64_t *lda, const T **x, - int64_t *incx, T *beta, T **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +inline sycl::event gemv_batch(Func func, sycl::queue& queue, transpose* trans, int64_t* m, + int64_t* n, T* alpha, const T** a, int64_t* lda, const T** x, + int64_t* incx, T* beta, T** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; for (int64_t i = 0; i < group_count; i++) { overflow_check(m[i], n[i], lda[i], incx[i], incy[i], group_size[i]); } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; for (int64_t i = 0; i < group_count; i++) { - auto **a_ = reinterpret_cast(a); - auto **x_ = reinterpret_cast(x); - auto **y_ = reinterpret_cast(y); - rocblas_native_func( - func, err, handle, get_rocblas_operation(trans[i]), (int)m[i], (int)n[i], - (rocDataType *)&alpha[i], a_ + offset, (int)lda[i], x_ + offset, (int)incx[i], - (rocDataType *)&beta[i], y_ + offset, (int)incy[i], (int)group_size[i]); + auto** a_ = reinterpret_cast(a); + auto** x_ = reinterpret_cast(x); + auto** y_ = reinterpret_cast(y); + rocblas_native_func(func, err, handle, get_rocblas_operation(trans[i]), (int)m[i], + (int)n[i], (rocDataType*)&alpha[i], a_ + offset, (int)lda[i], + x_ + offset, (int)incx[i], (rocDataType*)&beta[i], y_ + offset, + (int)incy[i], (int)group_size[i]); offset += group_size[i]; } }); @@ -745,9 +743,9 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, i #define GEMV_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ sycl::event gemv_batch( \ - sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, TYPE *alpha, const TYPE **a, \ - int64_t *lda, const TYPE **x, int64_t *incx, TYPE *beta, TYPE **y, int64_t *incy, \ - int64_t group_count, int64_t *group_size, const std::vector &dependencies) { \ + sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, TYPE* alpha, const TYPE** a, \ + int64_t* lda, const TYPE** x, int64_t* incx, TYPE* beta, TYPE** y, int64_t* incy, \ + int64_t group_count, int64_t* group_size, const std::vector& dependencies) { \ return gemv_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, x, incx, beta, y, \ incy, group_count, group_size, dependencies); \ } @@ -760,24 +758,24 @@ GEMV_BATCH_LAUNCHER_USM(std::complex, rocblas_zgemv_batched) #undef GEMV_BATCH_LAUNCHER_USM template -inline sycl::event dgmm_batch(Func func, sycl::queue &queue, side left_right, int64_t m, int64_t n, - const T *a, int64_t lda, int64_t stridea, const T *x, int64_t incx, - int64_t stridex, T *c, int64_t ldc, int64_t stridec, - int64_t batch_size, const std::vector &dependencies) { +inline sycl::event dgmm_batch(Func func, sycl::queue& queue, side left_right, int64_t m, int64_t n, + const T* a, int64_t lda, int64_t stridea, const T* x, int64_t incx, + int64_t stridex, T* c, int64_t ldc, int64_t stridec, + int64_t batch_size, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, incx, stridea, stridex, stridec, batch_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto c_ = reinterpret_cast(c); rocblas_status err; - rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), m, n, a_, - lda, stridea, x_, incx, stridex, c_, ldc, stridec, batch_size); + rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), m, n, a_, lda, + stridea, x_, incx, stridex, c_, ldc, stridec, batch_size); }); }); @@ -785,10 +783,10 @@ inline sycl::event dgmm_batch(Func func, sycl::queue &queue, side left_right, in } #define DGMM_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, \ - const TYPE *a, int64_t lda, int64_t stridea, const TYPE *x, \ - int64_t incx, int64_t stridex, TYPE *c, int64_t ldc, int64_t stridec, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, \ + const TYPE* a, int64_t lda, int64_t stridea, const TYPE* x, \ + int64_t incx, int64_t stridex, TYPE* c, int64_t ldc, int64_t stridec, \ + int64_t batch_size, const std::vector& dependencies) { \ return dgmm_batch(ROCBLAS_ROUTINE, queue, left_right, m, n, a, lda, stridea, x, incx, \ stridex, c, ldc, stridec, batch_size, dependencies); \ } @@ -801,29 +799,29 @@ DGMM_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zdgmm_strided_batc #undef DGMM_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event dgmm_batch(Func func, sycl::queue &queue, side *left_right, int64_t *m, - int64_t *n, const T **a, int64_t *lda, const T **x, int64_t *incx, - T **c, int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event dgmm_batch(Func func, sycl::queue& queue, side* left_right, int64_t* m, + int64_t* n, const T** a, int64_t* lda, const T** x, int64_t* incx, + T** c, int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; for (int64_t i = 0; i < group_count; i++) { overflow_check(m[i], n[i], lda[i], ldc[i], incx[i], group_size[i]); } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; for (int64_t i = 0; i < group_count; i++) { - auto **a_ = reinterpret_cast(a); - auto **x_ = reinterpret_cast(x); - auto **c_ = reinterpret_cast(c); + auto** a_ = reinterpret_cast(a); + auto** x_ = reinterpret_cast(x); + auto** c_ = reinterpret_cast(c); rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right[i]), - (int)m[i], (int)n[i], a_ + offset, (int)lda[i], x_ + offset, - (int)incx[i], c_ + offset, (int)ldc[i], (int)group_size[i]); + (int)m[i], (int)n[i], a_ + offset, (int)lda[i], x_ + offset, + (int)incx[i], c_ + offset, (int)ldc[i], (int)group_size[i]); offset += group_size[i]; } }); @@ -833,10 +831,10 @@ inline sycl::event dgmm_batch(Func func, sycl::queue &queue, side *left_right, i } #define DGMM_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, \ - const TYPE **a, int64_t *lda, const TYPE **x, int64_t *incx, TYPE **c, \ - int64_t *ldc, int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, \ + const TYPE** a, int64_t* lda, const TYPE** x, int64_t* incx, TYPE** c, \ + int64_t* ldc, int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return dgmm_batch(ROCBLAS_ROUTINE, queue, left_right, m, n, a, lda, x, incx, c, ldc, \ group_count, group_size, dependencies); \ } @@ -849,13 +847,13 @@ DGMM_BATCH_LAUNCHER_USM(std::complex, rocblas_zdgmm_batched) #undef DGMM_BATCH_LAUNCHER template -inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose transa, +inline sycl::event gemm_batch_strided_usm_impl(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - Ts alpha, const Ta *a, int64_t lda, int64_t stridea, - const Tb *b, int64_t ldb, int64_t strideb, Ts beta, - Tc *c, int64_t ldc, int64_t stridec, + Ts alpha, const Ta* a, int64_t lda, int64_t stridea, + const Tb* b, int64_t ldb, int64_t strideb, Ts beta, + Tc* c, int64_t ldc, int64_t stridec, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocTypeA = typename RocEquivalentType::Type; using rocTypeB = typename RocEquivalentType::Type; using rocTypeC = typename RocEquivalentType::Type; @@ -864,23 +862,23 @@ inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose tra int32_t solution_index = 0; rocblas_gemm_flags flags = rocblas_gemm_flags_none; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); rocblas_status err; rocblas_native_func(rocblas_gemm_strided_batched_ex, err, handle, - get_rocblas_operation(transa), get_rocblas_operation(transb), m, - n, k, &alpha, a_, get_rocblas_datatype(), lda, - stridea, b_, get_rocblas_datatype(), ldb, strideb, - &beta, c_, get_rocblas_datatype(), ldc, stridec, c_, - get_rocblas_datatype(), ldc, stridec, batch_size, - get_rocblas_datatype(), rocblas_gemm_algo_standard, - solution_index, flags); + get_rocblas_operation(transa), get_rocblas_operation(transb), m, n, + k, &alpha, a_, get_rocblas_datatype(), lda, stridea, b_, + get_rocblas_datatype(), ldb, strideb, &beta, c_, + get_rocblas_datatype(), ldc, stridec, c_, + get_rocblas_datatype(), ldc, stridec, batch_size, + get_rocblas_datatype(), rocblas_gemm_algo_standard, + solution_index, flags); }); }); @@ -888,11 +886,11 @@ inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose tra } #define GEMM_STRIDED_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, int64_t k, TYPE_S alpha, const TYPE_A *a, int64_t lda, \ - int64_t stridea, const TYPE_B *b, int64_t ldb, int64_t strideb, \ - TYPE_S beta, TYPE_C *c, int64_t ldc, int64_t stridec, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, int64_t k, TYPE_S alpha, const TYPE_A* a, int64_t lda, \ + int64_t stridea, const TYPE_B* b, int64_t ldb, int64_t strideb, \ + TYPE_S beta, TYPE_C* c, int64_t ldc, int64_t stridec, \ + int64_t batch_size, const std::vector& dependencies) { \ return gemm_batch_strided_usm_impl(queue, transa, transb, m, n, k, alpha, a, lda, stridea, \ b, ldb, strideb, beta, c, ldc, stridec, batch_size, \ dependencies); \ @@ -910,11 +908,11 @@ GEMM_STRIDED_BATCH_LAUNCHER_USM(sycl::half, sycl::half, float, float) #undef GEMM_STRIDED_BATCH_LAUNCHER_USM #define GEMM_STRIDED_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, int64_t k, TYPE_S alpha, const TYPE_A *a, int64_t lda, \ - int64_t stridea, const TYPE_B *b, int64_t ldb, int64_t strideb, \ - TYPE_S beta, TYPE_C *c, int64_t ldc, int64_t stridec, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, int64_t k, TYPE_S alpha, const TYPE_A* a, int64_t lda, \ + int64_t stridea, const TYPE_B* b, int64_t ldb, int64_t strideb, \ + TYPE_S beta, TYPE_C* c, int64_t ldc, int64_t stridec, \ + int64_t batch_size, const std::vector& dependencies) { \ throw unimplemented("blas", "gemm_batch", \ std::string("for dtype unimplemented dtype combination <") + \ dtype_string() + "," + dtype_string() + "," + \ @@ -927,11 +925,11 @@ GEMM_STRIDED_BATCH_LAUNCHER_USM(std::int8_t, std::int8_t, std::int32_t, float) #undef GEMM_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, transpose *transb, - int64_t *m, int64_t *n, int64_t *k, Ts *alpha, const Ta **a, - int64_t *lda, const Tb **b, int64_t *ldb, Ts *beta, Tc **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event gemm_batch_usm_impl(sycl::queue& queue, transpose* transa, transpose* transb, + int64_t* m, int64_t* n, int64_t* k, Ts* alpha, const Ta** a, + int64_t* lda, const Tb** b, int64_t* ldb, Ts* beta, Tc** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { using rocTypeA = typename RocEquivalentType::Type; using rocTypeB = typename RocEquivalentType::Type; using rocTypeC = typename RocEquivalentType::Type; @@ -942,17 +940,17 @@ inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, tr int32_t solution_index = 0; rocblas_gemm_flags flags = rocblas_gemm_flags_none; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; for (int64_t i = 0; i < group_count; i++) { - auto **a_ = reinterpret_cast(a); - auto **b_ = reinterpret_cast(b); - auto **c_ = reinterpret_cast(c); + auto** a_ = reinterpret_cast(a); + auto** b_ = reinterpret_cast(b); + auto** c_ = reinterpret_cast(c); rocblas_native_func( rocblas_gemm_batched_ex, err, handle, get_rocblas_operation(transa[i]), get_rocblas_operation(transb[i]), (int)m[i], (int)n[i], (int)k[i], &alpha[i], @@ -971,11 +969,11 @@ inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, tr } #define GEMM_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, \ - int64_t *n, int64_t *k, TYPE_S *alpha, const TYPE_A **a, int64_t *lda, \ - const TYPE_B **b, int64_t *ldb, TYPE_S *beta, TYPE_C **c, int64_t *ldc, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, \ + int64_t* n, int64_t* k, TYPE_S* alpha, const TYPE_A** a, int64_t* lda, \ + const TYPE_B** b, int64_t* ldb, TYPE_S* beta, TYPE_C** c, int64_t* ldc, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return gemm_batch_usm_impl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, \ ldc, group_count, group_size, dependencies); \ } @@ -992,11 +990,11 @@ GEMM_BATCH_LAUNCHER_USM(sycl::half, sycl::half, float, float) #undef GEMM_BATCH_LAUNCHER_USM #define GEMM_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, \ - int64_t *n, int64_t *k, TYPE_S *alpha, const TYPE_A **a, int64_t *lda, \ - const TYPE_B **b, int64_t *ldb, TYPE_S *beta, TYPE_C **c, int64_t *ldc, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, \ + int64_t* n, int64_t* k, TYPE_S* alpha, const TYPE_A** a, int64_t* lda, \ + const TYPE_B** b, int64_t* ldb, TYPE_S* beta, TYPE_C** c, int64_t* ldc, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ throw unimplemented("blas", "gemm_batch", \ std::string("for dtype unimplemented dtype combination <") + \ dtype_string() + "," + dtype_string() + "," + \ @@ -1009,27 +1007,26 @@ GEMM_BATCH_LAUNCHER_USM(std::int8_t, std::int8_t, std::int32_t, float) #undef GEMM_BATCH_LAUNCHER_USM template -inline sycl::event trsm_batch(Func func, sycl::queue &queue, side left_right, uplo upper_lower, +inline sycl::event trsm_batch(Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, - const T *a, int64_t lda, int64_t stridea, T *b, int64_t ldb, + const T* a, int64_t lda, int64_t stridea, T* b, int64_t ldb, int64_t strideb, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, stridea, strideb, batch_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), - get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - m, n, (rocDataType *)&alpha, a_, lda, stridea, b_, ldb, strideb, - batch_size); + get_rocblas_fill_mode(upper_lower), get_rocblas_operation(trans), + get_rocblas_diag_type(unit_diag), m, n, (rocDataType*)&alpha, a_, + lda, stridea, b_, ldb, strideb, batch_size); }); }); @@ -1037,10 +1034,10 @@ inline sycl::event trsm_batch(Func func, sycl::queue &queue, side left_right, up } #define TRSM_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE *a, \ - int64_t lda, int64_t stridea, TYPE *b, int64_t ldb, int64_t strideb, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE* a, \ + int64_t lda, int64_t stridea, TYPE* b, int64_t ldb, int64_t strideb, \ + int64_t batch_size, const std::vector& dependencies) { \ return trsm_batch(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, \ alpha, a, lda, stridea, b, ldb, strideb, batch_size, dependencies); \ } @@ -1053,31 +1050,31 @@ TRSM_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_ztrsm_strided_batc #undef TRSM_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event trsm_batch(Func func, sycl::queue &queue, side *left_right, uplo *upper_lower, - transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, T *alpha, - const T **a, int64_t *lda, T **b, int64_t *ldb, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +inline sycl::event trsm_batch(Func func, sycl::queue& queue, side* left_right, uplo* upper_lower, + transpose* trans, diag* unit_diag, int64_t* m, int64_t* n, T* alpha, + const T** a, int64_t* lda, T** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; for (int64_t i = 0; i < group_count; i++) { overflow_check(m[i], n[i], lda[i], ldb[i], group_size[i]); } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; for (int64_t i = 0; i < group_count; i++) { - auto **a_ = reinterpret_cast(a); - auto **b_ = reinterpret_cast(b); + auto** a_ = reinterpret_cast(a); + auto** b_ = reinterpret_cast(b); rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right[i]), - get_rocblas_fill_mode(upper_lower[i]), - get_rocblas_operation(trans[i]), - get_rocblas_diag_type(unit_diag[i]), (int)m[i], (int)n[i], - (rocDataType *)&alpha[i], a_ + offset, (int)lda[i], - b_ + offset, (int)ldb[i], (int)group_size[i]); + get_rocblas_fill_mode(upper_lower[i]), + get_rocblas_operation(trans[i]), + get_rocblas_diag_type(unit_diag[i]), (int)m[i], (int)n[i], + (rocDataType*)&alpha[i], a_ + offset, (int)lda[i], b_ + offset, + (int)ldb[i], (int)group_size[i]); offset += group_size[i]; } }); @@ -1087,11 +1084,11 @@ inline sycl::event trsm_batch(Func func, sycl::queue &queue, side *left_right, u } #define TRSM_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, \ - transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, TYPE *alpha, \ - const TYPE **a, int64_t *lda, TYPE **b, int64_t *ldb, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, \ + transpose* trans, diag* unit_diag, int64_t* m, int64_t* n, TYPE* alpha, \ + const TYPE** a, int64_t* lda, TYPE** b, int64_t* ldb, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return trsm_batch(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, \ alpha, a, lda, b, ldb, group_count, group_size, dependencies); \ } @@ -1104,30 +1101,30 @@ TRSM_BATCH_LAUNCHER_USM(std::complex, rocblas_ztrsm_batched) #undef TRSM_BATCH_LAUNCHER_USM template -inline sycl::event syrk_batch(Func func, sycl::queue &queue, uplo *upper_lower, transpose *trans, - int64_t *n, int64_t *k, T *alpha, const T **a, int64_t *lda, T *beta, - T **c, int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event syrk_batch(Func func, sycl::queue& queue, uplo* upper_lower, transpose* trans, + int64_t* n, int64_t* k, T* alpha, const T** a, int64_t* lda, T* beta, + T** c, int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; for (int64_t i = 0; i < group_count; i++) { overflow_check(n[i], k[i], lda[i], ldc[i], group_size[i]); } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; for (int64_t i = 0; i < group_count; i++) { - auto **a_ = reinterpret_cast(a); - auto **c_ = reinterpret_cast(c); + auto** a_ = reinterpret_cast(a); + auto** c_ = reinterpret_cast(c); rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower[i]), - get_rocblas_operation(trans[i]), (int)n[i], (int)k[i], - (rocDataType *)&alpha[i], a_ + offset, (int)lda[i], - (rocDataType *)&beta[i], c_ + offset, (int)ldc[i], - (int)group_size[i]); + get_rocblas_operation(trans[i]), (int)n[i], (int)k[i], + (rocDataType*)&alpha[i], a_ + offset, (int)lda[i], + (rocDataType*)&beta[i], c_ + offset, (int)ldc[i], + (int)group_size[i]); offset += group_size[i]; } }); @@ -1137,10 +1134,10 @@ inline sycl::event syrk_batch(Func func, sycl::queue &queue, uplo *upper_lower, } #define SYRK_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, \ - int64_t *k, TYPE *alpha, const TYPE **a, int64_t *lda, TYPE *beta, \ - TYPE **c, int64_t *ldc, int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, \ + int64_t* k, TYPE* alpha, const TYPE** a, int64_t* lda, TYPE* beta, \ + TYPE** c, int64_t* ldc, int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return syrk_batch(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, beta, \ c, ldc, group_count, group_size, dependencies); \ } @@ -1153,25 +1150,24 @@ SYRK_BATCH_LAUNCHER_USM(std::complex, rocblas_zsyrk_batched) #undef SYRK_BATCH_LAUNCHER_USM template -inline sycl::event syrk_batch(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - int64_t n, int64_t k, const T alpha, const T *a, int64_t lda, - int64_t stridea, const T beta, T *c, int64_t ldc, int64_t stridec, - int64_t batch_size, const std::vector &dependencies) { +inline sycl::event syrk_batch(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + int64_t n, int64_t k, const T alpha, const T* a, int64_t lda, + int64_t stridea, const T beta, T* c, int64_t ldc, int64_t stridec, + int64_t batch_size, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, ldc, stridea, stridec, batch_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto c_ = reinterpret_cast(c); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), n, k, (rocDataType *)&alpha, a_, - lda, stridea, (rocDataType *)&beta, c_, ldc, stridec, - batch_size); + get_rocblas_operation(trans), n, k, (rocDataType*)&alpha, a_, lda, + stridea, (rocDataType*)&beta, c_, ldc, stridec, batch_size); }); }); @@ -1179,11 +1175,11 @@ inline sycl::event syrk_batch(Func func, sycl::queue &queue, uplo upper_lower, t } #define SYRK_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, \ - int64_t k, const TYPE alpha, const TYPE *a, int64_t lda, \ - int64_t stridea, const TYPE beta, TYPE *c, int64_t ldc, \ + sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, \ + int64_t k, const TYPE alpha, const TYPE* a, int64_t lda, \ + int64_t stridea, const TYPE beta, TYPE* c, int64_t ldc, \ int64_t stridec, int64_t batch_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return syrk_batch(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, \ stridea, beta, c, ldc, stridec, batch_size, dependencies); \ } @@ -1196,11 +1192,11 @@ SYRK_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zsyrk_strided_batc #undef SYRK_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event omatcopy_batch(Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, const T alpha, const T *a, int64_t lda, - int64_t stridea, T *b, int64_t ldb, int64_t strideb, +inline sycl::event omatcopy_batch(Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, const T alpha, const T* a, int64_t lda, + int64_t stridea, T* b, int64_t ldb, int64_t strideb, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, stridea, strideb, batch_size); @@ -1208,18 +1204,18 @@ inline sycl::event omatcopy_batch(Func func, sycl::queue &queue, transpose trans const int64_t new_m = trans == oneapi::math::transpose::nontrans ? m : n; const int64_t new_n = trans == oneapi::math::transpose::nontrans ? n : m; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(trans), - get_rocblas_operation(trans), new_m, new_n, - (rocDataType *)&alpha, a_, lda, stridea, (rocDataType *)&beta, - nullptr, lda, stridea, b_, ldb, strideb, batch_size); + get_rocblas_operation(trans), new_m, new_n, (rocDataType*)&alpha, + a_, lda, stridea, (rocDataType*)&beta, nullptr, lda, stridea, b_, + ldb, strideb, batch_size); }); }); @@ -1227,10 +1223,10 @@ inline sycl::event omatcopy_batch(Func func, sycl::queue &queue, transpose trans } #define OMATCOPY_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, \ - const TYPE alpha, const TYPE *a, int64_t lda, int64_t stridea, \ - TYPE *b, int64_t ldb, int64_t strideb, int64_t batch_size, \ - const std::vector &dependencies) { \ + sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, \ + const TYPE alpha, const TYPE* a, int64_t lda, int64_t stridea, \ + TYPE* b, int64_t ldb, int64_t strideb, int64_t batch_size, \ + const std::vector& dependencies) { \ return omatcopy_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, stridea, b, ldb, \ strideb, batch_size, dependencies); \ } @@ -1242,54 +1238,54 @@ OMATCOPY_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zgeam_strided_ #undef OMATCOPY_STRIDED_BATCH_LAUNCHER_USM -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } template -inline sycl::event omatadd_batch(Func func, sycl::queue &queue, transpose transa, transpose transb, - int64_t m, int64_t n, const T alpha, const T *a, int64_t lda, - int64_t stridea, const T beta, const T *b, int64_t ldb, - int64_t strideb, T *c, int64_t ldc, int64_t stridec, - int64_t batch_size, const std::vector &dependencies) { +inline sycl::event omatadd_batch(Func func, sycl::queue& queue, transpose transa, transpose transb, + int64_t m, int64_t n, const T alpha, const T* a, int64_t lda, + int64_t stridea, const T beta, const T* b, int64_t ldb, + int64_t strideb, T* c, int64_t ldc, int64_t stridec, + int64_t batch_size, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc, stridea, strideb, stridec, batch_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(transa), - get_rocblas_operation(transb), m, n, (rocDataType *)&alpha, a_, - lda, stridea, (rocDataType *)&beta, b_, ldb, strideb, c_, ldc, - stridec, batch_size); + get_rocblas_operation(transb), m, n, (rocDataType*)&alpha, a_, lda, + stridea, (rocDataType*)&beta, b_, ldb, strideb, c_, ldc, stridec, + batch_size); }); }); @@ -1297,11 +1293,11 @@ inline sycl::event omatadd_batch(Func func, sycl::queue &queue, transpose transa } #define OMATADD_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, const TYPE alpha, const TYPE *a, int64_t lda, \ - int64_t stridea, const TYPE beta, const TYPE *b, int64_t ldb, \ - int64_t strideb, TYPE *c, int64_t ldc, int64_t stridec, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, const TYPE alpha, const TYPE* a, int64_t lda, \ + int64_t stridea, const TYPE beta, const TYPE* b, int64_t ldb, \ + int64_t strideb, TYPE* c, int64_t ldc, int64_t stridec, \ + int64_t batch_size, const std::vector& dependencies) { \ return omatadd_batch(ROCBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, lda, stridea, \ beta, b, ldb, strideb, c, ldc, stridec, batch_size, dependencies); \ } @@ -1314,35 +1310,35 @@ OMATADD_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zgeam_strided_b #undef OMATADD_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event omatcopy_batch(Func func, sycl::queue &queue, transpose *trans, int64_t *m, - int64_t *n, T *alpha, const T **a, int64_t *lda, T **b, - int64_t *ldb, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event omatcopy_batch(Func func, sycl::queue& queue, transpose* trans, int64_t* m, + int64_t* n, T* alpha, const T** a, int64_t* lda, T** b, + int64_t* ldb, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; for (int64_t i = 0; i < group_count; i++) { overflow_check(m[i], n[i], lda[i], ldb[i], group_size[i]); } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; for (int64_t i = 0; i < group_count; i++) { - auto **a_ = reinterpret_cast(a); - auto **b_ = reinterpret_cast(b); + auto** a_ = reinterpret_cast(a); + auto** b_ = reinterpret_cast(b); const T beta = 0; const auto new_m = trans[i] == oneapi::math::transpose::nontrans ? m[i] : n[i]; const auto new_n = trans[i] == oneapi::math::transpose::nontrans ? n[i] : m[i]; rocblas_native_func(func, err, handle, get_rocblas_operation(trans[i]), - get_rocblas_operation(trans[i]), (int)new_m, (int)new_n, - (rocDataType *)&alpha[i], a_ + offset, (int)lda[i], - (rocDataType *)&beta, nullptr, (int)lda[i], b_ + offset, - (int)ldb[i], (int)group_size[i]); + get_rocblas_operation(trans[i]), (int)new_m, (int)new_n, + (rocDataType*)&alpha[i], a_ + offset, (int)lda[i], + (rocDataType*)&beta, nullptr, (int)lda[i], b_ + offset, + (int)ldb[i], (int)group_size[i]); offset += group_size[i]; } }); @@ -1352,10 +1348,10 @@ inline sycl::event omatcopy_batch(Func func, sycl::queue &queue, transpose *tran } #define OMATCOPY_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, \ - TYPE *alpha, const TYPE **a, int64_t *lda, TYPE **b, int64_t *ldb, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, \ + TYPE* alpha, const TYPE** a, int64_t* lda, TYPE** b, int64_t* ldb, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return omatcopy_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, b, ldb, \ group_count, group_size, dependencies); \ } @@ -1367,31 +1363,31 @@ OMATCOPY_BATCH_LAUNCHER_USM(std::complex, rocblas_zgeam_batched) #undef OMATCOPY_BATCH_LAUNCHER_USM -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - float *alpha, float **ab, int64_t *lda, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + float* alpha, float** ab, int64_t* lda, int64_t* ldb, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - double *alpha, double **ab, int64_t *lda, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + double* alpha, double** ab, int64_t* lda, int64_t* ldb, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, std::complex **ab, int64_t *lda, - int64_t *ldb, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, std::complex** ab, int64_t* lda, + int64_t* ldb, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, std::complex **ab, int64_t *lda, - int64_t *ldb, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, std::complex** ab, int64_t* lda, + int64_t* ldb, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for column_major layout"); } @@ -1402,15 +1398,15 @@ namespace row_major { // Buffer APIs template -inline void copy_batch(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, +inline void copy_batch(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { column_major::copy_batch(func, queue, n, x, incx, stridex, y, incy, stridey, batch_size); } #define COPY_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void copy_batch(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - int64_t stridex, sycl::buffer &y, int64_t incy, int64_t stridey, \ + void copy_batch(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, \ int64_t batch_size) { \ copy_batch(ROCBLAS_ROUTINE, queue, n, x, incx, stridex, y, incy, stridey, batch_size); \ } @@ -1423,15 +1419,15 @@ COPY_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zcopy_strided_batched) #undef COPY_STRIDED_BATCH_LAUNCHER template -inline void axpy_batch(Func func, sycl::queue &queue, int64_t n, T alpha, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, +inline void axpy_batch(Func func, sycl::queue& queue, int64_t n, T alpha, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { column_major::axpy_batch(func, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } #define AXPY_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void axpy_batch(sycl::queue &queue, int64_t n, TYPE alpha, sycl::buffer &x, \ - int64_t incx, int64_t stridex, sycl::buffer &y, int64_t incy, \ + void axpy_batch(sycl::queue& queue, int64_t n, TYPE alpha, sycl::buffer& x, \ + int64_t incx, int64_t stridex, sycl::buffer& y, int64_t incy, \ int64_t stridey, int64_t batch_size) { \ axpy_batch(ROCBLAS_ROUTINE, queue, n, alpha, x, incx, stridex, y, incy, stridey, \ batch_size); \ @@ -1445,13 +1441,13 @@ AXPY_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zaxpy_strided_batched) #undef AXPY_STRIDED_BATCH_LAUNCHER template -inline void gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - int64_t stridea, sycl::buffer, 1> &x, int64_t incx, - int64_t stridex, std::complex beta, sycl::buffer, 1> &y, +inline void gemv_batch(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + int64_t stridea, sycl::buffer, 1>& x, int64_t incx, + int64_t stridex, std::complex beta, sycl::buffer, 1>& y, int64_t incy, int64_t stridey, int64_t batch_size) { auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); @@ -1459,11 +1455,11 @@ inline void gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m if (m > 0) { queue.submit( - [&](sycl::handler &cgh) { conj_vector(cgh, x, m, incx, stridex, batch_size); }); + [&](sycl::handler& cgh) { conj_vector(cgh, x, m, incx, stridex, batch_size); }); if (n > 0) { queue.submit( - [&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy, stridey, batch_size); }); + [&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy, stridey, batch_size); }); } } } @@ -1474,28 +1470,28 @@ inline void gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { queue.submit( - [&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy, stridey, batch_size); }); + [&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy, stridey, batch_size); }); } } } template -inline void gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - T alpha, sycl::buffer &a, int64_t lda, int64_t stridea, - sycl::buffer &x, int64_t incx, int64_t stridex, T beta, - sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { +inline void gemv_batch(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + T alpha, sycl::buffer& a, int64_t lda, int64_t stridea, + sycl::buffer& x, int64_t incx, int64_t stridex, T beta, + sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::gemv_batch(func, queue, new_trans, n, m, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } #define GEMV_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, int64_t stridea, \ - sycl::buffer &x, int64_t incx, int64_t stridex, TYPE beta, \ - sycl::buffer &y, int64_t incy, int64_t stridey, int64_t batch_size) { \ + void gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, int64_t stridea, \ + sycl::buffer& x, int64_t incx, int64_t stridex, TYPE beta, \ + sycl::buffer& y, int64_t incy, int64_t stridey, int64_t batch_size) { \ gemv_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, \ beta, y, incy, stridey, batch_size); \ } @@ -1508,22 +1504,22 @@ GEMV_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zgemv_strided_batched) #undef GEMV_STRIDED_BATCH_LAUNCHER template -inline void dgmm_batch(Func func, sycl::queue &queue, side left_right, int64_t m, int64_t n, - sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &x, - int64_t incx, int64_t stridex, sycl::buffer &c, int64_t ldc, +inline void dgmm_batch(Func func, sycl::queue& queue, side left_right, int64_t m, int64_t n, + sycl::buffer& a, int64_t lda, int64_t stridea, sycl::buffer& x, + int64_t incx, int64_t stridex, sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; column_major::dgmm_batch(func, queue, new_side, n, m, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } #define DGMM_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, \ - sycl::buffer &a, int64_t lda, int64_t stridea, \ - sycl::buffer &x, int64_t incx, int64_t stridex, \ - sycl::buffer &c, int64_t ldc, int64_t stridec, int64_t batch_size) { \ + void dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, \ + sycl::buffer& a, int64_t lda, int64_t stridea, \ + sycl::buffer& x, int64_t incx, int64_t stridex, \ + sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { \ dgmm_batch(ROCBLAS_ROUTINE, queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, \ ldc, stridec, batch_size); \ } @@ -1536,10 +1532,10 @@ DGMM_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zdgmm_strided_batched) #undef DGMM_STRIDED_BATCH_LAUNCHER template -inline void gemm_batch_impl(sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, Ts alpha, sycl::buffer &a, int64_t lda, - int64_t stridea, sycl::buffer &b, int64_t ldb, int64_t strideb, - Ts beta, sycl::buffer &c, int64_t ldc, int64_t stridec, +inline void gemm_batch_impl(sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, int64_t k, Ts alpha, sycl::buffer& a, int64_t lda, + int64_t stridea, sycl::buffer& b, int64_t ldb, int64_t strideb, + Ts beta, sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { auto new_transa = transb; auto new_transb = transa; @@ -1550,10 +1546,10 @@ inline void gemm_batch_impl(sycl::queue &queue, transpose transa, transpose tran #undef GEMM_STRIDED_BATCH_LAUNCHER #define GEMM_STRIDED_BATCH_LAUNCHER(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_S alpha, sycl::buffer &a, int64_t lda, \ - int64_t stridea, sycl::buffer &b, int64_t ldb, int64_t strideb, \ - TYPE_S beta, sycl::buffer &c, int64_t ldc, int64_t stridec, \ + void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_S alpha, sycl::buffer& a, int64_t lda, \ + int64_t stridea, sycl::buffer& b, int64_t ldb, int64_t strideb, \ + TYPE_S beta, sycl::buffer& c, int64_t ldc, int64_t stridec, \ int64_t batch_size) { \ gemm_batch_impl(queue, transa, transb, m, n, k, alpha, a, lda, stridea, b, ldb, strideb, \ beta, c, ldc, stridec, batch_size); \ @@ -1571,10 +1567,10 @@ GEMM_STRIDED_BATCH_LAUNCHER(sycl::half, sycl::half, float, float) #undef GEMM_STRIDED_BATCH_LAUNCHER #define GEMM_STRIDED_BATCH_LAUNCHER(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - void gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_S alpha, sycl::buffer &a, int64_t lda, \ - int64_t stridea, sycl::buffer &b, int64_t ldb, int64_t strideb, \ - TYPE_S beta, sycl::buffer &c, int64_t ldc, int64_t stridec, \ + void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_S alpha, sycl::buffer& a, int64_t lda, \ + int64_t stridea, sycl::buffer& b, int64_t ldb, int64_t strideb, \ + TYPE_S beta, sycl::buffer& c, int64_t ldc, int64_t stridec, \ int64_t batch_size) { \ throw unimplemented("blas", "gemm_batch", \ std::string("for dtype unimplemented dtype combination <") + \ @@ -1588,23 +1584,23 @@ GEMM_STRIDED_BATCH_LAUNCHER(std::int8_t, std::int8_t, std::int32_t, float) #undef GEMM_STRIDED_BATCH_LAUNCHER template -inline void trsm_batch(Func func, sycl::queue &queue, side left_right, uplo upper_lower, +inline void trsm_batch(Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, int64_t stridea, sycl::buffer &b, + sycl::buffer& a, int64_t lda, int64_t stridea, sycl::buffer& b, int64_t ldb, int64_t strideb, int64_t batch_size) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::trsm_batch(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, stridea, b, ldb, strideb, batch_size); } #define TRSM_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer &a, \ - int64_t lda, int64_t stridea, sycl::buffer &b, int64_t ldb, \ + void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer& a, \ + int64_t lda, int64_t stridea, sycl::buffer& b, int64_t ldb, \ int64_t strideb, int64_t batch_size) { \ trsm_batch(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, \ a, lda, stridea, b, ldb, strideb, batch_size); \ @@ -1618,23 +1614,23 @@ TRSM_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_ztrsm_strided_batched) #undef TRSM_STRIDED_BATCH_LAUNCHER template -inline void syrk_batch(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, T alpha, sycl::buffer &a, int64_t lda, int64_t stridea, - T beta, sycl::buffer &c, int64_t ldc, int64_t stridec, +inline void syrk_batch(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, T alpha, sycl::buffer& a, int64_t lda, int64_t stridea, + T beta, sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::syrk_batch(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, stridea, beta, c, ldc, stridec, batch_size); } #define SYRK_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, sycl::buffer &a, int64_t lda, int64_t stridea, TYPE beta, \ - sycl::buffer &c, int64_t ldc, int64_t stridec, int64_t batch_size) { \ + void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, sycl::buffer& a, int64_t lda, int64_t stridea, TYPE beta, \ + sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { \ syrk_batch(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, stridea, beta, \ c, ldc, stridec, batch_size); \ } @@ -1647,18 +1643,18 @@ SYRK_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zsyrk_strided_batched) #undef SYRK_STRIDED_BATCH_LAUNCHER template -inline void omatcopy_batch(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - const T alpha, sycl::buffer &a, int64_t lda, int64_t stridea, - sycl::buffer &b, int64_t ldb, int64_t strideb, +inline void omatcopy_batch(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + const T alpha, sycl::buffer& a, int64_t lda, int64_t stridea, + sycl::buffer& b, int64_t ldb, int64_t strideb, int64_t batch_size) { return column_major::omatcopy_batch(func, queue, trans, n, m, alpha, a, lda, stridea, b, ldb, strideb, batch_size); } #define OMATCOPY_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, \ - const TYPE alpha, sycl::buffer &a, int64_t lda, int64_t stridea, \ - sycl::buffer &b, int64_t ldb, int64_t strideb, \ + void omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, \ + const TYPE alpha, sycl::buffer& a, int64_t lda, int64_t stridea, \ + sycl::buffer& b, int64_t ldb, int64_t strideb, \ int64_t batch_size) { \ omatcopy_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, stridea, b, ldb, \ strideb, batch_size); \ @@ -1671,45 +1667,45 @@ OMATCOPY_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zgeam_strided_batc #undef OMATCOPY_STRIDED_BATCH_LAUNCHER -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb, int64_t stride, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -void imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } template -inline void omatadd_batch(Func func, sycl::queue &queue, transpose transa, transpose transb, - int64_t m, int64_t n, const T alpha, sycl::buffer &a, int64_t lda, - int64_t stridea, const T beta, sycl::buffer &b, int64_t ldb, - int64_t strideb, sycl::buffer &c, int64_t ldc, int64_t stridec, +inline void omatadd_batch(Func func, sycl::queue& queue, transpose transa, transpose transb, + int64_t m, int64_t n, const T alpha, sycl::buffer& a, int64_t lda, + int64_t stridea, const T beta, sycl::buffer& b, int64_t ldb, + int64_t strideb, sycl::buffer& c, int64_t ldc, int64_t stridec, int64_t batch_size) { return column_major::omatadd_batch(func, queue, transa, transb, n, m, alpha, a, lda, stridea, beta, b, ldb, strideb, c, ldc, stridec, batch_size); } #define OMATADD_STRIDED_BATCH_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, const TYPE alpha, sycl::buffer &a, int64_t lda, \ - int64_t stridea, const TYPE beta, sycl::buffer &b, int64_t ldb, \ - int64_t strideb, sycl::buffer &c, int64_t ldc, int64_t stridec, \ + void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, const TYPE alpha, sycl::buffer& a, int64_t lda, \ + int64_t stridea, const TYPE beta, sycl::buffer& b, int64_t ldb, \ + int64_t strideb, sycl::buffer& c, int64_t ldc, int64_t stridec, \ int64_t batch_size) { \ omatadd_batch(ROCBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, lda, stridea, beta, \ b, ldb, strideb, c, ldc, stridec, batch_size); \ @@ -1725,17 +1721,17 @@ OMATADD_STRIDED_BATCH_LAUNCHER(std::complex, rocblas_zgeam_strided_batch // USM APIs template -inline sycl::event copy_batch(Func func, sycl::queue &queue, int64_t *n, const T **x, int64_t *incx, - T **y, int64_t *incy, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event copy_batch(Func func, sycl::queue& queue, int64_t* n, const T** x, int64_t* incx, + T** y, int64_t* incy, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { return column_major::copy_batch(func, queue, n, x, incx, y, incy, group_count, group_size, dependencies); } #define COPY_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event copy_batch(sycl::queue &queue, int64_t *n, const TYPE **x, int64_t *incx, \ - TYPE **y, int64_t *incy, int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event copy_batch(sycl::queue& queue, int64_t* n, const TYPE** x, int64_t* incx, \ + TYPE** y, int64_t* incy, int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return copy_batch(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, group_count, group_size, \ dependencies); \ } @@ -1748,17 +1744,17 @@ COPY_BATCH_LAUNCHER_USM(std::complex, rocblas_zcopy_batched) #undef COPY_BATCH_LAUNCHER_USM template -inline sycl::event copy_batch(Func func, sycl::queue &queue, int64_t n, const T *x, int64_t incx, - int64_t stridex, T *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +inline sycl::event copy_batch(Func func, sycl::queue& queue, int64_t n, const T* x, int64_t incx, + int64_t stridex, T* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { return column_major::copy_batch(func, queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } #define COPY_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event copy_batch(sycl::queue &queue, int64_t n, const TYPE *x, int64_t incx, \ - int64_t stridex, TYPE *y, int64_t incy, int64_t stridey, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event copy_batch(sycl::queue& queue, int64_t n, const TYPE* x, int64_t incx, \ + int64_t stridex, TYPE* y, int64_t incy, int64_t stridey, \ + int64_t batch_size, const std::vector& dependencies) { \ return copy_batch(ROCBLAS_ROUTINE, queue, n, x, incx, stridex, y, incy, stridey, \ batch_size, dependencies); \ } @@ -1771,17 +1767,17 @@ COPY_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zcopy_strided_batc #undef COPY_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event axpy_batch(Func func, sycl::queue &queue, int64_t *n, T *alpha, const T **x, - int64_t *incx, T **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +inline sycl::event axpy_batch(Func func, sycl::queue& queue, int64_t* n, T* alpha, const T** x, + int64_t* incx, T** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { return column_major::axpy_batch(func, queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } #define AXPY_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event axpy_batch(sycl::queue &queue, int64_t *n, TYPE *alpha, const TYPE **x, \ - int64_t *incx, TYPE **y, int64_t *incy, int64_t group_count, \ - int64_t *group_size, const std::vector &dependencies) { \ + sycl::event axpy_batch(sycl::queue& queue, int64_t* n, TYPE* alpha, const TYPE** x, \ + int64_t* incx, TYPE** y, int64_t* incy, int64_t group_count, \ + int64_t* group_size, const std::vector& dependencies) { \ return axpy_batch(ROCBLAS_ROUTINE, queue, n, alpha, x, incx, y, incy, group_count, \ group_size, dependencies); \ } @@ -1794,17 +1790,17 @@ AXPY_BATCH_LAUNCHER_USM(std::complex, rocblas_zaxpy_batched) #undef AXPY_BATCH_LAUNCHER_USM template -inline sycl::event axpy_batch(Func func, sycl::queue &queue, int64_t n, T alpha, const T *x, - int64_t incx, int64_t stridex, T *y, int64_t incy, int64_t stridey, - int64_t batch_size, const std::vector &dependencies) { +inline sycl::event axpy_batch(Func func, sycl::queue& queue, int64_t n, T alpha, const T* x, + int64_t incx, int64_t stridex, T* y, int64_t incy, int64_t stridey, + int64_t batch_size, const std::vector& dependencies) { return column_major::axpy_batch(func, queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } #define AXPY_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event axpy_batch(sycl::queue &queue, int64_t n, TYPE alpha, const TYPE *x, int64_t incx, \ - int64_t stridex, TYPE *y, int64_t incy, int64_t stridey, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event axpy_batch(sycl::queue& queue, int64_t n, TYPE alpha, const TYPE* x, int64_t incx, \ + int64_t stridex, TYPE* y, int64_t incy, int64_t stridey, \ + int64_t batch_size, const std::vector& dependencies) { \ return axpy_batch(ROCBLAS_ROUTINE, queue, n, alpha, x, incx, stridex, y, incy, stridey, \ batch_size, dependencies); \ } @@ -1817,29 +1813,29 @@ AXPY_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zaxpy_strided_batc #undef AXPY_BATCH_LAUNCHER_USM template -inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - int64_t stridea, const std::complex *x, int64_t incx, - int64_t stridex, std::complex beta, std::complex *y, +inline sycl::event gemv_batch(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + int64_t stridea, const std::complex* x, int64_t incx, + int64_t stridex, std::complex beta, std::complex* y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { sycl::event done; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); beta = std::conj(beta); if (m > 0) { - done = queue.submit([&](sycl::handler &cgh) { - conj_vector(cgh, (std::complex *)x, m, incx, stridex, batch_size); + done = queue.submit([&](sycl::handler& cgh) { + conj_vector(cgh, (std::complex*)x, m, incx, stridex, batch_size); }); if (n > 0) { done = queue.submit( - [&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy, stridey, batch_size); }); + [&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy, stridey, batch_size); }); } } } @@ -1851,7 +1847,7 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose trans, in if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, y, n, incy, stridey, batch_size); }); @@ -1862,24 +1858,24 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose trans, in } template -inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - T alpha, const T *a, int64_t lda, int64_t stridea, const T *x, - int64_t incx, int64_t stridex, T beta, T *y, int64_t incy, +inline sycl::event gemv_batch(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + T alpha, const T* a, int64_t lda, int64_t stridea, const T* x, + int64_t incx, int64_t stridex, T beta, T* y, int64_t incy, int64_t stridey, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::gemv_batch(func, queue, new_trans, n, m, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); } #define GEMV_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event gemv_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, int64_t stridea, const TYPE *x, \ - int64_t incx, int64_t stridex, TYPE beta, TYPE *y, int64_t incy, \ + sycl::event gemv_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, int64_t stridea, const TYPE* x, \ + int64_t incx, int64_t stridex, TYPE beta, TYPE* y, int64_t incy, \ int64_t stridey, int64_t batch_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return gemv_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, stridea, x, incx, \ stridex, beta, y, incy, stridey, batch_size, dependencies); \ } @@ -1892,12 +1888,12 @@ GEMV_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zgemv_strided_batc #undef GEMV_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, int64_t *m, - int64_t *n, std::complex *alpha, const std::complex **a, - int64_t *lda, const std::complex **x, int64_t *incx, - std::complex *beta, std::complex **y, int64_t *incy, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event gemv_batch(Func func, sycl::queue& queue, transpose* trans, int64_t* m, + int64_t* n, std::complex* alpha, const std::complex** a, + int64_t* lda, const std::complex** x, int64_t* incx, + std::complex* beta, std::complex** y, int64_t* incy, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { sycl::event done; int64_t stride = 0; @@ -1907,12 +1903,12 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, i beta[i] = std::conj(beta[i]); if (m[i] > 0) { - done = queue.submit([&](sycl::handler &cgh) { - conj_vector(cgh, (std::complex **)x, m[i], incx[i], stride, group_size[i]); + done = queue.submit([&](sycl::handler& cgh) { + conj_vector(cgh, (std::complex**)x, m[i], incx[i], stride, group_size[i]); }); if (n[i] > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n[i], incy[i], stride, group_size[i]); }); } @@ -1942,7 +1938,7 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, i for (int64_t i = 0; i < group_count; i++) { if (trans[i] == oneapi::math::transpose::conjtrans) { if (n[i] > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n[i], incy[i], stride, group_size[i]); }); } @@ -1954,10 +1950,10 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, i } template -inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, int64_t *m, - int64_t *n, T *alpha, const T **a, int64_t *lda, const T **x, - int64_t *incx, T *beta, T **y, int64_t *incy, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +inline sycl::event gemv_batch(Func func, sycl::queue& queue, transpose* trans, int64_t* m, + int64_t* n, T* alpha, const T** a, int64_t* lda, const T** x, + int64_t* incx, T* beta, T** y, int64_t* incy, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { auto tmp_trans = std::vector{ static_cast(group_count) }; for (int64_t i = 0; i < group_count; i++) { @@ -1979,9 +1975,9 @@ inline sycl::event gemv_batch(Func func, sycl::queue &queue, transpose *trans, i #define GEMV_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ sycl::event gemv_batch( \ - sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, TYPE *alpha, const TYPE **a, \ - int64_t *lda, const TYPE **x, int64_t *incx, TYPE *beta, TYPE **y, int64_t *incy, \ - int64_t group_count, int64_t *group_size, const std::vector &dependencies) { \ + sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, TYPE* alpha, const TYPE** a, \ + int64_t* lda, const TYPE** x, int64_t* incx, TYPE* beta, TYPE** y, int64_t* incy, \ + int64_t group_count, int64_t* group_size, const std::vector& dependencies) { \ return gemv_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, x, incx, beta, y, \ incy, group_count, group_size, dependencies); \ } @@ -1994,22 +1990,22 @@ GEMV_BATCH_LAUNCHER_USM(std::complex, rocblas_zgemv_batched) #undef GEMV_BATCH_LAUNCHER_USM template -inline sycl::event dgmm_batch(Func func, sycl::queue &queue, side left_right, int64_t m, int64_t n, - const T *a, int64_t lda, int64_t stridea, const T *x, int64_t incx, - int64_t stridex, T *c, int64_t ldc, int64_t stridec, - int64_t batch_size, const std::vector &dependencies) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; +inline sycl::event dgmm_batch(Func func, sycl::queue& queue, side left_right, int64_t m, int64_t n, + const T* a, int64_t lda, int64_t stridea, const T* x, int64_t incx, + int64_t stridex, T* c, int64_t ldc, int64_t stridec, + int64_t batch_size, const std::vector& dependencies) { + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; return column_major::dgmm_batch(func, queue, new_side, n, m, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); } #define DGMM_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event dgmm_batch(sycl::queue &queue, side left_right, int64_t m, int64_t n, \ - const TYPE *a, int64_t lda, int64_t stridea, const TYPE *x, \ - int64_t incx, int64_t stridex, TYPE *c, int64_t ldc, int64_t stridec, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event dgmm_batch(sycl::queue& queue, side left_right, int64_t m, int64_t n, \ + const TYPE* a, int64_t lda, int64_t stridea, const TYPE* x, \ + int64_t incx, int64_t stridex, TYPE* c, int64_t ldc, int64_t stridec, \ + int64_t batch_size, const std::vector& dependencies) { \ return dgmm_batch(ROCBLAS_ROUTINE, queue, left_right, m, n, a, lda, stridea, x, incx, \ stridex, c, ldc, stridec, batch_size, dependencies); \ } @@ -2022,13 +2018,13 @@ DGMM_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zdgmm_strided_batc #undef DGMM_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event dgmm_batch(Func func, sycl::queue &queue, side *left_right, int64_t *m, - int64_t *n, const T **a, int64_t *lda, const T **x, int64_t *incx, - T **c, int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event dgmm_batch(Func func, sycl::queue& queue, side* left_right, int64_t* m, + int64_t* n, const T** a, int64_t* lda, const T** x, int64_t* incx, + T** c, int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { for (int64_t i = 0; i < group_count; i++) { const auto new_side = left_right[i] == oneapi::math::side::left ? oneapi::math::side::right - : oneapi::math::side::left; + : oneapi::math::side::left; left_right[i] = new_side; } @@ -2037,10 +2033,10 @@ inline sycl::event dgmm_batch(Func func, sycl::queue &queue, side *left_right, i } #define DGMM_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event dgmm_batch(sycl::queue &queue, side *left_right, int64_t *m, int64_t *n, \ - const TYPE **a, int64_t *lda, const TYPE **x, int64_t *incx, TYPE **c, \ - int64_t *ldc, int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event dgmm_batch(sycl::queue& queue, side* left_right, int64_t* m, int64_t* n, \ + const TYPE** a, int64_t* lda, const TYPE** x, int64_t* incx, TYPE** c, \ + int64_t* ldc, int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return dgmm_batch(ROCBLAS_ROUTINE, queue, left_right, m, n, a, lda, x, incx, c, ldc, \ group_count, group_size, dependencies); \ } @@ -2053,13 +2049,13 @@ DGMM_BATCH_LAUNCHER_USM(std::complex, rocblas_zdgmm_batched) #undef DGMM_BATCH_LAUNCHER template -inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose transa, +inline sycl::event gemm_batch_strided_usm_impl(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, int64_t k, - Ts alpha, const Ta *a, int64_t lda, int64_t stridea, - const Tb *b, int64_t ldb, int64_t strideb, Ts beta, - Tc *c, int64_t ldc, int64_t stridec, + Ts alpha, const Ta* a, int64_t lda, int64_t stridea, + const Tb* b, int64_t ldb, int64_t strideb, Ts beta, + Tc* c, int64_t ldc, int64_t stridec, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { auto new_transa = transb; auto new_transb = transa; @@ -2069,11 +2065,11 @@ inline sycl::event gemm_batch_strided_usm_impl(sycl::queue &queue, transpose tra } #define GEMM_STRIDED_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, int64_t k, TYPE_S alpha, const TYPE_A *a, int64_t lda, \ - int64_t stridea, const TYPE_B *b, int64_t ldb, int64_t strideb, \ - TYPE_S beta, TYPE_C *c, int64_t ldc, int64_t stridec, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, int64_t k, TYPE_S alpha, const TYPE_A* a, int64_t lda, \ + int64_t stridea, const TYPE_B* b, int64_t ldb, int64_t strideb, \ + TYPE_S beta, TYPE_C* c, int64_t ldc, int64_t stridec, \ + int64_t batch_size, const std::vector& dependencies) { \ return gemm_batch_strided_usm_impl(queue, transa, transb, m, n, k, alpha, a, lda, stridea, \ b, ldb, strideb, beta, c, ldc, stridec, batch_size, \ dependencies); \ @@ -2091,11 +2087,11 @@ GEMM_STRIDED_BATCH_LAUNCHER_USM(sycl::half, sycl::half, float, float) #undef GEMM_STRIDED_BATCH_LAUNCHER_USM #define GEMM_STRIDED_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, int64_t k, TYPE_S alpha, const TYPE_A *a, int64_t lda, \ - int64_t stridea, const TYPE_B *b, int64_t ldb, int64_t strideb, \ - TYPE_S beta, TYPE_C *c, int64_t ldc, int64_t stridec, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, int64_t k, TYPE_S alpha, const TYPE_A* a, int64_t lda, \ + int64_t stridea, const TYPE_B* b, int64_t ldb, int64_t strideb, \ + TYPE_S beta, TYPE_C* c, int64_t ldc, int64_t stridec, \ + int64_t batch_size, const std::vector& dependencies) { \ throw unimplemented("blas", "gemm_batch", \ std::string("for dtype unimplemented dtype combination <") + \ dtype_string() + "," + dtype_string() + "," + \ @@ -2108,11 +2104,11 @@ GEMM_STRIDED_BATCH_LAUNCHER_USM(std::int8_t, std::int8_t, std::int32_t, float) #undef GEMM_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, transpose *transb, - int64_t *m, int64_t *n, int64_t *k, Ts *alpha, const Ta **a, - int64_t *lda, const Tb **b, int64_t *ldb, Ts *beta, Tc **c, - int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event gemm_batch_usm_impl(sycl::queue& queue, transpose* transa, transpose* transb, + int64_t* m, int64_t* n, int64_t* k, Ts* alpha, const Ta** a, + int64_t* lda, const Tb** b, int64_t* ldb, Ts* beta, Tc** c, + int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { for (int64_t i = 0; i < group_count; i++) { std::swap(transa[i], transb[i]); } @@ -2122,11 +2118,11 @@ inline sycl::event gemm_batch_usm_impl(sycl::queue &queue, transpose *transa, tr } #define GEMM_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, \ - int64_t *n, int64_t *k, TYPE_S *alpha, const TYPE_A **a, int64_t *lda, \ - const TYPE_B **b, int64_t *ldb, TYPE_S *beta, TYPE_C **c, int64_t *ldc, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, \ + int64_t* n, int64_t* k, TYPE_S* alpha, const TYPE_A** a, int64_t* lda, \ + const TYPE_B** b, int64_t* ldb, TYPE_S* beta, TYPE_C** c, int64_t* ldc, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return gemm_batch_usm_impl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, \ ldc, group_count, group_size, dependencies); \ } @@ -2143,11 +2139,11 @@ GEMM_BATCH_LAUNCHER_USM(sycl::half, sycl::half, float, float) #undef GEMM_BATCH_LAUNCHER_USM #define GEMM_BATCH_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S) \ - sycl::event gemm_batch(sycl::queue &queue, transpose *transa, transpose *transb, int64_t *m, \ - int64_t *n, int64_t *k, TYPE_S *alpha, const TYPE_A **a, int64_t *lda, \ - const TYPE_B **b, int64_t *ldb, TYPE_S *beta, TYPE_C **c, int64_t *ldc, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event gemm_batch(sycl::queue& queue, transpose* transa, transpose* transb, int64_t* m, \ + int64_t* n, int64_t* k, TYPE_S* alpha, const TYPE_A** a, int64_t* lda, \ + const TYPE_B** b, int64_t* ldb, TYPE_S* beta, TYPE_C** c, int64_t* ldc, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ throw unimplemented("blas", "gemm_batch", \ std::string("for dtype unimplemented dtype combination <") + \ dtype_string() + "," + dtype_string() + "," + \ @@ -2160,25 +2156,25 @@ GEMM_BATCH_LAUNCHER_USM(std::int8_t, std::int8_t, std::int32_t, float) #undef GEMM_BATCH_LAUNCHER_USM template -inline sycl::event trsm_batch(Func func, sycl::queue &queue, side left_right, uplo upper_lower, +inline sycl::event trsm_batch(Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, - const T *a, int64_t lda, int64_t stridea, T *b, int64_t ldb, + const T* a, int64_t lda, int64_t stridea, T* b, int64_t ldb, int64_t strideb, int64_t batch_size, - const std::vector &dependencies) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; + const std::vector& dependencies) { + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::trsm_batch(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, stridea, b, ldb, strideb, batch_size, dependencies); } #define TRSM_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trsm_batch(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE *a, \ - int64_t lda, int64_t stridea, TYPE *b, int64_t ldb, int64_t strideb, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE* a, \ + int64_t lda, int64_t stridea, TYPE* b, int64_t ldb, int64_t strideb, \ + int64_t batch_size, const std::vector& dependencies) { \ return trsm_batch(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, \ alpha, a, lda, stridea, b, ldb, strideb, batch_size, dependencies); \ } @@ -2191,17 +2187,18 @@ TRSM_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_ztrsm_strided_batc #undef TRSM_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event trsm_batch(Func func, sycl::queue &queue, side *left_right, uplo *upper_lower, - transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, T *alpha, - const T **a, int64_t *lda, T **b, int64_t *ldb, int64_t group_count, - int64_t *group_size, const std::vector &dependencies) { +inline sycl::event trsm_batch(Func func, sycl::queue& queue, side* left_right, uplo* upper_lower, + transpose* trans, diag* unit_diag, int64_t* m, int64_t* n, T* alpha, + const T** a, int64_t* lda, T** b, int64_t* ldb, int64_t group_count, + int64_t* group_size, const std::vector& dependencies) { for (int64_t i = 0; i < group_count; i++) { const auto new_side = left_right[i] == oneapi::math::side::left ? oneapi::math::side::right - : oneapi::math::side::left; + : oneapi::math::side::left; left_right[i] = new_side; - const auto new_uplo = upper_lower[i] == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + const auto new_uplo = upper_lower[i] == oneapi::math::uplo::lower + ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; upper_lower[i] = new_uplo; } @@ -2210,11 +2207,11 @@ inline sycl::event trsm_batch(Func func, sycl::queue &queue, side *left_right, u } #define TRSM_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trsm_batch(sycl::queue &queue, side *left_right, uplo *upper_lower, \ - transpose *trans, diag *unit_diag, int64_t *m, int64_t *n, TYPE *alpha, \ - const TYPE **a, int64_t *lda, TYPE **b, int64_t *ldb, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event trsm_batch(sycl::queue& queue, side* left_right, uplo* upper_lower, \ + transpose* trans, diag* unit_diag, int64_t* m, int64_t* n, TYPE* alpha, \ + const TYPE** a, int64_t* lda, TYPE** b, int64_t* ldb, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return trsm_batch(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, \ alpha, a, lda, b, ldb, group_count, group_size, dependencies); \ } @@ -2227,13 +2224,14 @@ TRSM_BATCH_LAUNCHER_USM(std::complex, rocblas_ztrsm_batched) #undef TRSM_BATCH_LAUNCHER_USM template -inline sycl::event syrk_batch(Func func, sycl::queue &queue, uplo *upper_lower, transpose *trans, - int64_t *n, int64_t *k, T *alpha, const T **a, int64_t *lda, T *beta, - T **c, int64_t *ldc, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event syrk_batch(Func func, sycl::queue& queue, uplo* upper_lower, transpose* trans, + int64_t* n, int64_t* k, T* alpha, const T** a, int64_t* lda, T* beta, + T** c, int64_t* ldc, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { for (int64_t i = 0; i < group_count; i++) { - const auto new_uplo = upper_lower[i] == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + const auto new_uplo = upper_lower[i] == oneapi::math::uplo::lower + ? oneapi::math::uplo::upper + : oneapi::math::uplo::lower; upper_lower[i] = new_uplo; const auto new_trans = trans[i] == oneapi::math::transpose::nontrans @@ -2247,10 +2245,10 @@ inline sycl::event syrk_batch(Func func, sycl::queue &queue, uplo *upper_lower, } #define SYRK_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syrk_batch(sycl::queue &queue, uplo *upper_lower, transpose *trans, int64_t *n, \ - int64_t *k, TYPE *alpha, const TYPE **a, int64_t *lda, TYPE *beta, \ - TYPE **c, int64_t *ldc, int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event syrk_batch(sycl::queue& queue, uplo* upper_lower, transpose* trans, int64_t* n, \ + int64_t* k, TYPE* alpha, const TYPE** a, int64_t* lda, TYPE* beta, \ + TYPE** c, int64_t* ldc, int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return syrk_batch(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, beta, \ c, ldc, group_count, group_size, dependencies); \ } @@ -2263,25 +2261,25 @@ SYRK_BATCH_LAUNCHER_USM(std::complex, rocblas_zsyrk_batched) #undef SYRK_BATCH_LAUNCHER_USM template -inline sycl::event syrk_batch(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - int64_t n, int64_t k, const T alpha, const T *a, int64_t lda, - int64_t stridea, const T beta, T *c, int64_t ldc, int64_t stridec, - int64_t batch_size, const std::vector &dependencies) { +inline sycl::event syrk_batch(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + int64_t n, int64_t k, const T alpha, const T* a, int64_t lda, + int64_t stridea, const T beta, T* c, int64_t ldc, int64_t stridec, + int64_t batch_size, const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::syrk_batch(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, stridea, beta, c, ldc, stridec, batch_size, dependencies); } #define SYRK_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syrk_batch(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, \ - int64_t k, const TYPE alpha, const TYPE *a, int64_t lda, \ - int64_t stridea, const TYPE beta, TYPE *c, int64_t ldc, \ + sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, \ + int64_t k, const TYPE alpha, const TYPE* a, int64_t lda, \ + int64_t stridea, const TYPE beta, TYPE* c, int64_t ldc, \ int64_t stridec, int64_t batch_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return syrk_batch(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, \ stridea, beta, c, ldc, stridec, batch_size, dependencies); \ } @@ -2294,20 +2292,20 @@ SYRK_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zsyrk_strided_batc #undef SYRK_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event omatcopy_batch(Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, const T alpha, const T *a, int64_t lda, - int64_t stridea, T *b, int64_t ldb, int64_t strideb, +inline sycl::event omatcopy_batch(Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, const T alpha, const T* a, int64_t lda, + int64_t stridea, T* b, int64_t ldb, int64_t strideb, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return column_major::omatcopy_batch(func, queue, trans, n, m, alpha, a, lda, stridea, b, ldb, strideb, batch_size, dependencies); } #define OMATCOPY_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, \ - const TYPE alpha, const TYPE *a, int64_t lda, int64_t stridea, \ - TYPE *b, int64_t ldb, int64_t strideb, int64_t batch_size, \ - const std::vector &dependencies) { \ + sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, \ + const TYPE alpha, const TYPE* a, int64_t lda, int64_t stridea, \ + TYPE* b, int64_t ldb, int64_t strideb, int64_t batch_size, \ + const std::vector& dependencies) { \ return omatcopy_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, stridea, b, ldb, \ strideb, batch_size, dependencies); \ } @@ -2319,49 +2317,49 @@ OMATCOPY_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zgeam_strided_ #undef OMATCOPY_STRIDED_BATCH_LAUNCHER_USM -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, +sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, int64_t stride, int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } template -inline sycl::event omatadd_batch(Func func, sycl::queue &queue, transpose transa, transpose transb, - int64_t m, int64_t n, const T alpha, const T *a, int64_t lda, - int64_t stridea, const T beta, const T *b, int64_t ldb, - int64_t strideb, T *c, int64_t ldc, int64_t stridec, - int64_t batch_size, const std::vector &dependencies) { +inline sycl::event omatadd_batch(Func func, sycl::queue& queue, transpose transa, transpose transb, + int64_t m, int64_t n, const T alpha, const T* a, int64_t lda, + int64_t stridea, const T beta, const T* b, int64_t ldb, + int64_t strideb, T* c, int64_t ldc, int64_t stridec, + int64_t batch_size, const std::vector& dependencies) { return column_major::omatadd_batch(func, queue, transa, transb, n, m, alpha, a, lda, stridea, beta, b, ldb, strideb, c, ldc, stridec, batch_size, dependencies); } #define OMATADD_STRIDED_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatadd_batch(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, const TYPE alpha, const TYPE *a, int64_t lda, \ - int64_t stridea, const TYPE beta, const TYPE *b, int64_t ldb, \ - int64_t strideb, TYPE *c, int64_t ldc, int64_t stridec, \ - int64_t batch_size, const std::vector &dependencies) { \ + sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, const TYPE alpha, const TYPE* a, int64_t lda, \ + int64_t stridea, const TYPE beta, const TYPE* b, int64_t ldb, \ + int64_t strideb, TYPE* c, int64_t ldc, int64_t stridec, \ + int64_t batch_size, const std::vector& dependencies) { \ return omatadd_batch(ROCBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, lda, stridea, \ beta, b, ldb, strideb, c, ldc, stridec, batch_size, dependencies); \ } @@ -2374,19 +2372,19 @@ OMATADD_STRIDED_BATCH_LAUNCHER_USM(std::complex, rocblas_zgeam_strided_b #undef OMATADD_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event omatcopy_batch(Func func, sycl::queue &queue, transpose *trans, int64_t *m, - int64_t *n, T *alpha, const T **a, int64_t *lda, T **b, - int64_t *ldb, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +inline sycl::event omatcopy_batch(Func func, sycl::queue& queue, transpose* trans, int64_t* m, + int64_t* n, T* alpha, const T** a, int64_t* lda, T** b, + int64_t* ldb, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { return column_major::omatcopy_batch(func, queue, trans, n, m, alpha, a, lda, b, ldb, group_count, group_size, dependencies); } #define OMATCOPY_BATCH_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, \ - TYPE *alpha, const TYPE **a, int64_t *lda, TYPE **b, int64_t *ldb, \ - int64_t group_count, int64_t *group_size, \ - const std::vector &dependencies) { \ + sycl::event omatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, \ + TYPE* alpha, const TYPE** a, int64_t* lda, TYPE** b, int64_t* ldb, \ + int64_t group_count, int64_t* group_size, \ + const std::vector& dependencies) { \ return omatcopy_batch(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, b, ldb, \ group_count, group_size, dependencies); \ } @@ -2398,31 +2396,31 @@ OMATCOPY_BATCH_LAUNCHER_USM(std::complex, rocblas_zgeam_batched) #undef OMATCOPY_BATCH_LAUNCHER_USM -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - float *alpha, float **ab, int64_t *lda, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + float* alpha, float** ab, int64_t* lda, int64_t* ldb, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - double *alpha, double **ab, int64_t *lda, int64_t *ldb, - int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + double* alpha, double** ab, int64_t* lda, int64_t* ldb, + int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, std::complex **ab, int64_t *lda, - int64_t *ldb, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, std::complex** ab, int64_t* lda, + int64_t* ldb, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } -sycl::event imatcopy_batch(sycl::queue &queue, transpose *trans, int64_t *m, int64_t *n, - std::complex *alpha, std::complex **ab, int64_t *lda, - int64_t *ldb, int64_t group_count, int64_t *group_size, - const std::vector &dependencies) { +sycl::event imatcopy_batch(sycl::queue& queue, transpose* trans, int64_t* m, int64_t* n, + std::complex* alpha, std::complex** ab, int64_t* lda, + int64_t* ldb, int64_t group_count, int64_t* group_size, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy_batch", "for row_major layout"); } diff --git a/src/blas/backends/rocblas/rocblas_extensions.cpp b/src/blas/backends/rocblas/rocblas_extensions.cpp index a0b7d56e8..b649ed1ef 100644 --- a/src/blas/backends/rocblas/rocblas_extensions.cpp +++ b/src/blas/backends/rocblas/rocblas_extensions.cpp @@ -33,65 +33,65 @@ namespace column_major { // Buffer APIs -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, double beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for column_major layout"); } template -inline void omatcopy(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - const T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, +inline void omatcopy(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + const T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb); @@ -100,26 +100,26 @@ inline void omatcopy(Func func, sycl::queue &queue, transpose trans, int64_t m, const int64_t new_m = trans == oneapi::math::transpose::nontrans ? m : n; const int64_t new_n = trans == oneapi::math::transpose::nontrans ? n : m; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); rocblas_status err; ROCBLAS_ERROR_FUNC_SYNC(func, err, handle, get_rocblas_operation(trans), get_rocblas_operation(trans), new_m, new_n, - (rocDataType *)&alpha, a_, lda, (rocDataType *)&beta, nullptr, + (rocDataType*)&alpha, a_, lda, (rocDataType*)&beta, nullptr, lda, b_, ldb); }); }); } #define OMATCOPY_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, const TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, const TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { \ omatcopy(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, b, ldb); \ } @@ -131,16 +131,16 @@ OMATCOPY_LAUNCHER(std::complex, rocblas_zgeam) #undef OMATCOPY_LAUNCHER template -void omatcopy2(const char *func_name, Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb) { +void omatcopy2(const char* func_name, Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb) { throw unimplemented("blas", "omatcopy2", ""); } #define OMATCOPY2_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, int64_t stridea, \ - sycl::buffer &b, int64_t ldb, int64_t strideb) { \ + void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, int64_t stridea, \ + sycl::buffer& b, int64_t ldb, int64_t strideb) { \ omatcopy2(#ROCBLAS_ROUTINE, ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, stridea, lda, \ b, ldb, strideb); \ } @@ -152,55 +152,55 @@ OMATCOPY2_LAUNCHER(std::complex, "unimplemented") #undef OMATCOPY2_LAUNCHER -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } template -inline void omatadd(Func func, sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, const T alpha, sycl::buffer &a, int64_t lda, const T beta, - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { +inline void omatadd(Func func, sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, const T alpha, sycl::buffer& a, int64_t lda, const T beta, + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; ROCBLAS_ERROR_FUNC_SYNC(func, err, handle, get_rocblas_operation(transa), - get_rocblas_operation(transb), m, n, (rocDataType *)&alpha, a_, - lda, (rocDataType *)&beta, b_, ldb, c_, ldc); + get_rocblas_operation(transb), m, n, (rocDataType*)&alpha, a_, + lda, (rocDataType*)&beta, b_, ldb, c_, ldc); }); }); } #define OMATADD_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - const TYPE alpha, sycl::buffer &a, int64_t lda, const TYPE beta, \ - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { \ + void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + const TYPE alpha, sycl::buffer& a, int64_t lda, const TYPE beta, \ + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { \ omatadd(ROCBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, \ ldc); \ } @@ -214,72 +214,72 @@ OMATADD_LAUNCHER(std::complex, rocblas_zgeam) // USM APIs -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const int8_t *a, int64_t lda, - int8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const int8_t *a, int64_t lda, - int8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const uint8_t *a, int64_t lda, - uint8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const uint8_t *a, int64_t lda, - uint8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for column_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, float alpha, const float *a, int64_t lda, const float *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, float alpha, const float* a, int64_t lda, const float* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, double alpha, const double *a, int64_t lda, const double *b, - int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for column_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for column_major layout"); } template -inline sycl::event omatcopy(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - const T alpha, const T *a, int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { +inline sycl::event omatcopy(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + const T alpha, const T* a, int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb); @@ -287,17 +287,17 @@ inline sycl::event omatcopy(Func func, sycl::queue &queue, transpose trans, int6 const int64_t new_m = trans == oneapi::math::transpose::nontrans ? m : n; const int64_t new_n = trans == oneapi::math::transpose::nontrans ? n : m; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); rocblas_status err; ROCBLAS_ERROR_FUNC_SYNC(func, err, handle, get_rocblas_operation(trans), get_rocblas_operation(trans), new_m, new_n, - (rocDataType *)&alpha, a_, lda, (rocDataType *)&beta, nullptr, + (rocDataType*)&alpha, a_, lda, (rocDataType*)&beta, nullptr, lda, b_, ldb); }); }); @@ -306,9 +306,9 @@ inline sycl::event omatcopy(Func func, sycl::queue &queue, transpose trans, int6 } #define OMATCOPY_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, \ - const TYPE alpha, const TYPE *a, int64_t lda, TYPE *b, int64_t ldb, \ - const std::vector &dependencies) { \ + sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, \ + const TYPE alpha, const TYPE* a, int64_t lda, TYPE* b, int64_t ldb, \ + const std::vector& dependencies) { \ return omatcopy(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); \ } @@ -320,16 +320,16 @@ OMATCOPY_LAUNCHER_USM(std::complex, rocblas_zgeam) #undef OMATCOPY_LAUNCHER_USM template -sycl::event omatcopy2(const char *func_name, Func func, sycl::queue &queue, transpose trans, - int64_t m, int64_t n, T alpha, const T *a, int64_t lda, int64_t stridea, T *b, - int64_t ldb, int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(const char* func_name, Func func, sycl::queue& queue, transpose trans, + int64_t m, int64_t n, T alpha, const T* a, int64_t lda, int64_t stridea, T* b, + int64_t ldb, int64_t strideb, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy2", ""); } #define OMATCOPY2_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, int64_t stridea, TYPE *b, int64_t ldb, \ - int64_t strideb, const std::vector &dependencies) { \ + sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, int64_t stridea, TYPE* b, int64_t ldb, \ + int64_t strideb, const std::vector& dependencies) { \ return omatcopy2(#ROCBLAS_ROUTINE, ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, stridea, \ lda, b, ldb, strideb, dependencies); \ } @@ -341,50 +341,50 @@ OMATCOPY2_LAUNCHER_USM(std::complex, "unimplemented") #undef OMATCOPY2_LAUNCHER_USM -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for column_major layout"); } template -inline sycl::event omatadd(Func func, sycl::queue &queue, transpose transa, transpose transb, - int64_t m, int64_t n, const T alpha, const T *a, int64_t lda, - const T beta, const T *b, int64_t ldb, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event omatadd(Func func, sycl::queue& queue, transpose transa, transpose transb, + int64_t m, int64_t n, const T alpha, const T* a, int64_t lda, + const T beta, const T* b, int64_t ldb, T* c, int64_t ldc, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); rocblas_status err; ROCBLAS_ERROR_FUNC_SYNC(func, err, handle, get_rocblas_operation(transa), - get_rocblas_operation(transb), m, n, (rocDataType *)&alpha, a_, - lda, (rocDataType *)&beta, b_, ldb, c_, ldc); + get_rocblas_operation(transb), m, n, (rocDataType*)&alpha, a_, + lda, (rocDataType*)&beta, b_, ldb, c_, ldc); }); }); @@ -392,10 +392,10 @@ inline sycl::event omatadd(Func func, sycl::queue &queue, transpose transa, tran } #define OMATADD_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, const TYPE alpha, const TYPE *a, int64_t lda, const TYPE beta, \ - const TYPE *b, int64_t ldb, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, const TYPE alpha, const TYPE* a, int64_t lda, const TYPE beta, \ + const TYPE* b, int64_t ldb, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return omatadd(ROCBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, \ c, ldc, dependencies); \ } @@ -413,72 +413,72 @@ namespace row_major { // Buffer APIs -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - int8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + int8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -void gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, int64_t m, - int64_t n, int64_t k, float alpha, sycl::buffer &a, int64_t lda, - uint8_t ao, sycl::buffer &b, int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, int64_t ldc, sycl::buffer &co) { +void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, int64_t m, + int64_t n, int64_t k, float alpha, sycl::buffer& a, int64_t lda, + uint8_t ao, sycl::buffer& b, int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, int64_t ldc, sycl::buffer& co) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, float alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, float beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, float alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, float beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, double alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, double beta, sycl::buffer &c, +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, double alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, double beta, sycl::buffer& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -void gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, - int64_t k, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &b, int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, int64_t ldc) { +void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, int64_t n, + int64_t k, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& b, int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, int64_t ldc) { throw unimplemented("blas", "gemmt", "for row_major layout"); } template -inline void omatcopy(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - const T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, +inline void omatcopy(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + const T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { column_major::omatcopy(func, queue, trans, n, m, alpha, a, lda, b, ldb); } #define OMATCOPY_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, const TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, const TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& b, int64_t ldb) { \ omatcopy(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, b, ldb); \ } @@ -490,16 +490,16 @@ OMATCOPY_LAUNCHER(std::complex, rocblas_zgeam) #undef OMATCOPY_LAUNCHER template -void omatcopy2(const char *func_name, Func func, sycl::queue &queue, transpose trans, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, std::int64_t stridea, - sycl::buffer &b, int64_t ldb, std::int64_t strideb) { +void omatcopy2(const char* func_name, Func func, sycl::queue& queue, transpose trans, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, std::int64_t stridea, + sycl::buffer& b, int64_t ldb, std::int64_t strideb) { throw unimplemented("blas", "omatcopy2", ""); } #define OMATCOPY2_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, int64_t stridea, \ - sycl::buffer &b, int64_t ldb, int64_t strideb) { \ + void omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, int64_t stridea, \ + sycl::buffer& b, int64_t ldb, int64_t strideb) { \ omatcopy2(#ROCBLAS_ROUTINE, ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, stridea, lda, \ b, ldb, strideb); \ } @@ -511,37 +511,37 @@ OMATCOPY2_LAUNCHER(std::complex, "unimplemented") #undef OMATCOPY2_LAUNCHER -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - sycl::buffer &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + sycl::buffer& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -void imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, int64_t lda, int64_t ldb) { +void imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, int64_t lda, int64_t ldb) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } template -inline void omatadd(Func func, sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, const T alpha, sycl::buffer &a, int64_t lda, const T beta, - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { +inline void omatadd(Func func, sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, const T alpha, sycl::buffer& a, int64_t lda, const T beta, + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { column_major::omatadd(func, queue, transa, transb, n, m, alpha, a, lda, beta, b, ldb, c, ldc); } #define OMATADD_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - const TYPE alpha, sycl::buffer &a, int64_t lda, const TYPE beta, \ - sycl::buffer &b, int64_t ldb, sycl::buffer &c, int64_t ldc) { \ + void omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + const TYPE alpha, sycl::buffer& a, int64_t lda, const TYPE beta, \ + sycl::buffer& b, int64_t ldb, sycl::buffer& c, int64_t ldc) { \ omatadd(ROCBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, \ ldc); \ } @@ -555,79 +555,79 @@ OMATADD_LAUNCHER(std::complex, rocblas_zgeam) // USM APIs -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const int8_t *a, int64_t lda, - int8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const int8_t *a, int64_t lda, - int8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const int8_t* a, int64_t lda, + int8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const uint8_t *a, int64_t lda, - uint8_t ao, const int8_t *b, int64_t ldb, int8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const int8_t* b, int64_t ldb, int8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -sycl::event gemm_bias(sycl::queue &queue, transpose transa, transpose transb, offset offsetc, - int64_t m, int64_t n, int64_t k, float alpha, const uint8_t *a, int64_t lda, - uint8_t ao, const uint8_t *b, int64_t ldb, uint8_t bo, float beta, int32_t *c, - int64_t ldc, const int32_t *co, - const std::vector &dependencies) { +sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, + int64_t m, int64_t n, int64_t k, float alpha, const uint8_t* a, int64_t lda, + uint8_t ao, const uint8_t* b, int64_t ldb, uint8_t bo, float beta, int32_t* c, + int64_t ldc, const int32_t* co, + const std::vector& dependencies) { throw unimplemented("blas", "gemm_bias", "for row_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, float alpha, const float *a, int64_t lda, const float *b, - int64_t ldb, float beta, float *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, float alpha, const float* a, int64_t lda, const float* b, + int64_t ldb, float beta, float* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, double alpha, const double *a, int64_t lda, const double *b, - int64_t ldb, double beta, double *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for row_major layout"); } -sycl::event gemmt(sycl::queue &queue, uplo upper_lower, transpose transa, transpose transb, - int64_t n, int64_t k, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *b, int64_t ldb, - std::complex beta, std::complex *c, int64_t ldc, - const std::vector &dependencies) { +sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, + std::complex beta, std::complex* c, int64_t ldc, + const std::vector& dependencies) { throw unimplemented("blas", "gemmt", "for row_major layout"); } template -inline sycl::event omatcopy(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - const T alpha, const T *a, int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { +inline sycl::event omatcopy(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + const T alpha, const T* a, int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { return column_major::omatcopy(func, queue, trans, n, m, alpha, a, lda, b, ldb, dependencies); } #define OMATCOPY_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, \ - const TYPE alpha, const TYPE *a, int64_t lda, TYPE *b, int64_t ldb, \ - const std::vector &dependencies) { \ + sycl::event omatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, \ + const TYPE alpha, const TYPE* a, int64_t lda, TYPE* b, int64_t ldb, \ + const std::vector& dependencies) { \ return omatcopy(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); \ } @@ -639,16 +639,16 @@ OMATCOPY_LAUNCHER_USM(std::complex, rocblas_zgeam) #undef OMATCOPY_LAUNCHER_USM template -sycl::event omatcopy2(const char *func_name, Func func, sycl::queue &queue, transpose trans, - int64_t m, int64_t n, T alpha, const T *a, int64_t lda, int64_t stridea, T *b, - int64_t ldb, int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(const char* func_name, Func func, sycl::queue& queue, transpose trans, + int64_t m, int64_t n, T alpha, const T* a, int64_t lda, int64_t stridea, T* b, + int64_t ldb, int64_t strideb, const std::vector& dependencies) { throw unimplemented("blas", "omatcopy2", ""); } #define OMATCOPY2_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatcopy2(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, int64_t stridea, TYPE *b, int64_t ldb, \ - int64_t strideb, const std::vector &dependencies) { \ + sycl::event omatcopy2(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, int64_t stridea, TYPE* b, int64_t ldb, \ + int64_t strideb, const std::vector& dependencies) { \ return omatcopy2(#ROCBLAS_ROUTINE, ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, stridea, \ lda, b, ldb, strideb, dependencies); \ } @@ -660,44 +660,44 @@ OMATCOPY2_LAUNCHER_USM(std::complex, "unimplemented") #undef OMATCOPY2_LAUNCHER_USM -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, float alpha, - float *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, float alpha, + float* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, double alpha, - double *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, double alpha, + double* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } -sycl::event imatcopy(sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, std::complex *ab, int64_t lda, int64_t ldb, - const std::vector &dependencies) { +sycl::event imatcopy(sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, std::complex* ab, int64_t lda, int64_t ldb, + const std::vector& dependencies) { throw unimplemented("blas", "imatcopy", "for row_major layout"); } template -inline sycl::event omatadd(Func func, sycl::queue &queue, transpose transa, transpose transb, - int64_t m, int64_t n, const T alpha, const T *a, int64_t lda, - const T beta, const T *b, int64_t ldb, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event omatadd(Func func, sycl::queue& queue, transpose transa, transpose transb, + int64_t m, int64_t n, const T alpha, const T* a, int64_t lda, + const T beta, const T* b, int64_t ldb, T* c, int64_t ldc, + const std::vector& dependencies) { return column_major::omatadd(func, queue, transa, transb, n, m, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } #define OMATADD_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event omatadd(sycl::queue &queue, transpose transa, transpose transb, int64_t m, \ - int64_t n, const TYPE alpha, const TYPE *a, int64_t lda, const TYPE beta, \ - const TYPE *b, int64_t ldb, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, int64_t m, \ + int64_t n, const TYPE alpha, const TYPE* a, int64_t lda, const TYPE beta, \ + const TYPE* b, int64_t ldb, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return omatadd(ROCBLAS_ROUTINE, queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, \ c, ldc, dependencies); \ } diff --git a/src/blas/backends/rocblas/rocblas_handle.hpp b/src/blas/backends/rocblas/rocblas_handle.hpp index 66d348619..660be6371 100644 --- a/src/blas/backends/rocblas/rocblas_handle.hpp +++ b/src/blas/backends/rocblas/rocblas_handle.hpp @@ -30,10 +30,10 @@ namespace rocblas { template struct rocblas_handle_ { - using handle_container_t = std::unordered_map *>; + using handle_container_t = std::unordered_map*>; handle_container_t rocblas_handle_mapper_{}; ~rocblas_handle_() noexcept(false) { - for (auto &handle_pair : rocblas_handle_mapper_) { + for (auto& handle_pair : rocblas_handle_mapper_) { rocblas_status err; if (handle_pair.second != nullptr) { auto handle = handle_pair.second->exchange(nullptr); diff --git a/src/blas/backends/rocblas/rocblas_helper.hpp b/src/blas/backends/rocblas/rocblas_helper.hpp index f21d63655..5f9b03276 100644 --- a/src/blas/backends/rocblas/rocblas_helper.hpp +++ b/src/blas/backends/rocblas/rocblas_helper.hpp @@ -77,7 +77,7 @@ void overflow_check(Index index, Next... indices) { class rocblas_error : virtual public std::runtime_error { protected: - inline const char *rocblas_error_map(rocblas_status error) { + inline const char* rocblas_error_map(rocblas_status error) { switch (error) { case rocblas_status_success: return "rocblas_status_success"; case rocblas_status_invalid_handle: return "rocblas_status_invalid_handle"; @@ -124,7 +124,7 @@ class rocblas_error : virtual public std::runtime_error { class hip_error : virtual public std::runtime_error { protected: - inline const char *hip_error_map(hipError_t result) { + inline const char* hip_error_map(hipError_t result) { return hipGetErrorName(result); } int error_number; ///< error number @@ -174,12 +174,12 @@ class hip_error : virtual public std::runtime_error { HIP_ERROR_FUNC(hipStreamSynchronize, hip_err, currentStreamId); template -inline void rocblas_native_func(Func func, rocblas_status err, - rocblas_handle handle, Types... args) { +inline void rocblas_native_func(Func func, rocblas_status err, rocblas_handle handle, + Types... args) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - ROCBLAS_ERROR_FUNC(func, err, handle, args...) + ROCBLAS_ERROR_FUNC(func, err, handle, args...) #else - ROCBLAS_ERROR_FUNC_SYNC(func, err, handle, args...) + ROCBLAS_ERROR_FUNC_SYNC(func, err, handle, args...) #endif }; diff --git a/src/blas/backends/rocblas/rocblas_level1.cpp b/src/blas/backends/rocblas/rocblas_level1.cpp index b5966108d..31c96ea82 100644 --- a/src/blas/backends/rocblas/rocblas_level1.cpp +++ b/src/blas/backends/rocblas/rocblas_level1.cpp @@ -34,16 +34,16 @@ namespace column_major { // Buffer APIs template -inline void asum(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - const int64_t incx, sycl::buffer &result) { +inline void asum(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + const int64_t incx, sycl::buffer& result) { using rocDataType1 = typename RocEquivalentType::Type; using rocDataType2 = typename RocEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host // when the data is on buffer, it must be set to @@ -51,8 +51,8 @@ inline void asum(Func func, sycl::queue &queue, int64_t n, sycl::buffer & // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto x_ = sc.get_mem(x_acc); - auto res_ = sc.get_mem(res_acc); + auto x_ = sc.get_mem(x_acc); + auto res_ = sc.get_mem(res_acc); rocblas_status err; // ASUM does not support negative index rocblas_native_func(func, err, handle, n, x_, std::abs(incx), res_); @@ -65,8 +65,8 @@ inline void asum(Func func, sycl::queue &queue, int64_t n, sycl::buffer & } #define ASUM_LAUNCHER(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - void asum(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void asum(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ asum(ROCBLAS_ROUTINE, queue, n, x, incx, result); \ } @@ -78,26 +78,26 @@ ASUM_LAUNCHER(std::complex, double, rocblas_dzasum) #undef ASUM_LAUNCHER template -inline void scal(Func func, sycl::queue &queue, int64_t n, T1 a, sycl::buffer &x, +inline void scal(Func func, sycl::queue& queue, int64_t n, T1 a, sycl::buffer& x, int64_t incx) { using rocDataType1 = typename RocEquivalentType::Type; using rocDataType2 = typename RocEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = sc.get_mem(x_acc); + auto x_ = sc.get_mem(x_acc); rocblas_status err; // SCAL does not support negative incx - rocblas_native_func(func, err, handle, n, (rocDataType1 *)&a, x_, std::abs(incx)); + rocblas_native_func(func, err, handle, n, (rocDataType1*)&a, x_, std::abs(incx)); }); }); } #define SCAL_LAUNCHER(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - void scal(sycl::queue &queue, int64_t n, TYPE1 a, sycl::buffer &x, int64_t incx) { \ + void scal(sycl::queue& queue, int64_t n, TYPE1 a, sycl::buffer& x, int64_t incx) { \ scal(ROCBLAS_ROUTINE, queue, n, a, x, incx); \ } @@ -111,29 +111,28 @@ SCAL_LAUNCHER(double, std::complex, rocblas_zdscal) #undef SCAL_LAUNCHER template -inline void axpy(Func func, sycl::queue &queue, int64_t n, T alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy) { +inline void axpy(Func func, sycl::queue& queue, int64_t n, T alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; - rocblas_native_func(func, err, handle, n, (rocDataType *)&alpha, x_, incx, y_, - incy); + rocblas_native_func(func, err, handle, n, (rocDataType*)&alpha, x_, incx, y_, incy); }); }); } #define AXPY_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void axpy(sycl::queue &queue, int64_t n, TYPE alpha, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void axpy(sycl::queue& queue, int64_t n, TYPE alpha, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ axpy(ROCBLAS_ROUTINE, queue, n, alpha, x, incx, y, incy); \ } @@ -144,40 +143,40 @@ AXPY_LAUNCHER(std::complex, rocblas_zaxpy) #undef AXPY_LAUNCHER -void axpby(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - float beta, sycl::buffer &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + float beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "axpby", "for column_major layout"); } -void axpby(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx, - double beta, sycl::buffer &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, int64_t incx, + double beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "axpby", "for column_major layout"); } -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { throw unimplemented("blas", "axpby", "for column_major layout"); } -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { throw unimplemented("blas", "axpby", "for column_major layout"); } template -inline void rotg(Func func, sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s) { +inline void rotg(Func func, sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s) { using rocDataType1 = typename RocEquivalentType::Type; using rocDataType2 = typename RocEquivalentType::Type; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto s_acc = s.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host // when the data is on buffer, it must be set to @@ -185,10 +184,10 @@ inline void rotg(Func func, sycl::queue &queue, sycl::buffer &a, sycl::bu // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); - auto s_ = sc.get_mem(s_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); + auto s_ = sc.get_mem(s_acc); rocblas_status err; rocblas_native_func(func, err, handle, a_, b_, c_, s_); // Higher level BLAS functions expect rocblas_pointer_mode_host @@ -200,8 +199,8 @@ inline void rotg(Func func, sycl::queue &queue, sycl::buffer &a, sycl::bu } #define ROTG_LAUNCHER(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, \ - sycl::buffer &c, sycl::buffer &s) { \ + void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, \ + sycl::buffer& c, sycl::buffer& s) { \ rotg(ROCBLAS_ROUTINE, queue, a, b, c, s); \ } @@ -213,16 +212,16 @@ ROTG_LAUNCHER(std::complex, double, rocblas_zrotg) #undef ROTG_LAUNCHER template -inline void rotm(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { +inline void rotm(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& param) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); auto param_acc = param.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -231,9 +230,9 @@ inline void rotm(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); - auto param_ = sc.get_mem(param_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); + auto param_ = sc.get_mem(param_acc); rocblas_status err; rocblas_native_func(func, err, handle, n, x_, incx, y_, incy, param_); // Higher level BLAS functions expect rocblas_pointer_mode_host @@ -245,8 +244,8 @@ inline void rotm(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x } #define ROTM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void rotm(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { \ + void rotm(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy, sycl::buffer& param) { \ rotm(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, param); \ } @@ -256,19 +255,19 @@ ROTM_LAUNCHER(double, rocblas_drotm) #undef ROTM_LAUNCHER template -inline void copy(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy) { +inline void copy(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, n, x_, incx, y_, incy); }); @@ -276,8 +275,8 @@ inline void copy(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x } #define COPY_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void copy(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void copy(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ copy(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy); \ } @@ -289,16 +288,16 @@ COPY_LAUNCHER(std::complex, rocblas_zcopy) #undef COPY_LAUNCHER template -inline void dot(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { +inline void dot(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -307,9 +306,9 @@ inline void dot(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); - auto res_ = sc.get_mem(res_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); + auto res_ = sc.get_mem(res_acc); rocblas_status err; rocblas_native_func(func, err, handle, n, x_, incx, y_, incy, res_); // Higher level BLAS functions expect rocblas_pointer_mode_host @@ -321,8 +320,8 @@ inline void dot(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, } #define DOT_LAUNCHER(EXT, TYPE, ROCBLAS_ROUTINE) \ - void dot##EXT(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &y, const int64_t incy, sycl::buffer &result) { \ + void dot##EXT(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& y, const int64_t incy, sycl::buffer& result) { \ dot(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, result); \ } @@ -335,23 +334,23 @@ DOT_LAUNCHER(c, std::complex, rocblas_zdotc) #undef DOT_LAUNCHER -void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { +void dot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { throw unimplemented("blas", "dot", "for column_major layout"); } template -inline void rot(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - const int64_t incx, sycl::buffer &y, int64_t incy, T2 c, T3 s) { +inline void rot(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + const int64_t incx, sycl::buffer& y, int64_t incy, T2 c, T3 s) { using rocDataType1 = typename RocEquivalentType::Type; using rocDataType2 = typename RocEquivalentType::Type; using rocDataType3 = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host // when the data is on buffer, it must be set to @@ -359,18 +358,18 @@ inline void rot(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. // rocblas_set_pointer_mode(handle, rocblas_set_pointer_mode); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; - rocblas_native_func(func, err, handle, n, x_, incx, y_, incy, (rocDataType2 *)&c, - (rocDataType3 *)&s); + rocblas_native_func(func, err, handle, n, x_, incx, y_, incy, (rocDataType2*)&c, + (rocDataType3*)&s); }); }); } #define ROT_LAUNCHER(TYPE1, TYPE2, TYPE3, ROCBLAS_ROUTINE) \ - void rot(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &y, int64_t incy, TYPE2 c, TYPE3 s) { \ + void rot(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& y, int64_t incy, TYPE2 c, TYPE3 s) { \ rot(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, c, s); \ } @@ -381,16 +380,16 @@ ROT_LAUNCHER(std::complex, double, double, rocblas_zdrot) #undef ROT_LAUNCHER -void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { +void sdsdot(sycl::queue& queue, int64_t n, float sb, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { overflow_check(n, incx, incy); // rocBLAS does not support sdot so we need to mimic sdot. - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.get_access(cgh); auto y_acc = y.get_access(cgh); auto res_acc = result.get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -399,9 +398,9 @@ void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); - auto res_ = sc.get_mem(res_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); + auto res_ = sc.get_mem(res_acc); rocblas_status err; rocblas_native_func(rocblas_sdot, err, handle, n, x_, incx, y_, incy, res_); // Higher level BLAS functions expect rocblas_pointer_mode_host @@ -417,18 +416,18 @@ void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, } template -inline void rotmg(Func func, sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, T y1, sycl::buffer ¶m) { +inline void rotmg(Func func, sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, T y1, sycl::buffer& param) { using rocDataType = typename RocEquivalentType::Type; sycl::buffer y1_buff(&y1, sycl::range<1>(1)); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto d1_acc = d1.template get_access(cgh); auto d2_acc = d2.template get_access(cgh); auto x1_acc = x1.template get_access(cgh); auto y1_acc = y1_buff.template get_access(cgh); auto param_acc = param.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -437,11 +436,11 @@ inline void rotmg(Func func, sycl::queue &queue, sycl::buffer &d1, sycl::b // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto d1_ = sc.get_mem(d1_acc); - auto d2_ = sc.get_mem(d2_acc); - auto x1_ = sc.get_mem(x1_acc); - auto y1_ = sc.get_mem(y1_acc); - auto param_ = sc.get_mem(param_acc); + auto d1_ = sc.get_mem(d1_acc); + auto d2_ = sc.get_mem(d2_acc); + auto x1_ = sc.get_mem(x1_acc); + auto y1_ = sc.get_mem(y1_acc); + auto param_ = sc.get_mem(param_acc); rocblas_status err; rocblas_native_func(func, err, handle, d1_, d2_, x1_, y1_, param_); // Higher level BLAS functions expect rocblas_pointer_mode_host @@ -453,8 +452,8 @@ inline void rotmg(Func func, sycl::queue &queue, sycl::buffer &d1, sycl::b } #define ROTMG_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, \ - sycl::buffer &x1, TYPE y1, sycl::buffer ¶m) { \ + void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, \ + sycl::buffer& x1, TYPE y1, sycl::buffer& param) { \ rotmg(ROCBLAS_ROUTINE, queue, d1, d2, x1, y1, param); \ } @@ -464,8 +463,8 @@ ROTMG_LAUNCHER(double, rocblas_drotmg) #undef ROTMG_LAUNCHER template -inline void iamax(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - const int64_t incx, sycl::buffer &result) { +inline void iamax(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + const int64_t incx, sycl::buffer& result) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); @@ -477,10 +476,10 @@ inline void iamax(Func func, sycl::queue &queue, int64_t n, sycl::buffer & // to elementwise copy the data between two buffer, or allow reinterpret cast // to convert to different type with different typesize size. sycl::buffer int_res_buff{ sycl::range<1>(1) }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto int_res_acc = int_res_buff.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -489,8 +488,8 @@ inline void iamax(Func func, sycl::queue &queue, int64_t n, sycl::buffer & // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto x_ = sc.get_mem(x_acc); - auto int_res_ = sc.get_mem(int_res_acc); + auto x_ = sc.get_mem(x_acc); + auto int_res_ = sc.get_mem(int_res_acc); rocblas_status err; // For negative incx, iamax returns 0. This behaviour is similar to that of // reference netlib BLAS. @@ -502,7 +501,7 @@ inline void iamax(Func func, sycl::queue &queue, int64_t n, sycl::buffer & }); }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto int_res_acc = int_res_buff.template get_access(cgh); auto result_acc = result.template get_access(cgh); cgh.single_task( @@ -511,8 +510,8 @@ inline void iamax(Func func, sycl::queue &queue, int64_t n, sycl::buffer & } #define IAMAX_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void iamax(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void iamax(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ iamax(ROCBLAS_ROUTINE, queue, n, x, incx, result); \ } @@ -524,19 +523,19 @@ IAMAX_LAUNCHER(std::complex, rocblas_izamax) #undef IAMAX_LAUNCHER template -inline void swap(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy) { +inline void swap(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, n, x_, incx, y_, incy); }); @@ -544,8 +543,8 @@ inline void swap(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x } #define SWAP_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void swap(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void swap(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ swap(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy); \ } @@ -557,8 +556,8 @@ SWAP_LAUNCHER(std::complex, rocblas_zswap) #undef SWAP_LAUNCHER template -inline void iamin(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - const int64_t incx, sycl::buffer &result) { +inline void iamin(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + const int64_t incx, sycl::buffer& result) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); @@ -570,10 +569,10 @@ inline void iamin(Func func, sycl::queue &queue, int64_t n, sycl::buffer & // to elementwise copy the data between two buffer, or allow reinterpret cast // to convert to different type with different typesize size. sycl::buffer int_res_buff{ sycl::range<1>(1) }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto int_res_acc = int_res_buff.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -582,8 +581,8 @@ inline void iamin(Func func, sycl::queue &queue, int64_t n, sycl::buffer & // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto x_ = sc.get_mem(x_acc); - auto int_res_ = sc.get_mem(int_res_acc); + auto x_ = sc.get_mem(x_acc); + auto int_res_ = sc.get_mem(int_res_acc); rocblas_status err; // For negative incx, iamin returns 0. This behaviour is similar to that of // implemented as a reference IAMIN. @@ -595,7 +594,7 @@ inline void iamin(Func func, sycl::queue &queue, int64_t n, sycl::buffer & }); }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto int_res_acc = int_res_buff.template get_access(cgh); auto result_acc = result.template get_access(cgh); cgh.single_task( @@ -604,8 +603,8 @@ inline void iamin(Func func, sycl::queue &queue, int64_t n, sycl::buffer & } #define IAMIN_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void iamin(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void iamin(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ iamin(ROCBLAS_ROUTINE, queue, n, x, incx, result); \ } @@ -617,16 +616,16 @@ IAMIN_LAUNCHER(std::complex, rocblas_izamin) #undef IAMIN_LAUNCHER template -inline void nrm2(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - const int64_t incx, sycl::buffer &result) { +inline void nrm2(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + const int64_t incx, sycl::buffer& result) { using rocDataType1 = typename RocEquivalentType::Type; using rocDataType2 = typename RocEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto x_acc = x.template get_access(cgh); auto res_acc = result.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); // By default the pointer mode is the rocblas_pointer_mode_host @@ -635,8 +634,8 @@ inline void nrm2(Func func, sycl::queue &queue, int64_t n, sycl::buffer & // fault. When it is set to device it is users responsibility to // synchronise as the function is completely asynchronous. rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto x_ = sc.get_mem(x_acc); - auto res_ = sc.get_mem(res_acc); + auto x_ = sc.get_mem(x_acc); + auto res_ = sc.get_mem(res_acc); rocblas_status err; // NRM2 does not support negative index rocblas_native_func(func, err, handle, n, x_, std::abs(incx), res_); @@ -649,8 +648,8 @@ inline void nrm2(Func func, sycl::queue &queue, int64_t n, sycl::buffer & } #define NRM2_LAUNCHER(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - void nrm2(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void nrm2(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ nrm2(ROCBLAS_ROUTINE, queue, n, x, incx, result); \ } @@ -664,20 +663,20 @@ NRM2_LAUNCHER(std::complex, double, rocblas_dznrm2) // USM APIs template -inline sycl::event asum(Func func, sycl::queue &queue, int64_t n, const T1 *x, const int64_t incx, - T2 *result, const std::vector &dependencies) { +inline sycl::event asum(Func func, sycl::queue& queue, int64_t n, const T1* x, const int64_t incx, + T2* result, const std::vector& dependencies) { using rocDataType1 = typename RocEquivalentType::Type; using rocDataType2 = typename RocEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto x_ = reinterpret_cast(x); - auto res_ = reinterpret_cast(result); + auto x_ = reinterpret_cast(x); + auto res_ = reinterpret_cast(result); rocblas_status err; // ASUM does not support negative index rocblas_native_func(func, err, handle, n, x_, std::abs(incx), res_); @@ -689,8 +688,8 @@ inline sycl::event asum(Func func, sycl::queue &queue, int64_t n, const T1 *x, c } #define ASUM_LAUNCHER_USM(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - sycl::event asum(sycl::queue &queue, int64_t n, const TYPE1 *x, const int64_t incx, \ - TYPE2 *result, const std::vector &dependencies) { \ + sycl::event asum(sycl::queue& queue, int64_t n, const TYPE1* x, const int64_t incx, \ + TYPE2* result, const std::vector& dependencies) { \ return asum(ROCBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } @@ -702,21 +701,21 @@ ASUM_LAUNCHER_USM(std::complex, double, rocblas_dzasum) #undef ASUM_LAUNCHER_USM template -inline sycl::event scal(Func func, sycl::queue &queue, int64_t n, T1 a, T2 *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event scal(Func func, sycl::queue& queue, int64_t n, T1 a, T2* x, int64_t incx, + const std::vector& dependencies) { using rocDataType1 = typename RocEquivalentType::Type; using rocDataType2 = typename RocEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); + auto x_ = reinterpret_cast(x); rocblas_status err; // SCAL does not support negative incx - rocblas_native_func(func, err, handle, n, (rocDataType1 *)&a, x_, std::abs(incx)); + rocblas_native_func(func, err, handle, n, (rocDataType1*)&a, x_, std::abs(incx)); }); }); @@ -724,8 +723,8 @@ inline sycl::event scal(Func func, sycl::queue &queue, int64_t n, T1 a, T2 *x, i } #define SCAL_LAUNCHER_USM(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - sycl::event scal(sycl::queue &queue, int64_t n, TYPE1 a, TYPE2 *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event scal(sycl::queue& queue, int64_t n, TYPE1 a, TYPE2* x, int64_t incx, \ + const std::vector& dependencies) { \ return scal(ROCBLAS_ROUTINE, queue, n, a, x, incx, dependencies); \ } @@ -739,21 +738,20 @@ SCAL_LAUNCHER_USM(double, std::complex, rocblas_zdscal) #undef SCAL_LAUNCHER_USM template -inline sycl::event axpy(Func func, sycl::queue &queue, int64_t n, T alpha, const T *x, int64_t incx, - T *y, int64_t incy, const std::vector &dependencies) { +inline sycl::event axpy(Func func, sycl::queue& queue, int64_t n, T alpha, const T* x, int64_t incx, + T* y, int64_t incy, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; - rocblas_native_func(func, err, handle, n, (rocDataType *)&alpha, x_, incx, y_, - incy); + rocblas_native_func(func, err, handle, n, (rocDataType*)&alpha, x_, incx, y_, incy); }); }); @@ -761,8 +759,8 @@ inline sycl::event axpy(Func func, sycl::queue &queue, int64_t n, T alpha, const } #define AXPY_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event axpy(sycl::queue &queue, int64_t n, TYPE alpha, const TYPE *x, int64_t incx, \ - TYPE *y, int64_t incy, const std::vector &dependencies) { \ + sycl::event axpy(sycl::queue& queue, int64_t n, TYPE alpha, const TYPE* x, int64_t incx, \ + TYPE* y, int64_t incy, const std::vector& dependencies) { \ return axpy(ROCBLAS_ROUTINE, queue, n, alpha, x, incx, y, incy, dependencies); \ } @@ -773,44 +771,44 @@ AXPY_LAUNCHER_USM(std::complex, rocblas_zaxpy) #undef AXPY_LAUNCHER_USM -sycl::event axpby(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, - float beta, float *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, + float beta, float* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for column_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - double beta, double *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + double beta, double* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for column_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for column_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for column_major layout"); } template -inline sycl::event rotg(Func func, sycl::queue &queue, T1 *a, T1 *b, T2 *c, T1 *s, - const std::vector &dependencies) { +inline sycl::event rotg(Func func, sycl::queue& queue, T1* a, T1* b, T2* c, T1* s, + const std::vector& dependencies) { using rocDataType1 = typename RocEquivalentType::Type; using rocDataType2 = typename RocEquivalentType::Type; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); - auto s_ = reinterpret_cast(s); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); + auto s_ = reinterpret_cast(s); rocblas_status err; rocblas_native_func(func, err, handle, a_, b_, c_, s_); }); @@ -820,8 +818,8 @@ inline sycl::event rotg(Func func, sycl::queue &queue, T1 *a, T1 *b, T2 *c, T1 * } #define ROTG_LAUNCHER_USM(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - sycl::event rotg(sycl::queue &queue, TYPE1 *a, TYPE1 *b, TYPE2 *c, TYPE1 *s, \ - const std::vector &dependencies) { \ + sycl::event rotg(sycl::queue& queue, TYPE1* a, TYPE1* b, TYPE2* c, TYPE1* s, \ + const std::vector& dependencies) { \ return rotg(ROCBLAS_ROUTINE, queue, a, b, c, s, dependencies); \ } @@ -833,19 +831,19 @@ ROTG_LAUNCHER_USM(std::complex, double, rocblas_zrotg) #undef ROTG_LAUNCHER_USM template -inline sycl::event rotm(Func func, sycl::queue &queue, int64_t n, T *x, int64_t incx, T *y, - int64_t incy, T *param, const std::vector &dependencies) { +inline sycl::event rotm(Func func, sycl::queue& queue, int64_t n, T* x, int64_t incx, T* y, + int64_t incy, T* param, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); - auto param_ = reinterpret_cast(param); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); + auto param_ = reinterpret_cast(param); rocblas_status err; rocblas_native_func(func, err, handle, n, x_, incx, y_, incy, param_); }); @@ -855,8 +853,8 @@ inline sycl::event rotm(Func func, sycl::queue &queue, int64_t n, T *x, int64_t } #define ROTM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event rotm(sycl::queue &queue, int64_t n, TYPE *x, int64_t incx, TYPE *y, int64_t incy, \ - TYPE *param, const std::vector &dependencies) { \ + sycl::event rotm(sycl::queue& queue, int64_t n, TYPE* x, int64_t incx, TYPE* y, int64_t incy, \ + TYPE* param, const std::vector& dependencies) { \ return rotm(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, param, dependencies); \ } @@ -866,18 +864,18 @@ ROTM_LAUNCHER_USM(double, rocblas_drotm) #undef ROTM_LAUNCHER_USM template -inline sycl::event copy(Func func, sycl::queue &queue, int64_t n, const T *x, int64_t incx, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event copy(Func func, sycl::queue& queue, int64_t n, const T* x, int64_t incx, T* y, + int64_t incy, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, n, x_, incx, y_, incy); }); @@ -887,8 +885,8 @@ inline sycl::event copy(Func func, sycl::queue &queue, int64_t n, const T *x, in } #define COPY_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event copy(sycl::queue &queue, int64_t n, const TYPE *x, int64_t incx, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event copy(sycl::queue& queue, int64_t n, const TYPE* x, int64_t incx, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return copy(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, dependencies); \ } @@ -900,20 +898,20 @@ COPY_LAUNCHER_USM(std::complex, rocblas_zcopy) #undef COPY_LAUNCHER_USM template -inline sycl::event dot(Func func, sycl::queue &queue, int64_t n, const T *x, const int64_t incx, - const T *y, int64_t incy, T *result, - const std::vector &dependencies) { +inline sycl::event dot(Func func, sycl::queue& queue, int64_t n, const T* x, const int64_t incx, + const T* y, int64_t incy, T* result, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); - auto res_ = reinterpret_cast(result); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); + auto res_ = reinterpret_cast(result); rocblas_status err; rocblas_native_func(func, err, handle, n, x_, incx, y_, incy, res_); }); @@ -923,9 +921,9 @@ inline sycl::event dot(Func func, sycl::queue &queue, int64_t n, const T *x, con } #define DOT_LAUNCHER_USM(EXT, TYPE, ROCBLAS_ROUTINE) \ - sycl::event dot##EXT(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - const TYPE *y, const int64_t incy, TYPE *result, \ - const std::vector &dependencies) { \ + sycl::event dot##EXT(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + const TYPE* y, const int64_t incy, TYPE* result, \ + const std::vector& dependencies) { \ return dot(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, result, dependencies); \ } @@ -938,29 +936,29 @@ DOT_LAUNCHER_USM(c, std::complex, rocblas_zdotc) #undef DOT_LAUNCHER_USM -sycl::event dot(sycl::queue &queue, int64_t n, const float *x, int64_t incx, const float *y, - int64_t incy, double *result, const std::vector &dependencies) { +sycl::event dot(sycl::queue& queue, int64_t n, const float* x, int64_t incx, const float* y, + int64_t incy, double* result, const std::vector& dependencies) { throw unimplemented("blas", "dot", "for column_major layout"); } template -inline sycl::event rot(Func func, sycl::queue &queue, int64_t n, T1 *x, const int64_t incx, T1 *y, - int64_t incy, T2 c, T3 s, const std::vector &dependencies) { +inline sycl::event rot(Func func, sycl::queue& queue, int64_t n, T1* x, const int64_t incx, T1* y, + int64_t incy, T2 c, T3 s, const std::vector& dependencies) { using rocDataType1 = typename RocEquivalentType::Type; using rocDataType2 = typename RocEquivalentType::Type; using rocDataType3 = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; - rocblas_native_func(func, err, handle, n, x_, incx, y_, incy, (rocDataType2 *)&c, - (rocDataType3 *)&s); + rocblas_native_func(func, err, handle, n, x_, incx, y_, incy, (rocDataType2*)&c, + (rocDataType3*)&s); }); }); @@ -968,9 +966,9 @@ inline sycl::event rot(Func func, sycl::queue &queue, int64_t n, T1 *x, const in } #define ROT_LAUNCHER_USM(TYPE1, TYPE2, TYPE3, ROCBLAS_ROUTINE) \ - sycl::event rot(sycl::queue &queue, int64_t n, TYPE1 *x, const int64_t incx, TYPE1 *y, \ + sycl::event rot(sycl::queue& queue, int64_t n, TYPE1* x, const int64_t incx, TYPE1* y, \ int64_t incy, TYPE2 c, TYPE3 s, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return rot(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, c, s, dependencies); \ } @@ -981,20 +979,20 @@ ROT_LAUNCHER_USM(std::complex, double, double, rocblas_zdrot) #undef ROT_LAUNCHER_USM -sycl::event sdsdot(sycl::queue &queue, int64_t n, float sb, const float *x, int64_t incx, - const float *y, int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event sdsdot(sycl::queue& queue, int64_t n, float sb, const float* x, int64_t incx, + const float* y, int64_t incy, float* result, + const std::vector& dependencies) { overflow_check(n, incx, incy); // rocBLAS does not support sdot so we need to mimic sdot. - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); - auto res_ = reinterpret_cast(result); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); + auto res_ = reinterpret_cast(result); rocblas_status err; rocblas_native_func(rocblas_sdot, err, handle, n, x_, incx, y_, incy, res_); }); @@ -1006,20 +1004,20 @@ sycl::event sdsdot(sycl::queue &queue, int64_t n, float sb, const float *x, int6 } template -inline sycl::event rotmg(Func func, sycl::queue &queue, T *d1, T *d2, T *x1, T y1, T *param, - const std::vector &dependencies) { +inline sycl::event rotmg(Func func, sycl::queue& queue, T* d1, T* d2, T* x1, T y1, T* param, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto d1_ = reinterpret_cast(d1); - auto d2_ = reinterpret_cast(d2); - auto x1_ = reinterpret_cast(x1); - auto y1_ = reinterpret_cast(&y1); - auto param_ = reinterpret_cast(param); + auto d1_ = reinterpret_cast(d1); + auto d2_ = reinterpret_cast(d2); + auto x1_ = reinterpret_cast(x1); + auto y1_ = reinterpret_cast(&y1); + auto param_ = reinterpret_cast(param); rocblas_status err; rocblas_native_func(func, err, handle, d1_, d2_, x1_, y1_, param_); }); @@ -1029,8 +1027,8 @@ inline sycl::event rotmg(Func func, sycl::queue &queue, T *d1, T *d2, T *x1, T y } #define ROTMG_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event rotmg(sycl::queue &queue, TYPE *d1, TYPE *d2, TYPE *x1, TYPE y1, TYPE *param, \ - const std::vector &dependencies) { \ + sycl::event rotmg(sycl::queue& queue, TYPE* d1, TYPE* d2, TYPE* x1, TYPE y1, TYPE* param, \ + const std::vector& dependencies) { \ return rotmg(ROCBLAS_ROUTINE, queue, d1, d2, x1, y1, param, dependencies); \ } @@ -1040,8 +1038,8 @@ ROTMG_LAUNCHER_USM(double, rocblas_drotmg) #undef ROTMG_LAUNCHER_USM template -inline sycl::event iamax(Func func, sycl::queue &queue, int64_t n, const T *x, const int64_t incx, - int64_t *result, const std::vector &dependencies) { +inline sycl::event iamax(Func func, sycl::queue& queue, int64_t n, const T* x, const int64_t incx, + int64_t* result, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); // rocBLAS does not support int64_t as return type for the data by default. So we need to @@ -1049,17 +1047,17 @@ inline sycl::event iamax(Func func, sycl::queue &queue, int64_t n, const T *x, c // it back to the actual data on the host. // This change may cause failure as the result of integer overflow // based on the size. - auto int_res_p = (int *)sycl::aligned_alloc_shared(64, sizeof(rocblas_int), queue.get_device(), - queue.get_context()); + auto int_res_p = (int*)sycl::aligned_alloc_shared(64, sizeof(rocblas_int), queue.get_device(), + queue.get_context()); *int_res_p = 0; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto x_ = reinterpret_cast(x); - auto int_res_p_ = reinterpret_cast(int_res_p); + auto x_ = reinterpret_cast(x); + auto int_res_p_ = reinterpret_cast(int_res_p); rocblas_status err; // For negative incx, iamax returns 0. This behaviour is similar to that of // reference iamax. @@ -1074,8 +1072,8 @@ inline sycl::event iamax(Func func, sycl::queue &queue, int64_t n, const T *x, c } #define IAMAX_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event iamax(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - int64_t *result, const std::vector &dependencies) { \ + sycl::event iamax(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + int64_t* result, const std::vector& dependencies) { \ return iamax(ROCBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } @@ -1087,18 +1085,18 @@ IAMAX_LAUNCHER_USM(std::complex, rocblas_izamax) #undef IAMAX_LAUNCHER_USM template -inline sycl::event swap(Func func, sycl::queue &queue, int64_t n, T *x, int64_t incx, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event swap(Func func, sycl::queue& queue, int64_t n, T* x, int64_t incx, T* y, + int64_t incy, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, n, x_, incx, y_, incy); }); @@ -1108,8 +1106,8 @@ inline sycl::event swap(Func func, sycl::queue &queue, int64_t n, T *x, int64_t } #define SWAP_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event swap(sycl::queue &queue, int64_t n, TYPE *x, int64_t incx, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event swap(sycl::queue& queue, int64_t n, TYPE* x, int64_t incx, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return swap(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, dependencies); \ } @@ -1121,8 +1119,8 @@ SWAP_LAUNCHER_USM(std::complex, rocblas_zswap) #undef SWAP_LAUNCHER_USM template -inline sycl::event iamin(Func func, sycl::queue &queue, int64_t n, const T *x, const int64_t incx, - int64_t *result, const std::vector &dependencies) { +inline sycl::event iamin(Func func, sycl::queue& queue, int64_t n, const T* x, const int64_t incx, + int64_t* result, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); // rocBLAS does not support int64_t as return type for the data by default. So we need to @@ -1130,18 +1128,18 @@ inline sycl::event iamin(Func func, sycl::queue &queue, int64_t n, const T *x, c // it back to the actual data on the host. // This change may cause failure as the result of integer overflow // based on the size. - auto int_res_p = (int *)sycl::aligned_alloc_shared(64, sizeof(rocblas_int), queue.get_device(), - queue.get_context()); + auto int_res_p = (int*)sycl::aligned_alloc_shared(64, sizeof(rocblas_int), queue.get_device(), + queue.get_context()); *int_res_p = 0; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto x_ = reinterpret_cast(x); - auto int_res_p_ = reinterpret_cast(int_res_p); + auto x_ = reinterpret_cast(x); + auto int_res_p_ = reinterpret_cast(int_res_p); rocblas_status err; // For negative incx, iamin returns 0. This behaviour is similar to that of // implemented iamin. @@ -1156,8 +1154,8 @@ inline sycl::event iamin(Func func, sycl::queue &queue, int64_t n, const T *x, c } #define IAMIN_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event iamin(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - int64_t *result, const std::vector &dependencies) { \ + sycl::event iamin(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + int64_t* result, const std::vector& dependencies) { \ return iamin(ROCBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } @@ -1169,20 +1167,20 @@ IAMIN_LAUNCHER_USM(std::complex, rocblas_izamin) #undef IAMIN_LAUNCHER_USM template -inline sycl::event nrm2(Func func, sycl::queue &queue, int64_t n, const T1 *x, const int64_t incx, - T2 *result, const std::vector &dependencies) { +inline sycl::event nrm2(Func func, sycl::queue& queue, int64_t n, const T1* x, const int64_t incx, + T2* result, const std::vector& dependencies) { using rocDataType1 = typename RocEquivalentType::Type; using rocDataType2 = typename RocEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); rocblas_set_pointer_mode(handle, rocblas_pointer_mode_device); - auto x_ = reinterpret_cast(x); - auto res_ = reinterpret_cast(result); + auto x_ = reinterpret_cast(x); + auto res_ = reinterpret_cast(result); rocblas_status err; // NRM2 does not support negative index rocblas_native_func(func, err, handle, n, x_, std::abs(incx), res_); @@ -1194,8 +1192,8 @@ inline sycl::event nrm2(Func func, sycl::queue &queue, int64_t n, const T1 *x, c } #define NRM2_LAUNCHER_USM(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - sycl::event nrm2(sycl::queue &queue, int64_t n, const TYPE1 *x, const int64_t incx, \ - TYPE2 *result, const std::vector &dependencies) { \ + sycl::event nrm2(sycl::queue& queue, int64_t n, const TYPE1* x, const int64_t incx, \ + TYPE2* result, const std::vector& dependencies) { \ return nrm2(ROCBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } @@ -1212,14 +1210,14 @@ namespace row_major { // Buffer APIs template -inline void asum(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - const int64_t incx, sycl::buffer &result) { +inline void asum(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + const int64_t incx, sycl::buffer& result) { column_major::asum(func, queue, n, x, incx, result); } #define ASUM_LAUNCHER(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - void asum(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void asum(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ asum(ROCBLAS_ROUTINE, queue, n, x, incx, result); \ } @@ -1231,13 +1229,13 @@ ASUM_LAUNCHER(std::complex, double, rocblas_dzasum) #undef ASUM_LAUNCHER template -inline void scal(Func func, sycl::queue &queue, int64_t n, T1 a, sycl::buffer &x, +inline void scal(Func func, sycl::queue& queue, int64_t n, T1 a, sycl::buffer& x, int64_t incx) { column_major::scal(func, queue, n, a, x, incx); } #define SCAL_LAUNCHER(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - void scal(sycl::queue &queue, int64_t n, TYPE1 a, sycl::buffer &x, int64_t incx) { \ + void scal(sycl::queue& queue, int64_t n, TYPE1 a, sycl::buffer& x, int64_t incx) { \ scal(ROCBLAS_ROUTINE, queue, n, a, x, incx); \ } @@ -1251,14 +1249,14 @@ SCAL_LAUNCHER(double, std::complex, rocblas_zdscal) #undef SCAL_LAUNCHER template -inline void axpy(Func func, sycl::queue &queue, int64_t n, T alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy) { +inline void axpy(Func func, sycl::queue& queue, int64_t n, T alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy) { column_major::axpy(func, queue, n, alpha, x, incx, y, incy); } #define AXPY_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void axpy(sycl::queue &queue, int64_t n, TYPE alpha, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void axpy(sycl::queue& queue, int64_t n, TYPE alpha, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ axpy(ROCBLAS_ROUTINE, queue, n, alpha, x, incx, y, incy); \ } @@ -1269,37 +1267,37 @@ AXPY_LAUNCHER(std::complex, rocblas_zaxpy) #undef AXPY_LAUNCHER -void axpby(sycl::queue &queue, int64_t n, float alpha, sycl::buffer &x, int64_t incx, - float beta, sycl::buffer &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, float alpha, sycl::buffer& x, int64_t incx, + float beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "axpby", "for row_major layout"); } -void axpby(sycl::queue &queue, int64_t n, double alpha, sycl::buffer &x, int64_t incx, - double beta, sycl::buffer &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, double alpha, sycl::buffer& x, int64_t incx, + double beta, sycl::buffer& y, int64_t incy) { throw unimplemented("blas", "axpby", "for row_major layout"); } -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { throw unimplemented("blas", "axpby", "for row_major layout"); } -void axpby(sycl::queue &queue, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { +void axpby(sycl::queue& queue, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { throw unimplemented("blas", "axpby", "for row_major layout"); } template -inline void rotg(Func func, sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, - sycl::buffer &c, sycl::buffer &s) { +inline void rotg(Func func, sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, + sycl::buffer& c, sycl::buffer& s) { column_major::rotg(func, queue, a, b, c, s); } #define ROTG_LAUNCHER(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - void rotg(sycl::queue &queue, sycl::buffer &a, sycl::buffer &b, \ - sycl::buffer &c, sycl::buffer &s) { \ + void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, \ + sycl::buffer& c, sycl::buffer& s) { \ rotg(ROCBLAS_ROUTINE, queue, a, b, c, s); \ } @@ -1311,14 +1309,14 @@ ROTG_LAUNCHER(std::complex, double, rocblas_zrotg) #undef ROTG_LAUNCHER template -inline void rotm(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { +inline void rotm(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& param) { column_major::rotm(func, queue, n, x, incx, y, incy, param); } #define ROTM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void rotm(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy, sycl::buffer ¶m) { \ + void rotm(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy, sycl::buffer& param) { \ rotm(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, param); \ } @@ -1328,14 +1326,14 @@ ROTM_LAUNCHER(double, rocblas_drotm) #undef ROTM_LAUNCHER template -inline void copy(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy) { +inline void copy(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy) { column_major::copy(func, queue, n, x, incx, y, incy); } #define COPY_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void copy(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void copy(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ copy(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy); \ } @@ -1347,14 +1345,14 @@ COPY_LAUNCHER(std::complex, rocblas_zcopy) #undef COPY_LAUNCHER template -inline void dot(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { +inline void dot(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { column_major::dot(func, queue, n, x, incx, y, incy, result); } #define DOT_LAUNCHER(EXT, TYPE, ROCBLAS_ROUTINE) \ - void dot##EXT(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &y, const int64_t incy, sycl::buffer &result) { \ + void dot##EXT(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& y, const int64_t incy, sycl::buffer& result) { \ dot(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, result); \ } @@ -1367,20 +1365,20 @@ DOT_LAUNCHER(c, std::complex, rocblas_zdotc) #undef DOT_LAUNCHER -void dot(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { +void dot(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { throw unimplemented("blas", "dot", "for row_major layout"); } template -inline void rot(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - const int64_t incx, sycl::buffer &y, int64_t incy, T2 c, T3 s) { +inline void rot(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + const int64_t incx, sycl::buffer& y, int64_t incy, T2 c, T3 s) { column_major::rot(func, queue, n, x, incx, y, incy, c, s); } #define ROT_LAUNCHER(TYPE1, TYPE2, TYPE3, ROCBLAS_ROUTINE) \ - void rot(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &y, int64_t incy, TYPE2 c, TYPE3 s) { \ + void rot(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& y, int64_t incy, TYPE2 c, TYPE3 s) { \ rot(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, c, s); \ } @@ -1391,20 +1389,20 @@ ROT_LAUNCHER(std::complex, double, double, rocblas_zdrot) #undef ROT_LAUNCHER -void sdsdot(sycl::queue &queue, int64_t n, float sb, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy, sycl::buffer &result) { +void sdsdot(sycl::queue& queue, int64_t n, float sb, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy, sycl::buffer& result) { column_major::sdsdot(queue, n, sb, x, incx, y, incy, result); } template -inline void rotmg(Func func, sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, - sycl::buffer &x1, T y1, sycl::buffer ¶m) { +inline void rotmg(Func func, sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, + sycl::buffer& x1, T y1, sycl::buffer& param) { column_major::rotmg(func, queue, d1, d2, x1, y1, param); } #define ROTMG_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void rotmg(sycl::queue &queue, sycl::buffer &d1, sycl::buffer &d2, \ - sycl::buffer &x1, TYPE y1, sycl::buffer ¶m) { \ + void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, \ + sycl::buffer& x1, TYPE y1, sycl::buffer& param) { \ rotmg(ROCBLAS_ROUTINE, queue, d1, d2, x1, y1, param); \ } @@ -1414,14 +1412,14 @@ ROTMG_LAUNCHER(double, rocblas_drotmg) #undef ROTMG_LAUNCHER template -inline void iamax(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - const int64_t incx, sycl::buffer &result) { +inline void iamax(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + const int64_t incx, sycl::buffer& result) { column_major::iamax(func, queue, n, x, incx, result); } #define IAMAX_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void iamax(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void iamax(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ iamax(ROCBLAS_ROUTINE, queue, n, x, incx, result); \ } @@ -1433,14 +1431,14 @@ IAMAX_LAUNCHER(std::complex, rocblas_izamax) #undef IAMAX_LAUNCHER template -inline void swap(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, - sycl::buffer &y, int64_t incy) { +inline void swap(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, + sycl::buffer& y, int64_t incy) { column_major::swap(func, queue, n, x, incx, y, incy); } #define SWAP_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void swap(sycl::queue &queue, int64_t n, sycl::buffer &x, int64_t incx, \ - sycl::buffer &y, int64_t incy) { \ + void swap(sycl::queue& queue, int64_t n, sycl::buffer& x, int64_t incx, \ + sycl::buffer& y, int64_t incy) { \ swap(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy); \ } @@ -1452,14 +1450,14 @@ SWAP_LAUNCHER(std::complex, rocblas_zswap) #undef SWAP_LAUNCHER template -inline void iamin(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - const int64_t incx, sycl::buffer &result) { +inline void iamin(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + const int64_t incx, sycl::buffer& result) { column_major::iamin(func, queue, n, x, incx, result); } #define IAMIN_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void iamin(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void iamin(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ iamin(ROCBLAS_ROUTINE, queue, n, x, incx, result); \ } @@ -1471,14 +1469,14 @@ IAMIN_LAUNCHER(std::complex, rocblas_izamin) #undef IAMIN_LAUNCHER template -inline void nrm2(Func func, sycl::queue &queue, int64_t n, sycl::buffer &x, - const int64_t incx, sycl::buffer &result) { +inline void nrm2(Func func, sycl::queue& queue, int64_t n, sycl::buffer& x, + const int64_t incx, sycl::buffer& result) { column_major::nrm2(func, queue, n, x, incx, result); } #define NRM2_LAUNCHER(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - void nrm2(sycl::queue &queue, int64_t n, sycl::buffer &x, const int64_t incx, \ - sycl::buffer &result) { \ + void nrm2(sycl::queue& queue, int64_t n, sycl::buffer& x, const int64_t incx, \ + sycl::buffer& result) { \ nrm2(ROCBLAS_ROUTINE, queue, n, x, incx, result); \ } @@ -1492,14 +1490,14 @@ NRM2_LAUNCHER(std::complex, double, rocblas_dznrm2) // USM APIs template -inline sycl::event asum(Func func, sycl::queue &queue, int64_t n, const T1 *x, const int64_t incx, - T2 *result, const std::vector &dependencies) { +inline sycl::event asum(Func func, sycl::queue& queue, int64_t n, const T1* x, const int64_t incx, + T2* result, const std::vector& dependencies) { return column_major::asum(func, queue, n, x, incx, result, dependencies); } #define ASUM_LAUNCHER_USM(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - sycl::event asum(sycl::queue &queue, int64_t n, const TYPE1 *x, const int64_t incx, \ - TYPE2 *result, const std::vector &dependencies) { \ + sycl::event asum(sycl::queue& queue, int64_t n, const TYPE1* x, const int64_t incx, \ + TYPE2* result, const std::vector& dependencies) { \ return asum(ROCBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } @@ -1511,14 +1509,14 @@ ASUM_LAUNCHER_USM(std::complex, double, rocblas_dzasum) #undef ASUM_LAUNCHER_USM template -inline sycl::event scal(Func func, sycl::queue &queue, int64_t n, T1 a, T2 *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event scal(Func func, sycl::queue& queue, int64_t n, T1 a, T2* x, int64_t incx, + const std::vector& dependencies) { return column_major::scal(func, queue, n, a, x, incx, dependencies); } #define SCAL_LAUNCHER_USM(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - sycl::event scal(sycl::queue &queue, int64_t n, TYPE1 a, TYPE2 *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event scal(sycl::queue& queue, int64_t n, TYPE1 a, TYPE2* x, int64_t incx, \ + const std::vector& dependencies) { \ return scal(ROCBLAS_ROUTINE, queue, n, a, x, incx, dependencies); \ } @@ -1532,14 +1530,14 @@ SCAL_LAUNCHER_USM(double, std::complex, rocblas_zdscal) #undef SCAL_LAUNCHER_USM template -inline sycl::event axpy(Func func, sycl::queue &queue, int64_t n, T alpha, const T *x, int64_t incx, - T *y, int64_t incy, const std::vector &dependencies) { +inline sycl::event axpy(Func func, sycl::queue& queue, int64_t n, T alpha, const T* x, int64_t incx, + T* y, int64_t incy, const std::vector& dependencies) { return column_major::axpy(func, queue, n, alpha, x, incx, y, incy, dependencies); } #define AXPY_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event axpy(sycl::queue &queue, int64_t n, TYPE alpha, const TYPE *x, int64_t incx, \ - TYPE *y, int64_t incy, const std::vector &dependencies) { \ + sycl::event axpy(sycl::queue& queue, int64_t n, TYPE alpha, const TYPE* x, int64_t incx, \ + TYPE* y, int64_t incy, const std::vector& dependencies) { \ return axpy(ROCBLAS_ROUTINE, queue, n, alpha, x, incx, y, incy, dependencies); \ } @@ -1550,38 +1548,38 @@ AXPY_LAUNCHER_USM(std::complex, rocblas_zaxpy) #undef AXPY_LAUNCHER_USM -sycl::event axpby(sycl::queue &queue, int64_t n, float alpha, const float *x, int64_t incx, - float beta, float *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, float alpha, const float* x, int64_t incx, + float beta, float* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for row_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, double alpha, const double *x, int64_t incx, - double beta, double *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, double alpha, const double* x, int64_t incx, + double beta, double* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for row_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for row_major layout"); } -sycl::event axpby(sycl::queue &queue, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(sycl::queue& queue, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { throw unimplemented("blas", "axpby", "for row_major layout"); } template -inline sycl::event rotg(Func func, sycl::queue &queue, T1 *a, T1 *b, T2 *c, T1 *s, - const std::vector &dependencies) { +inline sycl::event rotg(Func func, sycl::queue& queue, T1* a, T1* b, T2* c, T1* s, + const std::vector& dependencies) { return column_major::rotg(func, queue, a, b, c, s, dependencies); } #define ROTG_LAUNCHER_USM(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - sycl::event rotg(sycl::queue &queue, TYPE1 *a, TYPE1 *b, TYPE2 *c, TYPE1 *s, \ - const std::vector &dependencies) { \ + sycl::event rotg(sycl::queue& queue, TYPE1* a, TYPE1* b, TYPE2* c, TYPE1* s, \ + const std::vector& dependencies) { \ return rotg(ROCBLAS_ROUTINE, queue, a, b, c, s, dependencies); \ } @@ -1593,14 +1591,14 @@ ROTG_LAUNCHER_USM(std::complex, double, rocblas_zrotg) #undef ROTG_LAUNCHER_USM template -inline sycl::event rotm(Func func, sycl::queue &queue, int64_t n, T *x, int64_t incx, T *y, - int64_t incy, T *param, const std::vector &dependencies) { +inline sycl::event rotm(Func func, sycl::queue& queue, int64_t n, T* x, int64_t incx, T* y, + int64_t incy, T* param, const std::vector& dependencies) { return column_major::rotm(func, queue, n, x, incx, y, incy, param, dependencies); } #define ROTM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event rotm(sycl::queue &queue, int64_t n, TYPE *x, int64_t incx, TYPE *y, int64_t incy, \ - TYPE *param, const std::vector &dependencies) { \ + sycl::event rotm(sycl::queue& queue, int64_t n, TYPE* x, int64_t incx, TYPE* y, int64_t incy, \ + TYPE* param, const std::vector& dependencies) { \ return rotm(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, param, dependencies); \ } @@ -1610,14 +1608,14 @@ ROTM_LAUNCHER_USM(double, rocblas_drotm) #undef ROTM_LAUNCHER_USM template -inline sycl::event copy(Func func, sycl::queue &queue, int64_t n, const T *x, int64_t incx, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event copy(Func func, sycl::queue& queue, int64_t n, const T* x, int64_t incx, T* y, + int64_t incy, const std::vector& dependencies) { return column_major::copy(func, queue, n, x, incx, y, incy, dependencies); } #define COPY_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event copy(sycl::queue &queue, int64_t n, const TYPE *x, int64_t incx, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event copy(sycl::queue& queue, int64_t n, const TYPE* x, int64_t incx, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return copy(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, dependencies); \ } @@ -1629,16 +1627,16 @@ COPY_LAUNCHER_USM(std::complex, rocblas_zcopy) #undef COPY_LAUNCHER_USM template -inline sycl::event dot(Func func, sycl::queue &queue, int64_t n, const T *x, const int64_t incx, - const T *y, int64_t incy, T *result, - const std::vector &dependencies) { +inline sycl::event dot(Func func, sycl::queue& queue, int64_t n, const T* x, const int64_t incx, + const T* y, int64_t incy, T* result, + const std::vector& dependencies) { return column_major::dot(func, queue, n, x, incx, y, incy, result, dependencies); } #define DOT_LAUNCHER_USM(EXT, TYPE, ROCBLAS_ROUTINE) \ - sycl::event dot##EXT(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - const TYPE *y, const int64_t incy, TYPE *result, \ - const std::vector &dependencies) { \ + sycl::event dot##EXT(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + const TYPE* y, const int64_t incy, TYPE* result, \ + const std::vector& dependencies) { \ return dot(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, result, dependencies); \ } @@ -1651,21 +1649,21 @@ DOT_LAUNCHER_USM(c, std::complex, rocblas_zdotc) #undef DOT_LAUNCHER_USM -sycl::event dot(sycl::queue &queue, int64_t n, const float *x, int64_t incx, const float *y, - int64_t incy, double *result, const std::vector &dependencies) { +sycl::event dot(sycl::queue& queue, int64_t n, const float* x, int64_t incx, const float* y, + int64_t incy, double* result, const std::vector& dependencies) { throw unimplemented("blas", "dot", "for row_major layout"); } template -inline sycl::event rot(Func func, sycl::queue &queue, int64_t n, T1 *x, const int64_t incx, T1 *y, - int64_t incy, T2 c, T3 s, const std::vector &dependencies) { +inline sycl::event rot(Func func, sycl::queue& queue, int64_t n, T1* x, const int64_t incx, T1* y, + int64_t incy, T2 c, T3 s, const std::vector& dependencies) { return column_major::rot(func, queue, n, x, incx, y, incy, c, s, dependencies); } #define ROT_LAUNCHER_USM(TYPE1, TYPE2, TYPE3, ROCBLAS_ROUTINE) \ - sycl::event rot(sycl::queue &queue, int64_t n, TYPE1 *x, const int64_t incx, TYPE1 *y, \ + sycl::event rot(sycl::queue& queue, int64_t n, TYPE1* x, const int64_t incx, TYPE1* y, \ int64_t incy, TYPE2 c, TYPE3 s, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return rot(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, c, s, dependencies); \ } @@ -1676,21 +1674,21 @@ ROT_LAUNCHER_USM(std::complex, double, double, rocblas_zdrot) #undef ROT_LAUNCHER_USM -sycl::event sdsdot(sycl::queue &queue, int64_t n, float sb, const float *x, int64_t incx, - const float *y, int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event sdsdot(sycl::queue& queue, int64_t n, float sb, const float* x, int64_t incx, + const float* y, int64_t incy, float* result, + const std::vector& dependencies) { return column_major::sdsdot(queue, n, sb, x, incx, y, incy, result); } template -inline sycl::event rotmg(Func func, sycl::queue &queue, T *d1, T *d2, T *x1, T y1, T *param, - const std::vector &dependencies) { +inline sycl::event rotmg(Func func, sycl::queue& queue, T* d1, T* d2, T* x1, T y1, T* param, + const std::vector& dependencies) { return column_major::rotmg(func, queue, d1, d2, x1, y1, param, dependencies); } #define ROTMG_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event rotmg(sycl::queue &queue, TYPE *d1, TYPE *d2, TYPE *x1, TYPE y1, TYPE *param, \ - const std::vector &dependencies) { \ + sycl::event rotmg(sycl::queue& queue, TYPE* d1, TYPE* d2, TYPE* x1, TYPE y1, TYPE* param, \ + const std::vector& dependencies) { \ return rotmg(ROCBLAS_ROUTINE, queue, d1, d2, x1, y1, param, dependencies); \ } @@ -1700,14 +1698,14 @@ ROTMG_LAUNCHER_USM(double, rocblas_drotmg) #undef ROTMG_LAUNCHER_USM template -inline sycl::event iamax(Func func, sycl::queue &queue, int64_t n, const T *x, const int64_t incx, - int64_t *result, const std::vector &dependencies) { +inline sycl::event iamax(Func func, sycl::queue& queue, int64_t n, const T* x, const int64_t incx, + int64_t* result, const std::vector& dependencies) { return column_major::iamax(func, queue, n, x, incx, result, dependencies); } #define IAMAX_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event iamax(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - int64_t *result, const std::vector &dependencies) { \ + sycl::event iamax(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + int64_t* result, const std::vector& dependencies) { \ return iamax(ROCBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } @@ -1719,14 +1717,14 @@ IAMAX_LAUNCHER_USM(std::complex, rocblas_izamax) #undef IAMAX_LAUNCHER_USM template -inline sycl::event swap(Func func, sycl::queue &queue, int64_t n, T *x, int64_t incx, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event swap(Func func, sycl::queue& queue, int64_t n, T* x, int64_t incx, T* y, + int64_t incy, const std::vector& dependencies) { return column_major::swap(func, queue, n, x, incx, y, incy, dependencies); } #define SWAP_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event swap(sycl::queue &queue, int64_t n, TYPE *x, int64_t incx, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event swap(sycl::queue& queue, int64_t n, TYPE* x, int64_t incx, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return swap(ROCBLAS_ROUTINE, queue, n, x, incx, y, incy, dependencies); \ } @@ -1738,14 +1736,14 @@ SWAP_LAUNCHER_USM(std::complex, rocblas_zswap) #undef SWAP_LAUNCHER_USM template -inline sycl::event iamin(Func func, sycl::queue &queue, int64_t n, const T *x, const int64_t incx, - int64_t *result, const std::vector &dependencies) { +inline sycl::event iamin(Func func, sycl::queue& queue, int64_t n, const T* x, const int64_t incx, + int64_t* result, const std::vector& dependencies) { return column_major::iamin(func, queue, n, x, incx, result, dependencies); } #define IAMIN_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event iamin(sycl::queue &queue, int64_t n, const TYPE *x, const int64_t incx, \ - int64_t *result, const std::vector &dependencies) { \ + sycl::event iamin(sycl::queue& queue, int64_t n, const TYPE* x, const int64_t incx, \ + int64_t* result, const std::vector& dependencies) { \ return iamin(ROCBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } @@ -1757,14 +1755,14 @@ IAMIN_LAUNCHER_USM(std::complex, rocblas_izamin) #undef IAMIN_LAUNCHER_USM template -inline sycl::event nrm2(Func func, sycl::queue &queue, int64_t n, const T1 *x, const int64_t incx, - T2 *result, const std::vector &dependencies) { +inline sycl::event nrm2(Func func, sycl::queue& queue, int64_t n, const T1* x, const int64_t incx, + T2* result, const std::vector& dependencies) { return column_major::nrm2(func, queue, n, x, incx, result, dependencies); } #define NRM2_LAUNCHER_USM(TYPE1, TYPE2, ROCBLAS_ROUTINE) \ - sycl::event nrm2(sycl::queue &queue, int64_t n, const TYPE1 *x, const int64_t incx, \ - TYPE2 *result, const std::vector &dependencies) { \ + sycl::event nrm2(sycl::queue& queue, int64_t n, const TYPE1* x, const int64_t incx, \ + TYPE2* result, const std::vector& dependencies) { \ return nrm2(ROCBLAS_ROUTINE, queue, n, x, incx, result, dependencies); \ } diff --git a/src/blas/backends/rocblas/rocblas_level2.cpp b/src/blas/backends/rocblas/rocblas_level2.cpp index dd62a8558..20c08c5ce 100644 --- a/src/blas/backends/rocblas/rocblas_level2.cpp +++ b/src/blas/backends/rocblas/rocblas_level2.cpp @@ -28,7 +28,7 @@ // Helper Functions template -static inline void conj_vector(sycl::handler &cgh, sycl::buffer &buf, const int64_t len, +static inline void conj_vector(sycl::handler& cgh, sycl::buffer& buf, const int64_t len, const int64_t inc) { const auto abs_inc = std::abs(inc); auto acc = buf.template get_access(cgh); @@ -38,7 +38,7 @@ static inline void conj_vector(sycl::handler &cgh, sycl::buffer &buf, const i }); } template -static inline void conj_vector(sycl::handler &cgh, T *ptr, const int64_t len, const int64_t inc) { +static inline void conj_vector(sycl::handler& cgh, T* ptr, const int64_t len, const int64_t inc) { const auto abs_inc = std::abs(inc); cgh.parallel_for(sycl::range{ (std::size_t)len }, [=](sycl::id<1> id) { const auto index = id * abs_inc; @@ -47,7 +47,7 @@ static inline void conj_vector(sycl::handler &cgh, T *ptr, const int64_t len, co } template -static inline void conj_vector(sycl::handler &cgh, sycl::buffer &buf_a, sycl::buffer &buf_b, +static inline void conj_vector(sycl::handler& cgh, sycl::buffer& buf_a, sycl::buffer& buf_b, const int64_t len, const int64_t inc_a, const int64_t inc_b) { const auto abs_inc_a = std::abs(inc_a); const auto abs_inc_b = std::abs(inc_b); @@ -61,7 +61,7 @@ static inline void conj_vector(sycl::handler &cgh, sycl::buffer &buf_a, sycl: }); } template -static inline void conj_vector(sycl::handler &cgh, T *ptr_a, T *ptr_b, const int64_t len, +static inline void conj_vector(sycl::handler& cgh, T* ptr_a, T* ptr_b, const int64_t len, const int64_t inc_a, const int64_t inc_b) { const auto abs_inc_a = std::abs(inc_a); const auto abs_inc_b = std::abs(inc_b); @@ -82,34 +82,34 @@ namespace column_major { // Buffer APIs template -inline void gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void gemv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, T alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, m, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(trans), m, n, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); } #define GEMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ gemv(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -121,34 +121,34 @@ GEMV_LAUNCHER(std::complex, rocblas_zgemv) #undef GEMV_LAUNCHER template -inline void gbmv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, - int64_t ku, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, T beta, sycl::buffer &y, int64_t incy) { +inline void gbmv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, + int64_t ku, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, T beta, sycl::buffer& y, int64_t incy) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, m, lda, kl, ku, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(trans), m, n, kl, ku, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); } #define GBMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ - int64_t incx, TYPE beta, sycl::buffer &y, int64_t incy) { \ + void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ + int64_t incx, TYPE beta, sycl::buffer& y, int64_t incy) { \ gbmv(ROCBLAS_ROUTINE, queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -160,32 +160,32 @@ GBMV_LAUNCHER(std::complex, rocblas_zgbmv) #undef GBMV_LAUNCHER template -inline void ger(Func func, sycl::queue &queue, int64_t m, int64_t n, T alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, +inline void ger(Func func, sycl::queue& queue, int64_t m, int64_t n, T alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, int64_t lda) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, m, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; - rocblas_native_func(func, err, handle, m, n, (rocDataType *)&alpha, x_, incx, y_, - incy, a_, lda); + rocblas_native_func(func, err, handle, m, n, (rocDataType*)&alpha, x_, incx, y_, incy, + a_, lda); }); }); } #define GER_LAUNCHER(EXT, TYPE, ROCBLAS_ROUTINE) \ - void ger##EXT(sycl::queue &queue, int64_t m, int64_t n, TYPE alpha, sycl::buffer &x, \ - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, \ + void ger##EXT(sycl::queue& queue, int64_t m, int64_t n, TYPE alpha, sycl::buffer& x, \ + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, \ int64_t lda) { \ ger(ROCBLAS_ROUTINE, queue, m, n, alpha, x, incx, y, incy, a, lda); \ } @@ -200,34 +200,34 @@ GER_LAUNCHER(c, std::complex, rocblas_zgerc) #undef GER_LAUNCHER template -inline void hbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void hbmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, T alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, k, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); } #define HBMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ hbmv(ROCBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -237,34 +237,34 @@ HBMV_LAUNCHER(std::complex, rocblas_zhbmv) #undef HBMV_LAUNCHER template -inline void hemv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void hemv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); } #define HEMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void hemv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void hemv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ hemv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -274,31 +274,31 @@ HEMV_LAUNCHER(std::complex, rocblas_zhemv) #undef HEMV_LAUNCHER template -inline void her(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, ScalarType alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a, +inline void her(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, ScalarType alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { using rocScalarType = typename RocEquivalentType::Type; using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocScalarType *)&alpha, x_, incx, a_, lda); + (rocScalarType*)&alpha, x_, incx, a_, lda); }); }); } #define HER_LAUNCHER(SCALAR_TYPE, DATA_TYPE, ROCBLAS_ROUTINE) \ - void her(sycl::queue &queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a, \ + void her(sycl::queue& queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a, \ int64_t lda) { \ her(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda); \ } @@ -309,33 +309,33 @@ HER_LAUNCHER(double, std::complex, rocblas_zher) #undef HER_LAUNCHER template -inline void her2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { +inline void her2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, y_, incy, a_, lda); + (rocDataType*)&alpha, x_, incx, y_, incy, a_, lda); }); }); } #define HER2_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void her2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a, int64_t lda) { \ + void her2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a, int64_t lda) { \ her2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); \ } @@ -345,34 +345,33 @@ HER2_LAUNCHER(std::complex, rocblas_zher2) #undef HER2_LAUNCHER template -inline void hpmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &a, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void hpmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& a, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, a_, x_, incx, (rocDataType *)&beta, y_, - incy); + (rocDataType*)&alpha, a_, x_, incx, (rocDataType*)&beta, y_, incy); }); }); } #define HPMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx, TYPE beta, \ - sycl::buffer &y, int64_t incy) { \ + void hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx, TYPE beta, \ + sycl::buffer& y, int64_t incy) { \ hpmv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); \ } @@ -382,30 +381,30 @@ HPMV_LAUNCHER(std::complex, rocblas_zhpmv) #undef HPMV_LAUNCHER template -inline void hpr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, ScalarType alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a) { +inline void hpr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, ScalarType alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& a) { using rocScalarType = typename RocEquivalentType::Type; using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocScalarType *)&alpha, x_, incx, a_); + (rocScalarType*)&alpha, x_, incx, a_); }); }); } #define HPR_LAUNCHER(SCALAR_TYPE, DATA_TYPE, ROCBLAS_ROUTINE) \ - void hpr(sycl::queue &queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a) { \ + void hpr(sycl::queue& queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a) { \ hpr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a); \ } @@ -415,33 +414,33 @@ HPR_LAUNCHER(double, std::complex, rocblas_zhpr) #undef HPR_LAUNCHER template -inline void hpr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a) { +inline void hpr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, y_, incy, a_); + (rocDataType*)&alpha, x_, incx, y_, incy, a_); }); }); } #define HPR2_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a) { \ + void hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a) { \ hpr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a); \ } @@ -451,34 +450,34 @@ HPR2_LAUNCHER(std::complex, rocblas_zhpr2) #undef HPR2_LAUNCHER template -inline void sbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void sbmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, T alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, k, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); } #define SBMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ sbmv(ROCBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -488,34 +487,34 @@ SBMV_LAUNCHER(double, rocblas_dsbmv) #undef SBMV_LAUNCHER template -inline void symv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void symv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); } #define SYMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void symv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void symv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ symv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -525,29 +524,29 @@ SYMV_LAUNCHER(double, rocblas_dsymv) #undef SYMV_LAUNCHER template -inline void syr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { +inline void syr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, a_, lda); + (rocDataType*)&alpha, x_, incx, a_, lda); }); }); } #define SYR_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void syr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { \ + void syr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { \ syr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda); \ } @@ -560,33 +559,33 @@ SYR_LAUNCHER(std::complex, rocblas_zsyr) #undef SYR_LAUNCHER template -inline void syr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { +inline void syr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, y_, incy, a_, lda); + (rocDataType*)&alpha, x_, incx, y_, incy, a_, lda); }); }); } #define SYR2_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void syr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a, int64_t lda) { \ + void syr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a, int64_t lda) { \ syr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); \ } @@ -599,34 +598,33 @@ SYR2_LAUNCHER(std::complex, rocblas_zsyr2) #undef SYR2_LAUNCHER template -inline void spmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &a, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void spmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& a, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, a_, x_, incx, (rocDataType *)&beta, y_, - incy); + (rocDataType*)&alpha, a_, x_, incx, (rocDataType*)&beta, y_, incy); }); }); } #define SPMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void spmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx, TYPE beta, \ - sycl::buffer &y, int64_t incy) { \ + void spmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx, TYPE beta, \ + sycl::buffer& y, int64_t incy) { \ spmv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); \ } @@ -636,29 +634,29 @@ SPMV_LAUNCHER(double, rocblas_dspmv) #undef SPMV_LAUNCHER template -inline void spr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a) { +inline void spr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& a) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, a_); + (rocDataType*)&alpha, x_, incx, a_); }); }); } #define SPR_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void spr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a) { \ + void spr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a) { \ spr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a); \ } @@ -668,33 +666,33 @@ SPR_LAUNCHER(double, rocblas_dspr) #undef SPR_LAUNCHER template -inline void spr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a) { +inline void spr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); auto y_acc = y.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); - auto y_ = sc.get_mem(y_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); + auto y_ = sc.get_mem(y_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, y_, incy, a_); + (rocDataType*)&alpha, x_, incx, y_, incy, a_); }); }); } #define SPR2_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void spr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a) { \ + void spr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a) { \ spr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a); \ } @@ -704,31 +702,31 @@ SPR2_LAUNCHER(double, rocblas_dspr2) #undef SPR2_LAUNCHER template -inline void tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +inline void tbmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, k, a_, lda, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + k, a_, lda, x_, incx); }); }); } #define TBMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ + void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ int64_t incx) { \ tbmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); \ } @@ -741,31 +739,31 @@ TBMV_LAUNCHER(std::complex, rocblas_ztbmv) #undef TBMV_LAUNCHER template -inline void tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +inline void tbsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, k, a_, lda, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + k, a_, lda, x_, incx); }); }); } #define TBSV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ + void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ int64_t incx) { \ tbsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); \ } @@ -778,30 +776,30 @@ TBSV_LAUNCHER(std::complex, rocblas_ztbsv) #undef TBSV_LAUNCHER template -inline void tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer &a, sycl::buffer &x, int64_t incx) { +inline void tpmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer& a, sycl::buffer& x, int64_t incx) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, a_, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + a_, x_, incx); }); }); } #define TPMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx) { \ + void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx) { \ tpmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, incx); \ } @@ -813,30 +811,30 @@ TPMV_LAUNCHER(std::complex, rocblas_ztpmv) #undef TPMV_LAUNCHER template -inline void tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer &a, sycl::buffer &x, int64_t incx) { +inline void tpsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer& a, sycl::buffer& x, int64_t incx) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, a_, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + a_, x_, incx); }); }); } #define TPSV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx) { \ + void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx) { \ tpsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, incx); \ } @@ -848,31 +846,31 @@ TPSV_LAUNCHER(std::complex, rocblas_ztpsv) #undef TPSV_LAUNCHER template -inline void trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer &a, int64_t lda, sycl::buffer &x, +inline void trmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, a_, lda, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + a_, lda, x_, incx); }); }); } #define TRMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { \ + void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { \ trmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); \ } @@ -884,31 +882,31 @@ TRMV_LAUNCHER(std::complex, rocblas_ztrmv) #undef TRMV_LAUNCHER template -inline void trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer &a, int64_t lda, sycl::buffer &x, +inline void trsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto x_acc = x.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto x_ = sc.get_mem(x_acc); + auto a_ = sc.get_mem(a_acc); + auto x_ = sc.get_mem(x_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, a_, lda, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + a_, lda, x_, incx); }); }); } #define TRSV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { \ + void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { \ trsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); \ } @@ -922,24 +920,24 @@ TRSV_LAUNCHER(std::complex, rocblas_ztrsv) // USM APIs template -inline sycl::event gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - T alpha, const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event gemv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + T alpha, const T* a, int64_t lda, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(trans), m, n, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); @@ -947,9 +945,9 @@ inline sycl::event gemv(Func func, sycl::queue &queue, transpose trans, int64_t } #define GEMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return gemv(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, \ dependencies); \ } @@ -962,25 +960,25 @@ GEMV_LAUNCHER_USM(std::complex, rocblas_zgemv) #undef GEMV_LAUNCHER_USM template -inline sycl::event gbmv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - int64_t kl, int64_t ku, T alpha, const T *a, int64_t lda, const T *x, - int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event gbmv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + int64_t kl, int64_t ku, T alpha, const T* a, int64_t lda, const T* x, + int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, m, lda, kl, ku, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(trans), m, n, kl, ku, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); @@ -988,10 +986,10 @@ inline sycl::event gbmv(Func func, sycl::queue &queue, transpose trans, int64_t } #define GBMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, \ - int64_t ku, TYPE alpha, const TYPE *a, int64_t lda, const TYPE *x, \ - int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, \ + int64_t ku, TYPE alpha, const TYPE* a, int64_t lda, const TYPE* x, \ + int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return gbmv(ROCBLAS_ROUTINE, queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, \ incy, dependencies); \ } @@ -1004,23 +1002,23 @@ GBMV_LAUNCHER_USM(std::complex, rocblas_zgbmv) #undef GBMV_LAUNCHER_USM template -inline sycl::event ger(Func func, sycl::queue &queue, int64_t m, int64_t n, T alpha, const T *x, - int64_t incx, const T *y, int64_t incy, T *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event ger(Func func, sycl::queue& queue, int64_t m, int64_t n, T alpha, const T* x, + int64_t incx, const T* y, int64_t incy, T* a, int64_t lda, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, m, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; - rocblas_native_func(func, err, handle, m, n, (rocDataType *)&alpha, x_, incx, y_, - incy, a_, lda); + rocblas_native_func(func, err, handle, m, n, (rocDataType*)&alpha, x_, incx, y_, incy, + a_, lda); }); }); @@ -1028,9 +1026,9 @@ inline sycl::event ger(Func func, sycl::queue &queue, int64_t m, int64_t n, T al } #define GER_LAUNCHER_USM(EXT, TYPE, ROCBLAS_ROUTINE) \ - sycl::event ger##EXT(sycl::queue &queue, int64_t m, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event ger##EXT(sycl::queue& queue, int64_t m, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return ger(ROCBLAS_ROUTINE, queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); \ } @@ -1044,24 +1042,24 @@ GER_LAUNCHER_USM(c, std::complex, rocblas_zgerc) #undef GER_LAUNCHER_USM template -inline sycl::event hbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, - T alpha, const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event hbmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, + T alpha, const T* a, int64_t lda, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, k, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); @@ -1069,9 +1067,9 @@ inline sycl::event hbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HBMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return hbmv(ROCBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, \ incy, dependencies); \ } @@ -1082,24 +1080,24 @@ HBMV_LAUNCHER_USM(std::complex, rocblas_zhbmv) #undef HBMV_LAUNCHER_USM template -inline sycl::event hemv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event hemv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* a, int64_t lda, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); @@ -1107,9 +1105,9 @@ inline sycl::event hemv(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HEMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event hemv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event hemv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return hemv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, \ dependencies); \ } @@ -1120,23 +1118,23 @@ HEMV_LAUNCHER_USM(std::complex, rocblas_zhemv) #undef HEMV_LAUNCHER_USM template -inline sycl::event her(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - const ScalarType alpha, const DataType *x, int64_t incx, DataType *a, - int64_t lda, const std::vector &dependencies) { +inline sycl::event her(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + const ScalarType alpha, const DataType* x, int64_t incx, DataType* a, + int64_t lda, const std::vector& dependencies) { using rocScalarType = typename RocEquivalentType::Type; using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocScalarType *)&alpha, x_, incx, a_, lda); + (rocScalarType*)&alpha, x_, incx, a_, lda); }); }); @@ -1144,9 +1142,9 @@ inline sycl::event her(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HER_LAUNCHER_USM(SCALAR_TYPE, DATA_TYPE, ROCBLAS_ROUTINE) \ - sycl::event her(sycl::queue &queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ - const DATA_TYPE *x, int64_t incx, DATA_TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event her(sycl::queue& queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ + const DATA_TYPE* x, int64_t incx, DATA_TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return her(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); \ } @@ -1156,23 +1154,23 @@ HER_LAUNCHER_USM(double, std::complex, rocblas_zher) #undef HER_LAUNCHER_USM template -inline sycl::event her2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, const T *y, int64_t incy, T *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event her2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, const T* y, int64_t incy, T* a, int64_t lda, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, y_, incy, a_, lda); + (rocDataType*)&alpha, x_, incx, y_, incy, a_, lda); }); }); @@ -1180,9 +1178,9 @@ inline sycl::event her2(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HER2_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event her2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event her2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return her2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, \ dependencies); \ } @@ -1193,24 +1191,23 @@ HER2_LAUNCHER_USM(std::complex, rocblas_zher2) #undef HER2_LAUNCHER_USM template -inline sycl::event hpmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *a, const T *x, int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event hpmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* a, const T* x, int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, a_, x_, incx, (rocDataType *)&beta, y_, - incy); + (rocDataType*)&alpha, a_, x_, incx, (rocDataType*)&beta, y_, incy); }); }); @@ -1218,9 +1215,9 @@ inline sycl::event hpmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HPMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return hpmv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, \ dependencies); \ } @@ -1231,23 +1228,23 @@ HPMV_LAUNCHER_USM(std::complex, rocblas_zhpmv) #undef HPMV_LAUNCHER_USM template -inline sycl::event hpr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - const ScalarType alpha, const DataType *x, int64_t incx, DataType *a, - const std::vector &dependencies) { +inline sycl::event hpr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + const ScalarType alpha, const DataType* x, int64_t incx, DataType* a, + const std::vector& dependencies) { using rocScalarType = typename RocEquivalentType::Type; using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocScalarType *)&alpha, x_, incx, a_); + (rocScalarType*)&alpha, x_, incx, a_); }); }); @@ -1255,9 +1252,9 @@ inline sycl::event hpr(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HPR_LAUNCHER_USM(SCALAR_TYPE, DATA_TYPE, ROCBLAS_ROUTINE) \ - sycl::event hpr(sycl::queue &queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ - const DATA_TYPE *x, int64_t incx, DATA_TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event hpr(sycl::queue& queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ + const DATA_TYPE* x, int64_t incx, DATA_TYPE* a, \ + const std::vector& dependencies) { \ return hpr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, dependencies); \ } @@ -1267,23 +1264,23 @@ HPR_LAUNCHER_USM(double, std::complex, rocblas_zhpr) #undef HPR_LAUNCHER_USM template -inline sycl::event hpr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, const T *y, int64_t incy, T *a, - const std::vector &dependencies) { +inline sycl::event hpr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, const T* y, int64_t incy, T* a, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, y_, incy, a_); + (rocDataType*)&alpha, x_, incx, y_, incy, a_); }); }); @@ -1291,9 +1288,9 @@ inline sycl::event hpr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HPR2_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, \ + const std::vector& dependencies) { \ return hpr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, \ dependencies); \ } @@ -1304,24 +1301,24 @@ HPR2_LAUNCHER_USM(std::complex, rocblas_zhpr2) #undef HPR2_LAUNCHER_USM template -inline sycl::event sbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, - T alpha, const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event sbmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, + T alpha, const T* a, int64_t lda, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, k, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); @@ -1329,9 +1326,9 @@ inline sycl::event sbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define SBMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return sbmv(ROCBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, \ incy, dependencies); \ } @@ -1342,24 +1339,24 @@ SBMV_LAUNCHER_USM(double, rocblas_dsbmv) #undef SBMV_LAUNCHER_USM template -inline sycl::event symv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event symv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* a, int64_t lda, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, a_, lda, x_, incx, (rocDataType *)&beta, - y_, incy); + (rocDataType*)&alpha, a_, lda, x_, incx, (rocDataType*)&beta, y_, + incy); }); }); @@ -1367,9 +1364,9 @@ inline sycl::event symv(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define SYMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event symv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event symv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return symv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, \ dependencies); \ } @@ -1380,22 +1377,22 @@ SYMV_LAUNCHER_USM(double, rocblas_dsymv) #undef SYMV_LAUNCHER_USM template -inline sycl::event syr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, T *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event syr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, T* a, int64_t lda, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, a_, lda); + (rocDataType*)&alpha, x_, incx, a_, lda); }); }); @@ -1403,9 +1400,9 @@ inline sycl::event syr(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define SYR_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event syr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return syr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); \ } @@ -1418,23 +1415,23 @@ SYR_LAUNCHER_USM(std::complex, rocblas_zsyr) #undef SYR_LAUNCHER_USM template -inline sycl::event syr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, const T *y, int64_t incy, T *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event syr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, const T* y, int64_t incy, T* a, int64_t lda, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, y_, incy, a_, lda); + (rocDataType*)&alpha, x_, incx, y_, incy, a_, lda); }); }); @@ -1442,9 +1439,9 @@ inline sycl::event syr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define SYR2_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event syr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return syr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, \ dependencies); \ } @@ -1458,24 +1455,23 @@ SYR2_LAUNCHER_USM(std::complex, rocblas_zsyr2) #undef SYR2_LAUNCHER_USM template -inline sycl::event spmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *a, const T *x, int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event spmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* a, const T* x, int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, a_, x_, incx, (rocDataType *)&beta, y_, - incy); + (rocDataType*)&alpha, a_, x_, incx, (rocDataType*)&beta, y_, incy); }); }); @@ -1483,9 +1479,9 @@ inline sycl::event spmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define SPMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event spmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event spmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return spmv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, \ dependencies); \ } @@ -1496,21 +1492,21 @@ SPMV_LAUNCHER_USM(double, rocblas_dspmv) #undef SPMV_LAUNCHER_USM template -inline sycl::event spr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, T *a, - const std::vector &dependencies) { +inline sycl::event spr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, T* a, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, a_); + (rocDataType*)&alpha, x_, incx, a_); }); }); @@ -1518,8 +1514,8 @@ inline sycl::event spr(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define SPR_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event spr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, TYPE *a, const std::vector &dependencies) { \ + sycl::event spr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, TYPE* a, const std::vector& dependencies) { \ return spr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, dependencies); \ } @@ -1529,23 +1525,23 @@ SPR_LAUNCHER_USM(double, rocblas_dspr) #undef SPR_LAUNCHER_USM template -inline sycl::event spr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, const T *y, int64_t incy, T *a, - const std::vector &dependencies) { +inline sycl::event spr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, const T* y, int64_t incy, T* a, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx, incy); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); - auto y_ = reinterpret_cast(y); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); + auto y_ = reinterpret_cast(y); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), n, - (rocDataType *)&alpha, x_, incx, y_, incy, a_); + (rocDataType*)&alpha, x_, incx, y_, incy, a_); }); }); @@ -1553,9 +1549,9 @@ inline sycl::event spr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define SPR2_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event spr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event spr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, \ + const std::vector& dependencies) { \ return spr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, \ dependencies); \ } @@ -1566,23 +1562,23 @@ SPR2_LAUNCHER_USM(double, rocblas_dspr2) #undef SPR2_LAUNCHER_USM template -inline sycl::event tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const T *a, int64_t lda, T *x, - int64_t incx, const std::vector &dependencies) { +inline sycl::event tbmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, int64_t k, const T* a, int64_t lda, T* x, + int64_t incx, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, k, a_, lda, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + k, a_, lda, x_, incx); }); }); @@ -1590,9 +1586,9 @@ inline sycl::event tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpo } #define TBMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, int64_t k, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, int64_t k, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tbmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, \ dependencies); \ } @@ -1605,23 +1601,23 @@ TBMV_LAUNCHER_USM(std::complex, rocblas_ztbmv) #undef TBMV_LAUNCHER_USM template -inline sycl::event tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const T *a, int64_t lda, T *x, - int64_t incx, const std::vector &dependencies) { +inline sycl::event tbsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, int64_t k, const T* a, int64_t lda, T* x, + int64_t incx, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, k, a_, lda, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + k, a_, lda, x_, incx); }); }); @@ -1629,9 +1625,9 @@ inline sycl::event tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpo } #define TBSV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, int64_t k, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, int64_t k, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tbsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, \ dependencies); \ } @@ -1644,23 +1640,23 @@ TBSV_LAUNCHER_USM(std::complex, rocblas_ztbsv) #undef TBSV_LAUNCHER_USM template -inline sycl::event tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const T *a, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tpmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const T* a, T* x, int64_t incx, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, a_, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + a_, x_, incx); }); }); @@ -1668,9 +1664,9 @@ inline sycl::event tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpo } #define TPMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tpmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, incx, \ dependencies); \ } @@ -1683,23 +1679,23 @@ TPMV_LAUNCHER_USM(std::complex, rocblas_ztpmv) #undef TPMV_LAUNCHER_USM template -inline sycl::event tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const T *a, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tpsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const T* a, T* x, int64_t incx, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, a_, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + a_, x_, incx); }); }); @@ -1707,9 +1703,9 @@ inline sycl::event tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpo } #define TPSV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tpsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, incx, \ dependencies); \ } @@ -1722,23 +1718,23 @@ TPSV_LAUNCHER_USM(std::complex, rocblas_ztpsv) #undef TPSV_LAUNCHER_USM template -inline sycl::event trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const T *a, int64_t lda, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event trmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const T* a, int64_t lda, T* x, int64_t incx, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, a_, lda, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + a_, lda, x_, incx); }); }); @@ -1746,9 +1742,9 @@ inline sycl::event trmv(Func func, sycl::queue &queue, uplo upper_lower, transpo } #define TRMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return trmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, \ dependencies); \ } @@ -1761,23 +1757,23 @@ TRMV_LAUNCHER_USM(std::complex, rocblas_ztrmv) #undef TRMV_LAUNCHER_USM template -inline sycl::event trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const T *a, int64_t lda, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event trsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const T* a, int64_t lda, T* x, int64_t incx, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, lda, incx); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto x_ = reinterpret_cast(x); + auto a_ = reinterpret_cast(a); + auto x_ = reinterpret_cast(x); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - n, a_, lda, x_, incx); + get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), n, + a_, lda, x_, incx); }); }); @@ -1785,9 +1781,9 @@ inline sycl::event trsv(Func func, sycl::queue &queue, uplo upper_lower, transpo } #define TRSV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return trsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, \ dependencies); \ } @@ -1806,22 +1802,22 @@ namespace row_major { // Buffer APIs template -inline void gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx, std::complex beta, - sycl::buffer, 1> &y, int64_t incy) { +inline void gemv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx, std::complex beta, + sycl::buffer, 1>& y, int64_t incy) { auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); beta = std::conj(beta); if (m > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, m, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, m, incx); }); if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy); }); } } } @@ -1830,25 +1826,25 @@ inline void gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int6 if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy); }); } } } template -inline void gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void gemv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, T alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::gemv(func, queue, new_trans, n, m, alpha, a, lda, x, incx, beta, y, incy); } #define GEMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ gemv(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -1860,22 +1856,22 @@ GEMV_LAUNCHER(std::complex, rocblas_zgemv) #undef GEMV_LAUNCHER template -inline void gbmv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, - int64_t ku, std::complex alpha, sycl::buffer, 1> &a, - int64_t lda, sycl::buffer, 1> &x, int64_t incx, - std::complex beta, sycl::buffer, 1> &y, int64_t incy) { +inline void gbmv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, + int64_t ku, std::complex alpha, sycl::buffer, 1>& a, + int64_t lda, sycl::buffer, 1>& x, int64_t incx, + std::complex beta, sycl::buffer, 1>& y, int64_t incy) { auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); beta = std::conj(beta); if (m > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, m, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, m, incx); }); if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy); }); } } } @@ -1884,25 +1880,25 @@ inline void gbmv(Func func, sycl::queue &queue, transpose trans, int64_t m, int6 if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy); }); } } } template -inline void gbmv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, - int64_t ku, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, - int64_t incx, T beta, sycl::buffer &y, int64_t incy) { +inline void gbmv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, + int64_t ku, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, + int64_t incx, T beta, sycl::buffer& y, int64_t incy) { auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::gbmv(func, queue, new_trans, n, m, ku, kl, alpha, a, lda, x, incx, beta, y, incy); } #define GBMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ - int64_t incx, TYPE beta, sycl::buffer &y, int64_t incy) { \ + void gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, int64_t ku, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ + int64_t incx, TYPE beta, sycl::buffer& y, int64_t incy) { \ gbmv(ROCBLAS_ROUTINE, queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -1914,35 +1910,35 @@ GBMV_LAUNCHER(std::complex, rocblas_zgbmv) #undef GBMV_LAUNCHER template -inline void gerc(Func func, sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { +inline void gerc(Func func, sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy); }); } column_major::ger(func, queue, n, m, alpha, y, incy, x, incx, a, lda); } template -inline void geru(Func func, sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - sycl::buffer, 1> &x, int64_t incx, - sycl::buffer, 1> &y, int64_t incy, - sycl::buffer, 1> &a, int64_t lda) { +inline void geru(Func func, sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + sycl::buffer, 1>& x, int64_t incx, + sycl::buffer, 1>& y, int64_t incy, + sycl::buffer, 1>& a, int64_t lda) { column_major::ger(func, queue, n, m, alpha, y, incy, x, incx, a, lda); } template -inline void ger(Func func, sycl::queue &queue, int64_t m, int64_t n, T alpha, sycl::buffer &x, - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, +inline void ger(Func func, sycl::queue& queue, int64_t m, int64_t n, T alpha, sycl::buffer& x, + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, int64_t lda) { column_major::ger(func, queue, n, m, alpha, y, incy, x, incx, a, lda); } #define GER_LAUNCHER(EXT, TYPE, ROCBLAS_ROUTINE) \ - void ger##EXT(sycl::queue &queue, int64_t m, int64_t n, TYPE alpha, sycl::buffer &x, \ - int64_t incx, sycl::buffer &y, int64_t incy, sycl::buffer &a, \ + void ger##EXT(sycl::queue& queue, int64_t m, int64_t n, TYPE alpha, sycl::buffer& x, \ + int64_t incx, sycl::buffer& y, int64_t incy, sycl::buffer& a, \ int64_t lda) { \ ger##EXT(ROCBLAS_ROUTINE, queue, m, n, alpha, x, incx, y, incy, a, lda); \ } @@ -1957,29 +1953,29 @@ GER_LAUNCHER(c, std::complex, rocblas_zgeru) #undef GER_LAUNCHER template -inline void hbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void hbmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, T alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, y, n, incx, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, y, n, incx, incy); }); } column_major::hbmv(func, queue, new_uplo, n, k, new_alpha, a, lda, x, incx, new_beta, y, incy); if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy); }); } } #define HBMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ hbmv(ROCBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -1989,29 +1985,29 @@ HBMV_LAUNCHER(std::complex, rocblas_zhbmv) #undef HBMV_LAUNCHER template -inline void hemv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void hemv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, y, n, incx, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, y, n, incx, incy); }); } column_major::hemv(func, queue, new_uplo, n, new_alpha, a, lda, x, incx, new_beta, y, incy); if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy); }); } } #define HEMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void hemv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void hemv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ hemv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -2021,22 +2017,22 @@ HEMV_LAUNCHER(std::complex, rocblas_zhemv) #undef HEMV_LAUNCHER template -inline void her(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, ScalarType alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a, +inline void her(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, ScalarType alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } column_major::her(func, queue, new_uplo, n, alpha, x, incx, a, lda); } #define HER_LAUNCHER(SCALAR_TYPE, DATA_TYPE, ROCBLAS_ROUTINE) \ - void her(sycl::queue &queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a, \ + void her(sycl::queue& queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a, \ int64_t lda) { \ her(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda); \ } @@ -2047,23 +2043,23 @@ HER_LAUNCHER(double, std::complex, rocblas_zher) #undef HER_LAUNCHER template -inline void her2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { +inline void her2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, y, n, incx, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, y, n, incx, incy); }); } column_major::her2(func, queue, new_uplo, n, alpha, y, incy, x, incx, a, lda); } #define HER2_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void her2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a, int64_t lda) { \ + void her2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a, int64_t lda) { \ her2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); \ } @@ -2073,29 +2069,29 @@ HER2_LAUNCHER(std::complex, rocblas_zher2) #undef HER2_LAUNCHER template -inline void hpmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &a, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void hpmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& a, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, y, n, incx, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, y, n, incx, incy); }); } column_major::hpmv(func, queue, new_uplo, n, new_alpha, a, x, incx, new_beta, y, incy); if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy); }); } } #define HPMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx, TYPE beta, \ - sycl::buffer &y, int64_t incy) { \ + void hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx, TYPE beta, \ + sycl::buffer& y, int64_t incy) { \ hpmv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); \ } @@ -2105,21 +2101,21 @@ HPMV_LAUNCHER(std::complex, rocblas_zhpmv) #undef HPMV_LAUNCHER template -inline void hpr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, ScalarType alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a) { +inline void hpr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, ScalarType alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& a) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } column_major::hpr(func, queue, new_uplo, n, alpha, x, incx, a); } #define HPR_LAUNCHER(SCALAR_TYPE, DATA_TYPE, ROCBLAS_ROUTINE) \ - void hpr(sycl::queue &queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a) { \ + void hpr(sycl::queue& queue, uplo upper_lower, int64_t n, SCALAR_TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a) { \ hpr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a); \ } @@ -2129,23 +2125,23 @@ HPR_LAUNCHER(double, std::complex, rocblas_zhpr) #undef HPR_LAUNCHER template -inline void hpr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a) { +inline void hpr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, y, n, incx, incy); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, y, n, incx, incy); }); } column_major::hpr2(func, queue, new_uplo, n, alpha, y, incy, x, incx, a); } #define HPR2_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a) { \ + void hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a) { \ hpr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a); \ } @@ -2155,19 +2151,19 @@ HPR2_LAUNCHER(std::complex, rocblas_zhpr2) #undef HPR2_LAUNCHER template -inline void sbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void sbmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, T alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::sbmv(func, queue, new_uplo, n, k, alpha, a, lda, x, incx, beta, y, incy); } #define SBMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ sbmv(ROCBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -2177,19 +2173,19 @@ SBMV_LAUNCHER(double, rocblas_dsbmv) #undef SBMV_LAUNCHER template -inline void symv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void symv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::symv(func, queue, new_uplo, n, alpha, a, lda, x, incx, beta, y, incy); } #define SYMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void symv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx, \ - TYPE beta, sycl::buffer &y, int64_t incy) { \ + void symv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx, \ + TYPE beta, sycl::buffer& y, int64_t incy) { \ symv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); \ } @@ -2199,17 +2195,17 @@ SYMV_LAUNCHER(double, rocblas_dsymv) #undef SYMV_LAUNCHER template -inline void syr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { +inline void syr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::syr(func, queue, new_uplo, n, alpha, x, incx, a, lda); } #define SYR_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void syr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a, int64_t lda) { \ + void syr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a, int64_t lda) { \ syr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda); \ } @@ -2222,19 +2218,19 @@ SYR_LAUNCHER(std::complex, rocblas_zsyr) #undef SYR_LAUNCHER template -inline void syr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a, int64_t lda) { +inline void syr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a, int64_t lda) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::syr2(func, queue, new_uplo, n, alpha, x, incx, y, incy, a, lda); } #define SYR2_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void syr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a, int64_t lda) { \ + void syr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a, int64_t lda) { \ syr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); \ } @@ -2247,19 +2243,19 @@ SYR2_LAUNCHER(std::complex, rocblas_zsyr2) #undef SYR2_LAUNCHER template -inline void spmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &a, sycl::buffer &x, int64_t incx, T beta, - sycl::buffer &y, int64_t incy) { +inline void spmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& a, sycl::buffer& x, int64_t incx, T beta, + sycl::buffer& y, int64_t incy) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::spmv(func, queue, new_uplo, n, alpha, a, x, incx, beta, y, incy); } #define SPMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void spmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx, TYPE beta, \ - sycl::buffer &y, int64_t incy) { \ + void spmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx, TYPE beta, \ + sycl::buffer& y, int64_t incy) { \ spmv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); \ } @@ -2269,17 +2265,17 @@ SPMV_LAUNCHER(double, rocblas_dspmv) #undef SPMV_LAUNCHER template -inline void spr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &a) { +inline void spr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& a) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::spr(func, queue, new_uplo, n, alpha, x, incx, a); } #define SPR_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void spr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &a) { \ + void spr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& a) { \ spr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a); \ } @@ -2289,19 +2285,19 @@ SPR_LAUNCHER(double, rocblas_dspr) #undef SPR_LAUNCHER template -inline void spr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, - sycl::buffer &a) { +inline void spr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, + sycl::buffer& a) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::spr2(func, queue, new_uplo, n, alpha, x, incx, y, incy, a); } #define SPR2_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void spr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, \ - sycl::buffer &x, int64_t incx, sycl::buffer &y, int64_t incy, \ - sycl::buffer &a) { \ + void spr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, \ + sycl::buffer& x, int64_t incx, sycl::buffer& y, int64_t incy, \ + sycl::buffer& a) { \ spr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a); \ } @@ -2311,17 +2307,17 @@ SPR2_LAUNCHER(double, rocblas_dspr2) #undef SPR2_LAUNCHER template -inline void tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, int64_t k, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { +inline void tbmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, int64_t k, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } @@ -2329,26 +2325,26 @@ inline void tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } } template -inline void tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +inline void tbmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::tbmv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx); } #define TBMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ + void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ int64_t incx) { \ tbmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); \ } @@ -2361,17 +2357,17 @@ TBMV_LAUNCHER(std::complex, rocblas_ztbmv) #undef TBMV_LAUNCHER template -inline void tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, int64_t k, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { +inline void tbsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, int64_t k, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } @@ -2379,26 +2375,26 @@ inline void tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } } template -inline void tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, +inline void tbsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::tbsv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx); } #define TBSV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - int64_t k, sycl::buffer &a, int64_t lda, sycl::buffer &x, \ + void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + int64_t k, sycl::buffer& a, int64_t lda, sycl::buffer& x, \ int64_t incx) { \ tbsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); \ } @@ -2411,17 +2407,17 @@ TBSV_LAUNCHER(std::complex, rocblas_ztbsv) #undef TBSV_LAUNCHER template -inline void tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, int64_t incx) { +inline void tpmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } @@ -2429,25 +2425,25 @@ inline void tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } } template -inline void tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer &a, sycl::buffer &x, int64_t incx) { +inline void tpmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer& a, sycl::buffer& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::tpmv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx); } #define TPMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx) { \ + void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx) { \ tpmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, incx); \ } @@ -2459,17 +2455,17 @@ TPMV_LAUNCHER(std::complex, rocblas_ztpmv) #undef TPMV_LAUNCHER template -inline void tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, int64_t incx) { +inline void tpsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } @@ -2477,25 +2473,25 @@ inline void tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } } template -inline void tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer &a, sycl::buffer &x, int64_t incx) { +inline void tpsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer& a, sycl::buffer& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::tpsv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx); } #define TPSV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, sycl::buffer &x, int64_t incx) { \ + void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, sycl::buffer& x, int64_t incx) { \ tpsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, incx); \ } @@ -2507,17 +2503,17 @@ TPSV_LAUNCHER(std::complex, rocblas_ztpsv) #undef TPSV_LAUNCHER template -inline void trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { +inline void trmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } @@ -2525,26 +2521,26 @@ inline void trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } } template -inline void trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer &a, int64_t lda, sycl::buffer &x, +inline void trmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::trmv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx); } #define TRMV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { \ + void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { \ trmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); \ } @@ -2556,17 +2552,17 @@ TRMV_LAUNCHER(std::complex, rocblas_ztrmv) #undef TRMV_LAUNCHER template -inline void trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer, 1> &a, int64_t lda, - sycl::buffer, 1> &x, int64_t incx) { +inline void trsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer, 1>& a, int64_t lda, + sycl::buffer, 1>& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } @@ -2574,26 +2570,26 @@ inline void trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose tran if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }); + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }); } } } template -inline void trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, - int64_t n, sycl::buffer &a, int64_t lda, sycl::buffer &x, +inline void trsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, + int64_t n, sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::trsv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx); } #define TRSV_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ - sycl::buffer &a, int64_t lda, sycl::buffer &x, int64_t incx) { \ + void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, int64_t n, \ + sycl::buffer& a, int64_t lda, sycl::buffer& x, int64_t incx) { \ trsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); \ } @@ -2607,15 +2603,15 @@ TRSV_LAUNCHER(std::complex, rocblas_ztrsv) // USM APIs template -inline sycl::event gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event gemv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + std::complex alpha, const std::complex* a, int64_t lda, + const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { sycl::event done; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); @@ -2623,10 +2619,10 @@ inline sycl::event gemv(Func func, sycl::queue &queue, transpose trans, int64_t if (m > 0) { done = queue.submit( - [&](sycl::handler &cgh) { conj_vector(cgh, (std::complex *)x, m, incx); }); + [&](sycl::handler& cgh) { conj_vector(cgh, (std::complex*)x, m, incx); }); if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); + done = queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy); }); } } } @@ -2638,7 +2634,7 @@ inline sycl::event gemv(Func func, sycl::queue &queue, transpose trans, int64_t if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, y, n, incy); }); @@ -2649,20 +2645,20 @@ inline sycl::event gemv(Func func, sycl::queue &queue, transpose trans, int64_t } template -inline sycl::event gemv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - T alpha, const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event gemv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + T alpha, const T* a, int64_t lda, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::gemv(func, queue, new_trans, n, m, alpha, a, lda, x, incx, beta, y, incy, dependencies); } #define GEMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event gemv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event gemv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return gemv(ROCBLAS_ROUTINE, queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, \ dependencies); \ } @@ -2675,15 +2671,15 @@ GEMV_LAUNCHER_USM(std::complex, rocblas_zgemv) #undef GEMV_LAUNCHER_USM template -inline sycl::event gbmv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - int64_t kl, int64_t ku, std::complex alpha, const std::complex *a, - int64_t lda, const std::complex *x, int64_t incx, std::complex beta, - std::complex *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event gbmv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + int64_t kl, int64_t ku, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* x, int64_t incx, std::complex beta, + std::complex* y, int64_t incy, + const std::vector& dependencies) { sycl::event done; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { alpha = std::conj(alpha); @@ -2691,10 +2687,10 @@ inline sycl::event gbmv(Func func, sycl::queue &queue, transpose trans, int64_t if (m > 0) { done = queue.submit( - [&](sycl::handler &cgh) { conj_vector(cgh, (std::complex *)x, m, incx); }); + [&](sycl::handler& cgh) { conj_vector(cgh, (std::complex*)x, m, incx); }); if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, y, n, incy); }); + done = queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, y, n, incy); }); } } } @@ -2706,7 +2702,7 @@ inline sycl::event gbmv(Func func, sycl::queue &queue, transpose trans, int64_t if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, y, n, incy); }); @@ -2717,22 +2713,22 @@ inline sycl::event gbmv(Func func, sycl::queue &queue, transpose trans, int64_t } template -inline sycl::event gbmv(Func func, sycl::queue &queue, transpose trans, int64_t m, int64_t n, - int64_t kl, int64_t ku, T alpha, const T *a, int64_t lda, const T *x, - int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event gbmv(Func func, sycl::queue& queue, transpose trans, int64_t m, int64_t n, + int64_t kl, int64_t ku, T alpha, const T* a, int64_t lda, const T* x, + int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::gbmv(func, queue, new_trans, n, m, ku, kl, alpha, a, lda, x, incx, beta, y, incy, dependencies); } #define GBMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event gbmv(sycl::queue &queue, transpose trans, int64_t m, int64_t n, int64_t kl, \ - int64_t ku, TYPE alpha, const TYPE *a, int64_t lda, const TYPE *x, \ - int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event gbmv(sycl::queue& queue, transpose trans, int64_t m, int64_t n, int64_t kl, \ + int64_t ku, TYPE alpha, const TYPE* a, int64_t lda, const TYPE* x, \ + int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return gbmv(ROCBLAS_ROUTINE, queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, \ incy, dependencies); \ } @@ -2745,12 +2741,12 @@ GBMV_LAUNCHER_USM(std::complex, rocblas_zgbmv) #undef GBMV_LAUNCHER_USM template -inline sycl::event gerc(Func func, sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event gerc(Func func, sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (std::complex *)y, n, incy); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, (std::complex*)y, n, incy); }) .wait_and_throw(); } @@ -2758,24 +2754,24 @@ inline sycl::event gerc(Func func, sycl::queue &queue, int64_t m, int64_t n, std } template -inline sycl::event geru(Func func, sycl::queue &queue, int64_t m, int64_t n, std::complex alpha, - const std::complex *x, int64_t incx, const std::complex *y, - int64_t incy, std::complex *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event geru(Func func, sycl::queue& queue, int64_t m, int64_t n, std::complex alpha, + const std::complex* x, int64_t incx, const std::complex* y, + int64_t incy, std::complex* a, int64_t lda, + const std::vector& dependencies) { return column_major::ger(func, queue, n, m, alpha, y, incy, x, incx, a, lda, dependencies); } template -inline sycl::event ger(Func func, sycl::queue &queue, int64_t m, int64_t n, T alpha, const T *x, - int64_t incx, const T *y, int64_t incy, T *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event ger(Func func, sycl::queue& queue, int64_t m, int64_t n, T alpha, const T* x, + int64_t incx, const T* y, int64_t incy, T* a, int64_t lda, + const std::vector& dependencies) { return column_major::ger(func, queue, n, m, alpha, y, incy, x, incx, a, lda, dependencies); } #define GER_LAUNCHER_USM(EXT, TYPE, ROCBLAS_ROUTINE) \ - sycl::event ger##EXT(sycl::queue &queue, int64_t m, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event ger##EXT(sycl::queue& queue, int64_t m, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return ger##EXT(ROCBLAS_ROUTINE, queue, m, n, alpha, x, incx, y, incy, a, lda, \ dependencies); \ } @@ -2790,18 +2786,18 @@ GER_LAUNCHER_USM(c, std::complex, rocblas_zgeru) #undef GER_LAUNCHER_USM template -inline sycl::event hbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, - T alpha, const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event hbmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, + T alpha, const T* a, int64_t lda, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { sycl::event done; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (T *)x, y, n, incx, incy); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, (T*)x, y, n, incx, incy); }) .wait_and_throw(); } @@ -2809,7 +2805,7 @@ inline sycl::event hbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t incy, dependencies); if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, y, n, incy); }); @@ -2819,9 +2815,9 @@ inline sycl::event hbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HBMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event hbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event hbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return hbmv(ROCBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, \ incy, dependencies); \ } @@ -2832,18 +2828,18 @@ HBMV_LAUNCHER_USM(std::complex, rocblas_zhbmv) #undef HBMV_LAUNCHER_USM template -inline sycl::event hemv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event hemv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* a, int64_t lda, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { sycl::event done; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (T *)x, y, n, incx, incy); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, (T*)x, y, n, incx, incy); }) .wait_and_throw(); } @@ -2851,7 +2847,7 @@ inline sycl::event hemv(Func func, sycl::queue &queue, uplo upper_lower, int64_t incy, dependencies); if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, y, n, incy); }); @@ -2861,9 +2857,9 @@ inline sycl::event hemv(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HEMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event hemv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event hemv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return hemv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, \ dependencies); \ } @@ -2874,14 +2870,14 @@ HEMV_LAUNCHER_USM(std::complex, rocblas_zhemv) #undef HEMV_LAUNCHER_USM template -inline sycl::event her(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - const ScalarType alpha, const DataType *x, int64_t incx, DataType *a, - int64_t lda, const std::vector &dependencies) { +inline sycl::event her(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + const ScalarType alpha, const DataType* x, int64_t incx, DataType* a, + int64_t lda, const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (DataType *)x, n, incx); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, (DataType*)x, n, incx); }) .wait_and_throw(); } @@ -2889,9 +2885,9 @@ inline sycl::event her(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HER_LAUNCHER_USM(SCALAR_TYPE, DATA_TYPE, ROCBLAS_ROUTINE) \ - sycl::event her(sycl::queue &queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ - const DATA_TYPE *x, int64_t incx, DATA_TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event her(sycl::queue& queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ + const DATA_TYPE* x, int64_t incx, DATA_TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return her(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); \ } @@ -2901,14 +2897,14 @@ HER_LAUNCHER_USM(double, std::complex, rocblas_zher) #undef HER_LAUNCHER_USM template -inline sycl::event her2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, const T *y, int64_t incy, T *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event her2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, const T* y, int64_t incy, T* a, int64_t lda, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (T *)x, (T *)y, n, incx, incy); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, (T*)x, (T*)y, n, incx, incy); }) .wait_and_throw(); } @@ -2917,9 +2913,9 @@ inline sycl::event her2(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HER2_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event her2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event her2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return her2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, \ dependencies); \ } @@ -2930,18 +2926,18 @@ HER2_LAUNCHER_USM(std::complex, rocblas_zher2) #undef HER2_LAUNCHER_USM template -inline sycl::event hpmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *a, const T *x, int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event hpmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* a, const T* x, int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { sycl::event done; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_alpha = std::conj(alpha); auto new_beta = std::conj(beta); if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (T *)x, y, n, incx, incy); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, (T*)x, y, n, incx, incy); }) .wait_and_throw(); } @@ -2949,7 +2945,7 @@ inline sycl::event hpmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t dependencies); if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, y, n, incy); }); @@ -2959,9 +2955,9 @@ inline sycl::event hpmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HPMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event hpmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event hpmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return hpmv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, \ dependencies); \ } @@ -2972,14 +2968,14 @@ HPMV_LAUNCHER_USM(std::complex, rocblas_zhpmv) #undef HPMV_LAUNCHER_USM template -inline sycl::event hpr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, - const ScalarType alpha, const DataType *x, int64_t incx, DataType *a, - const std::vector &dependencies) { +inline sycl::event hpr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, + const ScalarType alpha, const DataType* x, int64_t incx, DataType* a, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (DataType *)x, n, incx); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, (DataType*)x, n, incx); }) .wait_and_throw(); } @@ -2987,9 +2983,9 @@ inline sycl::event hpr(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HPR_LAUNCHER_USM(SCALAR_TYPE, DATA_TYPE, ROCBLAS_ROUTINE) \ - sycl::event hpr(sycl::queue &queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ - const DATA_TYPE *x, int64_t incx, DATA_TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event hpr(sycl::queue& queue, uplo upper_lower, int64_t n, const SCALAR_TYPE alpha, \ + const DATA_TYPE* x, int64_t incx, DATA_TYPE* a, \ + const std::vector& dependencies) { \ return hpr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, dependencies); \ } @@ -2999,14 +2995,14 @@ HPR_LAUNCHER_USM(double, std::complex, rocblas_zhpr) #undef HPR_LAUNCHER_USM template -inline sycl::event hpr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, const T *y, int64_t incy, T *a, - const std::vector &dependencies) { +inline sycl::event hpr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, const T* y, int64_t incy, T* a, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, (T *)x, (T *)y, n, incx, incy); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, (T*)x, (T*)y, n, incx, incy); }) .wait_and_throw(); } @@ -3014,9 +3010,9 @@ inline sycl::event hpr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t } #define HPR2_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event hpr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event hpr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, \ + const std::vector& dependencies) { \ return hpr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, \ dependencies); \ } @@ -3027,20 +3023,20 @@ HPR2_LAUNCHER_USM(std::complex, rocblas_zhpr2) #undef HPR2_LAUNCHER_USM template -inline sycl::event sbmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, - T alpha, const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event sbmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, + T alpha, const T* a, int64_t lda, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::sbmv(func, queue, new_uplo, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } #define SBMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event sbmv(sycl::queue &queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ - const TYPE *a, int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, \ - int64_t incy, const std::vector &dependencies) { \ + sycl::event sbmv(sycl::queue& queue, uplo upper_lower, int64_t n, int64_t k, TYPE alpha, \ + const TYPE* a, int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, \ + int64_t incy, const std::vector& dependencies) { \ return sbmv(ROCBLAS_ROUTINE, queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, \ incy, dependencies); \ } @@ -3051,20 +3047,20 @@ SBMV_LAUNCHER_USM(double, rocblas_dsbmv) #undef SBMV_LAUNCHER_USM template -inline sycl::event symv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *a, int64_t lda, const T *x, int64_t incx, T beta, T *y, - int64_t incy, const std::vector &dependencies) { +inline sycl::event symv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* a, int64_t lda, const T* x, int64_t incx, T beta, T* y, + int64_t incy, const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::symv(func, queue, new_uplo, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } #define SYMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event symv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - int64_t lda, const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event symv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + int64_t lda, const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return symv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, \ dependencies); \ } @@ -3075,19 +3071,19 @@ SYMV_LAUNCHER_USM(double, rocblas_dsymv) #undef SYMV_LAUNCHER_USM template -inline sycl::event syr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, T *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event syr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, T* a, int64_t lda, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::syr(func, queue, new_uplo, n, alpha, x, incx, a, lda, dependencies); } #define SYR_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event syr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return syr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); \ } @@ -3100,20 +3096,20 @@ SYR_LAUNCHER_USM(std::complex, rocblas_zsyr) #undef SYR_LAUNCHER_USM template -inline sycl::event syr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, const T *y, int64_t incy, T *a, int64_t lda, - const std::vector &dependencies) { +inline sycl::event syr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, const T* y, int64_t incy, T* a, int64_t lda, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::syr2(func, queue, new_uplo, n, alpha, x, incx, y, incy, a, lda, dependencies); } #define SYR2_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, int64_t lda, \ - const std::vector &dependencies) { \ + sycl::event syr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, int64_t lda, \ + const std::vector& dependencies) { \ return syr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, \ dependencies); \ } @@ -3127,20 +3123,20 @@ SYR2_LAUNCHER_USM(std::complex, rocblas_zsyr2) #undef SYR2_LAUNCHER_USM template -inline sycl::event spmv(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *a, const T *x, int64_t incx, T beta, T *y, int64_t incy, - const std::vector &dependencies) { +inline sycl::event spmv(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* a, const T* x, int64_t incx, T beta, T* y, int64_t incy, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::spmv(func, queue, new_uplo, n, alpha, a, x, incx, beta, y, incy, dependencies); } #define SPMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event spmv(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *a, \ - const TYPE *x, int64_t incx, TYPE beta, TYPE *y, int64_t incy, \ - const std::vector &dependencies) { \ + sycl::event spmv(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* a, \ + const TYPE* x, int64_t incx, TYPE beta, TYPE* y, int64_t incy, \ + const std::vector& dependencies) { \ return spmv(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, \ dependencies); \ } @@ -3151,18 +3147,18 @@ SPMV_LAUNCHER_USM(double, rocblas_dspmv) #undef SPMV_LAUNCHER_USM template -inline sycl::event spr(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, T *a, - const std::vector &dependencies) { +inline sycl::event spr(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, T* a, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::spr(func, queue, new_uplo, n, alpha, x, incx, a, dependencies); } #define SPR_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event spr(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, TYPE *a, const std::vector &dependencies) { \ + sycl::event spr(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, TYPE* a, const std::vector& dependencies) { \ return spr(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, a, dependencies); \ } @@ -3172,19 +3168,19 @@ SPR_LAUNCHER_USM(double, rocblas_dspr) #undef SPR_LAUNCHER_USM template -inline sycl::event spr2(Func func, sycl::queue &queue, uplo upper_lower, int64_t n, T alpha, - const T *x, int64_t incx, const T *y, int64_t incy, T *a, - const std::vector &dependencies) { +inline sycl::event spr2(Func func, sycl::queue& queue, uplo upper_lower, int64_t n, T alpha, + const T* x, int64_t incx, const T* y, int64_t incy, T* a, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::spr2(func, queue, new_uplo, n, alpha, x, incx, y, incy, a, dependencies); } #define SPR2_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event spr2(sycl::queue &queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE *x, \ - int64_t incx, const TYPE *y, int64_t incy, TYPE *a, \ - const std::vector &dependencies) { \ + sycl::event spr2(sycl::queue& queue, uplo upper_lower, int64_t n, TYPE alpha, const TYPE* x, \ + int64_t incx, const TYPE* y, int64_t incy, TYPE* a, \ + const std::vector& dependencies) { \ return spr2(ROCBLAS_ROUTINE, queue, upper_lower, n, alpha, x, incx, y, incy, a, \ dependencies); \ } @@ -3195,20 +3191,20 @@ SPR2_LAUNCHER_USM(double, rocblas_dspr2) #undef SPR2_LAUNCHER_USM template -inline sycl::event tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tbmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, int64_t k, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { sycl::event done; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); } } @@ -3218,7 +3214,7 @@ inline sycl::event tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpo if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, x, n, incx); }); @@ -3229,22 +3225,22 @@ inline sycl::event tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpo } template -inline sycl::event tbmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const T *a, int64_t lda, T *x, - int64_t incx, const std::vector &dependencies) { +inline sycl::event tbmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, int64_t k, const T* a, int64_t lda, T* x, + int64_t incx, const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::tbmv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx, dependencies); } #define TBMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event tbmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, int64_t k, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, int64_t k, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tbmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, \ dependencies); \ } @@ -3257,20 +3253,20 @@ TBMV_LAUNCHER_USM(std::complex, rocblas_ztbmv) #undef TBMV_LAUNCHER_USM template -inline sycl::event tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tbsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, int64_t k, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { sycl::event done; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); } } @@ -3280,7 +3276,7 @@ inline sycl::event tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpo if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, x, n, incx); }); @@ -3291,22 +3287,22 @@ inline sycl::event tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpo } template -inline sycl::event tbsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, int64_t k, const T *a, int64_t lda, T *x, - int64_t incx, const std::vector &dependencies) { +inline sycl::event tbsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, int64_t k, const T* a, int64_t lda, T* x, + int64_t incx, const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::tbsv(func, queue, new_uplo, new_trans, unit_diag, n, k, a, lda, x, incx, dependencies); } #define TBSV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event tbsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, int64_t k, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, int64_t k, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tbsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, \ dependencies); \ } @@ -3319,19 +3315,19 @@ TBSV_LAUNCHER_USM(std::complex, rocblas_ztbsv) #undef TBSV_LAUNCHER_USM template -inline sycl::event tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, std::complex *x, - int64_t incx, const std::vector &dependencies) { +inline sycl::event tpmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const std::complex* a, std::complex* x, + int64_t incx, const std::vector& dependencies) { sycl::event done; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); } } @@ -3343,7 +3339,7 @@ inline sycl::event tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpo if (n > 0) { incx = std::abs(incx); - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, x, n, incx); }); @@ -3354,22 +3350,22 @@ inline sycl::event tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpo } template -inline sycl::event tpmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const T *a, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tpmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const T* a, T* x, int64_t incx, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::tpmv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx, dependencies); } #define TPMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event tpmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tpmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, incx, \ dependencies); \ } @@ -3382,19 +3378,19 @@ TPMV_LAUNCHER_USM(std::complex, rocblas_ztpmv) #undef TPMV_LAUNCHER_USM template -inline sycl::event tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, std::complex *x, - int64_t incx, const std::vector &dependencies) { +inline sycl::event tpsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const std::complex* a, std::complex* x, + int64_t incx, const std::vector& dependencies) { sycl::event done; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); } } @@ -3406,7 +3402,7 @@ inline sycl::event tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpo if (n > 0) { incx = std::abs(incx); - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, x, n, incx); }); @@ -3417,22 +3413,22 @@ inline sycl::event tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpo } template -inline sycl::event tpsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const T *a, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event tpsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const T* a, T* x, int64_t incx, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::tpsv(func, queue, new_uplo, new_trans, unit_diag, n, a, x, incx, dependencies); } #define TPSV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event tpsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return tpsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, x, incx, \ dependencies); \ } @@ -3445,20 +3441,20 @@ TPSV_LAUNCHER_USM(std::complex, rocblas_ztpsv) #undef TPSV_LAUNCHER_USM template -inline sycl::event trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event trmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { sycl::event done; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); } } @@ -3468,7 +3464,7 @@ inline sycl::event trmv(Func func, sycl::queue &queue, uplo upper_lower, transpo if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, x, n, incx); }); @@ -3479,22 +3475,22 @@ inline sycl::event trmv(Func func, sycl::queue &queue, uplo upper_lower, transpo } template -inline sycl::event trmv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const T *a, int64_t lda, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event trmv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const T* a, int64_t lda, T* x, int64_t incx, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::trmv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx, dependencies); } #define TRMV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trmv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return trmv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, \ dependencies); \ } @@ -3507,20 +3503,20 @@ TRMV_LAUNCHER_USM(std::complex, rocblas_ztrmv) #undef TRMV_LAUNCHER_USM template -inline sycl::event trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const std::complex *a, int64_t lda, - std::complex *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event trsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const std::complex* a, int64_t lda, + std::complex* x, int64_t incx, + const std::vector& dependencies) { sycl::event done; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - queue.submit([&](sycl::handler &cgh) { conj_vector(cgh, x, n, incx); }) + queue.submit([&](sycl::handler& cgh) { conj_vector(cgh, x, n, incx); }) .wait_and_throw(); } } @@ -3530,7 +3526,7 @@ inline sycl::event trsv(Func func, sycl::queue &queue, uplo upper_lower, transpo if (trans == oneapi::math::transpose::conjtrans) { if (n > 0) { - done = queue.submit([&](sycl::handler &cgh) { + done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); conj_vector(cgh, x, n, incx); }); @@ -3541,22 +3537,22 @@ inline sycl::event trsv(Func func, sycl::queue &queue, uplo upper_lower, transpo } template -inline sycl::event trsv(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, int64_t n, const T *a, int64_t lda, T *x, int64_t incx, - const std::vector &dependencies) { +inline sycl::event trsv(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, int64_t n, const T* a, int64_t lda, T* x, int64_t incx, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::trsv(func, queue, new_uplo, new_trans, unit_diag, n, a, lda, x, incx, dependencies); } #define TRSV_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trsv(sycl::queue &queue, uplo upper_lower, transpose trans, diag unit_diag, \ - int64_t n, const TYPE *a, int64_t lda, TYPE *x, int64_t incx, \ - const std::vector &dependencies) { \ + sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, \ + int64_t n, const TYPE* a, int64_t lda, TYPE* x, int64_t incx, \ + const std::vector& dependencies) { \ return trsv(ROCBLAS_ROUTINE, queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, \ dependencies); \ } diff --git a/src/blas/backends/rocblas/rocblas_level3.cpp b/src/blas/backends/rocblas/rocblas_level3.cpp index 904d83114..1ec25bb29 100644 --- a/src/blas/backends/rocblas/rocblas_level3.cpp +++ b/src/blas/backends/rocblas/rocblas_level3.cpp @@ -34,34 +34,34 @@ namespace column_major { // Buffer APIs template -inline void gemm(Func func, sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, T alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, int64_t ldc) { +inline void gemm(Func func, sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, int64_t k, T alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, k, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(transa), - get_rocblas_operation(transb), m, n, k, (rocDataType *)&alpha, - a_, lda, b_, ldb, (rocDataType *)&beta, c_, ldc); + get_rocblas_operation(transb), m, n, k, (rocDataType*)&alpha, a_, + lda, b_, ldb, (rocDataType*)&beta, c_, ldc); }); }); } #define GEMM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, TYPE beta, sycl::buffer &c, \ + void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, TYPE beta, sycl::buffer& c, \ int64_t ldc) { \ gemm(ROCBLAS_ROUTINE, queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, \ ldc); \ @@ -77,9 +77,9 @@ GEMM_LAUNCHER(std::complex, rocblas_zgemm) template inline void gemm_ex(Func func, DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, COMPUTETYPE CT, - sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, T_S alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, T_S beta, sycl::buffer &c, + sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, T_S alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, T_S beta, sycl::buffer& c, int64_t ldc) { using rocDataType_A = typename RocEquivalentType::Type; using rocDataType_B = typename RocEquivalentType::Type; @@ -87,30 +87,30 @@ inline void gemm_ex(Func func, DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C using rocDataType_S = typename RocEquivalentType::Type; overflow_check(m, n, k, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(transa), - get_rocblas_operation(transb), m, n, k, (rocDataType_S *)&alpha, - a_, DT_A, lda, b_, DT_B, ldb, (rocDataType_S *)&beta, c_, DT_C, - ldc, c_, DT_C, ldc, CT, rocblas_gemm_algo_standard, 0, 0); + get_rocblas_operation(transb), m, n, k, (rocDataType_S*)&alpha, a_, + DT_A, lda, b_, DT_B, ldb, (rocDataType_S*)&beta, c_, DT_C, ldc, c_, + DT_C, ldc, CT, rocblas_gemm_algo_standard, 0, 0); }); }); } #define GEMM_EX_LAUNCHER(TYPE_A, TYPE_B, TYPE_C, TYPE_S, ROCBLAS_ROUTINE, ROCMDATATYPE_A, \ ROCMDATATYPE_B, ROCMDATATYPE_C, ROCMCOMPUTETYPE) \ - void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_S alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, TYPE_S beta, sycl::buffer &c, \ + void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_S alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, TYPE_S beta, sycl::buffer& c, \ int64_t ldc) { \ gemm_ex(ROCBLAS_ROUTINE, ROCMDATATYPE_A, ROCMDATATYPE_B, ROCMDATATYPE_C, ROCMCOMPUTETYPE, \ queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); \ @@ -130,34 +130,34 @@ GEMM_EX_LAUNCHER(bfloat16, bfloat16, bfloat16, float, rocblas_gemm_ex, rocblas_d #undef GEMM_EX_LAUNCHER template -inline void symm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, T beta, sycl::buffer &c, int64_t ldc) { +inline void symm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), - get_rocblas_fill_mode(upper_lower), m, n, (rocDataType *)&alpha, - a_, lda, b_, ldb, (rocDataType *)&beta, c_, ldc); + get_rocblas_fill_mode(upper_lower), m, n, (rocDataType*)&alpha, a_, + lda, b_, ldb, (rocDataType*)&beta, c_, ldc); }); }); } #define SYMM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ symm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, \ c, ldc); \ } @@ -170,34 +170,34 @@ SYMM_LAUNCHER(std::complex, rocblas_zsymm) #undef SYMM_LAUNCHER template -inline void hemm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, T beta, sycl::buffer &c, int64_t ldc) { +inline void hemm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), - get_rocblas_fill_mode(upper_lower), m, n, (rocDataType *)&alpha, - a_, lda, b_, ldb, (rocDataType *)&beta, c_, ldc); + get_rocblas_fill_mode(upper_lower), m, n, (rocDataType*)&alpha, a_, + lda, b_, ldb, (rocDataType*)&beta, c_, ldc); }); }); } #define HEMM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ hemm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, \ c, ldc); \ } @@ -208,32 +208,32 @@ HEMM_LAUNCHER(std::complex, rocblas_zhemm) #undef HEMM_LAUNCHER template -inline void syrk(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, T alpha, sycl::buffer &a, int64_t lda, T beta, - sycl::buffer &c, int64_t ldc) { +inline void syrk(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, T alpha, sycl::buffer& a, int64_t lda, T beta, + sycl::buffer& c, int64_t ldc) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), n, k, (rocDataType *)&alpha, a_, - lda, (rocDataType *)&beta, c_, ldc); + get_rocblas_operation(trans), n, k, (rocDataType*)&alpha, a_, lda, + (rocDataType*)&beta, c_, ldc); }); }); } #define SYRK_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, sycl::buffer &a, int64_t lda, TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, sycl::buffer& a, int64_t lda, TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ syrk(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); \ } @@ -245,33 +245,33 @@ SYRK_LAUNCHER(std::complex, rocblas_zsyrk) #undef SYRK_LAUNCHER template -inline void herk(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, ScalarType alpha, sycl::buffer &a, int64_t lda, - ScalarType beta, sycl::buffer &c, int64_t ldc) { +inline void herk(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, ScalarType alpha, sycl::buffer& a, int64_t lda, + ScalarType beta, sycl::buffer& c, int64_t ldc) { using rocDataType = typename RocEquivalentType::Type; using rocScalarType = typename RocEquivalentType::Type; overflow_check(n, k, lda, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), n, k, (rocScalarType *)&alpha, a_, - lda, (rocScalarType *)&beta, c_, ldc); + get_rocblas_operation(trans), n, k, (rocScalarType*)&alpha, a_, lda, + (rocScalarType*)&beta, c_, ldc); }); }); } #define HERK_LAUNCHER(DATA_TYPE, SCALAR_TYPE, ROCBLAS_ROUTINE) \ - void herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - SCALAR_TYPE alpha, sycl::buffer &a, int64_t lda, SCALAR_TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + SCALAR_TYPE alpha, sycl::buffer& a, int64_t lda, SCALAR_TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ herk(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); \ } @@ -281,34 +281,34 @@ HERK_LAUNCHER(std::complex, double, rocblas_zherk) #undef HERK_LAUNCHER template -inline void syr2k(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, T beta, sycl::buffer &c, int64_t ldc) { +inline void syr2k(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), n, k, (rocDataType *)&alpha, a_, - lda, b_, ldb, (rocDataType *)&beta, c_, ldc); + get_rocblas_operation(trans), n, k, (rocDataType*)&alpha, a_, lda, + b_, ldb, (rocDataType*)&beta, c_, ldc); }); }); } #define SYR2K_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ syr2k(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, \ ldc); \ } @@ -321,37 +321,37 @@ SYR2K_LAUNCHER(std::complex, rocblas_zsyr2k) #undef SYR2K_LAUNCHER template -inline void her2k(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, DataType alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, ScalarType beta, - sycl::buffer &c, int64_t ldc) { +inline void her2k(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, DataType alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, ScalarType beta, + sycl::buffer& c, int64_t ldc) { using rocDataType = typename RocEquivalentType::Type; using rocScalarType = typename RocEquivalentType::Type; overflow_check(n, k, lda, ldb, ldc); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), n, k, (rocDataType *)&alpha, a_, - lda, b_, ldb, (rocScalarType *)&beta, c_, ldc); + get_rocblas_operation(trans), n, k, (rocDataType*)&alpha, a_, lda, + b_, ldb, (rocScalarType*)&beta, c_, ldc); }); }); } #define HER2K_LAUNCHER(DATA_TYPE, SCALAR_TYPE, ROCBLAS_ROUTINE) \ - void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - DATA_TYPE alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, SCALAR_TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + DATA_TYPE alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, SCALAR_TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ her2k(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, \ ldc); \ } @@ -366,26 +366,26 @@ HER2K_LAUNCHER(std::complex, double, rocblas_zher2k) // separated from the B matrix. It is possible to use B instead of C, but this // will slow-down the code. template -inline void trmm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, T alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb) { +inline void trmm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, T alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); rocblas_status err; #if ROCBLAS_VERSION_MAJOR >= 4 rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), - get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - m, n, (rocDataType *)&alpha, a_, lda, b_, ldb, b_, ldb); + get_rocblas_fill_mode(upper_lower), get_rocblas_operation(trans), + get_rocblas_diag_type(unit_diag), m, n, (rocDataType*)&alpha, a_, + lda, b_, ldb, b_, ldb); #else rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), get_rocblas_fill_mode(upper_lower), @@ -397,9 +397,9 @@ inline void trmm(Func func, sycl::queue &queue, side left_right, uplo upper_lowe } #define TRMM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer &a, \ - int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer& a, \ + int64_t lda, sycl::buffer& b, int64_t ldb) { \ trmm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, \ lda, b, ldb); \ } @@ -412,33 +412,33 @@ TRMM_LAUNCHER(std::complex, rocblas_ztrmm) #undef TRMM_LAUNCHER template -inline void trsm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, T alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb) { +inline void trsm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, T alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), - get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - m, n, (rocDataType *)&alpha, a_, lda, b_, ldb); + get_rocblas_fill_mode(upper_lower), get_rocblas_operation(trans), + get_rocblas_diag_type(unit_diag), m, n, (rocDataType*)&alpha, a_, + lda, b_, ldb); }); }); } #define TRSM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer &a, \ - int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer& a, \ + int64_t lda, sycl::buffer& b, int64_t ldb) { \ trsm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, \ lda, b, ldb); \ } @@ -453,25 +453,25 @@ TRSM_LAUNCHER(std::complex, rocblas_ztrsm) // USM APIs template -inline sycl::event gemm(Func func, sycl::queue &queue, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, T alpha, const T *a, int64_t lda, - const T *b, int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event gemm(Func func, sycl::queue& queue, transpose transa, transpose transb, + int64_t m, int64_t n, int64_t k, T alpha, const T* a, int64_t lda, + const T* b, int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, k, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(transa), - get_rocblas_operation(transb), m, n, k, (rocDataType *)&alpha, - a_, lda, b_, ldb, (rocDataType *)&beta, c_, ldc); + get_rocblas_operation(transb), m, n, k, (rocDataType*)&alpha, a_, + lda, b_, ldb, (rocDataType*)&beta, c_, ldc); }); }); @@ -479,10 +479,10 @@ inline sycl::event gemm(Func func, sycl::queue &queue, transpose transa, transpo } #define GEMM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, \ - int64_t ldb, TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, \ + int64_t ldb, TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return gemm(ROCBLAS_ROUTINE, queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, \ c, ldc, dependencies); \ } @@ -497,29 +497,29 @@ GEMM_LAUNCHER_USM(std::complex, rocblas_zgemm) template inline sycl::event gemm_ex(Func func, DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, - COMPUTETYPE CT, sycl::queue &queue, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, T_S alpha, const T_A *a, int64_t lda, - const T_B *b, int64_t ldb, T_S beta, T_C *c, int64_t ldc, - const std::vector &dependencies) { + COMPUTETYPE CT, sycl::queue& queue, transpose transa, transpose transb, + int64_t m, int64_t n, int64_t k, T_S alpha, const T_A* a, int64_t lda, + const T_B* b, int64_t ldb, T_S beta, T_C* c, int64_t ldc, + const std::vector& dependencies) { using rocDataType_A = typename RocEquivalentType::Type; using rocDataType_B = typename RocEquivalentType::Type; using rocDataType_C = typename RocEquivalentType::Type; using rocDataType_S = typename RocEquivalentType::Type; overflow_check(m, n, k, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_operation(transa), - get_rocblas_operation(transb), m, n, k, (rocDataType_S *)&alpha, - a_, DT_A, lda, b_, DT_B, ldb, (rocDataType_S *)&beta, c_, DT_C, - ldc, c_, DT_C, ldc, CT, rocblas_gemm_algo_standard, 0, 0); + get_rocblas_operation(transb), m, n, k, (rocDataType_S*)&alpha, a_, + DT_A, lda, b_, DT_B, ldb, (rocDataType_S*)&beta, c_, DT_C, ldc, c_, + DT_C, ldc, CT, rocblas_gemm_algo_standard, 0, 0); }); }); @@ -528,10 +528,10 @@ inline sycl::event gemm_ex(Func func, DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE #define GEMM_EX_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S, ROCBLAS_ROUTINE, ROCMDATATYPE_A, \ ROCMDATATYPE_B, ROCMDATATYPE_C, ROCMCOMPUTETYPE) \ - sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_S alpha, const TYPE_A *a, int64_t lda, const TYPE_B *b, \ - int64_t ldb, TYPE_S beta, TYPE_C *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_S alpha, const TYPE_A* a, int64_t lda, const TYPE_B* b, \ + int64_t ldb, TYPE_S beta, TYPE_C* c, int64_t ldc, \ + const std::vector& dependencies) { \ return gemm_ex(ROCBLAS_ROUTINE, ROCMDATATYPE_A, ROCMDATATYPE_B, ROCMDATATYPE_C, \ ROCMCOMPUTETYPE, queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, \ beta, c, ldc, dependencies); \ @@ -551,24 +551,24 @@ GEMM_EX_LAUNCHER_USM(bfloat16, bfloat16, bfloat16, float, rocblas_gemm_ex, rocbl #undef GEMM_EX_LAUNCHER_USM template -inline sycl::event symm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, - int64_t n, T alpha, const T *a, int64_t lda, const T *b, int64_t ldb, - T beta, T *c, int64_t ldc, const std::vector &dependencies) { +inline sycl::event symm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, + int64_t n, T alpha, const T* a, int64_t lda, const T* b, int64_t ldb, + T beta, T* c, int64_t ldc, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), - get_rocblas_fill_mode(upper_lower), m, n, (rocDataType *)&alpha, - a_, lda, b_, ldb, (rocDataType *)&beta, c_, ldc); + get_rocblas_fill_mode(upper_lower), m, n, (rocDataType*)&alpha, a_, + lda, b_, ldb, (rocDataType*)&beta, c_, ldc); }); }); @@ -576,10 +576,10 @@ inline sycl::event symm(Func func, sycl::queue &queue, side left_right, uplo upp } #define SYMM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return symm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, \ beta, c, ldc, dependencies); \ } @@ -592,24 +592,24 @@ SYMM_LAUNCHER_USM(std::complex, rocblas_zsymm) #undef SYMM_LAUNCHER_USM template -inline sycl::event hemm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, - int64_t n, T alpha, const T *a, int64_t lda, const T *b, int64_t ldb, - T beta, T *c, int64_t ldc, const std::vector &dependencies) { +inline sycl::event hemm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, + int64_t n, T alpha, const T* a, int64_t lda, const T* b, int64_t ldb, + T beta, T* c, int64_t ldc, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), - get_rocblas_fill_mode(upper_lower), m, n, (rocDataType *)&alpha, - a_, lda, b_, ldb, (rocDataType *)&beta, c_, ldc); + get_rocblas_fill_mode(upper_lower), m, n, (rocDataType*)&alpha, a_, + lda, b_, ldb, (rocDataType*)&beta, c_, ldc); }); }); @@ -617,10 +617,10 @@ inline sycl::event hemm(Func func, sycl::queue &queue, side left_right, uplo upp } #define HEMM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return hemm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, \ beta, c, ldc, dependencies); \ } @@ -631,23 +631,23 @@ HEMM_LAUNCHER_USM(std::complex, rocblas_zhemm) #undef HEMM_LAUNCHER_USM template -inline sycl::event syrk(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, T alpha, const T *a, int64_t lda, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event syrk(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, T alpha, const T* a, int64_t lda, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto c_ = reinterpret_cast(c); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), n, k, (rocDataType *)&alpha, a_, - lda, (rocDataType *)&beta, c_, ldc); + get_rocblas_operation(trans), n, k, (rocDataType*)&alpha, a_, lda, + (rocDataType*)&beta, c_, ldc); }); }); @@ -655,9 +655,9 @@ inline sycl::event syrk(Func func, sycl::queue &queue, uplo upper_lower, transpo } #define SYRK_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, const TYPE *a, int64_t lda, TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, const TYPE* a, int64_t lda, TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return syrk(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, \ dependencies); \ } @@ -670,25 +670,25 @@ SYRK_LAUNCHER_USM(std::complex, rocblas_zsyrk) #undef SYRK_LAUNCHER_USM template -inline sycl::event herk(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, const ScalarType alpha, const DataType *a, int64_t lda, - const ScalarType beta, DataType *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event herk(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, const ScalarType alpha, const DataType* a, int64_t lda, + const ScalarType beta, DataType* c, int64_t ldc, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; using rocScalarType = typename RocEquivalentType::Type; overflow_check(n, k, lda, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto c_ = reinterpret_cast(c); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), n, k, (rocScalarType *)&alpha, a_, - lda, (rocScalarType *)&beta, c_, ldc); + get_rocblas_operation(trans), n, k, (rocScalarType*)&alpha, a_, lda, + (rocScalarType*)&beta, c_, ldc); }); }); @@ -696,10 +696,10 @@ inline sycl::event herk(Func func, sycl::queue &queue, uplo upper_lower, transpo } #define HERK_LAUNCHER_USM(DATA_TYPE, SCALAR_TYPE, ROCBLAS_ROUTINE) \ - sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - const SCALAR_TYPE alpha, const DATA_TYPE *a, int64_t lda, \ - const SCALAR_TYPE beta, DATA_TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + const SCALAR_TYPE alpha, const DATA_TYPE* a, int64_t lda, \ + const SCALAR_TYPE beta, DATA_TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return herk(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, \ dependencies); \ } @@ -710,25 +710,25 @@ HERK_LAUNCHER_USM(std::complex, double, rocblas_zherk) #undef HERK_LAUNCHER_USM template -inline sycl::event syr2k(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - int64_t n, int64_t k, T alpha, const T *a, int64_t lda, const T *b, - int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event syr2k(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + int64_t n, int64_t k, T alpha, const T* a, int64_t lda, const T* b, + int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(n, k, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), n, k, (rocDataType *)&alpha, a_, - lda, b_, ldb, (rocDataType *)&beta, c_, ldc); + get_rocblas_operation(trans), n, k, (rocDataType*)&alpha, a_, lda, + b_, ldb, (rocDataType*)&beta, c_, ldc); }); }); @@ -736,10 +736,10 @@ inline sycl::event syr2k(Func func, sycl::queue &queue, uplo upper_lower, transp } #define SYR2K_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return syr2k(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, \ beta, c, ldc, dependencies); \ } @@ -752,26 +752,26 @@ SYR2K_LAUNCHER_USM(std::complex, rocblas_zsyr2k) #undef SYR2K_LAUNCHER_USM template -inline sycl::event her2k(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - int64_t n, int64_t k, const DataType alpha, const DataType *a, int64_t lda, - const DataType *b, int64_t ldb, const ScalarType beta, DataType *c, - int64_t ldc, const std::vector &dependencies) { +inline sycl::event her2k(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + int64_t n, int64_t k, const DataType alpha, const DataType* a, int64_t lda, + const DataType* b, int64_t ldb, const ScalarType beta, DataType* c, + int64_t ldc, const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; using rocScalarType = typename RocEquivalentType::Type; overflow_check(n, k, lda, ldb, ldc); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto c_ = reinterpret_cast(c); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), n, k, (rocDataType *)&alpha, a_, - lda, b_, ldb, (rocScalarType *)&beta, c_, ldc); + get_rocblas_operation(trans), n, k, (rocDataType*)&alpha, a_, lda, + b_, ldb, (rocScalarType*)&beta, c_, ldc); }); }); @@ -779,10 +779,10 @@ inline sycl::event her2k(Func func, sycl::queue &queue, uplo upper_lower, transp } #define HER2K_LAUNCHER_USM(DATA_TYPE, SCALAR_TYPE, ROCBLAS_ROUTINE) \ - sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - const DATA_TYPE alpha, const DATA_TYPE *a, int64_t lda, const DATA_TYPE *b, \ - int64_t ldb, const SCALAR_TYPE beta, DATA_TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + const DATA_TYPE alpha, const DATA_TYPE* a, int64_t lda, const DATA_TYPE* b, \ + int64_t ldb, const SCALAR_TYPE beta, DATA_TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return her2k(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, \ beta, c, ldc, dependencies); \ } @@ -797,26 +797,26 @@ HER2K_LAUNCHER_USM(std::complex, double, rocblas_zher2k) // separated from the B matrix. It is possible to use B instead of C, but this // will slow-down the code. template -inline sycl::event trmm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, const T *a, - int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { +inline sycl::event trmm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, const T* a, + int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); rocblas_status err; #if ROCBLAS_VERSION_MAJOR >= 4 rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), - get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - m, n, (rocDataType *)&alpha, a_, lda, b_, ldb, b_, ldb); + get_rocblas_fill_mode(upper_lower), get_rocblas_operation(trans), + get_rocblas_diag_type(unit_diag), m, n, (rocDataType*)&alpha, a_, + lda, b_, ldb, b_, ldb); #else rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), get_rocblas_fill_mode(upper_lower), @@ -830,9 +830,9 @@ inline sycl::event trmm(Func func, sycl::queue &queue, side left_right, uplo upp } #define TRMM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE *a, int64_t lda, \ - TYPE *b, int64_t ldb, const std::vector &dependencies) { \ + sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE* a, int64_t lda, \ + TYPE* b, int64_t ldb, const std::vector& dependencies) { \ return trmm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, \ alpha, a, lda, b, ldb, dependencies); \ } @@ -845,25 +845,25 @@ TRMM_LAUNCHER_USM(std::complex, rocblas_ztrmm) #undef TRMM_LAUNCHER_USM template -inline sycl::event trsm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, const T *a, - int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { +inline sycl::event trsm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, const T* a, + int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { using rocDataType = typename RocEquivalentType::Type; overflow_check(m, n, lda, ldb); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler &sc) { + onemath_rocblas_host_task(cgh, queue, [=](RocblasScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); rocblas_status err; rocblas_native_func(func, err, handle, get_rocblas_side_mode(left_right), - get_rocblas_fill_mode(upper_lower), - get_rocblas_operation(trans), get_rocblas_diag_type(unit_diag), - m, n, (rocDataType *)&alpha, a_, lda, b_, ldb); + get_rocblas_fill_mode(upper_lower), get_rocblas_operation(trans), + get_rocblas_diag_type(unit_diag), m, n, (rocDataType*)&alpha, a_, + lda, b_, ldb); }); }); @@ -871,9 +871,9 @@ inline sycl::event trsm(Func func, sycl::queue &queue, side left_right, uplo upp } #define TRSM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE *a, int64_t lda, \ - TYPE *b, int64_t ldb, const std::vector &dependencies) { \ + sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE* a, int64_t lda, \ + TYPE* b, int64_t ldb, const std::vector& dependencies) { \ return trsm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, \ alpha, a, lda, b, ldb, dependencies); \ } @@ -892,9 +892,9 @@ namespace row_major { // Buffer APIs template -inline void gemm(Func func, sycl::queue &queue, transpose transa, transpose transb, int64_t m, - int64_t n, int64_t k, T alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, T beta, sycl::buffer &c, int64_t ldc) { +inline void gemm(Func func, sycl::queue& queue, transpose transa, transpose transb, int64_t m, + int64_t n, int64_t k, T alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { auto new_transa = transb; auto new_transb = transa; @@ -903,9 +903,9 @@ inline void gemm(Func func, sycl::queue &queue, transpose transa, transpose tran } #define GEMM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, TYPE beta, sycl::buffer &c, \ + void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, TYPE beta, sycl::buffer& c, \ int64_t ldc) { \ gemm(ROCBLAS_ROUTINE, queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, \ ldc); \ @@ -921,9 +921,9 @@ GEMM_LAUNCHER(std::complex, rocblas_zgemm) template inline void gemm_ex(Func func, DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, COMPUTETYPE CT, - sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, - int64_t k, T_S alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, T_S beta, sycl::buffer &c, + sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, + int64_t k, T_S alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, T_S beta, sycl::buffer& c, int64_t ldc) { auto new_transa = transb; auto new_transb = transa; @@ -934,9 +934,9 @@ inline void gemm_ex(Func func, DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C #define GEMM_EX_LAUNCHER(TYPE_A, TYPE_B, TYPE_C, TYPE_S, ROCBLAS_ROUTINE, ROCMDATATYPE_A, \ ROCMDATATYPE_B, ROCMDATATYPE_C, ROCMCOMPUTETYPE) \ - void gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_S alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, TYPE_S beta, sycl::buffer &c, \ + void gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_S alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, TYPE_S beta, sycl::buffer& c, \ int64_t ldc) { \ gemm_ex(ROCBLAS_ROUTINE, ROCMDATATYPE_A, ROCMDATATYPE_B, ROCMDATATYPE_C, ROCMCOMPUTETYPE, \ queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); \ @@ -956,21 +956,21 @@ GEMM_EX_LAUNCHER(bfloat16, bfloat16, bfloat16, float, rocblas_gemm_ex, rocblas_d #undef GEMM_EX_LAUNCHER template -inline void symm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, T beta, sycl::buffer &c, int64_t ldc) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; +inline void symm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::symm(func, queue, new_side, new_uplo, n, m, alpha, a, lda, b, ldb, beta, c, ldc); } #define SYMM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ symm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, \ c, ldc); \ } @@ -983,21 +983,21 @@ SYMM_LAUNCHER(std::complex, rocblas_zsymm) #undef SYMM_LAUNCHER template -inline void hemm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, - int64_t n, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, T beta, sycl::buffer &c, int64_t ldc) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; +inline void hemm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, + int64_t n, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::hemm(func, queue, new_side, new_uplo, n, m, alpha, a, lda, b, ldb, beta, c, ldc); } #define HEMM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ hemm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, \ c, ldc); \ } @@ -1008,21 +1008,21 @@ HEMM_LAUNCHER(std::complex, rocblas_zhemm) #undef HEMM_LAUNCHER template -inline void syrk(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, T alpha, sycl::buffer &a, int64_t lda, T beta, - sycl::buffer &c, int64_t ldc) { +inline void syrk(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, T alpha, sycl::buffer& a, int64_t lda, T beta, + sycl::buffer& c, int64_t ldc) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::syrk(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, beta, c, ldc); } #define SYRK_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, sycl::buffer &a, int64_t lda, TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, sycl::buffer& a, int64_t lda, TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ syrk(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); \ } @@ -1034,21 +1034,21 @@ SYRK_LAUNCHER(std::complex, rocblas_zsyrk) #undef SYRK_LAUNCHER template -inline void herk(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, ScalarType alpha, sycl::buffer &a, int64_t lda, - ScalarType beta, sycl::buffer &c, int64_t ldc) { +inline void herk(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, ScalarType alpha, sycl::buffer& a, int64_t lda, + ScalarType beta, sycl::buffer& c, int64_t ldc) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::conjtrans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::herk(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, beta, c, ldc); } #define HERK_LAUNCHER(DATA_TYPE, SCALAR_TYPE, ROCBLAS_ROUTINE) \ - void herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - SCALAR_TYPE alpha, sycl::buffer &a, int64_t lda, SCALAR_TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + SCALAR_TYPE alpha, sycl::buffer& a, int64_t lda, SCALAR_TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ herk(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); \ } @@ -1058,22 +1058,22 @@ HERK_LAUNCHER(std::complex, double, rocblas_zherk) #undef HERK_LAUNCHER template -inline void syr2k(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, T alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, - int64_t ldb, T beta, sycl::buffer &c, int64_t ldc) { +inline void syr2k(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, T alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, + int64_t ldb, T beta, sycl::buffer& c, int64_t ldc) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; column_major::syr2k(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } #define SYR2K_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, sycl::buffer &a, int64_t lda, sycl::buffer &b, \ - int64_t ldb, TYPE beta, sycl::buffer &c, int64_t ldc) { \ + void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, sycl::buffer& a, int64_t lda, sycl::buffer& b, \ + int64_t ldb, TYPE beta, sycl::buffer& c, int64_t ldc) { \ syr2k(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, \ ldc); \ } @@ -1086,14 +1086,14 @@ SYR2K_LAUNCHER(std::complex, rocblas_zsyr2k) #undef SYR2K_LAUNCHER template -inline void her2k(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, DataType alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb, ScalarType beta, - sycl::buffer &c, int64_t ldc) { +inline void her2k(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, DataType alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb, ScalarType beta, + sycl::buffer& c, int64_t ldc) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::conjtrans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; auto new_alpha = std::conj(alpha); column_major::her2k(func, queue, new_uplo, new_trans, n, k, new_alpha, a, lda, b, ldb, beta, c, @@ -1101,10 +1101,10 @@ inline void her2k(Func func, sycl::queue &queue, uplo upper_lower, transpose tra } #define HER2K_LAUNCHER(DATA_TYPE, SCALAR_TYPE, ROCBLAS_ROUTINE) \ - void her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - DATA_TYPE alpha, sycl::buffer &a, int64_t lda, \ - sycl::buffer &b, int64_t ldb, SCALAR_TYPE beta, \ - sycl::buffer &c, int64_t ldc) { \ + void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + DATA_TYPE alpha, sycl::buffer& a, int64_t lda, \ + sycl::buffer& b, int64_t ldb, SCALAR_TYPE beta, \ + sycl::buffer& c, int64_t ldc) { \ her2k(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, \ ldc); \ } @@ -1119,22 +1119,22 @@ HER2K_LAUNCHER(std::complex, double, rocblas_zher2k) // separated from the B matrix. It is possible to use B instead of C, but this // will slow-down the code. template -inline void trmm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, T alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; +inline void trmm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, T alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb) { + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::trmm(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, b, ldb); } #define TRMM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer &a, \ - int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer& a, \ + int64_t lda, sycl::buffer& b, int64_t ldb) { \ trmm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, \ lda, b, ldb); \ } @@ -1147,22 +1147,22 @@ TRMM_LAUNCHER(std::complex, rocblas_ztrmm) #undef TRMM_LAUNCHER template -inline void trsm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, int64_t m, int64_t n, T alpha, sycl::buffer &a, int64_t lda, - sycl::buffer &b, int64_t ldb) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; +inline void trsm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, + diag unit_diag, int64_t m, int64_t n, T alpha, sycl::buffer& a, int64_t lda, + sycl::buffer& b, int64_t ldb) { + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; column_major::trsm(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, b, ldb); } #define TRSM_LAUNCHER(TYPE, ROCBLAS_ROUTINE) \ - void trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer &a, \ - int64_t lda, sycl::buffer &b, int64_t ldb) { \ + void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, sycl::buffer& a, \ + int64_t lda, sycl::buffer& b, int64_t ldb) { \ trsm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, \ lda, b, ldb); \ } @@ -1177,10 +1177,10 @@ TRSM_LAUNCHER(std::complex, rocblas_ztrsm) // USM APIs template -inline sycl::event gemm(Func func, sycl::queue &queue, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, T alpha, const T *a, int64_t lda, - const T *b, int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event gemm(Func func, sycl::queue& queue, transpose transa, transpose transb, + int64_t m, int64_t n, int64_t k, T alpha, const T* a, int64_t lda, + const T* b, int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { auto new_transa = transb; auto new_transb = transa; @@ -1189,10 +1189,10 @@ inline sycl::event gemm(Func func, sycl::queue &queue, transpose transa, transpo } #define GEMM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, \ - int64_t ldb, TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, \ + int64_t ldb, TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return gemm(ROCBLAS_ROUTINE, queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, \ c, ldc, dependencies); \ } @@ -1207,10 +1207,10 @@ GEMM_LAUNCHER_USM(std::complex, rocblas_zgemm) template inline sycl::event gemm_ex(Func func, DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE_C DT_C, - COMPUTETYPE CT, sycl::queue &queue, transpose transa, transpose transb, - int64_t m, int64_t n, int64_t k, T_S alpha, const T_A *a, int64_t lda, - const T_B *b, int64_t ldb, T_S beta, T_C *c, int64_t ldc, - const std::vector &dependencies) { + COMPUTETYPE CT, sycl::queue& queue, transpose transa, transpose transb, + int64_t m, int64_t n, int64_t k, T_S alpha, const T_A* a, int64_t lda, + const T_B* b, int64_t ldb, T_S beta, T_C* c, int64_t ldc, + const std::vector& dependencies) { auto new_transa = transb; auto new_transb = transa; @@ -1220,10 +1220,10 @@ inline sycl::event gemm_ex(Func func, DATATYPE_A DT_A, DATATYPE_B DT_B, DATATYPE #define GEMM_EX_LAUNCHER_USM(TYPE_A, TYPE_B, TYPE_C, TYPE_S, ROCBLAS_ROUTINE, ROCMDATATYPE_A, \ ROCMDATATYPE_B, ROCMDATATYPE_C, ROCMCOMPUTETYPE) \ - sycl::event gemm(sycl::queue &queue, transpose transa, transpose transb, int64_t m, int64_t n, \ - int64_t k, TYPE_S alpha, const TYPE_A *a, int64_t lda, const TYPE_B *b, \ - int64_t ldb, TYPE_S beta, TYPE_C *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, int64_t m, int64_t n, \ + int64_t k, TYPE_S alpha, const TYPE_A* a, int64_t lda, const TYPE_B* b, \ + int64_t ldb, TYPE_S beta, TYPE_C* c, int64_t ldc, \ + const std::vector& dependencies) { \ return gemm_ex(ROCBLAS_ROUTINE, ROCMDATATYPE_A, ROCMDATATYPE_B, ROCMDATATYPE_C, \ ROCMCOMPUTETYPE, queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, \ beta, c, ldc, dependencies); \ @@ -1243,23 +1243,23 @@ GEMM_EX_LAUNCHER_USM(bfloat16, bfloat16, bfloat16, float, rocblas_gemm_ex, rocbl #undef GEMM_EX_LAUNCHER_USM template -inline sycl::event symm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, - int64_t n, T alpha, const T *a, int64_t lda, const T *b, int64_t ldb, - T beta, T *c, int64_t ldc, const std::vector &dependencies) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; +inline sycl::event symm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, + int64_t n, T alpha, const T* a, int64_t lda, const T* b, int64_t ldb, + T beta, T* c, int64_t ldc, const std::vector& dependencies) { + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::symm(func, queue, new_side, new_uplo, n, m, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } #define SYMM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event symm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return symm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, \ beta, c, ldc, dependencies); \ } @@ -1272,23 +1272,23 @@ SYMM_LAUNCHER_USM(std::complex, rocblas_zsymm) #undef SYMM_LAUNCHER_USM template -inline sycl::event hemm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, - int64_t n, T alpha, const T *a, int64_t lda, const T *b, int64_t ldb, - T beta, T *c, int64_t ldc, const std::vector &dependencies) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; +inline sycl::event hemm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, + int64_t n, T alpha, const T* a, int64_t lda, const T* b, int64_t ldb, + T beta, T* c, int64_t ldc, const std::vector& dependencies) { + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::hemm(func, queue, new_side, new_uplo, n, m, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } #define HEMM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event hemm(sycl::queue &queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, int64_t m, int64_t n, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return hemm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, \ beta, c, ldc, dependencies); \ } @@ -1299,22 +1299,22 @@ HEMM_LAUNCHER_USM(std::complex, rocblas_zhemm) #undef HEMM_LAUNCHER_USM template -inline sycl::event syrk(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, T alpha, const T *a, int64_t lda, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event syrk(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, T alpha, const T* a, int64_t lda, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::syrk(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } #define SYRK_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syrk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, const TYPE *a, int64_t lda, TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, const TYPE* a, int64_t lda, TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return syrk(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, \ dependencies); \ } @@ -1327,24 +1327,24 @@ SYRK_LAUNCHER_USM(std::complex, rocblas_zsyrk) #undef SYRK_LAUNCHER_USM template -inline sycl::event herk(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, - int64_t k, const ScalarType alpha, const DataType *a, int64_t lda, - const ScalarType beta, DataType *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event herk(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, + int64_t k, const ScalarType alpha, const DataType* a, int64_t lda, + const ScalarType beta, DataType* c, int64_t ldc, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::conjtrans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::herk(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } #define HERK_LAUNCHER_USM(DATA_TYPE, SCALAR_TYPE, ROCBLAS_ROUTINE) \ - sycl::event herk(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - const SCALAR_TYPE alpha, const DATA_TYPE *a, int64_t lda, \ - const SCALAR_TYPE beta, DATA_TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + const SCALAR_TYPE alpha, const DATA_TYPE* a, int64_t lda, \ + const SCALAR_TYPE beta, DATA_TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return herk(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, \ dependencies); \ } @@ -1355,24 +1355,24 @@ HERK_LAUNCHER_USM(std::complex, double, rocblas_zherk) #undef HERK_LAUNCHER_USM template -inline sycl::event syr2k(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - int64_t n, int64_t k, T alpha, const T *a, int64_t lda, const T *b, - int64_t ldb, T beta, T *c, int64_t ldc, - const std::vector &dependencies) { +inline sycl::event syr2k(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + int64_t n, int64_t k, T alpha, const T* a, int64_t lda, const T* b, + int64_t ldb, T beta, T* c, int64_t ldc, + const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::trans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; return column_major::syr2k(func, queue, new_uplo, new_trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } #define SYR2K_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event syr2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - TYPE alpha, const TYPE *a, int64_t lda, const TYPE *b, int64_t ldb, \ - TYPE beta, TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + TYPE alpha, const TYPE* a, int64_t lda, const TYPE* b, int64_t ldb, \ + TYPE beta, TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return syr2k(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, \ beta, c, ldc, dependencies); \ } @@ -1385,14 +1385,14 @@ SYR2K_LAUNCHER_USM(std::complex, rocblas_zsyr2k) #undef SYR2K_LAUNCHER_USM template -inline sycl::event her2k(Func func, sycl::queue &queue, uplo upper_lower, transpose trans, - int64_t n, int64_t k, const DataType alpha, const DataType *a, int64_t lda, - const DataType *b, int64_t ldb, const ScalarType beta, DataType *c, - int64_t ldc, const std::vector &dependencies) { +inline sycl::event her2k(Func func, sycl::queue& queue, uplo upper_lower, transpose trans, + int64_t n, int64_t k, const DataType alpha, const DataType* a, int64_t lda, + const DataType* b, int64_t ldb, const ScalarType beta, DataType* c, + int64_t ldc, const std::vector& dependencies) { auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; auto new_trans = trans == oneapi::math::transpose::nontrans ? oneapi::math::transpose::conjtrans - : oneapi::math::transpose::nontrans; + : oneapi::math::transpose::nontrans; auto new_alpha = std::conj(alpha); return column_major::her2k(func, queue, new_uplo, new_trans, n, k, new_alpha, a, lda, b, ldb, @@ -1400,10 +1400,10 @@ inline sycl::event her2k(Func func, sycl::queue &queue, uplo upper_lower, transp } #define HER2K_LAUNCHER_USM(DATA_TYPE, SCALAR_TYPE, ROCBLAS_ROUTINE) \ - sycl::event her2k(sycl::queue &queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ - const DATA_TYPE alpha, const DATA_TYPE *a, int64_t lda, const DATA_TYPE *b, \ - int64_t ldb, const SCALAR_TYPE beta, DATA_TYPE *c, int64_t ldc, \ - const std::vector &dependencies) { \ + sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, \ + const DATA_TYPE alpha, const DATA_TYPE* a, int64_t lda, const DATA_TYPE* b, \ + int64_t ldb, const SCALAR_TYPE beta, DATA_TYPE* c, int64_t ldc, \ + const std::vector& dependencies) { \ return her2k(ROCBLAS_ROUTINE, queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, \ beta, c, ldc, dependencies); \ } @@ -1418,23 +1418,23 @@ HER2K_LAUNCHER_USM(std::complex, double, rocblas_zher2k) // separated from the B matrix. It is possible to use B instead of C, but this // will slow-down the code. template -inline sycl::event trmm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, const T *a, - int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; +inline sycl::event trmm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, const T* a, + int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::trmm(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, b, ldb, dependencies); } #define TRMM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trmm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE *a, int64_t lda, \ - TYPE *b, int64_t ldb, const std::vector &dependencies) { \ + sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE* a, int64_t lda, \ + TYPE* b, int64_t ldb, const std::vector& dependencies) { \ return trmm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, \ alpha, a, lda, b, ldb, dependencies); \ } @@ -1447,23 +1447,23 @@ TRMM_LAUNCHER_USM(std::complex, rocblas_ztrmm) #undef TRMM_LAUNCHER_USM template -inline sycl::event trsm(Func func, sycl::queue &queue, side left_right, uplo upper_lower, - transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, const T *a, - int64_t lda, T *b, int64_t ldb, - const std::vector &dependencies) { - auto new_side = - left_right == oneapi::math::side::left ? oneapi::math::side::right : oneapi::math::side::left; +inline sycl::event trsm(Func func, sycl::queue& queue, side left_right, uplo upper_lower, + transpose trans, diag unit_diag, int64_t m, int64_t n, T alpha, const T* a, + int64_t lda, T* b, int64_t ldb, + const std::vector& dependencies) { + auto new_side = left_right == oneapi::math::side::left ? oneapi::math::side::right + : oneapi::math::side::left; auto new_uplo = upper_lower == oneapi::math::uplo::lower ? oneapi::math::uplo::upper - : oneapi::math::uplo::lower; + : oneapi::math::uplo::lower; return column_major::trsm(func, queue, new_side, new_uplo, trans, unit_diag, n, m, alpha, a, lda, b, ldb, dependencies); } #define TRSM_LAUNCHER_USM(TYPE, ROCBLAS_ROUTINE) \ - sycl::event trsm(sycl::queue &queue, side left_right, uplo upper_lower, transpose trans, \ - diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE *a, int64_t lda, \ - TYPE *b, int64_t ldb, const std::vector &dependencies) { \ + sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, \ + diag unit_diag, int64_t m, int64_t n, TYPE alpha, const TYPE* a, int64_t lda, \ + TYPE* b, int64_t ldb, const std::vector& dependencies) { \ return trsm(ROCBLAS_ROUTINE, queue, left_right, upper_lower, trans, unit_diag, m, n, \ alpha, a, lda, b, ldb, dependencies); \ } diff --git a/src/blas/backends/rocblas/rocblas_scope_handle.cpp b/src/blas/backends/rocblas/rocblas_scope_handle.cpp index 65187223e..8b59e3f99 100644 --- a/src/blas/backends/rocblas/rocblas_scope_handle.cpp +++ b/src/blas/backends/rocblas/rocblas_scope_handle.cpp @@ -26,7 +26,7 @@ namespace rocblas { template rocblas_handle_container::~rocblas_handle_container() noexcept(false) { - for (auto &handle_pair : rocblas_handle_container_mapper_) { + for (auto& handle_pair : rocblas_handle_container_mapper_) { rocblas_status err; if (handle_pair.second != nullptr) { auto handle = handle_pair.second->exchange(nullptr); @@ -59,7 +59,7 @@ thread_local rocblas_handle_container RocblasScopedContextHandler::h #endif RocblasScopedContextHandler::RocblasScopedContextHandler(sycl::queue queue, - sycl::interop_handle &ih) + sycl::interop_handle& ih) : interop_h(ih), needToRecover_(false) { placedContext_ = new sycl::context(queue.get_context()); @@ -89,8 +89,8 @@ RocblasScopedContextHandler::~RocblasScopedContextHandler() noexcept(false) { delete placedContext_; } -void ContextCallback(void *userData) { - auto *ptr = static_cast *>(userData); +void ContextCallback(void* userData) { + auto* ptr = static_cast*>(userData); if (!ptr) { return; } @@ -108,7 +108,7 @@ void ContextCallback(void *userData) { } } -rocblas_handle RocblasScopedContextHandler::get_handle(const sycl::queue &queue) { +rocblas_handle RocblasScopedContextHandler::get_handle(const sycl::queue& queue) { auto hipDevice = interop_h.get_native_device(); hipError_t hipErr; hipCtx_t desired; @@ -154,10 +154,10 @@ rocblas_handle RocblasScopedContextHandler::get_handle(const sycl::queue &queue) return handle; } -hipStream_t RocblasScopedContextHandler::get_stream(const sycl::queue &queue) { +hipStream_t RocblasScopedContextHandler::get_stream(const sycl::queue& queue) { return sycl::get_native(queue); } -sycl::context RocblasScopedContextHandler::get_context(const sycl::queue &queue) { +sycl::context RocblasScopedContextHandler::get_context(const sycl::queue& queue) { return queue.get_context(); } diff --git a/src/blas/backends/rocblas/rocblas_scope_handle.hpp b/src/blas/backends/rocblas/rocblas_scope_handle.hpp index c08db4a78..f5dbd6f23 100644 --- a/src/blas/backends/rocblas/rocblas_scope_handle.hpp +++ b/src/blas/backends/rocblas/rocblas_scope_handle.hpp @@ -45,29 +45,29 @@ namespace rocblas { template struct rocblas_handle_container { - using handle_container_t = std::unordered_map *>; + using handle_container_t = std::unordered_map*>; handle_container_t rocblas_handle_container_mapper_{}; ~rocblas_handle_container() noexcept(false); }; class RocblasScopedContextHandler { HIPcontext original_; - sycl::context *placedContext_; + sycl::context* placedContext_; bool needToRecover_; - sycl::interop_handle &interop_h; + sycl::interop_handle& interop_h; #ifdef ONEMATH_PI_INTERFACE_REMOVED static thread_local rocblas_handle_container handle_helper; #else static thread_local rocblas_handle_container handle_helper; #endif - sycl::context get_context(const sycl::queue &queue); - hipStream_t get_stream(const sycl::queue &queue); + sycl::context get_context(const sycl::queue& queue); + hipStream_t get_stream(const sycl::queue& queue); public: - RocblasScopedContextHandler(sycl::queue queue, sycl::interop_handle &ih); + RocblasScopedContextHandler(sycl::queue queue, sycl::interop_handle& ih); ~RocblasScopedContextHandler() noexcept(false); - rocblas_handle get_handle(const sycl::queue &queue); + rocblas_handle get_handle(const sycl::queue& queue); // This is a work-around function for reinterpret_casting the memory. This // will be fixed when SYCL-2020 has been implemented for Pi backend. diff --git a/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.cpp b/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.cpp index bc23cf26a..5ac12ca0e 100644 --- a/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.cpp +++ b/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.cpp @@ -27,7 +27,7 @@ namespace blas { namespace rocblas { rocblas_handle_container::~rocblas_handle_container() noexcept(false) { - for (auto &handle_pair : rocblas_handle_mapper_) { + for (auto& handle_pair : rocblas_handle_mapper_) { rocblas_status err; if (handle_pair.second != nullptr) { auto handle = handle_pair.second->exchange(nullptr); @@ -46,10 +46,10 @@ thread_local rocblas_handle_container RocblasScopedContextHandler::handle_helper rocblas_handle_container{}; RocblasScopedContextHandler::RocblasScopedContextHandler(sycl::queue queue, - sycl::interop_handle &ih) + sycl::interop_handle& ih) : interop_h(ih) {} -rocblas_handle RocblasScopedContextHandler::get_handle(const sycl::queue &queue) { +rocblas_handle RocblasScopedContextHandler::get_handle(const sycl::queue& queue) { sycl::device device = queue.get_device(); int current_device = interop_h.get_native_device(); hipStream_t streamId = get_stream(queue); @@ -84,7 +84,7 @@ rocblas_handle RocblasScopedContextHandler::get_handle(const sycl::queue &queue) return handle; } -hipStream_t RocblasScopedContextHandler::get_stream(const sycl::queue &queue) { +hipStream_t RocblasScopedContextHandler::get_stream(const sycl::queue& queue) { return interop_h.get_native_queue(); } diff --git a/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.hpp b/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.hpp index d6fe2a7e7..48dfd433b 100644 --- a/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.hpp +++ b/src/blas/backends/rocblas/rocblas_scope_handle_hipsycl.hpp @@ -35,7 +35,7 @@ namespace blas { namespace rocblas { struct rocblas_handle_container { - using handle_container_t = std::unordered_map *>; + using handle_container_t = std::unordered_map*>; handle_container_t rocblas_handle_mapper_{}; ~rocblas_handle_container() noexcept(false); }; @@ -43,13 +43,13 @@ struct rocblas_handle_container { class RocblasScopedContextHandler { sycl::interop_handle interop_h; static thread_local rocblas_handle_container handle_helper; - sycl::context get_context(const sycl::queue &queue); - hipStream_t get_stream(const sycl::queue &queue); + sycl::context get_context(const sycl::queue& queue); + hipStream_t get_stream(const sycl::queue& queue); public: - RocblasScopedContextHandler(sycl::queue queue, sycl::interop_handle &ih); + RocblasScopedContextHandler(sycl::queue queue, sycl::interop_handle& ih); - rocblas_handle get_handle(const sycl::queue &queue); + rocblas_handle get_handle(const sycl::queue& queue); // This is a work-around function for reinterpret_casting the memory. This // will be fixed when SYCL-2020 has been implemented for Pi backend. diff --git a/src/blas/backends/rocblas/rocblas_task.hpp b/src/blas/backends/rocblas/rocblas_task.hpp index bf1197897..ae48720c9 100644 --- a/src/blas/backends/rocblas/rocblas_task.hpp +++ b/src/blas/backends/rocblas/rocblas_task.hpp @@ -53,7 +53,7 @@ namespace rocblas { #ifdef __HIPSYCL__ template -static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { +static inline void host_task_internal(H& cgh, sycl::queue queue, F f) { cgh.hipSYCL_enqueue_custom_operation([f, queue](sycl::interop_handle ih) { auto sc = RocblasScopedContextHandler(queue, ih); f(sc); @@ -61,9 +61,9 @@ static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { } #else template -static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { +static inline void host_task_internal(H& cgh, sycl::queue queue, F f) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - cgh.ext_codeplay_enqueue_native_command([f, queue](sycl::interop_handle ih){ + cgh.ext_codeplay_enqueue_native_command([f, queue](sycl::interop_handle ih) { #else cgh.host_task([f, queue](sycl::interop_handle ih) { #endif @@ -73,7 +73,7 @@ static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { } #endif template -static inline void onemath_rocblas_host_task(H &cgh, sycl::queue queue, F f) { +static inline void onemath_rocblas_host_task(H& cgh, sycl::queue queue, F f) { (void)host_task_internal(cgh, queue, f); } diff --git a/src/blas/blas_loader.cpp b/src/blas/blas_loader.cpp index d92e188c6..33c3ca185 100644 --- a/src/blas/blas_loader.cpp +++ b/src/blas/blas_loader.cpp @@ -32,3962 +32,3971 @@ static oneapi::math::detail::table_initializer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_scasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_dzasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_sasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_dasum_sycl(queue, n, x, incx, result); } -void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, +void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_saxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, +void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_daxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_caxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zaxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_saxpy_batch_strided_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_daxpy_batch_strided_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, +void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_caxpy_batch_strided_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, +void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_zaxpy_batch_strided_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, +void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_saxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, +void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_daxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_caxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zaxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy) { function_tables[{ libkey, queue }].column_major_scopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, +void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_dcopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_ccopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zcopy_sycl(queue, n, x, incx, y, incy); } -void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_scopy_batch_strided_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_dcopy_batch_strided_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, +void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_ccopy_batch_strided_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, +void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_zcopy_batch_strided_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_sdot_sycl(queue, n, x, incx, y, incy, result); } -void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_ddot_sycl(queue, n, x, incx, y, incy, result); } -void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_dsdot_sycl(queue, n, x, incx, y, incy, result); } -void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { function_tables[{ libkey, queue }].column_major_cdotc_sycl(queue, n, x, incx, y, incy, result); } -void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { function_tables[{ libkey, queue }].column_major_zdotc_sycl(queue, n, x, incx, y, incy, result); } -void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { function_tables[{ libkey, queue }].column_major_cdotu_sycl(queue, n, x, incx, y, incy, result); } -void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { function_tables[{ libkey, queue }].column_major_zdotu_sycl(queue, n, x, incx, y, incy, result); } -void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_isamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_idamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_icamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_izamin_sycl(queue, n, x, incx, result); } -void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_isamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_idamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_icamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_izamax_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_scnrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_dznrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_snrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_dnrm2_sycl(queue, n, x, incx, result); } -void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { +void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, float c, float s) { function_tables[{ libkey, queue }].column_major_srot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { +void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, double s) { function_tables[{ libkey, queue }].column_major_drot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { +void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, float c, float s) { function_tables[{ libkey, queue }].column_major_csrot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { +void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, double c, double s) { function_tables[{ libkey, queue }].column_major_zdrot_sycl(queue, n, x, incx, y, incy, c, s); } -void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { +void rotg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { function_tables[{ libkey, queue }].column_major_srotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { +void rotg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { function_tables[{ libkey, queue }].column_major_drotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { function_tables[{ libkey, queue }].column_major_crotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { function_tables[{ libkey, queue }].column_major_zrotg_sycl(queue, a, b, c, s); } -void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m) { +void rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& param) { function_tables[{ libkey, queue }].column_major_srotm_sycl(queue, n, x, incx, y, incy, param); } -void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer ¶m) { +void rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& param) { function_tables[{ libkey, queue }].column_major_drotm_sycl(queue, n, x, incx, y, incy, param); } -void rotmg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, float y1, - sycl::buffer ¶m) { +void rotmg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, float y1, + sycl::buffer& param) { function_tables[{ libkey, queue }].column_major_srotmg_sycl(queue, d1, d2, x1, y1, param); } -void rotmg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, double y1, - sycl::buffer ¶m) { +void rotmg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, double y1, + sycl::buffer& param) { function_tables[{ libkey, queue }].column_major_drotmg_sycl(queue, d1, d2, x1, y1, param); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_sscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_cscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_csscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_zscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_zdscal_sycl(queue, n, alpha, x, incx); } -void sdsdot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &result) { +void sdsdot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float sb, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result) { function_tables[{ libkey, queue }].column_major_sdsdot_sycl(queue, n, sb, x, incx, y, incy, result); } -void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy) { function_tables[{ libkey, queue }].column_major_sswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, +void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_dswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_cswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zswap_sycl(queue, n, x, incx, y, incy); } -void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { +void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_sgbmv_sycl(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_dgbmv_sycl(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_cgbmv_sycl(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zgbmv_sycl(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, +void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_sgemv_sycl(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, +void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_dgemv_sycl(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_cgemv_sycl(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zgemv_sycl(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, +void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, float beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_sgemv_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, +void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, double beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_dgemv_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, +void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_cgemv_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_zgemv_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_sdgmm_batch_strided_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_ddgmm_batch_strided_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, +void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_cdgmm_batch_strided_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, +void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_zdgmm_batch_strided_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { +void ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_sger_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { +void ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_dger_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_cgerc_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_zgerc_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_cgeru_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_zgeru_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_chbmv_sycl(queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zhbmv_sycl(queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_chemv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zhemv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { +void her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_cher_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { +void her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_zher_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_cher2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_zher2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_chpmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_zhpmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { +void hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { function_tables[{ libkey, queue }].column_major_chpr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { +void hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { function_tables[{ libkey, queue }].column_major_zhpr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { +void hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { function_tables[{ libkey, queue }].column_major_chpr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { +void hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { function_tables[{ libkey, queue }].column_major_zhpr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, +void sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_ssbmv_sycl(queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, +void sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_dsbmv_sycl(queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy) { +void spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_sspmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy) { +void spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_dspmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { +void spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { function_tables[{ libkey, queue }].column_major_sspr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { +void spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { function_tables[{ libkey, queue }].column_major_dspr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a) { +void spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { function_tables[{ libkey, queue }].column_major_sspr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a) { +void spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { function_tables[{ libkey, queue }].column_major_dspr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { +void symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_ssymv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { +void symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].column_major_dsymv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, +void syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_ssyr_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, +void syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_dsyr_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { +void syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_ssyr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { +void syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].column_major_dsyr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_stbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_stbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, +void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_stpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, +void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, +void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_stpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, +void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_strmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { +void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_strsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_dtrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { +void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ctrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].column_major_ztrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_sgemm_sycl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_dgemm_sycl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_cgemm_sycl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_zgemm_sycl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::half beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_hgemm_sycl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_gemm_f16f16f32_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_gemm_bf16bf16f32_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_chemm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_zhemm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, - std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { +void herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1>& a, + std::int64_t lda, float beta, sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_cherk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, - std::int64_t lda, double beta, sycl::buffer, 1> &c, +void herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1>& a, + std::int64_t lda, double beta, sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_zherk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_cher2k_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, double beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_zher2k_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, +void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_ssymm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { +void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_dsymm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_csymm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_zsymm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - float beta, sycl::buffer &c, std::int64_t ldc) { +void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_ssyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { +void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_dsyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_csyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_zsyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, +void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_ssyrk_batch_strided_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, +void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_dsyrk_batch_strided_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_csyrk_batch_strided_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_zsyrk_batch_strided_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, +void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_ssyr2k_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { +void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_dsyr2k_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_csyr2k_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_zsyr2k_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_strmm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_dtrmm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_ctrmm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_ztrmm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_strsm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_dtrsm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_ctrsm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_ztrsm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_sgemm_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, double beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_dgemm_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_cgemm_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_zgemm_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::half beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::half beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_hgemm_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_gemm_f16f16f32_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_gemm_s8s8f32_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_gemm_s8s8s32_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_strsm_batch_strided_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_dtrsm_batch_strided_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_ctrsm_batch_strided_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_ztrsm_batch_strided_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { +void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_sgemmt_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_dgemmt_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_cgemmt_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_zgemmt_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { function_tables[{ libkey, queue }].column_major_gemm_s8u8s32_bias_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { function_tables[{ libkey, queue }].column_major_gemm_s8s8s32_bias_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { function_tables[{ libkey, queue }].column_major_gemm_u8s8s32_bias_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { function_tables[{ libkey, queue }].column_major_gemm_u8u8s32_bias_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { +void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_somatcopy_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { +void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_domatcopy_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_comatcopy_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_zomatcopy_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { +void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& ab, + std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_simatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { +void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& ab, + std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_dimatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_cimatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_zimatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, float beta, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_somatadd_batch_strided_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, double beta, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_domatadd_batch_strided_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_comatadd_batch_strided_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].column_major_zomatadd_batch_strided_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { +void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_somatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { +void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_domatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { +void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_comatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { +void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_zomatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, +void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { function_tables[{ libkey, queue }].column_major_somatcopy2_sycl(queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, +void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { function_tables[{ libkey, queue }].column_major_domatcopy2_sycl(queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, +void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { function_tables[{ libkey, queue }].column_major_comatcopy2_sycl(queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, +void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { function_tables[{ libkey, queue }].column_major_zomatcopy2_sycl(queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, +void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_simatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, +void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_dimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_cimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].column_major_zimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { +void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, float beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_somatadd_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { +void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_domatadd_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_comatadd_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].column_major_zomatadd_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } // USM APIs -sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_scasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dzasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, const std::vector &dependencies) { +sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, const std::vector &dependencies) { +sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_saxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + const double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_daxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_caxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zaxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - float *alpha, const float **x, std::int64_t *incx, float **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + float* alpha, const float** x, std::int64_t* incx, float** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_saxpy_batch_group_usm_sycl( queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - double *alpha, const double **x, std::int64_t *incx, double **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + double* alpha, const double** x, std::int64_t* incx, double** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_daxpy_batch_group_usm_sycl( queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_caxpy_batch_group_usm_sycl( queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zaxpy_batch_group_usm_sycl( queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_saxpy_batch_strided_usm_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double alpha, const double* x, std::int64_t incx, std::int64_t stridex, + double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_daxpy_batch_strided_usm_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, std::int64_t incy, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_caxpy_batch_strided_usm_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, std::int64_t incy, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zaxpy_batch_strided_usm_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, const float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_saxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double beta, double *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + const double* x, std::int64_t incx, const double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_daxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_caxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zaxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_scopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dcopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ccopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zcopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - const float **x, std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + const float** x, std::int64_t* incx, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_scopy_batch_group_usm_sycl( queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - const double **x, std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dcopy_batch_group_usm_sycl( queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ccopy_batch_group_usm_sycl( queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zcopy_batch_group_usm_sycl( queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const float* x, std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_scopy_batch_strided_usm_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const double* x, std::int64_t incx, std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dcopy_batch_strided_usm_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ccopy_batch_strided_usm_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zcopy_batch_strided_usm_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sdot_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ddot_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, double* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dsdot_usm_sycl( queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { +sycl::event dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cdotc_usm_sycl( queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { +sycl::event dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zdotc_usm_sycl( queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { +sycl::event dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cdotu_usm_sycl( queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { +sycl::event dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zdotu_usm_sycl( queue, n, x, incx, y, incy, result, dependencies); } -sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_isamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_idamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_icamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_izamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_isamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_idamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_icamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_izamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_scnrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dznrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, const std::vector &dependencies) { +sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_snrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, const std::vector &dependencies) { +sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dnrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, float c, float s, const std::vector &dependencies) { +sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, float c, float s, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_srot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, +sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_drot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_csrot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zdrot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, float *a, float *b, float *c, - float *s, const std::vector &dependencies) { +sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, float* a, float* b, float* c, + float* s, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_srotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, double *a, double *b, double *c, - double *s, const std::vector &dependencies) { +sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, double* a, double* b, double* c, + double* s, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_drotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, std::complex *a, - std::complex *b, float *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, std::complex* a, + std::complex* b, float* c, std::complex* s, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_crotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, std::complex *a, - std::complex *b, double *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, std::complex* a, + std::complex* b, double* c, std::complex* s, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zrotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float *param, - const std::vector &dependencies) { +sycl::event rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float* param, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_srotm_usm_sycl( queue, n, x, incx, y, incy, param, dependencies); } -sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double *param, - const std::vector &dependencies) { +sycl::event rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double* param, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_drotm_usm_sycl( queue, n, x, incx, y, incy, param, dependencies); } -sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, float *d1, float *d2, float *x1, - float y1, float *param, const std::vector &dependencies) { +sycl::event rotmg(oneapi::math::device libkey, sycl::queue& queue, float* d1, float* d2, float* x1, + float y1, float* param, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_srotmg_usm_sycl(queue, d1, d2, x1, y1, param, dependencies); } -sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, double *d1, double *d2, - double *x1, double y1, double *param, - const std::vector &dependencies) { +sycl::event rotmg(oneapi::math::device libkey, sycl::queue& queue, double* d1, double* d2, + double* x1, double y1, double* param, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_drotmg_usm_sycl(queue, d1, d2, x1, y1, param, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - float *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + float* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - double *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + double* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_csscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zdscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event sdsdot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float sb, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *result, const std::vector &dependencies) { +sycl::event sdsdot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float sb, + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* result, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sdsdot_usm_sycl( queue, n, sb, x, incx, y, incy, result, dependencies); } -sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sgbmv_usm_sycl( queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dgbmv_usm_sycl( queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cgbmv_usm_sycl( queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zgbmv_usm_sycl( queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sgemv_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dgemv_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cgemv_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zgemv_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float beta, float *y, std::int64_t incy, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float beta, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sgemv_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double beta, double *y, std::int64_t incy, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double beta, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dgemv_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cgemv_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zgemv_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, - float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, const float** x, std::int64_t* incx, float* beta, + float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sgemv_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, - double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, const double** x, std::int64_t* incx, double* beta, + double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dgemv_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cgemv_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zgemv_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, const float* a, std::int64_t lda, + std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sdgmm_batch_strided_usm_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, const double* a, std::int64_t lda, + std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ddgmm_batch_strided_usm_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cdgmm_batch_strided_usm_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zdgmm_batch_strided_usm_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + std::int64_t* m, std::int64_t* n, const float** a, std::int64_t* lda, + const float** x, std::int64_t* incx, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sdgmm_batch_group_usm_sycl( queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + std::int64_t* m, std::int64_t* n, const double** a, std::int64_t* lda, + const double** x, std::int64_t* incx, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ddgmm_batch_group_usm_sycl( queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cdgmm_batch_group_usm_sycl( queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zdgmm_batch_group_usm_sycl( queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *a, std::int64_t lda, const std::vector &dependencies) { +sycl::event ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sger_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dger_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cgerc_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zgerc_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cgeru_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zgeru_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_chbmv_usm_sycl( queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zhbmv_usm_sycl( queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_chemv_usm_sycl( queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zhemv_usm_sycl( queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cher_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zher_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cher2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zher2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_chpmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zhpmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { +sycl::event hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_chpr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { +sycl::event hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zhpr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { +sycl::event hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_chpr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { +sycl::event hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zhpr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ssbmv_usm_sycl( queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dsbmv_usm_sycl( queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *a, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, const std::vector &dependencies) { +sycl::event spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* a, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sspmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *a, const double *x, std::int64_t incx, double beta, - double *y, std::int64_t incy, const std::vector &dependencies) { +sycl::event spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* a, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dspmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, - const std::vector &dependencies) { +sycl::event spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, float* a, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sspr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, - const std::vector &dependencies) { +sycl::event spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, double* a, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dspr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *a, const std::vector &dependencies) { +sycl::event spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sspr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, const std::vector &dependencies) { +sycl::event spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dspr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, - float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* a, std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ssymv_usm_sycl( queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dsymv_usm_sycl( queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ssyr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dsyr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *a, std::int64_t lda, const std::vector &dependencies) { +sycl::event syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ssyr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dsyr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_stbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, - double *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dtbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ctbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ztbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_stbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, - double *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dtbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ctbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ztbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_stpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dtpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ctpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ztpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_stpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dtpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ctpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ztpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_strmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dtrmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ctrmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ztrmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_strsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dtrsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ctrsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ztrsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, const float* b, std::int64_t ldb, float beta, + float* c, std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sgemm_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, const double* b, std::int64_t ldb, double beta, + double* c, std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dgemm_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cgemm_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zgemm_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - sycl::half beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_hgemm_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_f16f16f32_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, - std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const bfloat16* a, std::int64_t lda, const bfloat16* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_bf16bf16f32_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_chemm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zhemm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const std::complex *a, - std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, const std::complex* a, + std::int64_t lda, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cherk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const std::complex *a, - std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, const std::complex* a, + std::int64_t lda, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zherk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cher2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zher2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ssymm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dsymm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_csymm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zsymm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ssyrk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dsyrk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_csyrk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zsyrk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, - const float **a, std::int64_t *lda, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, float* alpha, + const float** a, std::int64_t* lda, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ssyrk_batch_group_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, - const double **a, std::int64_t *lda, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, double* alpha, + const double** a, std::int64_t* lda, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dsyrk_batch_group_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_csyrk_batch_group_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zsyrk_batch_group_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float beta, float *c, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ssyrk_batch_strided_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dsyrk_batch_strided_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, std::complex *c, std::int64_t ldc, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_csyrk_batch_strided_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, std::complex *c, std::int64_t ldc, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zsyrk_batch_strided_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ssyr2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dsyr2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_csyr2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zsyr2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_strmm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dtrmm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ctrmm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ztrmm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_strsm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dtrsm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ctrsm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ztrsm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_strsm_batch_strided_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dtrsm_batch_strided_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ctrsm_batch_strided_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ztrsm_batch_strided_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_strsm_batch_group_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, - double **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dtrsm_batch_group_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ctrsm_batch_group_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_ztrsm_batch_group_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const float **a, std::int64_t *lda, const float **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const float** a, std::int64_t* lda, const float** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sgemm_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - double *alpha, const double **a, std::int64_t *lda, const double **b, - std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + double* alpha, const double** a, std::int64_t* lda, const double** b, + std::int64_t* ldb, double* beta, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dgemm_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **b, std::int64_t *ldb, std::complex *beta, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, std::complex* beta, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cgemm_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zgemm_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - sycl::half *alpha, const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, sycl::half *beta, sycl::half **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + sycl::half* alpha, const sycl::half** a, std::int64_t* lda, + const sycl::half** b, std::int64_t* ldb, sycl::half* beta, sycl::half** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_hgemm_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const sycl::half** a, std::int64_t* lda, const sycl::half** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_f16f16f32_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_s8s8f32_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, std::int32_t** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_s8s8s32_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, - const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, + const float* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sgemm_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, - const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, + const double* b, std::int64_t ldb, std::int64_t stride_b, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dgemm_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, const std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::complex beta, std::complex *c, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cgemm_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, const std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::complex beta, std::complex *c, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zgemm_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, const sycl::half *a, std::int64_t lda, - std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, + sycl::half alpha, const sycl::half* a, std::int64_t lda, + std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, sycl::half* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_hgemm_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_f16f16f32_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_s8s8f32_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_s8s8s32_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, - float *c, std::int64_t ldc, const std::vector &dependencies) { + const float* a, std::int64_t lda, const float* b, std::int64_t ldb, float beta, + float* c, std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_sgemmt_usm_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, - double *c, std::int64_t ldc, const std::vector &dependencies) { + const double* a, std::int64_t lda, const double* b, std::int64_t ldb, double beta, + double* c, std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dgemmt_usm_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cgemmt_usm_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zgemmt_usm_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_s8u8s32_bias_usm_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } -sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_s8s8s32_bias_usm_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } -sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_u8s8s32_bias_usm_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } -sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_gemm_u8u8s32_bias_usm_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_somatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_domatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_comatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zomatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_simatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dimatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cimatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zimatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, + const float* a, std::int64_t lda, std::int64_t stride_a, float beta, + const float* b, std::int64_t ldb, std::int64_t stride_b, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_somatadd_batch_strided_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + const double* b, std::int64_t ldb, std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_domatadd_batch_strided_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_comatadd_batch_strided_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zomatadd_batch_strided_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } -sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - float *b, std::int64_t ldb, const std::vector &dependencies) { +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + float* b, std::int64_t ldb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_somatcopy_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_domatcopy_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_comatcopy_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zomatcopy_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_somatcopy2_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, - std::int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stridea, double* b, std::int64_t ldb, + std::int64_t strideb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_domatcopy2_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - std::complex *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stridea, + std::complex* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_comatcopy2_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - std::complex *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stridea, + std::complex* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zomatcopy2_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, - std::int64_t ldb, const std::vector &dependencies) { +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, float* ab, std::int64_t lda, + std::int64_t ldb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_simatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, - std::int64_t ldb, const std::vector &dependencies) { +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, double* ab, std::int64_t lda, + std::int64_t ldb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dimatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + std::complex* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cimatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + std::complex* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zimatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, float beta, const float* b, std::int64_t ldb, float* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_somatadd_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double beta, const double* b, std::int64_t ldb, double* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_domatadd_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_comatadd_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zomatadd_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_somatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_domatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_comatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zomatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, float **ab, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, float** ab, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_simatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, double **ab, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, double** ab, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_dimatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - std::complex **ab, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + std::complex** ab, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_cimatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - std::complex **ab, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + std::complex** ab, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].column_major_zimatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } @@ -4001,3962 +4010,3971 @@ static oneapi::math::detail::table_initializer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_scasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_dzasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_sasum_sycl(queue, n, x, incx, result); } -void asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_dasum_sycl(queue, n, x, incx, result); } -void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, +void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_saxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, +void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_daxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_caxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zaxpy_sycl(queue, n, alpha, x, incx, y, incy); } -void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_saxpy_batch_strided_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_daxpy_batch_strided_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, +void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_caxpy_batch_strided_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1> &y, +void axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_zaxpy_batch_strided_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size); } -void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, +void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_saxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, +void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_daxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_caxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zaxpby_sycl(queue, n, alpha, x, incx, beta, y, incy); } -void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy) { function_tables[{ libkey, queue }].row_major_scopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, +void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_dcopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_ccopy_sycl(queue, n, x, incx, y, incy); } -void copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zcopy_sycl(queue, n, x, incx, y, incy); } -void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_scopy_batch_strided_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &y, std::int64_t incy, std::int64_t stridey, +void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_dcopy_batch_strided_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, +void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_ccopy_batch_strided_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, +void copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_zcopy_batch_strided_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size); } -void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_sdot_sycl(queue, n, x, incx, y, incy, result); } -void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_ddot_sycl(queue, n, x, incx, y, incy, result); } -void dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result) { +void dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_dsdot_sycl(queue, n, x, incx, y, incy, result); } -void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { function_tables[{ libkey, queue }].row_major_cdotc_sycl(queue, n, x, incx, y, incy, result); } -void dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { function_tables[{ libkey, queue }].row_major_zdotc_sycl(queue, n, x, incx, y, incy, result); } -void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { function_tables[{ libkey, queue }].row_major_cdotu_sycl(queue, n, x, incx, y, incy, result); } -void dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result) { +void dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result) { function_tables[{ libkey, queue }].row_major_zdotu_sycl(queue, n, x, incx, y, incy, result); } -void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_isamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_idamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_icamin_sycl(queue, n, x, incx, result); } -void iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_izamin_sycl(queue, n, x, incx, result); } -void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_isamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_idamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_icamax_sycl(queue, n, x, incx, result); } -void iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_izamax_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_scnrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result) { +void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_dznrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result) { +void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_snrm2_sycl(queue, n, x, incx, result); } -void nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &result) { +void nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_dnrm2_sycl(queue, n, x, incx, result); } -void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, float s) { +void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, float c, float s) { function_tables[{ libkey, queue }].row_major_srot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, double c, double s) { +void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, double s) { function_tables[{ libkey, queue }].row_major_drot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, float c, float s) { +void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, float c, float s) { function_tables[{ libkey, queue }].row_major_csrot_sycl(queue, n, x, incx, y, incy, c, s); } -void rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, double c, double s) { +void rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, double c, double s) { function_tables[{ libkey, queue }].row_major_zdrot_sycl(queue, n, x, incx, y, incy, c, s); } -void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { +void rotg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { function_tables[{ libkey, queue }].row_major_srotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, sycl::buffer &s) { +void rotg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, sycl::buffer& s) { function_tables[{ libkey, queue }].row_major_drotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { function_tables[{ libkey, queue }].row_major_crotg_sycl(queue, a, b, c, s); } -void rotg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s) { +void rotg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s) { function_tables[{ libkey, queue }].row_major_zrotg_sycl(queue, a, b, c, s); } -void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m) { +void rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& param) { function_tables[{ libkey, queue }].row_major_srotm_sycl(queue, n, x, incx, y, incy, param); } -void rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer ¶m) { +void rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& param) { function_tables[{ libkey, queue }].row_major_drotm_sycl(queue, n, x, incx, y, incy, param); } -void rotmg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, float y1, - sycl::buffer ¶m) { +void rotmg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, float y1, + sycl::buffer& param) { function_tables[{ libkey, queue }].row_major_srotmg_sycl(queue, d1, d2, x1, y1, param); } -void rotmg(oneapi::math::device libkey, sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, double y1, - sycl::buffer ¶m) { +void rotmg(oneapi::math::device libkey, sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, double y1, + sycl::buffer& param) { function_tables[{ libkey, queue }].row_major_drotmg_sycl(queue, d1, d2, x1, y1, param); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_sscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_cscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_csscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_zscal_sycl(queue, n, alpha, x, incx); } -void scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx) { +void scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_zdscal_sycl(queue, n, alpha, x, incx); } -void sdsdot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &result) { +void sdsdot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float sb, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& result) { function_tables[{ libkey, queue }].row_major_sdsdot_sycl(queue, n, sb, x, incx, y, incy, result); } -void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy) { +void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy) { function_tables[{ libkey, queue }].row_major_sswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, sycl::buffer &y, +void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_dswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_cswap_sycl(queue, n, x, incx, y, incy); } -void swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy) { +void swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zswap_sycl(queue, n, x, incx, y, incy); } -void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy) { +void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_sgbmv_sycl(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_dgbmv_sycl(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_cgbmv_sycl(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -void gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zgbmv_sycl(queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); } -void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, +void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_sgemv_sycl(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, +void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_dgemv_sycl(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_cgemv_sycl(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -void gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zgemv_sycl(queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } -void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, float beta, sycl::buffer &y, std::int64_t incy, +void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, float beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_sgemv_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, double beta, sycl::buffer &y, std::int64_t incy, +void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, double beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_dgemv_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &x, +void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy, std::int64_t stridey, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_cgemv_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -void gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +void gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_zgemv_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size); } -void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_sdgmm_batch_strided_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, std::int64_t stridec, +void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_ddgmm_batch_strided_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, +void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_cdgmm_batch_strided_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, std::int64_t m, - std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer, 1> &c, std::int64_t ldc, +void dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, std::int64_t m, + std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_zdgmm_batch_strided_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size); } -void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { +void ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_sger_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { +void ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_dger_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_cgerc_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_zgerc_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_cgeru_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_zgeru_sycl(queue, m, n, alpha, x, incx, y, incy, a, lda); } -void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_chbmv_sycl(queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -void hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, std::int64_t incy) { +void hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zhbmv_sycl(queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_chemv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zhemv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { +void her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_cher_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda) { +void her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_zher_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_cher2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda) { +void her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_zher2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_chpmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy) { +void hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, + sycl::buffer, 1>& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_zhpmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { +void hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { function_tables[{ libkey, queue }].row_major_chpr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a) { +void hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a) { function_tables[{ libkey, queue }].row_major_zhpr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { +void hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { function_tables[{ libkey, queue }].row_major_chpr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a) { +void hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a) { function_tables[{ libkey, queue }].row_major_zhpr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, sycl::buffer &y, +void sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_ssbmv_sycl(queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -void sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, sycl::buffer &y, +void sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_dsbmv_sycl(queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } -void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy) { +void spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_sspmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy) { +void spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& a, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_dspmv_sycl(queue, upper_lower, n, alpha, a, x, incx, beta, y, incy); } -void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { +void spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { function_tables[{ libkey, queue }].row_major_sspr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a) { +void spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a) { function_tables[{ libkey, queue }].row_major_dspr_sycl(queue, upper_lower, n, alpha, x, incx, a); } -void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a) { +void spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { function_tables[{ libkey, queue }].row_major_sspr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a) { +void spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a) { function_tables[{ libkey, queue }].row_major_dspr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a); } -void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, std::int64_t incy) { +void symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_ssymv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, sycl::buffer &x, - std::int64_t incx, double beta, sycl::buffer &y, std::int64_t incy) { +void symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, + std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy) { function_tables[{ libkey, queue }].row_major_dsymv_sycl(queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); } -void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, +void syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_ssyr_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &a, +void syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_dsyr_sycl(queue, upper_lower, n, alpha, x, incx, a, lda); } -void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { +void syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_ssyr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, std::int64_t lda) { +void syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda) { function_tables[{ libkey, queue }].row_major_dsyr2_sycl(queue, upper_lower, n, alpha, x, incx, y, incy, a, lda); } -void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_stbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztbmv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_stbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx) { +void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztbsv_sycl(queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); } -void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, +void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_stpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, +void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +void tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztpmv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, +void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_stpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, sycl::buffer &x, +void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx) { +void tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztpsv_sycl(queue, upper_lower, trans, unit_diag, n, a, x, incx); } -void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_strmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { +void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztrmv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_strsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx) { +void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_dtrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx) { +void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ctrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, std::int64_t incx) { +void trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx) { function_tables[{ libkey, queue }].row_major_ztrsv_sycl(queue, upper_lower, trans, unit_diag, n, a, lda, x, incx); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_sgemm_sycl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_dgemm_sycl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_cgemm_sycl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_zgemm_sycl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::half beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::half beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_hgemm_sycl(queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_gemm_f16f16f32_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { +void gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_gemm_bf16bf16f32_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_chemm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_zhemm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1> &a, - std::int64_t lda, float beta, sycl::buffer, 1> &c, std::int64_t ldc) { +void herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer, 1>& a, + std::int64_t lda, float beta, sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_cherk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1> &a, - std::int64_t lda, double beta, sycl::buffer, 1> &c, +void herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer, 1>& a, + std::int64_t lda, double beta, sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_zherk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, float beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, float beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_cher2k_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, double beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, double beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_zher2k_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, +void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_ssymm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { +void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_dsymm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_csymm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_zsymm_sycl(queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } -void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - float beta, sycl::buffer &c, std::int64_t ldc) { +void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_ssyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &c, std::int64_t ldc) { +void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_dsyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_csyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_zsyrk_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc); } -void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer &c, +void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_ssyrk_batch_strided_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer &c, +void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, double beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_dsyrk_batch_strided_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_csyrk_batch_strided_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -void syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_zsyrk_batch_strided_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size); } -void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, sycl::buffer &c, +void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_ssyr2k_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc) { +void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_dsyr2k_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_csyr2k_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +void syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_zsyr2k_sycl(queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_strmm_sycl(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_dtrmm_sycl(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_ctrmm_sycl(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_ztrmm_sycl(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_strsm_sycl(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_dtrsm_sycl(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_ctrsm_sycl(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb) { + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_ztrsm_sycl(queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_sgemm_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, double beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, double beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_dgemm_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_cgemm_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_zgemm_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::half beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::half beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_hgemm_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_gemm_f16f16f32_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_gemm_s8s8f32_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - float beta, sycl::buffer &c, std::int64_t ldc, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_gemm_s8s8s32_batch_strided_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size); } -void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_strsm_batch_strided_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_dtrsm_batch_strided_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_ctrsm_batch_strided_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +void trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_ztrsm_batch_strided_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, - transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc) { +void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, + transpose transb, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_sgemmt_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, double beta, sycl::buffer &c, std::int64_t ldc) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_dgemmt_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_cgemmt_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose transa, +void gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_zgemmt_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { function_tables[{ libkey, queue }].row_major_gemm_s8u8s32_bias_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& co) { function_tables[{ libkey, queue }].row_major_gemm_s8s8s32_bias_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, int8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, int8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { function_tables[{ libkey, queue }].row_major_gemm_u8s8s32_bias_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, - sycl::buffer &b, std::int64_t ldb, uint8_t bo, float beta, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &co) { + sycl::buffer& a, std::int64_t lda, uint8_t ao, + sycl::buffer& b, std::int64_t ldb, uint8_t bo, float beta, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& co) { function_tables[{ libkey, queue }].row_major_gemm_u8u8s32_bias_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co); } -void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { +void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_somatcopy_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size) { +void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, + std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_domatcopy_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_comatcopy_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, +void omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_zomatcopy_batch_strided_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size); } -void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { +void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& ab, + std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_simatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { +void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& ab, + std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_dimatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_cimatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, std::int64_t ldb, +void imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_zimatcopy_batch_strided_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size); } -void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, float beta, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, float beta, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_somatadd_batch_strided_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, double beta, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, double beta, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_domatadd_batch_strided_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, + std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_comatadd_batch_strided_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +void omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, + sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size) { function_tables[{ libkey, queue }].row_major_zomatadd_batch_strided_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size); } -void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { +void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_somatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb) { +void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_domatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { +void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_comatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, std::int64_t ldb) { +void omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_zomatcopy_sycl(queue, trans, m, n, alpha, a, lda, b, ldb); } -void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, +void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { function_tables[{ libkey, queue }].row_major_somatcopy2_sycl(queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, std::int64_t ldb, +void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb) { function_tables[{ libkey, queue }].row_major_domatcopy2_sycl(queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, +void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { function_tables[{ libkey, queue }].row_major_comatcopy2_sycl(queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -void omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1> &b, +void omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb) { function_tables[{ libkey, queue }].row_major_zomatcopy2_sycl(queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb); } -void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &ab, std::int64_t lda, +void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_simatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &ab, std::int64_t lda, +void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_dimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_cimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, sycl::buffer, 1> &ab, +void imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb) { function_tables[{ libkey, queue }].row_major_zimatcopy_sycl(queue, trans, m, n, alpha, ab, lda, ldb); } -void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { +void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, float beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_somatadd_sycl(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc) { +void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_domatadd_sycl(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_comatadd_sycl(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } -void omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, +void omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc) { + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc) { function_tables[{ libkey, queue }].row_major_zomatadd_sycl(queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc); } // USM APIs -sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_scasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dzasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, const std::vector &dependencies) { +sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event asum(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, const std::vector &dependencies) { +sycl::event asum(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dasum_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_saxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + const double* x, std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_daxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_caxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zaxpy_usm_sycl(queue, n, alpha, x, incx, y, incy, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - float *alpha, const float **x, std::int64_t *incx, float **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + float* alpha, const float** x, std::int64_t* incx, float** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_saxpy_batch_group_usm_sycl( queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - double *alpha, const double **x, std::int64_t *incx, double **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + double* alpha, const double** x, std::int64_t* incx, double** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_daxpy_batch_group_usm_sycl( queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_caxpy_batch_group_usm_sycl( queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - std::complex *alpha, const std::complex **x, - std::int64_t *incx, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + std::complex* alpha, const std::complex** x, + std::int64_t* incx, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zaxpy_batch_group_usm_sycl( queue, n, alpha, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_saxpy_batch_strided_usm_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + double alpha, const double* x, std::int64_t incx, std::int64_t stridex, + double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_daxpy_batch_strided_usm_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, std::int64_t incy, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_caxpy_batch_strided_usm_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, std::int64_t incy, +sycl::event axpy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zaxpy_batch_strided_usm_sycl( queue, n, alpha, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, const float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_saxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double beta, double *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + const double* x, std::int64_t incx, const double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_daxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_caxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event axpby(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event axpby(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zaxpby_usm_sycl( queue, n, alpha, x, incx, beta, y, incy, dependencies); } -sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_scopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dcopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ccopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event copy(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zcopy_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - const float **x, std::int64_t *incx, float **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + const float** x, std::int64_t* incx, float** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_scopy_batch_group_usm_sycl( queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - const double **x, std::int64_t *incx, double **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + const double** x, std::int64_t* incx, double** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dcopy_batch_group_usm_sycl( queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ccopy_batch_group_usm_sycl( queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zcopy_batch_group_usm_sycl( queue, n, x, incx, y, incy, group_count, group_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const float *x, std::int64_t incx, std::int64_t stridex, float *y, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const float* x, std::int64_t incx, std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_scopy_batch_strided_usm_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const double *x, std::int64_t incx, std::int64_t stridex, double *y, +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const double* x, std::int64_t incx, std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dcopy_batch_strided_usm_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ccopy_batch_strided_usm_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event copy_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event copy_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zcopy_batch_strided_usm_sycl( queue, n, x, incx, stridex, y, incy, stridey, batch_size, dependencies); } -sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, float *result, - const std::vector &dependencies) { +sycl::event dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, float* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sdot_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, double* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ddot_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, double *result, - const std::vector &dependencies) { +sycl::event dot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, double* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dsdot_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { +sycl::event dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cdotc_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { +sycl::event dotc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zdotc_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { +sycl::event dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cdotu_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event dotu(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *result, - const std::vector &dependencies) { +sycl::event dotu(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zdotu_usm_sycl(queue, n, x, incx, y, incy, result, dependencies); } -sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_isamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_idamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_icamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamin(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamin(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_izamin_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_isamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_idamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_icamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event iamax(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, std::int64_t *result, - const std::vector &dependencies) { +sycl::event iamax(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, std::int64_t* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_izamax_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, float *result, - const std::vector &dependencies) { +sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, float* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_scnrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, double *result, - const std::vector &dependencies) { +sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, double* result, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dznrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, const std::vector &dependencies) { +sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_snrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event nrm2(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, const std::vector &dependencies) { +sycl::event nrm2(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dnrm2_usm_sycl(queue, n, x, incx, result, dependencies); } -sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, float c, float s, const std::vector &dependencies) { +sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, float c, float s, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_srot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, +sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_drot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float c, float s, - const std::vector &dependencies) { +sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float c, float s, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_csrot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double c, double s, - const std::vector &dependencies) { +sycl::event rot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double c, double s, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zdrot_usm_sycl(queue, n, x, incx, y, incy, c, s, dependencies); } -sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, float *a, float *b, float *c, - float *s, const std::vector &dependencies) { +sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, float* a, float* b, float* c, + float* s, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_srotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, double *a, double *b, double *c, - double *s, const std::vector &dependencies) { +sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, double* a, double* b, double* c, + double* s, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_drotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, std::complex *a, - std::complex *b, float *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, std::complex* a, + std::complex* b, float* c, std::complex* s, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_crotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotg(oneapi::math::device libkey, sycl::queue &queue, std::complex *a, - std::complex *b, double *c, std::complex *s, - const std::vector &dependencies) { +sycl::event rotg(oneapi::math::device libkey, sycl::queue& queue, std::complex* a, + std::complex* b, double* c, std::complex* s, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zrotg_usm_sycl(queue, a, b, c, s, dependencies); } -sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float *param, - const std::vector &dependencies) { +sycl::event rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float* param, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_srotm_usm_sycl(queue, n, x, incx, y, incy, param, dependencies); } -sycl::event rotm(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, double *param, - const std::vector &dependencies) { +sycl::event rotm(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double* param, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_drotm_usm_sycl(queue, n, x, incx, y, incy, param, dependencies); } -sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, float *d1, float *d2, float *x1, - float y1, float *param, const std::vector &dependencies) { +sycl::event rotmg(oneapi::math::device libkey, sycl::queue& queue, float* d1, float* d2, float* x1, + float y1, float* param, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_srotmg_usm_sycl(queue, d1, d2, x1, y1, param, dependencies); } -sycl::event rotmg(oneapi::math::device libkey, sycl::queue &queue, double *d1, double *d2, - double *x1, double y1, double *param, - const std::vector &dependencies) { +sycl::event rotmg(oneapi::math::device libkey, sycl::queue& queue, double* d1, double* d2, + double* x1, double y1, double* param, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_drotmg_usm_sycl(queue, d1, d2, x1, y1, param, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - float *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + float* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - double *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + double* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex alpha, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_csscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event scal(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event scal(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zdscal_usm_sycl(queue, n, alpha, x, incx, dependencies); } -sycl::event sdsdot(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float sb, - const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *result, const std::vector &dependencies) { +sycl::event sdsdot(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float sb, + const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* result, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sdsdot_usm_sycl(queue, n, sb, x, incx, y, incy, result, dependencies); } -sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event swap(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event swap(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zswap_usm_sycl(queue, n, x, incx, y, incy, dependencies); } -sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, float beta, float *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, float beta, float* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sgbmv_usm_sycl( queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, double beta, double *y, - std::int64_t incy, const std::vector &dependencies) { +sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, double beta, double* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dgbmv_usm_sycl( queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cgbmv_usm_sycl( queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gbmv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, +sycl::event gbmv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *x, - std::int64_t incx, std::complex beta, std::complex *y, - std::int64_t incy, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* x, + std::int64_t incx, std::complex beta, std::complex* y, + std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zgbmv_usm_sycl( queue, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sgemv_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dgemv_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cgemv_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv(oneapi::math::device libkey, sycl::queue &queue, transpose trans, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event gemv(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zgemv_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float beta, float *y, std::int64_t incy, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float beta, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sgemv_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double beta, double *y, std::int64_t incy, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double beta, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dgemv_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cgemv_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zgemv_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, x, incx, stridex, beta, y, incy, stridey, batch_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, const float **x, std::int64_t *incx, float *beta, - float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, const float** x, std::int64_t* incx, float* beta, + float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sgemv_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, const double **x, std::int64_t *incx, double *beta, - double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, const double** x, std::int64_t* incx, double* beta, + double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dgemv_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cgemv_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); } -sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, - std::complex *beta, std::complex **y, std::int64_t *incy, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemv_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, + std::complex* beta, std::complex** y, std::int64_t* incy, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zgemv_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, x, incx, beta, y, incy, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, const float *a, std::int64_t lda, - std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, const float* a, std::int64_t lda, + std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sdgmm_batch_strided_usm_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, const double *a, std::int64_t lda, - std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, const double* a, std::int64_t lda, + std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ddgmm_batch_strided_usm_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cdgmm_batch_strided_usm_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, - std::int64_t m, std::int64_t n, const std::complex *a, - std::int64_t lda, std::int64_t stridea, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *c, +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, + std::int64_t m, std::int64_t n, const std::complex* a, + std::int64_t lda, std::int64_t stridea, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zdgmm_batch_strided_usm_sycl( queue, left_right, m, n, a, lda, stridea, x, incx, stridex, c, ldc, stridec, batch_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - std::int64_t *m, std::int64_t *n, const float **a, std::int64_t *lda, - const float **x, std::int64_t *incx, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + std::int64_t* m, std::int64_t* n, const float** a, std::int64_t* lda, + const float** x, std::int64_t* incx, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sdgmm_batch_group_usm_sycl( queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - std::int64_t *m, std::int64_t *n, const double **a, std::int64_t *lda, - const double **x, std::int64_t *incx, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + std::int64_t* m, std::int64_t* n, const double** a, std::int64_t* lda, + const double** x, std::int64_t* incx, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ddgmm_batch_group_usm_sycl( queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cdgmm_batch_group_usm_sycl( queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - std::int64_t *m, std::int64_t *n, const std::complex **a, - std::int64_t *lda, const std::complex **x, std::int64_t *incx, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event dgmm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + std::int64_t* m, std::int64_t* n, const std::complex** a, + std::int64_t* lda, const std::complex** x, std::int64_t* incx, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zdgmm_batch_group_usm_sycl( queue, left_right, m, n, a, lda, x, incx, c, ldc, group_count, group_size, dependencies); } -sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *a, std::int64_t lda, const std::vector &dependencies) { +sycl::event ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sger_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event ger(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event ger(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dger_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cgerc_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event gerc(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event gerc(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zgerc_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cgeru_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event geru(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event geru(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zgeru_usm_sycl( queue, m, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_chbmv_usm_sycl( queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zhbmv_usm_sycl( queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_chemv_usm_sycl( queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event hemv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hemv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zhemv_usm_sycl( queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cher_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event her(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event her(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zher_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cher2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event her2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - std::int64_t lda, const std::vector &dependencies) { +sycl::event her2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zher2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_chpmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event hpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *a, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event hpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* a, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zhpmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { +sycl::event hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_chpr_usm_sycl(queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event hpr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const std::complex *x, std::int64_t incx, - std::complex *a, const std::vector &dependencies) { +sycl::event hpr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const std::complex* x, std::int64_t incx, + std::complex* a, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zhpr_usm_sycl(queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { +sycl::event hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_chpr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event hpr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::complex alpha, const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, std::complex *a, - const std::vector &dependencies) { +sycl::event hpr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::complex alpha, const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, std::complex* a, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zhpr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, float alpha, const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ssbmv_usm_sycl( queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event sbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event sbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + std::int64_t k, double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dsbmv_usm_sycl( queue, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *a, const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, const std::vector &dependencies) { +sycl::event spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* a, const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sspmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event spmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *a, const double *x, std::int64_t incx, double beta, - double *y, std::int64_t incy, const std::vector &dependencies) { +sycl::event spmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* a, const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dspmv_usm_sycl( queue, upper_lower, n, alpha, a, x, incx, beta, y, incy, dependencies); } -sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, - const std::vector &dependencies) { +sycl::event spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, float* a, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sspr_usm_sycl(queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event spr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, - const std::vector &dependencies) { +sycl::event spr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, double* a, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dspr_usm_sycl(queue, upper_lower, n, alpha, x, incx, a, dependencies); } -sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *a, const std::vector &dependencies) { +sycl::event spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sspr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event spr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, const std::vector &dependencies) { +sycl::event spr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dspr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, dependencies); } -sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *a, std::int64_t lda, const float *x, std::int64_t incx, - float beta, float *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* a, std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ssymv_usm_sycl( queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event symv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, std::int64_t incy, - const std::vector &dependencies) { +sycl::event symv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dsymv_usm_sycl( queue, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy, dependencies); } -sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ssyr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event syr(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event syr(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dsyr_usm_sycl( queue, upper_lower, n, alpha, x, incx, a, lda, dependencies); } -sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - float alpha, const float *x, std::int64_t incx, const float *y, std::int64_t incy, - float *a, std::int64_t lda, const std::vector &dependencies) { +sycl::event syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + float alpha, const float* x, std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ssyr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event syr2(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, std::int64_t n, - double alpha, const double *x, std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies) { +sycl::event syr2(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, std::int64_t n, + double alpha, const double* x, std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dsyr2_usm_sycl( queue, upper_lower, n, alpha, x, incx, y, incy, a, lda, dependencies); } -sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_stbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, - double *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dtbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ctbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tbmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ztbmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_stbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const double *a, std::int64_t lda, - double *x, std::int64_t incx, const std::vector &dependencies) { +sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dtbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ctbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tbsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tbsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ztbsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, k, a, lda, x, incx, dependencies); } -sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_stpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dtpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ctpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ztpmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, float *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, float* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_stpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, double *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, double* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dtpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ctpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event tpsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event tpsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ztpsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, x, incx, dependencies); } -sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_strmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dtrmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ctrmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trmv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event trmv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ztrmv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const float *a, std::int64_t lda, float *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const float* a, std::int64_t lda, float* x, + std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_strsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const double *a, std::int64_t lda, double *x, - std::int64_t incx, const std::vector &dependencies) { +sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const double* a, std::int64_t lda, double* x, + std::int64_t incx, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dtrsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ctrsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event trsv(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t n, const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies) { +sycl::event trsv(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + diag unit_diag, std::int64_t n, const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ztrsv_usm_sycl( queue, upper_lower, trans, unit_diag, n, a, lda, x, incx, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const float* a, std::int64_t lda, const float* b, std::int64_t ldb, float beta, + float* c, std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sgemm_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, double beta, double *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, + const double* a, std::int64_t lda, const double* b, std::int64_t ldb, double beta, + double* c, std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dgemm_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cgemm_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zgemm_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, const sycl::half *b, std::int64_t ldb, - sycl::half beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_hgemm_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_f16f16f32_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm(oneapi::math::device libkey, sycl::queue &queue, transpose transa, transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const bfloat16 *a, - std::int64_t lda, const bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event gemm(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const bfloat16* a, std::int64_t lda, const bfloat16* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_bf16bf16f32_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_chemm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event hemm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event hemm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zhemm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const std::complex *a, - std::int64_t lda, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, const std::complex* a, + std::int64_t lda, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cherk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event herk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const std::complex *a, - std::int64_t lda, double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event herk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, const std::complex* a, + std::int64_t lda, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zherk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cher2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event her2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event her2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zher2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + const float* b, std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ssymm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, - std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + const double* b, std::int64_t ldb, double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dsymm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_csymm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event symm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event symm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zsymm_usm_sycl( queue, left_right, upper_lower, m, n, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ssyrk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dsyrk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_csyrk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, +sycl::event syrk(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zsyrk_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, float *alpha, - const float **a, std::int64_t *lda, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, float* alpha, + const float** a, std::int64_t* lda, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ssyrk_batch_group_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, double *alpha, - const double **a, std::int64_t *lda, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, double* alpha, + const double** a, std::int64_t* lda, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dsyrk_batch_group_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_csyrk_batch_group_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo *upper_lower, - transpose *trans, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo* upper_lower, + transpose* trans, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zsyrk_batch_group_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, - transpose trans, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float beta, float *c, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ssyrk_batch_strided_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dsyrk_batch_strided_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, std::complex *c, std::int64_t ldc, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_csyrk_batch_strided_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event syrk_batch(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, std::complex *c, std::int64_t ldc, + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zsyrk_batch_strided_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, stride_a, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - const float *b, std::int64_t ldb, float beta, float *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ssyr2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, - const double *b, std::int64_t ldb, double beta, double *c, std::int64_t ldc, - const std::vector &dependencies) { +sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, double beta, double* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dsyr2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_csyr2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event syr2k(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event syr2k(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, + transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zsyr2k_usm_sycl( queue, upper_lower, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_strmm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dtrmm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ctrmm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trmm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trmm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ztrmm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies) { + const float* a, std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_strsm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { + const double* a, std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dtrsm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ctrsm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm(oneapi::math::device libkey, sycl::queue &queue, side left_right, uplo upper_lower, +sycl::event trsm(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ztrsm_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_strsm_batch_strided_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dtrsm_batch_strided_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ctrsm_batch_strided_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side left_right, +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t m, - std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, + std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ztrsm_batch_strided_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_strsm_batch_group_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, - double **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dtrsm_batch_group_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ctrsm_batch_group_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); } -sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue &queue, side *left_right, - uplo *upper_lower, transpose *trans, diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event trsm_batch(oneapi::math::device libkey, sycl::queue& queue, side* left_right, + uplo* upper_lower, transpose* trans, diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_ztrsm_batch_group_usm_sycl( queue, left_right, upper_lower, trans, unit_diag, m, n, alpha, a, lda, b, ldb, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const float **a, std::int64_t *lda, const float **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const float** a, std::int64_t* lda, const float** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sgemm_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - double *alpha, const double **a, std::int64_t *lda, const double **b, - std::int64_t *ldb, double *beta, double **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + double* alpha, const double** a, std::int64_t* lda, const double** b, + std::int64_t* ldb, double* beta, double** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dgemm_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **b, std::int64_t *ldb, std::complex *beta, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** b, std::int64_t* ldb, std::complex* beta, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cgemm_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex *alpha, const std::complex **a, - std::int64_t *lda, const std::complex **b, std::int64_t *ldb, - std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex* alpha, const std::complex** a, + std::int64_t* lda, const std::complex** b, std::int64_t* ldb, + std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zgemm_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - sycl::half *alpha, const sycl::half **a, std::int64_t *lda, - const sycl::half **b, std::int64_t *ldb, sycl::half *beta, sycl::half **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + sycl::half* alpha, const sycl::half** a, std::int64_t* lda, + const sycl::half** b, std::int64_t* ldb, sycl::half* beta, sycl::half** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_hgemm_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const sycl::half **a, std::int64_t *lda, const sycl::half **b, - std::int64_t *ldb, float *beta, float **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const sycl::half** a, std::int64_t* lda, const sycl::half** b, + std::int64_t* ldb, float* beta, float** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_f16f16f32_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_s8s8f32_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *transa, - transpose *transb, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float *alpha, const std::int8_t **a, std::int64_t *lda, - const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies) { +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* transa, + transpose* transb, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float* alpha, const std::int8_t** a, std::int64_t* lda, + const std::int8_t** b, std::int64_t* ldb, float* beta, std::int32_t** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_s8s8s32_batch_group_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc, group_count, group_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, - const float *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, + const float* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sgemm_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, - const double *b, std::int64_t ldb, std::int64_t stride_b, double beta, - double *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, + const double* b, std::int64_t ldb, std::int64_t stride_b, double beta, + double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dgemm_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, const std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::complex beta, std::complex *c, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cgemm_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, const std::complex *b, std::int64_t ldb, - std::int64_t stride_b, std::complex beta, std::complex *c, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, const std::complex* b, std::int64_t ldb, + std::int64_t stride_b, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zgemm_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::half alpha, const sycl::half *a, std::int64_t lda, - std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, + sycl::half alpha, const sycl::half* a, std::int64_t lda, + std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, sycl::half* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_hgemm_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const sycl::half *a, std::int64_t lda, std::int64_t stride_a, - const sycl::half *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const sycl::half* a, std::int64_t lda, std::int64_t stride_a, + const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_f16f16f32_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_s8s8f32_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - float alpha, const std::int8_t *a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t *b, std::int64_t ldb, std::int64_t stride_b, float beta, - std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + float alpha, const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, + const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, float beta, + std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_s8s8s32_batch_strided_usm_sycl( queue, transa, transb, m, n, k, alpha, a, lda, stride_a, b, ldb, stride_b, beta, c, ldc, stride_c, batch_size, dependencies); } -sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, const float *b, std::int64_t ldb, float beta, - float *c, std::int64_t ldc, const std::vector &dependencies) { + const float* a, std::int64_t lda, const float* b, std::int64_t ldb, float beta, + float* c, std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_sgemmt_usm_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, const double *b, std::int64_t ldb, double beta, - double *c, std::int64_t ldc, const std::vector &dependencies) { + const double* a, std::int64_t lda, const double* b, std::int64_t ldb, double beta, + double* c, std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dgemmt_usm_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cgemmt_usm_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemmt(oneapi::math::device libkey, sycl::queue &queue, uplo upper_lower, +sycl::event gemmt(oneapi::math::device libkey, sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies) { + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zgemmt_usm_sycl( queue, upper_lower, transa, transb, n, k, alpha, a, lda, b, ldb, beta, c, ldc, dependencies); } -sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_s8u8s32_bias_usm_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } -sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::int8_t *a, std::int64_t lda, - std::int8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::int8_t* a, std::int64_t lda, + std::int8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_s8s8s32_bias_usm_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } -sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::int8_t *b, std::int64_t ldb, std::int8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::int8_t* b, std::int64_t ldb, std::int8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_u8s8s32_bias_usm_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } -sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event gemm_bias(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, offset offsetc, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, const std::uint8_t *a, std::int64_t lda, - std::uint8_t ao, const std::uint8_t *b, std::int64_t ldb, std::uint8_t bo, - float beta, std::int32_t *c, std::int64_t ldc, const std::int32_t *co, - const std::vector &dependencies) { + std::int64_t k, float alpha, const std::uint8_t* a, std::int64_t lda, + std::uint8_t ao, const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, + float beta, std::int32_t* c, std::int64_t ldc, const std::int32_t* co, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_gemm_u8u8s32_bias_usm_sycl( queue, transa, transb, offsetc, m, n, k, alpha, a, lda, ao, b, ldb, bo, beta, c, ldc, co, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, float *b, std::int64_t ldb, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_somatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, double *b, std::int64_t ldb, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_domatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_comatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zomatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, a, lda, stride_a, b, ldb, stride_b, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_simatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies) { + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dimatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cimatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, + std::complex* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zimatcopy_batch_strided_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, stride, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, std::int64_t stride_a, float beta, - const float *b, std::int64_t ldb, std::int64_t stride_b, float *c, + const float* a, std::int64_t lda, std::int64_t stride_a, float beta, + const float* b, std::int64_t ldb, std::int64_t stride_b, float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_somatadd_batch_strided_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, std::int64_t stride_a, double beta, - const double *b, std::int64_t ldb, std::int64_t stride_b, double *c, + const double* a, std::int64_t lda, std::int64_t stride_a, double beta, + const double* b, std::int64_t ldb, std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_domatadd_batch_strided_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_comatadd_batch_strided_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } -sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd_batch(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, + std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies) { + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zomatadd_batch_strided_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, stride_a, beta, b, ldb, stride_b, c, ldc, stride_c, batch_size, dependencies); } -sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - float *b, std::int64_t ldb, const std::vector &dependencies) { +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + float* b, std::int64_t ldb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_somatcopy_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies) { +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_domatcopy_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_comatcopy_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zomatcopy_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, dependencies); } -sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stridea, float *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_somatcopy2_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, std::int64_t stridea, double *b, std::int64_t ldb, - std::int64_t strideb, const std::vector &dependencies) { +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, std::int64_t stridea, double* b, std::int64_t ldb, + std::int64_t strideb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_domatcopy2_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - std::complex *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stridea, + std::complex* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_comatcopy2_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event omatcopy2(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - std::complex *b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::int64_t stridea, + std::complex* b, std::int64_t ldb, std::int64_t strideb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zomatcopy2_usm_sycl( queue, trans, m, n, alpha, a, lda, stridea, b, ldb, strideb, dependencies); } -sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, float alpha, float *ab, std::int64_t lda, - std::int64_t ldb, const std::vector &dependencies) { +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, float alpha, float* ab, std::int64_t lda, + std::int64_t ldb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_simatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, - std::int64_t m, std::int64_t n, double alpha, double *ab, std::int64_t lda, - std::int64_t ldb, const std::vector &dependencies) { +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, + std::int64_t m, std::int64_t n, double alpha, double* ab, std::int64_t lda, + std::int64_t ldb, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dimatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + std::complex* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cimatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event imatcopy(oneapi::math::device libkey, sycl::queue &queue, transpose trans, +sycl::event imatcopy(oneapi::math::device libkey, sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies) { + std::complex* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zimatcopy_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, dependencies); } -sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, float alpha, const float *a, - std::int64_t lda, float beta, const float *b, std::int64_t ldb, float *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, float alpha, const float* a, + std::int64_t lda, float beta, const float* b, std::int64_t ldb, float* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_somatadd_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, - transpose transb, std::int64_t m, std::int64_t n, double alpha, const double *a, - std::int64_t lda, double beta, const double *b, std::int64_t ldb, double *c, - std::int64_t ldc, const std::vector &dependencies) { +sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, + transpose transb, std::int64_t m, std::int64_t n, double alpha, const double* a, + std::int64_t lda, double beta, const double* b, std::int64_t ldb, double* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_domatadd_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_comatadd_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatadd(oneapi::math::device libkey, sycl::queue &queue, transpose transa, +sycl::event omatadd(oneapi::math::device libkey, sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, std::complex *c, - std::int64_t ldc, const std::vector &dependencies) { + const std::complex* a, std::int64_t lda, std::complex beta, + const std::complex* b, std::int64_t ldb, std::complex* c, + std::int64_t ldc, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zomatadd_usm_sycl( queue, transa, transb, m, n, alpha, a, lda, beta, b, ldb, c, ldc, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, const float **a, - std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, const float** a, + std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_somatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, const double **a, - std::int64_t *lda, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, const double** a, + std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_domatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_comatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event omatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zomatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, a, lda, b, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, float *alpha, float **ab, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, float* alpha, float** ab, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_simatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, double *alpha, double **ab, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies) { +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, double* alpha, double** ab, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_dimatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - std::complex **ab, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + std::complex** ab, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_cimatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } -sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue &queue, transpose *trans, - std::int64_t *m, std::int64_t *n, std::complex *alpha, - std::complex **ab, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies) { +sycl::event imatcopy_batch(oneapi::math::device libkey, sycl::queue& queue, transpose* trans, + std::int64_t* m, std::int64_t* n, std::complex* alpha, + std::complex** ab, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].row_major_zimatcopy_batch_group_usm_sycl( queue, trans, m, n, alpha, ab, lda, ldb, group_count, groupsize, dependencies); } diff --git a/src/blas/function_table.hpp b/src/blas/function_table.hpp index 71321bdc7..cd07f9dca 100644 --- a/src/blas/function_table.hpp +++ b/src/blas/function_table.hpp @@ -34,4940 +34,4942 @@ typedef struct { // Buffer APIs - void (*column_major_scasum_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*column_major_dzasum_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*column_major_sasum_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*column_major_dasum_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*column_major_saxpy_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); - void (*column_major_daxpy_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); - void (*column_major_caxpy_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zaxpy_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_saxpy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, + void (*column_major_scasum_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*column_major_dzasum_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*column_major_sasum_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*column_major_dasum_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*column_major_saxpy_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); + void (*column_major_daxpy_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); + void (*column_major_caxpy_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_zaxpy_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_saxpy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_daxpy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, + void (*column_major_daxpy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_caxpy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, + void (*column_major_caxpy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_zaxpy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, + void (*column_major_zaxpy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_saxpby_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); - void (*column_major_daxpby_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - void (*column_major_caxpby_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, + void (*column_major_saxpby_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); + void (*column_major_daxpby_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + void (*column_major_caxpby_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zaxpby_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_zaxpby_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_scopy_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_scopy_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); - void (*column_major_dcopy_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, + void (*column_major_dcopy_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); - void (*column_major_ccopy_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zcopy_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_scopy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, + void (*column_major_ccopy_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_zcopy_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_scopy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_dcopy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, + void (*column_major_dcopy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_ccopy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, + void (*column_major_ccopy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_zcopy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, + void (*column_major_zcopy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_sdot_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); - void (*column_major_ddot_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); - void (*column_major_dsdot_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); - void (*column_major_cdotc_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - void (*column_major_zdotc_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - void (*column_major_cdotu_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - void (*column_major_zdotu_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - void (*column_major_isamin_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*column_major_idamin_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*column_major_icamin_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*column_major_izamin_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*column_major_isamax_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*column_major_idamax_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*column_major_icamax_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*column_major_izamax_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*column_major_scnrm2_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*column_major_dznrm2_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*column_major_snrm2_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*column_major_dnrm2_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*column_major_srot_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, + void (*column_major_sdot_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + void (*column_major_ddot_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + void (*column_major_dsdot_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + void (*column_major_cdotc_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + void (*column_major_zdotc_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + void (*column_major_cdotu_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + void (*column_major_zdotu_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + void (*column_major_isamin_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*column_major_idamin_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*column_major_icamin_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*column_major_izamin_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*column_major_isamax_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*column_major_idamax_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*column_major_icamax_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*column_major_izamax_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*column_major_scnrm2_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*column_major_dznrm2_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*column_major_snrm2_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*column_major_dnrm2_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*column_major_srot_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, float c, float s); - void (*column_major_drot_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, + void (*column_major_drot_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, double s); - void (*column_major_csrot_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, + void (*column_major_csrot_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, float c, float s); - void (*column_major_zdrot_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, + void (*column_major_zdrot_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, double c, double s); - void (*column_major_srotg_sycl)(sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s); - void (*column_major_drotg_sycl)(sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s); - void (*column_major_crotg_sycl)(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, - sycl::buffer &c, - sycl::buffer, 1> &s); - void (*column_major_zrotg_sycl)(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, - sycl::buffer &c, - sycl::buffer, 1> &s); - void (*column_major_srotm_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m); - void (*column_major_drotm_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer ¶m); - void (*column_major_srotmg_sycl)(sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, - float y1, sycl::buffer ¶m); - void (*column_major_drotmg_sycl)(sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, - double y1, sycl::buffer ¶m); - void (*column_major_sscal_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx); - void (*column_major_dscal_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx); - void (*column_major_cscal_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_csscal_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_zscal_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_zdscal_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_sdsdot_sycl)(sycl::queue &queue, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); - void (*column_major_sswap_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, + void (*column_major_srotg_sycl)(sycl::queue& queue, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, + sycl::buffer& s); + void (*column_major_drotg_sycl)(sycl::queue& queue, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, + sycl::buffer& s); + void (*column_major_crotg_sycl)(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, + sycl::buffer& c, + sycl::buffer, 1>& s); + void (*column_major_zrotg_sycl)(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, + sycl::buffer& c, + sycl::buffer, 1>& s); + void (*column_major_srotm_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param); + void (*column_major_drotm_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& param); + void (*column_major_srotmg_sycl)(sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, + float y1, sycl::buffer& param); + void (*column_major_drotmg_sycl)(sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, + double y1, sycl::buffer& param); + void (*column_major_sscal_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx); + void (*column_major_dscal_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx); + void (*column_major_cscal_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_csscal_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_zscal_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_zdscal_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_sdsdot_sycl)(sycl::queue& queue, std::int64_t n, float sb, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + void (*column_major_sswap_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); - void (*column_major_dswap_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, + void (*column_major_dswap_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); - void (*column_major_cswap_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zswap_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_sgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*column_major_cswap_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_zswap_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_sgbmv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_dgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + std::int64_t ku, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy); + void (*column_major_dgbmv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_cgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + std::int64_t ku, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy); + void (*column_major_cgbmv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_zgbmv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_sgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_sgemv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); - void (*column_major_dgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); + void (*column_major_dgemv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - void (*column_major_cgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + void (*column_major_cgemv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_zgemv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_sgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_sgemv_batch_strided_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, std::int64_t stridex, - float beta, sycl::buffer &y, + float beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_dgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*column_major_dgemv_batch_strided_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, std::int64_t stridex, - double beta, sycl::buffer &y, + double beta, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); void (*column_major_cgemv_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, std::complex beta, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); void (*column_major_zgemv_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, std::complex beta, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*column_major_sdgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::side left_right, + void (*column_major_sdgmm_batch_strided_sycl)(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); - void (*column_major_ddgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::side left_right, + void (*column_major_ddgmm_batch_strided_sycl)(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); void (*column_major_cdgmm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); void (*column_major_zdgmm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); - void (*column_major_sger_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - void (*column_major_dger_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - void (*column_major_cgerc_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + void (*column_major_sger_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda); + void (*column_major_dger_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda); + void (*column_major_cgerc_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_zgerc_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + void (*column_major_zgerc_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_cgeru_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + void (*column_major_cgeru_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_zgeru_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + void (*column_major_zgeru_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_chbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + void (*column_major_chbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zhbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_zhbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_chemv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_chemv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zhemv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_zhemv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_cher_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_cher_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_zher_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda); + void (*column_major_zher_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_cher2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda); + void (*column_major_cher2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_zher2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + void (*column_major_zher2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - void (*column_major_chpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + void (*column_major_chpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_zhpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_zhpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*column_major_chpr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& y, std::int64_t incy); + void (*column_major_chpr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a); - void (*column_major_zhpr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a); + void (*column_major_zhpr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &a); - void (*column_major_chpr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& a); + void (*column_major_chpr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a); - void (*column_major_zhpr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a); + void (*column_major_zhpr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a); - void (*column_major_ssbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a); + void (*column_major_ssbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); - void (*column_major_dsbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); + void (*column_major_dsbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - void (*column_major_sspmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); - void (*column_major_dspmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - void (*column_major_sspr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a); - void (*column_major_dspr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a); - void (*column_major_sspr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a); - void (*column_major_dspr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a); - void (*column_major_ssymv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_dsymv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy); - void (*column_major_ssyr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a, std::int64_t lda); - void (*column_major_dsyr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &a, std::int64_t lda); - void (*column_major_ssyr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - void (*column_major_dsyr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, - std::int64_t incy, sycl::buffer &a, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + void (*column_major_sspmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); + void (*column_major_dspmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + void (*column_major_sspr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& a); + void (*column_major_dspr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& a); + void (*column_major_sspr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a); + void (*column_major_dspr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a); + void (*column_major_ssymv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy); + void (*column_major_dsymv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy); + void (*column_major_ssyr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& a, std::int64_t lda); + void (*column_major_dsyr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& a, std::int64_t lda); + void (*column_major_ssyr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda); + void (*column_major_dsyr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, + std::int64_t incy, sycl::buffer& a, std::int64_t lda); - void (*column_major_stbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*column_major_stbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*column_major_dtbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); + void (*column_major_dtbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); - void (*column_major_ctbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*column_major_ctbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_ztbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_ztbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_stbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_stbsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*column_major_dtbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); + void (*column_major_dtbsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); - void (*column_major_ctbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*column_major_ctbsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_ztbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_ztbsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_stpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_stpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx); - void (*column_major_dtpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); + void (*column_major_dtpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx); - void (*column_major_ctpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); + void (*column_major_ctpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_ztpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_ztpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_stpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_stpsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx); - void (*column_major_dtpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); + void (*column_major_dtpsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx); - void (*column_major_ctpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); + void (*column_major_ctpsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_ztpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_ztpsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - void (*column_major_strmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + void (*column_major_strmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - void (*column_major_dtrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + void (*column_major_dtrmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - void (*column_major_ctrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + void (*column_major_ctrmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx); - void (*column_major_ztrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*column_major_ztrmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx); - void (*column_major_strsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*column_major_strsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - void (*column_major_dtrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + void (*column_major_dtrsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - void (*column_major_ctrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + void (*column_major_ctrsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx); - void (*column_major_ztrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*column_major_ztrsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx); - void (*column_major_sgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + void (*column_major_sgemm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - float beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_dgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + float beta, sycl::buffer& c, std::int64_t ldc); + void (*column_major_dgemm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - double beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_cgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + double beta, sycl::buffer& c, std::int64_t ldc); + void (*column_major_cgemm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_zgemm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_hgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_hgemm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, - sycl::half beta, sycl::buffer &c, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + sycl::half beta, sycl::buffer& c, std::int64_t ldc); - void (*column_major_gemm_f16f16f32_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + void (*column_major_gemm_f16f16f32_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, - float beta, sycl::buffer &c, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + float beta, sycl::buffer& c, std::int64_t ldc); - void (*column_major_gemm_bf16bf16f32_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + void (*column_major_gemm_bf16bf16f32_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, + sycl::buffer& a, std::int64_t lda, - sycl::buffer &b, + sycl::buffer& b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - void (*column_major_chemm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::buffer& c, std::int64_t ldc); + void (*column_major_chemm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zhemm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_zhemm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_cherk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_cherk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer, 1> &a, + float alpha, sycl::buffer, 1>& a, std::int64_t lda, float beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zherk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_zherk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer, 1> &a, + double alpha, sycl::buffer, 1>& a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_cher2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_cher2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - float beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + float beta, sycl::buffer, 1>& c, std::int64_t ldc); - void (*column_major_zher2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*column_major_zher2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - double beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + double beta, sycl::buffer, 1>& c, std::int64_t ldc); - void (*column_major_ssymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + void (*column_major_ssymm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - void (*column_major_dsymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); + void (*column_major_dsymm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - void (*column_major_csymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); + void (*column_major_csymm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zsymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_zsymm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_ssyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_ssyrk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - float beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_dsyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& c, std::int64_t ldc); + void (*column_major_dsyrk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - double beta, sycl::buffer &c, std::int64_t ldc); - void (*column_major_csyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + double alpha, sycl::buffer& a, std::int64_t lda, + double beta, sycl::buffer& c, std::int64_t ldc); + void (*column_major_csyrk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zsyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_zsyrk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_ssyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, float beta, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - void (*column_major_dsyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_ssyrk_batch_strided_sycl)( + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, float beta, sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); + void (*column_major_dsyrk_batch_strided_sycl)( + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, double beta, sycl::buffer& c, std::int64_t ldc, + std::int64_t stride_c, std::int64_t batch_size); void (*column_major_csyrk_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_zsyrk_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*column_major_ssyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*column_major_ssyr2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - void (*column_major_dsyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); + void (*column_major_dsyr2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - void (*column_major_csyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); + void (*column_major_csyr2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zsyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_zsyr2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_strmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_strmm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*column_major_dtrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*column_major_dtrmm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*column_major_ctrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*column_major_ctrmm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - void (*column_major_ztrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + void (*column_major_ztrmm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - void (*column_major_strsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + void (*column_major_strsm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*column_major_dtrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*column_major_dtrsm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*column_major_ctrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*column_major_ctrsm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - void (*column_major_ztrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + void (*column_major_ztrsm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); void (*column_major_sgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_dgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, double beta, sycl::buffer &c, std::int64_t ldc, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, double beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_cgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_zgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*column_major_hgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, - sycl::buffer &c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); + void (*column_major_hgemm_batch_strided_sycl)( + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size); void (*column_major_gemm_f16f16f32_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_gemm_s8s8f32_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_gemm_s8s8s32_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_strsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + float alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_dtrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + double alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_ctrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_ztrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - void (*column_major_sgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*column_major_sgemmt_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - void (*column_major_dgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); + void (*column_major_dgemmt_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - void (*column_major_cgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); + void (*column_major_cgemmt_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_zgemmt_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& c, std::int64_t ldc); void (*column_major_gemm_s8u8s32_bias_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); void (*column_major_gemm_s8s8s32_bias_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); void (*column_major_gemm_u8s8s32_bias_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, - std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); void (*column_major_gemm_u8u8s32_bias_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, - std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); void (*column_major_somatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_domatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_comatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*column_major_zomatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - void (*column_major_simatcopy_batch_strided_sycl)(sycl::queue &queue, + void (*column_major_simatcopy_batch_strided_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &ab, std::int64_t lda, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - void (*column_major_dimatcopy_batch_strided_sycl)(sycl::queue &queue, + void (*column_major_dimatcopy_batch_strided_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &ab, std::int64_t lda, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - void (*column_major_cimatcopy_batch_strided_sycl)(sycl::queue &queue, + void (*column_major_cimatcopy_batch_strided_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - void (*column_major_zimatcopy_batch_strided_sycl)(sycl::queue &queue, + void (*column_major_zimatcopy_batch_strided_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); void (*column_major_somatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, float beta, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_domatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, double beta, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_comatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*column_major_zomatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*column_major_somatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*column_major_somatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*column_major_domatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*column_major_domatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*column_major_comatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*column_major_comatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - void (*column_major_zomatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + void (*column_major_zomatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - void (*column_major_somatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + void (*column_major_somatcopy2_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); - void (*column_major_domatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*column_major_domatcopy2_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); - void (*column_major_comatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*column_major_comatcopy2_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb); - void (*column_major_zomatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*column_major_zomatcopy2_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb); - void (*column_major_simatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*column_major_simatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &ab, std::int64_t lda, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); - void (*column_major_dimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*column_major_dimatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &ab, std::int64_t lda, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); - void (*column_major_cimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*column_major_cimatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb); - void (*column_major_zimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*column_major_zimatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb); - void (*column_major_somatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + void (*column_major_somatadd_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, float beta, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &c, + std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, float beta, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& c, std::int64_t ldc); - void (*column_major_domatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + void (*column_major_domatadd_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, double beta, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &c, + std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, double beta, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& c, std::int64_t ldc); - void (*column_major_comatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + void (*column_major_comatadd_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*column_major_zomatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*column_major_zomatadd_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); // USM APIs - sycl::event (*column_major_scasum_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - float *result, - const std::vector &dependencies); - sycl::event (*column_major_dzasum_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - double *result, - const std::vector &dependencies); - sycl::event (*column_major_sasum_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies); - sycl::event (*column_major_dasum_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies); - sycl::event (*column_major_saxpy_usm_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *y, + sycl::event (*column_major_scasum_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + float* result, + const std::vector& dependencies); + sycl::event (*column_major_dzasum_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + double* result, + const std::vector& dependencies); + sycl::event (*column_major_sasum_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, + const std::vector& dependencies); + sycl::event (*column_major_dasum_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, + const std::vector& dependencies); + sycl::event (*column_major_saxpy_usm_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, float* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_daxpy_usm_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *y, + const std::vector& dependencies); + sycl::event (*column_major_daxpy_usm_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + const double* x, std::int64_t incx, double* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_caxpy_usm_sycl)(sycl::queue &queue, std::int64_t n, + const std::vector& dependencies); + sycl::event (*column_major_caxpy_usm_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_zaxpy_usm_sycl)(sycl::queue &queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_zaxpy_usm_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); sycl::event (*column_major_saxpy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, float *alpha, const float **x, std::int64_t *incx, - float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, float* alpha, const float** x, std::int64_t* incx, + float** y, std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_daxpy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, double *alpha, const double **x, std::int64_t *incx, - double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, double* alpha, const double** x, std::int64_t* incx, + double** y, std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_caxpy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, std::complex* alpha, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_zaxpy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, std::complex* alpha, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_saxpy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, - std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t n, float alpha, const float* x, std::int64_t incx, + std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_daxpy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, - std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t n, double alpha, const double* x, std::int64_t incx, + std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_caxpy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, + sycl::queue& queue, std::int64_t n, std::complex alpha, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_zaxpy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); - sycl::event (*column_major_saxpby_usm_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float beta, - float *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_daxpby_usm_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, - const double beta, double *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_caxpby_usm_sycl)(sycl::queue &queue, std::int64_t n, + sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies); + sycl::event (*column_major_saxpby_usm_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, const float beta, + float* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_daxpby_usm_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + const double* x, std::int64_t incx, + const double beta, double* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_caxpby_usm_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, + const std::complex* x, std::int64_t incx, const std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_zaxpby_usm_sycl)(sycl::queue &queue, std::int64_t n, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_zaxpby_usm_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, + const std::complex* x, std::int64_t incx, const std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_scopy_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_dcopy_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_ccopy_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_zcopy_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); + std::complex* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_scopy_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_dcopy_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_ccopy_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_zcopy_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); sycl::event (*column_major_scopy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, const float **x, std::int64_t *incx, float **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, const float** x, std::int64_t* incx, float** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_dcopy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, const double **x, std::int64_t *incx, double **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, const double** x, std::int64_t* incx, double** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_ccopy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_zcopy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_scopy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t stridex, - float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, std::int64_t stridex, + float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies); sycl::event (*column_major_dcopy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_ccopy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t n, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_zcopy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); - sycl::event (*column_major_sdot_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - float *result, - const std::vector &dependencies); - sycl::event (*column_major_ddot_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, - double *result, - const std::vector &dependencies); - sycl::event (*column_major_dsdot_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - double *result, - const std::vector &dependencies); - sycl::event (*column_major_cdotc_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies); - sycl::event (*column_major_zdotc_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies); - sycl::event (*column_major_cdotu_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies); - sycl::event (*column_major_zdotu_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies); - sycl::event (*column_major_isamin_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies); - sycl::event (*column_major_idamin_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies); - sycl::event (*column_major_icamin_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies); - sycl::event (*column_major_izamin_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies); - sycl::event (*column_major_isamax_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies); - sycl::event (*column_major_idamax_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies); - sycl::event (*column_major_icamax_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies); - sycl::event (*column_major_izamax_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies); - sycl::event (*column_major_scnrm2_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - float *result, - const std::vector &dependencies); - sycl::event (*column_major_dznrm2_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - double *result, - const std::vector &dependencies); - sycl::event (*column_major_snrm2_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies); - sycl::event (*column_major_dnrm2_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies); - sycl::event (*column_major_srot_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, float c, + sycl::queue& queue, std::int64_t n, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); + sycl::event (*column_major_sdot_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, + float* result, + const std::vector& dependencies); + sycl::event (*column_major_ddot_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, + double* result, + const std::vector& dependencies); + sycl::event (*column_major_dsdot_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, + double* result, + const std::vector& dependencies); + sycl::event (*column_major_cdotc_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies); + sycl::event (*column_major_zdotc_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies); + sycl::event (*column_major_cdotu_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies); + sycl::event (*column_major_zdotu_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies); + sycl::event (*column_major_isamin_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies); + sycl::event (*column_major_idamin_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies); + sycl::event (*column_major_icamin_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies); + sycl::event (*column_major_izamin_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies); + sycl::event (*column_major_isamax_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies); + sycl::event (*column_major_idamax_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies); + sycl::event (*column_major_icamax_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies); + sycl::event (*column_major_izamax_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies); + sycl::event (*column_major_scnrm2_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + float* result, + const std::vector& dependencies); + sycl::event (*column_major_dznrm2_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + double* result, + const std::vector& dependencies); + sycl::event (*column_major_snrm2_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, + const std::vector& dependencies); + sycl::event (*column_major_dnrm2_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, + const std::vector& dependencies); + sycl::event (*column_major_srot_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, float c, float s, - const std::vector &dependencies); - sycl::event (*column_major_drot_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, double c, + const std::vector& dependencies); + sycl::event (*column_major_drot_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, double c, double s, - const std::vector &dependencies); - sycl::event (*column_major_csrot_usm_sycl)(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_csrot_usm_sycl)(sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float c, float s, - const std::vector &dependencies); - sycl::event (*column_major_zdrot_usm_sycl)(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_zdrot_usm_sycl)(sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double c, double s, - const std::vector &dependencies); - sycl::event (*column_major_srotg_usm_sycl)(sycl::queue &queue, float *a, float *b, float *c, - float *s, - const std::vector &dependencies); - sycl::event (*column_major_drotg_usm_sycl)(sycl::queue &queue, double *a, double *b, double *c, - double *s, - const std::vector &dependencies); - sycl::event (*column_major_crotg_usm_sycl)(sycl::queue &queue, std::complex *a, - std::complex *b, float *c, - std::complex *s, - const std::vector &dependencies); - sycl::event (*column_major_zrotg_usm_sycl)(sycl::queue &queue, std::complex *a, - std::complex *b, double *c, - std::complex *s, - const std::vector &dependencies); - sycl::event (*column_major_srotm_usm_sycl)(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - float *param, - const std::vector &dependencies); - sycl::event (*column_major_drotm_usm_sycl)(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - double *param, - const std::vector &dependencies); - sycl::event (*column_major_srotmg_usm_sycl)(sycl::queue &queue, float *d1, float *d2, float *x1, - float y1, float *param, - const std::vector &dependencies); - sycl::event (*column_major_drotmg_usm_sycl)(sycl::queue &queue, double *d1, double *d2, - double *x1, double y1, double *param, - const std::vector &dependencies); - sycl::event (*column_major_sscal_usm_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - float *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_dscal_usm_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - double *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_cscal_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, + const std::vector& dependencies); + sycl::event (*column_major_srotg_usm_sycl)(sycl::queue& queue, float* a, float* b, float* c, + float* s, + const std::vector& dependencies); + sycl::event (*column_major_drotg_usm_sycl)(sycl::queue& queue, double* a, double* b, double* c, + double* s, + const std::vector& dependencies); + sycl::event (*column_major_crotg_usm_sycl)(sycl::queue& queue, std::complex* a, + std::complex* b, float* c, + std::complex* s, + const std::vector& dependencies); + sycl::event (*column_major_zrotg_usm_sycl)(sycl::queue& queue, std::complex* a, + std::complex* b, double* c, + std::complex* s, + const std::vector& dependencies); + sycl::event (*column_major_srotm_usm_sycl)(sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, + float* param, + const std::vector& dependencies); + sycl::event (*column_major_drotm_usm_sycl)(sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, + double* param, + const std::vector& dependencies); + sycl::event (*column_major_srotmg_usm_sycl)(sycl::queue& queue, float* d1, float* d2, float* x1, + float y1, float* param, + const std::vector& dependencies); + sycl::event (*column_major_drotmg_usm_sycl)(sycl::queue& queue, double* d1, double* d2, + double* x1, double y1, double* param, + const std::vector& dependencies); + sycl::event (*column_major_sscal_usm_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + float* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_dscal_usm_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + double* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_cscal_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex alpha, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_csscal_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, + const std::vector& dependencies); + sycl::event (*column_major_csscal_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex alpha, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_zscal_usm_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_zdscal_usm_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_sdsdot_usm_sycl)(sycl::queue &queue, std::int64_t n, float sb, - const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *result, - const std::vector &dependencies); - sycl::event (*column_major_sswap_usm_sycl)(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_dswap_usm_sycl)(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_cswap_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_zswap_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_sgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*column_major_zscal_usm_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_zdscal_usm_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_sdsdot_usm_sycl)(sycl::queue& queue, std::int64_t n, float sb, + const float* x, std::int64_t incx, const float* y, + std::int64_t incy, float* result, + const std::vector& dependencies); + sycl::event (*column_major_sswap_usm_sycl)(sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_dswap_usm_sycl)(sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_cswap_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_zswap_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_sgbmv_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, - float beta, float *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_dgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + std::int64_t ku, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_dgbmv_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies); + std::int64_t ku, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, + double beta, double* y, std::int64_t incy, + const std::vector& dependencies); sycl::event (*column_major_cgbmv_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_zgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, const std::vector& dependencies); + sycl::event (*column_major_zgbmv_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_sgemv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*column_major_sgemv_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, + const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_dgemv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*column_major_dgemv_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, + const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_cgemv_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, const std::vector& dependencies); sycl::event (*column_major_zgemv_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, const std::vector& dependencies); sycl::event (*column_major_sgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, - std::int64_t incx, std::int64_t stridex, float beta, float *y, std::int64_t incy, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, std::int64_t stridea, const float* x, + std::int64_t incx, std::int64_t stridex, float beta, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_dgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, - std::int64_t incx, std::int64_t stridex, double beta, double *y, std::int64_t incy, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, std::int64_t stridea, const double* x, + std::int64_t incx, std::int64_t stridex, double beta, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_cgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_zgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_sgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - float *alpha, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, - float *beta, float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + float* alpha, const float** a, std::int64_t* lda, const float** x, std::int64_t* incx, + float* beta, float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_dgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - double *alpha, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, - double *beta, double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + double* alpha, const double** a, std::int64_t* lda, const double** x, std::int64_t* incx, + double* beta, double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_cgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_zgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_sdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + const float* a, std::int64_t lda, std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_ddgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + const double* a, std::int64_t lda, std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_cdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies); sycl::event (*column_major_zdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies); sycl::event (*column_major_sdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, - const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, std::int64_t* n, + const float** a, std::int64_t* lda, const float** x, std::int64_t* incx, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_ddgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, - const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, std::int64_t* n, + const double** a, std::int64_t* lda, const double** x, std::int64_t* incx, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_cdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, - const std::complex **a, std::int64_t *lda, const std::complex **x, - std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, const std::complex** x, + std::int64_t* incx, std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_zdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, - const std::complex **a, std::int64_t *lda, const std::complex **x, - std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); - sycl::event (*column_major_sger_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - float alpha, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *a, + sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, const std::complex** x, + std::int64_t* incx, std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); + sycl::event (*column_major_sger_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + float alpha, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_dger_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - double alpha, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *a, + const std::vector& dependencies); + sycl::event (*column_major_dger_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + double alpha, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_cgerc_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + const std::vector& dependencies); + sycl::event (*column_major_cgerc_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_zgerc_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_zgerc_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_cgeru_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_cgeru_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_zgeru_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_zgeru_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); sycl::event (*column_major_chbmv_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, const std::vector& dependencies); sycl::event (*column_major_zhbmv_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*column_major_chemv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, const std::vector& dependencies); + sycl::event (*column_major_chemv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_zhemv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_zhemv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_cher_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_cher_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_zher_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_zher_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_cher2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_cher2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_zher2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_zher2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_chpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_chpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, + const std::complex* a, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_zhpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_zhpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, + const std::complex* a, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_chpr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_chpr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, - const std::vector &dependencies); - sycl::event (*column_major_zhpr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + std::complex* a, + const std::vector& dependencies); + sycl::event (*column_major_zhpr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, - const std::vector &dependencies); - sycl::event (*column_major_chpr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + std::complex* a, + const std::vector& dependencies); + sycl::event (*column_major_chpr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, - const std::vector &dependencies); - sycl::event (*column_major_zhpr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies); + sycl::event (*column_major_zhpr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, - const std::vector &dependencies); - sycl::event (*column_major_ssbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies); + sycl::event (*column_major_ssbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, + const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_dsbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_dsbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, + const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_sspmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *a, - const float *x, std::int64_t incx, float beta, - float *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_dspmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *a, - const double *x, std::int64_t incx, double beta, - double *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_sspr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, - std::int64_t incx, float *a, - const std::vector &dependencies); - sycl::event (*column_major_dspr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, - std::int64_t incx, double *a, - const std::vector &dependencies); - sycl::event (*column_major_sspr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - float *a, - const std::vector &dependencies); - sycl::event (*column_major_dspr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double *y, - std::int64_t incy, double *a, - const std::vector &dependencies); - sycl::event (*column_major_ssymv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, - float beta, float *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_dsymv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*column_major_ssyr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, - std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_dsyr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, - std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_ssyr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - float *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_dsyr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double *y, - std::int64_t incy, double *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*column_major_stbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_sspmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* a, + const float* x, std::int64_t incx, float beta, + float* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_dspmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* a, + const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_sspr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* x, + std::int64_t incx, float* a, + const std::vector& dependencies); + sycl::event (*column_major_dspr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* a, + const std::vector& dependencies); + sycl::event (*column_major_sspr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, + float* a, + const std::vector& dependencies); + sycl::event (*column_major_dspr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* x, + std::int64_t incx, const double* y, + std::int64_t incy, double* a, + const std::vector& dependencies); + sycl::event (*column_major_ssymv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_dsymv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, + double beta, double* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*column_major_ssyr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* x, + std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_dsyr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_ssyr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_dsyr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* x, + std::int64_t incx, const double* y, + std::int64_t incy, double* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_stbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_dtbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_dtbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const double *a, std::int64_t lda, - double *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ctbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_ctbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, + std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ztbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_ztbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, + std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_stbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_stbsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_dtbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_dtbsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const double *a, std::int64_t lda, - double *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ctbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_ctbsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, + std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ztbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_ztbsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, + std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_stpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_stpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const float *a, float *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_dtpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const float* a, float* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_dtpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const double *a, double *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ctpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const double* a, double* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_ctpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::complex *x, + const std::complex* a, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ztpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_ztpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_stpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_stpsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const float *a, float *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_dtpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const float* a, float* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_dtpsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const double *a, double *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ctpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const double* a, double* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_ctpsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::complex *x, + const std::complex* a, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ztpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_ztpsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_strmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_strmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const float *a, std::int64_t lda, float *x, + const float* a, std::int64_t lda, float* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_dtrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_dtrmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const double *a, std::int64_t lda, double *x, + const double* a, std::int64_t lda, double* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ctrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_ctrmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ztrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_ztrmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_strsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_strsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const float *a, std::int64_t lda, float *x, + const float* a, std::int64_t lda, float* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_dtrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_dtrsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const double *a, std::int64_t lda, double *x, + const double* a, std::int64_t lda, double* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ctrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_ctrsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_ztrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_ztrsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*column_major_sgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*column_major_sgemm_usm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, + const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_dgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + const std::vector& dependencies); + sycl::event (*column_major_dgemm_usm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, + const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_cgemm_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*column_major_zgemm_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_hgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*column_major_hgemm_usm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, - const sycl::half *b, std::int64_t ldb, - sycl::half beta, sycl::half *c, std::int64_t ldc, - const std::vector &dependencies); + const sycl::half* a, std::int64_t lda, + const sycl::half* b, std::int64_t ldb, + sycl::half beta, sycl::half* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*column_major_gemm_f16f16f32_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies); sycl::event (*column_major_gemm_bf16bf16f32_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const oneapi::math::bfloat16 *a, - std::int64_t lda, const oneapi::math::bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const oneapi::math::bfloat16* a, std::int64_t lda, const oneapi::math::bfloat16* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*column_major_chemm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_zhemm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, const std::vector& dependencies); + sycl::event (*column_major_zhemm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_cherk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_cherk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, - const std::complex *a, std::int64_t lda, - float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_zherk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, std::int64_t lda, + float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*column_major_zherk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, - const std::complex *a, std::int64_t lda, - double beta, std::complex *c, + const std::complex* a, std::int64_t lda, + double beta, std::complex* c, std::int64_t ldc, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_cher2k_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, float beta, - std::complex *c, std::int64_t ldc, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, float beta, + std::complex* c, std::int64_t ldc, const std::vector& dependencies); sycl::event (*column_major_zher2k_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, double beta, - std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_ssymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, double beta, + std::complex* c, std::int64_t ldc, const std::vector& dependencies); + sycl::event (*column_major_ssymm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_dsymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*column_major_dsymm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies); + std::int64_t n, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*column_major_csymm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_zsymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, const std::vector& dependencies); + sycl::event (*column_major_zsymm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_ssyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_ssyrk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, - std::int64_t lda, float beta, float *c, + std::int64_t k, float alpha, const float* a, + std::int64_t lda, float beta, float* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_dsyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_dsyrk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, - std::int64_t lda, double beta, double *c, + std::int64_t k, double alpha, const double* a, + std::int64_t lda, double beta, double* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_csyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_csyrk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex beta, std::complex *c, + const std::complex* a, std::int64_t lda, + std::complex beta, std::complex* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_zsyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_zsyrk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex beta, std::complex *c, + const std::complex* a, std::int64_t lda, + std::complex beta, std::complex* c, std::int64_t ldc, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_ssyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, - float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* k, float* alpha, const float** a, std::int64_t* lda, + float* beta, float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_dsyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, - double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* k, double* alpha, const double** a, std::int64_t* lda, + double* beta, double** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_csyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* k, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_zsyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, std::complex *beta, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, std::complex* beta, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_ssyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_dsyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double beta, double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_csyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_zsyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); - sycl::event (*column_major_ssyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*column_major_ssyr2k_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_dsyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*column_major_dsyr2k_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies); + std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*column_major_csyr2k_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_zsyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, const std::vector& dependencies); + sycl::event (*column_major_zsyr2k_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_strmm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + const std::vector& dependencies); + sycl::event (*column_major_strmm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*column_major_dtrmm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*column_major_dtrmm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies); + std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies); sycl::event (*column_major_ctrmm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, const std::vector &dependencies); + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, const std::vector& dependencies); sycl::event (*column_major_ztrmm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*column_major_strsm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, const std::vector& dependencies); + sycl::event (*column_major_strsm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*column_major_dtrsm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*column_major_dtrsm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies); + std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies); sycl::event (*column_major_ctrsm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, const std::vector &dependencies); + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, const std::vector& dependencies); sycl::event (*column_major_ztrsm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, const std::vector &dependencies); + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, const std::vector& dependencies); sycl::event (*column_major_strsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_dtrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_ctrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies); + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_ztrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies); + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_strsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_dtrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, double** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_ctrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_ztrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*column_major_sgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, - std::int64_t *lda, const float **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, float* alpha, const float** a, + std::int64_t* lda, const float** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_dgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, - std::int64_t *lda, const double **b, std::int64_t *ldb, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, double* alpha, const double** a, + std::int64_t* lda, const double** b, std::int64_t* ldb, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_cgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, const std::complex **b, - std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, const std::complex** b, + std::int64_t* ldb, std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_zgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, const std::complex **b, - std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, const std::complex** b, + std::int64_t* ldb, std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_hgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, const sycl::half **a, - std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, sycl::half *beta, - sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, sycl::half* alpha, const sycl::half** a, + std::int64_t* lda, const sycl::half** b, std::int64_t* ldb, sycl::half* beta, + sycl::half** c, std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_gemm_f16f16f32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, - std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, float* alpha, const sycl::half** a, + std::int64_t* lda, const sycl::half** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_gemm_s8s8f32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, - std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, float* alpha, const std::int8_t** a, + std::int64_t* lda, const std::int8_t** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_gemm_s8s8s32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, - std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, float* alpha, const std::int8_t** a, + std::int64_t* lda, const std::int8_t** b, std::int64_t* ldb, float* beta, std::int32_t** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*column_major_sgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, const float *b, std::int64_t ldb, - std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, const float* b, std::int64_t ldb, + std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_dgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, const double *b, std::int64_t ldb, - std::int64_t stride_b, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, const double* b, std::int64_t ldb, + std::int64_t stride_b, double beta, double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_cgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_zgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_hgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, - std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half* a, + std::int64_t lda, std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, sycl::half* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_gemm_f16f16f32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_gemm_s8s8f32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, - std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t* a, + std::int64_t lda, std::int64_t stride_a, const std::int8_t* b, std::int64_t ldb, + std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_gemm_s8s8s32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, - std::int64_t stride_b, float beta, std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); - sycl::event (*column_major_sgemmt_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t* a, + std::int64_t lda, std::int64_t stride_a, const std::int8_t* b, std::int64_t ldb, + std::int64_t stride_b, float beta, std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); + sycl::event (*column_major_sgemmt_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, - std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_dgemmt_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*column_major_dgemmt_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, - std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies); + std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*column_major_cgemmt_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*column_major_zgemmt_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*column_major_gemm_s8u8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::uint8_t *b, - std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies); + const std::int8_t* a, std::int64_t lda, std::int8_t ao, const std::uint8_t* b, + std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies); sycl::event (*column_major_gemm_s8s8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::int8_t *b, - std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies); + const std::int8_t* a, std::int64_t lda, std::int8_t ao, const std::int8_t* b, + std::int64_t ldb, std::int8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies); sycl::event (*column_major_gemm_u8s8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::int8_t *b, - std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies); + const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, const std::int8_t* b, + std::int64_t ldb, std::int8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies); sycl::event (*column_major_gemm_u8u8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::uint8_t *b, - std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies); + const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, const std::uint8_t* b, + std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies); sycl::event (*column_major_somatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_domatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*column_major_comatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_zomatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_simatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_dimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_cimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, - std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, + std::int64_t stride, std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_zimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, - std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, + std::int64_t stride, std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_somatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float beta, const float *b, std::int64_t ldb, std::int64_t stride_b, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float beta, const float* b, std::int64_t ldb, std::int64_t stride_b, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies); sycl::event (*column_major_domatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double beta, const double *b, std::int64_t ldb, - std::int64_t stride_b, double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double beta, const double* b, std::int64_t ldb, + std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*column_major_comatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies); sycl::event (*column_major_zomatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies); - sycl::event (*column_major_somatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, - std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*column_major_domatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, - std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*column_major_comatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*column_major_zomatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies); + sycl::event (*column_major_somatcopy_usm_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*column_major_domatcopy_usm_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*column_major_comatcopy_usm_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*column_major_zomatcopy_usm_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies); sycl::event (*column_major_somatcopy2_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, std::int64_t stridea, float *b, - std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, std::int64_t stridea, float* b, + std::int64_t ldb, std::int64_t strideb, const std::vector& dependencies); sycl::event (*column_major_domatcopy2_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, std::int64_t stridea, double *b, - std::int64_t ldb, std::int64_t strideb, const std::vector &dependencies); - sycl::event (*column_major_comatcopy2_usm_sycl)(sycl::queue &queue, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, std::int64_t stridea, double* b, + std::int64_t ldb, std::int64_t strideb, const std::vector& dependencies); + sycl::event (*column_major_comatcopy2_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies); - sycl::event (*column_major_zomatcopy2_usm_sycl)(sycl::queue &queue, + const std::vector& dependencies); + sycl::event (*column_major_zomatcopy2_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies); - sycl::event (*column_major_simatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, float alpha, - float *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*column_major_dimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, double alpha, - double *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*column_major_cimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::complex alpha, - std::complex *ab, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_simatcopy_usm_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, float* ab, + std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*column_major_dimatcopy_usm_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, double* ab, + std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*column_major_cimatcopy_usm_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*column_major_zimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::complex alpha, - std::complex *ab, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*column_major_zimatcopy_usm_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*column_major_somatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + const std::vector& dependencies); + sycl::event (*column_major_somatadd_usm_sycl)(sycl::queue& queue, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, float beta, const float *b, - std::int64_t ldb, float *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*column_major_domatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, float beta, const float* b, + std::int64_t ldb, float* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*column_major_domatadd_usm_sycl)(sycl::queue& queue, + oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, double beta, const double *b, - std::int64_t ldb, double *c, std::int64_t ldc, - const std::vector &dependencies); + std::int64_t n, double alpha, const double* a, + std::int64_t lda, double beta, const double* b, + std::int64_t ldb, double* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*column_major_comatadd_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, - std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*column_major_zomatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, - oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex beta, - const std::complex *b, std::int64_t ldb, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, const std::complex* b, std::int64_t ldb, + std::complex* c, std::int64_t ldc, const std::vector& dependencies); + sycl::event (*column_major_zomatadd_usm_sycl)( + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, const std::complex* b, + std::int64_t ldb, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*column_major_somatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + float* alpha, const float** a, std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies); sycl::event (*column_major_domatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + double* alpha, const double** a, std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies); sycl::event (*column_major_comatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies); sycl::event (*column_major_zomatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies); sycl::event (*column_major_simatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - float *alpha, float **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + float* alpha, float** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies); sycl::event (*column_major_dimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - double *alpha, double **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + double* alpha, double** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies); sycl::event (*column_major_cimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, std::complex **ab, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, std::complex** ab, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies); sycl::event (*column_major_zimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, std::complex **ab, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, std::complex** ab, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies); // Buffer APIs - void (*row_major_scasum_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*row_major_dzasum_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*row_major_sasum_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*row_major_dasum_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*row_major_saxpy_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); - void (*row_major_daxpy_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy); - void (*row_major_caxpy_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_zaxpy_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_saxpy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, + void (*row_major_scasum_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*row_major_dzasum_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*row_major_sasum_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*row_major_dasum_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*row_major_saxpy_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); + void (*row_major_daxpy_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy); + void (*row_major_caxpy_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_zaxpy_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_saxpy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_daxpy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, + void (*row_major_daxpy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_caxpy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, + void (*row_major_caxpy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_zaxpy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, + void (*row_major_zaxpy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_saxpby_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); - void (*row_major_daxpby_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - void (*row_major_caxpby_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + void (*row_major_saxpby_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); + void (*row_major_daxpby_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + void (*row_major_caxpby_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); - void (*row_major_zaxpby_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, + void (*row_major_zaxpby_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_scopy_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); - void (*row_major_dcopy_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); - void (*row_major_ccopy_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_zcopy_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_scopy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_scopy_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); + void (*row_major_dcopy_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); + void (*row_major_ccopy_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_zcopy_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_scopy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_dcopy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer &x, std::int64_t incx, - std::int64_t stridex, sycl::buffer &y, + void (*row_major_dcopy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer& x, std::int64_t incx, + std::int64_t stridex, sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_ccopy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, + void (*row_major_ccopy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_zcopy_batch_strided_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, + void (*row_major_zcopy_batch_strided_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &y, + sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_sdot_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); - void (*row_major_ddot_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); - void (*row_major_dsdot_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); - void (*row_major_cdotc_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - void (*row_major_zdotc_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - void (*row_major_cdotu_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - void (*row_major_zdotu_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &result); - void (*row_major_isamin_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*row_major_idamin_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*row_major_icamin_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*row_major_izamin_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*row_major_isamax_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*row_major_idamax_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*row_major_icamax_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*row_major_izamax_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*row_major_scnrm2_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*row_major_dznrm2_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer &result); - void (*row_major_snrm2_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*row_major_dnrm2_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &result); - void (*row_major_srot_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, float c, + void (*row_major_sdot_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + void (*row_major_ddot_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + void (*row_major_dsdot_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + void (*row_major_cdotc_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + void (*row_major_zdotc_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + void (*row_major_cdotu_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + void (*row_major_zdotu_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& result); + void (*row_major_isamin_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*row_major_idamin_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*row_major_icamin_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*row_major_izamin_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*row_major_isamax_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*row_major_idamax_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*row_major_icamax_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*row_major_izamax_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*row_major_scnrm2_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*row_major_dznrm2_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer& result); + void (*row_major_snrm2_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*row_major_dnrm2_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& result); + void (*row_major_srot_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, float c, float s); - void (*row_major_drot_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, + void (*row_major_drot_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, double c, double s); - void (*row_major_csrot_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, + void (*row_major_csrot_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, float c, float s); - void (*row_major_zdrot_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, + void (*row_major_zdrot_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, double c, double s); - void (*row_major_srotg_sycl)(sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s); - void (*row_major_drotg_sycl)(sycl::queue &queue, sycl::buffer &a, - sycl::buffer &b, sycl::buffer &c, - sycl::buffer &s); - void (*row_major_crotg_sycl)(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, sycl::buffer &c, - sycl::buffer, 1> &s); - void (*row_major_zrotg_sycl)(sycl::queue &queue, sycl::buffer, 1> &a, - sycl::buffer, 1> &b, - sycl::buffer &c, - sycl::buffer, 1> &s); - void (*row_major_srotm_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m); - void (*row_major_drotm_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy, - sycl::buffer ¶m); - void (*row_major_srotmg_sycl)(sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, float y1, - sycl::buffer ¶m); - void (*row_major_drotmg_sycl)(sycl::queue &queue, sycl::buffer &d1, - sycl::buffer &d2, sycl::buffer &x1, - double y1, sycl::buffer ¶m); - void (*row_major_sscal_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx); - void (*row_major_dscal_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx); - void (*row_major_cscal_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_csscal_sycl)(sycl::queue &queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_zscal_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_zdscal_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_sdsdot_sycl)(sycl::queue &queue, std::int64_t n, float sb, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &result); - void (*row_major_sswap_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); - void (*row_major_dswap_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &x, - std::int64_t incx, sycl::buffer &y, std::int64_t incy); - void (*row_major_cswap_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_zswap_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_sgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, + void (*row_major_srotg_sycl)(sycl::queue& queue, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, + sycl::buffer& s); + void (*row_major_drotg_sycl)(sycl::queue& queue, sycl::buffer& a, + sycl::buffer& b, sycl::buffer& c, + sycl::buffer& s); + void (*row_major_crotg_sycl)(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, sycl::buffer& c, + sycl::buffer, 1>& s); + void (*row_major_zrotg_sycl)(sycl::queue& queue, sycl::buffer, 1>& a, + sycl::buffer, 1>& b, + sycl::buffer& c, + sycl::buffer, 1>& s); + void (*row_major_srotm_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param); + void (*row_major_drotm_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy, + sycl::buffer& param); + void (*row_major_srotmg_sycl)(sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, float y1, + sycl::buffer& param); + void (*row_major_drotmg_sycl)(sycl::queue& queue, sycl::buffer& d1, + sycl::buffer& d2, sycl::buffer& x1, + double y1, sycl::buffer& param); + void (*row_major_sscal_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx); + void (*row_major_dscal_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx); + void (*row_major_cscal_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_csscal_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_zscal_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_zdscal_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_sdsdot_sycl)(sycl::queue& queue, std::int64_t n, float sb, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, + sycl::buffer& result); + void (*row_major_sswap_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); + void (*row_major_dswap_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& x, + std::int64_t incx, sycl::buffer& y, std::int64_t incy); + void (*row_major_cswap_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_zswap_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_sgbmv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); - void (*row_major_dgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); + void (*row_major_dgbmv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - void (*row_major_cgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + void (*row_major_cgbmv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_zgbmv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_zgbmv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_sgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_dgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_cgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_sgemv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy); + void (*row_major_dgemv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy); + void (*row_major_cgemv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); - void (*row_major_zgemv_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, + void (*row_major_zgemv_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_sgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_sgemv_batch_strided_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, std::int64_t stridex, float beta, - sycl::buffer &y, std::int64_t incy, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_dgemv_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*row_major_dgemv_batch_strided_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, std::int64_t stridex, double beta, - sycl::buffer &y, std::int64_t incy, + sycl::buffer& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); void (*row_major_cgemv_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, std::complex beta, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); void (*row_major_zgemv_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &x, std::int64_t incx, - std::int64_t stridex, std::complex beta, sycl::buffer, 1> &y, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& x, std::int64_t incx, + std::int64_t stridex, std::complex beta, sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - void (*row_major_sdgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::side left_right, + void (*row_major_sdgmm_batch_strided_sycl)(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); - void (*row_major_ddgmm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::side left_right, + void (*row_major_ddgmm_batch_strided_sycl)(sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &x, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); void (*row_major_cdgmm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); void (*row_major_zdgmm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1> &c, std::int64_t ldc, std::int64_t stridec, + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, + sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, + sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size); - void (*row_major_sger_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - void (*row_major_dger_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - void (*row_major_cgerc_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, + void (*row_major_sger_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda); + void (*row_major_dger_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, + sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda); + void (*row_major_cgerc_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda); - void (*row_major_zgerc_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + void (*row_major_zgerc_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - void (*row_major_cgeru_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + void (*row_major_cgeru_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda); - void (*row_major_zgeru_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + void (*row_major_zgeru_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - void (*row_major_chbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + void (*row_major_chbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); - void (*row_major_zhbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, + void (*row_major_zhbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_chemv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_chemv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_zhemv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_zhemv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_cher_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_cher_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a, std::int64_t lda); - void (*row_major_zher_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a, + void (*row_major_zher_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a, std::int64_t lda); - void (*row_major_cher2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a, + void (*row_major_cher2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a, std::int64_t lda); - void (*row_major_zher2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, + void (*row_major_zher2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a, std::int64_t lda); - void (*row_major_chpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, - std::complex beta, sycl::buffer, 1> &y, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a, std::int64_t lda); + void (*row_major_chpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, + std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); - void (*row_major_zhpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, + void (*row_major_zhpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx, + sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1> &y, std::int64_t incy); - void (*row_major_chpr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a); - void (*row_major_zhpr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &a); - void (*row_major_chpr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &x, - std::int64_t incx, sycl::buffer, 1> &y, - std::int64_t incy, sycl::buffer, 1> &a); - void (*row_major_zhpr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, + sycl::buffer, 1>& y, std::int64_t incy); + void (*row_major_chpr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a); + void (*row_major_zhpr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& a); + void (*row_major_chpr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& x, + std::int64_t incx, sycl::buffer, 1>& y, + std::int64_t incy, sycl::buffer, 1>& a); + void (*row_major_zhpr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &x, std::int64_t incx, - sycl::buffer, 1> &y, std::int64_t incy, - sycl::buffer, 1> &a); - void (*row_major_ssbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - float beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_dsbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx, - double beta, sycl::buffer &y, std::int64_t incy); - void (*row_major_sspmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &a, sycl::buffer &x, - std::int64_t incx, float beta, sycl::buffer &y, + sycl::buffer, 1>& x, std::int64_t incx, + sycl::buffer, 1>& y, std::int64_t incy, + sycl::buffer, 1>& a); + void (*row_major_ssbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + float beta, sycl::buffer& y, std::int64_t incy); + void (*row_major_dsbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx, + double beta, sycl::buffer& y, std::int64_t incy); + void (*row_major_sspmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& a, sycl::buffer& x, + std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy); - void (*row_major_dspmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - void (*row_major_sspr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &a); - void (*row_major_dspr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &a); - void (*row_major_sspr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a); - void (*row_major_dspr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a); - void (*row_major_ssymv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, float beta, - sycl::buffer &y, std::int64_t incy); - void (*row_major_dsymv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx, double beta, - sycl::buffer &y, std::int64_t incy); - void (*row_major_ssyr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &a, std::int64_t lda); - void (*row_major_dsyr_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &a, std::int64_t lda); - void (*row_major_ssyr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - float alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - void (*row_major_dsyr2_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, - double alpha, sycl::buffer &x, std::int64_t incx, - sycl::buffer &y, std::int64_t incy, - sycl::buffer &a, std::int64_t lda); - void (*row_major_stbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*row_major_dspmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + void (*row_major_sspr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a); + void (*row_major_dspr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a); + void (*row_major_sspr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, + sycl::buffer& a); + void (*row_major_dspr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, + sycl::buffer& a); + void (*row_major_ssymv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, float beta, + sycl::buffer& y, std::int64_t incy); + void (*row_major_dsymv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx, double beta, + sycl::buffer& y, std::int64_t incy); + void (*row_major_ssyr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a, std::int64_t lda); + void (*row_major_dsyr_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& a, std::int64_t lda); + void (*row_major_ssyr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + float alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda); + void (*row_major_dsyr2_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, + double alpha, sycl::buffer& x, std::int64_t incx, + sycl::buffer& y, std::int64_t incy, + sycl::buffer& a, std::int64_t lda); + void (*row_major_stbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_dtbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); + void (*row_major_dtbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_ctbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); + void (*row_major_ctbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_ztbmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_ztbmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_stbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_stbsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_dtbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); + void (*row_major_dtbsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &x, std::int64_t incx); - void (*row_major_ctbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& x, std::int64_t incx); + void (*row_major_ctbsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_ztbsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_ztbsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, std::int64_t k, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_stpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_stpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx); - void (*row_major_dtpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); + void (*row_major_dtpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx); - void (*row_major_ctpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); + void (*row_major_ctpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_ztpmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_ztpmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_stpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_stpsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx); - void (*row_major_dtpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); + void (*row_major_dtpsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, - sycl::buffer &x, std::int64_t incx); - void (*row_major_ctpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, + sycl::buffer& x, std::int64_t incx); + void (*row_major_ctpsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_ztpsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_ztpsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - sycl::buffer, 1> &x, std::int64_t incx); - void (*row_major_strmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer, 1>& a, + sycl::buffer, 1>& x, std::int64_t incx); + void (*row_major_strmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - void (*row_major_dtrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + void (*row_major_dtrmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - void (*row_major_ctrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + void (*row_major_ctrmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx); - void (*row_major_ztrmv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*row_major_ztrmv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx); - void (*row_major_strsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*row_major_strsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - void (*row_major_dtrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + void (*row_major_dtrsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &x, std::int64_t incx); - void (*row_major_ctrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& x, std::int64_t incx); + void (*row_major_ctrsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx); - void (*row_major_ztrsv_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*row_major_ztrsv_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, - std::int64_t n, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &x, + std::int64_t n, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx); - void (*row_major_sgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + void (*row_major_sgemm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - float beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_dgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + float beta, sycl::buffer& c, std::int64_t ldc); + void (*row_major_dgemm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - double beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_cgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + std::int64_t k, double alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + double beta, sycl::buffer& c, std::int64_t ldc); + void (*row_major_cgemm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); - void (*row_major_zgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + void (*row_major_zgemm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_hgemm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*row_major_hgemm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::half alpha, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::half beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_gemm_f16f16f32_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + std::int64_t k, sycl::half alpha, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::half beta, sycl::buffer& c, std::int64_t ldc); + void (*row_major_gemm_f16f16f32_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, - float beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_gemm_bf16bf16f32_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, + float beta, sycl::buffer& c, std::int64_t ldc); + void (*row_major_gemm_bf16bf16f32_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, + sycl::buffer& a, std::int64_t lda, - sycl::buffer &b, - std::int64_t ldb, float beta, sycl::buffer &c, + sycl::buffer& b, + std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); - void (*row_major_chemm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + void (*row_major_chemm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zhemm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*row_major_zhemm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_cherk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*row_major_cherk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer, 1> &a, + float alpha, sycl::buffer, 1>& a, std::int64_t lda, float beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zherk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*row_major_zherk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer, 1> &a, + double alpha, sycl::buffer, 1>& a, std::int64_t lda, double beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_cher2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*row_major_cher2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - float beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + float beta, sycl::buffer, 1>& c, std::int64_t ldc); - void (*row_major_zher2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*row_major_zher2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - double beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + double beta, sycl::buffer, 1>& c, std::int64_t ldc); - void (*row_major_ssymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + void (*row_major_ssymm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - void (*row_major_dsymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); + void (*row_major_dsymm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - void (*row_major_csymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); + void (*row_major_csymm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zsymm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*row_major_zsymm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_ssyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*row_major_ssyrk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - float beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_dsyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& c, std::int64_t ldc); + void (*row_major_dsyrk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - double beta, sycl::buffer &c, std::int64_t ldc); - void (*row_major_csyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + double alpha, sycl::buffer& a, std::int64_t lda, + double beta, sycl::buffer& c, std::int64_t ldc); + void (*row_major_csyrk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1> &a, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zsyrk_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*row_major_zsyrk_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_ssyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*row_major_ssyrk_batch_strided_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, float beta, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*row_major_dsyrk_batch_strided_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*row_major_dsyrk_batch_strided_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, double beta, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_csyrk_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_zsyrk_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*row_major_ssyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*row_major_ssyr2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - void (*row_major_dsyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); + void (*row_major_dsyr2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - void (*row_major_csyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); + void (*row_major_csyr2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); - void (*row_major_zsyr2k_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*row_major_zsyr2k_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_strmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*row_major_strmm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*row_major_dtrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*row_major_dtrmm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*row_major_ctrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*row_major_ctrmm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb); - void (*row_major_ztrmm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + void (*row_major_ztrmm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - void (*row_major_strsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + void (*row_major_strsm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*row_major_dtrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + float alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*row_major_dtrsm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*row_major_ctrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + double alpha, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*row_major_ctrsm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, - std::int64_t lda, sycl::buffer, 1> &b, + std::complex alpha, sycl::buffer, 1>& a, + std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb); - void (*row_major_ztrsm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + void (*row_major_ztrsm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - void (*row_major_sgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + void (*row_major_sgemm_batch_strided_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*row_major_dgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + void (*row_major_dgemm_batch_strided_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, double beta, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_cgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_zgemm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1> &b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*row_major_hgemm_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + void (*row_major_hgemm_batch_strided_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - sycl::buffer &a, std::int64_t lda, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, sycl::half beta, - sycl::buffer &c, std::int64_t ldc, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_gemm_f16f16f32_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, float beta, sycl::buffer &c, std::int64_t ldc, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, float beta, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_gemm_s8s8f32_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_gemm_s8s8s32_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, float beta, + sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_strsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + float alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*row_major_dtrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, + double alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*row_major_ctrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*row_major_ztrsm_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - void (*row_major_sgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*row_major_sgemmt_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, float beta, - sycl::buffer &c, std::int64_t ldc); - void (*row_major_dgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, float beta, + sycl::buffer& c, std::int64_t ldc); + void (*row_major_dgemmt_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, double beta, - sycl::buffer &c, std::int64_t ldc); - void (*row_major_cgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, double beta, + sycl::buffer& c, std::int64_t ldc); + void (*row_major_cgemmt_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1> &c, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, + std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); - void (*row_major_zgemmt_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + void (*row_major_zgemmt_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& c, std::int64_t ldc); void (*row_major_gemm_s8u8s32_bias_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); void (*row_major_gemm_s8s8s32_bias_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, int8_t ao, sycl::buffer &b, - std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, int8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); void (*row_major_gemm_u8s8s32_bias_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, - std::int64_t ldb, int8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); + sycl::buffer& a, std::int64_t lda, uint8_t ao, sycl::buffer& b, + std::int64_t ldb, int8_t bo, float beta, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); void (*row_major_gemm_u8u8s32_bias_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer &a, std::int64_t lda, uint8_t ao, sycl::buffer &b, - std::int64_t ldb, uint8_t bo, float beta, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &co); - void (*row_major_somatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, uint8_t ao, sycl::buffer& b, + std::int64_t ldb, uint8_t bo, float beta, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& co); + void (*row_major_somatcopy_batch_strided_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, + sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - void (*row_major_domatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, + void (*row_major_domatcopy_batch_strided_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*row_major_comatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); void (*row_major_zomatcopy_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer, 1> &b, std::int64_t ldb, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - void (*row_major_simatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &ab, std::int64_t lda, + void (*row_major_simatcopy_batch_strided_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, float alpha, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - void (*row_major_dimatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &ab, std::int64_t lda, + void (*row_major_dimatcopy_batch_strided_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, double alpha, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - void (*row_major_cimatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &ab, + void (*row_major_cimatcopy_batch_strided_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); - void (*row_major_zimatcopy_batch_strided_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::complex alpha, - sycl::buffer, 1> &ab, + void (*row_major_zimatcopy_batch_strided_sycl)(sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::complex alpha, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, std::int64_t batch_size); void (*row_major_somatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, float beta, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, float beta, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_domatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer &c, std::int64_t ldc, std::int64_t stride_c, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, double beta, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_comatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); void (*row_major_zomatadd_batch_strided_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1> &b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1> &c, std::int64_t ldc, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, + std::complex beta, sycl::buffer, 1>& b, std::int64_t ldb, + std::int64_t stride_b, sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - void (*row_major_somatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*row_major_somatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*row_major_domatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*row_major_domatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb); - void (*row_major_comatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb); + void (*row_major_comatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - void (*row_major_zomatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + void (*row_major_zomatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - sycl::buffer, 1> &b, std::int64_t ldb); - void (*row_major_somatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::buffer, 1>& a, std::int64_t lda, + sycl::buffer, 1>& b, std::int64_t ldb); + void (*row_major_somatcopy2_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); - void (*row_major_domatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*row_major_domatcopy2_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer &b, + sycl::buffer& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); - void (*row_major_comatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*row_major_comatcopy2_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1> &b, + sycl::buffer, 1>& a, std::int64_t lda, + std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb); - void (*row_major_zomatcopy2_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*row_major_zomatcopy2_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1> &b, std::int64_t ldb, + sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t strideb); - void (*row_major_simatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*row_major_simatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer &ab, std::int64_t lda, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); - void (*row_major_dimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*row_major_dimatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer &ab, std::int64_t lda, + sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); - void (*row_major_cimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*row_major_cimatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb); - void (*row_major_zimatcopy_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + void (*row_major_zimatcopy_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &ab, std::int64_t lda, + sycl::buffer, 1>& ab, std::int64_t lda, std::int64_t ldb); - void (*row_major_somatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + void (*row_major_somatadd_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer &a, std::int64_t lda, - float beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc); - void (*row_major_domatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + float alpha, sycl::buffer& a, std::int64_t lda, + float beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc); + void (*row_major_domatadd_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer &a, std::int64_t lda, - double beta, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &c, std::int64_t ldc); - void (*row_major_comatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + double alpha, sycl::buffer& a, std::int64_t lda, + double beta, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& c, std::int64_t ldc); + void (*row_major_comatadd_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); - void (*row_major_zomatadd_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); + void (*row_major_zomatadd_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1> &a, std::int64_t lda, + sycl::buffer, 1>& a, std::int64_t lda, std::complex beta, - sycl::buffer, 1> &b, std::int64_t ldb, - sycl::buffer, 1> &c, std::int64_t ldc); + sycl::buffer, 1>& b, std::int64_t ldb, + sycl::buffer, 1>& c, std::int64_t ldc); // USM APIs - sycl::event (*row_major_scasum_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - float *result, - const std::vector &dependencies); - sycl::event (*row_major_dzasum_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - double *result, - const std::vector &dependencies); - sycl::event (*row_major_sasum_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies); - sycl::event (*row_major_dasum_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies); - sycl::event (*row_major_saxpy_usm_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, float *y, + sycl::event (*row_major_scasum_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + float* result, + const std::vector& dependencies); + sycl::event (*row_major_dzasum_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + double* result, + const std::vector& dependencies); + sycl::event (*row_major_sasum_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, + const std::vector& dependencies); + sycl::event (*row_major_dasum_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, + const std::vector& dependencies); + sycl::event (*row_major_saxpy_usm_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, float* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_daxpy_usm_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, double *y, + const std::vector& dependencies); + sycl::event (*row_major_daxpy_usm_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + const double* x, std::int64_t incx, double* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_caxpy_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, std::complex *y, + const std::vector& dependencies); + sycl::event (*row_major_caxpy_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_zaxpy_usm_sycl)(sycl::queue &queue, std::int64_t n, + const std::vector& dependencies); + sycl::event (*row_major_zaxpy_usm_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); sycl::event (*row_major_saxpy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, float *alpha, const float **x, std::int64_t *incx, - float **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, float* alpha, const float** x, std::int64_t* incx, + float** y, std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_daxpy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, double *alpha, const double **x, std::int64_t *incx, - double **y, std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, double* alpha, const double** x, std::int64_t* incx, + double** y, std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_caxpy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, std::complex* alpha, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_zaxpy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, std::complex *alpha, - const std::complex **x, std::int64_t *incx, std::complex **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, std::complex* alpha, + const std::complex** x, std::int64_t* incx, std::complex** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_saxpy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, float alpha, const float *x, std::int64_t incx, - std::int64_t stridex, float *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t n, float alpha, const float* x, std::int64_t incx, + std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_daxpy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, double alpha, const double *x, std::int64_t incx, - std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t n, double alpha, const double* x, std::int64_t incx, + std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_caxpy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, std::complex alpha, const std::complex *x, - std::int64_t incx, std::int64_t stridex, std::complex *y, std::int64_t incy, + sycl::queue& queue, std::int64_t n, std::complex alpha, const std::complex* x, + std::int64_t incx, std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*row_major_zaxpy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); - sycl::event (*row_major_saxpby_usm_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - const float *x, std::int64_t incx, const float beta, - float *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_daxpby_usm_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - const double *x, std::int64_t incx, const double beta, - double *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_caxpby_usm_sycl)(sycl::queue &queue, std::int64_t n, + sycl::queue& queue, std::int64_t n, std::complex alpha, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies); + sycl::event (*row_major_saxpby_usm_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + const float* x, std::int64_t incx, const float beta, + float* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_daxpby_usm_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + const double* x, std::int64_t incx, const double beta, + double* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_caxpby_usm_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex beta, std::complex *y, + const std::complex* x, std::int64_t incx, + const std::complex beta, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_zaxpby_usm_sycl)(sycl::queue &queue, std::int64_t n, + const std::vector& dependencies); + sycl::event (*row_major_zaxpby_usm_sycl)(sycl::queue& queue, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, + const std::complex* x, std::int64_t incx, const std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_scopy_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_dcopy_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_ccopy_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_zcopy_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); + std::complex* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_scopy_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_dcopy_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_ccopy_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_zcopy_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); sycl::event (*row_major_scopy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, const float **x, std::int64_t *incx, float **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, const float** x, std::int64_t* incx, float** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_dcopy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, const double **x, std::int64_t *incx, double **y, - std::int64_t *incy, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, const double** x, std::int64_t* incx, double** y, + std::int64_t* incy, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_ccopy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_zcopy_batch_group_usm_sycl)( - sycl::queue &queue, std::int64_t *n, const std::complex **x, std::int64_t *incx, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t* n, const std::complex** x, std::int64_t* incx, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_scopy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, const float *x, std::int64_t incx, std::int64_t stridex, - float *y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); + sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, std::int64_t stridex, + float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, + const std::vector& dependencies); sycl::event (*row_major_dcopy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, const double *x, std::int64_t incx, - std::int64_t stridex, double *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, + std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_ccopy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, std::int64_t n, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_zcopy_batch_strided_usm_sycl)( - sycl::queue &queue, std::int64_t n, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); - sycl::event (*row_major_sdot_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - float *result, - const std::vector &dependencies); - sycl::event (*row_major_ddot_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, - double *result, - const std::vector &dependencies); - sycl::event (*row_major_dsdot_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - double *result, - const std::vector &dependencies); - sycl::event (*row_major_cdotc_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies); - sycl::event (*row_major_zdotc_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies); - sycl::event (*row_major_cdotu_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies); - sycl::event (*row_major_zdotu_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *result, - const std::vector &dependencies); - sycl::event (*row_major_isamin_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies); - sycl::event (*row_major_idamin_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies); - sycl::event (*row_major_icamin_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies); - sycl::event (*row_major_izamin_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies); - sycl::event (*row_major_isamax_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies); - sycl::event (*row_major_idamax_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, std::int64_t *result, - const std::vector &dependencies); - sycl::event (*row_major_icamax_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies); - sycl::event (*row_major_izamax_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - std::int64_t *result, - const std::vector &dependencies); - sycl::event (*row_major_scnrm2_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - float *result, - const std::vector &dependencies); - sycl::event (*row_major_dznrm2_usm_sycl)(sycl::queue &queue, std::int64_t n, - const std::complex *x, std::int64_t incx, - double *result, - const std::vector &dependencies); - sycl::event (*row_major_snrm2_usm_sycl)(sycl::queue &queue, std::int64_t n, const float *x, - std::int64_t incx, float *result, - const std::vector &dependencies); - sycl::event (*row_major_dnrm2_usm_sycl)(sycl::queue &queue, std::int64_t n, const double *x, - std::int64_t incx, double *result, - const std::vector &dependencies); - sycl::event (*row_major_srot_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, float c, - float s, const std::vector &dependencies); - sycl::event (*row_major_drot_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, double c, - double s, const std::vector &dependencies); - sycl::event (*row_major_csrot_usm_sycl)(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, float c, - float s, const std::vector &dependencies); - sycl::event (*row_major_zdrot_usm_sycl)(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, + sycl::queue& queue, std::int64_t n, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); + sycl::event (*row_major_sdot_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, + float* result, + const std::vector& dependencies); + sycl::event (*row_major_ddot_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, + double* result, + const std::vector& dependencies); + sycl::event (*row_major_dsdot_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, + double* result, + const std::vector& dependencies); + sycl::event (*row_major_cdotc_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies); + sycl::event (*row_major_zdotc_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies); + sycl::event (*row_major_cdotu_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies); + sycl::event (*row_major_zdotu_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* result, + const std::vector& dependencies); + sycl::event (*row_major_isamin_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies); + sycl::event (*row_major_idamin_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies); + sycl::event (*row_major_icamin_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies); + sycl::event (*row_major_izamin_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies); + sycl::event (*row_major_isamax_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies); + sycl::event (*row_major_idamax_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, std::int64_t* result, + const std::vector& dependencies); + sycl::event (*row_major_icamax_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies); + sycl::event (*row_major_izamax_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + std::int64_t* result, + const std::vector& dependencies); + sycl::event (*row_major_scnrm2_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + float* result, + const std::vector& dependencies); + sycl::event (*row_major_dznrm2_usm_sycl)(sycl::queue& queue, std::int64_t n, + const std::complex* x, std::int64_t incx, + double* result, + const std::vector& dependencies); + sycl::event (*row_major_snrm2_usm_sycl)(sycl::queue& queue, std::int64_t n, const float* x, + std::int64_t incx, float* result, + const std::vector& dependencies); + sycl::event (*row_major_dnrm2_usm_sycl)(sycl::queue& queue, std::int64_t n, const double* x, + std::int64_t incx, double* result, + const std::vector& dependencies); + sycl::event (*row_major_srot_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, float c, + float s, const std::vector& dependencies); + sycl::event (*row_major_drot_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, double c, + double s, const std::vector& dependencies); + sycl::event (*row_major_csrot_usm_sycl)(sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, float c, + float s, const std::vector& dependencies); + sycl::event (*row_major_zdrot_usm_sycl)(sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, double c, double s, - const std::vector &dependencies); - sycl::event (*row_major_srotg_usm_sycl)(sycl::queue &queue, float *a, float *b, float *c, - float *s, const std::vector &dependencies); - sycl::event (*row_major_drotg_usm_sycl)(sycl::queue &queue, double *a, double *b, double *c, - double *s, - const std::vector &dependencies); - sycl::event (*row_major_crotg_usm_sycl)(sycl::queue &queue, std::complex *a, - std::complex *b, float *c, - std::complex *s, - const std::vector &dependencies); - sycl::event (*row_major_zrotg_usm_sycl)(sycl::queue &queue, std::complex *a, - std::complex *b, double *c, - std::complex *s, - const std::vector &dependencies); - sycl::event (*row_major_srotm_usm_sycl)(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - float *param, - const std::vector &dependencies); - sycl::event (*row_major_drotm_usm_sycl)(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - double *param, - const std::vector &dependencies); - sycl::event (*row_major_srotmg_usm_sycl)(sycl::queue &queue, float *d1, float *d2, float *x1, - float y1, float *param, - const std::vector &dependencies); - sycl::event (*row_major_drotmg_usm_sycl)(sycl::queue &queue, double *d1, double *d2, double *x1, - double y1, double *param, - const std::vector &dependencies); - sycl::event (*row_major_sscal_usm_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - float *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_dscal_usm_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - double *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_cscal_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, + const std::vector& dependencies); + sycl::event (*row_major_srotg_usm_sycl)(sycl::queue& queue, float* a, float* b, float* c, + float* s, const std::vector& dependencies); + sycl::event (*row_major_drotg_usm_sycl)(sycl::queue& queue, double* a, double* b, double* c, + double* s, + const std::vector& dependencies); + sycl::event (*row_major_crotg_usm_sycl)(sycl::queue& queue, std::complex* a, + std::complex* b, float* c, + std::complex* s, + const std::vector& dependencies); + sycl::event (*row_major_zrotg_usm_sycl)(sycl::queue& queue, std::complex* a, + std::complex* b, double* c, + std::complex* s, + const std::vector& dependencies); + sycl::event (*row_major_srotm_usm_sycl)(sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, + float* param, + const std::vector& dependencies); + sycl::event (*row_major_drotm_usm_sycl)(sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, + double* param, + const std::vector& dependencies); + sycl::event (*row_major_srotmg_usm_sycl)(sycl::queue& queue, float* d1, float* d2, float* x1, + float y1, float* param, + const std::vector& dependencies); + sycl::event (*row_major_drotmg_usm_sycl)(sycl::queue& queue, double* d1, double* d2, double* x1, + double y1, double* param, + const std::vector& dependencies); + sycl::event (*row_major_sscal_usm_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + float* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_dscal_usm_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + double* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_cscal_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex alpha, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_csscal_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex alpha, std::complex *x, + const std::vector& dependencies); + sycl::event (*row_major_csscal_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex alpha, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_zscal_usm_sycl)(sycl::queue &queue, std::int64_t n, float alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_zdscal_usm_sycl)(sycl::queue &queue, std::int64_t n, double alpha, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_sdsdot_usm_sycl)(sycl::queue &queue, std::int64_t n, float sb, - const float *x, std::int64_t incx, const float *y, - std::int64_t incy, float *result, - const std::vector &dependencies); - sycl::event (*row_major_sswap_usm_sycl)(sycl::queue &queue, std::int64_t n, float *x, - std::int64_t incx, float *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_dswap_usm_sycl)(sycl::queue &queue, std::int64_t n, double *x, - std::int64_t incx, double *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_cswap_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_zswap_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex *x, std::int64_t incx, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_sgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*row_major_zscal_usm_sycl)(sycl::queue& queue, std::int64_t n, float alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_zdscal_usm_sycl)(sycl::queue& queue, std::int64_t n, double alpha, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_sdsdot_usm_sycl)(sycl::queue& queue, std::int64_t n, float sb, + const float* x, std::int64_t incx, const float* y, + std::int64_t incy, float* result, + const std::vector& dependencies); + sycl::event (*row_major_sswap_usm_sycl)(sycl::queue& queue, std::int64_t n, float* x, + std::int64_t incx, float* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_dswap_usm_sycl)(sycl::queue& queue, std::int64_t n, double* x, + std::int64_t incx, double* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_cswap_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_zswap_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex* x, std::int64_t incx, + std::complex* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_sgbmv_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, - float beta, float *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_dgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + std::int64_t ku, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_dgbmv_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies); + std::int64_t ku, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, + double beta, double* y, std::int64_t incy, + const std::vector& dependencies); sycl::event (*row_major_cgbmv_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_zgbmv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t kl, std::int64_t ku, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, const std::vector& dependencies); + sycl::event (*row_major_zgbmv_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, std::int64_t ku, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_sgemv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*row_major_sgemv_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, + const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_dgemv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*row_major_dgemv_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, + const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_cgemv_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*row_major_cgemv_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); + std::complex* y, std::int64_t incy, + const std::vector& dependencies); sycl::event (*row_major_zgemv_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, const std::vector& dependencies); sycl::event (*row_major_sgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, std::int64_t stridea, const float *x, - std::int64_t incx, std::int64_t stridex, float beta, float *y, std::int64_t incy, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, std::int64_t stridea, const float* x, + std::int64_t incx, std::int64_t stridex, float beta, float* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*row_major_dgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, std::int64_t stridea, const double *x, - std::int64_t incx, std::int64_t stridex, double beta, double *y, std::int64_t incy, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, std::int64_t stridea, const double* x, + std::int64_t incx, std::int64_t stridex, double beta, double* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*row_major_cgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex beta, std::complex *y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_zgemv_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stridea, const std::complex *x, std::int64_t incx, - std::int64_t stridex, std::complex beta, std::complex *y, std::int64_t incy, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stridea, const std::complex* x, std::int64_t incx, + std::int64_t stridex, std::complex beta, std::complex* y, std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*row_major_sgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - float *alpha, const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, - float *beta, float **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + float* alpha, const float** a, std::int64_t* lda, const float** x, std::int64_t* incx, + float* beta, float** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_dgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - double *alpha, const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, - double *beta, double **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + double* alpha, const double** a, std::int64_t* lda, const double** x, std::int64_t* incx, + double* beta, double** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_cgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_zgemv_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - const std::complex **x, std::int64_t *incx, std::complex *beta, - std::complex **y, std::int64_t *incy, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + const std::complex** x, std::int64_t* incx, std::complex* beta, + std::complex** y, std::int64_t* incy, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_sdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - const float *a, std::int64_t lda, std::int64_t stridea, const float *x, std::int64_t incx, - std::int64_t stridex, float *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + const float* a, std::int64_t lda, std::int64_t stridea, const float* x, std::int64_t incx, + std::int64_t stridex, float* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_ddgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - const double *a, std::int64_t lda, std::int64_t stridea, const double *x, std::int64_t incx, - std::int64_t stridex, double *c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + const double* a, std::int64_t lda, std::int64_t stridea, const double* x, std::int64_t incx, + std::int64_t stridex, double* c, std::int64_t ldc, std::int64_t stridec, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_cdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies); sycl::event (*row_major_zdgmm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, - const std::complex *a, std::int64_t lda, std::int64_t stridea, - const std::complex *x, std::int64_t incx, std::int64_t stridex, - std::complex *c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side left_right, std::int64_t m, std::int64_t n, + const std::complex* a, std::int64_t lda, std::int64_t stridea, + const std::complex* x, std::int64_t incx, std::int64_t stridex, + std::complex* c, std::int64_t ldc, std::int64_t stridec, std::int64_t batch_size, + const std::vector& dependencies); sycl::event (*row_major_sdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, - const float **a, std::int64_t *lda, const float **x, std::int64_t *incx, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, std::int64_t* n, + const float** a, std::int64_t* lda, const float** x, std::int64_t* incx, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_ddgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, - const double **a, std::int64_t *lda, const double **x, std::int64_t *incx, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, std::int64_t* n, + const double** a, std::int64_t* lda, const double** x, std::int64_t* incx, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_cdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, - const std::complex **a, std::int64_t *lda, const std::complex **x, - std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, const std::complex** x, + std::int64_t* incx, std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_zdgmm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, std::int64_t *m, std::int64_t *n, - const std::complex **a, std::int64_t *lda, const std::complex **x, - std::int64_t *incx, std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); - sycl::event (*row_major_sger_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - float alpha, const float *x, std::int64_t incx, - const float *y, std::int64_t incy, float *a, + sycl::queue& queue, oneapi::math::side* left_right, std::int64_t* m, std::int64_t* n, + const std::complex** a, std::int64_t* lda, const std::complex** x, + std::int64_t* incx, std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); + sycl::event (*row_major_sger_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + float alpha, const float* x, std::int64_t incx, + const float* y, std::int64_t incy, float* a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_dger_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - double alpha, const double *x, std::int64_t incx, - const double *y, std::int64_t incy, double *a, + const std::vector& dependencies); + sycl::event (*row_major_dger_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + double alpha, const double* x, std::int64_t incx, + const double* y, std::int64_t incy, double* a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_cgerc_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, + const std::vector& dependencies); + sycl::event (*row_major_cgerc_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_zgerc_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + const std::vector& dependencies); + sycl::event (*row_major_zgerc_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_cgeru_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *x, - std::int64_t incx, const std::complex *y, - std::int64_t incy, std::complex *a, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*row_major_cgeru_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* x, + std::int64_t incx, const std::complex* y, + std::int64_t incy, std::complex* a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_zgeru_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + const std::vector& dependencies); + sycl::event (*row_major_zgeru_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_chbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*row_major_chbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *x, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, - const std::vector &dependencies); + std::complex* y, std::int64_t incy, + const std::vector& dependencies); sycl::event (*row_major_zhbmv_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, std::complex beta, - std::complex *y, std::int64_t incy, const std::vector &dependencies); - sycl::event (*row_major_chemv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, + std::complex alpha, const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, std::complex beta, + std::complex* y, std::int64_t incy, const std::vector& dependencies); + sycl::event (*row_major_chemv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_zhemv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_zhemv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, + const std::complex* a, std::int64_t lda, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_cher_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_cher_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_zher_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*row_major_zher_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_cher2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*row_major_cher2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_zher2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*row_major_zher2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_chpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*row_major_chpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, + const std::complex* a, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_zhpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_zhpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *a, - const std::complex *x, std::int64_t incx, - std::complex beta, std::complex *y, + const std::complex* a, + const std::complex* x, std::int64_t incx, + std::complex beta, std::complex* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_chpr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_chpr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, float alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, - const std::vector &dependencies); - sycl::event (*row_major_zhpr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + std::complex* a, + const std::vector& dependencies); + sycl::event (*row_major_zhpr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, double alpha, - const std::complex *x, std::int64_t incx, - std::complex *a, - const std::vector &dependencies); - sycl::event (*row_major_chpr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + std::complex* a, + const std::vector& dependencies); + sycl::event (*row_major_chpr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, - const std::vector &dependencies); - sycl::event (*row_major_zhpr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies); + sycl::event (*row_major_zhpr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::complex alpha, - const std::complex *x, std::int64_t incx, - const std::complex *y, std::int64_t incy, - std::complex *a, - const std::vector &dependencies); - sycl::event (*row_major_ssbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* x, std::int64_t incx, + const std::complex* y, std::int64_t incy, + std::complex* a, + const std::vector& dependencies); + sycl::event (*row_major_ssbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, const float *x, - std::int64_t incx, float beta, float *y, + const float* a, std::int64_t lda, const float* x, + std::int64_t incx, float beta, float* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_dsbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_dsbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, const double *x, - std::int64_t incx, double beta, double *y, + const double* a, std::int64_t lda, const double* x, + std::int64_t incx, double beta, double* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_sspmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *a, - const float *x, std::int64_t incx, float beta, float *y, + const std::vector& dependencies); + sycl::event (*row_major_sspmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* a, + const float* x, std::int64_t incx, float beta, float* y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_dspmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *a, - const double *x, std::int64_t incx, double beta, - double *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_sspr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, - std::int64_t incx, float *a, - const std::vector &dependencies); - sycl::event (*row_major_dspr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, - std::int64_t incx, double *a, - const std::vector &dependencies); - sycl::event (*row_major_sspr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - float *a, const std::vector &dependencies); - sycl::event (*row_major_dspr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, - double *a, - const std::vector &dependencies); - sycl::event (*row_major_ssymv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, const float *x, std::int64_t incx, - float beta, float *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_dsymv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *x, std::int64_t incx, - double beta, double *y, std::int64_t incy, - const std::vector &dependencies); - sycl::event (*row_major_ssyr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, - std::int64_t incx, float *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_dsyr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, - std::int64_t incx, double *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_ssyr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, float alpha, const float *x, - std::int64_t incx, const float *y, std::int64_t incy, - float *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_dsyr2_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, - std::int64_t n, double alpha, const double *x, - std::int64_t incx, const double *y, std::int64_t incy, - double *a, std::int64_t lda, - const std::vector &dependencies); - sycl::event (*row_major_stbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_dspmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* a, + const double* x, std::int64_t incx, double beta, + double* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_sspr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* x, + std::int64_t incx, float* a, + const std::vector& dependencies); + sycl::event (*row_major_dspr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* a, + const std::vector& dependencies); + sycl::event (*row_major_sspr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, + float* a, const std::vector& dependencies); + sycl::event (*row_major_dspr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, + double* a, + const std::vector& dependencies); + sycl::event (*row_major_ssymv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, const float* x, std::int64_t incx, + float beta, float* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_dsymv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* a, + std::int64_t lda, const double* x, std::int64_t incx, + double beta, double* y, std::int64_t incy, + const std::vector& dependencies); + sycl::event (*row_major_ssyr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* x, + std::int64_t incx, float* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*row_major_dsyr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* x, + std::int64_t incx, double* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*row_major_ssyr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, float alpha, const float* x, + std::int64_t incx, const float* y, std::int64_t incy, + float* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*row_major_dsyr2_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, + std::int64_t n, double alpha, const double* x, + std::int64_t incx, const double* y, std::int64_t incy, + double* a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*row_major_stbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_dtbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_dtbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const double *a, std::int64_t lda, - double *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ctbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_ctbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, + std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ztbmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_ztbmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, + std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_stbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_stbsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const float *a, std::int64_t lda, - float *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_dtbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, const float* a, std::int64_t lda, + float* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_dtbsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const double *a, std::int64_t lda, - double *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ctbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, const double* a, std::int64_t lda, + double* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_ctbsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, + std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ztbsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_ztbsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - std::int64_t k, const std::complex *a, - std::int64_t lda, std::complex *x, + std::int64_t k, const std::complex* a, + std::int64_t lda, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_stpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_stpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const float *a, float *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_dtpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const float* a, float* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_dtpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const double *a, double *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ctpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const double* a, double* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_ctpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::complex *x, + const std::complex* a, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ztpmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_ztpmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::complex *x, + const std::complex* a, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_stpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_stpsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const float *a, float *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_dtpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const float* a, float* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_dtpsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const double *a, double *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ctpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const double* a, double* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_ctpsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::complex *x, + const std::complex* a, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ztpsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_ztpsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::complex *x, + const std::complex* a, std::complex* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_strmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_strmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const float *a, std::int64_t lda, float *x, + const float* a, std::int64_t lda, float* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_dtrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_dtrmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const double *a, std::int64_t lda, double *x, + const double* a, std::int64_t lda, double* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ctrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_ctrmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ztrmv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_ztrmv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_strsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_strsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const float *a, std::int64_t lda, float *x, + const float* a, std::int64_t lda, float* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_dtrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_dtrsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const double *a, std::int64_t lda, double *x, + const double* a, std::int64_t lda, double* x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ctrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_ctrsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_ztrsv_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_ztrsv_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t n, - const std::complex *a, std::int64_t lda, - std::complex *x, std::int64_t incx, - const std::vector &dependencies); - sycl::event (*row_major_sgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + const std::complex* a, std::int64_t lda, + std::complex* x, std::int64_t incx, + const std::vector& dependencies); + sycl::event (*row_major_sgemm_usm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const float *a, std::int64_t lda, const float *b, - std::int64_t ldb, float beta, float *c, + const float* a, std::int64_t lda, const float* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_dgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + const std::vector& dependencies); + sycl::event (*row_major_dgemm_usm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, double alpha, - const double *a, std::int64_t lda, const double *b, - std::int64_t ldb, double beta, double *c, + const double* a, std::int64_t lda, const double* b, + std::int64_t ldb, double beta, double* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_cgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + const std::vector& dependencies); + sycl::event (*row_major_cgemm_usm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, + std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies); + std::complex* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*row_major_zgemm_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_hgemm_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*row_major_hgemm_usm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, - const sycl::half *a, std::int64_t lda, - const sycl::half *b, std::int64_t ldb, sycl::half beta, - sycl::half *c, std::int64_t ldc, - const std::vector &dependencies); + const sycl::half* a, std::int64_t lda, + const sycl::half* b, std::int64_t ldb, sycl::half beta, + sycl::half* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*row_major_gemm_f16f16f32_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, const sycl::half *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, const sycl::half* b, std::int64_t ldb, float beta, float* c, + std::int64_t ldc, const std::vector& dependencies); sycl::event (*row_major_gemm_bf16bf16f32_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const oneapi::math::bfloat16 *a, - std::int64_t lda, const oneapi::math::bfloat16 *b, std::int64_t ldb, float beta, float *c, - std::int64_t ldc, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, + const oneapi::math::bfloat16* a, std::int64_t lda, const oneapi::math::bfloat16* b, + std::int64_t ldb, float beta, float* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*row_major_chemm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_zhemm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, const std::vector& dependencies); + sycl::event (*row_major_zhemm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_cherk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_cherk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, float alpha, - const std::complex *a, std::int64_t lda, - float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_zherk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, std::int64_t lda, + float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*row_major_zherk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, double alpha, - const std::complex *a, std::int64_t lda, - double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_cher2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, std::int64_t lda, + double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*row_major_cher2k_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - float beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_zher2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + float beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*row_major_zher2k_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - double beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_ssymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + double beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*row_major_ssymm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_dsymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*row_major_dsymm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies); + std::int64_t n, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*row_major_csymm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_zsymm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, const std::vector& dependencies); + sycl::event (*row_major_zsymm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_ssyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_ssyrk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, - std::int64_t lda, float beta, float *c, + std::int64_t k, float alpha, const float* a, + std::int64_t lda, float beta, float* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_dsyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_dsyrk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, - std::int64_t lda, double beta, double *c, + std::int64_t k, double alpha, const double* a, + std::int64_t lda, double beta, double* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_csyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_csyrk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex beta, std::complex *c, + const std::complex* a, std::int64_t lda, + std::complex beta, std::complex* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_zsyrk_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_zsyrk_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex beta, std::complex *c, + const std::complex* a, std::int64_t lda, + std::complex beta, std::complex* c, std::int64_t ldc, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*row_major_ssyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *k, float *alpha, const float **a, std::int64_t *lda, - float *beta, float **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* k, float* alpha, const float** a, std::int64_t* lda, + float* beta, float** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_dsyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *k, double *alpha, const double **a, std::int64_t *lda, - double *beta, double **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* k, double* alpha, const double** a, std::int64_t* lda, + double* beta, double** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_csyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *k, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* k, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_zsyrk_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo *upper_lower, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, std::complex *beta, - std::complex **c, std::int64_t *ldc, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo* upper_lower, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, std::complex* beta, + std::complex** c, std::int64_t* ldc, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_ssyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_dsyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double beta, double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_csyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*row_major_zsyrk_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex *c, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); - sycl::event (*row_major_ssyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + const std::vector& dependencies); + sycl::event (*row_major_ssyr2k_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_dsyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*row_major_dsyr2k_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, - std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies); + std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*row_major_csyr2k_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, - std::int64_t n, std::int64_t k, std::complex alpha, const std::complex *a, - std::int64_t lda, const std::complex *b, std::int64_t ldb, std::complex beta, - std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_zsyr2k_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, + std::int64_t n, std::int64_t k, std::complex alpha, const std::complex* a, + std::int64_t lda, const std::complex* b, std::int64_t ldb, std::complex beta, + std::complex* c, std::int64_t ldc, const std::vector& dependencies); + sycl::event (*row_major_zsyr2k_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, - const std::complex *b, std::int64_t ldb, - std::complex beta, std::complex *c, + const std::complex* a, std::int64_t lda, + const std::complex* b, std::int64_t ldb, + std::complex beta, std::complex* c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_strmm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + const std::vector& dependencies); + sycl::event (*row_major_strmm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*row_major_dtrmm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*row_major_dtrmm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies); + std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies); sycl::event (*row_major_ctrmm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, const std::vector &dependencies); + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, const std::vector& dependencies); sycl::event (*row_major_ztrmm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, const std::vector &dependencies); - sycl::event (*row_major_strsm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, const std::vector& dependencies); + sycl::event (*row_major_strsm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, float *b, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*row_major_dtrsm_usm_sycl)(sycl::queue &queue, oneapi::math::side left_right, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, float* b, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*row_major_dtrsm_usm_sycl)(sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, double *b, std::int64_t ldb, - const std::vector &dependencies); + std::int64_t n, double alpha, const double* a, + std::int64_t lda, double* b, std::int64_t ldb, + const std::vector& dependencies); sycl::event (*row_major_ctrsm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, const std::vector &dependencies); + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, const std::vector& dependencies); sycl::event (*row_major_ztrsm_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, const std::vector &dependencies); + std::complex alpha, const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, const std::vector& dependencies); sycl::event (*row_major_strsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*row_major_dtrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*row_major_ctrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies); + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_ztrsm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::side left_right, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, oneapi::math::diag unit_diag, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies); + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_strsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, - std::int64_t *n, float *alpha, const float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, std::int64_t* m, + std::int64_t* n, float* alpha, const float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_dtrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, - std::int64_t *n, double *alpha, const double **a, std::int64_t *lda, double **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, std::int64_t* m, + std::int64_t* n, double* alpha, const double** a, std::int64_t* lda, double** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_ctrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_ztrsm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::side *left_right, oneapi::math::uplo *upper_lower, - oneapi::math::transpose *trans, oneapi::math::diag *unit_diag, std::int64_t *m, - std::int64_t *n, std::complex *alpha, const std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::side* left_right, oneapi::math::uplo* upper_lower, + oneapi::math::transpose* trans, oneapi::math::diag* unit_diag, std::int64_t* m, + std::int64_t* n, std::complex* alpha, const std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_size, const std::vector& dependencies); sycl::event (*row_major_sgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const float **a, - std::int64_t *lda, const float **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, float* alpha, const float** a, + std::int64_t* lda, const float** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_dgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, double *alpha, const double **a, - std::int64_t *lda, const double **b, std::int64_t *ldb, double *beta, double **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, double* alpha, const double** a, + std::int64_t* lda, const double** b, std::int64_t* ldb, double* beta, double** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_cgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, const std::complex **b, - std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, const std::complex** b, + std::int64_t* ldb, std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_zgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, std::complex *alpha, - const std::complex **a, std::int64_t *lda, const std::complex **b, - std::int64_t *ldb, std::complex *beta, std::complex **c, std::int64_t *ldc, - std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, std::complex* alpha, + const std::complex** a, std::int64_t* lda, const std::complex** b, + std::int64_t* ldb, std::complex* beta, std::complex** c, std::int64_t* ldc, + std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_hgemm_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, sycl::half *alpha, const sycl::half **a, - std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, sycl::half *beta, - sycl::half **c, std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, sycl::half* alpha, const sycl::half** a, + std::int64_t* lda, const sycl::half** b, std::int64_t* ldb, sycl::half* beta, + sycl::half** c, std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_gemm_f16f16f32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const sycl::half **a, - std::int64_t *lda, const sycl::half **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, float* alpha, const sycl::half** a, + std::int64_t* lda, const sycl::half** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_gemm_s8s8f32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, - std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, float **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, float* alpha, const std::int8_t** a, + std::int64_t* lda, const std::int8_t** b, std::int64_t* ldb, float* beta, float** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_gemm_s8s8s32_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *transa, oneapi::math::transpose *transb, - std::int64_t *m, std::int64_t *n, std::int64_t *k, float *alpha, const std::int8_t **a, - std::int64_t *lda, const std::int8_t **b, std::int64_t *ldb, float *beta, std::int32_t **c, - std::int64_t *ldc, std::int64_t group_count, std::int64_t *group_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* transa, oneapi::math::transpose* transb, + std::int64_t* m, std::int64_t* n, std::int64_t* k, float* alpha, const std::int8_t** a, + std::int64_t* lda, const std::int8_t** b, std::int64_t* ldb, float* beta, std::int32_t** c, + std::int64_t* ldc, std::int64_t group_count, std::int64_t* group_size, + const std::vector& dependencies); sycl::event (*row_major_sgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float *a, - std::int64_t lda, std::int64_t stride_a, const float *b, std::int64_t ldb, - std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const float* a, + std::int64_t lda, std::int64_t stride_a, const float* b, std::int64_t ldb, + std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_dgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double *a, - std::int64_t lda, std::int64_t stride_a, const double *b, std::int64_t ldb, - std::int64_t stride_b, double beta, double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, double alpha, const double* a, + std::int64_t lda, std::int64_t stride_a, const double* b, std::int64_t ldb, + std::int64_t stride_b, double beta, double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_cgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_zgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, std::int64_t stride_a, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, std::complex *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + const std::complex* a, std::int64_t lda, std::int64_t stride_a, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex beta, std::complex* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_hgemm_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half *a, - std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, sycl::half beta, sycl::half *c, std::int64_t ldc, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::half alpha, const sycl::half* a, + std::int64_t lda, std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, sycl::half beta, sycl::half* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*row_major_gemm_f16f16f32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half *a, - std::int64_t lda, std::int64_t stride_a, const sycl::half *b, std::int64_t ldb, - std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const sycl::half* a, + std::int64_t lda, std::int64_t stride_a, const sycl::half* b, std::int64_t ldb, + std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_gemm_s8s8f32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, - std::int64_t stride_b, float beta, float *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t* a, + std::int64_t lda, std::int64_t stride_a, const std::int8_t* b, std::int64_t ldb, + std::int64_t stride_b, float beta, float* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_gemm_s8s8s32_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t *a, - std::int64_t lda, std::int64_t stride_a, const std::int8_t *b, std::int64_t ldb, - std::int64_t stride_b, float beta, std::int32_t *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); - sycl::event (*row_major_sgemmt_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::int64_t k, float alpha, const std::int8_t* a, + std::int64_t lda, std::int64_t stride_a, const std::int8_t* b, std::int64_t ldb, + std::int64_t stride_b, float beta, std::int32_t* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); + sycl::event (*row_major_sgemmt_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, - std::int64_t k, float alpha, const float *a, - std::int64_t lda, const float *b, std::int64_t ldb, - float beta, float *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_dgemmt_usm_sycl)(sycl::queue &queue, oneapi::math::uplo upper_lower, + std::int64_t k, float alpha, const float* a, + std::int64_t lda, const float* b, std::int64_t ldb, + float beta, float* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*row_major_dgemmt_usm_sycl)(sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, - std::int64_t k, double alpha, const double *a, - std::int64_t lda, const double *b, std::int64_t ldb, - double beta, double *c, std::int64_t ldc, - const std::vector &dependencies); + std::int64_t k, double alpha, const double* a, + std::int64_t lda, const double* b, std::int64_t ldb, + double beta, double* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*row_major_cgemmt_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*row_major_zgemmt_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, + sycl::queue& queue, oneapi::math::uplo upper_lower, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t n, std::int64_t k, std::complex alpha, - const std::complex *a, std::int64_t lda, const std::complex *b, - std::int64_t ldb, std::complex beta, std::complex *c, std::int64_t ldc, - const std::vector &dependencies); + const std::complex* a, std::int64_t lda, const std::complex* b, + std::int64_t ldb, std::complex beta, std::complex* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*row_major_gemm_s8u8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::uint8_t *b, - std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies); + const std::int8_t* a, std::int64_t lda, std::int8_t ao, const std::uint8_t* b, + std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies); sycl::event (*row_major_gemm_s8s8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::int8_t *a, std::int64_t lda, std::int8_t ao, const std::int8_t *b, - std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies); + const std::int8_t* a, std::int64_t lda, std::int8_t ao, const std::int8_t* b, + std::int64_t ldb, std::int8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies); sycl::event (*row_major_gemm_u8s8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::int8_t *b, - std::int64_t ldb, std::int8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies); + const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, const std::int8_t* b, + std::int64_t ldb, std::int8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies); sycl::event (*row_major_gemm_u8u8s32_bias_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, oneapi::math::offset offsetc, std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - const std::uint8_t *a, std::int64_t lda, std::uint8_t ao, const std::uint8_t *b, - std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t *c, std::int64_t ldc, - const std::int32_t *co, const std::vector &dependencies); + const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, const std::uint8_t* b, + std::int64_t ldb, std::uint8_t bo, float beta, std::int32_t* c, std::int64_t ldc, + const std::int32_t* co, const std::vector& dependencies); sycl::event (*row_major_somatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float alpha, const float *a, std::int64_t lda, std::int64_t stride_a, float *b, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + float alpha, const float* a, std::int64_t lda, std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*row_major_domatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double alpha, const double *a, std::int64_t lda, std::int64_t stride_a, double *b, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + double alpha, const double* a, std::int64_t lda, std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*row_major_comatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_zomatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, const std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, const std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_simatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - float alpha, float *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + float alpha, float* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_dimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - double alpha, double *ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + double alpha, double* ab, std::int64_t lda, std::int64_t ldb, std::int64_t stride, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_cimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, - std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, + std::int64_t stride, std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_zimatcopy_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, std::complex *ab, std::int64_t lda, std::int64_t ldb, - std::int64_t stride, std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, + std::int64_t stride, std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_somatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, float alpha, const float *a, std::int64_t lda, - std::int64_t stride_a, float beta, const float *b, std::int64_t ldb, std::int64_t stride_b, - float *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, float alpha, const float* a, std::int64_t lda, + std::int64_t stride_a, float beta, const float* b, std::int64_t ldb, std::int64_t stride_b, + float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies); sycl::event (*row_major_domatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, double alpha, const double *a, std::int64_t lda, - std::int64_t stride_a, double beta, const double *b, std::int64_t ldb, - std::int64_t stride_b, double *c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, double alpha, const double* a, std::int64_t lda, + std::int64_t stride_a, double beta, const double* b, std::int64_t ldb, + std::int64_t stride_b, double* c, std::int64_t ldc, std::int64_t stride_c, + std::int64_t batch_size, const std::vector& dependencies); sycl::event (*row_major_comatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies); sycl::event (*row_major_zomatadd_batch_strided_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, std::int64_t lda, std::int64_t stride_a, std::complex beta, - const std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::complex *c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector &dependencies); - sycl::event (*row_major_somatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::complex* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, + const std::vector& dependencies); + sycl::event (*row_major_somatcopy_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, float *b, + const float* a, std::int64_t lda, float* b, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*row_major_domatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*row_major_domatcopy_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, double *b, + const double* a, std::int64_t lda, double* b, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*row_major_comatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*row_major_comatcopy_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*row_major_zomatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*row_major_zomatcopy_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*row_major_somatcopy2_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*row_major_somatcopy2_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - const float *a, std::int64_t lda, - std::int64_t stridea, float *b, std::int64_t ldb, + const float* a, std::int64_t lda, + std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies); - sycl::event (*row_major_domatcopy2_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*row_major_domatcopy2_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - const double *a, std::int64_t lda, - std::int64_t stridea, double *b, std::int64_t ldb, + const double* a, std::int64_t lda, + std::int64_t stridea, double* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies); - sycl::event (*row_major_comatcopy2_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*row_major_comatcopy2_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies); - sycl::event (*row_major_zomatcopy2_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*row_major_zomatcopy2_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, - std::int64_t stridea, std::complex *b, + const std::complex* a, std::int64_t lda, + std::int64_t stridea, std::complex* b, std::int64_t ldb, std::int64_t strideb, - const std::vector &dependencies); - sycl::event (*row_major_simatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*row_major_simatcopy_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float alpha, - float *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*row_major_dimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + float* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*row_major_dimatcopy_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double alpha, - double *ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*row_major_cimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + double* ab, std::int64_t lda, std::int64_t ldb, + const std::vector& dependencies); + sycl::event (*row_major_cimatcopy_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, std::complex *ab, + std::complex alpha, std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*row_major_zimatcopy_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + const std::vector& dependencies); + sycl::event (*row_major_zimatcopy_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::complex alpha, - std::complex *ab, std::int64_t lda, + std::complex* ab, std::int64_t lda, std::int64_t ldb, - const std::vector &dependencies); - sycl::event (*row_major_somatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + const std::vector& dependencies); + sycl::event (*row_major_somatadd_usm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, float alpha, const float *a, - std::int64_t lda, float beta, const float *b, - std::int64_t ldb, float *c, std::int64_t ldc, - const std::vector &dependencies); - sycl::event (*row_major_domatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + std::int64_t n, float alpha, const float* a, + std::int64_t lda, float beta, const float* b, + std::int64_t ldb, float* c, std::int64_t ldc, + const std::vector& dependencies); + sycl::event (*row_major_domatadd_usm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, - std::int64_t n, double alpha, const double *a, - std::int64_t lda, double beta, const double *b, - std::int64_t ldb, double *c, std::int64_t ldc, - const std::vector &dependencies); + std::int64_t n, double alpha, const double* a, + std::int64_t lda, double beta, const double* b, + std::int64_t ldb, double* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*row_major_comatadd_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose transa, oneapi::math::transpose transb, - std::int64_t m, std::int64_t n, std::complex alpha, const std::complex *a, - std::int64_t lda, std::complex beta, const std::complex *b, std::int64_t ldb, - std::complex *c, std::int64_t ldc, const std::vector &dependencies); - sycl::event (*row_major_zomatadd_usm_sycl)(sycl::queue &queue, oneapi::math::transpose transa, + sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, + std::int64_t m, std::int64_t n, std::complex alpha, const std::complex* a, + std::int64_t lda, std::complex beta, const std::complex* b, std::int64_t ldb, + std::complex* c, std::int64_t ldc, const std::vector& dependencies); + sycl::event (*row_major_zomatadd_usm_sycl)(sycl::queue& queue, oneapi::math::transpose transa, oneapi::math::transpose transb, std::int64_t m, std::int64_t n, std::complex alpha, - const std::complex *a, std::int64_t lda, + const std::complex* a, std::int64_t lda, std::complex beta, - const std::complex *b, std::int64_t ldb, - std::complex *c, std::int64_t ldc, - const std::vector &dependencies); + const std::complex* b, std::int64_t ldb, + std::complex* c, std::int64_t ldc, + const std::vector& dependencies); sycl::event (*row_major_somatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - float *alpha, const float **a, std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + float* alpha, const float** a, std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies); sycl::event (*row_major_domatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - double *alpha, const double **a, std::int64_t *lda, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + double* alpha, const double** a, std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies); sycl::event (*row_major_comatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies); sycl::event (*row_major_zomatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, const std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, const std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies); sycl::event (*row_major_simatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - float *alpha, float **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + float* alpha, float** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies); sycl::event (*row_major_dimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - double *alpha, double **ab, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *groupsize, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + double* alpha, double** ab, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* groupsize, const std::vector& dependencies); sycl::event (*row_major_cimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, std::complex **ab, std::int64_t *lda, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, std::complex** ab, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies); sycl::event (*row_major_zimatcopy_batch_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *m, std::int64_t *n, - std::complex *alpha, std::complex **ab, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *groupsize, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* m, std::int64_t* n, + std::complex* alpha, std::complex** ab, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* groupsize, + const std::vector& dependencies); } blas_function_table_t; diff --git a/src/dft/backends/backend_backward_instantiations.cxx b/src/dft/backends/backend_backward_instantiations.cxx index eab281a96..e475ff55c 100644 --- a/src/dft/backends/backend_backward_instantiations.cxx +++ b/src/dft/backends/backend_backward_instantiations.cxx @@ -25,29 +25,29 @@ using desc_rd_t = dft::detail::descriptor; using desc_cd_t = dft::detail::descriptor; -using depends_vec_t = const std::vector &; +using depends_vec_t = const std::vector&; -#define ONEMATH_DFT_BACKWARD_INSTANTIATIONS(DESCRIPTOR_T, SCALAR_T, FORWARD_T, BACKWARD_T) \ +#define ONEMATH_DFT_BACKWARD_INSTANTIATIONS(DESCRIPTOR_T, SCALAR_T, FORWARD_T, BACKWARD_T) \ /* Buffer API */ \ - template ONEMATH_EXPORT void compute_backward(DESCRIPTOR_T &, \ - sycl::buffer &); \ - template ONEMATH_EXPORT void compute_backward( \ - DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &); \ - template ONEMATH_EXPORT void compute_backward( \ - DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &); \ - template ONEMATH_EXPORT void compute_backward( \ - DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &, \ - sycl::buffer &, sycl::buffer &); \ + template ONEMATH_EXPORT void compute_backward(DESCRIPTOR_T&, \ + sycl::buffer&); \ + template ONEMATH_EXPORT void compute_backward( \ + DESCRIPTOR_T&, sycl::buffer&, sycl::buffer&); \ + template ONEMATH_EXPORT void compute_backward( \ + DESCRIPTOR_T&, sycl::buffer&, sycl::buffer&); \ + template ONEMATH_EXPORT void compute_backward( \ + DESCRIPTOR_T&, sycl::buffer&, sycl::buffer&, sycl::buffer&, \ + sycl::buffer&); \ \ /* USM API */ \ - template ONEMATH_EXPORT sycl::event compute_backward(DESCRIPTOR_T &, FORWARD_T *, \ - depends_vec_t); \ - template ONEMATH_EXPORT sycl::event compute_backward(DESCRIPTOR_T &, SCALAR_T *, \ - SCALAR_T *, depends_vec_t); \ - template ONEMATH_EXPORT sycl::event compute_backward( \ - DESCRIPTOR_T &, BACKWARD_T *, FORWARD_T *, depends_vec_t); \ - template ONEMATH_EXPORT sycl::event compute_backward( \ - DESCRIPTOR_T &, SCALAR_T *, SCALAR_T *, SCALAR_T *, SCALAR_T *, depends_vec_t); + template ONEMATH_EXPORT sycl::event compute_backward(DESCRIPTOR_T&, FORWARD_T*, \ + depends_vec_t); \ + template ONEMATH_EXPORT sycl::event compute_backward(DESCRIPTOR_T&, SCALAR_T*, \ + SCALAR_T*, depends_vec_t); \ + template ONEMATH_EXPORT sycl::event compute_backward(DESCRIPTOR_T&, BACKWARD_T*, \ + FORWARD_T*, depends_vec_t); \ + template ONEMATH_EXPORT sycl::event compute_backward( \ + DESCRIPTOR_T&, SCALAR_T*, SCALAR_T*, SCALAR_T*, SCALAR_T*, depends_vec_t); ONEMATH_DFT_BACKWARD_INSTANTIATIONS(desc_rf_t, float, float, std::complex) ONEMATH_DFT_BACKWARD_INSTANTIATIONS(desc_cf_t, float, std::complex, std::complex) diff --git a/src/dft/backends/backend_compute_signature.cxx b/src/dft/backends/backend_compute_signature.cxx index 0a154eeb6..71fcd793a 100644 --- a/src/dft/backends/backend_compute_signature.cxx +++ b/src/dft/backends/backend_compute_signature.cxx @@ -63,7 +63,8 @@ void forward_op_cc(descriptor_type& desc, sycl::buffer& in, sycl::event forward_op_cc(descriptor_type& desc, fwd_type* in, bwd_type* out, const std::vector& dependencies) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_forward"); - return oneapi::math::dft::BACKEND::compute_forward(desc, in, out, dependencies); + return oneapi::math::dft::BACKEND::compute_forward(desc, in, out, + dependencies); } // forward out-of-place REAL_REAL @@ -79,7 +80,7 @@ sycl::event forward_op_rr(descriptor_type& desc, scalar_type* in_re, scalar_type const std::vector& dependencies) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_forward"); return oneapi::math::dft::BACKEND::compute_forward(desc, in_re, in_im, out_re, out_im, - dependencies); + dependencies); } // backward inplace COMPLEX_COMPLEX @@ -133,5 +134,5 @@ sycl::event backward_op_rr(descriptor_type& desc, scalar_type* in_re, scalar_typ const std::vector& dependencies) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_backward"); return oneapi::math::dft::BACKEND::compute_backward(desc, in_re, in_im, out_re, out_im, - dependencies); + dependencies); } diff --git a/src/dft/backends/backend_forward_instantiations.cxx b/src/dft/backends/backend_forward_instantiations.cxx index fc4084633..17ac748f4 100644 --- a/src/dft/backends/backend_forward_instantiations.cxx +++ b/src/dft/backends/backend_forward_instantiations.cxx @@ -25,29 +25,29 @@ using desc_rd_t = dft::detail::descriptor; using desc_cd_t = dft::detail::descriptor; -using depends_vec_t = const std::vector &; +using depends_vec_t = const std::vector&; -#define ONEMATH_DFT_FORWARD_INSTANTIATIONS(DESCRIPTOR_T, SCALAR_T, FORWARD_T, BACKWARD_T) \ +#define ONEMATH_DFT_FORWARD_INSTANTIATIONS(DESCRIPTOR_T, SCALAR_T, FORWARD_T, BACKWARD_T) \ /* Buffer API */ \ - template ONEMATH_EXPORT void compute_forward(DESCRIPTOR_T &, \ - sycl::buffer &); \ - template ONEMATH_EXPORT void compute_forward( \ - DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &); \ - template ONEMATH_EXPORT void compute_forward( \ - DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &); \ - template ONEMATH_EXPORT void compute_forward( \ - DESCRIPTOR_T &, sycl::buffer &, sycl::buffer &, \ - sycl::buffer &, sycl::buffer &); \ + template ONEMATH_EXPORT void compute_forward(DESCRIPTOR_T&, \ + sycl::buffer&); \ + template ONEMATH_EXPORT void compute_forward( \ + DESCRIPTOR_T&, sycl::buffer&, sycl::buffer&); \ + template ONEMATH_EXPORT void compute_forward( \ + DESCRIPTOR_T&, sycl::buffer&, sycl::buffer&); \ + template ONEMATH_EXPORT void compute_forward( \ + DESCRIPTOR_T&, sycl::buffer&, sycl::buffer&, sycl::buffer&, \ + sycl::buffer&); \ \ /* USM API */ \ - template ONEMATH_EXPORT sycl::event compute_forward(DESCRIPTOR_T &, FORWARD_T *, \ - depends_vec_t); \ - template ONEMATH_EXPORT sycl::event compute_forward(DESCRIPTOR_T &, SCALAR_T *, \ - SCALAR_T *, depends_vec_t); \ - template ONEMATH_EXPORT sycl::event compute_forward(DESCRIPTOR_T &, FORWARD_T *, \ - BACKWARD_T *, depends_vec_t); \ - template ONEMATH_EXPORT sycl::event compute_forward( \ - DESCRIPTOR_T &, SCALAR_T *, SCALAR_T *, SCALAR_T *, SCALAR_T *, depends_vec_t); + template ONEMATH_EXPORT sycl::event compute_forward(DESCRIPTOR_T&, FORWARD_T*, \ + depends_vec_t); \ + template ONEMATH_EXPORT sycl::event compute_forward(DESCRIPTOR_T&, SCALAR_T*, \ + SCALAR_T*, depends_vec_t); \ + template ONEMATH_EXPORT sycl::event compute_forward(DESCRIPTOR_T&, FORWARD_T*, \ + BACKWARD_T*, depends_vec_t); \ + template ONEMATH_EXPORT sycl::event compute_forward( \ + DESCRIPTOR_T&, SCALAR_T*, SCALAR_T*, SCALAR_T*, SCALAR_T*, depends_vec_t); ONEMATH_DFT_FORWARD_INSTANTIATIONS(desc_rf_t, float, float, std::complex) ONEMATH_DFT_FORWARD_INSTANTIATIONS(desc_cf_t, float, std::complex, std::complex) diff --git a/src/dft/backends/cufft/backward.cpp b/src/dft/backends/cufft/backward.cpp index 4dfff0c8b..475f1ea49 100644 --- a/src/dft/backends/cufft/backward.cpp +++ b/src/dft/backends/cufft/backward.cpp @@ -38,19 +38,19 @@ namespace oneapi::math::dft::cufft { namespace detail { //forward declaration template -std::array get_offsets_bwd(dft::detail::commit_impl *commit); +std::array get_offsets_bwd(dft::detail::commit_impl* commit); template -cufftHandle get_bwd_plan(dft::detail::commit_impl *commit) { - return static_cast *>(commit->get_handle())[1].value(); +cufftHandle get_bwd_plan(dft::detail::commit_impl* commit) { + return static_cast*>(commit->get_handle())[1].value(); } } // namespace detail // BUFFER version //In-place transform template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &inout) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& inout) { const std::string func_name = "compute_backward(desc, inout)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -68,35 +68,35 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, } } - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto inout_acc = inout.template get_access(cgh); commit->add_buffer_workspace_dependency_if_rqd("compute_backward", cgh); dft::detail::fft_enqueue_task(cgh, [=](sycl::interop_handle ih) { auto stream = detail::setup_stream(func_name, ih, plan); - auto inout_native = reinterpret_cast *>( + auto inout_native = reinterpret_cast*>( ih.get_native_mem(inout_acc)); detail::cufft_execute>( - func_name, stream, plan, reinterpret_cast(inout_native + offsets[0]), - reinterpret_cast(inout_native + offsets[1])); + func_name, stream, plan, reinterpret_cast(inout_native + offsets[0]), + reinterpret_cast(inout_native + offsets[1])); }); }); } //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_backward(descriptor_type &, sycl::buffer, 1> &, - sycl::buffer, 1> &) { +ONEMATH_EXPORT void compute_backward(descriptor_type&, sycl::buffer, 1>&, + sycl::buffer, 1>&) { throw oneapi::math::unimplemented("DFT", "compute_backward(desc, inout_re, inout_im)", - "cuFFT does not support real-real complex storage."); + "cuFFT does not support real-real complex storage."); } //Out-of-place transform template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &in, - sycl::buffer, 1> &out) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& in, + sycl::buffer, 1>& out) { const std::string func_name = "compute_backward(desc, in, out)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -113,7 +113,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, } } - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto in_acc = in.template get_access(cgh); auto out_acc = out.template get_access(cgh); commit->add_buffer_workspace_dependency_if_rqd("compute_backward", cgh); @@ -121,12 +121,12 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, dft::detail::fft_enqueue_task(cgh, [=](sycl::interop_handle ih) { auto stream = detail::setup_stream(func_name, ih, plan); - auto in_native = reinterpret_cast( - reinterpret_cast *>( + auto in_native = reinterpret_cast( + reinterpret_cast*>( ih.get_native_mem(in_acc)) + offsets[0]); - auto out_native = reinterpret_cast( - reinterpret_cast *>( + auto out_native = reinterpret_cast( + reinterpret_cast*>( ih.get_native_mem(out_acc)) + offsets[1]); detail::cufft_execute>( @@ -137,20 +137,20 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_backward(descriptor_type &, sycl::buffer, 1> &, - sycl::buffer, 1> &, - sycl::buffer, 1> &, - sycl::buffer, 1> &) { +ONEMATH_EXPORT void compute_backward(descriptor_type&, sycl::buffer, 1>&, + sycl::buffer, 1>&, + sycl::buffer, 1>&, + sycl::buffer, 1>&) { throw oneapi::math::unimplemented("DFT", "compute_backward(desc, in_re, in_im, out_re, out_im)", - "cuFFT does not support real-real complex storage."); + "cuFFT does not support real-real complex storage."); } //USM version //In-place transform template -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, fwd* inout, + const std::vector& dependencies) { const std::string func_name = "compute_backward(desc, inout, dependencies)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -168,7 +168,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwddepend_on_last_usm_workspace_event_if_rqd(cgh); @@ -185,19 +185,19 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &, scalar *, - scalar *, - const std::vector &) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type&, scalar*, + scalar*, + const std::vector&) { throw oneapi::math::unimplemented("DFT", - "compute_backward(desc, inout_re, inout_im, dependencies)", - "cuFFT does not support real-real complex storage."); + "compute_backward(desc, inout_re, inout_im, dependencies)", + "cuFFT does not support real-real complex storage."); } //Out-of-place transform template -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, - fwd *out, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, bwd* in, + fwd* out, + const std::vector& dependencies) { const std::string func_name = "compute_backward(desc, in, out, dependencies)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -214,7 +214,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwddepend_on_last_usm_workspace_event_if_rqd(cgh); @@ -231,13 +231,13 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &, scalar *, - scalar *, scalar *, - scalar *, - const std::vector &) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type&, scalar*, + scalar*, scalar*, + scalar*, + const std::vector&) { throw oneapi::math::unimplemented("DFT", - "compute_backward(desc, in_re, in_im, out_re, out_im, deps)", - "cuFFT does not support real-real complex storage."); + "compute_backward(desc, in_re, in_im, out_re, out_im, deps)", + "cuFFT does not support real-real complex storage."); } // Template function instantiations diff --git a/src/dft/backends/cufft/commit.cpp b/src/dft/backends/cufft/commit.cpp index 61be17e1f..b6d2164ff 100644 --- a/src/dft/backends/cufft/commit.cpp +++ b/src/dft/backends/cufft/commit.cpp @@ -57,7 +57,7 @@ class cufft_commit final : public dft::detail::commit_impl { public: cufft_commit(sycl::queue& queue, const dft::detail::dft_values& config_values) : oneapi::math::dft::detail::commit_impl(queue, backend::cufft, - config_values) { + config_values) { if constexpr (prec == dft::detail::precision::DOUBLE) { if (!queue.get_device().has(sycl::aspect::fp64)) { throw math::exception("DFT", "commit", "Device does not support double precision."); @@ -70,14 +70,14 @@ class cufft_commit final : public dft::detail::commit_impl { if (plans[0]) { if (cufftDestroy(plans[0].value()) != CUFFT_SUCCESS) { throw math::exception("dft/backends/cufft", __FUNCTION__, - "Failed to destroy forward cuFFT plan."); + "Failed to destroy forward cuFFT plan."); } plans[0] = std::nullopt; } if (plans[1]) { if (cufftDestroy(plans[1].value()) != CUFFT_SUCCESS) { throw math::exception("dft/backends/cufft", __FUNCTION__, - "Failed to destroy backward cuFFT plan."); + "Failed to destroy backward cuFFT plan."); } plans[1] = std::nullopt; } @@ -88,7 +88,7 @@ class cufft_commit final : public dft::detail::commit_impl { CUcontext interopContext; if (cuDevicePrimaryCtxRetain(&interopContext, interopDevice) != CUDA_SUCCESS) { throw math::exception("dft/backends/cufft", __FUNCTION__, - "Failed to change cuda context."); + "Failed to change cuda context."); } } } @@ -277,7 +277,7 @@ class cufft_commit final : public dft::detail::commit_impl { if (res != CUFFT_SUCCESS) { throw math::exception("dft/backends/cufft", __FUNCTION__, - "Failed to create forward cuFFT plan."); + "Failed to create forward cuFFT plan."); } plans[0] = fwd_plan; @@ -305,7 +305,7 @@ class cufft_commit final : public dft::detail::commit_impl { ); if (res != CUFFT_SUCCESS) { throw math::exception("dft/backends/cufft", __FUNCTION__, - "Failed to create backward cuFFT plan."); + "Failed to create backward cuFFT plan."); } plans[1] = bwd_plan; } @@ -321,7 +321,7 @@ class cufft_commit final : public dft::detail::commit_impl { auto res = cufftSetAutoAllocation(handle, 0); if (res != CUFFT_SUCCESS) { throw math::exception("dft/backends/cufft", "commit", - "cufftSetAutoAllocation(plan, 0) failed."); + "cufftSetAutoAllocation(plan, 0) failed."); } } } diff --git a/src/dft/backends/cufft/execute_helper.hpp b/src/dft/backends/cufft/execute_helper.hpp index 877136bfe..da485fea2 100644 --- a/src/dft/backends/cufft/execute_helper.hpp +++ b/src/dft/backends/cufft/execute_helper.hpp @@ -37,12 +37,12 @@ namespace oneapi::math::dft::cufft::detail { template -inline dft::detail::commit_impl *checked_get_commit( - dft::detail::descriptor &desc) { +inline dft::detail::commit_impl* checked_get_commit( + dft::detail::descriptor& desc) { auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::cufft) { throw math::invalid_argument("dft/backends/cufft", "get_commit", - "DFT descriptor has not been commited for cuFFT"); + "DFT descriptor has not been commited for cuFFT"); } return commit_handle; } @@ -50,7 +50,7 @@ inline dft::detail::commit_impl *checked_get_commit( /// Throw an math::invalid_argument if the runtime param in the descriptor does not match /// the expected value. template -inline auto expect_config(DescT &desc, const char *message) { +inline auto expect_config(DescT& desc, const char* message) { dft::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { @@ -61,8 +61,8 @@ inline auto expect_config(DescT &desc, const char *message) { enum class Direction { Forward = CUFFT_FORWARD, Backward = CUFFT_INVERSE }; template -void cufft_execute(const std::string &func, CUstream stream, cufftHandle plan, void *input, - void *output) { +void cufft_execute(const std::string& func, CUstream stream, cufftHandle plan, void* input, + void* output) { constexpr bool is_real = std::is_floating_point_v; using single_type = std::conditional_t>; constexpr bool is_single = std::is_same_v; @@ -70,37 +70,41 @@ void cufft_execute(const std::string &func, CUstream stream, cufftHandle plan, v if constexpr (is_real) { if constexpr (dir == Direction::Forward) { if constexpr (is_single) { - auto result = cufftExecR2C(plan, reinterpret_cast(input), - reinterpret_cast(output)); + auto result = cufftExecR2C(plan, reinterpret_cast(input), + reinterpret_cast(output)); if (result != CUFFT_SUCCESS) { - throw oneapi::math::exception("dft/backends/cufft", func, - "cufftExecR2C returned " + std::to_string(result)); + throw oneapi::math::exception( + "dft/backends/cufft", func, + "cufftExecR2C returned " + std::to_string(result)); } } else { - auto result = cufftExecD2Z(plan, reinterpret_cast(input), - reinterpret_cast(output)); + auto result = cufftExecD2Z(plan, reinterpret_cast(input), + reinterpret_cast(output)); if (result != CUFFT_SUCCESS) { - throw oneapi::math::exception("dft/backends/cufft", func, - "cufftExecD2Z returned " + std::to_string(result)); + throw oneapi::math::exception( + "dft/backends/cufft", func, + "cufftExecD2Z returned " + std::to_string(result)); } } } else { if constexpr (is_single) { - auto result = cufftExecC2R(plan, reinterpret_cast(input), - reinterpret_cast(output)); + auto result = cufftExecC2R(plan, reinterpret_cast(input), + reinterpret_cast(output)); if (result != CUFFT_SUCCESS) { - throw oneapi::math::exception("dft/backends/cufft", func, - "cufftExecC2R returned " + std::to_string(result)); + throw oneapi::math::exception( + "dft/backends/cufft", func, + "cufftExecC2R returned " + std::to_string(result)); } } else { - auto result = cufftExecZ2D(plan, reinterpret_cast(input), - reinterpret_cast(output)); + auto result = cufftExecZ2D(plan, reinterpret_cast(input), + reinterpret_cast(output)); if (result != CUFFT_SUCCESS) { - throw oneapi::math::exception("dft/backends/cufft", func, - "cufftExecZ2D returned " + std::to_string(result)); + throw oneapi::math::exception( + "dft/backends/cufft", func, + "cufftExecZ2D returned " + std::to_string(result)); } } } @@ -108,20 +112,20 @@ void cufft_execute(const std::string &func, CUstream stream, cufftHandle plan, v else { if constexpr (is_single) { auto result = - cufftExecC2C(plan, reinterpret_cast(input), - reinterpret_cast(output), static_cast(dir)); + cufftExecC2C(plan, reinterpret_cast(input), + reinterpret_cast(output), static_cast(dir)); if (result != CUFFT_SUCCESS) { throw oneapi::math::exception("dft/backends/cufft", func, - "cufftExecC2C returned " + std::to_string(result)); + "cufftExecC2C returned " + std::to_string(result)); } } else { auto result = - cufftExecZ2Z(plan, reinterpret_cast(input), - reinterpret_cast(output), static_cast(dir)); + cufftExecZ2Z(plan, reinterpret_cast(input), + reinterpret_cast(output), static_cast(dir)); if (result != CUFFT_SUCCESS) { throw oneapi::math::exception("dft/backends/cufft", func, - "cufftExecZ2Z returned " + std::to_string(result)); + "cufftExecZ2Z returned " + std::to_string(result)); } } } @@ -132,17 +136,17 @@ void cufft_execute(const std::string &func, CUstream stream, cufftHandle plan, v auto result = cuStreamSynchronize(stream); if (result != CUDA_SUCCESS) { throw oneapi::math::exception("dft/backends/cufft", func, - "cuStreamSynchronize returned " + std::to_string(result)); + "cuStreamSynchronize returned " + std::to_string(result)); } #endif } -inline CUstream setup_stream(const std::string &func, sycl::interop_handle ih, cufftHandle plan) { +inline CUstream setup_stream(const std::string& func, sycl::interop_handle ih, cufftHandle plan) { auto stream = ih.get_native_queue(); auto result = cufftSetStream(plan, stream); if (result != CUFFT_SUCCESS) { throw oneapi::math::exception("dft/backends/cufft", func, - "cufftSetStream returned " + std::to_string(result)); + "cufftSetStream returned " + std::to_string(result)); } return stream; } diff --git a/src/dft/backends/cufft/forward.cpp b/src/dft/backends/cufft/forward.cpp index ca7b2bc62..6b2867b5f 100644 --- a/src/dft/backends/cufft/forward.cpp +++ b/src/dft/backends/cufft/forward.cpp @@ -40,11 +40,11 @@ namespace oneapi::math::dft::cufft { namespace detail { //forward declaration template -std::array get_offsets_fwd(dft::detail::commit_impl *commit); +std::array get_offsets_fwd(dft::detail::commit_impl* commit); template -cufftHandle get_fwd_plan(dft::detail::commit_impl *commit) { - return static_cast *>(commit->get_handle())[0].value(); +cufftHandle get_fwd_plan(dft::detail::commit_impl* commit) { + return static_cast*>(commit->get_handle())[0].value(); } } // namespace detail @@ -52,8 +52,8 @@ cufftHandle get_fwd_plan(dft::detail::commit_impl *commit) { //In-place transform template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> &inout) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& inout) { const std::string func_name = "compute_forward(desc, inout)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -71,34 +71,35 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, offsets[1] *= 2; // offset is supplied in complex but we offset scalar pointer } - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto inout_acc = inout.template get_access(cgh); commit->add_buffer_workspace_dependency_if_rqd("compute_forward", cgh); dft::detail::fft_enqueue_task(cgh, [=](sycl::interop_handle ih) { auto stream = detail::setup_stream(func_name, ih, plan); - auto inout_native = reinterpret_cast *>( + auto inout_native = reinterpret_cast*>( ih.get_native_mem(inout_acc)); detail::cufft_execute>( - func_name, stream, plan, reinterpret_cast(inout_native + offsets[0]), - reinterpret_cast(inout_native + offsets[1])); + func_name, stream, plan, reinterpret_cast(inout_native + offsets[0]), + reinterpret_cast(inout_native + offsets[1])); }); }); } //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_forward(descriptor_type &, sycl::buffer, 1> &, - sycl::buffer, 1> &) { +ONEMATH_EXPORT void compute_forward(descriptor_type&, sycl::buffer, 1>&, + sycl::buffer, 1>&) { throw oneapi::math::unimplemented("DFT", "compute_forward(desc, inout_re, inout_im)", - "cuFFT does not support real-real complex storage."); + "cuFFT does not support real-real complex storage."); } //Out-of-place transform template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, - sycl::buffer, 1> &out) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& in, + sycl::buffer, 1>& out) { const std::string func_name = "compute_forward(desc, in, out)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -115,7 +116,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer(cgh); auto out_acc = out.template get_access(cgh); commit->add_buffer_workspace_dependency_if_rqd("compute_forward", cgh); @@ -123,12 +124,12 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer( - reinterpret_cast *>( + auto in_native = reinterpret_cast( + reinterpret_cast*>( ih.get_native_mem(in_acc)) + offsets[0]); - auto out_native = reinterpret_cast( - reinterpret_cast *>( + auto out_native = reinterpret_cast( + reinterpret_cast*>( ih.get_native_mem(out_acc)) + offsets[1]); detail::cufft_execute>( @@ -139,20 +140,20 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer -ONEMATH_EXPORT void compute_forward(descriptor_type &, sycl::buffer, 1> &, - sycl::buffer, 1> &, - sycl::buffer, 1> &, - sycl::buffer, 1> &) { +ONEMATH_EXPORT void compute_forward(descriptor_type&, sycl::buffer, 1>&, + sycl::buffer, 1>&, + sycl::buffer, 1>&, + sycl::buffer, 1>&) { throw oneapi::math::unimplemented("DFT", "compute_forward(desc, in_re, in_im, out_re, out_im)", - "cuFFT does not support real-real complex storage."); + "cuFFT does not support real-real complex storage."); } //USM version //In-place transform template -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, fwd* inout, + const std::vector& dependencies) { const std::string func_name = "compute_forward(desc, inout, dependencies)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -170,7 +171,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwddepend_on_last_usm_workspace_event_if_rqd(cgh); @@ -187,19 +188,19 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &, scalar *, - scalar *, - const std::vector &) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type&, scalar*, + scalar*, + const std::vector&) { throw oneapi::math::unimplemented("DFT", - "compute_forward(desc, inout_re, inout_im, dependencies)", - "cuFFT does not support real-real complex storage."); + "compute_forward(desc, inout_re, inout_im, dependencies)", + "cuFFT does not support real-real complex storage."); } //Out-of-place transform template -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, - bwd *out, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, fwd* in, + bwd* out, + const std::vector& dependencies) { const std::string func_name = "compute_forward(desc, in, out, dependencies)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -216,7 +217,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwddepend_on_last_usm_workspace_event_if_rqd(cgh); @@ -233,10 +234,10 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &, scalar *, - scalar *, scalar *, - scalar *, - const std::vector &) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type&, scalar*, + scalar*, scalar*, + scalar*, + const std::vector&) { throw oneapi::math::unimplemented( "DFT", "compute_forward(desc, in_re, in_im, out_re, out_im, dependencies)", "cuFFT does not support real-real complex storage."); diff --git a/src/dft/backends/descriptor.cpp b/src/dft/backends/descriptor.cpp index a12a0083d..9a3d12911 100644 --- a/src/dft/backends/descriptor.cpp +++ b/src/dft/backends/descriptor.cpp @@ -25,7 +25,7 @@ namespace oneapi::math::dft::detail { template -void descriptor::commit(sycl::queue &queue) { +void descriptor::commit(sycl::queue& queue) { if (!pimpl_ || pimpl_->get_queue() != queue) { if (pimpl_) { pimpl_->get_queue().wait(); @@ -34,9 +34,9 @@ void descriptor::commit(sycl::queue &queue) { } pimpl_->commit(values_); } -template void descriptor::commit(sycl::queue &); -template void descriptor::commit(sycl::queue &); -template void descriptor::commit(sycl::queue &); -template void descriptor::commit(sycl::queue &); +template void descriptor::commit(sycl::queue&); +template void descriptor::commit(sycl::queue&); +template void descriptor::commit(sycl::queue&); +template void descriptor::commit(sycl::queue&); } //namespace oneapi::math::dft::detail diff --git a/src/dft/backends/mklcpu/backward.cpp b/src/dft/backends/mklcpu/backward.cpp index cf62db084..a210c247e 100644 --- a/src/dft/backends/mklcpu/backward.cpp +++ b/src/dft/backends/mklcpu/backward.cpp @@ -40,26 +40,26 @@ namespace detail { // BUFFER version // backward a MKLCPU DFT call to the backend, checking that the commit impl is valid. template -inline void check_bwd_commit(dft::descriptor &desc) { +inline void check_bwd_commit(dft::descriptor& desc) { auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::mklcpu) { throw math::invalid_argument("DFT", "computer_backward", - "DFT descriptor has not been commited for MKLCPU"); + "DFT descriptor has not been commited for MKLCPU"); } - auto mklcpu_desc = reinterpret_cast(commit_handle->get_handle()); + auto mklcpu_desc = reinterpret_cast(commit_handle->get_handle()); MKL_LONG commit_status{ DFTI_UNCOMMITTED }; DftiGetValue(mklcpu_desc[1], DFTI_COMMIT_STATUS, &commit_status); if (commit_status != DFTI_COMMITTED) { throw math::invalid_argument("DFT", "compute_backward", - "MKLCPU DFT descriptor was not successfully committed."); + "MKLCPU DFT descriptor was not successfully committed."); } } // Throw an math::invalid_argument if the runtime param in the descriptor does not match // the expected value. template -inline auto expect_config(DescT &desc, const char *message) { +inline auto expect_config(DescT& desc, const char* message) { dft::detail::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { @@ -68,26 +68,26 @@ inline auto expect_config(DescT &desc, const char *message) { } // convert the base commit class to derived cpu commit class template -auto get_buffer(commit_t *commit_handle) { - commit_derived_t *derived_commit = - static_cast *>(commit_handle); +auto get_buffer(commit_t* commit_handle) { + commit_derived_t* derived_commit = + static_cast*>(commit_handle); return derived_commit->get_handle_buffer(); } } // namespace detail //In-place transform template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &inout) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& inout) { detail::expect_config( desc, "Unexpected value for placement"); auto commit_handle = dft::detail::get_commit(desc); detail::check_bwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - cpu_queue.submit([&](sycl::handler &cgh) { + cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); auto inout_acc = inout.template get_access(cgh); detail::host_task(cgh, [=]() { @@ -104,20 +104,20 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &inout_re, - sycl::buffer, 1> &inout_im) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& inout_re, + sycl::buffer, 1>& inout_im) { detail::expect_config( desc, "Unexpected value for complex storage"); auto commit_handle = dft::detail::get_commit(desc); detail::check_bwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - cpu_queue.submit([&](sycl::handler &cgh) { + cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); auto re_acc = inout_re.template get_access(cgh); auto im_acc = inout_im.template get_access(cgh); @@ -136,26 +136,26 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, //Out-of-place transform template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &in, - sycl::buffer, 1> &out) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& in, + sycl::buffer, 1>& out) { detail::expect_config(desc, "Unexpected value for placement"); auto commit_handle = dft::detail::get_commit(desc); detail::check_bwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - cpu_queue.submit([&](sycl::handler &cgh) { + cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); auto in_acc = in.template get_access(cgh); auto out_acc = out.template get_access(cgh); detail::host_task(cgh, [=]() { - auto in_ptr = const_cast *>(detail::acc_to_ptr(in_acc)); + auto in_ptr = const_cast*>(detail::acc_to_ptr(in_acc)); DFT_ERROR status = DftiComputeBackward(desc_acc[detail::DIR::bwd], in_ptr, detail::acc_to_ptr(out_acc)); if (status != DFTI_NO_ERROR) { @@ -169,22 +169,22 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &in_re, - sycl::buffer, 1> &in_im, - sycl::buffer, 1> &out_re, - sycl::buffer, 1> &out_im) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& in_re, + sycl::buffer, 1>& in_im, + sycl::buffer, 1>& out_re, + sycl::buffer, 1>& out_im) { detail::expect_config( desc, "Unexpected value for complex storage"); auto commit_handle = dft::detail::get_commit(desc); detail::check_bwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - cpu_queue.submit([&](sycl::handler &cgh) { + cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); auto inre_acc = in_re.template get_access(cgh); auto inim_acc = in_im.template get_access(cgh); @@ -192,8 +192,8 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, auto outim_acc = out_im.template get_access(cgh); detail::host_task(cgh, [=]() { - auto inre_ptr = const_cast *>(detail::acc_to_ptr(inre_acc)); - auto inim_ptr = const_cast *>(detail::acc_to_ptr(inim_acc)); + auto inre_ptr = const_cast*>(detail::acc_to_ptr(inre_acc)); + auto inim_ptr = const_cast*>(detail::acc_to_ptr(inim_acc)); DFT_ERROR status = DftiComputeBackward(desc_acc[detail::DIR::bwd], inre_ptr, inim_ptr, detail::acc_to_ptr(outre_acc), detail::acc_to_ptr(outim_acc)); @@ -210,18 +210,18 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, //In-place transform template -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, fwd* inout, + const std::vector& dependencies) { detail::expect_config( desc, "Unexpected value for placement"); auto commit_handle = dft::detail::get_commit(desc); detail::check_bwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - return cpu_queue.submit([&](sycl::handler &cgh) { + return cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); cgh.depends_on(dependencies); detail::host_task(cgh, [=]() { @@ -237,19 +237,20 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *inout_re, - scalar *inout_im, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, + scalar* inout_re, + scalar* inout_im, + const std::vector& dependencies) { detail::expect_config( desc, "Unexpected value for complex storage"); auto commit_handle = dft::detail::get_commit(desc); detail::check_bwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - return cpu_queue.submit([&](sycl::handler &cgh) { + return cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); cgh.depends_on(dependencies); detail::host_task(cgh, [=]() { @@ -265,9 +266,9 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, - fwd *out, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, bwd* in, + fwd* out, + const std::vector& dependencies) { // Check: inplace, complex storage detail::expect_config(desc, @@ -275,10 +276,10 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwdget_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - return cpu_queue.submit([&](sycl::handler &cgh) { + return cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); cgh.depends_on(dependencies); @@ -295,20 +296,20 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *in_re, - scalar *in_im, - scalar *out_re, - scalar *out_im, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, scalar* in_re, + scalar* in_im, + scalar* out_re, + scalar* out_im, + const std::vector& dependencies) { detail::expect_config( desc, "Unexpected value for complex storage"); auto commit_handle = dft::detail::get_commit(desc); detail::check_bwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - return cpu_queue.submit([&](sycl::handler &cgh) { + return cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); cgh.depends_on(dependencies); diff --git a/src/dft/backends/mklcpu/commit.cpp b/src/dft/backends/mklcpu/commit.cpp index ca5ecaa87..a8dca834b 100644 --- a/src/dft/backends/mklcpu/commit.cpp +++ b/src/dft/backends/mklcpu/commit.cpp @@ -125,7 +125,7 @@ void commit_derived_impl::set_value_item(mklcpu_desc_t hand, enum DFT DFT_ERROR value_err = DftiSetValue(hand, name, args...); if (value_err != DFTI_NO_ERROR) { throw oneapi::math::exception("dft/backends/mklcpu", "set_value_item", - DftiErrorMessage(value_err)); + DftiErrorMessage(value_err)); } } @@ -169,17 +169,17 @@ void commit_derived_impl::set_value(mklcpu_desc_t* descHandle, // Setting the workspace causes an FFT_INVALID_DESCRIPTOR. if (config.workspace != config_value::ALLOW) { throw math::invalid_argument("dft/backends/mklcpu", "commit", - "MKLCPU only supports workspace set to allow"); + "MKLCPU only supports workspace set to allow"); } // Setting the ordering causes an FFT_INVALID_DESCRIPTOR. Check that default is used: if (config.ordering != dft::detail::config_value::ORDERED) { throw math::invalid_argument("dft/backends/mklcpu", "commit", - "MKLCPU only supports ordered ordering."); + "MKLCPU only supports ordered ordering."); } // Setting the transpose causes an FFT_INVALID_DESCRIPTOR. Check that default is used: if (config.transpose != false) { throw math::invalid_argument("dft/backends/mklcpu", "commit", - "MKLCPU only supports non-transposed."); + "MKLCPU only supports non-transposed."); } } } diff --git a/src/dft/backends/mklcpu/forward.cpp b/src/dft/backends/mklcpu/forward.cpp index 8c8b17f85..304bb6d01 100644 --- a/src/dft/backends/mklcpu/forward.cpp +++ b/src/dft/backends/mklcpu/forward.cpp @@ -40,26 +40,26 @@ namespace detail { // BUFFER version // Forward a MKLCPU DFT call to the backend, checking that the commit impl is valid. template -inline void check_fwd_commit(dft::descriptor &desc) { +inline void check_fwd_commit(dft::descriptor& desc) { auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::mklcpu) { throw math::invalid_argument("DFT", "computer_forward", - "DFT descriptor has not been commited for MKLCPU"); + "DFT descriptor has not been commited for MKLCPU"); } - auto mklcpu_desc = reinterpret_cast(commit_handle->get_handle()); + auto mklcpu_desc = reinterpret_cast(commit_handle->get_handle()); MKL_LONG commit_status{ DFTI_UNCOMMITTED }; DftiGetValue(mklcpu_desc[0], DFTI_COMMIT_STATUS, &commit_status); if (commit_status != DFTI_COMMITTED) { throw math::invalid_argument("DFT", "compute_forward", - "MKLCPU DFT descriptor was not successfully committed."); + "MKLCPU DFT descriptor was not successfully committed."); } } // Throw an math::invalid_argument if the runtime param in the descriptor does not match // the expected value. template -inline auto expect_config(DescT &desc, const char *message) { +inline auto expect_config(DescT& desc, const char* message) { dft::detail::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { @@ -69,26 +69,26 @@ inline auto expect_config(DescT &desc, const char *message) { // convert the base commit class to derived cpu commit class template -auto get_buffer(commit_t *commit_handle) { - commit_derived_t *derived_commit = - static_cast *>(commit_handle); +auto get_buffer(commit_t* commit_handle) { + commit_derived_t* derived_commit = + static_cast*>(commit_handle); return derived_commit->get_handle_buffer(); } } // namespace detail //In-place transform template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> &inout) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& inout) { detail::expect_config( desc, "Unexpected value for placement"); auto commit_handle = dft::detail::get_commit(desc); detail::check_fwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - cpu_queue.submit([&](sycl::handler &cgh) { + cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); auto inout_acc = inout.template get_access(cgh); detail::host_task(cgh, [=]() { @@ -105,20 +105,20 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> &inout_re, - sycl::buffer, 1> &inout_im) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& inout_re, + sycl::buffer, 1>& inout_im) { detail::expect_config( desc, "Unexpected value for complex storage"); auto commit_handle = dft::detail::get_commit(desc); detail::check_fwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - cpu_queue.submit([&](sycl::handler &cgh) { + cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); auto re_acc = inout_re.template get_access(cgh); auto im_acc = inout_im.template get_access(cgh); @@ -137,25 +137,26 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, //Out-of-place transform template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, - sycl::buffer, 1> &out) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& in, + sycl::buffer, 1>& out) { detail::expect_config(desc, "Unexpected value for placement"); auto commit_handle = dft::detail::get_commit(desc); detail::check_fwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - cpu_queue.submit([&](sycl::handler &cgh) { + cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); auto in_acc = in.template get_access(cgh); auto out_acc = out.template get_access(cgh); detail::host_task(cgh, [=]() { - auto in_ptr = const_cast *>(detail::acc_to_ptr(in_acc)); + auto in_ptr = const_cast*>(detail::acc_to_ptr(in_acc)); DFT_ERROR status = DftiComputeForward(desc_acc[detail::DIR::fwd], in_ptr, detail::acc_to_ptr(out_acc)); if (status != DFTI_NO_ERROR) { @@ -169,22 +170,22 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> &in_re, - sycl::buffer, 1> &in_im, - sycl::buffer, 1> &out_re, - sycl::buffer, 1> &out_im) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& in_re, + sycl::buffer, 1>& in_im, + sycl::buffer, 1>& out_re, + sycl::buffer, 1>& out_im) { detail::expect_config( desc, "Unexpected value for complex storage"); auto commit_handle = dft::detail::get_commit(desc); detail::check_fwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - cpu_queue.submit([&](sycl::handler &cgh) { + cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); auto inre_acc = in_re.template get_access(cgh); auto inim_acc = in_im.template get_access(cgh); @@ -192,8 +193,8 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, auto outim_acc = out_im.template get_access(cgh); detail::host_task(cgh, [=]() { - auto inre_ptr = const_cast *>(detail::acc_to_ptr(inre_acc)); - auto inim_ptr = const_cast *>(detail::acc_to_ptr(inim_acc)); + auto inre_ptr = const_cast*>(detail::acc_to_ptr(inre_acc)); + auto inim_ptr = const_cast*>(detail::acc_to_ptr(inim_acc)); DFT_ERROR status = DftiComputeForward(desc_acc[detail::DIR::fwd], inre_ptr, inim_ptr, detail::acc_to_ptr(outre_acc), detail::acc_to_ptr(outim_acc)); @@ -210,18 +211,18 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, //In-place transform template -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, fwd* inout, + const std::vector& dependencies) { detail::expect_config( desc, "Unexpected value for placement"); auto commit_handle = dft::detail::get_commit(desc); detail::check_fwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - return cpu_queue.submit([&](sycl::handler &cgh) { + return cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); cgh.depends_on(dependencies); @@ -238,20 +239,20 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *inout_re, - scalar *inout_im, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, scalar* inout_re, + scalar* inout_im, + const std::vector& dependencies) { detail::expect_config( desc, "Unexpected value for complex storage"); auto commit_handle = dft::detail::get_commit(desc); detail::check_fwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - return cpu_queue.submit([&](sycl::handler &cgh) { + return cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); cgh.depends_on(dependencies); @@ -268,9 +269,9 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, - bwd *out, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, fwd* in, + bwd* out, + const std::vector& dependencies) { // Check: inplace detail::expect_config(desc, @@ -278,11 +279,11 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwdget_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - return cpu_queue.submit([&](sycl::handler &cgh) { + return cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); cgh.depends_on(dependencies); @@ -299,22 +300,22 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *in_re, - scalar *in_im, - scalar *out_re, - scalar *out_im, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, scalar* in_re, + scalar* in_im, + scalar* out_re, + scalar* out_im, + const std::vector& dependencies) { detail::expect_config( desc, "Unexpected value for complex storage"); auto commit_handle = dft::detail::get_commit(desc); detail::check_fwd_commit(desc); - sycl::queue &cpu_queue{ commit_handle->get_queue() }; + sycl::queue& cpu_queue{ commit_handle->get_queue() }; auto mklcpu_desc_buffer{ detail::get_buffer(commit_handle) }; - return cpu_queue.submit([&](sycl::handler &cgh) { + return cpu_queue.submit([&](sycl::handler& cgh) { auto desc_acc = mklcpu_desc_buffer.template get_access(cgh); cgh.depends_on(dependencies); diff --git a/src/dft/backends/mklcpu/mklcpu_helpers.hpp b/src/dft/backends/mklcpu/mklcpu_helpers.hpp index e8f9ff951..c3dc4db4b 100644 --- a/src/dft/backends/mklcpu/mklcpu_helpers.hpp +++ b/src/dft/backends/mklcpu/mklcpu_helpers.hpp @@ -83,7 +83,7 @@ inline constexpr DFTI_CONFIG_PARAM to_mklcpu(dft::detail::config_param param) { case iparam::COMMIT_STATUS: return DFTI_COMMIT_STATUS; default: throw math::invalid_argument("dft", "MKLCPU descriptor set_value()", - "Invalid config param."); + "Invalid config param."); return static_cast(0); } } @@ -106,7 +106,7 @@ inline constexpr int to_mklcpu( } else { throw math::invalid_argument("dft", "MKLCPU descriptor set_value()", - "Invalid config value for complex storage."); + "Invalid config value for complex storage."); return 0; } } @@ -119,7 +119,7 @@ inline constexpr int to_mklcpu( } else { throw math::invalid_argument("dft", "MKLCPU descriptor set_value()", - "Invalid config value for real storage."); + "Invalid config value for real storage."); return 0; } } @@ -131,7 +131,7 @@ inline constexpr int to_mklcpu( } else { throw math::invalid_argument("dft", "MKLCPU descriptor set_value()", - "Invalid config value for inplace."); + "Invalid config value for inplace."); return 0; } } @@ -160,7 +160,7 @@ inline constexpr int to_mklcpu( } else { throw math::invalid_argument("dft", "MKLCPU descriptor set_value()", - "Invalid config value for packed format."); + "Invalid config value for packed format."); return 0; } } diff --git a/src/dft/backends/mklgpu/backward.cpp b/src/dft/backends/mklgpu/backward.cpp index f9d315d2f..c9a976ff0 100644 --- a/src/dft/backends/mklgpu/backward.cpp +++ b/src/dft/backends/mklgpu/backward.cpp @@ -45,33 +45,34 @@ namespace detail { /// Forward a MKLGPU DFT call to the backend, checking that the commit impl is valid. /// Assumes backend descriptor values match those of the frontend. template -inline auto compute_backward(dft::detail::descriptor &desc, ArgTs &&... args) { +inline auto compute_backward(dft::detail::descriptor& desc, ArgTs&&... args) { using mklgpu_desc_t = oneapi::mkl::dft::descriptor; using desc_shptr_t = std::shared_ptr; using handle_t = std::pair; auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::mklgpu) { throw math::invalid_argument("DFT", "compute_backward", - "DFT descriptor has not been commited for MKLGPU"); + "DFT descriptor has not been commited for MKLGPU"); } - auto handle = reinterpret_cast(commit_handle->get_handle()); + auto handle = reinterpret_cast(commit_handle->get_handle()); auto mklgpu_desc = handle->second; // Second because backward DFT. int commit_status{ DFTI_UNCOMMITTED }; mklgpu_desc->get_value(oneapi::mkl::dft::config_param::COMMIT_STATUS, &commit_status); if (commit_status != DFTI_COMMITTED) { throw math::invalid_argument("DFT", "compute_backward", - "MKLGPU DFT descriptor was not successfully committed."); + "MKLGPU DFT descriptor was not successfully committed."); } // The MKLGPU backend's interface contains fewer function signatures than in this // open-source library. Consequently, it is not required to forward template arguments // to resolve to the correct function. - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::dft::compute_backward(*mklgpu_desc, std::forward(args)...)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + oneapi::mkl::dft::compute_backward(*mklgpu_desc, std::forward(args)...)); } /// Throw an math::invalid_argument if the runtime param in the descriptor does not match /// the expected value. template -inline auto expect_config(DescT &desc, const char *message) { +inline auto expect_config(DescT& desc, const char* message) { dft::detail::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { @@ -84,8 +85,8 @@ inline auto expect_config(DescT &desc, const char *message) { //In-place transform template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &inout) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& inout) { detail::expect_config( desc, "Unexpected value for placement"); return detail::compute_backward(desc, inout); @@ -93,18 +94,19 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_backward(descriptor_type & /*desc*/, - sycl::buffer, 1> & /*inout_re*/, - sycl::buffer, 1> & /*inout_im*/) { - throw math::unimplemented("DFT", "compute_backward", - "MKLGPU does not support compute_backward(desc, inout_re, inout_im)."); +ONEMATH_EXPORT void compute_backward(descriptor_type& /*desc*/, + sycl::buffer, 1>& /*inout_re*/, + sycl::buffer, 1>& /*inout_im*/) { + throw math::unimplemented( + "DFT", "compute_backward", + "MKLGPU does not support compute_backward(desc, inout_re, inout_im)."); } //Out-of-place transform template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &in, - sycl::buffer, 1> &out) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& in, + sycl::buffer, 1>& out) { detail::expect_config(desc, "Unexpected value for placement"); @@ -113,11 +115,11 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> & /*in_re*/, - sycl::buffer, 1> & /*in_im*/, - sycl::buffer, 1> & /*out_re*/, - sycl::buffer, 1> & /*out_im*/) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& /*in_re*/, + sycl::buffer, 1>& /*in_im*/, + sycl::buffer, 1>& /*out_re*/, + sycl::buffer, 1>& /*out_im*/) { detail::expect_config( desc, "Unexpected value for complex storage"); @@ -130,8 +132,8 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, //In-place transform template -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, fwd* inout, + const std::vector& dependencies) { detail::expect_config( desc, "Unexpected value for placement"); return detail::compute_backward(desc, inout, dependencies); @@ -139,10 +141,10 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type & /*desc*/, - scalar * /*inout_re*/, - scalar * /*inout_im*/, - const std::vector & /*dependencies*/) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& /*desc*/, + scalar* /*inout_re*/, + scalar* /*inout_im*/, + const std::vector& /*dependencies*/) { throw math::unimplemented( "DFT", "compute_backward", "MKLGPU does not support compute_backward(desc, inout_re, inout_im, dependencies)."); @@ -150,9 +152,9 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type & /*desc*/, //Out-of-place transform template -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, - fwd *out, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, bwd* in, + fwd* out, + const std::vector& dependencies) { detail::expect_config(desc, "Unexpected value for placement"); @@ -161,12 +163,12 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, - scalar * /*in_re*/, - scalar * /*in_im*/, - scalar * /*out_re*/, - scalar * /*out_im*/, - const std::vector & /*dependencies*/) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, + scalar* /*in_re*/, + scalar* /*in_im*/, + scalar* /*out_re*/, + scalar* /*out_im*/, + const std::vector& /*dependencies*/) { detail::expect_config( desc, "Unexpected value for complex storage"); diff --git a/src/dft/backends/mklgpu/commit.cpp b/src/dft/backends/mklgpu/commit.cpp index 9742baded..e2550f828 100644 --- a/src/dft/backends/mklgpu/commit.cpp +++ b/src/dft/backends/mklgpu/commit.cpp @@ -79,14 +79,14 @@ class mklgpu_commit final : public dft::detail::commit_impl { public: mklgpu_commit(sycl::queue queue, const dft::detail::dft_values& config_values) : oneapi::math::dft::detail::commit_impl(queue, backend::mklgpu, - config_values), + config_values), handle(std::make_shared(config_values.dimensions), nullptr) { handle.second = handle.first; // Make sure the bwd pointer is valid. // MKLGPU does not throw an informative exception for the following: if constexpr (prec == dft::detail::precision::DOUBLE) { if (!queue.get_device().has(sycl::aspect::fp64)) { throw math::exception("dft/backends/mklgpu", "commit", - "Device does not support double precision."); + "Device does not support double precision."); } } } @@ -163,7 +163,7 @@ class mklgpu_commit final : public dft::detail::commit_impl { to_mklgpu(config.complex_storage)); if (config.real_storage != dft::detail::config_value::REAL_REAL) { throw math::invalid_argument("dft/backends/mklgpu", "commit", - "MKLGPU only supports real-real real storage."); + "MKLGPU only supports real-real real storage."); } desc.set_value(backend_param::CONJUGATE_EVEN_STORAGE, to_mklgpu(config.conj_even_storage)); @@ -173,7 +173,7 @@ class mklgpu_commit final : public dft::detail::commit_impl { if (stride_choice == dft::detail::stride_api::FB_STRIDES) { if (config.fwd_strides[0] != 0 || config.fwd_strides[0] != 0) { throw math::unimplemented("dft/backends/mklgpu", "commit", - "MKLGPU does not support nonzero offsets."); + "MKLGPU does not support nonzero offsets."); } desc.set_value(backend_param::FWD_STRIDES, config.fwd_strides.data()); desc.set_value(backend_param::BWD_STRIDES, config.bwd_strides.data()); @@ -181,7 +181,7 @@ class mklgpu_commit final : public dft::detail::commit_impl { else { if (config.input_strides[0] != 0 || config.output_strides[0] != 0) { throw math::unimplemented("dft/backends/mklgpu", "commit", - "MKLGPU does not support nonzero offsets."); + "MKLGPU does not support nonzero offsets."); } if (assume_fwd_dft) { desc.set_value(backend_param::FWD_STRIDES, config.input_strides.data()); @@ -203,12 +203,12 @@ class mklgpu_commit final : public dft::detail::commit_impl { // Setting the ordering causes an FFT_INVALID_DESCRIPTOR. Check that default is used: if (config.ordering != dft::detail::config_value::ORDERED) { throw math::invalid_argument("dft/backends/mklgpu", "commit", - "MKLGPU only supports ordered ordering."); + "MKLGPU only supports ordered ordering."); } // Setting the transpose causes an FFT_INVALID_DESCRIPTOR. Check that default is used: if (config.transpose != false) { throw math::invalid_argument("dft/backends/mklgpu", "commit", - "MKLGPU only supports non-transposed."); + "MKLGPU only supports non-transposed."); } } diff --git a/src/dft/backends/mklgpu/forward.cpp b/src/dft/backends/mklgpu/forward.cpp index 3ff8e9c9b..4f6522614 100644 --- a/src/dft/backends/mklgpu/forward.cpp +++ b/src/dft/backends/mklgpu/forward.cpp @@ -51,33 +51,34 @@ namespace detail { /// Forward a MKLGPU DFT call to the backend, checking that the commit impl is valid. /// Assumes backend descriptor values match those of the frontend. template -inline auto compute_forward(dft::detail::descriptor &desc, ArgTs &&... args) { +inline auto compute_forward(dft::detail::descriptor& desc, ArgTs&&... args) { using mklgpu_desc_t = oneapi::mkl::dft::descriptor; using desc_shptr_t = std::shared_ptr; using handle_t = std::pair; auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::mklgpu) { throw math::invalid_argument("DFT", "compute_forward", - "DFT descriptor has not been commited for MKLGPU"); + "DFT descriptor has not been commited for MKLGPU"); } - auto handle = reinterpret_cast(commit_handle->get_handle()); + auto handle = reinterpret_cast(commit_handle->get_handle()); auto mklgpu_desc = handle->first; // First because forward DFT. int commit_status{ DFTI_UNCOMMITTED }; mklgpu_desc->get_value(oneapi::mkl::dft::config_param::COMMIT_STATUS, &commit_status); if (commit_status != DFTI_COMMITTED) { throw math::invalid_argument("DFT", "compute_forward", - "MKLGPU DFT descriptor was not successfully committed."); + "MKLGPU DFT descriptor was not successfully committed."); } // The MKLGPU backend's interface contains fewer function signatures than in this // open-source library. Consequently, it is not required to forward template arguments // to resolve to the correct function. - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::dft::compute_forward(*mklgpu_desc, std::forward(args)...)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + oneapi::mkl::dft::compute_forward(*mklgpu_desc, std::forward(args)...)); } /// Throw an math::invalid_argument if the runtime param in the descriptor does not match /// the expected value. template -inline auto expect_config(DescT &desc, const char *message) { +inline auto expect_config(DescT& desc, const char* message) { dft::detail::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { @@ -90,8 +91,8 @@ inline auto expect_config(DescT &desc, const char *message) { //In-place transform template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> &inout) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& inout) { detail::expect_config( desc, "Unexpected value for placement"); return detail::compute_forward(desc, inout); @@ -99,17 +100,18 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_forward(descriptor_type & /*desc*/, - sycl::buffer, 1> & /*inout_re*/, - sycl::buffer, 1> & /*inout_im*/) { +ONEMATH_EXPORT void compute_forward(descriptor_type& /*desc*/, + sycl::buffer, 1>& /*inout_re*/, + sycl::buffer, 1>& /*inout_im*/) { throw math::unimplemented("DFT", "compute_forward", - "MKLGPU does not support compute_forward(desc, inout_re, inout_im)."); + "MKLGPU does not support compute_forward(desc, inout_re, inout_im)."); } //Out-of-place transform template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, - sycl::buffer, 1> &out) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& in, + sycl::buffer, 1>& out) { detail::expect_config(desc, "Unexpected value for placement"); @@ -118,11 +120,11 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> & /*in_re*/, - sycl::buffer, 1> & /*in_im*/, - sycl::buffer, 1> & /*out_re*/, - sycl::buffer, 1> & /*out_im*/) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& /*in_re*/, + sycl::buffer, 1>& /*in_im*/, + sycl::buffer, 1>& /*out_re*/, + sycl::buffer, 1>& /*out_im*/) { detail::expect_config( desc, "Unexpected value for complex storage"); @@ -135,8 +137,8 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, //In-place transform template -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, fwd* inout, + const std::vector& dependencies) { detail::expect_config( desc, "Unexpected value for placement"); return detail::compute_forward(desc, inout, dependencies); @@ -144,10 +146,10 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type & /*desc*/, - scalar * /*inout_re*/, - scalar * /*inout_im*/, - const std::vector & /*dependencies*/) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& /*desc*/, + scalar* /*inout_re*/, + scalar* /*inout_im*/, + const std::vector& /*dependencies*/) { throw math::unimplemented( "DFT", "compute_forward", "MKLGPU does not support compute_forward(desc, inout_re, inout_im, dependencies)."); @@ -155,9 +157,9 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type & /*desc*/, //Out-of-place transform template -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, - bwd *out, - const std::vector &dependencies) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, fwd* in, + bwd* out, + const std::vector& dependencies) { detail::expect_config(desc, "Unexpected value for placement"); @@ -166,12 +168,12 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, - scalar * /*in_re*/, - scalar * /*in_im*/, - scalar * /*out_re*/, - scalar * /*out_im*/, - const std::vector & /*dependencies*/) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, + scalar* /*in_re*/, + scalar* /*in_im*/, + scalar* /*out_re*/, + scalar* /*out_im*/, + const std::vector& /*dependencies*/) { detail::expect_config( desc, "Unexpected value for complex storage"); diff --git a/src/dft/backends/mklgpu/mklgpu_helpers.hpp b/src/dft/backends/mklgpu/mklgpu_helpers.hpp index 0687b26e7..0e440f5cf 100644 --- a/src/dft/backends/mklgpu/mklgpu_helpers.hpp +++ b/src/dft/backends/mklgpu/mklgpu_helpers.hpp @@ -79,7 +79,7 @@ inline constexpr oneapi::mkl::dft::config_param to_mklgpu(dft::detail::config_pa case iparam::COMMIT_STATUS: return oparam::COMMIT_STATUS; default: throw math::invalid_argument("dft", "MKLGPU descriptor set_value()", - "Invalid config param."); + "Invalid config param."); return static_cast(0); } } @@ -99,7 +99,7 @@ inline constexpr int to_mklgpu( } else { throw math::unimplemented("dft", "MKLGPU descriptor set_value()", - "MKLGPU only supports complex-complex for complex storage."); + "MKLGPU only supports complex-complex for complex storage."); return 0; } } @@ -112,7 +112,7 @@ inline constexpr int to_mklgpu( } else { throw math::invalid_argument("dft", "MKLGPU descriptor set_value()", - "Invalid config value for inplace."); + "Invalid config value for inplace."); return 0; } } @@ -141,7 +141,7 @@ inline constexpr int to_mklgpu( } else { throw math::invalid_argument("dft", "MKLGPU descriptor set_value()", - "Invalid config value for packed format."); + "Invalid config value for packed format."); return 0; } } @@ -151,7 +151,8 @@ inline constexpr int to_mklgpu( * @param value The config value to convert. **/ template -inline constexpr oneapi::mkl::dft::config_value to_mklgpu_config_value(dft::detail::config_value value); +inline constexpr oneapi::mkl::dft::config_value to_mklgpu_config_value( + dft::detail::config_value value); template <> inline constexpr oneapi::mkl::dft::config_value @@ -166,7 +167,7 @@ to_mklgpu_config_value( } else { throw math::invalid_argument("dft", "MKLGPU descriptor set_value()", - "Invalid config value for workspace placement."); + "Invalid config value for workspace placement."); return oneapi::mkl::dft::config_value::WORKSPACE_INTERNAL; } } diff --git a/src/dft/backends/portfft/commit.cpp b/src/dft/backends/portfft/commit.cpp index f7dbcbf83..628f93d02 100644 --- a/src/dft/backends/portfft/commit.cpp +++ b/src/dft/backends/portfft/commit.cpp @@ -62,7 +62,7 @@ class portfft_commit final : public dft::detail::commit_impl { public: portfft_commit(sycl::queue& queue, const dft::detail::dft_values& config_values) : oneapi::math::dft::detail::commit_impl(queue, backend::portfft, - config_values) { + config_values) { if constexpr (prec == dft::detail::precision::DOUBLE) { if (!queue.get_device().has(sycl::aspect::fp64)) { throw math::exception("DFT", "commit", "Device does not support double precision."); @@ -78,15 +78,15 @@ class portfft_commit final : public dft::detail::commit_impl { oneapi::math::dft::detail::config_value::WORKSPACE_EXTERNAL); if (config_values.workspace != config_value::ALLOW) { throw math::unimplemented("dft/backends/portfft", __FUNCTION__, - "portFFT only supports ALLOW for the WORKSPACE parameter"); + "portFFT only supports ALLOW for the WORKSPACE parameter"); } if (config_values.ordering != config_value::ORDERED) { throw math::unimplemented("dft/backends/portfft", __FUNCTION__, - "portFFT only supports ORDERED for the ORDERING parameter"); + "portFFT only supports ORDERED for the ORDERING parameter"); } if (config_values.transpose) { throw math::unimplemented("dft/backends/portfft", __FUNCTION__, - "portFFT does not supported transposed output"); + "portFFT does not supported transposed output"); } auto stride_api_choice = dft::detail::get_stride_api(config_values); @@ -182,14 +182,14 @@ class portfft_commit final : public dft::detail::commit_impl { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_forward"); throw oneapi::math::unimplemented("DFT", "compute_forward(desc, inout_re, inout_im)", - "portFFT does not support real-real complex storage."); + "portFFT does not support real-real complex storage."); } sycl::event forward_ip_rr(descriptor_type& desc, scalar_type*, scalar_type*, const std::vector&) override { dft::detail::get_commit(desc)->template compute_call_throw("compute_forward"); throw oneapi::math::unimplemented("DFT", - "compute_forward(desc, inout_re, inout_im, dependencies)", - "portFFT does not support real-real complex storage."); + "compute_forward(desc, inout_re, inout_im, dependencies)", + "portFFT does not support real-real complex storage."); } // forward out-of-place COMPLEX_COMPLEX @@ -223,8 +223,8 @@ class portfft_commit final : public dft::detail::commit_impl { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_forward"); throw oneapi::math::unimplemented("DFT", - "compute_forward(desc, in_re, in_im, out_re, out_im)", - "portFFT does not support real-real complex storage."); + "compute_forward(desc, in_re, in_im, out_re, out_im)", + "portFFT does not support real-real complex storage."); } sycl::event forward_op_rr(descriptor_type& desc, scalar_type*, scalar_type*, scalar_type*, scalar_type*, const std::vector&) override { @@ -263,15 +263,15 @@ class portfft_commit final : public dft::detail::commit_impl { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_backward"); throw oneapi::math::unimplemented("DFT", "compute_backward(desc, inout_re, inout_im)", - "portFFT does not support real-real complex storage."); + "portFFT does not support real-real complex storage."); } sycl::event backward_ip_rr(descriptor_type& desc, scalar_type*, scalar_type*, const std::vector&) override { dft::detail::get_commit(desc)->template compute_call_throw( "compute_backward"); - throw oneapi::math::unimplemented("DFT", - "compute_backward(desc, inout_re, inout_im, dependencies)", - "portFFT does not support real-real complex storage."); + throw oneapi::math::unimplemented( + "DFT", "compute_backward(desc, inout_re, inout_im, dependencies)", + "portFFT does not support real-real complex storage."); } // backward out-of-place COMPLEX_COMPLEX @@ -305,8 +305,8 @@ class portfft_commit final : public dft::detail::commit_impl { dft::detail::get_commit(desc)->template compute_call_throw>( "compute_backward"); throw oneapi::math::unimplemented("DFT", - "compute_backward(desc, in_re, in_im, out_re, out_im)", - "portFFT does not support real-real complex storage."); + "compute_backward(desc, in_re, in_im, out_re, out_im)", + "portFFT does not support real-real complex storage."); } sycl::event backward_op_rr(descriptor_type& desc, scalar_type*, scalar_type*, scalar_type*, scalar_type*, const std::vector&) override { diff --git a/src/dft/backends/portfft/portfft_helper.hpp b/src/dft/backends/portfft/portfft_helper.hpp index c4270cd3c..9900d6feb 100644 --- a/src/dft/backends/portfft/portfft_helper.hpp +++ b/src/dft/backends/portfft/portfft_helper.hpp @@ -31,12 +31,12 @@ namespace pfft = portfft; namespace oneapi::math::dft::portfft::detail { template -inline dft::detail::commit_impl *checked_get_commit( - dft::detail::descriptor &desc) { +inline dft::detail::commit_impl* checked_get_commit( + dft::detail::descriptor& desc) { auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::portfft) { throw math::invalid_argument("dft/backends/portfft", "get_commit", - "DFT descriptor has not been commited for portFFT"); + "DFT descriptor has not been commited for portFFT"); } return commit_handle; } @@ -53,9 +53,9 @@ using storage_type = detail::to_pfft_domain::type::value>>; template -auto get_descriptors(descriptor_type &desc) { +auto get_descriptors(descriptor_type& desc) { auto commit = detail::checked_get_commit(desc); - return reinterpret_cast *>(commit->get_handle()); + return reinterpret_cast*>(commit->get_handle()); } } // namespace oneapi::math::dft::portfft::detail diff --git a/src/dft/backends/rocfft/backward.cpp b/src/dft/backends/rocfft/backward.cpp index d0f47bd00..745ff666d 100644 --- a/src/dft/backends/rocfft/backward.cpp +++ b/src/dft/backends/rocfft/backward.cpp @@ -39,24 +39,24 @@ namespace oneapi::math::dft::rocfft { namespace detail { //forward declaration template -std::array get_offsets_bwd(dft::detail::commit_impl *commit); +std::array get_offsets_bwd(dft::detail::commit_impl* commit); template -rocfft_plan get_bwd_plan(dft::detail::commit_impl *commit) { - return static_cast(commit->get_handle())[1].plan.value(); +rocfft_plan get_bwd_plan(dft::detail::commit_impl* commit) { + return static_cast(commit->get_handle())[1].plan.value(); } template -rocfft_execution_info get_bwd_info(dft::detail::commit_impl *commit) { - return static_cast(commit->get_handle())[1].info.value(); +rocfft_execution_info get_bwd_info(dft::detail::commit_impl* commit) { + return static_cast(commit->get_handle())[1].info.value(); } } // namespace detail // BUFFER version //In-place transform template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &inout) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& inout) { const std::string func_name = "compute_backward(desc, inout)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -75,26 +75,26 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, "rocFFT requires input and output offsets (first value in strides) to be equal for in-place transforms!"); } - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto inout_acc = inout.template get_access(cgh); commit->add_buffer_workspace_dependency_if_rqd("compute_backward", cgh); dft::detail::fft_enqueue_task(cgh, [=](sycl::interop_handle ih) { auto stream = detail::setup_stream(func_name, ih, info); - auto inout_native = reinterpret_cast( - reinterpret_cast *>(detail::native_mem(ih, inout_acc)) + + auto inout_native = reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, inout_acc)) + offsets[0]); - detail::execute_checked(func_name, stream, plan, &inout_native, nullptr, info); + detail::execute_checked(func_name, stream, plan, &inout_native, nullptr, info); }); }); } //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &inout_re, - sycl::buffer, 1> &inout_im) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& inout_re, + sycl::buffer, 1>& inout_im) { const std::string func_name = "compute_backward(desc, inout_re, inout_im)"; auto commit = detail::checked_get_commit(desc); auto queue = commit->get_queue(); @@ -108,7 +108,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, "rocFFT requires input and output offsets (first value in strides) to be equal for in-place transforms!"); } - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto inout_re_acc = inout_re.template get_access(cgh); auto inout_im_acc = inout_im.template get_access(cgh); commit->add_buffer_workspace_dependency_if_rqd("compute_backward", cgh); @@ -116,24 +116,24 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, dft::detail::fft_enqueue_task(cgh, [=](sycl::interop_handle ih) { auto stream = detail::setup_stream(func_name, ih, info); - std::array inout_native{ - reinterpret_cast(reinterpret_cast *>( - detail::native_mem(ih, inout_re_acc)) + - offsets[0]), - reinterpret_cast(reinterpret_cast *>( - detail::native_mem(ih, inout_im_acc)) + - offsets[0]) + std::array inout_native{ + reinterpret_cast(reinterpret_cast*>( + detail::native_mem(ih, inout_re_acc)) + + offsets[0]), + reinterpret_cast(reinterpret_cast*>( + detail::native_mem(ih, inout_im_acc)) + + offsets[0]) }; - detail::execute_checked(func_name, stream, plan, inout_native.data(), nullptr, info); + detail::execute_checked(func_name, stream, plan, inout_native.data(), nullptr, info); }); }); } //Out-of-place transform template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &in, - sycl::buffer, 1> &out) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& in, + sycl::buffer, 1>& out) { detail::expect_config( desc, "Unexpected value for placement"); auto commit = detail::checked_get_commit(desc); @@ -142,7 +142,7 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, auto info = detail::get_bwd_info(commit); auto offsets = detail::get_offsets_bwd(commit); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto in_acc = in.template get_access(cgh); auto out_acc = out.template get_access(cgh); commit->add_buffer_workspace_dependency_if_rqd("compute_backward", cgh); @@ -151,31 +151,31 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, const std::string func_name = "compute_backward(desc, in, out)"; auto stream = detail::setup_stream(func_name, ih, info); - auto in_native = reinterpret_cast( - reinterpret_cast *>(detail::native_mem(ih, in_acc)) + + auto in_native = reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, in_acc)) + offsets[0]); - auto out_native = reinterpret_cast( - reinterpret_cast *>(detail::native_mem(ih, out_acc)) + + auto out_native = reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, out_acc)) + offsets[1]); - detail::execute_checked(func_name, stream, plan, &in_native, &out_native, info); + detail::execute_checked(func_name, stream, plan, &in_native, &out_native, info); }); }); } //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_backward(descriptor_type &desc, - sycl::buffer, 1> &in_re, - sycl::buffer, 1> &in_im, - sycl::buffer, 1> &out_re, - sycl::buffer, 1> &out_im) { +ONEMATH_EXPORT void compute_backward(descriptor_type& desc, + sycl::buffer, 1>& in_re, + sycl::buffer, 1>& in_im, + sycl::buffer, 1>& out_re, + sycl::buffer, 1>& out_im) { auto commit = detail::checked_get_commit(desc); auto queue = commit->get_queue(); auto plan = detail::get_bwd_plan(commit); auto info = detail::get_bwd_info(commit); auto offsets = detail::get_offsets_bwd(commit); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto in_re_acc = in_re.template get_access(cgh); auto in_im_acc = in_im.template get_access(cgh); auto out_re_acc = out_re.template get_access(cgh); @@ -186,23 +186,24 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, const std::string func_name = "compute_backward(desc, in_re, in_im, out_re, out_im)"; auto stream = detail::setup_stream(func_name, ih, info); - std::array in_native{ - reinterpret_cast( - reinterpret_cast *>(detail::native_mem(ih, in_re_acc)) + + std::array in_native{ + reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, in_re_acc)) + offsets[0]), - reinterpret_cast( - reinterpret_cast *>(detail::native_mem(ih, in_im_acc)) + + reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, in_im_acc)) + offsets[0]) }; - std::array out_native{ - reinterpret_cast(reinterpret_cast *>( - detail::native_mem(ih, out_re_acc)) + - offsets[1]), - reinterpret_cast(reinterpret_cast *>( - detail::native_mem(ih, out_im_acc)) + - offsets[1]) + std::array out_native{ + reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, out_re_acc)) + + offsets[1]), + reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, out_im_acc)) + + offsets[1]) }; - detail::execute_checked(func_name, stream, plan, in_native.data(), out_native.data(), info); + detail::execute_checked(func_name, stream, plan, in_native.data(), out_native.data(), + info); }); }); } @@ -211,8 +212,8 @@ ONEMATH_EXPORT void compute_backward(descriptor_type &desc, //In-place transform template -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd *inout, - const std::vector &deps) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, fwd* inout, + const std::vector& deps) { const std::string func_name = "compute_backward(desc, inout, deps)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -232,15 +233,15 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwddepend_on_last_usm_workspace_event_if_rqd(cgh); dft::detail::fft_enqueue_task(cgh, [=](sycl::interop_handle ih) { auto stream = detail::setup_stream(func_name, ih, info); - void *inout_ptr = inout; - detail::execute_checked(func_name, stream, plan, &inout_ptr, nullptr, info); + void* inout_ptr = inout; + detail::execute_checked(func_name, stream, plan, &inout_ptr, nullptr, info); }); }); commit->set_last_usm_workspace_event_if_rqd(sycl_event); @@ -249,9 +250,10 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *inout_re, - scalar *inout_im, - const std::vector &deps) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, + scalar* inout_re, + scalar* inout_im, + const std::vector& deps) { const std::string func_name = "compute_backward(desc, inout_re, inout_im, deps)"; auto commit = detail::checked_get_commit(desc); auto queue = commit->get_queue(); @@ -265,16 +267,15 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalardepend_on_last_usm_workspace_event_if_rqd(cgh); dft::detail::fft_enqueue_task(cgh, [=](sycl::interop_handle ih) { auto stream = detail::setup_stream(func_name, ih, info); - std::array inout_native{ inout_re + offsets[0], inout_im + offsets[0] }; - detail::execute_checked(func_name, stream, plan, inout_native.data(), nullptr, info); - + std::array inout_native{ inout_re + offsets[0], inout_im + offsets[0] }; + detail::execute_checked(func_name, stream, plan, inout_native.data(), nullptr, info); }); }); commit->set_last_usm_workspace_event_if_rqd(sycl_event); @@ -283,9 +284,9 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd *in, - fwd *out, - const std::vector &deps) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, bwd* in, + fwd* out, + const std::vector& deps) { detail::expect_config( desc, "Unexpected value for placement"); auto commit = detail::checked_get_commit(desc); @@ -297,7 +298,7 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwddepend_on_last_usm_workspace_event_if_rqd(cgh); @@ -305,9 +306,9 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwdset_last_usm_workspace_event_if_rqd(sycl_event); @@ -316,18 +317,18 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, bwd -ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar *in_re, - scalar *in_im, - scalar *out_re, - scalar *out_im, - const std::vector &deps) { +ONEMATH_EXPORT sycl::event compute_backward(descriptor_type& desc, scalar* in_re, + scalar* in_im, + scalar* out_re, + scalar* out_im, + const std::vector& deps) { auto commit = detail::checked_get_commit(desc); auto queue = commit->get_queue(); auto plan = detail::get_bwd_plan(commit); auto info = detail::get_bwd_info(commit); auto offsets = detail::get_offsets_bwd(commit); - sycl::event sycl_event = queue.submit([&](sycl::handler &cgh) { + sycl::event sycl_event = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(deps); commit->depend_on_last_usm_workspace_event_if_rqd(cgh); @@ -336,9 +337,10 @@ ONEMATH_EXPORT sycl::event compute_backward(descriptor_type &desc, scalar in_native{ in_re + offsets[0], in_im + offsets[0] }; - std::array out_native{ out_re + offsets[1], out_im + offsets[1] }; - detail::execute_checked(func_name, stream, plan, in_native.data(), out_native.data(), info); + std::array in_native{ in_re + offsets[0], in_im + offsets[0] }; + std::array out_native{ out_re + offsets[1], out_im + offsets[1] }; + detail::execute_checked(func_name, stream, plan, in_native.data(), out_native.data(), + info); }); }); commit->set_last_usm_workspace_event_if_rqd(sycl_event); diff --git a/src/dft/backends/rocfft/commit.cpp b/src/dft/backends/rocfft/commit.cpp index d650251e6..4c5d51d2f 100644 --- a/src/dft/backends/rocfft/commit.cpp +++ b/src/dft/backends/rocfft/commit.cpp @@ -94,7 +94,7 @@ class rocfft_commit final : public dft::detail::commit_impl { public: rocfft_commit(sycl::queue& queue, const dft::detail::dft_values& config_values) : oneapi::math::dft::detail::commit_impl(queue, backend::rocfft, - config_values) { + config_values) { if constexpr (prec == dft::detail::precision::DOUBLE) { if (!queue.get_device().has(sycl::aspect::fp64)) { throw math::exception("DFT", "commit", "Device does not support double precision."); @@ -108,14 +108,14 @@ class rocfft_commit final : public dft::detail::commit_impl { if (handles[0].plan) { if (rocfft_plan_destroy(handles[0].plan.value()) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to destroy forward plan."); + "Failed to destroy forward plan."); } handles[0].plan = std::nullopt; } if (handles[1].plan) { if (rocfft_plan_destroy(handles[1].plan.value()) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to destroy backward plan."); + "Failed to destroy backward plan."); } handles[1].plan = std::nullopt; } @@ -123,14 +123,14 @@ class rocfft_commit final : public dft::detail::commit_impl { if (handles[0].info) { if (rocfft_execution_info_destroy(handles[0].info.value()) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to destroy forward execution info ."); + "Failed to destroy forward execution info ."); } handles[0].info = std::nullopt; } if (handles[1].info) { if (rocfft_execution_info_destroy(handles[1].info.value()) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to destroy backward execution info ."); + "Failed to destroy backward execution info ."); } handles[1].info = std::nullopt; } @@ -280,17 +280,17 @@ class rocfft_commit final : public dft::detail::commit_impl { rocfft_plan_description plan_desc_fwd, plan_desc_bwd; // Can't reuse with ROCm 6 due to bug. if (rocfft_plan_description_create(&plan_desc_fwd) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to create plan description."); + "Failed to create plan description."); } if (rocfft_plan_description_create(&plan_desc_bwd) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to create plan description."); + "Failed to create plan description."); } // plan_description can be destroyed afted plan_create auto description_destroy = [](rocfft_plan_description p) { if (rocfft_plan_description_destroy(p) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to destroy plan description."); + "Failed to destroy plan description."); } }; std::unique_ptr @@ -359,13 +359,13 @@ class rocfft_commit final : public dft::detail::commit_impl { ); if (res != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to set forward data layout."); + "Failed to set forward data layout."); } if (rocfft_plan_description_set_scale_factor(plan_desc_fwd, config_values.fwd_scale) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to set forward scale factor."); + "Failed to set forward scale factor."); } rocfft_plan fwd_plan; @@ -374,7 +374,7 @@ class rocfft_commit final : public dft::detail::commit_impl { if (res != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to create forward plan."); + "Failed to create forward plan."); } handles[0].plan = fwd_plan; @@ -382,7 +382,7 @@ class rocfft_commit final : public dft::detail::commit_impl { rocfft_execution_info fwd_info; if (rocfft_execution_info_create(&fwd_info) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to create forward execution info."); + "Failed to create forward execution info."); } handles[0].info = fwd_info; @@ -392,7 +392,7 @@ class rocfft_commit final : public dft::detail::commit_impl { void* work_buf; if (hipMalloc(&work_buf, work_buf_size) != hipSuccess) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to get allocate forward work buffer."); + "Failed to get allocate forward work buffer."); } set_workspace_impl(handles[0], reinterpret_cast(work_buf), work_buf_size, "commit"); @@ -415,13 +415,13 @@ class rocfft_commit final : public dft::detail::commit_impl { ); if (res != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to set backward data layout."); + "Failed to set backward data layout."); } if (rocfft_plan_description_set_scale_factor(plan_desc_bwd, config_values.bwd_scale) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to set backward scale factor."); + "Failed to set backward scale factor."); } rocfft_plan bwd_plan; @@ -429,14 +429,14 @@ class rocfft_commit final : public dft::detail::commit_impl { lengths.data(), number_of_transforms, plan_desc_bwd); if (res != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to create backward rocFFT plan."); + "Failed to create backward rocFFT plan."); } handles[1].plan = bwd_plan; rocfft_execution_info bwd_info; if (rocfft_execution_info_create(&bwd_info) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to create backward execution info."); + "Failed to create backward execution info."); } handles[1].info = bwd_info; @@ -446,7 +446,7 @@ class rocfft_commit final : public dft::detail::commit_impl { void* work_buf; if (hipMalloc(&work_buf, work_buf_size) != hipSuccess) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to get allocate backward work buffer."); + "Failed to get allocate backward work buffer."); } set_workspace_impl(handles[1], reinterpret_cast(work_buf), work_buf_size, "commit"); @@ -489,7 +489,7 @@ class rocfft_commit final : public dft::detail::commit_impl { std::size_t size = 0; if (rocfft_plan_get_work_buffer_size(*handle.plan, &size) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", function, - "Failed to get rocfft work buffer size."); + "Failed to get rocfft work buffer size."); } return static_cast(size); } @@ -530,7 +530,7 @@ class rocfft_commit final : public dft::detail::commit_impl { if (handle.buffer) { if (hipFree(*handle.buffer) != hipSuccess) { throw math::exception("dft/backends/rocfft", function, - "Failed to free internal buffer."); + "Failed to free internal buffer."); } handle.buffer = std::nullopt; } @@ -584,12 +584,12 @@ class rocfft_commit final : public dft::detail::commit_impl { // plan work buffer if (plan == nullptr) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Missing internal rocFFT plan."); + "Missing internal rocFFT plan."); } std::size_t work_buf_size; if (rocfft_plan_get_work_buffer_size(plan, &work_buf_size) != rocfft_status_success) { throw math::exception("dft/backends/rocfft", __FUNCTION__, - "Failed to get work buffer size."); + "Failed to get work buffer size."); } return static_cast(work_buf_size); } diff --git a/src/dft/backends/rocfft/execute_helper.hpp b/src/dft/backends/rocfft/execute_helper.hpp index b73bc3a32..c1ee6302b 100644 --- a/src/dft/backends/rocfft/execute_helper.hpp +++ b/src/dft/backends/rocfft/execute_helper.hpp @@ -37,12 +37,12 @@ namespace oneapi::math::dft::rocfft::detail { template -inline dft::detail::commit_impl *checked_get_commit( - dft::detail::descriptor &desc) { +inline dft::detail::commit_impl* checked_get_commit( + dft::detail::descriptor& desc) { auto commit_handle = dft::detail::get_commit(desc); if (commit_handle == nullptr || commit_handle->get_backend() != backend::rocfft) { throw math::invalid_argument("dft/backends/rocfft", "get_commit", - "DFT descriptor has not been commited for rocFFT"); + "DFT descriptor has not been commited for rocFFT"); } return commit_handle; } @@ -50,7 +50,7 @@ inline dft::detail::commit_impl *checked_get_commit( /// Throw an math::invalid_argument if the runtime param in the descriptor does not match /// the expected value. template -inline auto expect_config(DescT &desc, const char *message) { +inline auto expect_config(DescT& desc, const char* message) { dft::config_value actual{ 0 }; desc.get_value(Param, &actual); if (actual != Expected) { @@ -59,11 +59,11 @@ inline auto expect_config(DescT &desc, const char *message) { } template -inline void *native_mem(sycl::interop_handle &ih, Acc &buf) { +inline void* native_mem(sycl::interop_handle& ih, Acc& buf) { return ih.get_native_mem(buf); } -inline hipStream_t setup_stream(const std::string &func, sycl::interop_handle &ih, +inline hipStream_t setup_stream(const std::string& func, sycl::interop_handle& ih, rocfft_execution_info info) { auto stream = ih.get_native_queue(); auto result = rocfft_execution_info_set_stream(info, stream); @@ -75,16 +75,16 @@ inline hipStream_t setup_stream(const std::string &func, sycl::interop_handle &i return stream; } -inline void sync_checked(const std::string &func, hipStream_t stream) { - auto result = hipStreamSynchronize(stream); - if (result != hipSuccess) { - throw oneapi::math::exception("dft/backends/rocfft", func, - "hipStreamSynchronize returned " + std::to_string(result)); - } +inline void sync_checked(const std::string& func, hipStream_t stream) { + auto result = hipStreamSynchronize(stream); + if (result != hipSuccess) { + throw oneapi::math::exception("dft/backends/rocfft", func, + "hipStreamSynchronize returned " + std::to_string(result)); + } } -inline void execute_checked(const std::string &func, hipStream_t stream, const rocfft_plan plan, void *in_buffer[], - void *out_buffer[], rocfft_execution_info info) { +inline void execute_checked(const std::string& func, hipStream_t stream, const rocfft_plan plan, + void* in_buffer[], void* out_buffer[], rocfft_execution_info info) { auto result = rocfft_execute(plan, in_buffer, out_buffer, info); if (result != rocfft_status_success) { throw oneapi::math::exception("dft/backends/rocfft", func, diff --git a/src/dft/backends/rocfft/forward.cpp b/src/dft/backends/rocfft/forward.cpp index dfec5fbd2..40f8834a3 100644 --- a/src/dft/backends/rocfft/forward.cpp +++ b/src/dft/backends/rocfft/forward.cpp @@ -41,16 +41,16 @@ namespace oneapi::math::dft::rocfft { namespace detail { //forward declaration template -std::array get_offsets_fwd(dft::detail::commit_impl *commit); +std::array get_offsets_fwd(dft::detail::commit_impl* commit); template -rocfft_plan get_fwd_plan(dft::detail::commit_impl *commit) { - return static_cast(commit->get_handle())[0].plan.value(); +rocfft_plan get_fwd_plan(dft::detail::commit_impl* commit) { + return static_cast(commit->get_handle())[0].plan.value(); } template -rocfft_execution_info get_fwd_info(dft::detail::commit_impl *commit) { - return static_cast(commit->get_handle())[0].info.value(); +rocfft_execution_info get_fwd_info(dft::detail::commit_impl* commit) { + return static_cast(commit->get_handle())[0].info.value(); } } // namespace detail @@ -58,8 +58,8 @@ rocfft_execution_info get_fwd_info(dft::detail::commit_impl *commit) //In-place transform template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> &inout) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& inout) { const std::string func_name = "compute_forward(desc, inout)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -78,26 +78,26 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, "rocFFT requires input and output offsets (first value in strides) to be equal for in-place transforms!"); } - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto inout_acc = inout.template get_access(cgh); commit->add_buffer_workspace_dependency_if_rqd("compute_forward", cgh); dft::detail::fft_enqueue_task(cgh, [=](sycl::interop_handle ih) { auto stream = detail::setup_stream(func_name, ih, info); - auto inout_native = reinterpret_cast( - reinterpret_cast *>(detail::native_mem(ih, inout_acc)) + + auto inout_native = reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, inout_acc)) + offsets[0]); - detail::execute_checked(func_name, stream, plan, &inout_native, nullptr, info); + detail::execute_checked(func_name, stream, plan, &inout_native, nullptr, info); }); }); } //In-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> &inout_re, - sycl::buffer, 1> &inout_im) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& inout_re, + sycl::buffer, 1>& inout_im) { const std::string func_name = "compute_forward(desc, inout_re, inout_im)"; auto commit = detail::checked_get_commit(desc); auto queue = commit->get_queue(); @@ -111,7 +111,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, "rocFFT requires input and output offsets (first value in strides) to be equal for in-place transforms!"); } - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto inout_re_acc = inout_re.template get_access(cgh); auto inout_im_acc = inout_im.template get_access(cgh); commit->add_buffer_workspace_dependency_if_rqd("compute_forward", cgh); @@ -119,23 +119,24 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, dft::detail::fft_enqueue_task(cgh, [=](sycl::interop_handle ih) { auto stream = detail::setup_stream(func_name, ih, info); - std::array inout_native{ - reinterpret_cast(reinterpret_cast *>( - detail::native_mem(ih, inout_re_acc)) + - offsets[0]), - reinterpret_cast(reinterpret_cast *>( - detail::native_mem(ih, inout_im_acc)) + - offsets[0]) + std::array inout_native{ + reinterpret_cast(reinterpret_cast*>( + detail::native_mem(ih, inout_re_acc)) + + offsets[0]), + reinterpret_cast(reinterpret_cast*>( + detail::native_mem(ih, inout_im_acc)) + + offsets[0]) }; - detail::execute_checked(func_name, stream, plan, inout_native.data(), nullptr, info); + detail::execute_checked(func_name, stream, plan, inout_native.data(), nullptr, info); }); }); } //Out-of-place transform template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer, 1> &in, - sycl::buffer, 1> &out) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& in, + sycl::buffer, 1>& out) { detail::expect_config( desc, "Unexpected value for placement"); auto commit = detail::checked_get_commit(desc); @@ -144,7 +145,7 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer(cgh); auto out_acc = out.template get_access(cgh); commit->add_buffer_workspace_dependency_if_rqd("compute_forward", cgh); @@ -153,31 +154,31 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, sycl::buffer( - reinterpret_cast *>(detail::native_mem(ih, in_acc)) + + auto in_native = reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, in_acc)) + offsets[0]); - auto out_native = reinterpret_cast( - reinterpret_cast *>(detail::native_mem(ih, out_acc)) + + auto out_native = reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, out_acc)) + offsets[1]); - detail::execute_checked(func_name, stream, plan, &in_native, &out_native, info); + detail::execute_checked(func_name, stream, plan, &in_native, &out_native, info); }); }); } //Out-of-place transform, using config_param::COMPLEX_STORAGE=config_value::REAL_REAL data format template -ONEMATH_EXPORT void compute_forward(descriptor_type &desc, - sycl::buffer, 1> &in_re, - sycl::buffer, 1> &in_im, - sycl::buffer, 1> &out_re, - sycl::buffer, 1> &out_im) { +ONEMATH_EXPORT void compute_forward(descriptor_type& desc, + sycl::buffer, 1>& in_re, + sycl::buffer, 1>& in_im, + sycl::buffer, 1>& out_re, + sycl::buffer, 1>& out_im) { auto commit = detail::checked_get_commit(desc); auto queue = commit->get_queue(); auto plan = detail::get_fwd_plan(commit); auto info = detail::get_fwd_info(commit); auto offsets = detail::get_offsets_fwd(commit); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto in_re_acc = in_re.template get_access(cgh); auto in_im_acc = in_im.template get_access(cgh); auto out_re_acc = out_re.template get_access(cgh); @@ -188,23 +189,24 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, const std::string func_name = "compute_forward(desc, in_re, in_im, out_re, out_im)"; auto stream = detail::setup_stream(func_name, ih, info); - std::array in_native{ - reinterpret_cast( - reinterpret_cast *>(detail::native_mem(ih, in_re_acc)) + + std::array in_native{ + reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, in_re_acc)) + offsets[0]), - reinterpret_cast( - reinterpret_cast *>(detail::native_mem(ih, in_im_acc)) + + reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, in_im_acc)) + offsets[0]) }; - std::array out_native{ - reinterpret_cast(reinterpret_cast *>( - detail::native_mem(ih, out_re_acc)) + - offsets[1]), - reinterpret_cast(reinterpret_cast *>( - detail::native_mem(ih, out_im_acc)) + - offsets[1]) + std::array out_native{ + reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, out_re_acc)) + + offsets[1]), + reinterpret_cast( + reinterpret_cast*>(detail::native_mem(ih, out_im_acc)) + + offsets[1]) }; - detail::execute_checked(func_name, stream, plan, in_native.data(), out_native.data(), info); + detail::execute_checked(func_name, stream, plan, in_native.data(), out_native.data(), + info); }); }); } @@ -213,8 +215,8 @@ ONEMATH_EXPORT void compute_forward(descriptor_type &desc, //In-place transform template -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *inout, - const std::vector &deps) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, fwd* inout, + const std::vector& deps) { const std::string func_name = "compute_forward(desc, inout, deps)"; detail::expect_config( desc, "Unexpected value for placement"); @@ -234,15 +236,15 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwddepend_on_last_usm_workspace_event_if_rqd(cgh); dft::detail::fft_enqueue_task(cgh, [=](sycl::interop_handle ih) { auto stream = detail::setup_stream(func_name, ih, info); - void *inout_ptr = inout; - detail::execute_checked(func_name, stream, plan, &inout_ptr, nullptr, info); + void* inout_ptr = inout; + detail::execute_checked(func_name, stream, plan, &inout_ptr, nullptr, info); }); }); commit->set_last_usm_workspace_event_if_rqd(sycl_event); @@ -251,9 +253,9 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *inout_re, - scalar *inout_im, - const std::vector &deps) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, scalar* inout_re, + scalar* inout_im, + const std::vector& deps) { const std::string func_name = "compute_forward(desc, inout_re, inout_im, deps)"; auto commit = detail::checked_get_commit(desc); auto queue = commit->get_queue(); @@ -267,14 +269,14 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalardepend_on_last_usm_workspace_event_if_rqd(cgh); dft::detail::fft_enqueue_task(cgh, [=](sycl::interop_handle ih) { auto stream = detail::setup_stream(func_name, ih, info); - std::array inout_native{ inout_re + offsets[0], inout_im + offsets[0] }; - detail::execute_checked(func_name, stream, plan, inout_native.data(), nullptr, info); + std::array inout_native{ inout_re + offsets[0], inout_im + offsets[0] }; + detail::execute_checked(func_name, stream, plan, inout_native.data(), nullptr, info); }); }); commit->set_last_usm_workspace_event_if_rqd(sycl_event); @@ -283,9 +285,9 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd *in, - bwd *out, - const std::vector &deps) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, fwd* in, + bwd* out, + const std::vector& deps) { detail::expect_config( desc, "Unexpected value for placement"); auto commit = detail::checked_get_commit(desc); @@ -297,7 +299,7 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwddepend_on_last_usm_workspace_event_if_rqd(cgh); @@ -305,9 +307,9 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwdset_last_usm_workspace_event_if_rqd(sycl_event); @@ -316,18 +318,18 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, fwd -ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar *in_re, - scalar *in_im, - scalar *out_re, - scalar *out_im, - const std::vector &deps) { +ONEMATH_EXPORT sycl::event compute_forward(descriptor_type& desc, scalar* in_re, + scalar* in_im, + scalar* out_re, + scalar* out_im, + const std::vector& deps) { auto commit = detail::checked_get_commit(desc); auto queue = commit->get_queue(); auto plan = detail::get_fwd_plan(commit); auto info = detail::get_fwd_info(commit); auto offsets = detail::get_offsets_fwd(commit); - sycl::event sycl_event = queue.submit([&](sycl::handler &cgh) { + sycl::event sycl_event = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(deps); commit->depend_on_last_usm_workspace_event_if_rqd(cgh); @@ -336,9 +338,10 @@ ONEMATH_EXPORT sycl::event compute_forward(descriptor_type &desc, scalar in_native{ in_re + offsets[0], in_im + offsets[0] }; - std::array out_native{ out_re + offsets[1], out_im + offsets[1] }; - detail::execute_checked(func_name, stream, plan, in_native.data(), out_native.data(), info); + std::array in_native{ in_re + offsets[0], in_im + offsets[0] }; + std::array out_native{ out_re + offsets[1], out_im + offsets[1] }; + detail::execute_checked(func_name, stream, plan, in_native.data(), out_native.data(), + info); }); }); commit->set_last_usm_workspace_event_if_rqd(sycl_event); diff --git a/src/dft/backends/stride_helper.hpp b/src/dft/backends/stride_helper.hpp index e00d0528e..8375b472e 100644 --- a/src/dft/backends/stride_helper.hpp +++ b/src/dft/backends/stride_helper.hpp @@ -76,7 +76,7 @@ struct stride_vectors { bwd_out(stride_choice == stride_api::FB_STRIDES ? vec_a : vec_b) { if (stride_choice == stride_api::INVALID) { throw math::exception("DFT", "detail::stride_vector constructor", - "Internal error: invalid stride API"); + "Internal error: invalid stride API"); } auto& v1 = stride_choice == stride_api::FB_STRIDES ? config_values.fwd_strides : config_values.input_strides; @@ -89,7 +89,7 @@ struct stride_vectors { if constexpr (std::is_unsigned_v) { if (v1[i] < 0 || v2[i] < 0) { throw math::unimplemented("DFT", "commit", - "Backend does not support negative strides."); + "Backend does not support negative strides."); } } vec_a[i] = static_cast(v1[i]); diff --git a/src/dft/descriptor.cxx b/src/dft/descriptor.cxx index df90d9f0e..57ed1f086 100644 --- a/src/dft/descriptor.cxx +++ b/src/dft/descriptor.cxx @@ -157,7 +157,7 @@ descriptor::descriptor(std::vector dimensions) { for (const auto& dim : dimensions) { if (dim <= 0) { throw math::invalid_argument("DFT", "descriptor", - "Invalid dimension value (negative or 0)."); + "Invalid dimension value (negative or 0)."); } } compute_default_strides(dimensions, values_.fwd_strides, values_.bwd_strides); @@ -283,7 +283,7 @@ void descriptor::set_workspace(scalar_type* usm_workspace) { } else { throw math::uninitialized("DFT", "set_workspace", - "Can only set workspace on committed descriptor."); + "Can only set workspace on committed descriptor."); } } @@ -294,7 +294,7 @@ void descriptor::set_workspace(sycl::buffer& buffer_work } else { throw math::uninitialized("DFT", "set_workspace", - "Can only set workspace on committed descriptor."); + "Can only set workspace on committed descriptor."); } } diff --git a/src/dft/descriptor_config_helper.hpp b/src/dft/descriptor_config_helper.hpp index c4d319096..00e3c8c79 100644 --- a/src/dft/descriptor_config_helper.hpp +++ b/src/dft/descriptor_config_helper.hpp @@ -118,7 +118,7 @@ void set_value(dft_values& vals, for (std::size_t i{ 0 }; i < vals.dimensions.size(); ++i) { if (set_val[i] <= 0) { throw math::invalid_argument("DFT", "set_value", - "Invalid length value (negative or 0)."); + "Invalid length value (negative or 0)."); } } std::copy(set_val, set_val + vals.dimensions.size(), vals.dimensions.begin()); @@ -135,7 +135,7 @@ void set_value(dft_values& vals, else if constexpr (Param == config_param::NUMBER_OF_TRANSFORMS) { if (set_val <= 0) { throw math::invalid_argument("DFT", "set_value", - "Number of transforms must be positive."); + "Number of transforms must be positive."); } vals.number_of_transforms = set_val; } @@ -145,7 +145,7 @@ void set_value(dft_values& vals, } else { throw math::invalid_argument("DFT", "set_value", - "Complex storage must be complex_complex or real_real."); + "Complex storage must be complex_complex or real_real."); } } else if constexpr (Param == config_param::REAL_STORAGE) { @@ -162,7 +162,7 @@ void set_value(dft_values& vals, } else { throw math::invalid_argument("DFT", "set_value", - "Conjugate even storage must be complex_complex."); + "Conjugate even storage must be complex_complex."); } } else if constexpr (Param == config_param::PLACEMENT) { @@ -171,7 +171,7 @@ void set_value(dft_values& vals, } else { throw math::invalid_argument("DFT", "set_value", - "Placement must be inplace or not inplace."); + "Placement must be inplace or not inplace."); } } #pragma clang diagnostic push @@ -224,7 +224,7 @@ void set_value(dft_values& vals, } else { throw math::invalid_argument("DFT", "set_value", - "Ordering must be ordered or backwards scrambled."); + "Ordering must be ordered or backwards scrambled."); } } else if constexpr (Param == config_param::TRANSPOSE) { diff --git a/src/dft/execute_helper_generic.hpp b/src/dft/execute_helper_generic.hpp index ba4cc709a..4f5ae6727 100644 --- a/src/dft/execute_helper_generic.hpp +++ b/src/dft/execute_helper_generic.hpp @@ -40,9 +40,9 @@ namespace oneapi::math::dft::detail { template static inline void fft_enqueue_task(HandlerT&& cgh, FnT&& f) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - cgh.ext_codeplay_enqueue_native_command([=](sycl::interop_handle ih){ + cgh.ext_codeplay_enqueue_native_command([=](sycl::interop_handle ih) { #else - cgh.host_task([=](sycl::interop_handle ih){ + cgh.host_task([=](sycl::interop_handle ih) { #endif f(std::move(ih)); }); diff --git a/src/dft/function_table.hpp b/src/dft/function_table.hpp index 244a06808..d1c647eb9 100644 --- a/src/dft/function_table.hpp +++ b/src/dft/function_table.hpp @@ -36,26 +36,28 @@ typedef struct { int version; oneapi::math::dft::detail::commit_impl* ( + oneapi::math::dft::domain::COMPLEX>* ( *create_commit_sycl_fz)( const oneapi::math::dft::descriptor& desc, + oneapi::math::dft::domain::COMPLEX>& desc, sycl::queue& sycl_queue); oneapi::math::dft::detail::commit_impl* ( + oneapi::math::dft::domain::COMPLEX>* ( *create_commit_sycl_dz)( const oneapi::math::dft::descriptor& desc, + oneapi::math::dft::domain::COMPLEX>& desc, sycl::queue& sycl_queue); oneapi::math::dft::detail::commit_impl* (*create_commit_sycl_fr)( + oneapi::math::dft::domain::REAL>* ( + *create_commit_sycl_fr)( const oneapi::math::dft::descriptor& desc, + oneapi::math::dft::domain::REAL>& desc, sycl::queue& sycl_queue); oneapi::math::dft::detail::commit_impl* (*create_commit_sycl_dr)( + oneapi::math::dft::domain::REAL>* ( + *create_commit_sycl_dr)( const oneapi::math::dft::descriptor& desc, + oneapi::math::dft::domain::REAL>& desc, sycl::queue& sycl_queue); } dft_function_table_t; diff --git a/src/include/allocator_helper.hpp b/src/include/allocator_helper.hpp index 9dfd6e36e..433e0468e 100644 --- a/src/include/allocator_helper.hpp +++ b/src/include/allocator_helper.hpp @@ -29,7 +29,7 @@ namespace oneapi { namespace math { -static inline void *aligned_alloc(size_t align, size_t size) { +static inline void* aligned_alloc(size_t align, size_t size) { #ifdef _WIN64 return ::_aligned_malloc(size, align); #else @@ -37,7 +37,7 @@ static inline void *aligned_alloc(size_t align, size_t size) { #endif } -static inline void aligned_free(void *p) { +static inline void aligned_free(void* p) { #ifdef _WIN64 ::_aligned_free(p); #else diff --git a/src/include/common_onemkl_conversion.hpp b/src/include/common_onemkl_conversion.hpp index a54669e71..47d9bb0a4 100644 --- a/src/include/common_onemkl_conversion.hpp +++ b/src/include/common_onemkl_conversion.hpp @@ -33,73 +33,120 @@ namespace oneapi { namespace math { namespace detail { -inline auto get_onemkl_transpose(oneapi::math::transpose* param_ptr) { return reinterpret_cast(param_ptr); } -inline auto get_onemkl_transpose(oneapi::math::transpose param) { return *get_onemkl_transpose(¶m); } - -inline auto get_onemkl_uplo(oneapi::math::uplo* param_ptr) { return reinterpret_cast(param_ptr); } -inline auto get_onemkl_uplo(oneapi::math::uplo param) { return *get_onemkl_uplo(¶m); } - -inline auto get_onemkl_diag(oneapi::math::diag* param_ptr) { return reinterpret_cast(param_ptr); } -inline auto get_onemkl_diag(oneapi::math::diag param) { return *get_onemkl_diag(¶m); } - -inline auto get_onemkl_side(oneapi::math::side* param_ptr) { return reinterpret_cast(param_ptr); } -inline auto get_onemkl_side(oneapi::math::side param) { return *get_onemkl_side(¶m); } - -inline auto get_onemkl_offset(oneapi::math::offset param) { return *reinterpret_cast(¶m); } - -inline auto get_onemkl_layout(oneapi::math::layout param) { return *reinterpret_cast(¶m); } - -inline auto get_onemkl_index_base(oneapi::math::index_base param) { return *reinterpret_cast(¶m); } - -inline auto get_onemkl_job(oneapi::math::job param) { return *reinterpret_cast(¶m); } - -inline auto get_onemkl_jobsvd(oneapi::math::jobsvd param) { return *reinterpret_cast(¶m); } - -inline auto get_onemkl_generate(oneapi::math::generate param) { return *reinterpret_cast(¶m); } - -inline auto get_onemkl_compz(oneapi::math::compz param) { return *reinterpret_cast(¶m); } - -inline auto get_onemkl_direct(oneapi::math::direct param) { return *reinterpret_cast(¶m); } - -inline auto get_onemkl_storev(oneapi::math::storev param) { return *reinterpret_cast(¶m); } - -inline auto get_onemkl_rangev(oneapi::math::rangev param) { return *reinterpret_cast(¶m); } - -inline auto get_onemkl_order(oneapi::math::order param) { return *reinterpret_cast(¶m); } +inline auto get_onemkl_transpose(oneapi::math::transpose* param_ptr) { + return reinterpret_cast(param_ptr); +} +inline auto get_onemkl_transpose(oneapi::math::transpose param) { + return *get_onemkl_transpose(¶m); +} + +inline auto get_onemkl_uplo(oneapi::math::uplo* param_ptr) { + return reinterpret_cast(param_ptr); +} +inline auto get_onemkl_uplo(oneapi::math::uplo param) { + return *get_onemkl_uplo(¶m); +} + +inline auto get_onemkl_diag(oneapi::math::diag* param_ptr) { + return reinterpret_cast(param_ptr); +} +inline auto get_onemkl_diag(oneapi::math::diag param) { + return *get_onemkl_diag(¶m); +} + +inline auto get_onemkl_side(oneapi::math::side* param_ptr) { + return reinterpret_cast(param_ptr); +} +inline auto get_onemkl_side(oneapi::math::side param) { + return *get_onemkl_side(¶m); +} + +inline auto get_onemkl_offset(oneapi::math::offset param) { + return *reinterpret_cast(¶m); +} + +inline auto get_onemkl_layout(oneapi::math::layout param) { + return *reinterpret_cast(¶m); +} + +inline auto get_onemkl_index_base(oneapi::math::index_base param) { + return *reinterpret_cast(¶m); +} + +inline auto get_onemkl_job(oneapi::math::job param) { + return *reinterpret_cast(¶m); +} + +inline auto get_onemkl_jobsvd(oneapi::math::jobsvd param) { + return *reinterpret_cast(¶m); +} + +inline auto get_onemkl_generate(oneapi::math::generate param) { + return *reinterpret_cast(¶m); +} + +inline auto get_onemkl_compz(oneapi::math::compz param) { + return *reinterpret_cast(¶m); +} + +inline auto get_onemkl_direct(oneapi::math::direct param) { + return *reinterpret_cast(¶m); +} + +inline auto get_onemkl_storev(oneapi::math::storev param) { + return *reinterpret_cast(¶m); +} + +inline auto get_onemkl_rangev(oneapi::math::rangev param) { + return *reinterpret_cast(¶m); +} + +inline auto get_onemkl_order(oneapi::math::order param) { + return *reinterpret_cast(¶m); +} // Rethrow Intel(R) oneMKL exceptions as oneMath exceptions -#define RETHROW_ONEMKL_EXCEPTIONS(EXPRESSION) \ -do { \ - try { \ - EXPRESSION; \ - } catch(const oneapi::mkl::unsupported_device& e) { \ - throw unsupported_device(e.what()); \ - } catch(const oneapi::mkl::host_bad_alloc& e) { \ - throw host_bad_alloc(e.what()); \ - } catch(const oneapi::mkl::device_bad_alloc& e) { \ - throw device_bad_alloc(e.what()); \ - } catch(const oneapi::mkl::unimplemented& e) { \ - throw unimplemented(e.what()); \ - } catch(const oneapi::mkl::invalid_argument& e) { \ - throw invalid_argument(e.what()); \ - } catch(const oneapi::mkl::uninitialized& e) { \ - throw uninitialized(e.what()); \ - } catch(const oneapi::mkl::computation_error& e) { \ - throw computation_error(e.what()); \ - } catch(const oneapi::mkl::batch_error& e) { \ - throw batch_error(e.what()); \ - } catch(const oneapi::mkl::exception& e) { \ - throw exception(e.what()); \ - } \ -} while (0) - -#define RETHROW_ONEMKL_EXCEPTIONS_RET(EXPRESSION) \ -do { \ -RETHROW_ONEMKL_EXCEPTIONS(return EXPRESSION); \ -} while(0) - -} // namespace detail -} // namespace math -} // namespace oneapi +#define RETHROW_ONEMKL_EXCEPTIONS(EXPRESSION) \ + do { \ + try { \ + EXPRESSION; \ + } \ + catch (const oneapi::mkl::unsupported_device& e) { \ + throw unsupported_device(e.what()); \ + } \ + catch (const oneapi::mkl::host_bad_alloc& e) { \ + throw host_bad_alloc(e.what()); \ + } \ + catch (const oneapi::mkl::device_bad_alloc& e) { \ + throw device_bad_alloc(e.what()); \ + } \ + catch (const oneapi::mkl::unimplemented& e) { \ + throw unimplemented(e.what()); \ + } \ + catch (const oneapi::mkl::invalid_argument& e) { \ + throw invalid_argument(e.what()); \ + } \ + catch (const oneapi::mkl::uninitialized& e) { \ + throw uninitialized(e.what()); \ + } \ + catch (const oneapi::mkl::computation_error& e) { \ + throw computation_error(e.what()); \ + } \ + catch (const oneapi::mkl::batch_error& e) { \ + throw batch_error(e.what()); \ + } \ + catch (const oneapi::mkl::exception& e) { \ + throw exception(e.what()); \ + } \ + } while (0) + +#define RETHROW_ONEMKL_EXCEPTIONS_RET(EXPRESSION) \ + do { \ + RETHROW_ONEMKL_EXCEPTIONS(return EXPRESSION); \ + } while (0) + +} // namespace detail +} // namespace math +} // namespace oneapi #endif // _ONEMATH_SRC_INCLUDE_COMMON_ONEMKL_TYPES_CONVERSION_HPP_ diff --git a/src/include/function_table_initializer.hpp b/src/include/function_table_initializer.hpp index c32cb691d..3e1f0824b 100644 --- a/src/include/function_table_initializer.hpp +++ b/src/include/function_table_initializer.hpp @@ -30,7 +30,7 @@ #ifdef __linux__ #include -#define LIB_TYPE void * +#define LIB_TYPE void* #define GET_LIB_HANDLE(libname) dlopen((libname), RTLD_LAZY | RTLD_GLOBAL) #define GET_FUNC(lib, fn) dlsym(lib, (fn)) #define FREE_LIB_HANDLE(libname) dlclose(libname) @@ -59,7 +59,7 @@ class table_initializer { using dlhandle = std::unique_ptr; public: - function_table_t &operator[](std::pair device_queue_pair) { + function_table_t& operator[](std::pair device_queue_pair) { auto lib = tables.find(device_queue_pair.first); if (lib != tables.end()) return lib->second; @@ -96,10 +96,10 @@ class table_initializer { } #endif - function_table_t &add_table(oneapi::math::device key, sycl::queue &q) { + function_table_t& add_table(oneapi::math::device key, sycl::queue& q) { dlhandle handle; // check all available libraries for the key(device) - for (const char *libname : libraries[domain_id][key]) { + for (const char* libname : libraries[domain_id][key]) { handle = dlhandle{ ::GET_LIB_HANDLE(libname) }; if (handle) break; @@ -114,7 +114,7 @@ class table_initializer { } } auto t = - reinterpret_cast(::GET_FUNC(handle.get(), table_names[domain_id])); + reinterpret_cast(::GET_FUNC(handle.get(), table_names[domain_id])); if (!t) { std::cerr << ERROR_MSG << '\n'; diff --git a/src/lapack/backends/cusolver/cusolver_batch.cpp b/src/lapack/backends/cusolver/cusolver_batch.cpp index f2c990963..3ad137912 100644 --- a/src/lapack/backends/cusolver/cusolver_batch.cpp +++ b/src/lapack/backends/cusolver/cusolver_batch.cpp @@ -31,24 +31,24 @@ namespace cusolver { // BATCH BUFFER API template -inline void geqrf_batch(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void geqrf_batch(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, stride_a, stride_tau, batch_size, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; // Uses scratch so sync between each cuSolver call @@ -62,10 +62,10 @@ inline void geqrf_batch(const char *func_name, Func func, sycl::queue &queue, st } #define GEQRF_STRIDED_BATCH_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, \ + void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, \ std::int64_t stride_tau, std::int64_t batch_size, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ return geqrf_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, stride_a, \ tau, stride_tau, batch_size, scratchpad, scratchpad_size); \ } @@ -78,10 +78,10 @@ GEQRF_STRIDED_BATCH_LAUNCHER(std::complex, cusolverDnZgeqrf) #undef GEQRF_STRIDED_BATCH_LAUNCHER template -inline void getri_batch(const char *func_name, Func func, sycl::queue &queue, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +inline void getri_batch(const char* func_name, Func func, sycl::queue& queue, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -91,7 +91,7 @@ inline void getri_batch(const char *func_name, Func func, sycl::queue &queue, st sycl::buffer ipiv32(sycl::range<1>{ ipiv32_size }); sycl::buffer devInfo{ batch_size }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto ipiv_acc = sycl::accessor{ ipiv, cgh, sycl::read_only }; auto ipiv32_acc = sycl::accessor{ ipiv32, cgh, sycl::write_only }; cgh.parallel_for(sycl::range<1>{ ipiv32_size }, [=](sycl::id<1> index) { @@ -102,7 +102,7 @@ inline void getri_batch(const char *func_name, Func func, sycl::queue &queue, st // getri_batched is contained within cublas, not cusolver. For this reason // we need to use cublas types instead of cusolver types (as is needed for // other lapack routines) - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { using blas::cublas::cublas_error; sycl::accessor a_acc{ a, cgh, sycl::read_only }; @@ -110,7 +110,7 @@ inline void getri_batch(const char *func_name, Func func, sycl::queue &queue, st sycl::accessor ipiv32_acc{ ipiv32, cgh }; sycl::accessor devInfo_acc{ devInfo, cgh, sycl::write_only }; - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { cublasStatus_t err; CUresult cuda_result; cublasHandle_t cublas_handle; @@ -118,27 +118,28 @@ inline void getri_batch(const char *func_name, Func func, sycl::queue &queue, st CUstream cu_stream = sycl::get_native(queue); CUBLAS_ERROR_FUNC(cublasSetStream, err, cublas_handle, cu_stream); - auto a_ = sc.get_mem(a_acc); - auto scratch_ = sc.get_mem(scratch_acc); - auto ipiv32_ = sc.get_mem(ipiv32_acc); - auto info_ = sc.get_mem(devInfo_acc); + auto a_ = sc.get_mem(a_acc); + auto scratch_ = sc.get_mem(scratch_acc); + auto ipiv32_ = sc.get_mem(ipiv32_acc); + auto info_ = sc.get_mem(devInfo_acc); CUdeviceptr a_dev; - cuDataType **a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); - CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T *) * batch_size); - CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T *) * batch_size); - auto **a_dev_ = reinterpret_cast(a_dev); + cuDataType** a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); + CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T*) * batch_size); + CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T*) * batch_size); + auto** a_dev_ = reinterpret_cast(a_dev); CUdeviceptr scratch_dev; - cuDataType **scratch_batched = + cuDataType** scratch_batched = create_ptr_list_from_stride(scratch_, stride_a, batch_size); - CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &scratch_dev, sizeof(T *) * batch_size); + CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &scratch_dev, sizeof(T*) * batch_size); CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, scratch_dev, scratch_batched, - sizeof(T *) * batch_size); - auto **scratch_dev_ = reinterpret_cast(scratch_dev); + sizeof(T*) * batch_size); + auto** scratch_dev_ = reinterpret_cast(scratch_dev); - blas::cublas::cublas_native_named_func(func_name, func, err, cublas_handle, n, a_dev_, lda, ipiv32_, - scratch_dev_, lda, info_, batch_size); + blas::cublas::cublas_native_named_func(func_name, func, err, cublas_handle, n, a_dev_, + lda, ipiv32_, scratch_dev_, lda, info_, + batch_size); free(a_batched); free(scratch_batched); @@ -148,7 +149,7 @@ inline void getri_batch(const char *func_name, Func func, sycl::queue &queue, st }); // The inverted matrices stored in scratch_ need to be stored in a_ - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { sycl::accessor a_acc{ a, cgh, sycl::write_only }; sycl::accessor scratch_acc{ scratchpad, cgh, sycl::read_only }; cgh.parallel_for(sycl::range<1>{ static_cast( @@ -156,7 +157,7 @@ inline void getri_batch(const char *func_name, Func func, sycl::queue &queue, st [=](sycl::id<1> index) { a_acc[index] = scratch_acc[index]; }); }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { sycl::accessor ipiv32_acc{ ipiv32, cgh, sycl::read_only }; sycl::accessor ipiv_acc{ ipiv, cgh, sycl::write_only }; cgh.parallel_for(sycl::range<1>{ static_cast(ipiv32_size) }, @@ -168,10 +169,10 @@ inline void getri_batch(const char *func_name, Func func, sycl::queue &queue, st } #define GETRI_STRIDED_BATCH_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ - std::int64_t stride_a, sycl::buffer &ipiv, \ + void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, \ + std::int64_t stride_a, sycl::buffer& ipiv, \ std::int64_t stride_ipiv, std::int64_t batch_size, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ return getri_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, n, a, lda, stride_a, ipiv, \ stride_ipiv, batch_size, scratchpad, scratchpad_size); \ } @@ -184,12 +185,12 @@ GETRI_STRIDED_BATCH_LAUNCHER(std::complex, cublasZgetriBatched) #undef GETRI_STRIDED_BATCH_LAUNCHER template -inline void getrs_batch(const char *func_name, Func func, sycl::queue &queue, +inline void getrs_batch(const char* func_name, Func func, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -201,7 +202,7 @@ inline void getrs_batch(const char *func_name, Func func, sycl::queue &queue, std::uint64_t ipiv_size = stride_ipiv * batch_size; sycl::buffer ipiv32(sycl::range<1>{ ipiv_size }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto ipiv32_acc = ipiv32.template get_access(cgh); auto ipiv_acc = ipiv.template get_access(cgh); cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { @@ -209,16 +210,16 @@ inline void getrs_batch(const char *func_name, Func func, sycl::queue &queue, }); }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto ipiv_acc = ipiv32.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto ipiv_ = sc.get_mem(ipiv_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto ipiv_ = sc.get_mem(ipiv_acc); + auto b_ = sc.get_mem(b_acc); cusolverStatus_t err; // Does not use scratch so call cuSolver asynchronously and sync at end @@ -235,12 +236,12 @@ inline void getrs_batch(const char *func_name, Func func, sycl::queue &queue, } #define GETRS_STRIDED_BATCH_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, \ - std::int64_t stride_a, sycl::buffer &ipiv, \ - std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, \ + void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, \ + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, \ + std::int64_t stride_a, sycl::buffer& ipiv, \ + std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, \ std::int64_t stride_b, std::int64_t batch_size, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ return getrs_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, trans, n, nrhs, a, lda, \ stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, \ scratchpad_size); \ @@ -254,10 +255,10 @@ GETRS_STRIDED_BATCH_LAUNCHER(std::complex, cusolverDnZgetrs) #undef GETRS_STRIDED_BATCH_LAUNCHER template -inline void getrf_batch(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +inline void getrf_batch(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -270,17 +271,17 @@ inline void getrf_batch(const char *func_name, Func func, sycl::queue &queue, st sycl::buffer ipiv32(sycl::range<1>{ ipiv_size }); sycl::buffer devInfo{ batch_size }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto ipiv32_acc = ipiv32.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto ipiv_ = sc.get_mem(ipiv32_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto ipiv_ = sc.get_mem(ipiv32_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; // Uses scratch so sync between each cuSolver call @@ -292,7 +293,7 @@ inline void getrf_batch(const char *func_name, Func func, sycl::queue &queue, st }); // Copy from 32-bit USM to 64-bit - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto ipiv32_acc = ipiv32.template get_access(cgh); auto ipiv_acc = ipiv.template get_access(cgh); cgh.parallel_for(sycl::range<1>{ ipiv_size }, @@ -303,10 +304,10 @@ inline void getrf_batch(const char *func_name, Func func, sycl::queue &queue, st } #define GETRF_STRIDED_BATCH_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, \ + void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, \ std::int64_t stride_ipiv, std::int64_t batch_size, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ return getrf_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, stride_a, \ ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); \ } @@ -319,25 +320,25 @@ GETRF_STRIDED_BATCH_LAUNCHER(std::complex, cusolverDnZgetrf) #undef GETRF_STRIDED_BATCH_LAUNCHER template -inline void orgqr_batch(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, +inline void orgqr_batch(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, stride_a, stride_tau, batch_size, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; // Uses scratch so sync between each cuSolver call @@ -351,10 +352,10 @@ inline void orgqr_batch(const char *func_name, Func func, sycl::queue &queue, st } #define ORGQR_STRIDED_BATCH_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, \ - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, \ - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, \ + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, \ + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ return orgqr_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, a, lda, stride_a, \ tau, stride_tau, batch_size, scratchpad, scratchpad_size); \ } @@ -365,30 +366,30 @@ ORGQR_STRIDED_BATCH_LAUNCHER(double, cusolverDnDorgqr) #undef ORGQR_STRIDED_BATCH_LAUNCHER template -inline void potrf_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, +inline void potrf_batch(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, stride_a, batch_size, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); CUdeviceptr a_dev; CUresult cuda_result; cusolverStatus_t err; - auto a_ = sc.get_mem(a_acc); + auto a_ = sc.get_mem(a_acc); // Transform ptr and stride to list of ptr's - cuDataType **a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); - CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T *) * batch_size); - CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T *) * batch_size); + cuDataType** a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); + CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T*) * batch_size); + CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T*) * batch_size); - auto **a_dev_ = reinterpret_cast(a_dev); + auto** a_dev_ = reinterpret_cast(a_dev); cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), (int)n, a_dev_, (int)lda, nullptr, (int)batch_size); @@ -401,9 +402,9 @@ inline void potrf_batch(const char *func_name, Func func, sycl::queue &queue, // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRF_STRIDED_BATCH_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, \ - std::int64_t batch_size, sycl::buffer &scratchpad, \ + void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, \ + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, \ + std::int64_t batch_size, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ return potrf_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, stride_a, \ batch_size, scratchpad, scratchpad_size); \ @@ -417,11 +418,11 @@ POTRF_STRIDED_BATCH_LAUNCHER(std::complex, cusolverDnZpotrfBatched) #undef POTRF_STRIDED_BATCH_LAUNCHER template -inline void potrs_batch(const char *func_name, Func func, sycl::queue &queue, +inline void potrs_batch(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; @@ -431,28 +432,28 @@ inline void potrs_batch(const char *func_name, Func func, sycl::queue &queue, if (nrhs != 1) throw unimplemented("lapack", "potrs_batch", "cusolver potrs_batch only supports nrhs = 1"); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); CUdeviceptr a_dev, b_dev; cusolverStatus_t err; CUresult cuda_result; - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); // Transform ptr and stride to list of ptr's - cuDataType **a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); - cuDataType **b_batched = create_ptr_list_from_stride(b_, stride_b, batch_size); - CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T *) * batch_size); - CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T *) * batch_size); - CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &b_dev, sizeof(T *) * batch_size); - CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, b_dev, b_batched, sizeof(T *) * batch_size); + cuDataType** a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); + cuDataType** b_batched = create_ptr_list_from_stride(b_, stride_b, batch_size); + CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T*) * batch_size); + CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T*) * batch_size); + CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &b_dev, sizeof(T*) * batch_size); + CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, b_dev, b_batched, sizeof(T*) * batch_size); - auto **a_dev_ = reinterpret_cast(a_dev); - auto **b_dev_ = reinterpret_cast(b_dev); + auto** a_dev_ = reinterpret_cast(a_dev); + auto** b_dev_ = reinterpret_cast(b_dev); cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), (int)n, (int)nrhs, a_dev_, (int)lda, b_dev_, ldb, nullptr, @@ -468,11 +469,11 @@ inline void potrs_batch(const char *func_name, Func func, sycl::queue &queue, // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRS_STRIDED_BATCH_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, \ - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, \ + void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, \ + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, \ + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, \ std::int64_t stride_b, std::int64_t batch_size, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ return potrs_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, nrhs, a, lda, \ stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); \ } @@ -485,25 +486,25 @@ POTRS_STRIDED_BATCH_LAUNCHER(std::complex, cusolverDnZpotrsBatched) #undef POTRS_STRIDED_BATCH_LAUNCHER template -inline void ungqr_batch(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, +inline void ungqr_batch(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, stride_a, stride_tau, batch_size, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; // Uses scratch so sync between each cuSolver call @@ -517,10 +518,10 @@ inline void ungqr_batch(const char *func_name, Func func, sycl::queue &queue, st } #define UNGQR_STRIDED_BATCH_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, \ - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, \ - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, \ + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, \ + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ return ungqr_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, a, lda, stride_a, \ tau, stride_tau, batch_size, scratchpad, scratchpad_size); \ } @@ -533,22 +534,22 @@ UNGQR_STRIDED_BATCH_LAUNCHER(std::complex, cusolverDnZungqr) // BATCH USM API template -inline sycl::event geqrf_batch(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, T *a, std::int64_t lda, std::int64_t stride_a, - T *tau, std::int64_t stride_tau, std::int64_t batch_size, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event geqrf_batch(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, T* a, std::int64_t lda, std::int64_t stride_a, + T* tau, std::int64_t stride_tau, std::int64_t batch_size, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, stride_a, stride_tau, batch_size, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; // Uses scratch so sync between each cuSolver call @@ -564,11 +565,11 @@ inline sycl::event geqrf_batch(const char *func_name, Func func, sycl::queue &qu } #define GEQRF_STRIDED_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, TYPE *a, \ - std::int64_t lda, std::int64_t stride_a, TYPE *tau, \ - std::int64_t stride_tau, std::int64_t batch_size, TYPE *scratchpad, \ + sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, TYPE* a, \ + std::int64_t lda, std::int64_t stride_a, TYPE* tau, \ + std::int64_t stride_tau, std::int64_t batch_size, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return geqrf_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, stride_a, \ tau, stride_tau, batch_size, scratchpad, scratchpad_size, \ dependencies); \ @@ -582,24 +583,24 @@ GEQRF_STRIDED_BATCH_LAUNCHER_USM(std::complex, cusolverDnZgeqrf) #undef GEQRF_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event geqrf_batch(const char *func_name, Func func, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, T **a, std::int64_t *lda, T **tau, - std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, +inline sycl::event geqrf_batch(const char* func_name, Func func, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, T** a, std::int64_t* lda, T** tau, + std::int64_t group_count, std::int64_t* group_sizes, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(group_count, scratchpad_size); for (int64_t i = 0; i < group_count; ++i) overflow_check(m[i], n[i], lda[i], group_sizes[i]); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); int64_t global_id = 0; cusolverStatus_t err; @@ -620,9 +621,9 @@ inline sycl::event geqrf_batch(const char *func_name, Func func, sycl::queue &qu #define GEQRF_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ sycl::event geqrf_batch( \ - sycl::queue &queue, std::int64_t *m, std::int64_t *n, TYPE **a, std::int64_t *lda, \ - TYPE **tau, std::int64_t group_count, std::int64_t *group_sizes, TYPE *scratchpad, \ - std::int64_t scratchpad_size, const std::vector &dependencies) { \ + sycl::queue& queue, std::int64_t* m, std::int64_t* n, TYPE** a, std::int64_t* lda, \ + TYPE** tau, std::int64_t group_count, std::int64_t* group_sizes, TYPE* scratchpad, \ + std::int64_t scratchpad_size, const std::vector& dependencies) { \ return geqrf_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, tau, \ group_count, group_sizes, scratchpad, scratchpad_size, dependencies); \ } @@ -635,11 +636,11 @@ GEQRF_BATCH_LAUNCHER_USM(std::complex, cusolverDnZgeqrf) #undef GEQRF_BATCH_LAUNCHER_USM template -inline sycl::event getrf_batch(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, T *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event getrf_batch(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, T* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, stride_a, stride_ipiv, batch_size, scratchpad_size); @@ -648,17 +649,17 @@ inline sycl::event getrf_batch(const char *func_name, Func func, sycl::queue &qu // To get around the limitation. // Allocate memory with 32-bit ints then copy over results std::uint64_t ipiv_size = stride_ipiv * batch_size; - int *ipiv32 = (int *)malloc_device(sizeof(int) * ipiv_size, queue); - int *devInfo = (int *)malloc_device(sizeof(int) * batch_size, queue); + int* ipiv32 = (int*)malloc_device(sizeof(int) * ipiv_size, queue); + int* devInfo = (int*)malloc_device(sizeof(int) * batch_size, queue); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratchpad_ = reinterpret_cast(scratchpad); - auto ipiv_ = reinterpret_cast(ipiv32); + auto a_ = reinterpret_cast(a); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratchpad_ = reinterpret_cast(scratchpad); + auto ipiv_ = reinterpret_cast(ipiv32); cusolverStatus_t err; // Uses scratch so sync between each cuSolver call @@ -670,14 +671,14 @@ inline sycl::event getrf_batch(const char *func_name, Func func, sycl::queue &qu }); // Copy from 32-bit USM to 64-bit - sycl::event done_casting = queue.submit([&](sycl::handler &cgh) { + sycl::event done_casting = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { ipiv[index] = ipiv32[index]; }); }); // Enqueue free memory, don't return event as not-neccessary for user to wait for ipiv32 being released - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done_casting); cgh.host_task([=](sycl::interop_handle ih) { sycl::free(ipiv32, queue); }); }); @@ -690,11 +691,11 @@ inline sycl::event getrf_batch(const char *func_name, Func func, sycl::queue &qu } #define GETRF_STRIDED_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, TYPE *a, \ - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, \ - std::int64_t stride_ipiv, std::int64_t batch_size, TYPE *scratchpad, \ + sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, TYPE* a, \ + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, \ + std::int64_t stride_ipiv, std::int64_t batch_size, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return getrf_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, stride_a, \ ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, \ dependencies); \ @@ -708,12 +709,12 @@ GETRF_STRIDED_BATCH_LAUNCHER_USM(std::complex, cusolverDnZgetrf) #undef GETRF_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event getrf_batch(const char *func_name, Func func, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, T **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, T *scratchpad, +inline sycl::event getrf_batch(const char* func_name, Func func, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, T** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; int64_t batch_size = 0; @@ -726,19 +727,19 @@ inline sycl::event getrf_batch(const char *func_name, Func func, sycl::queue &qu // cuSolver legacy api does not accept 64-bit ints. // To get around the limitation. // Allocate memory with 32-bit ints then copy over results - int **ipiv32 = (int **)malloc(sizeof(int *) * batch_size); + int** ipiv32 = (int**)malloc(sizeof(int*) * batch_size); int64_t global_id = 0; for (int64_t group_id = 0; group_id < group_count; ++group_id) for (int64_t local_id = 0; local_id < group_sizes[group_id]; ++local_id, ++global_id) - ipiv32[global_id] = (int *)malloc_device(sizeof(int) * n[group_id], queue); - int *devInfo = (int *)malloc_device(sizeof(int) * batch_size, queue); + ipiv32[global_id] = (int*)malloc_device(sizeof(int) * n[group_id], queue); + int* devInfo = (int*)malloc_device(sizeof(int) * batch_size, queue); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto scratch_ = reinterpret_cast(scratchpad); int64_t global_id = 0; cusolverStatus_t err; @@ -759,10 +760,10 @@ inline sycl::event getrf_batch(const char *func_name, Func func, sycl::queue &qu for (int64_t group_id = 0, global_id = 0; group_id < group_count; ++group_id) { uint64_t ipiv_size = n[group_id]; for (int64_t local_id = 0; local_id < group_sizes[group_id]; ++local_id, ++global_id) { - int64_t *d_ipiv = ipiv[global_id]; - int *d_ipiv32 = ipiv32[global_id]; + int64_t* d_ipiv = ipiv[global_id]; + int* d_ipiv32 = ipiv32[global_id]; - sycl::event e = queue.submit([&](sycl::handler &cgh) { + sycl::event e = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { d_ipiv[index] = d_ipiv32[index]; }); @@ -772,7 +773,7 @@ inline sycl::event getrf_batch(const char *func_name, Func func, sycl::queue &qu } // Enqueue free memory - sycl::event done_freeing = queue.submit([&](sycl::handler &cgh) { + sycl::event done_freeing = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(casting_dependencies); cgh.host_task([=](sycl::interop_handle ih) { for (int64_t global_id = 0; global_id < batch_size; ++global_id) @@ -789,11 +790,11 @@ inline sycl::event getrf_batch(const char *func_name, Func func, sycl::queue &qu } #define GETRF_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, TYPE **a, \ - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, \ - std::int64_t *group_sizes, TYPE *scratchpad, \ + sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, TYPE** a, \ + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, \ + std::int64_t* group_sizes, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return getrf_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, ipiv, \ group_count, group_sizes, scratchpad, scratchpad_size, dependencies); \ } @@ -806,20 +807,20 @@ GETRF_BATCH_LAUNCHER_USM(std::complex, cusolverDnZgetrf) #undef GETRS_BATCH_LAUNCHER_USM template -sycl::event getri_batch(const char *func_name, Func func, sycl::queue &queue, std::int64_t n, T *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, T *scratchpad, +sycl::event getri_batch(const char* func_name, Func func, sycl::queue& queue, std::int64_t n, T* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, stride_a, stride_ipiv, batch_size, scratchpad_size); std::uint64_t ipiv32_size = n * batch_size; - int *ipiv32 = sycl::malloc_device(ipiv32_size, queue); - int *devInfo = sycl::malloc_device(batch_size, queue); + int* ipiv32 = sycl::malloc_device(ipiv32_size, queue); + int* devInfo = sycl::malloc_device(batch_size, queue); - sycl::event done_casting = queue.submit([&](sycl::handler &cgh) { + sycl::event done_casting = queue.submit([&](sycl::handler& cgh) { cgh.parallel_for( sycl::range<1>{ static_cast(ipiv32_size) }, [=](sycl::id<1> index) { ipiv32[index] = static_cast(ipiv[(index / n) * stride_ipiv + index % n]); @@ -829,13 +830,13 @@ sycl::event getri_batch(const char *func_name, Func func, sycl::queue &queue, st // getri_batched is contained within cublas, not cusolver. For this reason // we need to use cublas types instead of cusolver types (as is needed for // other lapack routines) - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { using blas::cublas::cublas_error; cgh.depends_on(done_casting); cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { cublasStatus_t err; CUresult cuda_result; cublasHandle_t cublas_handle; @@ -844,23 +845,24 @@ sycl::event getri_batch(const char *func_name, Func func, sycl::queue &queue, st CUBLAS_ERROR_FUNC(cublasSetStream, err, cublas_handle, cu_stream); CUdeviceptr a_dev; - auto *a_ = reinterpret_cast(a); - cuDataType **a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); - CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T *) * batch_size); - CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T *) * batch_size); - auto **a_dev_ = reinterpret_cast(a_dev); + auto* a_ = reinterpret_cast(a); + cuDataType** a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); + CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T*) * batch_size); + CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T*) * batch_size); + auto** a_dev_ = reinterpret_cast(a_dev); CUdeviceptr scratch_dev; - auto *scratch_ = reinterpret_cast(scratchpad); - cuDataType **scratch_batched = + auto* scratch_ = reinterpret_cast(scratchpad); + cuDataType** scratch_batched = create_ptr_list_from_stride(scratch_, stride_a, batch_size); - CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &scratch_dev, sizeof(T *) * batch_size); + CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &scratch_dev, sizeof(T*) * batch_size); CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, scratch_dev, scratch_batched, - sizeof(T *) * batch_size); - auto **scratch_dev_ = reinterpret_cast(scratch_dev); + sizeof(T*) * batch_size); + auto** scratch_dev_ = reinterpret_cast(scratch_dev); - blas::cublas::cublas_native_named_func(func_name, func, err, cublas_handle, n, a_dev_, lda, ipiv32, - scratch_dev_, lda, devInfo, batch_size); + blas::cublas::cublas_native_named_func(func_name, func, err, cublas_handle, n, a_dev_, + lda, ipiv32, scratch_dev_, lda, devInfo, + batch_size); free(a_batched); free(scratch_batched); @@ -870,14 +872,14 @@ sycl::event getri_batch(const char *func_name, Func func, sycl::queue &queue, st }); // The inverted matrices stored in scratch_ need to be stored in a_ - auto copy1 = queue.submit([&](sycl::handler &cgh) { + auto copy1 = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); cgh.parallel_for( sycl::range<1>{ static_cast(stride_a * (batch_size - 1) + lda * n) }, [=](sycl::id<1> index) { a[index] = scratchpad[index]; }); }); - auto copy2 = queue.submit([&](sycl::handler &cgh) { + auto copy2 = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); cgh.parallel_for( sycl::range<1>{ static_cast(ipiv32_size) }, [=](sycl::id<1> index) { @@ -893,9 +895,9 @@ sycl::event getri_batch(const char *func_name, Func func, sycl::queue &queue, st #define GETRI_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ sycl::event getri_batch( \ - sycl::queue &queue, std::int64_t n, TYPE *a, std::int64_t lda, std::int64_t stride_a, \ - std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, TYPE *scratchpad, \ - std::int64_t scratchpad_size, const std::vector &dependencies) { \ + sycl::queue& queue, std::int64_t n, TYPE* a, std::int64_t lda, std::int64_t stride_a, \ + std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, TYPE* scratchpad, \ + std::int64_t scratchpad_size, const std::vector& dependencies) { \ return getri_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, n, a, lda, stride_a, ipiv, \ stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); \ } @@ -907,41 +909,41 @@ GETRI_BATCH_LAUNCHER_USM(std::complex, cublasZgetriBatched) #undef GETRI_BATCH_LAUNCHER_USM -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } template -inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event getrs_batch(const char* func_name, Func func, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - T *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, T *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, T *scratchpad, + T* a, std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, T* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, nrhs, lda, ldb, stride_ipiv, stride_b, batch_size, scratchpad_size); @@ -950,22 +952,22 @@ inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &qu // To get around the limitation. // Create new memory and convert 64-bit values. std::uint64_t ipiv_size = stride_ipiv * batch_size; - int *ipiv32 = (int *)malloc_device(sizeof(int) * ipiv_size, queue); + int* ipiv32 = (int*)malloc_device(sizeof(int) * ipiv_size, queue); - auto done_casting = queue.submit([&](sycl::handler &cgh) { + auto done_casting = queue.submit([&](sycl::handler& cgh) { cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { ipiv32[index] = static_cast(ipiv[index]); }); }); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); cgh.depends_on(done_casting); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto ipiv_ = reinterpret_cast(ipiv32); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto ipiv_ = reinterpret_cast(ipiv32); + auto b_ = reinterpret_cast(b); cusolverStatus_t err; // Does not use scratch so call cuSolver asynchronously and sync at end @@ -986,12 +988,12 @@ inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &qu } #define GETRS_STRIDED_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ - std::int64_t nrhs, TYPE *a, std::int64_t lda, std::int64_t stride_a, \ - std::int64_t *ipiv, std::int64_t stride_ipiv, TYPE *b, \ + sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, \ + std::int64_t nrhs, TYPE* a, std::int64_t lda, std::int64_t stride_a, \ + std::int64_t* ipiv, std::int64_t stride_ipiv, TYPE* b, \ std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, \ - TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return getrs_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, trans, n, nrhs, a, lda, \ stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, \ scratchpad_size, dependencies); \ @@ -1005,13 +1007,13 @@ GETRS_STRIDED_BATCH_LAUNCHER_USM(std::complex, cusolverDnZgetrs) #undef GETRS_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - T **a, std::int64_t *lda, std::int64_t **ipiv, T **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, T *scratchpad, +inline sycl::event getrs_batch(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + T** a, std::int64_t* lda, std::int64_t** ipiv, T** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; int64_t batch_size = 0; @@ -1026,17 +1028,17 @@ inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &qu // an array of 64-bit ints in device memory. Each vec of ipiv // values need to be converted from 64-bit to 32-bit. The list // must stay on host. - int **ipiv32 = (int **)malloc(sizeof(int *) * batch_size); + int** ipiv32 = (int**)malloc(sizeof(int*) * batch_size); std::vector casting_dependencies(batch_size); int64_t global_id = 0; for (int64_t group_id = 0; group_id < group_count; ++group_id) { for (int64_t local_id = 0; local_id < group_sizes[group_id]; ++local_id, ++global_id) { uint64_t ipiv_size = n[group_id]; - int *d_group_ipiv32 = (int *)malloc_device(sizeof(int) * ipiv_size, queue); + int* d_group_ipiv32 = (int*)malloc_device(sizeof(int) * ipiv_size, queue); ipiv32[global_id] = d_group_ipiv32; - int64_t *d_group_ipiv = ipiv[global_id]; + int64_t* d_group_ipiv = ipiv[global_id]; - auto e = queue.submit([&](sycl::handler &cgh) { + auto e = queue.submit([&](sycl::handler& cgh) { cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { d_group_ipiv32[index] = static_cast(d_group_ipiv[index]); }); @@ -1045,14 +1047,14 @@ inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &qu } } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); cgh.depends_on(casting_dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); cusolverStatus_t err; int64_t global_id = 0; @@ -1079,15 +1081,15 @@ inline sycl::event getrs_batch(const char *func_name, Func func, sycl::queue &qu return done; } -#define GETRS_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event getrs_batch( \ - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, \ - TYPE **a, std::int64_t *lda, std::int64_t **ipiv, TYPE **b, std::int64_t *ldb, \ - std::int64_t group_count, std::int64_t *group_sizes, TYPE *scratchpad, \ - std::int64_t scratchpad_size, const std::vector &dependencies) { \ - return getrs_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, trans, n, nrhs, a, lda, \ - ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, \ - dependencies); \ +#define GETRS_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ + sycl::event getrs_batch( \ + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, \ + TYPE** a, std::int64_t* lda, std::int64_t** ipiv, TYPE** b, std::int64_t* ldb, \ + std::int64_t group_count, std::int64_t* group_sizes, TYPE* scratchpad, \ + std::int64_t scratchpad_size, const std::vector& dependencies) { \ + return getrs_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, trans, n, nrhs, a, lda, \ + ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, \ + dependencies); \ } GETRS_BATCH_LAUNCHER_USM(float, cusolverDnSgetrs) @@ -1098,22 +1100,22 @@ GETRS_BATCH_LAUNCHER_USM(std::complex, cusolverDnZgetrs) #undef GETRS_BATCH_LAUNCHER_USM template -inline sycl::event orgqr_batch(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, T *a, std::int64_t lda, - std::int64_t stride_a, T *tau, std::int64_t stride_tau, - std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event orgqr_batch(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, T* a, std::int64_t lda, + std::int64_t stride_a, T* tau, std::int64_t stride_tau, + std::int64_t batch_size, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, stride_a, stride_tau, batch_size, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; // Uses scratch so sync between each cuSolver call @@ -1129,11 +1131,11 @@ inline sycl::event orgqr_batch(const char *func_name, Func func, sycl::queue &qu } #define ORGQR_STRIDED_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, \ - TYPE *a, std::int64_t lda, std::int64_t stride_a, TYPE *tau, \ - std::int64_t stride_tau, std::int64_t batch_size, TYPE *scratchpad, \ + sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, \ + TYPE* a, std::int64_t lda, std::int64_t stride_a, TYPE* tau, \ + std::int64_t stride_tau, std::int64_t batch_size, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return orgqr_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, a, lda, stride_a, \ tau, stride_tau, batch_size, scratchpad, scratchpad_size, \ dependencies); \ @@ -1145,25 +1147,25 @@ ORGQR_STRIDED_BATCH_LAUNCHER_USM(double, cusolverDnDorgqr) #undef ORGQR_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event orgqr_batch(const char *func_name, Func func, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, T **a, - std::int64_t *lda, T **tau, std::int64_t group_count, - std::int64_t *group_sizes, T *scratchpad, +inline sycl::event orgqr_batch(const char* func_name, Func func, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, T** a, + std::int64_t* lda, T** tau, std::int64_t group_count, + std::int64_t* group_sizes, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(group_count, scratchpad_size); for (int64_t i = 0; i < group_count; ++i) overflow_check(m[i], n[i], k[i], lda[i], group_sizes[i]); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); int64_t global_id = 0; cusolverStatus_t err; @@ -1184,11 +1186,11 @@ inline sycl::event orgqr_batch(const char *func_name, Func func, sycl::queue &qu } #define ORGQR_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, \ - TYPE **a, std::int64_t *lda, TYPE **tau, std::int64_t group_count, \ - std::int64_t *group_sizes, TYPE *scratchpad, \ + sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, \ + TYPE** a, std::int64_t* lda, TYPE** tau, std::int64_t group_count, \ + std::int64_t* group_sizes, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return orgqr_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, a, lda, tau, \ group_count, group_sizes, scratchpad, scratchpad_size, dependencies); \ } @@ -1199,31 +1201,31 @@ ORGQR_BATCH_LAUNCHER_USM(double, cusolverDnDorgqr) #undef ORGQR_BATCH_LAUNCHER_USM template -inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t batch_size, T *scratchpad, +inline sycl::event potrf_batch(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, stride_a, batch_size, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); CUdeviceptr a_dev; cusolverStatus_t err; CUresult cuda_result; - auto *a_ = reinterpret_cast(a); + auto* a_ = reinterpret_cast(a); // Transform ptr and stride to list of ptr's - cuDataType **a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); - CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T *) * batch_size); - CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T *) * batch_size); + cuDataType** a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); + CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T*) * batch_size); + CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T*) * batch_size); - auto **a_dev_ = reinterpret_cast(a_dev); + auto** a_dev_ = reinterpret_cast(a_dev); cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), (int)n, a_dev_, (int)lda, nullptr, (int)batch_size); @@ -1237,10 +1239,10 @@ inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &qu // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRF_STRIDED_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ + sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, \ - TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return potrf_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, stride_a, \ batch_size, scratchpad, scratchpad_size, dependencies); \ } @@ -1253,11 +1255,11 @@ POTRF_STRIDED_BATCH_LAUNCHER_USM(std::complex, cusolverDnZpotrfBatched) #undef POTRF_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, T **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, +inline sycl::event potrf_batch(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, T** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; int64_t batch_size = 0; @@ -1266,19 +1268,19 @@ inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &qu batch_size += group_sizes[i]; } - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; CUdeviceptr a_dev; CUresult cuda_result; cusolverStatus_t err; - CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T *) * batch_size); - CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a, sizeof(T *) * batch_size); + CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T*) * batch_size); + CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a, sizeof(T*) * batch_size); - auto **a_dev_ = reinterpret_cast(a_dev); + auto** a_dev_ = reinterpret_cast(a_dev); // Does not use scratch so call cuSolver asynchronously and sync at end for (int64_t i = 0; i < group_count; i++) { @@ -1298,13 +1300,13 @@ inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &qu } // Scratchpad memory not needed as parts of buffer a is used as workspace memory -#define POTRF_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event potrf_batch( \ - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, TYPE **a, std::int64_t *lda, \ - std::int64_t group_count, std::int64_t *group_sizes, TYPE *scratchpad, \ - std::int64_t scratchpad_size, const std::vector &dependencies) { \ - return potrf_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, \ - group_count, group_sizes, scratchpad, scratchpad_size, dependencies); \ +#define POTRF_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ + sycl::event potrf_batch( \ + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, TYPE** a, \ + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes, TYPE* scratchpad, \ + std::int64_t scratchpad_size, const std::vector& dependencies) { \ + return potrf_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, \ + group_count, group_sizes, scratchpad, scratchpad_size, dependencies); \ } POTRF_BATCH_LAUNCHER_USM(float, cusolverDnSpotrfBatched) @@ -1315,12 +1317,12 @@ POTRF_BATCH_LAUNCHER_USM(std::complex, cusolverDnZpotrfBatched) #undef POTRF_BATCH_LAUNCHER_USM template -inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, T *a, - std::int64_t lda, std::int64_t stride_a, T *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, T *scratchpad, +inline sycl::event potrs_batch(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, T* a, + std::int64_t lda, std::int64_t stride_a, T* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, nrhs, lda, ldb, stride_a, stride_b, batch_size, scratchpad_size); @@ -1329,26 +1331,26 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu if (nrhs != 1) throw unimplemented("lapack", "potrs_batch", "cusolver potrs_batch only supports nrhs = 1"); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); CUresult cuda_result; CUdeviceptr a_dev, b_dev; - auto *a_ = reinterpret_cast(a); - auto *b_ = reinterpret_cast(b); + auto* a_ = reinterpret_cast(a); + auto* b_ = reinterpret_cast(b); cusolverStatus_t err; // Transform ptr and stride to list of ptr's - cuDataType **a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); - cuDataType **b_batched = create_ptr_list_from_stride(b_, stride_b, batch_size); - CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T *) * batch_size); - CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &b_dev, sizeof(T *) * batch_size); - CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T *) * batch_size); - CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, b_dev, b_batched, sizeof(T *) * batch_size); + cuDataType** a_batched = create_ptr_list_from_stride(a_, stride_a, batch_size); + cuDataType** b_batched = create_ptr_list_from_stride(b_, stride_b, batch_size); + CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &a_dev, sizeof(T*) * batch_size); + CUDA_ERROR_FUNC(cuMemAlloc, cuda_result, &b_dev, sizeof(T*) * batch_size); + CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, a_dev, a_batched, sizeof(T*) * batch_size); + CUDA_ERROR_FUNC(cuMemcpyHtoD, cuda_result, b_dev, b_batched, sizeof(T*) * batch_size); - auto **a_dev_ = reinterpret_cast(a_dev); - auto **b_dev_ = reinterpret_cast(b_dev); + auto** a_dev_ = reinterpret_cast(a_dev); + auto** b_dev_ = reinterpret_cast(b_dev); cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), (int)n, (int)nrhs, a_dev_, (int)lda, b_dev_, ldb, nullptr, @@ -1365,10 +1367,10 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRS_STRIDED_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ sycl::event potrs_batch( \ - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, TYPE *a, \ - std::int64_t lda, std::int64_t stride_a, TYPE *b, std::int64_t ldb, std::int64_t stride_b, \ - std::int64_t batch_size, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, TYPE* a, \ + std::int64_t lda, std::int64_t stride_a, TYPE* b, std::int64_t ldb, std::int64_t stride_b, \ + std::int64_t batch_size, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return potrs_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, nrhs, a, lda, \ stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, \ dependencies); \ @@ -1382,12 +1384,12 @@ POTRS_STRIDED_BATCH_LAUNCHER_USM(std::complex, cusolverDnZpotrsBatched) #undef POTRS_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, T **a, - std::int64_t *lda, T **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, +inline sycl::event potrs_batch(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, T** a, + std::int64_t* lda, T** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; int64_t batch_size = 0; @@ -1401,29 +1403,29 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu "cusolver potrs_batch only supports nrhs = 1"); } - int *info = (int *)malloc_device(sizeof(int *) * batch_size, queue); - T **a_dev = (T **)malloc_device(sizeof(T *) * batch_size, queue); - T **b_dev = (T **)malloc_device(sizeof(T *) * batch_size, queue); + int* info = (int*)malloc_device(sizeof(int*) * batch_size, queue); + T** a_dev = (T**)malloc_device(sizeof(T*) * batch_size, queue); + T** b_dev = (T**)malloc_device(sizeof(T*) * batch_size, queue); auto done_cpy_a = - queue.submit([&](sycl::handler &h) { h.memcpy(a_dev, a, batch_size * sizeof(T *)); }); + queue.submit([&](sycl::handler& h) { h.memcpy(a_dev, a, batch_size * sizeof(T*)); }); auto done_cpy_b = - queue.submit([&](sycl::handler &h) { h.memcpy(b_dev, b, batch_size * sizeof(T *)); }); + queue.submit([&](sycl::handler& h) { h.memcpy(b_dev, b, batch_size * sizeof(T*)); }); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); cgh.depends_on(done_cpy_a); cgh.depends_on(done_cpy_b); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; cusolverStatus_t err; // Does not use scratch so call cuSolver asynchronously and sync at end for (int64_t i = 0; i < group_count; i++) { - auto **a_ = reinterpret_cast(a_dev); - auto **b_ = reinterpret_cast(b_dev); - auto info_ = reinterpret_cast(info); + auto** a_ = reinterpret_cast(a_dev); + auto** b_ = reinterpret_cast(b_dev); + auto info_ = reinterpret_cast(info); CUSOLVER_ERROR_FUNC_T(func_name, func, err, handle, get_cublas_fill_mode(uplo[i]), (int)n[i], (int)nrhs[i], a_ + offset, (int)lda[i], b_ + offset, (int)ldb[i], info_, (int)group_sizes[i]); @@ -1440,10 +1442,10 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRS_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ sycl::event potrs_batch( \ - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, \ - TYPE **a, std::int64_t *lda, TYPE **b, std::int64_t *ldb, std::int64_t group_count, \ - std::int64_t *group_sizes, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, \ + TYPE** a, std::int64_t* lda, TYPE** b, std::int64_t* ldb, std::int64_t group_count, \ + std::int64_t* group_sizes, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return potrs_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, nrhs, a, lda, b, \ ldb, group_count, group_sizes, scratchpad, scratchpad_size, \ dependencies); \ @@ -1457,22 +1459,22 @@ POTRS_BATCH_LAUNCHER_USM(std::complex, cusolverDnZpotrsBatched) #undef POTRS_BATCH_LAUNCHER_USM template -inline sycl::event ungqr_batch(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, T *a, std::int64_t lda, - std::int64_t stride_a, T *tau, std::int64_t stride_tau, - std::int64_t batch_size, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event ungqr_batch(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, T* a, std::int64_t lda, + std::int64_t stride_a, T* tau, std::int64_t stride_tau, + std::int64_t batch_size, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, stride_a, stride_tau, batch_size, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; // Uses scratch so sync between each cuSolver call @@ -1488,11 +1490,11 @@ inline sycl::event ungqr_batch(const char *func_name, Func func, sycl::queue &qu } #define UNGQR_STRIDED_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, \ - TYPE *a, std::int64_t lda, std::int64_t stride_a, TYPE *tau, \ - std::int64_t stride_tau, std::int64_t batch_size, TYPE *scratchpad, \ + sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, \ + TYPE* a, std::int64_t lda, std::int64_t stride_a, TYPE* tau, \ + std::int64_t stride_tau, std::int64_t batch_size, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return ungqr_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, a, lda, stride_a, \ tau, stride_tau, batch_size, scratchpad, scratchpad_size, \ dependencies); \ @@ -1504,25 +1506,25 @@ UNGQR_STRIDED_BATCH_LAUNCHER_USM(std::complex, cusolverDnZungqr) #undef UNGQR_STRIDED_BATCH_LAUNCHER_USM template -inline sycl::event ungqr_batch(const char *func_name, Func func, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, T **a, - std::int64_t *lda, T **tau, std::int64_t group_count, - std::int64_t *group_sizes, T *scratchpad, +inline sycl::event ungqr_batch(const char* func_name, Func func, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, T** a, + std::int64_t* lda, T** tau, std::int64_t group_count, + std::int64_t* group_sizes, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(group_count, scratchpad_size); for (int64_t i = 0; i < group_count; ++i) overflow_check(m[i], n[i], k[i], lda[i], group_sizes[i]); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); int64_t global_id = 0; cusolverStatus_t err; @@ -1543,11 +1545,11 @@ inline sycl::event ungqr_batch(const char *func_name, Func func, sycl::queue &qu } #define UNGQR_BATCH_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, \ - TYPE **a, std::int64_t *lda, TYPE **tau, std::int64_t group_count, \ - std::int64_t *group_sizes, TYPE *scratchpad, \ + sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, \ + TYPE** a, std::int64_t* lda, TYPE** tau, std::int64_t group_count, \ + std::int64_t* group_sizes, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return ungqr_batch(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, a, lda, tau, \ group_count, group_sizes, scratchpad, scratchpad_size, dependencies); \ } @@ -1560,12 +1562,12 @@ UNGQR_BATCH_LAUNCHER_USM(std::complex, cusolverDnZungqr) // BATCH SCRATCHPAD API template -inline void getrf_batch_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void getrf_batch_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, - std::int64_t batch_size, int *scratch_size) { - auto e = queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + std::int64_t batch_size, int* scratch_size) { + auto e = queue.submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; @@ -1611,13 +1613,13 @@ GETRI_STRIDED_BATCH_LAUNCHER_SCRATCH(std::complex) #undef GETRI_STRIDED_BATCH_LAUNCHER_SCRATCH // cusolverDnXgetrs does not use scratchpad memory -#define GETRS_STRIDED_BATCH_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t getrs_batch_scratchpad_size( \ +#define GETRS_STRIDED_BATCH_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t getrs_batch_scratchpad_size( \ sycl::queue & queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, \ - std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, \ - std::int64_t stride_b, std::int64_t batch_size) { \ - return 0; \ + std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, \ + std::int64_t stride_b, std::int64_t batch_size) { \ + return 0; \ } GETRS_STRIDED_BATCH_LAUNCHER_SCRATCH(float) @@ -1628,12 +1630,12 @@ GETRS_STRIDED_BATCH_LAUNCHER_SCRATCH(std::complex) #undef GETRS_STRIDED_BATCH_LAUNCHER_SCRATCH template -inline void geqrf_batch_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void geqrf_batch_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, - std::int64_t batch_size, int *scratch_size) { - auto e = queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + std::int64_t batch_size, int* scratch_size) { + auto e = queue.submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; @@ -1662,12 +1664,12 @@ GEQRF_STRIDED_BATCH_LAUNCHER_SCRATCH(std::complex, cusolverDnZgeqrf_buff #undef GEQRF_STRIDED_BATCH_LAUNCHER_SCRATCH // cusolverDnXpotrfBatched does not use scratchpad memory -#define POTRF_STRIDED_BATCH_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t potrf_batch_scratchpad_size( \ +#define POTRF_STRIDED_BATCH_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t potrf_batch_scratchpad_size( \ sycl::queue & queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, \ - std::int64_t stride_a, std::int64_t batch_size) { \ - return 0; \ + std::int64_t stride_a, std::int64_t batch_size) { \ + return 0; \ } POTRF_STRIDED_BATCH_LAUNCHER_SCRATCH(float) @@ -1681,7 +1683,7 @@ POTRF_STRIDED_BATCH_LAUNCHER_SCRATCH(std::complex) #define POTRS_STRIDED_BATCH_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t potrs_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, \ + sycl::queue & queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, \ std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, \ std::int64_t batch_size) { \ return 0; \ @@ -1695,13 +1697,13 @@ POTRS_STRIDED_BATCH_LAUNCHER_SCRATCH(std::complex) #undef POTRS_STRIDED_BATCH_LAUNCHER_SCRATCH template -inline void orgqr_batch_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void orgqr_batch_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size, - int *scratch_size) { - auto e = queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + int* scratch_size) { + auto e = queue.submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; @@ -1729,13 +1731,13 @@ ORGQR_STRIDED_BATCH_LAUNCHER_SCRATCH(double, cusolverDnDorgqr_bufferSize) #undef ORGQR_STRIDED_BATCH_LAUNCHER_SCRATCH template -inline void ungqr_batch_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void ungqr_batch_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size, - int *scratch_size) { - auto e = queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + int* scratch_size) { + auto e = queue.submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); cusolverStatus_t err; @@ -1763,12 +1765,12 @@ ORGQR_STRIDED_BATCH_LAUNCHER_SCRATCH(std::complex, cusolverDnZungqr_buff #undef ORGQR_STRIDED_BATCH_LAUNCHER_SCRATCH template -inline void getrf_batch_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - int *scratch_size) { - auto e = queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { +inline void getrf_batch_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + int* scratch_size) { + auto e = queue.submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int group_scratch_size = 0; *scratch_size = 0; @@ -1789,8 +1791,8 @@ inline void getrf_batch_scratchpad_size(const char *func_name, Func func, sycl:: #define GETRF_GROUP_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ std::int64_t getrf_batch_scratchpad_size( \ - sycl::queue & queue, std::int64_t * m, std::int64_t * n, std::int64_t * lda, \ - std::int64_t group_count, std::int64_t * group_sizes) { \ + sycl::queue & queue, std::int64_t* m, std::int64_t* n, std::int64_t* lda, \ + std::int64_t group_count, std::int64_t* group_sizes) { \ int scratch_size; \ getrf_batch_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, lda, \ group_count, group_sizes, &scratch_size); \ @@ -1804,18 +1806,18 @@ GETRF_GROUP_LAUNCHER_SCRATCH(std::complex, cusolverDnZgetrf_bufferSize) #undef GETRF_GROUP_LAUNCHER_SCRATCH -#define GETRI_GROUP_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t getri_batch_scratchpad_size(sycl::queue & queue, std::int64_t * n, \ - std::int64_t * lda, std::int64_t group_count, \ - std::int64_t * group_sizes) { \ - std::int64_t max_scratch_sz = 0; \ - for (auto group_id = 0; group_id < group_count; ++group_id) { \ - auto scratch_sz = lda[group_id] * n[group_id]; \ - if (scratch_sz > max_scratch_sz) \ - max_scratch_sz = scratch_sz; \ - } \ - return max_scratch_sz; \ +#define GETRI_GROUP_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t getri_batch_scratchpad_size(sycl::queue & queue, std::int64_t* n, \ + std::int64_t* lda, std::int64_t group_count, \ + std::int64_t* group_sizes) { \ + std::int64_t max_scratch_sz = 0; \ + for (auto group_id = 0; group_id < group_count; ++group_id) { \ + auto scratch_sz = lda[group_id] * n[group_id]; \ + if (scratch_sz > max_scratch_sz) \ + max_scratch_sz = scratch_sz; \ + } \ + return max_scratch_sz; \ } GETRI_GROUP_LAUNCHER_SCRATCH(float) @@ -1825,13 +1827,13 @@ GETRI_GROUP_LAUNCHER_SCRATCH(std::complex) #undef GETRI_GROUP_LAUNCHER_SCRATCH -#define GETRS_GROUP_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t getrs_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::math::transpose * trans, std::int64_t * n, \ - std::int64_t * nrhs, std::int64_t * lda, std::int64_t * ldb, std::int64_t group_count, \ - std::int64_t * group_sizes) { \ - return 0; \ +#define GETRS_GROUP_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t getrs_batch_scratchpad_size( \ + sycl::queue & queue, oneapi::math::transpose * trans, std::int64_t* n, std::int64_t* nrhs, \ + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, \ + std::int64_t* group_sizes) { \ + return 0; \ } GETRS_GROUP_LAUNCHER_SCRATCH(float) @@ -1842,12 +1844,12 @@ GETRS_GROUP_LAUNCHER_SCRATCH(std::complex) #undef GETRS_GROUP_LAUNCHER_SCRATCH template -inline void geqrf_batch_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - int *scratch_size) { - auto e = queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { +inline void geqrf_batch_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + int* scratch_size) { + auto e = queue.submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int group_scratch_size = 0; *scratch_size = 0; @@ -1868,8 +1870,8 @@ inline void geqrf_batch_scratchpad_size(const char *func_name, Func func, sycl:: #define GEQRF_GROUP_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ std::int64_t geqrf_batch_scratchpad_size( \ - sycl::queue & queue, std::int64_t * m, std::int64_t * n, std::int64_t * lda, \ - std::int64_t group_count, std::int64_t * group_sizes) { \ + sycl::queue & queue, std::int64_t* m, std::int64_t* n, std::int64_t* lda, \ + std::int64_t group_count, std::int64_t* group_sizes) { \ int scratch_size; \ geqrf_batch_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, lda, \ group_count, group_sizes, &scratch_size); \ @@ -1884,12 +1886,12 @@ GEQRF_GROUP_LAUNCHER_SCRATCH(std::complex, cusolverDnZgeqrf_bufferSize) #undef GEQRF_GROUP_LAUNCHER_SCRATCH template -inline void orgqr_batch_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, int *scratch_size) { - auto e = queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { +inline void orgqr_batch_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, int* scratch_size) { + auto e = queue.submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int group_scratch_size = 0; *scratch_size = 0; @@ -1908,15 +1910,15 @@ inline void orgqr_batch_scratchpad_size(const char *func_name, Func func, sycl:: e.wait(); } -#define ORGQR_GROUP_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ - std::int64_t orgqr_batch_scratchpad_size( \ - sycl::queue & queue, std::int64_t * m, std::int64_t * n, std::int64_t * k, \ - std::int64_t * lda, std::int64_t group_count, std::int64_t * group_sizes) { \ - int scratch_size; \ - orgqr_batch_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, lda, \ - group_count, group_sizes, &scratch_size); \ - return scratch_size; \ +#define ORGQR_GROUP_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ + std::int64_t orgqr_batch_scratchpad_size( \ + sycl::queue & queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, std::int64_t* lda, \ + std::int64_t group_count, std::int64_t* group_sizes) { \ + int scratch_size; \ + orgqr_batch_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, lda, \ + group_count, group_sizes, &scratch_size); \ + return scratch_size; \ } ORGQR_GROUP_LAUNCHER_SCRATCH(float, cusolverDnSorgqr_bufferSize) @@ -1925,12 +1927,12 @@ ORGQR_GROUP_LAUNCHER_SCRATCH(double, cusolverDnDorgqr_bufferSize) #undef ORGQR_GROUP_LAUNCHER_SCRATCH // cusolverDnXpotrfBatched does not use scratchpad memory -#define POTRF_GROUP_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t potrf_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t * n, std::int64_t * lda, \ - std::int64_t group_count, std::int64_t * group_sizes) { \ - return 0; \ +#define POTRF_GROUP_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t potrf_batch_scratchpad_size( \ + sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t* n, std::int64_t* lda, \ + std::int64_t group_count, std::int64_t* group_sizes) { \ + return 0; \ } POTRF_GROUP_LAUNCHER_SCRATCH(float) @@ -1941,13 +1943,13 @@ POTRF_GROUP_LAUNCHER_SCRATCH(std::complex) #undef POTRF_GROUP_LAUNCHER_SCRATCH // cusolverDnXpotrsBatched does not use scratchpad memory -#define POTRS_GROUP_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t potrs_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t * n, std::int64_t * nrhs, \ - std::int64_t * lda, std::int64_t * ldb, std::int64_t group_count, \ - std::int64_t * group_sizes) { \ - return 0; \ +#define POTRS_GROUP_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t potrs_batch_scratchpad_size( \ + sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t* n, std::int64_t* nrhs, \ + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, \ + std::int64_t* group_sizes) { \ + return 0; \ } POTRS_GROUP_LAUNCHER_SCRATCH(float) @@ -1958,12 +1960,12 @@ POTRS_GROUP_LAUNCHER_SCRATCH(std::complex) #undef POTRS_GROUP_LAUNCHER_SCRATCH template -inline void ungqr_batch_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, int *scratch_size) { - auto e = queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { +inline void ungqr_batch_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, int* scratch_size) { + auto e = queue.submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int group_scratch_size = 0; *scratch_size = 0; @@ -1982,15 +1984,15 @@ inline void ungqr_batch_scratchpad_size(const char *func_name, Func func, sycl:: e.wait(); } -#define UNGQR_GROUP_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ - std::int64_t ungqr_batch_scratchpad_size( \ - sycl::queue & queue, std::int64_t * m, std::int64_t * n, std::int64_t * k, \ - std::int64_t * lda, std::int64_t group_count, std::int64_t * group_sizes) { \ - int scratch_size; \ - ungqr_batch_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, lda, \ - group_count, group_sizes, &scratch_size); \ - return scratch_size; \ +#define UNGQR_GROUP_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ + std::int64_t ungqr_batch_scratchpad_size( \ + sycl::queue & queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, std::int64_t* lda, \ + std::int64_t group_count, std::int64_t* group_sizes) { \ + int scratch_size; \ + ungqr_batch_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, lda, \ + group_count, group_sizes, &scratch_size); \ + return scratch_size; \ } UNGQR_GROUP_LAUNCHER_SCRATCH(std::complex, cusolverDnCungqr_bufferSize) diff --git a/src/lapack/backends/cusolver/cusolver_handle.hpp b/src/lapack/backends/cusolver/cusolver_handle.hpp index 269295cd5..53be426f5 100644 --- a/src/lapack/backends/cusolver/cusolver_handle.hpp +++ b/src/lapack/backends/cusolver/cusolver_handle.hpp @@ -28,10 +28,10 @@ namespace cusolver { template struct cusolver_handle { - using handle_container_t = std::unordered_map *>; + using handle_container_t = std::unordered_map*>; handle_container_t cusolver_handle_mapper_{}; ~cusolver_handle() noexcept(false) { - for (auto &handle_pair : cusolver_handle_mapper_) { + for (auto& handle_pair : cusolver_handle_mapper_) { cusolverStatus_t err; if (handle_pair.second != nullptr) { auto handle = handle_pair.second->exchange(nullptr); diff --git a/src/lapack/backends/cusolver/cusolver_helper.hpp b/src/lapack/backends/cusolver/cusolver_helper.hpp index f51827f9f..52da00bb6 100644 --- a/src/lapack/backends/cusolver/cusolver_helper.hpp +++ b/src/lapack/backends/cusolver/cusolver_helper.hpp @@ -82,7 +82,7 @@ void overflow_check(Index index, Next... indices) { class cusolver_error : virtual public std::runtime_error { protected: - inline const char *cusolver_error_map(cusolverStatus_t error) { + inline const char* cusolver_error_map(cusolverStatus_t error) { switch (error) { case CUSOLVER_STATUS_SUCCESS: return "CUSOLVER_STATUS_SUCCESS"; @@ -131,7 +131,7 @@ class cusolver_error : virtual public std::runtime_error { class cuda_error : virtual public std::runtime_error { protected: - inline const char *cuda_error_map(CUresult result) { + inline const char* cuda_error_map(CUresult result) { switch (result) { case CUDA_SUCCESS: return "CUDA_SUCCESS"; case CUDA_ERROR_NOT_PERMITTED: return "CUDA_ERROR_NOT_PERMITTED"; @@ -201,9 +201,8 @@ class cuda_error : virtual public std::runtime_error { CUSOLVER_SYNC(err, handle) template -inline void cusolver_native_named_func(const char *func_name, Func func, - cusolverStatus_t err, - cusolverDnHandle_t handle, Types... args){ +inline void cusolver_native_named_func(const char* func_name, Func func, cusolverStatus_t err, + cusolverDnHandle_t handle, Types... args) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND CUSOLVER_ERROR_FUNC_T(func_name, func, err, handle, args...) #else @@ -291,25 +290,25 @@ struct CudaEquivalentType> { /* devinfo */ -inline void get_cusolver_devinfo(sycl::queue &queue, sycl::buffer &devInfo, - std::vector &dev_info_) { +inline void get_cusolver_devinfo(sycl::queue& queue, sycl::buffer& devInfo, + std::vector& dev_info_) { sycl::host_accessor dev_info_acc{ devInfo }; for (unsigned int i = 0; i < dev_info_.size(); ++i) dev_info_[i] = dev_info_acc[i]; } -inline void get_cusolver_devinfo(sycl::queue &queue, const int *devInfo, - std::vector &dev_info_) { +inline void get_cusolver_devinfo(sycl::queue& queue, const int* devInfo, + std::vector& dev_info_) { queue.wait(); queue.memcpy(dev_info_.data(), devInfo, sizeof(int)); } template -inline void lapack_info_check(sycl::queue &queue, DEVINFO_T devinfo, const char *func_name, - const char *cufunc_name, int dev_info_size = 1) { +inline void lapack_info_check(sycl::queue& queue, DEVINFO_T devinfo, const char* func_name, + const char* cufunc_name, int dev_info_size = 1) { std::vector dev_info_(dev_info_size); get_cusolver_devinfo(queue, devinfo, dev_info_); - for (const auto &val : dev_info_) { + for (const auto& val : dev_info_) { if (val > 0) throw oneapi::math::lapack::computation_error( func_name, std::string(cufunc_name) + " failed with info = " + std::to_string(val), @@ -322,8 +321,8 @@ inline void lapack_info_check(sycl::queue &queue, DEVINFO_T devinfo, const char // Creates list of matrix/vector pointers from initial ptr and stride // Note: user is responsible for deallocating memory template -T **create_ptr_list_from_stride(T *ptr, int64_t ptr_stride, int64_t batch_size) { - T **ptr_list = (T **)malloc(sizeof(T *) * batch_size); +T** create_ptr_list_from_stride(T* ptr, int64_t ptr_stride, int64_t batch_size) { + T** ptr_list = (T**)malloc(sizeof(T*) * batch_size); for (int64_t i = 0; i < batch_size; i++) ptr_list[i] = ptr + i * ptr_stride; diff --git a/src/lapack/backends/cusolver/cusolver_lapack.cpp b/src/lapack/backends/cusolver/cusolver_lapack.cpp index 52d28edc3..6a5427712 100644 --- a/src/lapack/backends/cusolver/cusolver_lapack.cpp +++ b/src/lapack/backends/cusolver/cusolver_lapack.cpp @@ -30,10 +30,10 @@ namespace cusolver { // BUFFER APIs template -inline void gebrd(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void gebrd(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tauq, sycl::buffer& taup, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); @@ -41,21 +41,21 @@ inline void gebrd(const char *func_name, Func func, sycl::queue &queue, std::int if (m < n) throw unimplemented("lapack", "gebrd", "cusolver gebrd does not support m < n"); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto d_acc = d.template get_access(cgh); auto e_acc = e.template get_access(cgh); auto tauq_acc = tauq.template get_access(cgh); auto taup_acc = taup.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto d_ = sc.get_mem(d_acc); - auto e_ = sc.get_mem(e_acc); - auto tauq_ = sc.get_mem(tauq_acc); - auto taup_ = sc.get_mem(taup_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto d_ = sc.get_mem(d_acc); + auto e_ = sc.get_mem(e_acc); + auto tauq_ = sc.get_mem(tauq_acc); + auto taup_ = sc.get_mem(taup_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, d_, e_, tauq_, taup_, scratch_, scratchpad_size, nullptr); @@ -64,10 +64,10 @@ inline void gebrd(const char *func_name, Func func, sycl::queue &queue, std::int } #define GEBRD_LAUNCHER(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, \ - sycl::buffer &tauq, sycl::buffer &taup, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, \ + sycl::buffer& tauq, sycl::buffer& taup, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ gebrd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, d, e, tauq, taup, \ scratchpad, scratchpad_size); \ } @@ -79,43 +79,43 @@ GEBRD_LAUNCHER(std::complex, double, cusolverDnZgebrd) #undef GEBRD_LAUNCHER -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "gerqf"); } -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "gerqf"); } -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "gerqf"); } -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "gerqf"); } template -inline void geqrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void geqrf(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -124,8 +124,8 @@ inline void geqrf(const char *func_name, Func func, sycl::queue &queue, std::int } #define GEQRF_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, \ + void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ geqrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, tau, scratchpad, \ scratchpad_size); \ @@ -139,9 +139,9 @@ GEQRF_LAUNCHER(std::complex, cusolverDnZgeqrf) #undef GEQRF_LAUNCHER template -void getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void getrf(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); @@ -152,17 +152,17 @@ void getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, sycl::buffer ipiv32(sycl::range<1>{ ipiv_size }); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto ipiv32_acc = ipiv32.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto ipiv32_ = sc.get_mem(ipiv32_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto ipiv32_ = sc.get_mem(ipiv32_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, scratch_, ipiv32_, devInfo_); @@ -170,7 +170,7 @@ void getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, }); // Copy from 32-bit buffer to 64-bit - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto ipiv32_acc = ipiv32.template get_access(cgh); auto ipiv_acc = ipiv.template get_access(cgh); cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { @@ -181,8 +181,8 @@ void getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, } #define GETRF_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, \ + void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ getrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, ipiv, scratchpad, \ scratchpad_size); \ @@ -196,8 +196,8 @@ GETRF_LAUNCHER(std::complex, cusolverDnZgetrf) #undef GETRF_LAUNCHER #define GETRI_LAUNCHER(TYPE) \ - void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ - sycl::buffer &ipiv, sycl::buffer &scratchpad, \ + void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, \ + sycl::buffer& ipiv, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ return getri_batch(queue, n, a, lda, lda * n, ipiv, n, 1, scratchpad, scratchpad_size); \ } @@ -211,10 +211,10 @@ GETRI_LAUNCHER(std::complex) // cusolverDnXgetrs does not use scratchpad memory template -inline void getrs(const char *func_name, Func func, sycl::queue &queue, +inline void getrs(const char* func_name, Func func, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, nrhs, lda, ldb); @@ -225,7 +225,7 @@ inline void getrs(const char *func_name, Func func, sycl::queue &queue, std::uint64_t ipiv_size = ipiv.size(); sycl::buffer ipiv32(sycl::range<1>{ ipiv_size }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto ipiv32_acc = ipiv32.template get_access(cgh); auto ipiv_acc = ipiv.template get_access(cgh); cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { @@ -233,15 +233,15 @@ inline void getrs(const char *func_name, Func func, sycl::queue &queue, }); }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto ipiv_acc = ipiv32.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto ipiv_ = sc.get_mem(ipiv_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto ipiv_ = sc.get_mem(ipiv_acc); + auto b_ = sc.get_mem(b_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_operation(trans), n, nrhs, a_, lda, ipiv_, b_, ldb, nullptr); @@ -250,10 +250,10 @@ inline void getrs(const char *func_name, Func func, sycl::queue &queue, } #define GETRS_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, \ - sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, \ + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, \ + sycl::buffer& ipiv, sycl::buffer& b, std::int64_t ldb, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ getrs(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, trans, n, nrhs, a, lda, ipiv, b, ldb, \ scratchpad, scratchpad_size); \ } @@ -266,30 +266,30 @@ GETRS_LAUNCHER(std::complex, cusolverDnZgetrs) #undef GETRS_LAUNCHER template -inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, +inline void gesvd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; overflow_check(n, m, lda, ldu, ldvt, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto s_acc = s.template get_access(cgh); auto u_acc = u.template get_access(cgh); auto vt_acc = vt.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto s_ = sc.get_mem(s_acc); - auto u_ = sc.get_mem(u_acc); - auto vt_ = sc.get_mem(vt_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto s_ = sc.get_mem(s_acc); + auto u_ = sc.get_mem(u_acc); + auto vt_ = sc.get_mem(vt_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; // rwork is set to nullptr. If set it is filled with information from the superdiagonal. cusolver_native_named_func(func_name, func, err, handle, get_cusolver_jobsvd(jobu), @@ -301,10 +301,10 @@ inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define GESVD_LAUNCHER(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, \ - sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, \ + void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, \ + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, \ + sycl::buffer& vt, std::int64_t ldvt, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ gesvd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobu, jobvt, m, n, a, lda, s, u, ldu, \ vt, ldvt, scratchpad, scratchpad_size); \ @@ -318,25 +318,25 @@ GESVD_LAUNCHER(std::complex, double, cusolverDnZgesvd) #undef GESVD_LAUNCHER template -inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, +inline void heevd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto w_ = sc.get_mem(w_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto w_ = sc.get_mem(w_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, a_, lda, w_, scratch_, @@ -346,12 +346,12 @@ inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: lapack_info_check(queue, devInfo, __func__, func_name); } -#define HEEVD_LAUNCHER(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &w, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ - heevd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, scratchpad, \ - scratchpad_size); \ +#define HEEVD_LAUNCHER(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ + void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& w, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ + heevd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, scratchpad, \ + scratchpad_size); \ } HEEVD_LAUNCHER(std::complex, float, cusolverDnCheevd) @@ -360,28 +360,28 @@ HEEVD_LAUNCHER(std::complex, double, cusolverDnZheevd) #undef HEEVD_LAUNCHER template -inline void hegvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, +inline void hegvd(const char* func_name, Func func, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; overflow_check(n, lda, ldb, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto w_ = sc.get_mem(w_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto w_ = sc.get_mem(w_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cusolver_itype(itype), get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, a_, @@ -392,10 +392,10 @@ inline void hegvd(const char *func_name, Func func, sycl::queue &queue, std::int } #define HEGVD_LAUNCHER(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ - sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, \ + sycl::buffer& b, std::int64_t ldb, sycl::buffer& w, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ hegvd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, ldb, w, \ scratchpad, scratchpad_size); \ } @@ -406,29 +406,29 @@ HEGVD_LAUNCHER(std::complex, double, cusolverDnZhegvd) #undef HEGVD_LAUNCHER template -inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, +inline void hetrd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto d_acc = d.template get_access(cgh); auto e_acc = e.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto d_ = sc.get_mem(d_acc); - auto e_ = sc.get_mem(e_acc); - auto tau_ = sc.get_mem(tau_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto d_ = sc.get_mem(d_acc); + auto e_ = sc.get_mem(e_acc); + auto tau_ = sc.get_mem(tau_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, d_, e_, tau_, scratch_, scratchpad_size, devInfo_); @@ -438,10 +438,10 @@ inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define HETRD_LAUNCHER(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, \ - sycl::buffer &e, sycl::buffer &tau, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, \ + sycl::buffer& e, sycl::buffer& tau, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ hetrd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, d, e, tau, scratchpad, \ scratchpad_size); \ } @@ -451,34 +451,34 @@ HETRD_LAUNCHER(std::complex, double, cusolverDnZhetrd) #undef HETRD_LAUNCHER -void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "hetrf"); } -void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, +void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "hetrf"); } template -inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +inline void orgbr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_generate(vec), m, n, k, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -487,9 +487,9 @@ inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORGBR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ orgbr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, scratchpad, \ scratchpad_size); \ } @@ -500,20 +500,20 @@ ORGBR_LAUNCHER(double, cusolverDnDorgbr) #undef ORGBR_LAUNCHER template -inline void orgqr(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void orgqr(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, m, n, k, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -522,9 +522,9 @@ inline void orgqr(const char *func_name, Func func, sycl::queue &queue, std::int } #define ORGQR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ orgqr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, a, lda, tau, scratchpad, \ scratchpad_size); \ } @@ -535,20 +535,20 @@ ORGQR_LAUNCHER(double, cusolverDnDorgqr) #undef ORGQR_LAUNCHER template -inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void orgtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -556,12 +556,12 @@ inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: }); } -#define ORGTR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, \ - std::int64_t scratchpad_size) { \ - orgtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ - scratchpad_size); \ +#define ORGTR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ + void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, \ + std::int64_t scratchpad_size) { \ + orgtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ + scratchpad_size); \ } ORGTR_LAUNCHER(float, cusolverDnSorgtr) @@ -570,24 +570,24 @@ ORGTR_LAUNCHER(double, cusolverDnDorgtr) #undef ORGTR_LAUNCHER template -inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, +inline void ormtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldc, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto c_ = sc.get_mem(c_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto c_ = sc.get_mem(c_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_side_mode(side), get_cublas_fill_mode(uplo), get_cublas_operation(trans), m, @@ -598,10 +598,10 @@ inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORMTR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, \ + void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ ormtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, uplo, trans, m, n, a, lda, tau, c, \ ldc, scratchpad, scratchpad_size); \ @@ -612,37 +612,37 @@ ORMTR_LAUNCHER(double, cusolverDnDormtr) #undef ORMTR_LAUNCHER -void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "ormrq"); } -void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "ormrq"); } template -inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, +inline void ormqr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, ldc, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto c_ = sc.get_mem(c_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto c_ = sc.get_mem(c_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_side_mode(side), get_cublas_operation(trans), m, n, k, a_, lda, tau_, c_, ldc, @@ -652,10 +652,10 @@ inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORMQR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ ormqr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, tau, c, \ ldc, scratchpad, scratchpad_size); \ } @@ -666,21 +666,21 @@ ORMQR_LAUNCHER(double, cusolverDnDormqr) #undef ORMQR_LAUNCHER template -inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, +inline void potrf(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, scratch_, scratchpad_size, devInfo_); @@ -689,11 +689,11 @@ inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: lapack_info_check(queue, devInfo, __func__, func_name); } -#define POTRF_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ - potrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ - scratchpad_size); \ +#define POTRF_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ + void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ + potrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ + scratchpad_size); \ } POTRF_LAUNCHER(float, cusolverDnSpotrf) @@ -704,21 +704,21 @@ POTRF_LAUNCHER(std::complex, cusolverDnZpotrf) #undef POTRF_LAUNCHER template -inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, +inline void potri(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, scratch_, scratchpad_size, devInfo_); @@ -727,11 +727,11 @@ inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi:: lapack_info_check(queue, devInfo, __func__, func_name); } -#define POTRI_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ - potri(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ - scratchpad_size); \ +#define POTRI_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ + void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ + potri(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ + scratchpad_size); \ } POTRI_LAUNCHER(float, cusolverDnSpotri) @@ -743,19 +743,19 @@ POTRI_LAUNCHER(std::complex, cusolverDnZpotri) // cusolverDnXpotrs does not use scratchpad memory template -inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, +inline void potrs(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, nrhs, lda, ldb, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, nrhs, a_, lda, b_, ldb, nullptr); @@ -764,9 +764,9 @@ inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define POTRS_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ potrs(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, nrhs, a, lda, b, ldb, \ scratchpad, scratchpad_size); \ } @@ -779,23 +779,23 @@ POTRS_LAUNCHER(std::complex, cusolverDnZpotrs) #undef POTRS_LAUNCHER template -inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void syevd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto w_ = sc.get_mem(w_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto w_ = sc.get_mem(w_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, a_, lda, w_, scratch_, @@ -805,12 +805,12 @@ inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: lapack_info_check(queue, devInfo, __func__, func_name); } -#define SYEVD_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &w, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ - syevd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, scratchpad, \ - scratchpad_size); \ +#define SYEVD_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ + void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& w, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ + syevd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, scratchpad, \ + scratchpad_size); \ } SYEVD_LAUNCHER(float, cusolverDnSsyevd) @@ -819,26 +819,26 @@ SYEVD_LAUNCHER(double, cusolverDnDsyevd) #undef SYEVD_LAUNCHER template -inline void sygvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void sygvd(const char* func_name, Func func, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, ldb, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto w_ = sc.get_mem(w_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto w_ = sc.get_mem(w_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cusolver_itype(itype), get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, a_, @@ -849,10 +849,10 @@ inline void sygvd(const char *func_name, Func func, sycl::queue &queue, std::int } #define SYGVD_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ - sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, \ + sycl::buffer& b, std::int64_t ldb, sycl::buffer& w, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ sygvd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, ldb, w, \ scratchpad, scratchpad_size); \ } @@ -863,28 +863,28 @@ SYGVD_LAUNCHER(double, cusolverDnDsygvd) #undef SYGVD_LAUNCH template -inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, +inline void sytrd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto d_acc = d.template get_access(cgh); auto e_acc = e.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto d_ = sc.get_mem(d_acc); - auto e_ = sc.get_mem(e_acc); - auto tau_ = sc.get_mem(tau_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto d_ = sc.get_mem(d_acc); + auto e_ = sc.get_mem(e_acc); + auto tau_ = sc.get_mem(tau_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, d_, e_, tau_, scratch_, scratchpad_size, devInfo_); @@ -893,13 +893,13 @@ inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: lapack_info_check(queue, devInfo, __func__, func_name); } -#define SYTRD_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, \ - sycl::buffer &tau, sycl::buffer &scratchpad, \ - std::int64_t scratchpad_size) { \ - sytrd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, d, e, tau, scratchpad, \ - scratchpad_size); \ +#define SYTRD_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ + void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, \ + sycl::buffer& tau, sycl::buffer& scratchpad, \ + std::int64_t scratchpad_size) { \ + sytrd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, d, e, tau, scratchpad, \ + scratchpad_size); \ } SYTRD_LAUNCHER(float, cusolverDnSsytrd) @@ -908,9 +908,9 @@ SYTRD_LAUNCHER(double, cusolverDnDsytrd) #undef SYTRD_LAUNCHER template -inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, +inline void sytrf(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); @@ -922,17 +922,17 @@ inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: std::uint64_t ipiv_size = n; sycl::buffer ipiv32(sycl::range<1>{ ipiv_size }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto ipiv32_acc = ipiv32.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto ipiv32_ = sc.get_mem(ipiv32_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto ipiv32_ = sc.get_mem(ipiv32_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, ipiv32_, scratch_, scratchpad_size, devInfo_); @@ -940,7 +940,7 @@ inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: }); // Copy from 32-bit buffer to 64-bit - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto ipiv32_acc = ipiv32.template get_access(cgh); auto ipiv_acc = ipiv.template get_access(cgh); cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { @@ -951,8 +951,8 @@ inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define SYTRF_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, \ + void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ sytrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, ipiv, scratchpad, \ scratchpad_size); \ @@ -965,49 +965,49 @@ SYTRF_LAUNCHER(std::complex, cusolverDnZsytrf) #undef SYTRF_LAUNCHER -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } template -inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +inline void ungbr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_generate(vec), m, n, k, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -1016,9 +1016,9 @@ inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNGBR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ ungbr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, scratchpad, \ scratchpad_size); \ } @@ -1029,20 +1029,20 @@ UNGBR_LAUNCHER(std::complex, cusolverDnZungbr) #undef UNGBR_LAUNCHER template -inline void ungqr(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void ungqr(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, m, n, k, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -1051,9 +1051,9 @@ inline void ungqr(const char *func_name, Func func, sycl::queue &queue, std::int } #define UNGQR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ ungqr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, a, lda, tau, scratchpad, \ scratchpad_size); \ } @@ -1064,20 +1064,20 @@ UNGQR_LAUNCHER(std::complex, cusolverDnZungqr) #undef UNGQR_LAUNCHER template -inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void ungtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -1085,12 +1085,12 @@ inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: }); } -#define UNGTR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, \ - std::int64_t scratchpad_size) { \ - ungtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ - scratchpad_size); \ +#define UNGTR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ + void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, \ + std::int64_t scratchpad_size) { \ + ungtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ + scratchpad_size); \ } UNGTR_LAUNCHER(std::complex, cusolverDnCungtr) @@ -1098,39 +1098,39 @@ UNGTR_LAUNCHER(std::complex, cusolverDnZungtr) #undef UNGTR_LAUNCHER -void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size) { +void unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "unmrq"); } -void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size) { +void unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "unmrq"); } template -inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, +inline void unmqr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto c_ = sc.get_mem(c_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto c_ = sc.get_mem(c_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_side_mode(side), get_cublas_operation(trans), m, n, k, a_, lda, tau_, c_, ldc, @@ -1140,10 +1140,10 @@ inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNMQR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ unmqr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, tau, c, \ ldc, scratchpad, scratchpad_size); \ } @@ -1154,24 +1154,24 @@ UNMQR_LAUNCHER(std::complex, cusolverDnZunmqr) #undef UNMQR_LAUNCHER template -inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, +inline void unmtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldc, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto c_ = sc.get_mem(c_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto c_ = sc.get_mem(c_acc); + auto scratch_ = sc.get_mem(scratch_acc); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_side_mode(side), get_cublas_fill_mode(uplo), get_cublas_operation(trans), m, @@ -1182,10 +1182,10 @@ inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNMTR_LAUNCHER(TYPE, CUSOLVER_ROUTINE) \ - void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, \ + void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ unmtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, uplo, trans, m, n, a, lda, tau, c, \ ldc, scratchpad, scratchpad_size); \ @@ -1199,10 +1199,10 @@ UNMTR_LAUNCHER(std::complex, cusolverDnZunmtr) // USM APIs template -inline sycl::event gebrd(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, T_A *a, std::int64_t lda, T_B *d, T_B *e, T_A *tauq, - T_A *taup, T_A *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event gebrd(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, T_A* a, std::int64_t lda, T_B* d, T_B* e, T_A* tauq, + T_A* taup, T_A* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); @@ -1210,19 +1210,19 @@ inline sycl::event gebrd(const char *func_name, Func func, sycl::queue &queue, s if (m < n) throw unimplemented("lapack", "gebrd", "cusolver gebrd does not support m < n"); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto d_ = reinterpret_cast(d); - auto e_ = reinterpret_cast(e); - auto tauq_ = reinterpret_cast(tauq); - auto taup_ = reinterpret_cast(taup); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto d_ = reinterpret_cast(d); + auto e_ = reinterpret_cast(e); + auto tauq_ = reinterpret_cast(tauq); + auto taup_ = reinterpret_cast(taup); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, d_, e_, tauq_, taup_, scratch_, scratchpad_size, nullptr); @@ -1232,10 +1232,10 @@ inline sycl::event gebrd(const char *func_name, Func func, sycl::queue &queue, s } #define GEBRD_LAUNCHER_USM(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, TYPE_A *a, \ - std::int64_t lda, TYPE_B *d, TYPE_B *e, TYPE_A *tauq, TYPE_A *taup, \ - TYPE_A *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, TYPE_A* a, \ + std::int64_t lda, TYPE_B* d, TYPE_B* e, TYPE_A* tauq, TYPE_A* taup, \ + TYPE_A* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return gebrd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, d, e, tauq, taup, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1247,44 +1247,44 @@ GEBRD_LAUNCHER_USM(std::complex, double, cusolverDnZgebrd) #undef GEBRD_LAUNCHER_USM -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "gerqf"); } -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "gerqf"); } -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "gerqf"); } -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "gerqf"); } template -inline sycl::event geqrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, T *a, std::int64_t lda, T *tau, T *scratchpad, +inline sycl::event geqrf(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, T* a, std::int64_t lda, T* tau, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -1294,9 +1294,9 @@ inline sycl::event geqrf(const char *func_name, Func func, sycl::queue &queue, s } #define GEQRF_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return geqrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, tau, scratchpad, \ scratchpad_size, dependencies); \ } @@ -1309,10 +1309,10 @@ GEQRF_LAUNCHER_USM(std::complex, cusolverDnZgeqrf) #undef GEQRF_LAUNCHER_USM template -inline sycl::event getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, T *a, std::int64_t lda, std::int64_t *ipiv, T *scratchpad, +inline sycl::event getrf(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, T* a, std::int64_t lda, std::int64_t* ipiv, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); @@ -1320,20 +1320,20 @@ inline sycl::event getrf(const char *func_name, Func func, sycl::queue &queue, s // To get around the limitation. // Allocate memory with 32-bit ints then copy over results std::uint64_t ipiv_size = std::min(n, m); - int *ipiv32 = (int *)malloc_device(sizeof(int) * ipiv_size, queue); + int* ipiv32 = (int*)malloc_device(sizeof(int) * ipiv_size, queue); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); - auto ipiv_ = reinterpret_cast(ipiv32); + auto a_ = reinterpret_cast(a); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); + auto ipiv_ = reinterpret_cast(ipiv32); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, scratch_, ipiv_, devInfo_); @@ -1341,7 +1341,7 @@ inline sycl::event getrf(const char *func_name, Func func, sycl::queue &queue, s }); // Copy from 32-bit USM to 64-bit - auto done_casting = queue.submit([&](sycl::handler &cgh) { + auto done_casting = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { ipiv[index] = static_cast(ipiv32[index]); @@ -1358,10 +1358,10 @@ inline sycl::event getrf(const char *func_name, Func func, sycl::queue &queue, s } #define GETRF_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, TYPE *a, \ - std::int64_t lda, std::int64_t *ipiv, TYPE *scratchpad, \ + sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, TYPE* a, \ + std::int64_t lda, std::int64_t* ipiv, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return getrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, a, lda, ipiv, scratchpad, \ scratchpad_size, dependencies); \ } @@ -1374,9 +1374,9 @@ GETRF_LAUNCHER_USM(std::complex, cusolverDnZgetrf) #undef GETRF_LAUNCHER_USM #define GETRI_LAUNCHER_USM(TYPE) \ - sycl::event getri(sycl::queue &queue, std::int64_t n, TYPE *a, std::int64_t lda, \ - std::int64_t *ipiv, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event getri(sycl::queue& queue, std::int64_t n, TYPE* a, std::int64_t lda, \ + std::int64_t* ipiv, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return getri_batch(queue, n, a, lda, lda * n, ipiv, n, 1, scratchpad, scratchpad_size, \ dependencies); \ } @@ -1390,11 +1390,11 @@ GETRI_LAUNCHER_USM(std::complex) // cusolverDnXgetrs does not use scratchpad memory template -inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, T *a, - std::int64_t lda, std::int64_t *ipiv, T *b, std::int64_t ldb, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event getrs(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, T* a, + std::int64_t lda, std::int64_t* ipiv, T* b, std::int64_t ldb, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, nrhs, lda, ldb, scratchpad_size); @@ -1402,25 +1402,25 @@ inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, // To get around the limitation. // Create new buffer and convert 64-bit values. std::uint64_t ipiv_size = n; - int *ipiv32 = (int *)malloc_device(sizeof(int) * ipiv_size, queue); + int* ipiv32 = (int*)malloc_device(sizeof(int) * ipiv_size, queue); - auto done_casting = queue.submit([&](sycl::handler &cgh) { + auto done_casting = queue.submit([&](sycl::handler& cgh) { cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { ipiv32[index] = static_cast(ipiv[index]); }); }); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } cgh.depends_on(done_casting); - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto ipiv_ = reinterpret_cast(ipiv32); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto ipiv_ = reinterpret_cast(ipiv32); + auto b_ = reinterpret_cast(b); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_operation(trans), n, nrhs, a_, lda, ipiv_, b_, ldb, nullptr); @@ -1435,10 +1435,10 @@ inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, } #define GETRS_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ - std::int64_t nrhs, TYPE *a, std::int64_t lda, std::int64_t *ipiv, TYPE *b, \ - std::int64_t ldb, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, \ + std::int64_t nrhs, TYPE* a, std::int64_t lda, std::int64_t* ipiv, TYPE* b, \ + std::int64_t ldb, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return getrs(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, trans, n, nrhs, a, lda, ipiv, b, \ ldb, scratchpad, scratchpad_size, dependencies); \ } @@ -1451,28 +1451,28 @@ GETRS_LAUNCHER_USM(std::complex, cusolverDnZgetrs) #undef GETRS_LAUNCHER_USM template -inline sycl::event gesvd(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event gesvd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, T_A *a, std::int64_t lda, T_B *s, T_A *u, std::int64_t ldu, - T_A *vt, std::int64_t ldvt, T_A *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t n, T_A* a, std::int64_t lda, T_B* s, T_A* u, std::int64_t ldu, + T_A* vt, std::int64_t ldvt, T_A* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldu, ldvt, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto s_ = reinterpret_cast(s); - auto u_ = reinterpret_cast(u); - auto vt_ = reinterpret_cast(vt); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto s_ = reinterpret_cast(s); + auto u_ = reinterpret_cast(u); + auto vt_ = reinterpret_cast(vt); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; // rwork is set to nullptr. If set it is filled with information from the superdiagonal. cusolver_native_named_func(func_name, func, err, handle, get_cusolver_jobsvd(jobu), @@ -1486,11 +1486,11 @@ inline sycl::event gesvd(const char *func_name, Func func, sycl::queue &queue, } #define GESVD_LAUNCHER_USM(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ - std::int64_t m, std::int64_t n, TYPE_A *a, std::int64_t lda, TYPE_B *s, \ - TYPE_A *u, std::int64_t ldu, TYPE_A *vt, std::int64_t ldvt, \ - TYPE_A *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ + std::int64_t m, std::int64_t n, TYPE_A* a, std::int64_t lda, TYPE_B* s, \ + TYPE_A* u, std::int64_t ldu, TYPE_A* vt, std::int64_t ldvt, \ + TYPE_A* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return gesvd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobu, jobvt, m, n, a, lda, s, u, \ ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); \ } @@ -1503,25 +1503,25 @@ GESVD_LAUNCHER_USM(std::complex, double, cusolverDnZgesvd) #undef GESVD_LAUNCHER_USM template -inline sycl::event heevd(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A *&a, - std::int64_t lda, T_B *&w, T_A *&scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event heevd(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A*& a, + std::int64_t lda, T_B*& w, T_A*& scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto w_ = reinterpret_cast(w); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto w_ = reinterpret_cast(w); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, a_, lda, w_, scratch_, @@ -1534,10 +1534,10 @@ inline sycl::event heevd(const char *func_name, Func func, sycl::queue &queue, } #define HEEVD_LAUNCHER_USM(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ - std::int64_t n, TYPE_A *a, std::int64_t lda, TYPE_B *w, TYPE_A *scratchpad, \ + sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ + std::int64_t n, TYPE_A* a, std::int64_t lda, TYPE_B* w, TYPE_A* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return heevd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1548,27 +1548,27 @@ HEEVD_LAUNCHER_USM(std::complex, double, cusolverDnZheevd) #undef HEEVD_LAUNCHER_USM template -inline sycl::event hegvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A *&a, - std::int64_t lda, T_A *&b, std::int64_t ldb, T_B *&w, T_A *&scratchpad, +inline sycl::event hegvd(const char* func_name, Func func, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A*& a, + std::int64_t lda, T_A*& b, std::int64_t ldb, T_B*& w, T_A*& scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; overflow_check(n, lda, ldb, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto w_ = reinterpret_cast(w); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto w_ = reinterpret_cast(w); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cusolver_itype(itype), get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, a_, @@ -1581,11 +1581,11 @@ inline sycl::event hegvd(const char *func_name, Func func, sycl::queue &queue, s } #define HEGVD_LAUNCHER_USM(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ - oneapi::math::uplo uplo, std::int64_t n, TYPE_A *a, std::int64_t lda, \ - TYPE_A *b, std::int64_t ldb, TYPE_B *w, TYPE_A *scratchpad, \ + sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, TYPE_A* a, std::int64_t lda, \ + TYPE_A* b, std::int64_t ldb, TYPE_B* w, TYPE_A* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return hegvd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, \ ldb, w, scratchpad, scratchpad_size, dependencies); \ } @@ -1596,27 +1596,27 @@ HEGVD_LAUNCHER_USM(std::complex, double, cusolverDnZhegvd) #undef HEGVD_LAUNCHER_USM template -inline sycl::event hetrd(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T_A *a, std::int64_t lda, T_B *d, - T_B *e, T_A *tau, T_A *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event hetrd(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T_A* a, std::int64_t lda, T_B* d, + T_B* e, T_A* tau, T_A* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType_A = typename CudaEquivalentType::Type; using cuDataType_B = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto d_ = reinterpret_cast(d); - auto e_ = reinterpret_cast(e); - auto tau_ = reinterpret_cast(tau); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto d_ = reinterpret_cast(d); + auto e_ = reinterpret_cast(e); + auto tau_ = reinterpret_cast(tau); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, d_, e_, tau_, scratch_, scratchpad_size, devInfo_); @@ -1628,10 +1628,10 @@ inline sycl::event hetrd(const char *func_name, Func func, sycl::queue &queue, } #define HETRD_LAUNCHER_USM(TYPE_A, TYPE_B, CUSOLVER_ROUTINE) \ - sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE_A *a, \ - std::int64_t lda, TYPE_B *d, TYPE_B *e, TYPE_A *tau, TYPE_A *scratchpad, \ + sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE_A* a, \ + std::int64_t lda, TYPE_B* d, TYPE_B* e, TYPE_A* tau, TYPE_A* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return hetrd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, d, e, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1641,37 +1641,37 @@ HETRD_LAUNCHER_USM(std::complex, double, cusolverDnZhetrd) #undef HETRD_LAUNCHER_USM -sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "hetrf"); } -sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "hetrf"); } template -inline sycl::event orgbr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event orgbr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - T *a, std::int64_t lda, T *tau, T *scratchpad, + T* a, std::int64_t lda, T* tau, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_generate(vec), m, n, k, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -1681,10 +1681,10 @@ inline sycl::event orgbr(const char *func_name, Func func, sycl::queue &queue, } #define ORGBR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, \ - std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, TYPE *tau, \ - TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, \ + std::int64_t n, std::int64_t k, TYPE* a, std::int64_t lda, TYPE* tau, \ + TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return orgbr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1695,22 +1695,22 @@ ORGBR_LAUNCHER_USM(double, cusolverDnDorgbr) #undef ORGBR_LAUNCHER_USM template -inline sycl::event orgqr(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event orgqr(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, T* a, std::int64_t lda, T* tau, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, m, n, k, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -1720,9 +1720,9 @@ inline sycl::event orgqr(const char *func_name, Func func, sycl::queue &queue, s } #define ORGQR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return orgqr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, a, lda, tau, scratchpad, \ scratchpad_size, dependencies); \ } @@ -1733,22 +1733,22 @@ ORGQR_LAUNCHER_USM(double, cusolverDnDorgqr) #undef ORGQR_LAUNCHER_USM template -inline sycl::event orgtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event orgtr(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, T* tau, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -1758,9 +1758,9 @@ inline sycl::event orgtr(const char *func_name, Func func, sycl::queue &queue, } #define ORGTR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return orgtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ scratchpad_size, dependencies); \ } @@ -1771,25 +1771,25 @@ ORGTR_LAUNCHER_USM(double, cusolverDnDorgtr) #undef ORGTR_LAUNCHER_USM template -inline sycl::event ormtr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event ormtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T *a, - std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T* a, + std::int64_t lda, T* tau, T* c, std::int64_t ldc, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldc, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto c_ = reinterpret_cast(c); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto c_ = reinterpret_cast(c); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_side_mode(side), get_cublas_fill_mode(uplo), get_cublas_operation(trans), m, @@ -1801,11 +1801,11 @@ inline sycl::event ormtr(const char *func_name, Func func, sycl::queue &queue, } #define ORMTR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ + sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* c, std::int64_t ldc, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return ormtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, uplo, trans, m, n, a, lda, \ tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ } @@ -1815,38 +1815,38 @@ ORMTR_LAUNCHER_USM(double, cusolverDnDormtr) #undef ORMTR_LAUNCHER_USM -sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *c, std::int64_t ldc, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "ormrq"); } -sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *c, std::int64_t ldc, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "ormrq"); } template -inline sycl::event ormqr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event ormqr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, - std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, T* a, std::int64_t lda, T* tau, T* c, + std::int64_t ldc, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, ldc, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto c_ = reinterpret_cast(c); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto c_ = reinterpret_cast(c); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_side_mode(side), get_cublas_operation(trans), m, n, k, a_, lda, tau_, c_, ldc, @@ -1856,14 +1856,14 @@ inline sycl::event ormqr(const char *func_name, Func func, sycl::queue &queue, return done; } -#define ORMQR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, \ - TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ - std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ - return ormqr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, \ - tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ +#define ORMQR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ + sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, TYPE* a, std::int64_t lda, \ + TYPE* tau, TYPE* c, std::int64_t ldc, TYPE* scratchpad, \ + std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ + return ormqr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, \ + tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ } ORMQR_LAUNCHER_USM(float, cusolverDnSormqr) @@ -1872,23 +1872,23 @@ ORMQR_LAUNCHER_USM(double, cusolverDnDormqr) #undef ORMQR_LAUNCHER_USM template -inline sycl::event potrf(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event potrf(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, scratch_, scratchpad_size, devInfo_); @@ -1900,9 +1900,9 @@ inline sycl::event potrf(const char *func_name, Func func, sycl::queue &queue, } #define POTRF_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return potrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ scratchpad_size, dependencies); \ } @@ -1915,23 +1915,23 @@ POTRF_LAUNCHER_USM(std::complex, cusolverDnZpotrf) #undef POTRF_LAUNCHER_USM template -inline sycl::event potri(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event potri(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto scratch_ = reinterpret_cast(scratchpad); - auto devInfo_ = reinterpret_cast(devInfo); + auto a_ = reinterpret_cast(a); + auto scratch_ = reinterpret_cast(scratchpad); + auto devInfo_ = reinterpret_cast(devInfo); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, scratch_, scratchpad_size, devInfo_); @@ -1943,9 +1943,9 @@ inline sycl::event potri(const char *func_name, Func func, sycl::queue &queue, } #define POTRI_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return potri(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ scratchpad_size, dependencies); \ } @@ -1959,22 +1959,22 @@ POTRI_LAUNCHER_USM(std::complex, cusolverDnZpotri) // cusolverDnXpotrs does not use scratchpad memory template -inline sycl::event potrs(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, T *a, - std::int64_t lda, T *b, std::int64_t ldb, T *scratchpad, +inline sycl::event potrs(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, T* a, + std::int64_t lda, T* b, std::int64_t ldb, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, nrhs, lda, ldb, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, nrhs, a_, lda, b_, ldb, nullptr); @@ -1984,10 +1984,10 @@ inline sycl::event potrs(const char *func_name, Func func, sycl::queue &queue, } #define POTRS_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ - std::int64_t nrhs, TYPE *a, std::int64_t lda, TYPE *b, std::int64_t ldb, \ - TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, \ + std::int64_t nrhs, TYPE* a, std::int64_t lda, TYPE* b, std::int64_t ldb, \ + TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return potrs(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, nrhs, a, lda, b, ldb, \ scratchpad, scratchpad_size, dependencies); \ } @@ -2000,24 +2000,24 @@ POTRS_LAUNCHER_USM(std::complex, cusolverDnZpotrs) #undef POTRS_LAUNCHER_USM template -inline sycl::event syevd(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T *a, - std::int64_t lda, T *w, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event syevd(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T* a, + std::int64_t lda, T* w, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto w_ = reinterpret_cast(w); - auto scratch_ = reinterpret_cast(scratchpad); - auto devInfo_ = reinterpret_cast(devInfo); + auto a_ = reinterpret_cast(a); + auto w_ = reinterpret_cast(w); + auto scratch_ = reinterpret_cast(scratchpad); + auto devInfo_ = reinterpret_cast(devInfo); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, a_, lda, w_, scratch_, @@ -2030,10 +2030,10 @@ inline sycl::event syevd(const char *func_name, Func func, sycl::queue &queue, } #define SYEVD_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ - std::int64_t n, TYPE *a, std::int64_t lda, TYPE *w, TYPE *scratchpad, \ + sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ + std::int64_t n, TYPE* a, std::int64_t lda, TYPE* w, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return syevd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, \ scratchpad, scratchpad_size, dependencies); \ } @@ -2044,26 +2044,26 @@ SYEVD_LAUNCHER_USM(double, cusolverDnDsyevd) #undef SYEVD_LAUNCHER_USM template -inline sycl::event sygvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T *a, - std::int64_t lda, T *b, std::int64_t ldb, T *w, T *scratchpad, +inline sycl::event sygvd(const char* func_name, Func func, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T* a, + std::int64_t lda, T* b, std::int64_t ldb, T* w, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, ldb, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto w_ = reinterpret_cast(w); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto w_ = reinterpret_cast(w); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cusolver_itype(itype), get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, a_, @@ -2075,13 +2075,13 @@ inline sycl::event sygvd(const char *func_name, Func func, sycl::queue &queue, s return done; } -#define SYGVD_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ - oneapi::math::uplo uplo, std::int64_t n, TYPE *a, std::int64_t lda, TYPE *b, \ - std::int64_t ldb, TYPE *w, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ - return sygvd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, \ - ldb, w, scratchpad, scratchpad_size, dependencies); \ +#define SYGVD_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ + sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, TYPE* a, std::int64_t lda, TYPE* b, \ + std::int64_t ldb, TYPE* w, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ + return sygvd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, \ + ldb, w, scratchpad, scratchpad_size, dependencies); \ } SYGVD_LAUNCHER_USM(float, cusolverDnSsygvd) @@ -2090,26 +2090,26 @@ SYGVD_LAUNCHER_USM(double, cusolverDnDsygvd) #undef SYGVD_LAUNCHER_USM template -inline sycl::event sytrd(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *d, T *e, - T *tau, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event sytrd(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, T* d, + T* e, T* tau, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto d_ = reinterpret_cast(d); - auto e_ = reinterpret_cast(e); - auto tau_ = reinterpret_cast(tau); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto d_ = reinterpret_cast(d); + auto e_ = reinterpret_cast(e); + auto tau_ = reinterpret_cast(tau); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, d_, e_, tau_, scratch_, scratchpad_size, devInfo_); @@ -2121,10 +2121,10 @@ inline sycl::event sytrd(const char *func_name, Func func, sycl::queue &queue, } #define SYTRD_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *d, TYPE *e, TYPE *tau, TYPE *scratchpad, \ + sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* d, TYPE* e, TYPE* tau, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return sytrd(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, d, e, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -2135,31 +2135,31 @@ SYTRD_LAUNCHER_USM(double, cusolverDnDsytrd) #undef SYTRD_LAUNCHER_USM template -inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, - std::int64_t *ipiv, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event sytrf(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, + std::int64_t* ipiv, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); + int* devInfo = (int*)malloc_device(sizeof(int), queue); // cuSolver legacy api does not accept 64-bit ints. // To get around the limitation. // Allocate memory with 32-bit ints then copy over results std::uint64_t ipiv_size = n; - int *ipiv32 = (int *)malloc_device(sizeof(int) * ipiv_size, queue); + int* ipiv32 = (int*)malloc_device(sizeof(int) * ipiv_size, queue); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto scratch_ = reinterpret_cast(scratchpad); - auto ipiv_ = reinterpret_cast(ipiv32); - auto devInfo_ = reinterpret_cast(devInfo); + auto a_ = reinterpret_cast(a); + auto scratch_ = reinterpret_cast(scratchpad); + auto ipiv_ = reinterpret_cast(ipiv32); + auto devInfo_ = reinterpret_cast(devInfo); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, ipiv_, scratch_, scratchpad_size, devInfo_); @@ -2167,7 +2167,7 @@ inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, }); // Copy from 32-bit USM to 64-bit - auto done_casting = queue.submit([&](sycl::handler &cgh) { + auto done_casting = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { ipiv[index] = static_cast(ipiv32[index]); @@ -2183,13 +2183,13 @@ inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, return done_casting; } -#define SYTRF_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, std::int64_t *ipiv, TYPE *scratchpad, \ - std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ - return sytrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, ipiv, \ - scratchpad, scratchpad_size, dependencies); \ +#define SYTRF_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ + sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, std::int64_t* ipiv, TYPE* scratchpad, \ + std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ + return sytrf(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, ipiv, \ + scratchpad, scratchpad_size, dependencies); \ } SYTRF_LAUNCHER_USM(float, cusolverDnSsytrf) @@ -2199,51 +2199,51 @@ SYTRF_LAUNCHER_USM(std::complex, cusolverDnZsytrf) #undef SYTRF_LAUNCHER_USM -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, - std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double* a, + std::int64_t lda, double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float* a, + std::int64_t lda, float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "trtrs"); } template -inline sycl::event ungbr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event ungbr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - T *a, std::int64_t lda, T *tau, T *scratchpad, + T* a, std::int64_t lda, T* tau, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_generate(vec), m, n, k, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -2253,10 +2253,10 @@ inline sycl::event ungbr(const char *func_name, Func func, sycl::queue &queue, } #define UNGBR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, \ - std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, TYPE *tau, \ - TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, \ + std::int64_t n, std::int64_t k, TYPE* a, std::int64_t lda, TYPE* tau, \ + TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return ungbr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -2267,22 +2267,22 @@ UNGBR_LAUNCHER_USM(std::complex, cusolverDnZungbr) #undef UNGBR_LAUNCHER_USM template -inline sycl::event ungqr(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event ungqr(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, T* a, std::int64_t lda, T* tau, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, m, n, k, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -2292,9 +2292,9 @@ inline sycl::event ungqr(const char *func_name, Func func, sycl::queue &queue, s } #define UNGQR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return ungqr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, m, n, k, a, lda, tau, scratchpad, \ scratchpad_size, dependencies); \ } @@ -2305,22 +2305,22 @@ UNGQR_LAUNCHER_USM(std::complex, cusolverDnZungqr) #undef UNGQR_LAUNCHER_USM template -inline sycl::event ungtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event ungtr(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, T* tau, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, a_, lda, tau_, scratch_, scratchpad_size, nullptr); @@ -2330,9 +2330,9 @@ inline sycl::event ungtr(const char *func_name, Func func, sycl::queue &queue, } #define UNGTR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return ungtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ scratchpad_size, dependencies); \ } @@ -2342,40 +2342,40 @@ UNGTR_LAUNCHER_USM(std::complex, cusolverDnZungtr) #undef UNGTR_LAUNCHER_USM -sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "unmrq"); } -sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "unmrq"); } template -inline sycl::event unmqr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event unmqr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, - std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, T* a, std::int64_t lda, T* tau, T* c, + std::int64_t ldc, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto c_ = reinterpret_cast(c); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto c_ = reinterpret_cast(c); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_side_mode(side), get_cublas_operation(trans), m, n, k, a_, lda, tau_, c_, ldc, @@ -2385,14 +2385,14 @@ inline sycl::event unmqr(const char *func_name, Func func, sycl::queue &queue, return done; } -#define UNMQR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, \ - TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ - std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ - return unmqr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, \ - tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ +#define UNMQR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ + sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, TYPE* a, std::int64_t lda, \ + TYPE* tau, TYPE* c, std::int64_t ldc, TYPE* scratchpad, \ + std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ + return unmqr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, \ + tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ } UNMQR_LAUNCHER_USM(std::complex, cusolverDnCunmqr) @@ -2401,25 +2401,25 @@ UNMQR_LAUNCHER_USM(std::complex, cusolverDnZunmqr) #undef UNMQR_LAUNCHER_USM template -inline sycl::event unmtr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event unmtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T *a, - std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T* a, + std::int64_t lda, T* tau, T* c, std::int64_t ldc, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using cuDataType = typename CudaEquivalentType::Type; overflow_check(m, n, lda, ldc, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto c_ = reinterpret_cast(c); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto c_ = reinterpret_cast(c); + auto scratch_ = reinterpret_cast(scratchpad); cusolverStatus_t err; cusolver_native_named_func(func_name, func, err, handle, get_cublas_side_mode(side), get_cublas_fill_mode(uplo), get_cublas_operation(trans), m, @@ -2431,11 +2431,11 @@ inline sycl::event unmtr(const char *func_name, Func func, sycl::queue &queue, } #define UNMTR_LAUNCHER_USM(TYPE, CUSOLVER_ROUTINE) \ - sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ + sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* c, std::int64_t ldc, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return unmtr(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, uplo, trans, m, n, a, lda, \ tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ } @@ -2448,16 +2448,18 @@ UNMTR_LAUNCHER_USM(std::complex, cusolverDnZunmtr) // SCRATCHPAD APIs template -inline void gebrd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void gebrd_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, scratch_size); - }); - }).wait(); + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, scratch_size); + }); + }) + .wait(); } #define GEBRD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ @@ -2478,37 +2480,40 @@ GEBRD_LAUNCHER_SCRATCH(std::complex, cusolverDnZgebrd_bufferSize) #undef GEBRD_LAUNCHER_SCRATCH template <> -std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "gerqf_scratchpad_size"); } template <> -std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "gerqf_scratchpad_size"); } template <> -std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t gerqf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "gerqf_scratchpad_size"); } template <> -std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t gerqf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "gerqf_scratchpad_size"); } template -inline void geqrf_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void geqrf_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, nullptr, lda, scratch_size); - }); - }).wait(); + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, nullptr, lda, + scratch_size); + }); + }) + .wait(); } #define GEQRF_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ @@ -2529,24 +2534,26 @@ GEQRF_LAUNCHER_SCRATCH(std::complex, cusolverDnZgeqrf_bufferSize) #undef GEQRF_LAUNCHER_SCRATCH template -inline void gesvd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void gesvd_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, - std::int64_t ldu, std::int64_t ldvt, int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, scratch_size); - }); - }).wait(); + std::int64_t ldu, std::int64_t ldvt, int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, scratch_size); + }); + }) + .wait(); } #define GESVD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ std::int64_t gesvd_scratchpad_size( \ - sycl::queue & queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, \ - std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { \ + sycl::queue & queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ + std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { \ int scratch_size; \ gesvd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobu, jobvt, m, n, lda, \ ldu, ldvt, &scratch_size); \ @@ -2561,16 +2568,19 @@ GESVD_LAUNCHER_SCRATCH(std::complex, cusolverDnZgesvd_bufferSize) #undef GESVD_LAUNCHER_SCRATCH template -inline void getrf_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void getrf_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, nullptr, lda, scratch_size); - }); - }).wait(); + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, nullptr, lda, + scratch_size); + }); + }) + .wait(); } #define GETRF_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ @@ -2607,7 +2617,7 @@ GETRI_LAUNCHER_SCRATCH(std::complex) // cusolverDnXgetrs does not use scratchpad memory #define GETRS_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t getrs_scratchpad_size(sycl::queue & queue, oneapi::math::transpose trans, \ + std::int64_t getrs_scratchpad_size(sycl::queue & queue, oneapi::math::transpose trans, \ std::int64_t n, std::int64_t nrhs, std::int64_t lda, \ std::int64_t ldb) { \ return 0; \ @@ -2621,24 +2631,26 @@ GETRS_LAUNCHER_SCRATCH(std::complex) #undef GETRS_LAUNCHER_SCRATCH template -inline void heevd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void heevd_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda, int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_job(jobz), - get_cublas_fill_mode(uplo), n, nullptr, lda, nullptr, - scratch_size); - }); - }).wait(); + std::int64_t lda, int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_job(jobz), + get_cublas_fill_mode(uplo), n, nullptr, lda, nullptr, + scratch_size); + }); + }) + .wait(); } #define HEEVD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t heevd_scratchpad_size(sycl::queue & queue, oneapi::math::job jobz, \ - oneapi::math::uplo uplo, std::int64_t n, \ + std::int64_t heevd_scratchpad_size(sycl::queue & queue, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, \ std::int64_t lda) { \ int scratch_size; \ heevd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, lda, \ @@ -2652,25 +2664,27 @@ HEEVD_LAUNCHER_SCRATCH(std::complex, cusolverDnZheevd_bufferSize) #undef HEEVD_LAUNCHER_SCRATCH template -inline void hegvd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda, std::int64_t ldb, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_itype(itype), - get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, nullptr, - lda, nullptr, ldb, nullptr, scratch_size); - }); - }).wait(); +inline void hegvd_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, + std::int64_t ldb, int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_itype(itype), + get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, + nullptr, lda, nullptr, ldb, nullptr, scratch_size); + }); + }) + .wait(); } #define HEGVD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ std::int64_t hegvd_scratchpad_size(sycl::queue & queue, std::int64_t itype, \ - oneapi::math::job jobz, oneapi::math::uplo uplo, \ + oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda, std::int64_t ldb) { \ int scratch_size; \ hegvd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, \ @@ -2684,27 +2698,30 @@ HEGVD_LAUNCHER_SCRATCH(std::complex, cusolverDnZhegvd_bufferSize) #undef HEGVD_LAUNCHER_SCRATCH template -inline void hetrd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void hetrd_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, - nullptr, lda, nullptr, nullptr, nullptr, scratch_size); - }); - }).wait(); -} - -#define HETRD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), + n, nullptr, lda, nullptr, nullptr, nullptr, + scratch_size); + }); + }) + .wait(); +} + +#define HETRD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ std::int64_t hetrd_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - int scratch_size; \ - hetrd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ - &scratch_size); \ - return scratch_size; \ + std::int64_t n, std::int64_t lda) { \ + int scratch_size; \ + hetrd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ + &scratch_size); \ + return scratch_size; \ } HETRD_LAUNCHER_SCRATCH(std::complex, cusolverDnChetrd_bufferSize) @@ -2713,39 +2730,42 @@ HETRD_LAUNCHER_SCRATCH(std::complex, cusolverDnZhetrd_bufferSize) #undef HETRD_LAUNCHER_SCRATCH template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t hetrf_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "hetrf_scratchpad_size"); } template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { +std::int64_t hetrf_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { throw unimplemented("lapack", "hetrf_scratchpad_size"); } template -inline void orgbr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void orgbr_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_generate(vec), m, n, k, - nullptr, lda, nullptr, scratch_size); - }); - }).wait(); -} - -#define ORGBR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ + std::int64_t k, std::int64_t lda, int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_generate(vec), + m, n, k, nullptr, lda, nullptr, scratch_size); + }); + }) + .wait(); +} + +#define ORGBR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ std::int64_t orgbr_scratchpad_size(sycl::queue & queue, oneapi::math::generate vec, \ - std::int64_t m, std::int64_t n, std::int64_t k, \ - std::int64_t lda) { \ - int scratch_size; \ - orgbr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, vec, m, n, k, lda, \ - &scratch_size); \ - return scratch_size; \ + std::int64_t m, std::int64_t n, std::int64_t k, \ + std::int64_t lda) { \ + int scratch_size; \ + orgbr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, vec, m, n, k, lda, \ + &scratch_size); \ + return scratch_size; \ } ORGBR_LAUNCHER_SCRATCH(float, cusolverDnSorgbr_bufferSize) @@ -2754,27 +2774,29 @@ ORGBR_LAUNCHER_SCRATCH(double, cusolverDnDorgbr_bufferSize) #undef ORGBR_LAUNCHER_SCRATCH template -inline void orgtr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void orgtr_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, - nullptr, lda, nullptr, scratch_size); - }); - }).wait(); -} - -#define ORGTR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), + n, nullptr, lda, nullptr, scratch_size); + }); + }) + .wait(); +} + +#define ORGTR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ std::int64_t orgtr_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - int scratch_size; \ - orgtr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ - &scratch_size); \ - return scratch_size; \ + std::int64_t n, std::int64_t lda) { \ + int scratch_size; \ + orgtr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ + &scratch_size); \ + return scratch_size; \ } ORGTR_LAUNCHER_SCRATCH(float, cusolverDnSorgtr_bufferSize) @@ -2783,17 +2805,19 @@ ORGTR_LAUNCHER_SCRATCH(double, cusolverDnDorgtr_bufferSize) #undef ORGTR_LAUNCHER_SCRATCH template -inline void orgqr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void orgqr_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, k, nullptr, lda, nullptr, - scratch_size); - }); - }).wait(); + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, k, nullptr, lda, + nullptr, scratch_size); + }); + }) + .wait(); } #define ORGQR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ @@ -2812,14 +2836,14 @@ ORGQR_LAUNCHER_SCRATCH(double, cusolverDnDorgqr_bufferSize) #undef ORGQR_LAUNCHER_SCRATCH template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { throw unimplemented("lapack", "ormrq_scratchpad_size"); } template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { @@ -2827,30 +2851,32 @@ std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::sid } template -inline void ormqr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void ormqr_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, - std::int64_t ldc, int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), - get_cublas_operation(trans), m, n, k, nullptr, lda, nullptr, - nullptr, ldc, scratch_size); - }); - }).wait(); -} - -#define ORMQRF_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ - std::int64_t ormqr_scratchpad_size( \ - sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, \ - std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ - int scratch_size; \ - ormqr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, \ - lda, ldc, &scratch_size); \ - return scratch_size; \ + std::int64_t ldc, int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), + get_cublas_operation(trans), m, n, k, nullptr, lda, + nullptr, nullptr, ldc, scratch_size); + }); + }) + .wait(); +} + +#define ORMQRF_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ + std::int64_t ormqr_scratchpad_size( \ + sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ + int scratch_size; \ + ormqr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, \ + lda, ldc, &scratch_size); \ + return scratch_size; \ } ORMQRF_LAUNCHER_SCRATCH(float, cusolverDnSormqr_bufferSize) @@ -2859,27 +2885,29 @@ ORMQRF_LAUNCHER_SCRATCH(double, cusolverDnDormqr_bufferSize) #undef ORMQRF_LAUNCHER_SCRATCH template -inline void ormtr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void ormtr_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t lda, std::int64_t ldc, int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), - get_cublas_fill_mode(uplo), get_cublas_operation(trans), m, n, - nullptr, lda, nullptr, nullptr, ldc, scratch_size); - }); - }).wait(); + std::int64_t lda, std::int64_t ldc, int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), + get_cublas_fill_mode(uplo), get_cublas_operation(trans), + m, n, nullptr, lda, nullptr, nullptr, ldc, scratch_size); + }); + }) + .wait(); } #define ORMTR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t ormtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ - oneapi::math::uplo uplo, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t lda, \ - std::int64_t ldc) { \ + std::int64_t ormtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ + oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, \ + std::int64_t n, std::int64_t lda, std::int64_t ldc) { \ int scratch_size; \ ormtr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, uplo, trans, m, n, \ lda, ldc, &scratch_size); \ @@ -2892,27 +2920,29 @@ ORMTR_LAUNCHER_SCRATCH(double, cusolverDnDormtr_bufferSize) #undef ORMTR_LAUNCHER_SCRATCH template -inline void potrf_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void potrf_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, - nullptr, lda, scratch_size); - }); - }).wait(); -} - -#define POTRF_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), + n, nullptr, lda, scratch_size); + }); + }) + .wait(); +} + +#define POTRF_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ std::int64_t potrf_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - int scratch_size; \ - potrf_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ - &scratch_size); \ - return scratch_size; \ + std::int64_t n, std::int64_t lda) { \ + int scratch_size; \ + potrf_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ + &scratch_size); \ + return scratch_size; \ } POTRF_LAUNCHER_SCRATCH(float, cusolverDnSpotrf_bufferSize) @@ -2925,7 +2955,7 @@ POTRF_LAUNCHER_SCRATCH(std::complex, cusolverDnZpotrf_bufferSize) // cusolverDnXpotrs does not use scratchpad memory #define POTRS_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t potrs_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ + std::int64_t potrs_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t nrhs, std::int64_t lda, \ std::int64_t ldb) { \ return 0; \ @@ -2939,27 +2969,29 @@ POTRS_LAUNCHER_SCRATCH(std::complex) #undef POTRS_LAUNCHER_SCRATCH template -inline void potri_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void potri_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, - nullptr, lda, scratch_size); - }); - }).wait(); -} - -#define POTRI_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), + n, nullptr, lda, scratch_size); + }); + }) + .wait(); +} + +#define POTRI_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ std::int64_t potri_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - int scratch_size; \ - potri_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ - &scratch_size); \ - return scratch_size; \ + std::int64_t n, std::int64_t lda) { \ + int scratch_size; \ + potri_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ + &scratch_size); \ + return scratch_size; \ } POTRI_LAUNCHER_SCRATCH(float, cusolverDnSpotri_bufferSize) @@ -2970,26 +3002,29 @@ POTRI_LAUNCHER_SCRATCH(std::complex, cusolverDnZpotri_bufferSize) #undef POTRI_LAUNCHER_SCRATCH template -inline void sytrf_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void sytrf_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, n, nullptr, lda, scratch_size); - }); - }).wait(); -} - -#define SYTRF_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, n, nullptr, lda, + scratch_size); + }); + }) + .wait(); +} + +#define SYTRF_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ std::int64_t sytrf_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - int scratch_size; \ - sytrf_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ - &scratch_size); \ - return scratch_size; \ + std::int64_t n, std::int64_t lda) { \ + int scratch_size; \ + sytrf_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ + &scratch_size); \ + return scratch_size; \ } SYTRF_LAUNCHER_SCRATCH(float, cusolverDnSsytrf_bufferSize) @@ -3000,24 +3035,26 @@ SYTRF_LAUNCHER_SCRATCH(std::complex, cusolverDnZsytrf_bufferSize) #undef SYTRF_LAUNCHER_SCRATCH template -inline void syevd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void syevd_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda, int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_job(jobz), - get_cublas_fill_mode(uplo), n, nullptr, lda, nullptr, - scratch_size); - }); - }).wait(); + std::int64_t lda, int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_job(jobz), + get_cublas_fill_mode(uplo), n, nullptr, lda, nullptr, + scratch_size); + }); + }) + .wait(); } #define SYEVD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t syevd_scratchpad_size(sycl::queue & queue, oneapi::math::job jobz, \ - oneapi::math::uplo uplo, std::int64_t n, \ + std::int64_t syevd_scratchpad_size(sycl::queue & queue, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, \ std::int64_t lda) { \ int scratch_size; \ syevd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, jobz, uplo, n, lda, \ @@ -3031,25 +3068,27 @@ SYEVD_LAUNCHER_SCRATCH(double, cusolverDnDsyevd_bufferSize) #undef SYEVD_LAUNCHER_SCRATCH template -inline void sygvd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, - std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda, std::int64_t ldb, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_itype(itype), - get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, nullptr, - lda, nullptr, ldb, nullptr, scratch_size); - }); - }).wait(); +inline void sygvd_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, + std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, + std::int64_t ldb, int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cusolver_itype(itype), + get_cusolver_job(jobz), get_cublas_fill_mode(uplo), n, + nullptr, lda, nullptr, ldb, nullptr, scratch_size); + }); + }) + .wait(); } #define SYGVD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ std::int64_t sygvd_scratchpad_size(sycl::queue & queue, std::int64_t itype, \ - oneapi::math::job jobz, oneapi::math::uplo uplo, \ + oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda, std::int64_t ldb) { \ int scratch_size; \ sygvd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, itype, jobz, uplo, n, \ @@ -3063,27 +3102,30 @@ SYGVD_LAUNCHER_SCRATCH(double, cusolverDnDsygvd_bufferSize) #undef SYGVD_LAUNCHER_SCRATCH template -inline void sytrd_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void sytrd_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, - nullptr, lda, nullptr, nullptr, nullptr, scratch_size); - }); - }).wait(); -} - -#define SYTRD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), + n, nullptr, lda, nullptr, nullptr, nullptr, + scratch_size); + }); + }) + .wait(); +} + +#define SYTRD_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ std::int64_t sytrd_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - int scratch_size; \ - sytrd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ - &scratch_size); \ - return scratch_size; \ + std::int64_t n, std::int64_t lda) { \ + int scratch_size; \ + sytrd_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ + &scratch_size); \ + return scratch_size; \ } SYTRD_LAUNCHER_SCRATCH(float, cusolverDnSsytrd_bufferSize) @@ -3092,21 +3134,21 @@ SYTRD_LAUNCHER_SCRATCH(double, cusolverDnDsytrd_bufferSize) #undef SYTRD_LAUNCHER_SCRATCH template <> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t trtrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { throw unimplemented("lapack", "trtrs_scratchpad_size"); } template <> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t trtrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { throw unimplemented("lapack", "trtrs_scratchpad_size"); } template <> -std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t trtrs_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, @@ -3114,7 +3156,8 @@ std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, onea throw unimplemented("lapack", "trtrs_scratchpad_size"); } template <> -std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t trtrs_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, @@ -3123,28 +3166,30 @@ std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, one } template -inline void ungbr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void ungbr_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_generate(vec), m, n, k, - nullptr, lda, nullptr, scratch_size); - }); - }).wait(); -} - -#define UNGBR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ + std::int64_t k, std::int64_t lda, int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_generate(vec), + m, n, k, nullptr, lda, nullptr, scratch_size); + }); + }) + .wait(); +} + +#define UNGBR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ std::int64_t ungbr_scratchpad_size(sycl::queue & queue, oneapi::math::generate vec, \ - std::int64_t m, std::int64_t n, std::int64_t k, \ - std::int64_t lda) { \ - int scratch_size; \ - ungbr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, vec, m, n, k, lda, \ - &scratch_size); \ - return scratch_size; \ + std::int64_t m, std::int64_t n, std::int64_t k, \ + std::int64_t lda) { \ + int scratch_size; \ + ungbr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, vec, m, n, k, lda, \ + &scratch_size); \ + return scratch_size; \ } UNGBR_LAUNCHER_SCRATCH(std::complex, cusolverDnCungbr_bufferSize) @@ -3153,17 +3198,19 @@ UNGBR_LAUNCHER_SCRATCH(std::complex, cusolverDnZungbr_bufferSize) #undef UNGBR_LAUNCHER_SCRATCH template -inline void ungqr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void ungqr_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, k, nullptr, lda, nullptr, - scratch_size); - }); - }).wait(); + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, m, n, k, nullptr, lda, + nullptr, scratch_size); + }); + }) + .wait(); } #define UNGQR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ @@ -3182,27 +3229,29 @@ UNGQR_LAUNCHER_SCRATCH(std::complex, cusolverDnZungqr_bufferSize) #undef UNGQR_LAUNCHER_SCRATCH template -inline void ungtr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void ungtr_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, - int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), n, - nullptr, lda, nullptr, scratch_size); - }); - }).wait(); -} - -#define UNGTR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ + int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_fill_mode(uplo), + n, nullptr, lda, nullptr, scratch_size); + }); + }) + .wait(); +} + +#define UNGTR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ std::int64_t ungtr_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - int scratch_size; \ - ungtr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ - &scratch_size); \ - return scratch_size; \ + std::int64_t n, std::int64_t lda) { \ + int scratch_size; \ + ungtr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, uplo, n, lda, \ + &scratch_size); \ + return scratch_size; \ } UNGTR_LAUNCHER_SCRATCH(std::complex, cusolverDnCungtr_bufferSize) @@ -3211,7 +3260,7 @@ UNGTR_LAUNCHER_SCRATCH(std::complex, cusolverDnZungtr_bufferSize) #undef UNGTR_LAUNCHER_SCRATCH template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmrq_scratchpad_size>(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, @@ -3219,39 +3268,39 @@ std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, onea throw unimplemented("lapack", "unmrq_scratchpad_size"); } template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc) { +std::int64_t unmrq_scratchpad_size>( + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { throw unimplemented("lapack", "unmrq_scratchpad_size"); } template -inline void unmqr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void unmqr_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, - std::int64_t ldc, int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), - get_cublas_operation(trans), m, n, k, nullptr, lda, nullptr, - nullptr, ldc, scratch_size); - }); - }).wait(); -} - -#define UNMQR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ - template <> \ - std::int64_t unmqr_scratchpad_size( \ - sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, \ - std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ - int scratch_size; \ - unmqr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, \ - lda, ldc, &scratch_size); \ - return scratch_size; \ + std::int64_t ldc, int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), + get_cublas_operation(trans), m, n, k, nullptr, lda, + nullptr, nullptr, ldc, scratch_size); + }); + }) + .wait(); +} + +#define UNMQR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ + template <> \ + std::int64_t unmqr_scratchpad_size( \ + sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ + int scratch_size; \ + unmqr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, trans, m, n, k, \ + lda, ldc, &scratch_size); \ + return scratch_size; \ } UNMQR_LAUNCHER_SCRATCH(std::complex, cusolverDnCunmqr_bufferSize) @@ -3260,27 +3309,29 @@ UNMQR_LAUNCHER_SCRATCH(std::complex, cusolverDnZunmqr_bufferSize) #undef UNMQR_LAUNCHER_SCRATCH template -inline void unmtr_scratchpad_size(const char *func_name, Func func, sycl::queue &queue, +inline void unmtr_scratchpad_size(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t lda, std::int64_t ldc, int *scratch_size) { - queue.submit([&](sycl::handler &cgh) { - onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler &sc) { - auto handle = sc.get_handle(queue); - cusolverStatus_t err; - CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), - get_cublas_fill_mode(uplo), get_cublas_operation(trans), m, n, - nullptr, lda, nullptr, nullptr, ldc, scratch_size); - }); - }).wait(); + std::int64_t lda, std::int64_t ldc, int* scratch_size) { + queue + .submit([&](sycl::handler& cgh) { + onemath_cusolver_host_task(cgh, queue, [=](CusolverScopedContextHandler& sc) { + auto handle = sc.get_handle(queue); + cusolverStatus_t err; + CUSOLVER_ERROR_FUNC_T_SYNC(func_name, func, err, handle, get_cublas_side_mode(side), + get_cublas_fill_mode(uplo), get_cublas_operation(trans), + m, n, nullptr, lda, nullptr, nullptr, ldc, scratch_size); + }); + }) + .wait(); } #define UNMTR_LAUNCHER_SCRATCH(TYPE, CUSOLVER_ROUTINE) \ template <> \ - std::int64_t unmtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ - oneapi::math::uplo uplo, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t lda, \ - std::int64_t ldc) { \ + std::int64_t unmtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ + oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, \ + std::int64_t n, std::int64_t lda, std::int64_t ldc) { \ int scratch_size; \ unmtr_scratchpad_size(#CUSOLVER_ROUTINE, CUSOLVER_ROUTINE, queue, side, uplo, trans, m, n, \ lda, ldc, &scratch_size); \ diff --git a/src/lapack/backends/cusolver/cusolver_scope_handle.cpp b/src/lapack/backends/cusolver/cusolver_scope_handle.cpp index 67175ad8d..af0881c10 100644 --- a/src/lapack/backends/cusolver/cusolver_scope_handle.cpp +++ b/src/lapack/backends/cusolver/cusolver_scope_handle.cpp @@ -44,7 +44,7 @@ thread_local cusolver_handle CusolverScopedContextHandler::handle_he #endif CusolverScopedContextHandler::CusolverScopedContextHandler(sycl::queue queue, - sycl::interop_handle &ih) + sycl::interop_handle& ih) : ih(ih), needToRecover_(false) { placedContext_ = new sycl::context(queue.get_context()); @@ -74,8 +74,8 @@ CusolverScopedContextHandler::~CusolverScopedContextHandler() noexcept(false) { delete placedContext_; } -void ContextCallback(void *userData) { - auto *ptr = static_cast *>(userData); +void ContextCallback(void* userData) { + auto* ptr = static_cast*>(userData); if (!ptr) { return; } @@ -93,7 +93,7 @@ void ContextCallback(void *userData) { } } -cusolverDnHandle_t CusolverScopedContextHandler::get_handle(const sycl::queue &queue) { +cusolverDnHandle_t CusolverScopedContextHandler::get_handle(const sycl::queue& queue) { auto cudaDevice = ih.get_native_device(); CUresult cuErr; CUcontext desired; @@ -140,10 +140,10 @@ cusolverDnHandle_t CusolverScopedContextHandler::get_handle(const sycl::queue &q return handle; } -CUstream CusolverScopedContextHandler::get_stream(const sycl::queue &queue) { +CUstream CusolverScopedContextHandler::get_stream(const sycl::queue& queue) { return sycl::get_native(queue); } -sycl::context CusolverScopedContextHandler::get_context(const sycl::queue &queue) { +sycl::context CusolverScopedContextHandler::get_context(const sycl::queue& queue) { return queue.get_context(); } diff --git a/src/lapack/backends/cusolver/cusolver_scope_handle.hpp b/src/lapack/backends/cusolver/cusolver_scope_handle.hpp index b552b4af2..d6443b41a 100644 --- a/src/lapack/backends/cusolver/cusolver_scope_handle.hpp +++ b/src/lapack/backends/cusolver/cusolver_scope_handle.hpp @@ -89,19 +89,19 @@ cuSolver handle to the SYCL context. class CusolverScopedContextHandler { CUcontext original_; - sycl::context *placedContext_; + sycl::context* placedContext_; bool needToRecover_; - sycl::interop_handle &ih; + sycl::interop_handle& ih; #ifdef ONEMATH_PI_INTERFACE_REMOVED static thread_local cusolver_handle handle_helper; #else static thread_local cusolver_handle handle_helper; #endif - CUstream get_stream(const sycl::queue &queue); - sycl::context get_context(const sycl::queue &queue); + CUstream get_stream(const sycl::queue& queue); + sycl::context get_context(const sycl::queue& queue); public: - CusolverScopedContextHandler(sycl::queue queue, sycl::interop_handle &ih); + CusolverScopedContextHandler(sycl::queue queue, sycl::interop_handle& ih); ~CusolverScopedContextHandler() noexcept(false); /** @@ -111,7 +111,7 @@ class CusolverScopedContextHandler { * @param queue sycl queue. * @return cusolverDnHandle_t a handle to construct cusolver routines */ - cusolverDnHandle_t get_handle(const sycl::queue &queue); + cusolverDnHandle_t get_handle(const sycl::queue& queue); // This is a work-around function for reinterpret_casting the memory. This // will be fixed when SYCL-2020 has been implemented for Pi backend. template @@ -120,7 +120,7 @@ class CusolverScopedContextHandler { return reinterpret_cast(cudaPtr); } - void wait_stream(const sycl::queue &queue) { + void wait_stream(const sycl::queue& queue) { cuStreamSynchronize(get_stream(queue)); } }; diff --git a/src/lapack/backends/cusolver/cusolver_task.hpp b/src/lapack/backends/cusolver/cusolver_task.hpp index 497b29052..802230c9a 100644 --- a/src/lapack/backends/cusolver/cusolver_task.hpp +++ b/src/lapack/backends/cusolver/cusolver_task.hpp @@ -49,9 +49,9 @@ namespace lapack { namespace cusolver { template -static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { +static inline void host_task_internal(H& cgh, sycl::queue queue, F f) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - cgh.ext_codeplay_enqueue_native_command([f, queue](sycl::interop_handle ih){ + cgh.ext_codeplay_enqueue_native_command([f, queue](sycl::interop_handle ih) { #else cgh.host_task([f, queue](sycl::interop_handle ih) { #endif @@ -61,7 +61,7 @@ static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { } template -static inline void onemath_cusolver_host_task(H &cgh, sycl::queue queue, F f) { +static inline void onemath_cusolver_host_task(H& cgh, sycl::queue queue, F f) { (void)host_task_internal(cgh, queue, f); } diff --git a/src/lapack/backends/mkl_common/mkl_lapack.cxx b/src/lapack/backends/mkl_common/mkl_lapack.cxx index a0c85f82b..6c440b1c6 100644 --- a/src/lapack/backends/mkl_common/mkl_lapack.cxx +++ b/src/lapack/backends/mkl_common/mkl_lapack.cxx @@ -17,2495 +17,2817 @@ * SPDX-License-Identifier: Apache-2.0 *******************************************************************************/ -void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, sycl::buffer> &taup, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size)); -} -void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, +void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, sycl::buffer>& taup, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, + scratchpad, scratchpad_size)); +} +void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, + scratchpad, scratchpad_size)); } -void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, +void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, + scratchpad, scratchpad_size)); } -void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer> &tauq, - sycl::buffer> &taup, sycl::buffer> &scratchpad, +void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer>& tauq, + sycl::buffer>& taup, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, + scratchpad, scratchpad_size)); } -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); -} -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); -} -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); +} +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); +} +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } -void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); +void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } -void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } -void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } -void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size)); } -void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); +void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); } -void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); } -void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); } -void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, +void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size)); } -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); } -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); } -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); -} -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); -} -void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size)); -} -void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); +} +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size)); +} +void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size)); +} +void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size)); } -void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, +void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size)); -} -void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size)); +} +void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size)); -} -void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size)); -} -void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size)); -} -void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size)); -} -void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, - scratchpad_size)); -} -void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size)); -} -void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size)); -} -void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size)); -} -void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size)); -} -void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); -} -void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); -} -void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); -} -void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size)); +} +void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); -} -void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size)); -} -void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size)); -} -void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); -} -void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); -} -void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, + u, ldu, vt, ldvt, scratchpad, scratchpad_size)); +} +void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); -} -void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); -} -void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); -} -void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); -} -void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); -} -void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); -} -void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); -} -void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); -} -void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); -} -void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); -} -void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); -} -void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); -} -void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); -} -void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); -} -void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); -} -void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size)); -} -void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); -} -void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); -} -void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); -} -void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); -} -void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size)); -} -void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, scratchpad_size)); -} -void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size)); -} -void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, + u, ldu, vt, ldvt, scratchpad, scratchpad_size)); +} +void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, + u, ldu, vt, ldvt, scratchpad, scratchpad_size)); +} +void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::gesvd( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, + u, ldu, vt, ldvt, scratchpad, scratchpad_size)); +} +void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& w, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), + detail::get_onemkl_uplo(uplo), n, a, lda, + w, scratchpad, scratchpad_size)); +} +void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& w, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), + detail::get_onemkl_uplo(uplo), n, a, lda, + w, scratchpad, scratchpad_size)); +} +void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, sycl::buffer& w, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hegvd( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, + ldb, w, scratchpad, scratchpad_size)); +} +void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, sycl::buffer& w, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hegvd( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, + ldb, w, scratchpad, scratchpad_size)); +} +void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrd( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); +} +void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrd( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); +} +void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrf( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); +} +void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size)); -} -void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tau, sycl::buffer &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::hetrf( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); +} +void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), + m, n, k, a, lda, tau, scratchpad, + scratchpad_size)); +} +void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), + m, n, k, a, lda, tau, scratchpad, + scratchpad_size)); +} +void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); +} +void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); +} +void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgtr( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); } -void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tau, sycl::buffer &scratchpad, +void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); -} -void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgtr( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); +} +void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormtr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); +} +void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormtr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); +} +void ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormrq( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size)); +} +void ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormrq( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size)); +} +void ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormqr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size)); +} +void ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ormqr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size)); +} +void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, + a, lda, scratchpad, scratchpad_size)); +} +void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, + a, lda, scratchpad, scratchpad_size)); +} +void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, + a, lda, scratchpad, scratchpad_size)); +} +void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, + a, lda, scratchpad, scratchpad_size)); +} +void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, + a, lda, scratchpad, scratchpad_size)); +} +void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, + a, lda, scratchpad, scratchpad_size)); +} +void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, + a, lda, scratchpad, scratchpad_size)); +} +void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, + a, lda, scratchpad, scratchpad_size)); +} +void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, + nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size)); +} +void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, + nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size)); +} +void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, + nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size)); +} +void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, + nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size)); +} +void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), + detail::get_onemkl_uplo(uplo), n, a, lda, + w, scratchpad, scratchpad_size)); +} +void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), + detail::get_onemkl_uplo(uplo), n, a, lda, + w, scratchpad, scratchpad_size)); +} +void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sygvd( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, + ldb, w, scratchpad, scratchpad_size)); } -void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, +void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sygvd( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, + ldb, w, scratchpad, scratchpad_size)); } -void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); -} -void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, +void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrd( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); } -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size)); -} -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size)); -} -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size)); -} -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size)); -} -void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrd( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, scratchpad_size)); } -void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); } -void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); -} -void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); +} +void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); +} +void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::sytrf( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size)); } -void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); +} +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); +} +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); +} +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::trtrs( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size)); +} +void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), + m, n, k, a, lda, tau, scratchpad, + scratchpad_size)); } -void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), + m, n, k, a, lda, tau, scratchpad, + scratchpad_size)); } -void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, +void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); } -void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, +void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size)); } -void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, +void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungtr( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); } -void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, +void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); -} -void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungtr( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size)); +} +void unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmrq( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size)); +} +void unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmrq( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size)); +} +void unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmqr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size)); +} +void unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmqr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size)); +} +void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& c, + std::int64_t ldc, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmtr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); } -void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, +void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& c, + std::int64_t ldc, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, - scratchpad_size)); -} -sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, float *d, float *e, std::complex *tauq, - std::complex *taup, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - double *d, double *e, double *tauq, double *taup, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - float *d, float *e, float *tauq, float *taup, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, double *d, double *e, std::complex *tauq, - std::complex *taup, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, - std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, - std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, - std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, - std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, - float *s, std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, - double *s, std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, float *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, double *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::heevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, float *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, double *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hegvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *c, std::int64_t ldc, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *c, std::int64_t ldc, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *c, std::int64_t ldc, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *c, std::int64_t ldc, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), n, a, lda, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *w, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *w, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, - std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, - std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, ldb, w, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), n, a, lda, d, e, tau, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), n, a, lda, ipiv, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, - std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::unmtr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, + scratchpad_size)); +} +sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, float* d, float* e, std::complex* tauq, + std::complex* taup, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd( + queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies)); +} +sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* d, double* e, double* tauq, double* taup, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd( + queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies)); +} +sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* d, float* e, float* tauq, float* taup, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd( + queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies)); +} +sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, double* d, double* e, std::complex* tauq, + std::complex* taup, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd( + queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies)); +} +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf(queue, m, n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf(queue, m, n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf( + queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf( + queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf( + queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf( + queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event getri(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, + std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event getri(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri(queue, n, a, lda, ipiv, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t* ipiv, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t* ipiv, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrs(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, + ipiv, b, ldb, scratchpad, scratchpad_size, dependencies)); +} +sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, double* s, double* u, + std::int64_t ldu, double* vt, std::int64_t ldvt, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, + u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies)); +} +sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, float* s, float* u, + std::int64_t ldu, float* vt, std::int64_t ldvt, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, + u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies)); +} +sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, std::int64_t lda, + float* s, std::complex* u, std::int64_t ldu, std::complex* vt, + std::int64_t ldvt, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, + u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies)); +} +sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, std::int64_t lda, + double* s, std::complex* u, std::int64_t ldu, std::complex* vt, + std::int64_t ldvt, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, a, lda, s, + u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies)); +} +sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, float* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::heevd( + queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, double* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::heevd( + queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, float* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hegvd( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, + ldb, w, scratchpad, scratchpad_size, dependencies)); +} +sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, double* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hegvd( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, + ldb, w, scratchpad, scratchpad_size, dependencies)); +} +sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, d, e, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrd(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, d, e, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, ipiv, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrf(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, ipiv, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::orgbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr( + queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies)); +} +sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr( + queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies)); +} +sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies)); +} +sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies)); +} +sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size, dependencies)); +} +sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size, dependencies)); +} +sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size, dependencies)); +} +sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size, dependencies)); +} +sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), + n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + double* a, std::int64_t lda, double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), + n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), + n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs(queue, detail::get_onemkl_uplo(uplo), + n, nrhs, a, lda, b, ldb, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, double* a, std::int64_t lda, double* w, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd( + queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, float* a, std::int64_t lda, float* w, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd( + queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, w, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, std::int64_t lda, double* b, + std::int64_t ldb, double* w, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, + ldb, w, scratchpad, scratchpad_size, dependencies)); +} +sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, float* b, + std::int64_t ldb, float* w, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, a, lda, b, + ldb, w, scratchpad, scratchpad_size, dependencies)); +} +sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* d, double* e, double* tau, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, d, e, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* d, float* e, float* tau, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, d, e, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, ipiv, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, ipiv, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, ipiv, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, ipiv, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), n, a, lda, tau, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmrq(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmqr(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies)); +} +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double* a, + std::int64_t lda, double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies)); +} +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float* a, + std::int64_t lda, float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies)); +} +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, + dependencies)); +} +sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungbr(queue, detail::get_onemkl_generate(vec), m, n, k, a, lda, tau, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr( + queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies)); +} +sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr( + queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies)); +} +sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungtr(queue, detail::get_onemkl_uplo(uplo), + n, a, lda, tau, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmrq( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size, dependencies)); +} +sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmrq( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size, dependencies)); +} +sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmqr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size, dependencies)); +} +sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmqr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, a, lda, + tau, c, ldc, scratchpad, scratchpad_size, dependencies)); +} +sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmtr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies)); +} +sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmtr(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, - scratchpad, scratchpad_size, dependencies)); -} -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, - std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmtr( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, + dependencies)); +} +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, + std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch( + queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size)); } -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, - std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, + std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch( + queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size)); } -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch( + queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size)); } -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, - scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::geqrf_batch( + queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size)); } -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch( + queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size)); } -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch( + queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size)); } -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size)); + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch( + queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size)); } -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size)); + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getri_batch( + queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size)); } -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, - b, ldb, stride_b, batch_size, scratchpad, scratchpad_size)); -} -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, + ldb, stride_b, batch_size, scratchpad, scratchpad_size)); +} +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, - b, ldb, stride_b, batch_size, scratchpad, scratchpad_size)); -} -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, + ldb, stride_b, batch_size, scratchpad, scratchpad_size)); +} +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, - b, ldb, stride_b, batch_size, scratchpad, scratchpad_size)); -} -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, + ldb, stride_b, batch_size, scratchpad, scratchpad_size)); +} +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, - b, ldb, stride_b, batch_size, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, + ldb, stride_b, batch_size, scratchpad, scratchpad_size)); } -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch( + queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size)); } -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size)); -} -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer> &scratchpad, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch( + queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size)); +} +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch( + queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size)); } -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, - scratchpad, scratchpad_size)); -} -void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size)); -} -void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size)); -} -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, - scratchpad_size)); -} -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::getrf_batch( + queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size)); +} +void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, + scratchpad, scratchpad_size)); +} +void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, + scratchpad, scratchpad_size)); +} +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, - scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, + stride_a, batch_size, scratchpad, scratchpad_size)); } -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, - scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, + stride_a, batch_size, scratchpad, scratchpad_size)); } -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, - scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, + stride_a, batch_size, scratchpad, scratchpad_size)); } -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size)); -} -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS( + ::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, + stride_a, batch_size, scratchpad, scratchpad_size)); +} +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size)); -} -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size)); +} +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size)); -} -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size)); +} +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, - batch_size, scratchpad, scratchpad_size)); -} -void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size)); +} +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size)); -} -void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size)); +} +void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size) { + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, + scratchpad, scratchpad_size)); +} +void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { - RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size)); + RETHROW_ONEMKL_EXCEPTIONS(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, + tau, stride_tau, batch_size, + scratchpad, scratchpad_size)); } -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, float *tau, - std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, float* tau, + std::int64_t stride_tau, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, double *tau, - std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, double* tau, + std::int64_t stride_tau, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size, - dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size, dependencies)); } -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *tau, +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, - std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, - std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch(queue, m, n, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, - batch_size, scratchpad, scratchpad_size, - dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, + batch_size, scratchpad, scratchpad_size, dependencies)); } -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch(queue, m, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, stride_a, ipiv, stride_ipiv, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch(queue, n, a, lda, ipiv, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, - stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, + ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, - stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, + ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, - stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, + ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, - stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, - scratchpad_size, dependencies)); -} -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, - float **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, - group_count, group_sizes, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, - double **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, - group_count, group_sizes, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, - group_count, group_sizes, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch(queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, - group_count, group_sizes, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - float *a, std::int64_t lda, std::int64_t stride_a, float *tau, - std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, + ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, float** a, std::int64_t* lda, std::int64_t** ipiv, + float** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, + group_sizes, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, double** a, std::int64_t* lda, std::int64_t** ipiv, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, + group_sizes, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, + group_sizes, scratchpad, scratchpad_size, dependencies)); +} +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch( + queue, detail::get_onemkl_transpose(trans), n, nrhs, a, lda, ipiv, b, ldb, group_count, + group_sizes, scratchpad, scratchpad_size, dependencies)); +} +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + float* a, std::int64_t lda, std::int64_t stride_a, float* tau, + std::int64_t stride_tau, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - double *a, std::int64_t lda, std::int64_t stride_a, double *tau, - std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + double* a, std::int64_t lda, std::int64_t stride_a, double* tau, + std::int64_t stride_tau, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float **a, std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - double **a, std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float** a, std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + double** a, std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::orgqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, - scratchpad, scratchpad_size, dependencies)); + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, + scratchpad_size, dependencies)); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, std::complex *scratchpad, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, std::complex *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, stride_a, batch_size, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch(queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, - float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, - stride_b, batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, - double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, - stride_b, batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch( + queue, detail::get_onemkl_uplo(uplo), n, a, lda, group_count, group_sizes, scratchpad, + scratchpad_size, dependencies)); +} +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t stride_a, + float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t stride_a, + double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, - stride_b, batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, - stride_b, batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, - group_sizes, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, - group_sizes, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, stride_a, b, ldb, stride_b, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, double** a, std::int64_t* lda, double** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, - group_sizes, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch(queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, - group_sizes, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, - batch_size, scratchpad, scratchpad_size, - dependencies)); -} -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); -} -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, - scratchpad, scratchpad_size, dependencies)); + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, a, lda, b, ldb, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, stride_a, tau, stride_tau, + batch_size, scratchpad, scratchpad_size, dependencies)); +} +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); +} +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungqr_batch(queue, m, n, k, a, lda, tau, group_count, group_sizes, + scratchpad, scratchpad_size, dependencies)); } template <> -std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gebrd_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd_scratchpad_size(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::gebrd_scratchpad_size(queue, m, n, lda)); } template <> -std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gebrd_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd_scratchpad_size(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::gebrd_scratchpad_size(queue, m, n, lda)); } template <> -std::int64_t gebrd_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t gebrd_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd_scratchpad_size>(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::gebrd_scratchpad_size>(queue, m, n, lda)); } template <> -std::int64_t gebrd_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t gebrd_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gebrd_scratchpad_size>(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::gebrd_scratchpad_size>(queue, m, n, lda)); } template <> -std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf_scratchpad_size(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::gerqf_scratchpad_size(queue, m, n, lda)); } template <> -std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf_scratchpad_size(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::gerqf_scratchpad_size(queue, m, n, lda)); } template <> -std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t gerqf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf_scratchpad_size>(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::gerqf_scratchpad_size>(queue, m, n, lda)); } template <> -std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t gerqf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gerqf_scratchpad_size>(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::gerqf_scratchpad_size>(queue, m, n, lda)); } template <> -std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_scratchpad_size(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_scratchpad_size(queue, m, n, lda)); } template <> -std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_scratchpad_size(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_scratchpad_size(queue, m, n, lda)); } template <> -std::int64_t geqrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t geqrf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_scratchpad_size>(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_scratchpad_size>(queue, m, n, lda)); } template <> -std::int64_t geqrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t geqrf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_scratchpad_size>(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_scratchpad_size>(queue, m, n, lda)); } template <> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd_scratchpad_size(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, - ldvt)); +std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, + std::int64_t n, std::int64_t lda, std::int64_t ldu, + std::int64_t ldvt) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd_scratchpad_size( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, + ldvt)); } template <> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, +std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd_scratchpad_size(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, - ldvt)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd_scratchpad_size( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, + ldvt)); } template <> -std::int64_t gesvd_scratchpad_size>(sycl::queue &queue, +std::int64_t gesvd_scratchpad_size>(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd_scratchpad_size>(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, - n, lda, ldu, ldvt)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd_scratchpad_size>( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, ldu, + ldvt)); } template <> -std::int64_t gesvd_scratchpad_size>(sycl::queue &queue, +std::int64_t gesvd_scratchpad_size>(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::gesvd_scratchpad_size>(queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, - n, lda, ldu, ldvt)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::gesvd_scratchpad_size>( + queue, detail::get_onemkl_jobsvd(jobu), detail::get_onemkl_jobsvd(jobvt), m, n, lda, + ldu, ldvt)); } template <> -std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_scratchpad_size(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_scratchpad_size(queue, m, n, lda)); } template <> -std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_scratchpad_size(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_scratchpad_size(queue, m, n, lda)); } template <> -std::int64_t getrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t getrf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_scratchpad_size>(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_scratchpad_size>(queue, m, n, lda)); } template <> -std::int64_t getrf_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t getrf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_scratchpad_size>(queue, m, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_scratchpad_size>(queue, m, n, lda)); } template <> -std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_scratchpad_size(queue, n, lda)); +std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_scratchpad_size(queue, n, lda)); } template <> -std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_scratchpad_size(queue, n, lda)); +std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_scratchpad_size(queue, n, lda)); } template <> -std::int64_t getri_scratchpad_size>(sycl::queue &queue, std::int64_t n, +std::int64_t getri_scratchpad_size>(sycl::queue& queue, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_scratchpad_size>(queue, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_scratchpad_size>(queue, n, lda)); } template <> -std::int64_t getri_scratchpad_size>(sycl::queue &queue, std::int64_t n, +std::int64_t getri_scratchpad_size>(sycl::queue& queue, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_scratchpad_size>(queue, n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_scratchpad_size>(queue, n, lda)); } template <> -std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, +std::int64_t getrs_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_scratchpad_size( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb)); } template <> -std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, +std::int64_t getrs_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_scratchpad_size( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb)); } template <> -std::int64_t getrs_scratchpad_size>(sycl::queue &queue, +std::int64_t getrs_scratchpad_size>(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_scratchpad_size>(queue, detail::get_onemkl_transpose(trans), n, nrhs, - lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_scratchpad_size>( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb)); } template <> -std::int64_t getrs_scratchpad_size>(sycl::queue &queue, +std::int64_t getrs_scratchpad_size>(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_scratchpad_size>(queue, detail::get_onemkl_transpose(trans), n, nrhs, - lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrs_scratchpad_size>( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb)); } template <> -std::int64_t heevd_scratchpad_size>(sycl::queue &queue, oneapi::math::job jobz, +std::int64_t heevd_scratchpad_size>(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::heevd_scratchpad_size>(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, - lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::heevd_scratchpad_size>( + queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t heevd_scratchpad_size>(sycl::queue &queue, oneapi::math::job jobz, +std::int64_t heevd_scratchpad_size>(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::heevd_scratchpad_size>(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, - lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::heevd_scratchpad_size>( + queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t hegvd_scratchpad_size>(sycl::queue &queue, std::int64_t itype, +std::int64_t hegvd_scratchpad_size>(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hegvd_scratchpad_size>(queue, itype, detail::get_onemkl_job(jobz), - detail::get_onemkl_uplo(uplo), n, lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hegvd_scratchpad_size>( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, ldb)); } template <> -std::int64_t hegvd_scratchpad_size>(sycl::queue &queue, std::int64_t itype, +std::int64_t hegvd_scratchpad_size>(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hegvd_scratchpad_size>(queue, itype, detail::get_onemkl_job(jobz), - detail::get_onemkl_uplo(uplo), n, lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::hegvd_scratchpad_size>( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, + ldb)); } template <> -std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t hetrd_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrd_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrd_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t hetrd_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrd_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); +std::int64_t hetrd_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::hetrd_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t hetrf_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrf_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::hetrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); +std::int64_t hetrf_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::hetrf_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, +std::int64_t orgbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgbr_scratchpad_size(queue, detail::get_onemkl_generate(vect), m, n, k, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgbr_scratchpad_size( + queue, detail::get_onemkl_generate(vect), m, n, k, lda)); } template <> -std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, +std::int64_t orgbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgbr_scratchpad_size(queue, detail::get_onemkl_generate(vect), m, n, k, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgbr_scratchpad_size( + queue, detail::get_onemkl_generate(vect), m, n, k, lda)); } template <> -std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t orgtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t orgtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgtr_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_scratchpad_size(queue, m, n, k, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::orgqr_scratchpad_size(queue, m, n, k, lda)); } template <> -std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_scratchpad_size(queue, m, n, k, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::orgqr_scratchpad_size(queue, m, n, k, lda)); } template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, - ldc)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq_scratchpad_size( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, - ldc)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormrq_scratchpad_size( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> -std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, - ldc)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr_scratchpad_size( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> -std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, - ldc)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormqr_scratchpad_size( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> -std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, - ldc)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr_scratchpad_size( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, lda, ldc)); } template <> -std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr_scratchpad_size(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, - ldc)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ormtr_scratchpad_size( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, lda, ldc)); } template <> -std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t potrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrf_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t potrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); +std::int64_t potrf_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::potrf_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb)); } template <> -std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb)); } template <> -std::int64_t potrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrs_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, nrhs, - lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb)); } template <> -std::int64_t potrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, nrhs, - lda, ldb)); +std::int64_t potrs_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::int64_t lda, + std::int64_t ldb) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::potrs_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb)); } template <> -std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potri_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potri_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t potri_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potri_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t potri_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potri_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); +std::int64_t potri_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::potri_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t sytrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t sytrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t sytrf_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t sytrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrf_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); +std::int64_t sytrf_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::sytrf_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, +std::int64_t syevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd_scratchpad_size(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd_scratchpad_size( + queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, +std::int64_t syevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd_scratchpad_size(queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::syevd_scratchpad_size( + queue, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, +std::int64_t sygvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, - ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd_scratchpad_size( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, ldb)); } template <> -std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, +std::int64_t sygvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd_scratchpad_size(queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, - ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sygvd_scratchpad_size( + queue, itype, detail::get_onemkl_job(jobz), detail::get_onemkl_uplo(uplo), n, lda, ldb)); } template <> -std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t sytrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t sytrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::sytrd_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t trtrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, - lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, lda, ldb)); } template <> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t trtrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, - lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, lda, ldb)); } template <> -std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t trtrs_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, lda, ldb)); + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, lda, ldb)); } template <> -std::int64_t trtrs_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t trtrs_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::trtrs_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), detail::get_onemkl_diag(diag), n, nrhs, lda, ldb)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::trtrs_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), + detail::get_onemkl_diag(diag), n, nrhs, lda, ldb)); } template <> -std::int64_t ungbr_scratchpad_size>(sycl::queue &queue, +std::int64_t ungbr_scratchpad_size>(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungbr_scratchpad_size>(queue, detail::get_onemkl_generate(vect), m, n, k, - lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungbr_scratchpad_size>( + queue, detail::get_onemkl_generate(vect), m, n, k, lda)); } template <> -std::int64_t ungbr_scratchpad_size>(sycl::queue &queue, - oneapi::math::generate vect, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungbr_scratchpad_size>(queue, detail::get_onemkl_generate(vect), m, n, k, - lda)); +std::int64_t ungbr_scratchpad_size>(sycl::queue& queue, + oneapi::math::generate vect, + std::int64_t m, std::int64_t n, + std::int64_t k, std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungbr_scratchpad_size>( + queue, detail::get_onemkl_generate(vect), m, n, k, lda)); } template <> -std::int64_t ungqr_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t ungqr_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_scratchpad_size>(queue, m, n, k, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungqr_scratchpad_size>(queue, m, n, k, lda)); } template <> -std::int64_t ungqr_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t ungqr_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_scratchpad_size>(queue, m, n, k, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungqr_scratchpad_size>(queue, m, n, k, lda)); } template <> -std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t ungtr_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungtr_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungtr_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t ungtr_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungtr_scratchpad_size>(queue, detail::get_onemkl_uplo(uplo), n, lda)); +std::int64_t ungtr_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungtr_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda)); } template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmrq_scratchpad_size>(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmrq_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, - n, k, lda, ldc)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmrq_scratchpad_size>( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmrq_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, - n, k, lda, ldc)); +std::int64_t unmrq_scratchpad_size>( + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::unmrq_scratchpad_size>( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> -std::int64_t unmqr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmqr_scratchpad_size>(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmqr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, - n, k, lda, ldc)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmqr_scratchpad_size>( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> -std::int64_t unmqr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmqr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, - n, k, lda, ldc)); +std::int64_t unmqr_scratchpad_size>( + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::unmqr_scratchpad_size>( + queue, detail::get_onemkl_side(side), detail::get_onemkl_transpose(trans), m, n, k, lda, + ldc)); } template <> -std::int64_t unmtr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmtr_scratchpad_size>(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmtr_scratchpad_size>(queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), - detail::get_onemkl_transpose(trans), m, n, lda, ldc)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmtr_scratchpad_size>( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, lda, ldc)); } template <> -std::int64_t unmtr_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmtr_scratchpad_size>(sycl::queue& queue, + oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::unmtr_scratchpad_size>( - queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), detail::get_onemkl_transpose(trans), m, n, lda, ldc)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::unmtr_scratchpad_size>( + queue, detail::get_onemkl_side(side), detail::get_onemkl_uplo(uplo), + detail::get_onemkl_transpose(trans), m, n, lda, ldc)); } template <> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, stride_a, - stride_ipiv, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size( + queue, m, n, lda, stride_a, stride_ipiv, batch_size)); } template <> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, stride_a, - stride_ipiv, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size( + queue, m, n, lda, stride_a, stride_ipiv, batch_size)); } template <> -std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t getrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( - queue, m, n, lda, stride_a, stride_ipiv, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( + queue, m, n, lda, stride_a, stride_ipiv, batch_size)); } template <> -std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t getrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( - queue, m, n, lda, stride_a, stride_ipiv, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( + queue, m, n, lda, stride_a, stride_ipiv, batch_size)); } template <> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, stride_a, - stride_ipiv, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size( + queue, n, lda, stride_a, stride_ipiv, batch_size)); } template <> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, stride_a, - stride_ipiv, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size( + queue, n, lda, stride_a, stride_ipiv, batch_size)); } template <> -std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t n, +std::int64_t getri_batch_scratchpad_size>(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size>( - queue, n, lda, stride_a, stride_ipiv, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch_scratchpad_size>( + queue, n, lda, stride_a, stride_ipiv, batch_size)); } template <> -std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t n, +std::int64_t getri_batch_scratchpad_size>(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size>( - queue, n, lda, stride_a, stride_ipiv, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch_scratchpad_size>( + queue, n, lda, stride_a, stride_ipiv, batch_size)); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size)); + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, + stride_b, batch_size)); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size)); + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, + stride_b, batch_size)); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, + stride_b, batch_size)); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, stride_a, stride_ipiv, ldb, + stride_b, batch_size)); } template <> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, stride_a, - stride_tau, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size( + queue, m, n, lda, stride_a, stride_tau, batch_size)); } template <> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, stride_a, - stride_tau, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size( + queue, m, n, lda, stride_a, stride_tau, batch_size)); } template <> -std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t geqrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( - queue, m, n, lda, stride_a, stride_tau, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( + queue, m, n, lda, stride_a, stride_tau, batch_size)); } template <> -std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t geqrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( - queue, m, n, lda, stride_a, stride_tau, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( + queue, m, n, lda, stride_a, stride_tau, batch_size)); } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, - batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size)); } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, - batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size)); } template <> -std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, +std::int64_t potrf_batch_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size)); } template <> -std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, +std::int64_t potrf_batch_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda, stride_a, batch_size)); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { @@ -2513,7 +2835,7 @@ std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size)); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, @@ -2523,271 +2845,295 @@ std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mat } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size)); + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, + batch_size)); } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, batch_size)); + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, stride_a, ldb, stride_b, + batch_size)); } template <> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, stride_a, - stride_tau, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size( + queue, m, n, k, lda, stride_a, stride_tau, batch_size)); } template <> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, stride_a, - stride_tau, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size( + queue, m, n, k, lda, stride_a, stride_tau, batch_size)); } template <> std::int64_t ungqr_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( - queue, m, n, k, lda, stride_a, stride_tau, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( + queue, m, n, k, lda, stride_a, stride_tau, batch_size)); } template <> std::int64_t ungqr_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( - queue, m, n, k, lda, stride_a, stride_tau, batch_size)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( + queue, m, n, k, lda, stride_a, stride_tau, batch_size)); } template <> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, group_count, - group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size( + queue, m, n, lda, group_count, group_sizes)); } template <> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size(queue, m, n, lda, group_count, - group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size( + queue, m, n, lda, group_count, group_sizes)); } template <> -std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t getrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( - queue, m, n, lda, group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( + queue, m, n, lda, group_count, group_sizes)); } template <> -std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t getrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( - queue, m, n, lda, group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrf_batch_scratchpad_size>( + queue, m, n, lda, group_count, group_sizes)); } template <> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, group_count, - group_sizes)); +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size( + queue, n, lda, group_count, group_sizes)); } template <> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size(queue, n, lda, group_count, - group_sizes)); +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size( + queue, n, lda, group_count, group_sizes)); } template <> -std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, +std::int64_t getri_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size>( - queue, n, lda, group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch_scratchpad_size>( + queue, n, lda, group_count, group_sizes)); } template <> -std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, +std::int64_t getri_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getri_batch_scratchpad_size>( - queue, n, lda, group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getri_batch_scratchpad_size>( + queue, n, lda, group_count, group_sizes)); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size(queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, - ldb, group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size( queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes)); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, + group_sizes)); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( - queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, group_sizes)); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::getrs_batch_scratchpad_size>( + queue, detail::get_onemkl_transpose(trans), n, nrhs, lda, ldb, group_count, + group_sizes)); } template <> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, group_count, - group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size( + queue, m, n, lda, group_count, group_sizes)); } template <> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size(queue, m, n, lda, group_count, - group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size( + queue, m, n, lda, group_count, group_sizes)); } template <> -std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t geqrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( - queue, m, n, lda, group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( + queue, m, n, lda, group_count, group_sizes)); } template <> -std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t geqrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( - queue, m, n, lda, group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::geqrf_batch_scratchpad_size>( + queue, m, n, lda, group_count, group_sizes)); } template <> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, std::int64_t *lda, +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, - group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size( + queue, m, n, k, lda, group_count, group_sizes)); } template <> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size(queue, m, n, k, lda, - group_count, group_sizes)); +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::orgqr_batch_scratchpad_size( + queue, m, n, k, lda, group_count, group_sizes)); } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, - group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes)); } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, lda, - group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes)); } template <> -std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, - oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, +std::int64_t potrf_batch_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes)); } template <> -std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, - oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, +std::int64_t potrf_batch_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::potrf_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, lda, group_count, group_sizes)); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size(queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, - group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size( queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes)); + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( - queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes)); + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::potrs_batch_scratchpad_size>( + queue, detail::get_onemkl_uplo(uplo), n, nrhs, lda, ldb, group_count, group_sizes)); } template <> -std::int64_t ungqr_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, +std::int64_t ungqr_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( - queue, m, n, k, lda, group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( + queue, m, n, k, lda, group_count, group_sizes)); } template <> -std::int64_t ungqr_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, +std::int64_t ungqr_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { - RETHROW_ONEMKL_EXCEPTIONS_RET(::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( - queue, m, n, k, lda, group_count, group_sizes)); + std::int64_t* group_sizes) { + RETHROW_ONEMKL_EXCEPTIONS_RET( + ::oneapi::mkl::lapack::ungqr_batch_scratchpad_size>( + queue, m, n, k, lda, group_count, group_sizes)); } diff --git a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp b/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp index eee9e42e0..c4d03f26b 100644 --- a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp +++ b/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp @@ -29,1235 +29,1239 @@ namespace oneapi { namespace math { namespace lapack { -void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, sycl::buffer> &taup, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, +void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, sycl::buffer>& taup, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, +void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer> &tauq, - sycl::buffer> &taup, sycl::buffer> &scratchpad, +void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer>& tauq, + sycl::buffer>& taup, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, +void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, +void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, +void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer &u, std::int64_t ldu, sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &w, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, +void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& w, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& w, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, sycl::buffer& w, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, sycl::buffer& w, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, +void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &w, sycl::buffer &scratchpad, +void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tau, sycl::buffer &scratchpad, +void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, +void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, +void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, +void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); -void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, +void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, +void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, +void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, +void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, +void unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& c, + std::int64_t ldc, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, +void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& c, + std::int64_t ldc, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, - std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, + std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, - std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, + std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size); -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer> &scratchpad, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, +void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, float *d, float *e, std::complex *tauq, - std::complex *taup, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - double *d, double *e, double *tauq, double *taup, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - float *d, float *e, float *tauq, float *taup, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, double *d, double *e, std::complex *tauq, - std::complex *taup, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t *ipiv, double *b, - std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, float *b, - std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *b, std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, double *a, std::int64_t lda, double *s, double *u, - std::int64_t ldu, double *vt, std::int64_t ldvt, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, float *a, std::int64_t lda, float *s, float *u, - std::int64_t ldu, float *vt, std::int64_t ldvt, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, - float *s, std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, std::complex *a, std::int64_t lda, - double *s, std::complex *u, std::int64_t ldu, std::complex *vt, - std::int64_t ldvt, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, float *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, double *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, float *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, double *w, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, double *a, - std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, float *a, - std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *c, std::int64_t ldc, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *c, std::int64_t ldc, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *c, std::int64_t ldc, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *c, std::int64_t ldc, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *c, std::int64_t ldc, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *c, std::int64_t ldc, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - float *a, std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - double *a, std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *w, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *w, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *b, - std::int64_t ldb, double *w, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *b, - std::int64_t ldb, float *w, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, double *d, double *e, double *tau, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, float *d, float *e, float *tau, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, - std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, +void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); +sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, float* d, float* e, std::complex* tauq, + std::complex* taup, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* d, double* e, double* tauq, double* taup, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* d, float* e, float* tauq, float* taup, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, double* d, double* e, std::complex* tauq, + std::complex* taup, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getri(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, + std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getri(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t* ipiv, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t* ipiv, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* b, std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, double* s, double* u, + std::int64_t ldu, double* vt, std::int64_t ldvt, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, float* s, float* u, + std::int64_t ldu, float* vt, std::int64_t ldvt, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, std::int64_t lda, + float* s, std::complex* u, std::int64_t ldu, std::complex* vt, + std::int64_t ldvt, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, std::complex* a, std::int64_t lda, + double* s, std::complex* u, std::int64_t ldu, std::complex* vt, + std::int64_t ldvt, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, float* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, double* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, float* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, double* w, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, double* a, + std::int64_t lda, double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, float* a, + std::int64_t lda, float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + float* a, std::int64_t lda, float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + double* a, std::int64_t lda, double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, double* a, std::int64_t lda, double* w, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, float* a, std::int64_t lda, float* w, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, std::int64_t lda, double* b, + std::int64_t ldb, double* w, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, float* b, + std::int64_t ldb, float* w, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* d, double* e, double* tau, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* d, float* e, float* tau, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double* a, + std::int64_t lda, double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float* a, + std::int64_t lda, float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies = {}); -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, float *tau, - std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies = {}); +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, float* tau, + std::int64_t stride_tau, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, double *tau, - std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, + const std::vector& dependencies = {}); +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, double* tau, + std::int64_t stride_tau, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *tau, + const std::vector& dependencies = {}); +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, - std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, - std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, + const std::vector& dependencies = {}); +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, + const std::vector& dependencies = {}); +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies = {}); +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, + const std::vector& dependencies = {}); +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, + const std::vector& dependencies = {}); +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + const std::vector& dependencies = {}); +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, + const std::vector& dependencies = {}); +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, - float **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, - double **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - float *a, std::int64_t lda, std::int64_t stride_a, float *tau, - std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, + const std::vector& dependencies = {}); +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, float** a, std::int64_t* lda, std::int64_t** ipiv, + float** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, double** a, std::int64_t* lda, std::int64_t** ipiv, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + float* a, std::int64_t lda, std::int64_t stride_a, float* tau, + std::int64_t stride_tau, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - double *a, std::int64_t lda, std::int64_t stride_a, double *tau, - std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, + const std::vector& dependencies = {}); +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + double* a, std::int64_t lda, std::int64_t stride_a, double* tau, + std::int64_t stride_tau, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float **a, std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - double **a, std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, + const std::vector& dependencies = {}); +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float** a, std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + double** a, std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, std::complex *scratchpad, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, std::complex *scratchpad, + const std::vector& dependencies = {}); +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies = {}); +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies = {}); +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, - float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, - double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, + const std::vector& dependencies = {}); +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t stride_a, + float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t stride_a, + double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, float **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, double **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, float** a, std::int64_t* lda, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, double** a, std::int64_t* lda, double** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies = {}); +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies = {}); + const std::vector& dependencies = {}); +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies = {}); template = nullptr> -std::int64_t gebrd_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gebrd_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t geqrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, +std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue &queue, oneapi::math::jobsvd jobu, +std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); template = nullptr> -std::int64_t getrf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda); +std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t getrs_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); +std::int64_t getrs_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::int64_t lda, + std::int64_t ldb); template = nullptr> -std::int64_t heevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, +std::int64_t heevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hegvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, +std::int64_t hegvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t hetrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t hetrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t hetrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t hetrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t orgbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, +std::int64_t orgbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t orgtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t orgtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t orgqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t ormtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t potrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t potrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t potrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t potrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t potri_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t potri_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t sytrf_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t sytrf_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t syevd_scratchpad_size(sycl::queue &queue, oneapi::math::job jobz, +std::int64_t syevd_scratchpad_size(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t sygvd_scratchpad_size(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, +std::int64_t sygvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t sytrd_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t sytrd_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t trtrs_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t trtrs_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); template = nullptr> -std::int64_t ungbr_scratchpad_size(sycl::queue &queue, oneapi::math::generate vect, std::int64_t m, +std::int64_t ungbr_scratchpad_size(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ungqr_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t ungqr_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); template = nullptr> -std::int64_t ungtr_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, +std::int64_t ungtr_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); template = nullptr> -std::int64_t unmrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmqr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmqr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t unmtr_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmtr_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); template = nullptr> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template = nullptr> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda, +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t lda, std::int64_t stride_a, +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size); +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, std::int64_t lda, + std::int64_t stride_a, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size); template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); template = nullptr> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template = nullptr> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes); template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes); +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes); template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes); +std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes); } // namespace lapack } // namespace math diff --git a/src/lapack/backends/rocsolver/rocsolver_batch.cpp b/src/lapack/backends/rocsolver/rocsolver_batch.cpp index ae21f70af..a84be1fe6 100644 --- a/src/lapack/backends/rocsolver/rocsolver_batch.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_batch.cpp @@ -31,476 +31,477 @@ namespace rocsolver { // BATCH BUFFER API -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, - std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, + std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "geqrf_batch"); } -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &tau, - std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer &scratchpad, +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, + std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "geqrf_batch"); } -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "geqrf_batch"); } -void geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "geqrf_batch"); } -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getri_batch"); } -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getri_batch"); } -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getri_batch"); } -void getri_batch(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, +void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getri_batch"); } -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, sycl::buffer &b, +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getrs_batch"); } -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getrs_batch"); } -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getrs_batch"); } -void getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getrs_batch"); } -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer &scratchpad, +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getrf_batch"); } -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getrf_batch"); } -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getrf_batch"); } -void getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getrf_batch"); } -void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "orgqr_batch"); } -void orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "orgqr_batch"); } -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, + std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrf_batch"); } -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrf_batch"); } -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrf_batch"); } -void potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrf_batch"); } -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrs_batch"); } -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrs_batch"); } -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrs_batch"); } -void potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "potrs_batch"); } -void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "ungqr_batch"); } -void ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "ungqr_batch"); } // BATCH USM API -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, float *tau, - std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, float* tau, + std::int64_t stride_tau, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "geqrf_batch"); } -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, double *tau, - std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, double* tau, + std::int64_t stride_tau, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "geqrf_batch"); } -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *tau, +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "geqrf_batch"); } -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *tau, +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "geqrf_batch"); } -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, - std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "geqrf_batch"); } -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, - std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "geqrf_batch"); } -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "geqrf_batch"); } -sycl::event geqrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "geqrf_batch"); } -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "getrf_batch"); } -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "getrf_batch"); } -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getrf_batch"); } -sycl::event getrf_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, +sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getrf_batch"); } -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getrf_batch"); } -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getrf_batch"); } -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getrf_batch"); } -sycl::event getrf_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getrf_batch"); } -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getri_batch(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, +sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, float **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, double **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getri_batch(sycl::queue &queue, std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "getri_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, float *scratchpad, +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, float* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, float **a, std::int64_t *lda, std::int64_t **ipiv, - float **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, float** a, std::int64_t* lda, std::int64_t** ipiv, + float** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, double **a, std::int64_t *lda, std::int64_t **ipiv, - double **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, double** a, std::int64_t* lda, std::int64_t** ipiv, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event getrs_batch(sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrs_batch(sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getrs_batch"); } -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - float *a, std::int64_t lda, std::int64_t stride_a, float *tau, - std::int64_t stride_tau, std::int64_t batch_size, float *scratchpad, +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + float* a, std::int64_t lda, std::int64_t stride_a, float* tau, + std::int64_t stride_tau, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "orgqr_batch"); } -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - double *a, std::int64_t lda, std::int64_t stride_a, double *tau, - std::int64_t stride_tau, std::int64_t batch_size, double *scratchpad, +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + double* a, std::int64_t lda, std::int64_t stride_a, double* tau, + std::int64_t stride_tau, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "orgqr_batch"); } -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - float **a, std::int64_t *lda, float **tau, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + float** a, std::int64_t* lda, float** tau, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "orgqr_batch"); } -sycl::event orgqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - double **a, std::int64_t *lda, double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + double** a, std::int64_t* lda, double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "orgqr_batch"); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, float *a, +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "potrf_batch"); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, double *a, +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, double* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "potrf_batch"); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, std::complex *scratchpad, +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "potrf_batch"); } -sycl::event potrf_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, std::complex *scratchpad, +sycl::event potrf_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { throw unimplemented("lapack", "potrf_batch"); } template -inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, T **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, +inline sycl::event potrf_batch(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, T** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; int64_t batch_size = 0; @@ -509,27 +510,27 @@ inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &qu batch_size += group_sizes[i]; } - int *info = (int *)malloc_device(sizeof(int) * batch_size, queue); - T **a_dev = (T **)malloc_device(sizeof(T *) * batch_size, queue); + int* info = (int*)malloc_device(sizeof(int) * batch_size, queue); + T** a_dev = (T**)malloc_device(sizeof(T*) * batch_size, queue); auto done_cpy = - queue.submit([&](sycl::handler &h) { h.memcpy(a_dev, a, batch_size * sizeof(T *)); }); + queue.submit([&](sycl::handler& h) { h.memcpy(a_dev, a, batch_size * sizeof(T*)); }); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } cgh.depends_on(done_cpy); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; for (int64_t i = 0; i < group_count; i++) { - auto **a_ = reinterpret_cast(a_dev); - auto *info_ = reinterpret_cast(info); - rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo[i]), - (int)n[i], a_ + offset, (int)lda[i], info_ + offset, - (int)group_sizes[i]); + auto** a_ = reinterpret_cast(a_dev); + auto* info_ = reinterpret_cast(info); + rocsolver_native_named_func(func_name, func, err, handle, + get_rocblas_fill_mode(uplo[i]), (int)n[i], a_ + offset, + (int)lda[i], info_ + offset, (int)group_sizes[i]); offset += group_sizes[i]; } }); @@ -538,13 +539,13 @@ inline sycl::event potrf_batch(const char *func_name, Func func, sycl::queue &qu } // Scratchpad memory not needed as parts of buffer a is used as workspace memory -#define POTRF_BATCH_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event potrf_batch( \ - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, TYPE **a, std::int64_t *lda, \ - std::int64_t group_count, std::int64_t *group_sizes, TYPE *scratchpad, \ - std::int64_t scratchpad_size, const std::vector &dependencies) { \ - return potrf_batch(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, \ - group_count, group_sizes, scratchpad, scratchpad_size, dependencies); \ +#define POTRF_BATCH_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ + sycl::event potrf_batch( \ + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, TYPE** a, \ + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes, TYPE* scratchpad, \ + std::int64_t scratchpad_size, const std::vector& dependencies) { \ + return potrf_batch(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, \ + group_count, group_sizes, scratchpad, scratchpad_size, dependencies); \ } POTRF_BATCH_LAUNCHER_USM(float, rocsolver_spotrf_batched) @@ -554,44 +555,44 @@ POTRF_BATCH_LAUNCHER_USM(std::complex, rocsolver_zpotrf_batched) #undef POTRF_BATCH_LAUNCHER_USM -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t stride_a, - float *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t stride_a, + float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "potrs_batch"); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, std::int64_t stride_a, - double *b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t stride_a, + double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "potrs_batch"); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "potrs_batch"); } -sycl::event potrs_batch(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, +sycl::event potrs_batch(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "potrs_batch"); } template -inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, T **a, - std::int64_t *lda, T **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, T *scratchpad, +inline sycl::event potrs_batch(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, T** a, + std::int64_t* lda, T** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; int64_t batch_size = 0; @@ -605,31 +606,32 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu "rocsolver potrs_batch only supports nrhs = 1"); } - T **a_dev = (T **)malloc_device(sizeof(T *) * batch_size, queue); - T **b_dev = (T **)malloc_device(sizeof(T *) * batch_size, queue); + T** a_dev = (T**)malloc_device(sizeof(T*) * batch_size, queue); + T** b_dev = (T**)malloc_device(sizeof(T*) * batch_size, queue); auto done_cpy_a = - queue.submit([&](sycl::handler &h) { h.memcpy(a_dev, a, batch_size * sizeof(T *)); }); + queue.submit([&](sycl::handler& h) { h.memcpy(a_dev, a, batch_size * sizeof(T*)); }); auto done_cpy_b = - queue.submit([&](sycl::handler &h) { h.memcpy(b_dev, b, batch_size * sizeof(T *)); }); + queue.submit([&](sycl::handler& h) { h.memcpy(b_dev, b, batch_size * sizeof(T*)); }); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } cgh.depends_on(done_cpy_a); cgh.depends_on(done_cpy_b); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); int64_t offset = 0; rocblas_status err; for (int64_t i = 0; i < group_count; i++) { - auto **a_ = reinterpret_cast(a_dev); - auto **b_ = reinterpret_cast(b_dev); - rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo[i]), - (int)n[i], (int)nrhs[i], a_ + offset, (int)lda[i], - b_ + offset, (int)ldb[i], (int)group_sizes[i]); + auto** a_ = reinterpret_cast(a_dev); + auto** b_ = reinterpret_cast(b_dev); + rocsolver_native_named_func(func_name, func, err, handle, + get_rocblas_fill_mode(uplo[i]), (int)n[i], (int)nrhs[i], + a_ + offset, (int)lda[i], b_ + offset, (int)ldb[i], + (int)group_sizes[i]); offset += group_sizes[i]; } }); @@ -640,10 +642,10 @@ inline sycl::event potrs_batch(const char *func_name, Func func, sycl::queue &qu // Scratchpad memory not needed as parts of buffer a is used as workspace memory #define POTRS_BATCH_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ sycl::event potrs_batch( \ - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, \ - TYPE **a, std::int64_t *lda, TYPE **b, std::int64_t *ldb, std::int64_t group_count, \ - std::int64_t *group_sizes, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, \ + TYPE** a, std::int64_t* lda, TYPE** b, std::int64_t* ldb, std::int64_t group_count, \ + std::int64_t* group_sizes, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return potrs_batch(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, nrhs, a, lda, b, \ ldb, group_count, group_sizes, scratchpad, scratchpad_size, \ dependencies); \ @@ -656,52 +658,52 @@ POTRS_BATCH_LAUNCHER_USM(std::complex, rocsolver_zpotrs_batched) #undef POTRS_BATCH_LAUNCHER_USM -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "ungqr_batch"); } -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::complex *tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "ungqr_batch"); } -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "ungqr_batch"); } -sycl::event ungqr_batch(sycl::queue &queue, std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::complex **a, std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::complex** a, std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "ungqr_batch"); } // BATCH SCRATCHPAD API template <> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { throw unimplemented("lapack", "getrf_batch_scratchpad_size"); } template <> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { throw unimplemented("lapack", "getrf_batch_scratchpad_size"); } template <> -std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t getrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, @@ -709,7 +711,7 @@ std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue throw unimplemented("lapack", "getrf_batch_scratchpad_size"); } template <> -std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t getrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, @@ -717,20 +719,20 @@ std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queu throw unimplemented("lapack", "getrf_batch_scratchpad_size"); } template <> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { throw unimplemented("lapack", "getri_batch_scratchpad_size"); } template <> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t n, +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { throw unimplemented("lapack", "getri_batch_scratchpad_size"); } template <> -std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t n, +std::int64_t getri_batch_scratchpad_size>(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, @@ -738,7 +740,7 @@ std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue throw unimplemented("lapack", "getri_batch_scratchpad_size"); } template <> -std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t n, +std::int64_t getri_batch_scratchpad_size>(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, @@ -746,7 +748,7 @@ std::int64_t getri_batch_scratchpad_size>(sycl::queue &queu throw unimplemented("lapack", "getri_batch_scratchpad_size"); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, @@ -754,7 +756,7 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose trans, +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, @@ -763,32 +765,32 @@ std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mat } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } template <> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { throw unimplemented("lapack", "geqrf_batch_scratchpad_size"); } template <> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { throw unimplemented("lapack", "geqrf_batch_scratchpad_size"); } template <> -std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t geqrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, @@ -796,7 +798,7 @@ std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue throw unimplemented("lapack", "geqrf_batch_scratchpad_size"); } template <> -std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t geqrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, @@ -805,19 +807,19 @@ std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queu } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { throw unimplemented("lapack", "potrf_batch_scratchpad_size"); } template <> -std::int64_t potrf_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { throw unimplemented("lapack", "potrf_batch_scratchpad_size"); } template <> -std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, +std::int64_t potrf_batch_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, @@ -825,7 +827,7 @@ std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue throw unimplemented("lapack", "potrf_batch_scratchpad_size"); } template <> -std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queue, +std::int64_t potrf_batch_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, @@ -833,14 +835,14 @@ std::int64_t potrf_batch_scratchpad_size>(sycl::queue &queu throw unimplemented("lapack", "potrf_batch_scratchpad_size"); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { throw unimplemented("lapack", "potrs_batch_scratchpad_size"); } template <> -std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, @@ -849,25 +851,27 @@ std::int64_t potrs_batch_scratchpad_size(sycl::queue &queue, oneapi::mat } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size) { throw unimplemented("lapack", "potrs_batch_scratchpad_size"); } template <> std::int64_t potrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size) { throw unimplemented("lapack", "potrs_batch_scratchpad_size"); } template <> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { throw unimplemented("lapack", "orgqr_batch_scratchpad_size"); } template <> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { @@ -875,148 +879,148 @@ std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_ } template <> std::int64_t ungqr_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { throw unimplemented("lapack", "ungqr_batch_scratchpad_size"); } template <> std::int64_t ungqr_batch_scratchpad_size>( - sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { throw unimplemented("lapack", "ungqr_batch_scratchpad_size"); } template <> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "getrf_batch_scratchpad_size"); } template <> -std::int64_t getrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "getrf_batch_scratchpad_size"); } template <> -std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t getrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "getrf_batch_scratchpad_size"); } template <> -std::int64_t getrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t getrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "getrf_batch_scratchpad_size"); } template <> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { throw unimplemented("lapack", "getri_batch_scratchpad_size"); } template <> -std::int64_t getri_batch_scratchpad_size(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { throw unimplemented("lapack", "getri_batch_scratchpad_size"); } template <> -std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, +std::int64_t getri_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "getri_batch_scratchpad_size"); } template <> -std::int64_t getri_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, +std::int64_t getri_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "getri_batch_scratchpad_size"); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } template <> -std::int64_t getrs_batch_scratchpad_size(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, +std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes) { throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } template <> std::int64_t getrs_batch_scratchpad_size>( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes) { + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes) { throw unimplemented("lapack", "getrs_batch_scratchpad_size"); } template <> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "geqrf_batch_scratchpad_size"); } template <> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "geqrf_batch_scratchpad_size"); } template <> -std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t geqrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "geqrf_batch_scratchpad_size"); } template <> -std::int64_t geqrf_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, +std::int64_t geqrf_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "geqrf_batch_scratchpad_size"); } template <> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, std::int64_t *lda, +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "orgqr_batch_scratchpad_size"); } template <> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { throw unimplemented("lapack", "orgqr_batch_scratchpad_size"); } // rocsolverDnXpotrfBatched does not use scratchpad memory -#define POTRF_GROUP_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t potrf_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t * n, std::int64_t * lda, \ - std::int64_t group_count, std::int64_t * group_sizes) { \ - return 0; \ +#define POTRF_GROUP_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t potrf_batch_scratchpad_size( \ + sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t* n, std::int64_t* lda, \ + std::int64_t group_count, std::int64_t* group_sizes) { \ + return 0; \ } POTRF_GROUP_LAUNCHER_SCRATCH(float) @@ -1027,13 +1031,13 @@ POTRF_GROUP_LAUNCHER_SCRATCH(std::complex) #undef POTRF_GROUP_LAUNCHER_SCRATCH // rocsolverDnXpotrsBatched does not use scratchpad memory -#define POTRS_GROUP_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t potrs_batch_scratchpad_size( \ - sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t * n, std::int64_t * nrhs, \ - std::int64_t * lda, std::int64_t * ldb, std::int64_t group_count, \ - std::int64_t * group_sizes) { \ - return 0; \ +#define POTRS_GROUP_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t potrs_batch_scratchpad_size( \ + sycl::queue & queue, oneapi::math::uplo * uplo, std::int64_t* n, std::int64_t* nrhs, \ + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, \ + std::int64_t* group_sizes) { \ + return 0; \ } POTRS_GROUP_LAUNCHER_SCRATCH(float) @@ -1044,19 +1048,19 @@ POTRS_GROUP_LAUNCHER_SCRATCH(std::complex) #undef POTRS_GROUP_LAUNCHER_SCRATCH template <> -std::int64_t ungqr_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, +std::int64_t ungqr_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "ungqr_batch_scratchpad_size"); } template <> -std::int64_t ungqr_batch_scratchpad_size>(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, +std::int64_t ungqr_batch_scratchpad_size>(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { throw unimplemented("lapack", "ungqr_batch_scratchpad_size"); } diff --git a/src/lapack/backends/rocsolver/rocsolver_handle.hpp b/src/lapack/backends/rocsolver/rocsolver_handle.hpp index 3e1a14b34..4828c0254 100644 --- a/src/lapack/backends/rocsolver/rocsolver_handle.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_handle.hpp @@ -30,10 +30,10 @@ namespace rocsolver { template struct rocsolver_handle { - using handle_container_t = std::unordered_map *>; + using handle_container_t = std::unordered_map*>; handle_container_t rocsolver_handle_mapper_{}; ~rocsolver_handle() noexcept(false) { - for (auto &handle_pair : rocsolver_handle_mapper_) { + for (auto& handle_pair : rocsolver_handle_mapper_) { rocblas_status err; if (handle_pair.second != nullptr) { auto handle = handle_pair.second->exchange(nullptr); diff --git a/src/lapack/backends/rocsolver/rocsolver_helper.hpp b/src/lapack/backends/rocsolver/rocsolver_helper.hpp index 36464f2e6..5d4e6e821 100644 --- a/src/lapack/backends/rocsolver/rocsolver_helper.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_helper.hpp @@ -81,7 +81,7 @@ void overflow_check(Index index, Next... indices) { class rocsolver_error : virtual public std::runtime_error { protected: - inline const char *rocsolver_error_map(rocblas_status error) { + inline const char* rocsolver_error_map(rocblas_status error) { return rocblas_status_to_string(error); } @@ -111,7 +111,7 @@ class rocsolver_error : virtual public std::runtime_error { class hip_error : virtual public std::runtime_error { protected: - inline const char *hip_error_map(hipError_t result) { + inline const char* hip_error_map(hipError_t result) { return hipGetErrorName(result); } int error_number; ///< error number @@ -167,9 +167,8 @@ class hip_error : virtual public std::runtime_error { HIP_ERROR_FUNC(hipStreamSynchronize, hip_err, currentStreamId); template -inline void rocsolver_native_named_func(const char *func_name, Func func, - rocsolver_status err, - rocsolver_handle handle, Types... args){ +inline void rocsolver_native_named_func(const char* func_name, Func func, rocsolver_status err, + rocsolver_handle handle, Types... args) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND ROCSOLVER_ERROR_FUNC_T(func_name, func, err, handle, args...) #else @@ -258,12 +257,12 @@ struct RocmEquivalentType> { /* devinfo */ -inline int get_rocsolver_devinfo(sycl::queue &queue, sycl::buffer &devInfo) { +inline int get_rocsolver_devinfo(sycl::queue& queue, sycl::buffer& devInfo) { sycl::host_accessor dev_info_{ devInfo }; return dev_info_[0]; } -inline int get_rocsolver_devinfo(sycl::queue &queue, const int *devInfo) { +inline int get_rocsolver_devinfo(sycl::queue& queue, const int* devInfo) { int dev_info_; queue.memcpy(&dev_info_, devInfo, sizeof(int)); queue.wait(); @@ -271,8 +270,8 @@ inline int get_rocsolver_devinfo(sycl::queue &queue, const int *devInfo) { } template -inline void lapack_info_check(sycl::queue &queue, DEVINFO_T devinfo, const char *func_name, - const char *cufunc_name) { +inline void lapack_info_check(sycl::queue& queue, DEVINFO_T devinfo, const char* func_name, + const char* cufunc_name) { queue.wait(); const int devinfo_ = get_rocsolver_devinfo(queue, devinfo); if (devinfo_ > 0) diff --git a/src/lapack/backends/rocsolver/rocsolver_lapack.cpp b/src/lapack/backends/rocsolver/rocsolver_lapack.cpp index ea7a61937..5b0c265b2 100644 --- a/src/lapack/backends/rocsolver/rocsolver_lapack.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_lapack.cpp @@ -32,27 +32,27 @@ namespace rocsolver { // BUFFER APIs template -inline void gebrd(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void gebrd(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tauq, sycl::buffer& taup, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType_A = typename RocmEquivalentType::Type; using rocmDataType_B = typename RocmEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto d_acc = d.template get_access(cgh); auto e_acc = e.template get_access(cgh); auto tauq_acc = tauq.template get_access(cgh); auto taup_acc = taup.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto d_ = sc.get_mem(d_acc); - auto e_ = sc.get_mem(e_acc); - auto tauq_ = sc.get_mem(tauq_acc); - auto taup_ = sc.get_mem(taup_acc); + auto a_ = sc.get_mem(a_acc); + auto d_ = sc.get_mem(d_acc); + auto e_ = sc.get_mem(e_acc); + auto tauq_ = sc.get_mem(tauq_acc); + auto taup_ = sc.get_mem(taup_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, d_, e_, tauq_, taup_); @@ -61,10 +61,10 @@ inline void gebrd(const char *func_name, Func func, sycl::queue &queue, std::int } #define GEBRD_LAUNCHER(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - void gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, \ - sycl::buffer &tauq, sycl::buffer &taup, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, \ + sycl::buffer& tauq, sycl::buffer& taup, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ gebrd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, m, n, a, lda, d, e, tauq, taup, \ scratchpad, scratchpad_size); \ } @@ -76,41 +76,41 @@ GEBRD_LAUNCHER(std::complex, double, rocsolver_zgebrd) #undef GEBRD_LAUNCHER -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "gerqf"); } -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "gerqf"); } -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "gerqf"); } -void gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "gerqf"); } template -inline void geqrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void geqrf(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, tau_); }); @@ -118,8 +118,8 @@ inline void geqrf(const char *func_name, Func func, sycl::queue &queue, std::int } #define GEQRF_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, \ + void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ geqrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, m, n, a, lda, tau, scratchpad, \ scratchpad_size); \ @@ -133,9 +133,9 @@ GEQRF_LAUNCHER(std::complex, rocsolver_zgeqrf) #undef GEQRF_LAUNCHER template -void getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void getrf(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); @@ -146,15 +146,15 @@ void getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, sycl::buffer ipiv32(sycl::range<1>{ ipiv_size }); sycl::buffer devInfo{ 1 }; - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto ipiv32_acc = ipiv32.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto ipiv32_ = sc.get_mem(ipiv32_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); + auto a_ = sc.get_mem(a_acc); + auto ipiv32_ = sc.get_mem(ipiv32_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, ipiv32_, devInfo_); @@ -162,7 +162,7 @@ void getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, }); // Copy from 32-bit buffer to 64-bit - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); auto ipiv32_acc = ipiv32.template get_access(cgh); auto ipiv_acc = ipiv.template get_access(cgh); @@ -174,8 +174,8 @@ void getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, } #define GETRF_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, \ + void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ getrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, m, n, a, lda, ipiv, scratchpad, \ scratchpad_size); \ @@ -188,32 +188,32 @@ GETRF_LAUNCHER(std::complex, rocsolver_zgetrf) #undef GETRF_LAUNCHER -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getri"); } -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getri"); } -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getri"); } -void getri(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "getri"); } template -inline void getrs(const char *func_name, Func func, sycl::queue &queue, +inline void getrs(const char* func_name, Func func, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, nrhs, lda, ldb); @@ -224,7 +224,7 @@ inline void getrs(const char *func_name, Func func, sycl::queue &queue, std::uint64_t ipiv_size = ipiv.size(); sycl::buffer ipiv32(sycl::range<1>{ ipiv_size }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto ipiv32_acc = ipiv32.template get_access(cgh); auto ipiv_acc = ipiv.template get_access(cgh); cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { @@ -232,15 +232,15 @@ inline void getrs(const char *func_name, Func func, sycl::queue &queue, }); }); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto ipiv_acc = ipiv32.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto ipiv_ = sc.get_mem(ipiv_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto ipiv_ = sc.get_mem(ipiv_acc); + auto b_ = sc.get_mem(b_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_operation(trans), n, nrhs, a_, lda, ipiv_, b_, ldb); @@ -249,10 +249,10 @@ inline void getrs(const char *func_name, Func func, sycl::queue &queue, } #define GETRS_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, \ - sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, \ + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, \ + sycl::buffer& ipiv, sycl::buffer& b, std::int64_t ldb, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ getrs(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, trans, n, nrhs, a, lda, ipiv, b, ldb, \ scratchpad, scratchpad_size); \ } @@ -265,30 +265,30 @@ GETRS_LAUNCHER(std::complex, rocsolver_zgetrs) #undef GETRS_LAUNCHER template -inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, +inline void gesvd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType_A = typename RocmEquivalentType::Type; using rocmDataType_B = typename RocmEquivalentType::Type; overflow_check(n, m, lda, ldu, ldvt, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto s_acc = s.template get_access(cgh); auto u_acc = u.template get_access(cgh); auto vt_acc = vt.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto s_ = sc.get_mem(s_acc); - auto u_ = sc.get_mem(u_acc); - auto vt_ = sc.get_mem(vt_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto s_ = sc.get_mem(s_acc); + auto u_ = sc.get_mem(u_acc); + auto vt_ = sc.get_mem(vt_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocsolver_jobsvd(jobu), get_rocsolver_jobsvd(jobvt), m, n, a_, lda, s_, u_, ldu, @@ -300,10 +300,10 @@ inline void gesvd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define GESVD_LAUNCHER(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - void gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, \ - sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, \ + void gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, \ + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, \ + sycl::buffer& vt, std::int64_t ldvt, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ gesvd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, jobu, jobvt, m, n, a, lda, s, u, ldu, \ vt, ldvt, scratchpad, scratchpad_size); \ @@ -317,25 +317,25 @@ GESVD_LAUNCHER(std::complex, double, rocsolver_zgesvd) #undef GESVD_LAUNCHER template -inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, +inline void heevd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType_A = typename RocmEquivalentType::Type; using rocmDataType_B = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto w_ = sc.get_mem(w_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto w_ = sc.get_mem(w_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocsolver_job(jobz), get_rocblas_fill_mode(uplo), n, a_, lda, w_, scratch_, @@ -345,12 +345,12 @@ inline void heevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: lapack_info_check(queue, devInfo, __func__, func_name); } -#define HEEVD_LAUNCHER(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - void heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &w, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ - heevd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, scratchpad, \ - scratchpad_size); \ +#define HEEVD_LAUNCHER(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ + void heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& w, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ + heevd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, scratchpad, \ + scratchpad_size); \ } HEEVD_LAUNCHER(std::complex, float, rocsolver_cheevd) @@ -359,28 +359,28 @@ HEEVD_LAUNCHER(std::complex, double, rocsolver_zheevd) #undef HEEVD_LAUNCHER template -inline void hegvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, +inline void hegvd(const char* func_name, Func func, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType_A = typename RocmEquivalentType::Type; using rocmDataType_B = typename RocmEquivalentType::Type; overflow_check(n, lda, ldb, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto w_ = sc.get_mem(w_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto w_ = sc.get_mem(w_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocsolver_itype(itype), get_rocsolver_job(jobz), get_rocblas_fill_mode(uplo), n, a_, @@ -390,13 +390,13 @@ inline void hegvd(const char *func_name, Func func, sycl::queue &queue, std::int lapack_info_check(queue, devInfo, __func__, func_name); } -#define HEGVD_LAUNCHER(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - void hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ - sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ - hegvd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, ldb, \ - w, scratchpad, scratchpad_size); \ +#define HEGVD_LAUNCHER(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ + void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, \ + sycl::buffer& b, std::int64_t ldb, sycl::buffer& w, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ + hegvd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, ldb, \ + w, scratchpad, scratchpad_size); \ } HEGVD_LAUNCHER(std::complex, float, rocsolver_chegvd) @@ -405,24 +405,24 @@ HEGVD_LAUNCHER(std::complex, double, rocsolver_zhegvd) #undef HEGVD_LAUNCHER template -inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, +inline void hetrd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType_A = typename RocmEquivalentType::Type; using rocmDataType_B = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto d_acc = d.template get_access(cgh); auto e_acc = e.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto d_ = sc.get_mem(d_acc); - auto e_ = sc.get_mem(e_acc); - auto tau_ = sc.get_mem(tau_acc); + auto a_ = sc.get_mem(a_acc); + auto d_ = sc.get_mem(d_acc); + auto e_ = sc.get_mem(e_acc); + auto tau_ = sc.get_mem(tau_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, d_, e_, tau_); @@ -431,10 +431,10 @@ inline void hetrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define HETRD_LAUNCHER(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - void hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, \ - sycl::buffer &e, sycl::buffer &tau, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, \ + sycl::buffer& e, sycl::buffer& tau, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ hetrd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, d, e, tau, \ scratchpad, scratchpad_size); \ } @@ -444,32 +444,32 @@ HETRD_LAUNCHER(std::complex, double, rocsolver_zhetrd) #undef HETRD_LAUNCHER -void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "hetrf"); } -void hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, +void hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "hetrf"); } template -inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +inline void orgbr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_generate(vec), m, n, k, a_, lda, tau_); @@ -478,9 +478,9 @@ inline void orgbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORGBR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ orgbr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, scratchpad, \ scratchpad_size); \ } @@ -491,18 +491,18 @@ ORGBR_LAUNCHER(double, rocsolver_dorgbr) #undef ORGBR_LAUNCHER template -inline void orgqr(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void orgqr(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, m, n, k, a_, lda, tau_); }); @@ -510,9 +510,9 @@ inline void orgqr(const char *func_name, Func func, sycl::queue &queue, std::int } #define ORGQR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ orgqr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, m, n, k, a, lda, tau, scratchpad, \ scratchpad_size); \ } @@ -523,18 +523,18 @@ ORGQR_LAUNCHER(double, rocsolver_dorgqr) #undef ORGQR_LAUNCHER template -inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void orgtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, tau_); @@ -542,12 +542,12 @@ inline void orgtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: }); } -#define ORGTR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, \ - std::int64_t scratchpad_size) { \ - orgtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ - scratchpad_size); \ +#define ORGTR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ + void orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, \ + std::int64_t scratchpad_size) { \ + orgtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ + scratchpad_size); \ } ORGTR_LAUNCHER(float, rocsolver_sorgtr) @@ -556,22 +556,22 @@ ORGTR_LAUNCHER(double, rocsolver_dorgtr) #undef ORGTR_LAUNCHER template -inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, +inline void ormtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, lda, ldc, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_side_mode(side), get_rocblas_fill_mode(uplo), get_rocblas_operation(trans), @@ -581,10 +581,10 @@ inline void ormtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORMTR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, \ + void ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ ormtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, side, uplo, trans, m, n, a, lda, tau, \ c, ldc, scratchpad, scratchpad_size); \ @@ -595,35 +595,35 @@ ORMTR_LAUNCHER(double, rocsolver_dormtr) #undef ORMTR_LAUNCHER -void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "ormrq"); } -void ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "ormrq"); } template -inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, +inline void ormqr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, k, lda, ldc, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_side_mode(side), get_rocblas_operation(trans), m, n, k, a_, lda, tau_, c_, @@ -633,10 +633,10 @@ inline void ormqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define ORMQR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ ormqr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, tau, c, \ ldc, scratchpad, scratchpad_size); \ } @@ -647,19 +647,19 @@ ORMQR_LAUNCHER(double, rocsolver_dormqr) #undef ORMQR_LAUNCHER template -inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, +inline void potrf(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); + auto a_ = sc.get_mem(a_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, devInfo_); @@ -668,11 +668,11 @@ inline void potrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: lapack_info_check(queue, devInfo, __func__, func_name); } -#define POTRF_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ - potrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ - scratchpad_size); \ +#define POTRF_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ + void potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ + potrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ + scratchpad_size); \ } POTRF_LAUNCHER(float, rocsolver_spotrf) @@ -683,19 +683,19 @@ POTRF_LAUNCHER(std::complex, rocsolver_zpotrf) #undef POTRF_LAUNCHER template -inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, +inline void potri(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); + auto a_ = sc.get_mem(a_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, devInfo_); @@ -704,11 +704,11 @@ inline void potri(const char *func_name, Func func, sycl::queue &queue, oneapi:: lapack_info_check(queue, devInfo, __func__, func_name); } -#define POTRI_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ - potri(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ - scratchpad_size); \ +#define POTRI_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ + void potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ + potri(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ + scratchpad_size); \ } POTRI_LAUNCHER(float, rocsolver_spotri) @@ -719,19 +719,19 @@ POTRI_LAUNCHER(std::complex, rocsolver_zpotri) #undef POTRI_LAUNCHER template -inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, +inline void potrs(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, nrhs, lda, ldb, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, nrhs, a_, lda, b_, ldb); @@ -740,9 +740,9 @@ inline void potrs(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define POTRS_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ potrs(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, nrhs, a, lda, b, ldb, \ scratchpad, scratchpad_size); \ } @@ -755,23 +755,23 @@ POTRS_LAUNCHER(std::complex, rocsolver_zpotrs) #undef POTRS_LAUNCHER template -inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void syevd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto w_ = sc.get_mem(w_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto w_ = sc.get_mem(w_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocsolver_job(jobz), get_rocblas_fill_mode(uplo), n, a_, lda, w_, scratch_, @@ -782,9 +782,9 @@ inline void syevd(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define SYEVD_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &w, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& w, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ syevd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, scratchpad, \ scratchpad_size); \ } @@ -795,26 +795,26 @@ SYEVD_LAUNCHER(double, rocsolver_dsyevd) #undef SYEVD_LAUNCHER template -inline void sygvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void sygvd(const char* func_name, Func func, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, ldb, scratchpad_size); sycl::buffer devInfo{ 1 }; - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto b_acc = b.template get_access(cgh); auto w_acc = w.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); auto scratch_acc = scratchpad.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto b_ = sc.get_mem(b_acc); - auto w_ = sc.get_mem(w_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); - auto scratch_ = sc.get_mem(scratch_acc); + auto a_ = sc.get_mem(a_acc); + auto b_ = sc.get_mem(b_acc); + auto w_ = sc.get_mem(w_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); + auto scratch_ = sc.get_mem(scratch_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocsolver_itype(itype), get_rocsolver_job(jobz), get_rocblas_fill_mode(uplo), n, a_, @@ -825,10 +825,10 @@ inline void sygvd(const char *func_name, Func func, sycl::queue &queue, std::int } #define SYGVD_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, \ - sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, \ + sycl::buffer& b, std::int64_t ldb, sycl::buffer& w, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ sygvd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, b, ldb, \ w, scratchpad, scratchpad_size); \ } @@ -839,23 +839,23 @@ SYGVD_LAUNCHER(double, rocsolver_dsygvd) #undef SYGVD_LAUNCH template -inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, +inline void sytrd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto d_acc = d.template get_access(cgh); auto e_acc = e.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto d_ = sc.get_mem(d_acc); - auto e_ = sc.get_mem(e_acc); - auto tau_ = sc.get_mem(tau_acc); + auto a_ = sc.get_mem(a_acc); + auto d_ = sc.get_mem(d_acc); + auto e_ = sc.get_mem(e_acc); + auto tau_ = sc.get_mem(tau_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, d_, e_, tau_); @@ -863,13 +863,13 @@ inline void sytrd(const char *func_name, Func func, sycl::queue &queue, oneapi:: }); } -#define SYTRD_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, \ - sycl::buffer &tau, sycl::buffer &scratchpad, \ - std::int64_t scratchpad_size) { \ - sytrd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, d, e, tau, \ - scratchpad, scratchpad_size); \ +#define SYTRD_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ + void sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, \ + sycl::buffer& tau, sycl::buffer& scratchpad, \ + std::int64_t scratchpad_size) { \ + sytrd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, d, e, tau, \ + scratchpad, scratchpad_size); \ } SYTRD_LAUNCHER(float, rocsolver_ssytrd) @@ -878,9 +878,9 @@ SYTRD_LAUNCHER(double, rocsolver_dsytrd) #undef SYTRD_LAUNCHER template -inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &scratchpad, +inline void sytrf(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); @@ -892,15 +892,15 @@ inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: std::uint64_t ipiv_size = n; sycl::buffer ipiv32(sycl::range<1>{ ipiv_size }); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto ipiv32_acc = ipiv32.template get_access(cgh); auto devInfo_acc = devInfo.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto ipiv32_ = sc.get_mem(ipiv32_acc); - auto devInfo_ = sc.get_mem(devInfo_acc); + auto a_ = sc.get_mem(a_acc); + auto ipiv32_ = sc.get_mem(ipiv32_acc); + auto devInfo_ = sc.get_mem(devInfo_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, ipiv32_, devInfo_); @@ -908,7 +908,7 @@ inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: }); // Copy from 32-bit buffer to 64-bit - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); auto ipiv32_acc = ipiv32.template get_access(cgh); auto ipiv_acc = ipiv.template get_access(cgh); @@ -920,8 +920,8 @@ inline void sytrf(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define SYTRF_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, \ + void sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ sytrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, ipiv, scratchpad, \ scratchpad_size); \ @@ -934,47 +934,47 @@ SYTRF_LAUNCHER(std::complex, rocsolver_zsytrf) #undef SYTRF_LAUNCHER -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } -void trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, +void trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "trtrs"); } template -inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +inline void ungbr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_generate(vec), m, n, k, a_, lda, tau_); @@ -983,9 +983,9 @@ inline void ungbr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNGBR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, \ + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ ungbr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, scratchpad, \ scratchpad_size); \ } @@ -996,18 +996,18 @@ UNGBR_LAUNCHER(std::complex, rocsolver_zungbr) #undef UNGBR_LAUNCHER template -inline void ungqr(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void ungqr(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, m, n, k, a_, lda, tau_); }); @@ -1015,9 +1015,9 @@ inline void ungqr(const char *func_name, Func func, sycl::queue &queue, std::int } #define UNGQR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ ungqr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, m, n, k, a, lda, tau, scratchpad, \ scratchpad_size); \ } @@ -1028,18 +1028,18 @@ UNGQR_LAUNCHER(std::complex, rocsolver_zungqr) #undef UNGQR_LAUNCHER template -inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +inline void ungtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, tau_); @@ -1047,12 +1047,12 @@ inline void ungtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: }); } -#define UNGTR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, \ - std::int64_t scratchpad_size) { \ - ungtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ - scratchpad_size); \ +#define UNGTR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ + void ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, \ + std::int64_t scratchpad_size) { \ + ungtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, scratchpad, \ + scratchpad_size); \ } UNGTR_LAUNCHER(std::complex, rocsolver_cungtr) @@ -1060,37 +1060,37 @@ UNGTR_LAUNCHER(std::complex, rocsolver_zungtr) #undef UNGTR_LAUNCHER -void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size) { +void unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "unmrq"); } -void unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size) { +void unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { throw unimplemented("lapack", "unmrq"); } template -inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, +inline void unmqr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_side_mode(side), get_rocblas_operation(trans), m, n, k, a_, lda, tau_, c_, @@ -1100,10 +1100,10 @@ inline void unmqr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNMQR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, \ - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, std::int64_t ldc, \ - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { \ + void unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, \ + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, \ + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { \ unmqr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, tau, c, \ ldc, scratchpad, scratchpad_size); \ } @@ -1114,22 +1114,22 @@ UNMQR_LAUNCHER(std::complex, rocsolver_zunmqr) #undef UNMQR_LAUNCHER template -inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi::math::side side, +inline void unmtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, lda, ldc, scratchpad_size); - queue.submit([&](sycl::handler &cgh) { + queue.submit([&](sycl::handler& cgh) { auto a_acc = a.template get_access(cgh); auto tau_acc = tau.template get_access(cgh); auto c_acc = c.template get_access(cgh); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = sc.get_mem(a_acc); - auto tau_ = sc.get_mem(tau_acc); - auto c_ = sc.get_mem(c_acc); + auto a_ = sc.get_mem(a_acc); + auto tau_ = sc.get_mem(tau_acc); + auto c_ = sc.get_mem(c_acc); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_side_mode(side), get_rocblas_fill_mode(uplo), get_rocblas_operation(trans), @@ -1139,10 +1139,10 @@ inline void unmtr(const char *func_name, Func func, sycl::queue &queue, oneapi:: } #define UNMTR_LAUNCHER(TYPE, ROCSOLVER_ROUTINE) \ - void unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, \ - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, \ + void unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, \ + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, \ + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, \ std::int64_t scratchpad_size) { \ unmtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, side, uplo, trans, m, n, a, lda, tau, \ c, ldc, scratchpad, scratchpad_size); \ @@ -1156,26 +1156,26 @@ UNMTR_LAUNCHER(std::complex, rocsolver_zunmtr) // USM APIs template -inline sycl::event gebrd(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, T_A *a, std::int64_t lda, T_B *d, T_B *e, T_A *tauq, - T_A *taup, T_A *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event gebrd(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, T_A* a, std::int64_t lda, T_B* d, T_B* e, T_A* tauq, + T_A* taup, T_A* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType_A = typename RocmEquivalentType::Type; using rocmDataType_B = typename RocmEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto d_ = reinterpret_cast(d); - auto e_ = reinterpret_cast(e); - auto tauq_ = reinterpret_cast(tauq); - auto taup_ = reinterpret_cast(taup); + auto a_ = reinterpret_cast(a); + auto d_ = reinterpret_cast(d); + auto e_ = reinterpret_cast(e); + auto tauq_ = reinterpret_cast(tauq); + auto taup_ = reinterpret_cast(taup); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, d_, e_, tauq_, taup_); @@ -1185,10 +1185,10 @@ inline sycl::event gebrd(const char *func_name, Func func, sycl::queue &queue, s } #define GEBRD_LAUNCHER_USM(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - sycl::event gebrd(sycl::queue &queue, std::int64_t m, std::int64_t n, TYPE_A *a, \ - std::int64_t lda, TYPE_B *d, TYPE_B *e, TYPE_A *tauq, TYPE_A *taup, \ - TYPE_A *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, TYPE_A* a, \ + std::int64_t lda, TYPE_B* d, TYPE_B* e, TYPE_A* tauq, TYPE_A* taup, \ + TYPE_A* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return gebrd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, m, n, a, lda, d, e, tauq, taup, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1200,43 +1200,43 @@ GEBRD_LAUNCHER_USM(std::complex, double, rocsolver_zgebrd) #undef GEBRD_LAUNCHER_USM -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "gerqf"); } -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "gerqf"); } -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "gerqf"); } -sycl::event gerqf(sycl::queue &queue, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "gerqf"); } template -inline sycl::event geqrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, T *a, std::int64_t lda, T *tau, T *scratchpad, +inline sycl::event geqrf(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, T* a, std::int64_t lda, T* tau, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, tau_); }); @@ -1245,9 +1245,9 @@ inline sycl::event geqrf(const char *func_name, Func func, sycl::queue &queue, s } #define GEQRF_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event geqrf(sycl::queue &queue, std::int64_t m, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return geqrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, m, n, a, lda, tau, scratchpad, \ scratchpad_size, dependencies); \ } @@ -1260,10 +1260,10 @@ GEQRF_LAUNCHER_USM(std::complex, rocsolver_zgeqrf) #undef GEQRF_LAUNCHER_USM template -inline sycl::event getrf(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, T *a, std::int64_t lda, std::int64_t *ipiv, T *scratchpad, +inline sycl::event getrf(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, T* a, std::int64_t lda, std::int64_t* ipiv, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, lda, scratchpad_size); @@ -1271,19 +1271,19 @@ inline sycl::event getrf(const char *func_name, Func func, sycl::queue &queue, s // To get around the limitation. // Allocate memory with 32-bit ints then copy over results std::uint64_t ipiv_size = std::min(n, m); - int *ipiv32 = (int *)malloc_device(sizeof(int) * ipiv_size, queue); + int* ipiv32 = (int*)malloc_device(sizeof(int) * ipiv_size, queue); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto devInfo_ = reinterpret_cast(devInfo); - auto ipiv_ = reinterpret_cast(ipiv32); + auto a_ = reinterpret_cast(a); + auto devInfo_ = reinterpret_cast(devInfo); + auto ipiv_ = reinterpret_cast(ipiv32); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, m, n, a_, lda, ipiv_, devInfo_); @@ -1291,7 +1291,7 @@ inline sycl::event getrf(const char *func_name, Func func, sycl::queue &queue, s }); // Copy from 32-bit USM to 64-bit - auto done_casting = queue.submit([&](sycl::handler &cgh) { + auto done_casting = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { ipiv[index] = static_cast(ipiv32[index]); @@ -1305,10 +1305,10 @@ inline sycl::event getrf(const char *func_name, Func func, sycl::queue &queue, s } #define GETRF_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event getrf(sycl::queue &queue, std::int64_t m, std::int64_t n, TYPE *a, \ - std::int64_t lda, std::int64_t *ipiv, TYPE *scratchpad, \ + sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, TYPE* a, \ + std::int64_t lda, std::int64_t* ipiv, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return getrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, m, n, a, lda, ipiv, scratchpad, \ scratchpad_size, dependencies); \ } @@ -1320,33 +1320,33 @@ GETRF_LAUNCHER_USM(std::complex, rocsolver_zgetrf) #undef GETRF_LAUNCHER_USM -sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getri"); } -sycl::event getri(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, + std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getri"); } -sycl::event getri(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "getri"); } -sycl::event getri(sycl::queue &queue, std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "getri"); } template -inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, T *a, - std::int64_t lda, std::int64_t *ipiv, T *b, std::int64_t ldb, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event getrs(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, T* a, + std::int64_t lda, std::int64_t* ipiv, T* b, std::int64_t ldb, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, nrhs, lda, ldb, scratchpad_size); @@ -1354,25 +1354,25 @@ inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, // To get around the limitation. // Create new buffer and convert 64-bit values. std::uint64_t ipiv_size = n; - int *ipiv32 = (int *)malloc_device(sizeof(int) * ipiv_size, queue); + int* ipiv32 = (int*)malloc_device(sizeof(int) * ipiv_size, queue); - auto done_casting = queue.submit([&](sycl::handler &cgh) { + auto done_casting = queue.submit([&](sycl::handler& cgh) { cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { ipiv32[index] = static_cast(ipiv[index]); }); }); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } cgh.depends_on(done_casting); - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto ipiv_ = reinterpret_cast(ipiv32); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto ipiv_ = reinterpret_cast(ipiv32); + auto b_ = reinterpret_cast(b); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_operation(trans), n, nrhs, a_, lda, ipiv_, b_, ldb); @@ -1387,10 +1387,10 @@ inline sycl::event getrs(const char *func_name, Func func, sycl::queue &queue, } #define GETRS_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event getrs(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, \ - std::int64_t nrhs, TYPE *a, std::int64_t lda, std::int64_t *ipiv, TYPE *b, \ - std::int64_t ldb, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event getrs(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, \ + std::int64_t nrhs, TYPE* a, std::int64_t lda, std::int64_t* ipiv, TYPE* b, \ + std::int64_t ldb, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return getrs(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, trans, n, nrhs, a, lda, ipiv, \ b, ldb, scratchpad, scratchpad_size, dependencies); \ } @@ -1403,28 +1403,28 @@ GETRS_LAUNCHER_USM(std::complex, rocsolver_zgetrs) #undef GETRS_LAUNCHER_USM template -inline sycl::event gesvd(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event gesvd(const char* func_name, Func func, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, - std::int64_t n, T_A *a, std::int64_t lda, T_B *s, T_A *u, std::int64_t ldu, - T_A *vt, std::int64_t ldvt, T_A *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t n, T_A* a, std::int64_t lda, T_B* s, T_A* u, std::int64_t ldu, + T_A* vt, std::int64_t ldvt, T_A* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType_A = typename RocmEquivalentType::Type; using rocmDataType_B = typename RocmEquivalentType::Type; overflow_check(m, n, lda, ldu, ldvt, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto s_ = reinterpret_cast(s); - auto u_ = reinterpret_cast(u); - auto vt_ = reinterpret_cast(vt); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto s_ = reinterpret_cast(s); + auto u_ = reinterpret_cast(u); + auto vt_ = reinterpret_cast(vt); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocsolver_jobsvd(jobu), get_rocsolver_jobsvd(jobvt), m, n, a_, lda, s_, u_, ldu, @@ -1438,11 +1438,11 @@ inline sycl::event gesvd(const char *func_name, Func func, sycl::queue &queue, } #define GESVD_LAUNCHER_USM(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - sycl::event gesvd(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ - std::int64_t m, std::int64_t n, TYPE_A *a, std::int64_t lda, TYPE_B *s, \ - TYPE_A *u, std::int64_t ldu, TYPE_A *vt, std::int64_t ldvt, \ - TYPE_A *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event gesvd(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ + std::int64_t m, std::int64_t n, TYPE_A* a, std::int64_t lda, TYPE_B* s, \ + TYPE_A* u, std::int64_t ldu, TYPE_A* vt, std::int64_t ldvt, \ + TYPE_A* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return gesvd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, jobu, jobvt, m, n, a, lda, s, \ u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); \ } @@ -1455,25 +1455,25 @@ GESVD_LAUNCHER_USM(std::complex, double, rocsolver_zgesvd) #undef GESVD_LAUNCHER_USM template -inline sycl::event heevd(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A *&a, - std::int64_t lda, T_B *&w, T_A *&scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event heevd(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A*& a, + std::int64_t lda, T_B*& w, T_A*& scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType_A = typename RocmEquivalentType::Type; using rocmDataType_B = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto w_ = reinterpret_cast(w); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto w_ = reinterpret_cast(w); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocsolver_job(jobz), get_rocblas_fill_mode(uplo), n, a_, lda, w_, scratch_, @@ -1486,10 +1486,10 @@ inline sycl::event heevd(const char *func_name, Func func, sycl::queue &queue, } #define HEEVD_LAUNCHER_USM(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - sycl::event heevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ - std::int64_t n, TYPE_A *a, std::int64_t lda, TYPE_B *w, TYPE_A *scratchpad, \ + sycl::event heevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ + std::int64_t n, TYPE_A* a, std::int64_t lda, TYPE_B* w, TYPE_A* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return heevd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1500,27 +1500,27 @@ HEEVD_LAUNCHER_USM(std::complex, double, rocsolver_zheevd) #undef HEEVD_LAUNCHER_USM template -inline sycl::event hegvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A *&a, - std::int64_t lda, T_A *&b, std::int64_t ldb, T_B *&w, T_A *&scratchpad, +inline sycl::event hegvd(const char* func_name, Func func, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T_A*& a, + std::int64_t lda, T_A*& b, std::int64_t ldb, T_B*& w, T_A*& scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocmDataType_A = typename RocmEquivalentType::Type; using rocmDataType_B = typename RocmEquivalentType::Type; overflow_check(n, lda, ldb, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto w_ = reinterpret_cast(w); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto w_ = reinterpret_cast(w); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocsolver_itype(itype), get_rocsolver_job(jobz), get_rocblas_fill_mode(uplo), n, a_, @@ -1533,11 +1533,11 @@ inline sycl::event hegvd(const char *func_name, Func func, sycl::queue &queue, s } #define HEGVD_LAUNCHER_USM(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - sycl::event hegvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ - oneapi::math::uplo uplo, std::int64_t n, TYPE_A *a, std::int64_t lda, \ - TYPE_A *b, std::int64_t ldb, TYPE_B *w, TYPE_A *scratchpad, \ + sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, TYPE_A* a, std::int64_t lda, \ + TYPE_A* b, std::int64_t ldb, TYPE_B* w, TYPE_A* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return hegvd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, \ b, ldb, w, scratchpad, scratchpad_size, dependencies); \ } @@ -1548,24 +1548,24 @@ HEGVD_LAUNCHER_USM(std::complex, double, rocsolver_zhegvd) #undef HEGVD_LAUNCHER_USM template -inline sycl::event hetrd(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T_A *a, std::int64_t lda, T_B *d, - T_B *e, T_A *tau, T_A *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event hetrd(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T_A* a, std::int64_t lda, T_B* d, + T_B* e, T_A* tau, T_A* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType_A = typename RocmEquivalentType::Type; using rocmDataType_B = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto d_ = reinterpret_cast(d); - auto e_ = reinterpret_cast(e); - auto tau_ = reinterpret_cast(tau); + auto a_ = reinterpret_cast(a); + auto d_ = reinterpret_cast(d); + auto e_ = reinterpret_cast(e); + auto tau_ = reinterpret_cast(tau); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, d_, e_, tau_); @@ -1575,10 +1575,10 @@ inline sycl::event hetrd(const char *func_name, Func func, sycl::queue &queue, } #define HETRD_LAUNCHER_USM(TYPE_A, TYPE_B, ROCSOLVER_ROUTINE) \ - sycl::event hetrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE_A *a, \ - std::int64_t lda, TYPE_B *d, TYPE_B *e, TYPE_A *tau, TYPE_A *scratchpad, \ + sycl::event hetrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE_A* a, \ + std::int64_t lda, TYPE_B* d, TYPE_B* e, TYPE_A* tau, TYPE_A* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return hetrd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, d, e, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1588,36 +1588,36 @@ HETRD_LAUNCHER_USM(std::complex, double, rocsolver_zhetrd) #undef HETRD_LAUNCHER_USM -sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "hetrf"); } -sycl::event hetrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event hetrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "hetrf"); } template -inline sycl::event orgbr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event orgbr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - T *a, std::int64_t lda, T *tau, T *scratchpad, + T* a, std::int64_t lda, T* tau, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_generate(vec), m, n, k, a_, lda, tau_); @@ -1627,10 +1627,10 @@ inline sycl::event orgbr(const char *func_name, Func func, sycl::queue &queue, } #define ORGBR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event orgbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, \ - std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, TYPE *tau, \ - TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event orgbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, \ + std::int64_t n, std::int64_t k, TYPE* a, std::int64_t lda, TYPE* tau, \ + TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return orgbr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1641,21 +1641,21 @@ ORGBR_LAUNCHER_USM(double, rocsolver_dorgbr) #undef ORGBR_LAUNCHER_USM template -inline sycl::event orgqr(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event orgqr(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, T* a, std::int64_t lda, T* tau, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, m, n, k, a_, lda, tau_); }); @@ -1664,9 +1664,9 @@ inline sycl::event orgqr(const char *func_name, Func func, sycl::queue &queue, s } #define ORGQR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event orgqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return orgqr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, m, n, k, a, lda, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1677,21 +1677,21 @@ ORGQR_LAUNCHER_USM(double, rocsolver_dorgqr) #undef ORGQR_LAUNCHER_USM template -inline sycl::event orgtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event orgtr(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, T* tau, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, tau_); @@ -1701,9 +1701,9 @@ inline sycl::event orgtr(const char *func_name, Func func, sycl::queue &queue, } #define ORGTR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event orgtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event orgtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return orgtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1714,24 +1714,24 @@ ORGTR_LAUNCHER_USM(double, rocsolver_dorgtr) #undef ORGTR_LAUNCHER_USM template -inline sycl::event ormtr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event ormtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T *a, - std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T* a, + std::int64_t lda, T* tau, T* c, std::int64_t ldc, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, lda, ldc, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto c_ = reinterpret_cast(c); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_side_mode(side), get_rocblas_fill_mode(uplo), get_rocblas_operation(trans), @@ -1742,11 +1742,11 @@ inline sycl::event ormtr(const char *func_name, Func func, sycl::queue &queue, } #define ORMTR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event ormtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ + sycl::event ormtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* c, std::int64_t ldc, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return ormtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, side, uplo, trans, m, n, a, \ lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ } @@ -1756,37 +1756,37 @@ ORMTR_LAUNCHER_USM(double, rocsolver_dormtr) #undef ORMTR_LAUNCHER_USM -sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *c, std::int64_t ldc, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* c, std::int64_t ldc, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "ormrq"); } -sycl::event ormrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *c, std::int64_t ldc, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event ormrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* c, std::int64_t ldc, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "ormrq"); } template -inline sycl::event ormqr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event ormqr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, - std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, T* a, std::int64_t lda, T* tau, T* c, + std::int64_t ldc, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, k, lda, ldc, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto c_ = reinterpret_cast(c); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_side_mode(side), get_rocblas_operation(trans), m, n, k, a_, lda, tau_, c_, @@ -1796,14 +1796,14 @@ inline sycl::event ormqr(const char *func_name, Func func, sycl::queue &queue, return done; } -#define ORMQR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event ormqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, \ - TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ - std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ - return ormqr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, \ - tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ +#define ORMQR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ + sycl::event ormqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, TYPE* a, std::int64_t lda, \ + TYPE* tau, TYPE* c, std::int64_t ldc, TYPE* scratchpad, \ + std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ + return ormqr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, \ + tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ } ORMQR_LAUNCHER_USM(float, rocsolver_sormqr) @@ -1812,22 +1812,22 @@ ORMQR_LAUNCHER_USM(double, rocsolver_dormqr) #undef ORMQR_LAUNCHER_USM template -inline sycl::event potrf(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event potrf(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto devInfo_ = reinterpret_cast(devInfo); + auto a_ = reinterpret_cast(a); + auto devInfo_ = reinterpret_cast(devInfo); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, devInfo_); @@ -1839,9 +1839,9 @@ inline sycl::event potrf(const char *func_name, Func func, sycl::queue &queue, } #define POTRF_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event potrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event potrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return potrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ scratchpad_size, dependencies); \ } @@ -1854,23 +1854,23 @@ POTRF_LAUNCHER_USM(std::complex, rocsolver_zpotrf) #undef POTRF_LAUNCHER_USM template -inline sycl::event potri(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event potri(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto scratch_ = reinterpret_cast(scratchpad); - auto devInfo_ = reinterpret_cast(devInfo); + auto a_ = reinterpret_cast(a); + auto scratch_ = reinterpret_cast(scratchpad); + auto devInfo_ = reinterpret_cast(devInfo); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, devInfo_); @@ -1882,9 +1882,9 @@ inline sycl::event potri(const char *func_name, Func func, sycl::queue &queue, } #define POTRI_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event potri(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event potri(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return potri(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, scratchpad, \ scratchpad_size, dependencies); \ } @@ -1897,22 +1897,22 @@ POTRI_LAUNCHER_USM(std::complex, rocsolver_zpotri) #undef POTRI_LAUNCHER_USM template -inline sycl::event potrs(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, T *a, - std::int64_t lda, T *b, std::int64_t ldb, T *scratchpad, +inline sycl::event potrs(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, T* a, + std::int64_t lda, T* b, std::int64_t ldb, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, nrhs, lda, ldb, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, nrhs, a_, lda, b_, ldb); @@ -1922,10 +1922,10 @@ inline sycl::event potrs(const char *func_name, Func func, sycl::queue &queue, } #define POTRS_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event potrs(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, \ - std::int64_t nrhs, TYPE *a, std::int64_t lda, TYPE *b, std::int64_t ldb, \ - TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event potrs(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, \ + std::int64_t nrhs, TYPE* a, std::int64_t lda, TYPE* b, std::int64_t ldb, \ + TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return potrs(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, nrhs, a, lda, b, ldb, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1938,24 +1938,24 @@ POTRS_LAUNCHER_USM(std::complex, rocsolver_zpotrs) #undef POTRS_LAUNCHER_USM template -inline sycl::event syevd(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T *a, - std::int64_t lda, T *w, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event syevd(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T* a, + std::int64_t lda, T* w, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto w_ = reinterpret_cast(w); - auto scratch_ = reinterpret_cast(scratchpad); - auto devInfo_ = reinterpret_cast(devInfo); + auto a_ = reinterpret_cast(a); + auto w_ = reinterpret_cast(w); + auto scratch_ = reinterpret_cast(scratchpad); + auto devInfo_ = reinterpret_cast(devInfo); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocsolver_job(jobz), get_rocblas_fill_mode(uplo), n, a_, lda, w_, scratch_, @@ -1968,10 +1968,10 @@ inline sycl::event syevd(const char *func_name, Func func, sycl::queue &queue, } #define SYEVD_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event syevd(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ - std::int64_t n, TYPE *a, std::int64_t lda, TYPE *w, TYPE *scratchpad, \ + sycl::event syevd(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, \ + std::int64_t n, TYPE* a, std::int64_t lda, TYPE* w, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return syevd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, jobz, uplo, n, a, lda, w, \ scratchpad, scratchpad_size, dependencies); \ } @@ -1982,26 +1982,26 @@ SYEVD_LAUNCHER_USM(double, rocsolver_dsyevd) #undef SYEVD_LAUNCHER_USM template -inline sycl::event sygvd(const char *func_name, Func func, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T *a, - std::int64_t lda, T *b, std::int64_t ldb, T *w, T *scratchpad, +inline sycl::event sygvd(const char* func_name, Func func, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, T* a, + std::int64_t lda, T* b, std::int64_t ldb, T* w, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, ldb, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); - auto done = queue.submit([&](sycl::handler &cgh) { + int* devInfo = (int*)malloc_device(sizeof(int), queue); + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto b_ = reinterpret_cast(b); - auto w_ = reinterpret_cast(w); - auto devInfo_ = reinterpret_cast(devInfo); - auto scratch_ = reinterpret_cast(scratchpad); + auto a_ = reinterpret_cast(a); + auto b_ = reinterpret_cast(b); + auto w_ = reinterpret_cast(w); + auto devInfo_ = reinterpret_cast(devInfo); + auto scratch_ = reinterpret_cast(scratchpad); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocsolver_itype(itype), get_rocsolver_job(jobz), get_rocblas_fill_mode(uplo), n, a_, @@ -2013,13 +2013,13 @@ inline sycl::event sygvd(const char *func_name, Func func, sycl::queue &queue, s return done; } -#define SYGVD_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event sygvd(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, \ - oneapi::math::uplo uplo, std::int64_t n, TYPE *a, std::int64_t lda, TYPE *b, \ - std::int64_t ldb, TYPE *w, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ - return sygvd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, \ - b, ldb, w, scratchpad, scratchpad_size, dependencies); \ +#define SYGVD_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ + sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, \ + oneapi::math::uplo uplo, std::int64_t n, TYPE* a, std::int64_t lda, TYPE* b, \ + std::int64_t ldb, TYPE* w, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ + return sygvd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, itype, jobz, uplo, n, a, lda, \ + b, ldb, w, scratchpad, scratchpad_size, dependencies); \ } SYGVD_LAUNCHER_USM(float, rocsolver_ssygvd) @@ -2028,23 +2028,23 @@ SYGVD_LAUNCHER_USM(double, rocsolver_dsygvd) #undef SYGVD_LAUNCHER_USM template -inline sycl::event sytrd(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *d, T *e, - T *tau, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event sytrd(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, T* d, + T* e, T* tau, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto d_ = reinterpret_cast(d); - auto e_ = reinterpret_cast(e); - auto tau_ = reinterpret_cast(tau); + auto a_ = reinterpret_cast(a); + auto d_ = reinterpret_cast(d); + auto e_ = reinterpret_cast(e); + auto tau_ = reinterpret_cast(tau); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, d_, e_, tau_); @@ -2054,10 +2054,10 @@ inline sycl::event sytrd(const char *func_name, Func func, sycl::queue &queue, } #define SYTRD_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event sytrd(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *d, TYPE *e, TYPE *tau, TYPE *scratchpad, \ + sycl::event sytrd(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* d, TYPE* e, TYPE* tau, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return sytrd(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, d, e, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -2068,30 +2068,30 @@ SYTRD_LAUNCHER_USM(double, rocsolver_dsytrd) #undef SYTRD_LAUNCHER_USM template -inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, - std::int64_t *ipiv, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event sytrf(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, + std::int64_t* ipiv, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - int *devInfo = (int *)malloc_device(sizeof(int), queue); + int* devInfo = (int*)malloc_device(sizeof(int), queue); // rocsolver legacy api does not accept 64-bit ints. // To get around the limitation. // Allocate memory with 32-bit ints then copy over results std::uint64_t ipiv_size = n; - int *ipiv32 = (int *)malloc_device(sizeof(int) * ipiv_size, queue); + int* ipiv32 = (int*)malloc_device(sizeof(int) * ipiv_size, queue); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto ipiv_ = reinterpret_cast(ipiv32); - auto devInfo_ = reinterpret_cast(devInfo); + auto a_ = reinterpret_cast(a); + auto ipiv_ = reinterpret_cast(ipiv32); + auto devInfo_ = reinterpret_cast(devInfo); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, ipiv_, devInfo_); @@ -2099,7 +2099,7 @@ inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, }); // Copy from 32-bit USM to 64-bit - auto done_casting = queue.submit([&](sycl::handler &cgh) { + auto done_casting = queue.submit([&](sycl::handler& cgh) { cgh.depends_on(done); cgh.parallel_for(sycl::range<1>{ ipiv_size }, [=](sycl::id<1> index) { ipiv[index] = static_cast(ipiv32[index]); @@ -2112,13 +2112,13 @@ inline sycl::event sytrf(const char *func_name, Func func, sycl::queue &queue, return done_casting; } -#define SYTRF_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event sytrf(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, std::int64_t *ipiv, TYPE *scratchpad, \ - std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ - return sytrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, ipiv, \ - scratchpad, scratchpad_size, dependencies); \ +#define SYTRF_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ + sycl::event sytrf(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, std::int64_t* ipiv, TYPE* scratchpad, \ + std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ + return sytrf(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, ipiv, \ + scratchpad, scratchpad_size, dependencies); \ } SYTRF_LAUNCHER_USM(float, rocsolver_ssytrf) @@ -2128,50 +2128,50 @@ SYTRF_LAUNCHER_USM(std::complex, rocsolver_zsytrf) #undef SYTRF_LAUNCHER_USM -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double *a, - std::int64_t lda, double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, double* a, + std::int64_t lda, double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, float* a, + std::int64_t lda, float* b, std::int64_t ldb, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { throw unimplemented("lapack", "trtrs"); } -sycl::event trtrs(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, +sycl::event trtrs(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "trtrs"); } template -inline sycl::event ungbr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event ungbr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, std::int64_t n, std::int64_t k, - T *a, std::int64_t lda, T *tau, T *scratchpad, + T* a, std::int64_t lda, T* tau, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_generate(vec), m, n, k, a_, lda, tau_); @@ -2181,10 +2181,10 @@ inline sycl::event ungbr(const char *func_name, Func func, sycl::queue &queue, } #define UNGBR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event ungbr(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, \ - std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, TYPE *tau, \ - TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event ungbr(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, \ + std::int64_t n, std::int64_t k, TYPE* a, std::int64_t lda, TYPE* tau, \ + TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return ungbr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, vec, m, n, k, a, lda, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -2195,21 +2195,21 @@ UNGBR_LAUNCHER_USM(std::complex, rocsolver_zungbr) #undef UNGBR_LAUNCHER_USM template -inline sycl::event ungqr(const char *func_name, Func func, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event ungqr(const char* func_name, Func func, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, T* a, std::int64_t lda, T* tau, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, k, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, m, n, k, a_, lda, tau_); }); @@ -2218,9 +2218,9 @@ inline sycl::event ungqr(const char *func_name, Func func, sycl::queue &queue, s } #define UNGQR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event ungqr(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return ungqr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, m, n, k, a, lda, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -2231,21 +2231,21 @@ UNGQR_LAUNCHER_USM(std::complex, rocsolver_zungqr) #undef UNGQR_LAUNCHER_USM template -inline sycl::event ungtr(const char *func_name, Func func, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, T *a, std::int64_t lda, T *tau, - T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +inline sycl::event ungtr(const char* func_name, Func func, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, T* a, std::int64_t lda, T* tau, + T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_fill_mode(uplo), n, a_, lda, tau_); @@ -2255,9 +2255,9 @@ inline sycl::event ungtr(const char *func_name, Func func, sycl::queue &queue, } #define UNGTR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event ungtr(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *scratchpad, std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + sycl::event ungtr(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* scratchpad, std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ return ungtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, uplo, n, a, lda, tau, \ scratchpad, scratchpad_size, dependencies); \ } @@ -2267,39 +2267,39 @@ UNGTR_LAUNCHER_USM(std::complex, rocsolver_zungtr) #undef UNGTR_LAUNCHER_USM -sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "unmrq"); } -sycl::event unmrq(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event unmrq(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { throw unimplemented("lapack", "unmrq"); } template -inline sycl::event unmqr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event unmqr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, T *a, std::int64_t lda, T *tau, T *c, - std::int64_t ldc, T *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t n, std::int64_t k, T* a, std::int64_t lda, T* tau, T* c, + std::int64_t ldc, T* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(n, lda, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto c_ = reinterpret_cast(c); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_side_mode(side), get_rocblas_operation(trans), m, n, k, a_, lda, tau_, c_, @@ -2309,14 +2309,14 @@ inline sycl::event unmqr(const char *func_name, Func func, sycl::queue &queue, return done; } -#define UNMQR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event unmqr(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t k, TYPE *a, std::int64_t lda, \ - TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ - std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ - return unmqr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, \ - tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ +#define UNMQR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ + sycl::event unmqr(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, TYPE* a, std::int64_t lda, \ + TYPE* tau, TYPE* c, std::int64_t ldc, TYPE* scratchpad, \ + std::int64_t scratchpad_size, \ + const std::vector& dependencies) { \ + return unmqr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, side, trans, m, n, k, a, lda, \ + tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ } UNMQR_LAUNCHER_USM(std::complex, rocsolver_cunmqr) @@ -2325,24 +2325,24 @@ UNMQR_LAUNCHER_USM(std::complex, rocsolver_zunmqr) #undef UNMQR_LAUNCHER_USM template -inline sycl::event unmtr(const char *func_name, Func func, sycl::queue &queue, +inline sycl::event unmtr(const char* func_name, Func func, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T *a, - std::int64_t lda, T *tau, T *c, std::int64_t ldc, T *scratchpad, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, T* a, + std::int64_t lda, T* tau, T* c, std::int64_t ldc, T* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { using rocmDataType = typename RocmEquivalentType::Type; overflow_check(m, n, lda, ldc, scratchpad_size); - auto done = queue.submit([&](sycl::handler &cgh) { + auto done = queue.submit([&](sycl::handler& cgh) { int64_t num_events = dependencies.size(); for (int64_t i = 0; i < num_events; i++) { cgh.depends_on(dependencies[i]); } - onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler &sc) { + onemath_rocsolver_host_task(cgh, queue, [=](RocsolverScopedContextHandler& sc) { auto handle = sc.get_handle(queue); - auto a_ = reinterpret_cast(a); - auto tau_ = reinterpret_cast(tau); - auto c_ = reinterpret_cast(c); + auto a_ = reinterpret_cast(a); + auto tau_ = reinterpret_cast(tau); + auto c_ = reinterpret_cast(c); rocblas_status err; rocsolver_native_named_func(func_name, func, err, handle, get_rocblas_side_mode(side), get_rocblas_fill_mode(uplo), get_rocblas_operation(trans), @@ -2353,11 +2353,11 @@ inline sycl::event unmtr(const char *func_name, Func func, sycl::queue &queue, } #define UNMTR_LAUNCHER_USM(TYPE, ROCSOLVER_ROUTINE) \ - sycl::event unmtr(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, \ - oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE *a, \ - std::int64_t lda, TYPE *tau, TYPE *c, std::int64_t ldc, TYPE *scratchpad, \ + sycl::event unmtr(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, TYPE* a, \ + std::int64_t lda, TYPE* tau, TYPE* c, std::int64_t ldc, TYPE* scratchpad, \ std::int64_t scratchpad_size, \ - const std::vector &dependencies) { \ + const std::vector& dependencies) { \ return unmtr(#ROCSOLVER_ROUTINE, ROCSOLVER_ROUTINE, queue, side, uplo, trans, m, n, a, \ lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); \ } @@ -2384,22 +2384,22 @@ GEBRD_LAUNCHER_SCRATCH(std::complex) #undef GEBRD_LAUNCHER_SCRATCH template <> -std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "gerqf_scratchpad_size"); } template <> -std::int64_t gerqf_scratchpad_size(sycl::queue &queue, std::int64_t m, std::int64_t n, +std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "gerqf_scratchpad_size"); } template <> -std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t gerqf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "gerqf_scratchpad_size"); } template <> -std::int64_t gerqf_scratchpad_size>(sycl::queue &queue, std::int64_t m, +std::int64_t gerqf_scratchpad_size>(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "gerqf_scratchpad_size"); } @@ -2418,12 +2418,12 @@ GEQRF_LAUNCHER_SCRATCH(std::complex) #undef GEQRF_LAUNCHER_SCRATCH -#define GESVD_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t gesvd_scratchpad_size( \ - sycl::queue & queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, \ - std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { \ - return std::min(m, n) - 1; \ +#define GESVD_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t gesvd_scratchpad_size( \ + sycl::queue & queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, \ + std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { \ + return std::min(m, n) - 1; \ } GESVD_LAUNCHER_SCRATCH(float) @@ -2448,27 +2448,27 @@ GETRF_LAUNCHER_SCRATCH(std::complex) #undef GETRF_LAUNCHER_SCRATCH template <> -std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda) { +std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "getri_scratchpad_size"); } template <> -std::int64_t getri_scratchpad_size(sycl::queue &queue, std::int64_t n, std::int64_t lda) { +std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "getri_scratchpad_size"); } template <> -std::int64_t getri_scratchpad_size>(sycl::queue &queue, std::int64_t n, +std::int64_t getri_scratchpad_size>(sycl::queue& queue, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "getri_scratchpad_size"); } template <> -std::int64_t getri_scratchpad_size>(sycl::queue &queue, std::int64_t n, +std::int64_t getri_scratchpad_size>(sycl::queue& queue, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "getri_scratchpad_size"); } #define GETRS_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t getrs_scratchpad_size(sycl::queue & queue, oneapi::math::transpose trans, \ + std::int64_t getrs_scratchpad_size(sycl::queue & queue, oneapi::math::transpose trans, \ std::int64_t n, std::int64_t nrhs, std::int64_t lda, \ std::int64_t ldb) { \ return 0; \ @@ -2481,12 +2481,12 @@ GETRS_LAUNCHER_SCRATCH(std::complex) #undef GETRS_LAUNCHER_SCRATCH -#define HEEVD_LAUNCHER_SCRATCH(TYPE) \ - template <> \ +#define HEEVD_LAUNCHER_SCRATCH(TYPE) \ + template <> \ std::int64_t heevd_scratchpad_size(sycl::queue & queue, oneapi::math::job jobz, \ oneapi::math::uplo uplo, std::int64_t n, \ - std::int64_t lda) { \ - return n; \ + std::int64_t lda) { \ + return n; \ } HEEVD_LAUNCHER_SCRATCH(std::complex) @@ -2497,7 +2497,7 @@ HEEVD_LAUNCHER_SCRATCH(std::complex) #define HEGVD_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t hegvd_scratchpad_size(sycl::queue & queue, std::int64_t itype, \ - oneapi::math::job jobz, oneapi::math::uplo uplo, \ + oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda, std::int64_t ldb) { \ return n; \ } @@ -2507,11 +2507,11 @@ HEGVD_LAUNCHER_SCRATCH(std::complex) #undef HEGVD_LAUNCHER_SCRATCH -#define HETRD_LAUNCHER_SCRATCH(TYPE) \ - template <> \ +#define HETRD_LAUNCHER_SCRATCH(TYPE) \ + template <> \ std::int64_t hetrd_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - return 0; \ + std::int64_t n, std::int64_t lda) { \ + return 0; \ } HETRD_LAUNCHER_SCRATCH(std::complex) @@ -2520,22 +2520,23 @@ HETRD_LAUNCHER_SCRATCH(std::complex) #undef HETRD_LAUNCHER_SCRATCH template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, +std::int64_t hetrf_scratchpad_size>(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { throw unimplemented("lapack", "hetrf_scratchpad_size"); } template <> -std::int64_t hetrf_scratchpad_size>(sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { +std::int64_t hetrf_scratchpad_size>(sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { throw unimplemented("lapack", "hetrf_scratchpad_size"); } -#define ORGBR_LAUNCHER_SCRATCH(TYPE) \ - template <> \ +#define ORGBR_LAUNCHER_SCRATCH(TYPE) \ + template <> \ std::int64_t orgbr_scratchpad_size(sycl::queue & queue, oneapi::math::generate vec, \ - std::int64_t m, std::int64_t n, std::int64_t k, \ - std::int64_t lda) { \ - return 0; \ + std::int64_t m, std::int64_t n, std::int64_t k, \ + std::int64_t lda) { \ + return 0; \ } ORGBR_LAUNCHER_SCRATCH(float) @@ -2543,11 +2544,11 @@ ORGBR_LAUNCHER_SCRATCH(double) #undef ORGBR_LAUNCHER_SCRATCH -#define ORGTR_LAUNCHER_SCRATCH(TYPE) \ - template <> \ +#define ORGTR_LAUNCHER_SCRATCH(TYPE) \ + template <> \ std::int64_t orgtr_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - return 0; \ + std::int64_t n, std::int64_t lda) { \ + return 0; \ } ORGTR_LAUNCHER_SCRATCH(float) @@ -2568,26 +2569,26 @@ ORGQR_LAUNCHER_SCRATCH(double) #undef ORGQR_LAUNCHER_SCRATCH template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { throw unimplemented("lapack", "ormrq_scratchpad_size"); } template <> -std::int64_t ormrq_scratchpad_size(sycl::queue &queue, oneapi::math::side side, +std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { throw unimplemented("lapack", "ormrq_scratchpad_size"); } -#define ORMQRF_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t ormqr_scratchpad_size( \ - sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, \ - std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ - return 0; \ +#define ORMQRF_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t ormqr_scratchpad_size( \ + sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ + return 0; \ } ORMQRF_LAUNCHER_SCRATCH(float) @@ -2597,10 +2598,10 @@ ORMQRF_LAUNCHER_SCRATCH(double) #define ORMTR_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t ormtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ - oneapi::math::uplo uplo, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t lda, \ - std::int64_t ldc) { \ + std::int64_t ormtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ + oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, \ + std::int64_t n, std::int64_t lda, std::int64_t ldc) { \ return 0; \ } @@ -2609,11 +2610,11 @@ ORMTR_LAUNCHER_SCRATCH(double) #undef ORMTR_LAUNCHER_SCRATCH -#define POTRF_LAUNCHER_SCRATCH(TYPE) \ - template <> \ +#define POTRF_LAUNCHER_SCRATCH(TYPE) \ + template <> \ std::int64_t potrf_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - return 0; \ + std::int64_t n, std::int64_t lda) { \ + return 0; \ } POTRF_LAUNCHER_SCRATCH(float) @@ -2625,7 +2626,7 @@ POTRF_LAUNCHER_SCRATCH(std::complex) #define POTRS_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t potrs_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ + std::int64_t potrs_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t nrhs, std::int64_t lda, \ std::int64_t ldb) { \ return 0; \ @@ -2638,11 +2639,11 @@ POTRS_LAUNCHER_SCRATCH(std::complex) #undef POTRS_LAUNCHER_SCRATCH -#define POTRI_LAUNCHER_SCRATCH(TYPE) \ - template <> \ +#define POTRI_LAUNCHER_SCRATCH(TYPE) \ + template <> \ std::int64_t potri_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - return 0; \ + std::int64_t n, std::int64_t lda) { \ + return 0; \ } POTRI_LAUNCHER_SCRATCH(float) @@ -2652,11 +2653,11 @@ POTRI_LAUNCHER_SCRATCH(std::complex) #undef POTRI_LAUNCHER_SCRATCH -#define SYTRF_LAUNCHER_SCRATCH(TYPE) \ - template <> \ +#define SYTRF_LAUNCHER_SCRATCH(TYPE) \ + template <> \ std::int64_t sytrf_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - return 0; \ + std::int64_t n, std::int64_t lda) { \ + return 0; \ } SYTRF_LAUNCHER_SCRATCH(float) @@ -2666,12 +2667,12 @@ SYTRF_LAUNCHER_SCRATCH(std::complex) #undef SYTRF_LAUNCHER_SCRATCH -#define SYEVD_LAUNCHER_SCRATCH(TYPE) \ - template <> \ +#define SYEVD_LAUNCHER_SCRATCH(TYPE) \ + template <> \ std::int64_t syevd_scratchpad_size(sycl::queue & queue, oneapi::math::job jobz, \ oneapi::math::uplo uplo, std::int64_t n, \ - std::int64_t lda) { \ - return n; \ + std::int64_t lda) { \ + return n; \ } SYEVD_LAUNCHER_SCRATCH(float) @@ -2682,7 +2683,7 @@ SYEVD_LAUNCHER_SCRATCH(double) #define SYGVD_LAUNCHER_SCRATCH(TYPE) \ template <> \ std::int64_t sygvd_scratchpad_size(sycl::queue & queue, std::int64_t itype, \ - oneapi::math::job jobz, oneapi::math::uplo uplo, \ + oneapi::math::job jobz, oneapi::math::uplo uplo, \ std::int64_t n, std::int64_t lda, std::int64_t ldb) { \ return n; \ } @@ -2692,11 +2693,11 @@ SYGVD_LAUNCHER_SCRATCH(double) #undef SYGVD_LAUNCHER_SCRATCH -#define SYTRD_LAUNCHER_SCRATCH(TYPE) \ - template <> \ +#define SYTRD_LAUNCHER_SCRATCH(TYPE) \ + template <> \ std::int64_t sytrd_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - return 0; \ + std::int64_t n, std::int64_t lda) { \ + return 0; \ } SYTRD_LAUNCHER_SCRATCH(float) @@ -2704,13 +2705,13 @@ SYTRD_LAUNCHER_SCRATCH(double) #undef SYTRD_LAUNCHER_SCRATCH -#define TRTRS_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t trtrs_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - oneapi::math::transpose trans, oneapi::math::diag diag, \ - std::int64_t n, std::int64_t nrhs, std::int64_t lda, \ - std::int64_t ldb) { \ - return 0; \ +#define TRTRS_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t trtrs_scratchpad_size( \ + sycl::queue & queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, \ + oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, \ + std::int64_t ldb) { \ + return 0; \ } TRTRS_LAUNCHER_SCRATCH(float) @@ -2720,12 +2721,12 @@ TRTRS_LAUNCHER_SCRATCH(std::complex) #undef TRTRS_LAUNCHER_SCRATCH -#define UNGBR_LAUNCHER_SCRATCH(TYPE) \ - template <> \ +#define UNGBR_LAUNCHER_SCRATCH(TYPE) \ + template <> \ std::int64_t ungbr_scratchpad_size(sycl::queue & queue, oneapi::math::generate vec, \ - std::int64_t m, std::int64_t n, std::int64_t k, \ - std::int64_t lda) { \ - return 0; \ + std::int64_t m, std::int64_t n, std::int64_t k, \ + std::int64_t lda) { \ + return 0; \ } UNGBR_LAUNCHER_SCRATCH(std::complex) @@ -2745,11 +2746,11 @@ UNGQR_LAUNCHER_SCRATCH(std::complex) #undef UNGQR_LAUNCHER_SCRATCH -#define UNGTR_LAUNCHER_SCRATCH(TYPE) \ - template <> \ +#define UNGTR_LAUNCHER_SCRATCH(TYPE) \ + template <> \ std::int64_t ungtr_scratchpad_size(sycl::queue & queue, oneapi::math::uplo uplo, \ - std::int64_t n, std::int64_t lda) { \ - return 0; \ + std::int64_t n, std::int64_t lda) { \ + return 0; \ } UNGTR_LAUNCHER_SCRATCH(std::complex) @@ -2758,7 +2759,7 @@ UNGTR_LAUNCHER_SCRATCH(std::complex) #undef UNGTR_LAUNCHER_SCRATCH template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, +std::int64_t unmrq_scratchpad_size>(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, @@ -2766,20 +2767,18 @@ std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, onea throw unimplemented("lapack", "unmrq_scratchpad_size"); } template <> -std::int64_t unmrq_scratchpad_size>(sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc) { +std::int64_t unmrq_scratchpad_size>( + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, + std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { throw unimplemented("lapack", "unmrq_scratchpad_size"); } -#define UNMQR_LAUNCHER_SCRATCH(TYPE) \ - template <> \ - std::int64_t unmqr_scratchpad_size( \ - sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, \ - std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ - return 0; \ +#define UNMQR_LAUNCHER_SCRATCH(TYPE) \ + template <> \ + std::int64_t unmqr_scratchpad_size( \ + sycl::queue & queue, oneapi::math::side side, oneapi::math::transpose trans, \ + std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { \ + return 0; \ } UNMQR_LAUNCHER_SCRATCH(std::complex) @@ -2789,10 +2788,10 @@ UNMQR_LAUNCHER_SCRATCH(std::complex) #define UNMTR_LAUNCHER_SCRATCH(TYPE) \ template <> \ - std::int64_t unmtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ - oneapi::math::uplo uplo, oneapi::math::transpose trans, \ - std::int64_t m, std::int64_t n, std::int64_t lda, \ - std::int64_t ldc) { \ + std::int64_t unmtr_scratchpad_size(sycl::queue & queue, oneapi::math::side side, \ + oneapi::math::uplo uplo, \ + oneapi::math::transpose trans, std::int64_t m, \ + std::int64_t n, std::int64_t lda, std::int64_t ldc) { \ return 0; \ } diff --git a/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp b/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp index 306c43f38..264515d6e 100644 --- a/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp +++ b/src/lapack/backends/rocsolver/rocsolver_scope_handle.cpp @@ -46,7 +46,7 @@ thread_local rocsolver_handle RocsolverScopedContextHandler::handle_ #endif RocsolverScopedContextHandler::RocsolverScopedContextHandler(sycl::queue queue, - sycl::interop_handle &ih) + sycl::interop_handle& ih) : ih(ih), needToRecover_(false) { placedContext_ = new sycl::context(queue.get_context()); @@ -76,8 +76,8 @@ RocsolverScopedContextHandler::~RocsolverScopedContextHandler() noexcept(false) delete placedContext_; } -void ContextCallback(void *userData) { - auto *ptr = static_cast *>(userData); +void ContextCallback(void* userData) { + auto* ptr = static_cast*>(userData); if (!ptr) { return; } @@ -95,7 +95,7 @@ void ContextCallback(void *userData) { } } -rocblas_handle RocsolverScopedContextHandler::get_handle(const sycl::queue &queue) { +rocblas_handle RocsolverScopedContextHandler::get_handle(const sycl::queue& queue) { auto hipDevice = ih.get_native_device(); hipError_t hipErr; hipCtx_t desired; @@ -142,10 +142,10 @@ rocblas_handle RocsolverScopedContextHandler::get_handle(const sycl::queue &queu return handle; } -hipStream_t RocsolverScopedContextHandler::get_stream(const sycl::queue &queue) { +hipStream_t RocsolverScopedContextHandler::get_stream(const sycl::queue& queue) { return sycl::get_native(queue); } -sycl::context RocsolverScopedContextHandler::get_context(const sycl::queue &queue) { +sycl::context RocsolverScopedContextHandler::get_context(const sycl::queue& queue) { return queue.get_context(); } diff --git a/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp b/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp index e1036e034..4d3b12a8c 100644 --- a/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_scope_handle.hpp @@ -52,23 +52,23 @@ namespace rocsolver { class RocsolverScopedContextHandler { hipCtx_t original_; - sycl::context *placedContext_; + sycl::context* placedContext_; bool needToRecover_; - sycl::interop_handle &ih; + sycl::interop_handle& ih; #ifdef ONEMATH_PI_INTERFACE_REMOVED static thread_local rocsolver_handle handle_helper; #else static thread_local rocsolver_handle handle_helper; #endif - hipStream_t get_stream(const sycl::queue &queue); - sycl::context get_context(const sycl::queue &queue); + hipStream_t get_stream(const sycl::queue& queue); + sycl::context get_context(const sycl::queue& queue); public: - RocsolverScopedContextHandler(sycl::queue queue, sycl::interop_handle &ih); + RocsolverScopedContextHandler(sycl::queue queue, sycl::interop_handle& ih); ~RocsolverScopedContextHandler() noexcept(false); - rocblas_handle get_handle(const sycl::queue &queue); + rocblas_handle get_handle(const sycl::queue& queue); // This is a work-around function for reinterpret_casting the memory. This // will be fixed when SYCL-2020 has been implemented for Pi backend. template diff --git a/src/lapack/backends/rocsolver/rocsolver_task.hpp b/src/lapack/backends/rocsolver/rocsolver_task.hpp index 061e34504..c89734bd3 100644 --- a/src/lapack/backends/rocsolver/rocsolver_task.hpp +++ b/src/lapack/backends/rocsolver/rocsolver_task.hpp @@ -51,9 +51,9 @@ namespace lapack { namespace rocsolver { template -static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { +static inline void host_task_internal(H& cgh, sycl::queue queue, F f) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - cgh.ext_codeplay_enqueue_native_command([f, queue](sycl::interop_handle ih){ + cgh.ext_codeplay_enqueue_native_command([f, queue](sycl::interop_handle ih) { #else cgh.host_task([f, queue](cl::sycl::interop_handle ih) { #endif @@ -63,7 +63,7 @@ static inline void host_task_internal(H &cgh, sycl::queue queue, F f) { } template -static inline void onemath_rocsolver_host_task(H &cgh, sycl::queue queue, F f) { +static inline void onemath_rocsolver_host_task(H& cgh, sycl::queue queue, F f) { (void)host_task_internal(cgh, queue, f); } diff --git a/src/lapack/function_table.hpp b/src/lapack/function_table.hpp index 9f72752bd..93c56450f 100644 --- a/src/lapack/function_table.hpp +++ b/src/lapack/function_table.hpp @@ -32,1808 +32,1812 @@ typedef struct { int version; - void (*cgebrd_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*dgebrd_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sgebrd_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*zgebrd_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tauq, - sycl::buffer> &taup, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*sgerqf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*dgerqf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*cgerqf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zgerqf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*cgeqrf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*dgeqrf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sgeqrf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*zgeqrf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*cgetrf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*dgetrf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sgetrf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*zgetrf_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*cgetri_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*dgetri_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sgetri_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*zgetri_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*cgetrs_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*dgetrs_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sgetrs_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*zgetrs_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*dgesvd_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sgesvd_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*cgesvd_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zgesvd_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, - std::int64_t m, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*cheevd_sycl)(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zheevd_sycl)(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*chegvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + void (*cgebrd_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*dgebrd_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*sgebrd_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*zgebrd_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tauq, + sycl::buffer>& taup, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*sgerqf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*dgerqf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*cgerqf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zgerqf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*cgeqrf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*dgeqrf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*sgeqrf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*zgeqrf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*cgetrf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*dgetrf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*sgetrf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*zgetrf_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*cgetri_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*dgetri_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*sgetri_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*zgetri_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*cgetrs_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*dgetrs_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*sgetrs_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*zgetrs_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*dgesvd_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*sgesvd_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*cgesvd_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zgesvd_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, + std::int64_t m, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*cheevd_sycl)(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zheevd_sycl)(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*chegvd_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*zhegvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, + void (*zhegvd_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*chetrd_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zhetrd_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &d, sycl::buffer &e, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*chetrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zhetrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*sorgbr_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*dorgbr_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*dorgqr_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sorgqr_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sorgtr_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dorgtr_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*sormtr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*chetrd_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zhetrd_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& d, sycl::buffer& e, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*chetrf_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zhetrf_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*sorgbr_sycl)(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*dorgbr_sycl)(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*dorgqr_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*sorgqr_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*sorgtr_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*dorgtr_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*sormtr_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*dormtr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + void (*dormtr_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*sormrq_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*dormrq_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*dormqr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*sormqr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &c, - std::int64_t ldc, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*spotrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*dpotrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*cpotrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zpotrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*spotri_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*dpotri_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*cpotri_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zpotri_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*spotrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*dpotrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &b, std::int64_t ldb, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*cpotrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zpotrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*dsyevd_sycl)(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*ssyevd_sycl)(sycl::queue &queue, oneapi::math::job jobz, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*dsygvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*ssygvd_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &w, sycl::buffer &scratchpad, - std::int64_t scratchpad_size); - void (*dsytrd_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*ssytrd_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*ssytrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dsytrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*csytrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zsytrf_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*ctrtrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*sormrq_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*dormrq_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*dormqr_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*sormqr_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, + std::int64_t ldc, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*spotrf_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*dpotrf_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*cpotrf_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zpotrf_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*spotri_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*dpotri_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*cpotri_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zpotri_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*spotrs_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*dpotrs_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*cpotrs_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zpotrs_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*dsyevd_sycl)(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*ssyevd_sycl)(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*dsygvd_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*ssygvd_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& w, sycl::buffer& scratchpad, + std::int64_t scratchpad_size); + void (*dsytrd_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*ssytrd_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*ssytrf_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*dsytrf_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*csytrf_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zsytrf_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*ctrtrs_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*dtrtrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + void (*dtrtrs_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*strtrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + void (*strtrs_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*ztrtrs_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, + void (*ztrtrs_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*cungbr_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zungbr_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*cungqr_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zungqr_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*cungtr_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*zungtr_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - void (*cunmrq_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*cungbr_sycl)(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zungbr_sycl)(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*cungqr_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zungqr_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*cungtr_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*zungtr_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + void (*cunmrq_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*zunmrq_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, + void (*zunmrq_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*cunmqr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, + void (*cunmqr_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*zunmqr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::transpose trans, + void (*zunmqr_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*cunmtr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + void (*cunmtr_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*zunmtr_sycl)(sycl::queue &queue, oneapi::math::side side, oneapi::math::uplo uplo, + void (*zunmtr_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, - sycl::buffer> &c, std::int64_t ldc, - sycl::buffer> &scratchpad, - std::int64_t scratchpad_size); - sycl::event (*cgebrd_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgebrd_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *d, double *e, double *tauq, - double *taup, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgebrd_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *d, float *e, float *tauq, float *taup, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgebrd_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgerqf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, + sycl::buffer>& c, std::int64_t ldc, + sycl::buffer>& scratchpad, + std::int64_t scratchpad_size); + sycl::event (*cgebrd_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dgebrd_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* d, double* e, double* tauq, + double* taup, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*sgebrd_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* d, float* e, float* tauq, float* taup, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*zgebrd_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*sgerqf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgerqf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, + const std::vector& dependencies); + sycl::event (*dgerqf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgerqf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*cgerqf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgerqf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zgerqf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgeqrf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*cgeqrf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgeqrf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *tau, double *scratchpad, + const std::vector& dependencies); + sycl::event (*dgeqrf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* tau, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgeqrf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *tau, float *scratchpad, + const std::vector& dependencies); + sycl::event (*sgeqrf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgeqrf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zgeqrf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgetrf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgetrf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, std::int64_t *ipiv, double *scratchpad, + const std::vector& dependencies); + sycl::event (*cgetrf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dgetrf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgetrf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, std::int64_t *ipiv, float *scratchpad, + const std::vector& dependencies); + sycl::event (*sgetrf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgetrf_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgetri_usm_sycl)(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgetri_usm_sycl)(sycl::queue &queue, std::int64_t n, double *a, std::int64_t lda, - std::int64_t *ipiv, double *scratchpad, + const std::vector& dependencies); + sycl::event (*zgetrf_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*cgetri_usm_sycl)(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dgetri_usm_sycl)(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, + std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgetri_usm_sycl)(sycl::queue &queue, std::int64_t n, float *a, std::int64_t lda, - std::int64_t *ipiv, float *scratchpad, + const std::vector& dependencies); + sycl::event (*sgetri_usm_sycl)(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, + std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgetri_usm_sycl)(sycl::queue &queue, std::int64_t n, std::complex *a, - std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgetrs_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgetrs_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t *ipiv, double *b, std::int64_t ldb, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgetrs_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t *ipiv, float *b, std::int64_t ldb, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgetrs_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgesvd_usm_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + const std::vector& dependencies); + sycl::event (*zgetri_usm_sycl)(sycl::queue& queue, std::int64_t n, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*cgetrs_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dgetrs_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, double* a, std::int64_t lda, + std::int64_t* ipiv, double* b, std::int64_t ldb, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*sgetrs_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, float* a, std::int64_t lda, + std::int64_t* ipiv, float* b, std::int64_t ldb, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*zgetrs_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::int64_t* ipiv, std::complex* b, + std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dgesvd_usm_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *s, double *u, - std::int64_t ldu, double *vt, std::int64_t ldvt, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgesvd_usm_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + double* a, std::int64_t lda, double* s, double* u, + std::int64_t ldu, double* vt, std::int64_t ldvt, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*sgesvd_usm_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *s, float *u, std::int64_t ldu, - float *vt, std::int64_t ldvt, float *scratchpad, + float* a, std::int64_t lda, float* s, float* u, std::int64_t ldu, + float* vt, std::int64_t ldvt, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgesvd_usm_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + const std::vector& dependencies); + sycl::event (*cgesvd_usm_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, float *s, - std::complex *u, std::int64_t ldu, - std::complex *vt, std::int64_t ldvt, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgesvd_usm_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + std::complex* a, std::int64_t lda, float* s, + std::complex* u, std::int64_t ldu, + std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*zgesvd_usm_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *s, - std::complex *u, std::int64_t ldu, - std::complex *vt, std::int64_t ldvt, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cheevd_usm_sycl)(sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, float *w, std::complex *scratchpad, + std::complex* a, std::int64_t lda, double* s, + std::complex* u, std::int64_t ldu, + std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*cheevd_usm_sycl)(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, float* w, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zheevd_usm_sycl)(sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, double *w, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zheevd_usm_sycl)(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, double* w, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*chegvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - float *w, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*chegvd_usm_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + float* w, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zhegvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - double *w, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zhegvd_usm_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + double* w, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*chetrd_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tau, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*chetrd_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zhetrd_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tau, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zhetrd_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*chetrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zhetrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sorgbr_usm_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dorgbr_usm_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dorgqr_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sorgqr_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sorgtr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *scratchpad, + const std::vector& dependencies); + sycl::event (*chetrf_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*zhetrf_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*sorgbr_usm_sycl)(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dorgbr_usm_sycl)(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dorgqr_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*sorgqr_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*sorgtr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dorgtr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *scratchpad, + const std::vector& dependencies); + sycl::event (*dorgtr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* tau, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sormtr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + const std::vector& dependencies); + sycl::event (*sormtr_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, float *a, std::int64_t lda, - float *tau, float *c, std::int64_t ldc, float *scratchpad, + std::int64_t m, std::int64_t n, float* a, std::int64_t lda, + float* tau, float* c, std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dormtr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + const std::vector& dependencies); + sycl::event (*dormtr_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, double *a, std::int64_t lda, - double *tau, double *c, std::int64_t ldc, double *scratchpad, + std::int64_t m, std::int64_t n, double* a, std::int64_t lda, + double* tau, double* c, std::int64_t ldc, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sormrq_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + const std::vector& dependencies); + sycl::event (*sormrq_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, - std::int64_t ldc, float *scratchpad, + std::int64_t k, float* a, std::int64_t lda, float* tau, float* c, + std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dormrq_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + const std::vector& dependencies); + sycl::event (*dormrq_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, - double *c, std::int64_t ldc, double *scratchpad, + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* c, std::int64_t ldc, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dormqr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + const std::vector& dependencies); + sycl::event (*dormqr_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, - double *c, std::int64_t ldc, double *scratchpad, + std::int64_t k, double* a, std::int64_t lda, double* tau, + double* c, std::int64_t ldc, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sormqr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + const std::vector& dependencies); + sycl::event (*sormqr_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *c, - std::int64_t ldc, float *scratchpad, + std::int64_t k, float* a, std::int64_t lda, float* tau, float* c, + std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*spotrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *scratchpad, + const std::vector& dependencies); + sycl::event (*spotrf_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dpotrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *scratchpad, + const std::vector& dependencies); + sycl::event (*dpotrf_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cpotrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zpotrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*spotri_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *scratchpad, + const std::vector& dependencies); + sycl::event (*cpotrf_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*zpotrf_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*spotri_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dpotri_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *scratchpad, + const std::vector& dependencies); + sycl::event (*dpotri_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cpotri_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zpotri_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*spotrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, float *b, - std::int64_t ldb, float *scratchpad, + const std::vector& dependencies); + sycl::event (*cpotri_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*zpotri_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*spotrs_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, float* a, std::int64_t lda, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dpotrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, double *b, - std::int64_t ldb, double *scratchpad, + const std::vector& dependencies); + sycl::event (*dpotrs_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, double* a, std::int64_t lda, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cpotrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zpotrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dsyevd_usm_sycl)(sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, double *w, double *scratchpad, + const std::vector& dependencies); + sycl::event (*cpotrs_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*zpotrs_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dsyevd_usm_sycl)(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* w, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*ssyevd_usm_sycl)(sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, float *w, float *scratchpad, + const std::vector& dependencies); + sycl::event (*ssyevd_usm_sycl)(sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* w, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dsygvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, double *b, std::int64_t ldb, double *w, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*ssygvd_usm_sycl)(sycl::queue &queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *w, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dsytrd_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, double *d, double *e, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*ssytrd_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, float *d, float *e, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*ssytrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, std::int64_t *ipiv, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dsytrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, std::int64_t *ipiv, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*csytrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zsytrf_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*ctrtrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + const std::vector& dependencies); + sycl::event (*dsygvd_usm_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* b, std::int64_t ldb, double* w, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*ssygvd_usm_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* b, std::int64_t ldb, float* w, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dsytrd_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, double* d, double* e, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*ssytrd_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, float* d, float* e, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*ssytrf_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + float* a, std::int64_t lda, std::int64_t* ipiv, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dsytrf_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + double* a, std::int64_t lda, std::int64_t* ipiv, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*csytrf_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*zsytrf_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*ctrtrs_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dtrtrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dtrtrs_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, - double *b, std::int64_t ldb, double *scratchpad, + std::int64_t n, std::int64_t nrhs, double* a, std::int64_t lda, + double* b, std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*strtrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + const std::vector& dependencies); + sycl::event (*strtrs_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, - float *b, std::int64_t ldb, float *scratchpad, + std::int64_t n, std::int64_t nrhs, float* a, std::int64_t lda, + float* b, std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*ztrtrs_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + const std::vector& dependencies); + sycl::event (*ztrtrs_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cungbr_usm_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zungbr_usm_sycl)(sycl::queue &queue, oneapi::math::generate vec, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cungqr_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*cungbr_usm_sycl)(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*zungbr_usm_sycl)(sycl::queue& queue, oneapi::math::generate vec, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*cungqr_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zungqr_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zungqr_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cungtr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*cungtr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zungtr_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zungtr_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cunmrq_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + const std::vector& dependencies); + sycl::event (*cunmrq_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zunmrq_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + const std::vector& dependencies); + sycl::event (*zunmrq_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cunmqr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + const std::vector& dependencies); + sycl::event (*cunmqr_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zunmqr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + const std::vector& dependencies); + sycl::event (*zunmqr_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, - std::int64_t ldc, std::complex *scratchpad, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, + std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cunmtr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + const std::vector& dependencies); + sycl::event (*cunmtr_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zunmtr_usm_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*zunmtr_usm_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - void (*sgeqrf_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, + std::int64_t m, std::int64_t n, std::complex* a, + std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + void (*sgeqrf_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*dgeqrf_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, + void (*dgeqrf_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*cgeqrf_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, + void (*cgeqrf_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*zgeqrf_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, + void (*zgeqrf_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*sgetri_batch_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &a, + void (*sgetri_batch_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*dgetri_batch_sycl)(sycl::queue &queue, std::int64_t n, sycl::buffer &a, + void (*dgetri_batch_sycl)(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*cgetri_batch_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + void (*cgetri_batch_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*zgetri_batch_sycl)(sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + void (*zgetri_batch_sycl)(sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*sgetrs_batch_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, + void (*sgetrs_batch_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dgetrs_batch_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer &b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*dgetrs_batch_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*cgetrs_batch_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*cgetrs_batch_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*zgetrs_batch_sycl)(sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, + void (*zgetrs_batch_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer> &b, std::int64_t ldb, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*sgetrf_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + void (*sgetrf_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*dgetrf_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, + void (*dgetrf_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*cgetrf_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + void (*cgetrf_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*zgetrf_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, + void (*zgetrf_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*sorgqr_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, + void (*sorgqr_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*dorgqr_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer &scratchpad, + void (*dorgqr_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*spotrf_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, + void (*spotrf_batch_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*dpotrf_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, + void (*dpotrf_batch_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size); - void (*cpotrf_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, + void (*cpotrf_batch_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*zpotrf_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, + void (*zpotrf_batch_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*spotrs_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + void (*spotrs_batch_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*dpotrs_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*dpotrs_batch_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size); - void (*cpotrs_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, + sycl::buffer& scratchpad, std::int64_t scratchpad_size); + void (*cpotrs_batch_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*zpotrs_batch_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, + void (*zpotrs_batch_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &b, std::int64_t ldb, + sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*cungqr_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, + void (*cungqr_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - void (*zungqr_batch_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, + void (*zungqr_batch_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, + sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size); - sycl::event (*sgeqrf_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, std::int64_t stride_a, - float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, + sycl::event (*sgeqrf_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + float* a, std::int64_t lda, std::int64_t stride_a, + float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgeqrf_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, std::int64_t stride_a, - double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, + const std::vector& dependencies); + sycl::event (*dgeqrf_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgeqrf_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, + const std::vector& dependencies); + sycl::event (*cgeqrf_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgeqrf_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, + const std::vector& dependencies); + sycl::event (*zgeqrf_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgetrf_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, float *scratchpad, + const std::vector& dependencies); + sycl::event (*sgetrf_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgetrf_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, double *scratchpad, + const std::vector& dependencies); + sycl::event (*dgetrf_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgetrf_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + const std::vector& dependencies); + sycl::event (*cgetrf_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgetrf_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + const std::vector& dependencies); + sycl::event (*zgetrf_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgetri_batch_usm_sycl)(sycl::queue &queue, std::int64_t n, float *a, + const std::vector& dependencies); + sycl::event (*sgetri_batch_usm_sycl)(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, float *scratchpad, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgetri_batch_usm_sycl)(sycl::queue &queue, std::int64_t n, double *a, + const std::vector& dependencies); + sycl::event (*dgetri_batch_usm_sycl)(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, double *scratchpad, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgetri_batch_usm_sycl)(sycl::queue &queue, std::int64_t n, std::complex *a, + const std::vector& dependencies); + sycl::event (*cgetri_batch_usm_sycl)(sycl::queue& queue, std::int64_t n, std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, std::complex *scratchpad, + std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgetri_batch_usm_sycl)(sycl::queue &queue, std::int64_t n, - std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, + const std::vector& dependencies); + sycl::event (*zgetri_batch_usm_sycl)(sycl::queue& queue, std::int64_t n, + std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgetrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, float *a, + const std::vector& dependencies); + sycl::event (*sgetrs_batch_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, float *b, + std::int64_t* ipiv, std::int64_t stride_ipiv, float* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, float *scratchpad, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgetrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, double *a, + const std::vector& dependencies); + sycl::event (*dgetrs_batch_usm_sycl)(sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, double* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, double *b, + std::int64_t* ipiv, std::int64_t stride_ipiv, double* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, double *scratchpad, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*cgetrs_batch_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); sycl::event (*zgetrs_batch_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::complex *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sorgqr_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*sorgqr_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dorgqr_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, + const std::vector& dependencies); + sycl::event (*dorgqr_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*spotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, float *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dpotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, double *scratchpad, - std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cpotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*spotrf_batch_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, float* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dpotrf_batch_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, double* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*cpotrf_batch_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zpotrf_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, + const std::vector& dependencies); + sycl::event (*zpotrf_batch_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*spotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dpotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cpotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, + const std::vector& dependencies); + sycl::event (*spotrs_batch_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, float* a, + std::int64_t lda, std::int64_t stride_a, float* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, float* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dpotrs_batch_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, double* a, + std::int64_t lda, std::int64_t stride_a, double* b, + std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*cpotrs_batch_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zpotrs_batch_usm_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::complex *a, + const std::vector& dependencies); + sycl::event (*zpotrs_batch_usm_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex *b, std::int64_t ldb, + std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cungqr_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, + const std::vector& dependencies); + sycl::event (*cungqr_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zungqr_batch_usm_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, + const std::vector& dependencies); + sycl::event (*zungqr_batch_usm_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, std::int64_t batch_size, - std::complex *scratchpad, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgeqrf_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgeqrf_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgeqrf_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*sgeqrf_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dgeqrf_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*cgeqrf_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgeqrf_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, - std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zgeqrf_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgetrf_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - float **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgetrf_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - double **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgetrf_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*sgetrf_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + float** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dgetrf_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + double** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*cgetrf_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgetrf_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, - std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zgetrf_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgetri_group_usm_sycl)(sycl::queue &queue, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgetri_group_usm_sycl)(sycl::queue &queue, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgetri_group_usm_sycl)(sycl::queue &queue, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*sgetri_group_usm_sycl)(sycl::queue& queue, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dgetri_group_usm_sycl)(sycl::queue& queue, std::int64_t* n, double** a, + std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*cgetri_group_usm_sycl)(sycl::queue& queue, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zgetri_group_usm_sycl)(sycl::queue &queue, std::int64_t *n, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, - std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zgetri_group_usm_sycl)(sycl::queue& queue, std::int64_t* n, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sgetrs_group_usm_sycl)(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, float **a, - std::int64_t *lda, std::int64_t **ipiv, float **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, + const std::vector& dependencies); + sycl::event (*sgetrs_group_usm_sycl)(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, float** a, + std::int64_t* lda, std::int64_t** ipiv, float** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dgetrs_group_usm_sycl)(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, double **a, - std::int64_t *lda, std::int64_t **ipiv, double **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, + const std::vector& dependencies); + sycl::event (*dgetrs_group_usm_sycl)(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, double** a, + std::int64_t* lda, std::int64_t** ipiv, double** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cgetrs_group_usm_sycl)(sycl::queue &queue, oneapi::math::transpose *trans, - std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::complex **b, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*cgetrs_group_usm_sycl)(sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); + const std::vector& dependencies); sycl::event (*zgetrs_group_usm_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, std::int64_t **ipiv, std::complex **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*sorgqr_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dorgqr_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, double **a, std::int64_t *lda, - double **tau, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, std::int64_t** ipiv, std::complex** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*sorgqr_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dorgqr_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, double** a, std::int64_t* lda, + double** tau, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*spotrf_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, float **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dpotrf_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, double **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*cpotrf_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*spotrf_group_usm_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, float** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dpotrf_group_usm_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, double** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*cpotrf_group_usm_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zpotrf_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::complex **a, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, - std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zpotrf_group_usm_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::complex** a, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*spotrs_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, float **a, - std::int64_t *lda, float **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*dpotrs_group_usm_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, double **a, - std::int64_t *lda, double **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); + const std::vector& dependencies); + sycl::event (*spotrs_group_usm_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, float** a, + std::int64_t* lda, float** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); + sycl::event (*dpotrs_group_usm_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, double** a, + std::int64_t* lda, double** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies); sycl::event (*cpotrs_group_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies); + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies); sycl::event (*zpotrs_group_usm_sycl)( - sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies); - sycl::event (*cungqr_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex **a, - std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies); + sycl::event (*cungqr_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex** a, + std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); - sycl::event (*zungqr_group_usm_sycl)(sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::complex **a, - std::int64_t *lda, std::complex **tau, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, + const std::vector& dependencies); + sycl::event (*zungqr_group_usm_sycl)(sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::complex** a, + std::int64_t* lda, std::complex** tau, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies); + const std::vector& dependencies); - std::int64_t (*sgebrd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*sgebrd_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*dgebrd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*dgebrd_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*cgebrd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*cgebrd_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*zgebrd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*zgebrd_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*sgerqf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*sgerqf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*dgerqf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*dgerqf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*cgerqf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*cgerqf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*zgerqf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*zgerqf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*sgeqrf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*sgeqrf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*dgeqrf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*dgeqrf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*cgeqrf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*cgeqrf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*zgeqrf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*zgeqrf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*sgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + std::int64_t (*sgesvd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); - std::int64_t (*dgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + std::int64_t (*dgesvd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); - std::int64_t (*cgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + std::int64_t (*cgesvd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); - std::int64_t (*zgesvd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::jobsvd jobu, + std::int64_t (*zgesvd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); - std::int64_t (*sgetrf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*sgetrf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*dgetrf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*dgetrf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*cgetrf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*cgetrf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*zgetrf_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*zgetrf_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda); - std::int64_t (*sgetri_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t n, + std::int64_t (*sgetri_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t n, std::int64_t lda); - std::int64_t (*dgetri_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t n, + std::int64_t (*dgetri_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t n, std::int64_t lda); - std::int64_t (*cgetri_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t n, + std::int64_t (*cgetri_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t n, std::int64_t lda); - std::int64_t (*zgetri_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t n, + std::int64_t (*zgetri_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t n, std::int64_t lda); - std::int64_t (*sgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + std::int64_t (*sgetrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*dgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + std::int64_t (*dgetrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*cgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + std::int64_t (*cgetrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*zgetrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::transpose trans, + std::int64_t (*zgetrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*cheevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::job jobz, + std::int64_t (*cheevd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zheevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::job jobz, + std::int64_t (*zheevd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*chegvd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t itype, + std::int64_t (*chegvd_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); - std::int64_t (*zhegvd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t itype, + std::int64_t (*zhegvd_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); - std::int64_t (*chetrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*chetrd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zhetrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*zhetrd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*chetrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*chetrf_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zhetrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*zhetrf_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*sorgbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::generate vect, + std::int64_t (*sorgbr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*dorgbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::generate vect, + std::int64_t (*dorgbr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*sorgtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*sorgtr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dorgtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*dorgtr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*sorgqr_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*sorgqr_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*dorgqr_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*dorgqr_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*sormrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*sormrq_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*dormrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*dormrq_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*sormqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*sormqr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*dormqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*dormqr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*sormtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*sormtr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); - std::int64_t (*dormtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*dormtr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); - std::int64_t (*spotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*spotrf_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dpotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*dpotrf_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*cpotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*cpotrf_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zpotrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*zpotrf_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*spotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*spotrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*dpotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*dpotrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*cpotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*cpotrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*zpotrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*zpotrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*spotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*spotri_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dpotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*dpotri_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*cpotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*cpotri_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zpotri_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*zpotri_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*ssytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*ssytrf_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dsytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*dsytrf_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*csytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*csytrf_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zsytrf_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*zsytrf_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*ssyevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::job jobz, + std::int64_t (*ssyevd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dsyevd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::job jobz, + std::int64_t (*dsyevd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*ssygvd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t itype, + std::int64_t (*ssygvd_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); - std::int64_t (*dsygvd_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t itype, + std::int64_t (*dsygvd_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb); - std::int64_t (*ssytrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*ssytrd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*dsytrd_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*dsytrd_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*strtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*strtrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*dtrtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*dtrtrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*ctrtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*ctrtrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*ztrtrs_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*ztrtrs_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); - std::int64_t (*cungbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::generate vect, + std::int64_t (*cungbr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*zungbr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::generate vect, + std::int64_t (*zungbr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*cungqr_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*cungqr_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*zungqr_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, std::int64_t n, + std::int64_t (*zungqr_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda); - std::int64_t (*cungtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*cungtr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*zungtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*zungtr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda); - std::int64_t (*cunmrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*cunmrq_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*zunmrq_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*zunmrq_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*cunmqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*cunmqr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*zunmqr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*zunmqr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc); - std::int64_t (*cunmtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*cunmtr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); - std::int64_t (*zunmtr_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::side side, + std::int64_t (*zunmtr_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc); - std::int64_t (*sgetrf_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*sgetrf_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); - std::int64_t (*dgetrf_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*dgetrf_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); - std::int64_t (*cgetrf_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*cgetrf_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); - std::int64_t (*zgetrf_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*zgetrf_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); - std::int64_t (*sgetri_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t n, + std::int64_t (*sgetri_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); - std::int64_t (*dgetri_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t n, + std::int64_t (*dgetri_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); - std::int64_t (*cgetri_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t n, + std::int64_t (*cgetri_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); - std::int64_t (*zgetri_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t n, + std::int64_t (*zgetri_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size); std::int64_t (*sgetrs_batch_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); std::int64_t (*dgetrs_batch_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); std::int64_t (*cgetrs_batch_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); std::int64_t (*zgetrs_batch_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - std::int64_t (*sgeqrf_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*sgeqrf_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); - std::int64_t (*dgeqrf_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*dgeqrf_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); - std::int64_t (*cgeqrf_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*cgeqrf_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); - std::int64_t (*zgeqrf_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*zgeqrf_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); - std::int64_t (*spotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*spotrf_batch_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); - std::int64_t (*dpotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*dpotrf_batch_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); - std::int64_t (*cpotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*cpotrf_batch_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); - std::int64_t (*zpotrf_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*zpotrf_batch_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size); - std::int64_t (*spotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*spotrs_batch_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - std::int64_t (*dpotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*dpotrs_batch_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - std::int64_t (*cpotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*cpotrs_batch_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - std::int64_t (*zpotrs_batch_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo uplo, + std::int64_t (*zpotrs_batch_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size); - std::int64_t (*sorgqr_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*sorgqr_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); - std::int64_t (*dorgqr_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*dorgqr_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); - std::int64_t (*cungqr_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*cungqr_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); - std::int64_t (*zungqr_batch_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t m, + std::int64_t (*zungqr_batch_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size); - std::int64_t (*sgetrf_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + std::int64_t (*sgetrf_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*dgetrf_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + std::int64_t* group_sizes); + std::int64_t (*dgetrf_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*cgetrf_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + std::int64_t* group_sizes); + std::int64_t (*cgetrf_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*zgetrf_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + std::int64_t* group_sizes); + std::int64_t (*zgetrf_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*sgetri_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*dgetri_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*cgetri_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*zgetri_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); + std::int64_t* group_sizes); + std::int64_t (*sgetri_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); + std::int64_t (*dgetri_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); + std::int64_t (*cgetri_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); + std::int64_t (*zgetri_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); std::int64_t (*sgetrs_group_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); std::int64_t (*dgetrs_group_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); std::int64_t (*cgetrs_group_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); std::int64_t (*zgetrs_group_scratchpad_size_sycl)( - sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes); - std::int64_t (*sgeqrf_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + sycl::queue& queue, oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes); + std::int64_t (*sgeqrf_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*dgeqrf_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + std::int64_t* group_sizes); + std::int64_t (*dgeqrf_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*cgeqrf_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + std::int64_t* group_sizes); + std::int64_t (*cgeqrf_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*zgeqrf_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + std::int64_t* group_sizes); + std::int64_t (*zgeqrf_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*sorgqr_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*dorgqr_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*spotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, + std::int64_t* group_sizes); + std::int64_t (*sorgqr_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); + std::int64_t (*dorgqr_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); + std::int64_t (*spotrf_group_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*dpotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, + std::int64_t* group_sizes); + std::int64_t (*dpotrf_group_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*cpotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, + std::int64_t* group_sizes); + std::int64_t (*cpotrf_group_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*zpotrf_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *lda, + std::int64_t* group_sizes); + std::int64_t (*zpotrf_group_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*spotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, + std::int64_t* group_sizes); + std::int64_t (*spotrs_group_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*dpotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, + std::int64_t* group_sizes); + std::int64_t (*dpotrs_group_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*cpotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, + std::int64_t* group_sizes); + std::int64_t (*cpotrs_group_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*zpotrs_group_scratchpad_size_sycl)(sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, - std::int64_t *lda, std::int64_t *ldb, + std::int64_t* group_sizes); + std::int64_t (*zpotrs_group_scratchpad_size_sycl)(sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, + std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*cungqr_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); - std::int64_t (*zungqr_group_scratchpad_size_sycl)(sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes); + std::int64_t* group_sizes); + std::int64_t (*cungqr_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); + std::int64_t (*zungqr_group_scratchpad_size_sycl)(sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes); } lapack_function_table_t; diff --git a/src/lapack/lapack_loader.cpp b/src/lapack/lapack_loader.cpp index 3283d458c..410c04575 100644 --- a/src/lapack/lapack_loader.cpp +++ b/src/lapack/lapack_loader.cpp @@ -30,2063 +30,2065 @@ namespace detail { static oneapi::math::detail::table_initializer function_tables; -void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer> &tauq, - sycl::buffer> &taup, sycl::buffer> &scratchpad, +void gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer>& tauq, + sycl::buffer>& taup, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgebrd_sycl(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tauq, sycl::buffer &taup, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tauq, sycl::buffer& taup, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgebrd_sycl(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tauq, sycl::buffer &taup, sycl::buffer &scratchpad, +void gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tauq, sycl::buffer& taup, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgebrd_sycl(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -void gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer> &tauq, - sycl::buffer> &taup, sycl::buffer> &scratchpad, +void gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer>& tauq, + sycl::buffer>& taup, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgebrd_sycl(queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size); } -void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgerqf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgerqf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgerqf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgerqf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgeqrf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgeqrf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgeqrf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgeqrf_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size); } -void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgetrf_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgetrf_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgetrf_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, +void getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgetrf_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgetri_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgetri_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer &scratchpad, +void getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgetri_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, +void getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgetri_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, sycl::buffer> &b, - std::int64_t ldb, sycl::buffer> &scratchpad, +void getrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, sycl::buffer>& b, + std::int64_t ldb, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgetrs_sycl(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void getrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgetrs_sycl(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void getrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer& b, std::int64_t ldb, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgetrs_sycl(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void getrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgetrs_sycl(queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size); } -void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, +void gesvd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgesvd_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &s, sycl::buffer &u, std::int64_t ldu, - sycl::buffer &vt, std::int64_t ldvt, sycl::buffer &scratchpad, +void gesvd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& s, sycl::buffer& u, std::int64_t ldu, + sycl::buffer& vt, std::int64_t ldvt, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgesvd_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, +void gesvd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& a, std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgesvd_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, +void gesvd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &s, - sycl::buffer> &u, std::int64_t ldu, - sycl::buffer> &vt, std::int64_t ldvt, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& a, std::int64_t lda, sycl::buffer& s, + sycl::buffer>& u, std::int64_t ldu, + sycl::buffer>& vt, std::int64_t ldvt, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgesvd_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size); } -void heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &w, sycl::buffer> &scratchpad, +void heevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& w, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cheevd_sycl(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -void heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer> &a, - std::int64_t lda, sycl::buffer &w, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void heevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer>& a, + std::int64_t lda, sycl::buffer& w, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zheevd_sycl(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -void hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, +void hegvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, sycl::buffer& w, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].chegvd_sycl(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, +void hegvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, sycl::buffer &w, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, sycl::buffer& w, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zhegvd_sycl(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void hetrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].chetrd_sycl(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void hetrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zhetrd_sycl(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void hetrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].chetrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, +void hetrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zhetrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, std::int64_t lda, - sycl::buffer &tau, sycl::buffer &scratchpad, +void orgbr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sorgbr_sycl(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer &a, - std::int64_t lda, sycl::buffer &tau, sycl::buffer &scratchpad, +void orgbr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer& a, + std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dorgbr_sycl(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void orgqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dorgqr_sycl(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void orgqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sorgqr_sycl(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void orgtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sorgtr_sycl(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -void orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void orgtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dorgtr_sycl(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -void ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void ormtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sormtr_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void ormtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dormtr_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void ormrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sormrq_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void ormrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dormrq_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void ormqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dormqr_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void ormqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer &a, std::int64_t lda, sycl::buffer &tau, - sycl::buffer &c, std::int64_t ldc, sycl::buffer &scratchpad, + sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, + sycl::buffer& c, std::int64_t ldc, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sormqr_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, +void potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].spotrf_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, +void potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dpotrf_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cpotrf_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zpotrf_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, +void potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].spotri_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &scratchpad, +void potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dpotri_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cpotri_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zpotri_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size); } -void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].spotrs_sycl(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, sycl::buffer &b, - std::int64_t ldb, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dpotrs_sycl(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cpotrs_sycl(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - std::int64_t nrhs, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zpotrs_sycl(queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, +void syevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dsyevd_sycl(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -void syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, std::int64_t lda, - sycl::buffer &w, sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void syevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, std::int64_t lda, + sycl::buffer& w, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ssyevd_sycl(queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size); } -void sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void sygvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dsygvd_sycl(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer &a, - std::int64_t lda, sycl::buffer &b, std::int64_t ldb, sycl::buffer &w, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void sygvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, sycl::buffer& a, + std::int64_t lda, sycl::buffer& b, std::int64_t ldb, sycl::buffer& w, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ssygvd_sycl(queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size); } -void sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, - sycl::buffer &e, sycl::buffer &tau, sycl::buffer &scratchpad, +void sytrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, + sycl::buffer& e, sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dsytrd_sycl(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &d, sycl::buffer &e, - sycl::buffer &tau, sycl::buffer &scratchpad, +void sytrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& d, sycl::buffer& e, + sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ssytrd_sycl(queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size); } -void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ssytrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dsytrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, sycl::buffer &ipiv, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].csytrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer &ipiv, sycl::buffer> &scratchpad, +void sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer& ipiv, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zsytrf_sycl(queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size); } -void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ctrtrs_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dtrtrs_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer &a, std::int64_t lda, sycl::buffer &b, std::int64_t ldb, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, + std::int64_t ldb, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].strtrs_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &b, std::int64_t ldb, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& b, std::int64_t ldb, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].ztrtrs_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size); } -void ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void ungbr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cungbr_sycl(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer> &a, - std::int64_t lda, sycl::buffer> &tau, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { +void ungbr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, sycl::buffer>& a, + std::int64_t lda, sycl::buffer>& tau, + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zungbr_sycl(queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void ungqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cungqr_sycl(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void ungqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zungqr_sycl(queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size); } -void ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void ungtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cungtr_sycl(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -void ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &scratchpad, +void ungtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zungtr_sycl(queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size); } -void unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void unmrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& c, + std::int64_t ldc, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cunmrq_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void unmrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& c, + std::int64_t ldc, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zunmrq_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void unmqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& c, + std::int64_t ldc, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cunmqr_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void unmqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& c, + std::int64_t ldc, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zunmqr_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void unmtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& c, + std::int64_t ldc, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cunmtr_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -void unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +void unmtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, - sycl::buffer> &tau, sycl::buffer> &c, - std::int64_t ldc, sycl::buffer> &scratchpad, + sycl::buffer>& a, std::int64_t lda, + sycl::buffer>& tau, sycl::buffer>& c, + std::int64_t ldc, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zunmtr_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size); } -sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgebrd_usm_sycl( queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *d, double *e, double *tauq, double *taup, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, double* d, double* e, double* tauq, double* taup, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgebrd_usm_sycl( queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *d, float *e, float *tauq, float *taup, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + float* a, std::int64_t lda, float* d, float* e, float* tauq, float* taup, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgebrd_usm_sycl( queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -sycl::event gebrd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tauq, std::complex *taup, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gebrd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tauq, std::complex* taup, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgebrd_usm_sycl( queue, m, n, a, lda, d, e, tauq, taup, scratchpad, scratchpad_size, dependencies); } -sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgerqf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgerqf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgerqf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event gerqf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gerqf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgerqf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgeqrf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgeqrf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgeqrf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event geqrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgeqrf_usm_sycl(queue, m, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgetrf_usm_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - double *a, std::int64_t lda, std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + double* a, std::int64_t lda, std::int64_t* ipiv, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgetrf_usm_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + float* a, std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgetrf_usm_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrf(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgetrf_usm_sycl(queue, m, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgetri_usm_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *a, - std::int64_t lda, std::int64_t *ipiv, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double* a, + std::int64_t lda, std::int64_t* ipiv, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgetri_usm_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *a, - std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float* a, + std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgetri_usm_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getri(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgetri_usm_sycl(queue, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgetrs_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t *ipiv, double *b, std::int64_t ldb, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event getrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, double* a, std::int64_t lda, + std::int64_t* ipiv, double* b, std::int64_t ldb, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgetrs_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, std::int64_t *ipiv, - float *b, std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, float* a, std::int64_t lda, std::int64_t* ipiv, + float* b, std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgetrs_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t *ipiv, std::complex *b, std::int64_t ldb, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t* ipiv, std::complex* b, std::int64_t ldb, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgetrs_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double *a, - std::int64_t lda, double *s, double *u, std::int64_t ldu, double *vt, - std::int64_t ldvt, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gesvd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, double* a, + std::int64_t lda, double* s, double* u, std::int64_t ldu, double* vt, + std::int64_t ldvt, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgesvd_usm_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float *a, - std::int64_t lda, float *s, float *u, std::int64_t ldu, float *vt, - std::int64_t ldvt, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event gesvd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, float* a, + std::int64_t lda, float* s, float* u, std::int64_t ldu, float* vt, + std::int64_t ldvt, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgesvd_usm_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, - oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::complex *a, - std::int64_t lda, float *s, std::complex *u, std::int64_t ldu, - std::complex *vt, std::int64_t ldvt, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event gesvd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, + oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, + std::complex* a, std::int64_t lda, float* s, std::complex* u, + std::int64_t ldu, std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgesvd_usm_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event gesvd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::jobsvd jobu, +sycl::event gesvd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::complex *a, std::int64_t lda, double *s, std::complex *u, - std::int64_t ldu, std::complex *vt, std::int64_t ldvt, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* a, std::int64_t lda, double* s, std::complex* u, + std::int64_t ldu, std::complex* vt, std::int64_t ldvt, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgesvd_usm_sycl(queue, jobu, jobvt, m, n, a, lda, s, u, ldu, vt, ldvt, scratchpad, scratchpad_size, dependencies); } -sycl::event heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, - float *w, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event heevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, std::int64_t lda, + float* w, std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cheevd_usm_sycl( queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event heevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::complex *a, std::int64_t lda, - double *w, std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event heevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, std::complex* a, + std::int64_t lda, double* w, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].zheevd_usm_sycl( queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, +sycl::event hegvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, float *w, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, float* w, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].chegvd_usm_sycl( queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event hegvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, +sycl::event hegvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, - std::complex *a, std::int64_t lda, std::complex *b, - std::int64_t ldb, double *w, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { + std::complex* a, std::int64_t lda, std::complex* b, + std::int64_t ldb, double* w, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].zhegvd_usm_sycl( queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, float *d, float *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event hetrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, float* d, float* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].chetrd_usm_sycl( queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, double *d, double *e, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event hetrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, double* d, double* e, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].zhetrd_usm_sycl( queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event hetrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].chetrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event hetrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event hetrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zhetrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - float *tau, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event orgbr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + float* tau, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sorgbr_usm_sycl( queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - double *tau, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event orgbr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + double* tau, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dorgbr_usm_sycl( queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event orgqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].dorgqr_usm_sycl( queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event orgqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].sorgqr_usm_sycl( queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, float *a, std::int64_t lda, float *tau, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event orgtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, float* a, std::int64_t lda, float* tau, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].sorgtr_usm_sycl( queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event orgtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, double *a, std::int64_t lda, double *tau, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event orgtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, double* a, std::int64_t lda, double* tau, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].dorgtr_usm_sycl( queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event ormtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, float *tau, float *c, - std::int64_t ldc, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t n, float* a, std::int64_t lda, float* tau, float* c, + std::int64_t ldc, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sormtr_usm_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event ormtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, double *tau, double *c, - std::int64_t ldc, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t n, double* a, std::int64_t lda, double* tau, double* c, + std::int64_t ldc, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dormtr_usm_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event ormrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + float* a, std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sormrq_usm_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event ormrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + double* a, std::int64_t lda, double* tau, double* c, std::int64_t ldc, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dormrq_usm_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event ormqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - double *a, std::int64_t lda, double *tau, double *c, std::int64_t ldc, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + double* a, std::int64_t lda, double* tau, double* c, std::int64_t ldc, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dormqr_usm_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event ormqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event ormqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - float *a, std::int64_t lda, float *tau, float *c, std::int64_t ldc, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + float* a, std::int64_t lda, float* tau, float* c, std::int64_t ldc, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sormqr_usm_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, float *a, std::int64_t lda, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, float* a, std::int64_t lda, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].spotrf_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, double *a, std::int64_t lda, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, double* a, std::int64_t lda, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].dpotrf_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cpotrf_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zpotrf_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, float *a, std::int64_t lda, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, float* a, std::int64_t lda, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].spotri_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, double *a, std::int64_t lda, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, double* a, std::int64_t lda, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].dpotri_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cpotri_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potri(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potri(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zpotri_usm_sycl(queue, uplo, n, a, lda, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, float *b, - std::int64_t ldb, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, float* a, std::int64_t lda, float* b, + std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].spotrs_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, double *b, - std::int64_t ldb, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, double* a, std::int64_t lda, double* b, + std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dpotrs_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].cpotrs_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event potrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].zpotrs_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, double *a, std::int64_t lda, double *w, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event syevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, double* a, std::int64_t lda, double* w, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dsyevd_usm_sycl( queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event syevd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, float *a, std::int64_t lda, float *w, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event syevd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, + oneapi::math::uplo uplo, std::int64_t n, float* a, std::int64_t lda, float* w, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].ssyevd_usm_sycl( queue, jobz, uplo, n, a, lda, w, scratchpad, scratchpad_size, dependencies); } -sycl::event sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double *a, - std::int64_t lda, double *b, std::int64_t ldb, double *w, double *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event sygvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, double* a, + std::int64_t lda, double* b, std::int64_t ldb, double* w, double* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].dsygvd_usm_sycl( queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event sygvd(oneapi::math::device libkey, sycl::queue &queue, std::int64_t itype, - oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float *a, - std::int64_t lda, float *b, std::int64_t ldb, float *w, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event sygvd(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, + oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, float* a, + std::int64_t lda, float* b, std::int64_t ldb, float* w, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].ssygvd_usm_sycl( queue, itype, jobz, uplo, n, a, lda, b, ldb, w, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, double *a, std::int64_t lda, double *d, double *e, double *tau, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event sytrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, double* a, std::int64_t lda, double* d, double* e, double* tau, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dsytrd_usm_sycl( queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrd(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, float *a, std::int64_t lda, float *d, float *e, float *tau, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event sytrd(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, float* a, std::int64_t lda, float* d, float* e, float* tau, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].ssytrd_usm_sycl( queue, uplo, n, a, lda, d, e, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, float *a, std::int64_t lda, std::int64_t *ipiv, float *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, float* a, std::int64_t lda, std::int64_t* ipiv, float* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].ssytrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, double *a, std::int64_t lda, std::int64_t *ipiv, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, double* a, std::int64_t lda, std::int64_t* ipiv, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dsytrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].csytrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event sytrf(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, std::int64_t *ipiv, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event sytrf(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, std::int64_t* ipiv, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zsytrf_usm_sycl( queue, uplo, n, a, lda, ipiv, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, +sycl::event trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].ctrtrs_usm_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, +sycl::event trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, double *a, std::int64_t lda, double *b, std::int64_t ldb, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t nrhs, double* a, std::int64_t lda, double* b, std::int64_t ldb, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dtrtrs_usm_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, +sycl::event trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, float *a, std::int64_t lda, float *b, std::int64_t ldb, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t nrhs, float* a, std::int64_t lda, float* b, std::int64_t ldb, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].strtrs_usm_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event trtrs(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, +sycl::event trtrs(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::complex *b, std::int64_t ldb, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { + std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::complex* b, std::int64_t ldb, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].ztrtrs_usm_sycl(queue, uplo, trans, diag, n, nrhs, a, lda, b, ldb, scratchpad, scratchpad_size, dependencies); } -sycl::event ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event ungbr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].cungbr_usm_sycl( queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungbr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::generate vec, - std::int64_t m, std::int64_t n, std::int64_t k, std::complex *a, - std::int64_t lda, std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event ungbr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vec, + std::int64_t m, std::int64_t n, std::int64_t k, std::complex* a, + std::int64_t lda, std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].zungbr_usm_sycl( queue, vec, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event ungqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].cungqr_usm_sycl( queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event ungqr(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].zungqr_usm_sycl( queue, m, n, k, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event ungtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].cungtr_usm_sycl( queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event ungtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { +sycl::event ungtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].zungtr_usm_sycl( queue, uplo, n, a, lda, tau, scratchpad, scratchpad_size, dependencies); } -sycl::event unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event unmrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].cunmrq_usm_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmrq(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event unmrq(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].zunmrq_usm_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event unmqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].cunmqr_usm_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmqr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event unmqr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex *a, std::int64_t lda, std::complex *tau, - std::complex *c, std::int64_t ldc, std::complex *scratchpad, - std::int64_t scratchpad_size, const std::vector &dependencies) { + std::complex* a, std::int64_t lda, std::complex* tau, + std::complex* c, std::int64_t ldc, std::complex* scratchpad, + std::int64_t scratchpad_size, const std::vector& dependencies) { return function_tables[{ libkey, queue }].zunmqr_usm_sycl(queue, side, trans, m, n, k, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event unmtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cunmtr_usm_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -sycl::event unmtr(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::side side, +sycl::event unmtr(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::complex *tau, std::complex *c, std::int64_t ldc, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::int64_t n, std::complex* a, std::int64_t lda, + std::complex* tau, std::complex* c, std::int64_t ldc, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zunmtr_usm_sycl(queue, side, uplo, trans, m, n, a, lda, tau, c, ldc, scratchpad, scratchpad_size, dependencies); } -void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgeqrf_batch_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgeqrf_batch_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgeqrf_batch_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer> &tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer>& tau, std::int64_t stride_tau, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgeqrf_batch_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgetri_batch_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgetri_batch_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgetri_batch_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgetri_batch_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgetrs_batch_sycl(queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &ipiv, std::int64_t stride_ipiv, - sycl::buffer &b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, + sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgetrs_batch_sycl(queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, +void getrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgetrs_batch_sycl(queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void getrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer &ipiv, - std::int64_t stride_ipiv, sycl::buffer> &b, std::int64_t ldb, +void getrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, + std::int64_t stride_ipiv, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgetrs_batch_sycl(queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sgetrf_batch_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer &scratchpad, +void getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dgetrf_batch_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cgetrf_batch_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - sycl::buffer> &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer> &scratchpad, +void getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zgetrf_batch_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size); } -void orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].sorgqr_batch_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer &tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer &scratchpad, std::int64_t scratchpad_size) { +void orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, + sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dorgqr_batch_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, +void potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].spotrf_batch_sycl(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer &a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer &scratchpad, +void potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dpotrf_batch_sycl(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, +void potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cpotrf_batch_sycl(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, sycl::buffer> &a, std::int64_t lda, +void potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zpotrf_batch_sycl(queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, +void potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].spotrs_batch_sycl(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer &b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, sycl::buffer &scratchpad, +void potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, sycl::buffer& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].dpotrs_batch_sycl(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, +void potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cpotrs_batch_sycl(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, sycl::buffer> &a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer> &b, +void potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, sycl::buffer>& a, + std::int64_t lda, std::int64_t stride_a, sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zpotrs_batch_sycl(queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size); } -void ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, +void ungqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].cungqr_batch_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -void ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer> &a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer> &tau, +void ungqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, + std::int64_t k, sycl::buffer>& a, std::int64_t lda, + std::int64_t stride_a, sycl::buffer>& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer> &scratchpad, std::int64_t scratchpad_size) { + sycl::buffer>& scratchpad, std::int64_t scratchpad_size) { function_tables[{ libkey, queue }].zungqr_batch_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size); } -sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, - float *tau, std::int64_t stride_tau, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, float* a, std::int64_t lda, std::int64_t stride_a, + float* tau, std::int64_t stride_tau, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgeqrf_batch_usm_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, - double *tau, std::int64_t stride_tau, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, double* a, std::int64_t lda, std::int64_t stride_a, + double* tau, std::int64_t stride_tau, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgeqrf_batch_usm_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgeqrf_batch_usm_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgeqrf_batch_usm_sycl( queue, m, n, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgetrf_batch_usm_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgetrf_batch_usm_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, std::complex *scratchpad, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgetrf_batch_usm_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::int64_t *ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, std::complex *scratchpad, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::int64_t* ipiv, std::int64_t stride_ipiv, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgetrf_batch_usm_sycl( queue, m, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, float *scratchpad, +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgetri_batch_usm_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, double *scratchpad, +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgetri_batch_usm_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgetri_batch_usm_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::int64_t batch_size, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgetri_batch_usm_sycl( queue, n, a, lda, stride_a, ipiv, stride_ipiv, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, float* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgetrs_batch_usm_sycl( queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double *a, - std::int64_t lda, std::int64_t stride_a, std::int64_t *ipiv, - std::int64_t stride_ipiv, double *b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, double *scratchpad, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, double* a, + std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, + std::int64_t stride_ipiv, double* b, std::int64_t ldb, + std::int64_t stride_b, std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgetrs_batch_usm_sycl( queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgetrs_batch_usm_sycl( queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, - std::complex *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t *ipiv, std::int64_t stride_ipiv, std::complex *b, + std::complex* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t* ipiv, std::int64_t stride_ipiv, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgetrs_batch_usm_sycl( queue, trans, n, nrhs, a, lda, stride_a, ipiv, stride_ipiv, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, float *a, std::int64_t lda, - std::int64_t stride_a, float *tau, std::int64_t stride_tau, - std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, float* a, std::int64_t lda, + std::int64_t stride_a, float* tau, std::int64_t stride_tau, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sorgqr_batch_usm_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, double *a, std::int64_t lda, - std::int64_t stride_a, double *tau, std::int64_t stride_tau, - std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, double* a, std::int64_t lda, + std::int64_t stride_a, double* tau, std::int64_t stride_tau, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dorgqr_batch_usm_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, float *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, float* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].spotrf_batch_usm_sycl( queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, double *a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, double* a, std::int64_t lda, std::int64_t stride_a, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dpotrf_batch_usm_sycl( queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cpotrf_batch_usm_sycl( queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::complex *a, std::int64_t lda, +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::complex* a, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zpotrf_batch_usm_sycl( queue, uplo, n, a, lda, stride_a, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, float *a, std::int64_t lda, - std::int64_t stride_a, float *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, float* a, std::int64_t lda, + std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].spotrs_batch_usm_sycl( queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, double *a, std::int64_t lda, - std::int64_t stride_a, double *b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, double* a, std::int64_t lda, + std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, + std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dpotrs_batch_usm_sycl( queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *b, std::int64_t ldb, +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cpotrs_batch_usm_sycl( queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, std::complex *a, - std::int64_t lda, std::int64_t stride_a, std::complex *b, +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + std::int64_t n, std::int64_t nrhs, std::complex* a, + std::int64_t lda, std::int64_t stride_a, std::complex* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zpotrs_batch_usm_sycl( queue, uplo, n, nrhs, a, lda, stride_a, b, ldb, stride_b, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, +sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cungqr_batch_usm_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex *a, std::int64_t lda, - std::int64_t stride_a, std::complex *tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex *scratchpad, +sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, + std::int64_t n, std::int64_t k, std::complex* a, std::int64_t lda, + std::int64_t stride_a, std::complex* tau, std::int64_t stride_tau, + std::int64_t batch_size, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zungqr_batch_usm_sycl( queue, m, n, k, a, lda, stride_a, tau, stride_tau, batch_size, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgeqrf_group_usm_sycl( queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, double **a, std::int64_t *lda, double **tau, - std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, double** a, std::int64_t* lda, double** tau, + std::int64_t group_count, std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgeqrf_group_usm_sycl( queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgeqrf_group_usm_sycl( queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +sycl::event geqrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgeqrf_group_usm_sycl( queue, m, n, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, float **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, float** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgetrf_group_usm_sycl( queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, double **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, double *scratchpad, +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, double** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgetrf_group_usm_sycl( queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgetrf_group_usm_sycl( queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t **ipiv, std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrf_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t** ipiv, std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgetrf_group_usm_sycl( queue, m, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, float **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, float** a, + std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgetri_group_usm_sycl( queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, double **a, - std::int64_t *lda, std::int64_t **ipiv, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + double** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, double* scratchpad, + std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgetri_group_usm_sycl( queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgetri_group_usm_sycl( queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getri_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *n, - std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getri_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* n, + std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgetri_group_usm_sycl( queue, n, a, lda, ipiv, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - float **a, std::int64_t *lda, std::int64_t **ipiv, float **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, - float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + float** a, std::int64_t* lda, std::int64_t** ipiv, float** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes, + float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sgetrs_group_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - double **a, std::int64_t *lda, std::int64_t **ipiv, double **b, - std::int64_t *ldb, std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + double** a, std::int64_t* lda, std::int64_t** ipiv, double** b, + std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dgetrs_group_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cgetrs_group_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose *trans, std::int64_t *n, std::int64_t *nrhs, - std::complex **a, std::int64_t *lda, std::int64_t **ipiv, - std::complex **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +sycl::event getrs_batch(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose* trans, std::int64_t* n, std::int64_t* nrhs, + std::complex** a, std::int64_t* lda, std::int64_t** ipiv, + std::complex** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zgetrs_group_usm_sycl( queue, trans, n, nrhs, a, lda, ipiv, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, float **a, std::int64_t *lda, float **tau, - std::int64_t group_count, std::int64_t *group_sizes, float *scratchpad, +sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, float** a, std::int64_t* lda, float** tau, + std::int64_t group_count, std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].sorgqr_group_usm_sycl( queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, double **a, std::int64_t *lda, - double **tau, std::int64_t group_count, std::int64_t *group_sizes, - double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event orgqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, double** a, std::int64_t* lda, + double** tau, std::int64_t group_count, std::int64_t* group_sizes, + double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dorgqr_group_usm_sycl( queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, float **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, float** a, std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].spotrf_group_usm_sycl( queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, double **a, std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, double** a, std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dpotrf_group_usm_sycl( queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cpotrf_group_usm_sycl( queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::complex **a, std::int64_t *lda, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrf_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::complex** a, std::int64_t* lda, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zpotrf_group_usm_sycl( queue, uplo, n, a, lda, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, float **a, std::int64_t *lda, - float **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, float *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, float** a, std::int64_t* lda, + float** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].spotrs_group_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, double **a, std::int64_t *lda, - double **b, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes, double *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, double** a, std::int64_t* lda, + double** b, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].dpotrs_group_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cpotrs_group_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, - std::int64_t *n, std::int64_t *nrhs, std::complex **a, - std::int64_t *lda, std::complex **b, std::int64_t *ldb, - std::int64_t group_count, std::int64_t *group_sizes, - std::complex *scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { +sycl::event potrs_batch(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, + std::int64_t* n, std::int64_t* nrhs, std::complex** a, + std::int64_t* lda, std::complex** b, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes, + std::complex* scratchpad, std::int64_t scratchpad_size, + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zpotrs_group_usm_sycl( queue, uplo, n, nrhs, a, lda, b, ldb, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, std::complex **a, - std::int64_t *lda, std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::complex** a, + std::int64_t* lda, std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].cungqr_group_usm_sycl( queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } -sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *k, std::complex **a, - std::int64_t *lda, std::complex **tau, std::int64_t group_count, - std::int64_t *group_sizes, std::complex *scratchpad, +sycl::event ungqr_batch(oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* k, std::complex** a, + std::int64_t* lda, std::complex** tau, std::int64_t group_count, + std::int64_t* group_sizes, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector &dependencies) { + const std::vector& dependencies) { return function_tables[{ libkey, queue }].zungqr_group_usm_sycl( queue, m, n, k, a, lda, tau, group_count, group_sizes, scratchpad, scratchpad_size, dependencies); } template <> -std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sgebrd_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t gebrd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dgebrd_scratchpad_size_sycl(queue, m, n, lda); } template <> std::int64_t gebrd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cgebrd_scratchpad_size_sycl(queue, m, n, lda); } template <> std::int64_t gebrd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zgebrd_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sgerqf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t gerqf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dgerqf_scratchpad_size_sycl(queue, m, n, lda); } template <> std::int64_t gerqf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cgerqf_scratchpad_size_sycl(queue, m, n, lda); } template <> std::int64_t gerqf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zgerqf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sgeqrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t geqrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dgeqrf_scratchpad_size_sycl(queue, m, n, lda); } template <> std::int64_t geqrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cgeqrf_scratchpad_size_sycl(queue, m, n, lda); } template <> std::int64_t geqrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zgeqrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { @@ -2094,7 +2096,7 @@ std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::que lda, ldu, ldvt); } template <> -std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldu, std::int64_t ldvt) { @@ -2103,7 +2105,7 @@ std::int64_t gesvd_scratchpad_size(oneapi::math::device libkey, sycl::qu } template <> std::int64_t gesvd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, @@ -2113,7 +2115,7 @@ std::int64_t gesvd_scratchpad_size>(oneapi::math::device lib } template <> std::int64_t gesvd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, std::int64_t m, std::int64_t n, std::int64_t lda, @@ -2122,58 +2124,58 @@ std::int64_t gesvd_scratchpad_size>(oneapi::math::device li lda, ldu, ldvt); } template <> -std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sgetrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dgetrf_scratchpad_size_sycl(queue, m, n, lda); } template <> std::int64_t getrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cgetrf_scratchpad_size_sycl(queue, m, n, lda); } template <> std::int64_t getrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zgetrf_scratchpad_size_sycl(queue, m, n, lda); } template <> -std::int64_t getri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getri_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sgetri_scratchpad_size_sycl(queue, n, lda); } template <> -std::int64_t getri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getri_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dgetri_scratchpad_size_sycl(queue, n, lda); } template <> std::int64_t getri_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t n, + sycl::queue& queue, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cgetri_scratchpad_size_sycl(queue, n, lda); } template <> std::int64_t getri_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t n, + sycl::queue& queue, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zgetri_scratchpad_size_sycl(queue, n, lda); } template <> -std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].sgetrs_scratchpad_size_sycl(queue, trans, n, nrhs, lda, ldb); } template <> -std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].dgetrs_scratchpad_size_sycl(queue, trans, n, nrhs, @@ -2181,7 +2183,7 @@ std::int64_t getrs_scratchpad_size(oneapi::math::device libkey, sycl::qu } template <> std::int64_t getrs_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { @@ -2190,7 +2192,7 @@ std::int64_t getrs_scratchpad_size>(oneapi::math::device lib } template <> std::int64_t getrs_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { @@ -2199,7 +2201,7 @@ std::int64_t getrs_scratchpad_size>(oneapi::math::device li } template <> std::int64_t heevd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::job jobz, + sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cheevd_scratchpad_size_sycl(queue, jobz, uplo, n, @@ -2207,7 +2209,7 @@ std::int64_t heevd_scratchpad_size>(oneapi::math::device lib } template <> std::int64_t heevd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::job jobz, + sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].zheevd_scratchpad_size_sycl(queue, jobz, uplo, n, @@ -2215,7 +2217,7 @@ std::int64_t heevd_scratchpad_size>(oneapi::math::device li } template <> std::int64_t hegvd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t itype, + sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { @@ -2224,7 +2226,7 @@ std::int64_t hegvd_scratchpad_size>(oneapi::math::device lib } template <> std::int64_t hegvd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t itype, + sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { @@ -2233,68 +2235,70 @@ std::int64_t hegvd_scratchpad_size>(oneapi::math::device li } template <> std::int64_t hetrd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].chetrd_scratchpad_size_sycl(queue, uplo, n, lda); } template <> std::int64_t hetrd_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { return function_tables[{ libkey, queue }].zhetrd_scratchpad_size_sycl(queue, uplo, n, lda); } template <> std::int64_t hetrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].chetrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> std::int64_t hetrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { return function_tables[{ libkey, queue }].zhetrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].sorgbr_scratchpad_size_sycl(queue, vect, m, n, k, lda); } template <> -std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t orgbr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].dorgbr_scratchpad_size_sycl(queue, vect, m, n, k, lda); } template <> -std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].sorgtr_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t orgtr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dorgtr_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].sorgqr_scratchpad_size_sycl(queue, m, n, k, lda); } template <> -std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t orgqr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].dorgqr_scratchpad_size_sycl(queue, m, n, k, lda); } template <> -std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { @@ -2302,7 +2306,7 @@ std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, sycl::que k, lda, ldc); } template <> -std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { @@ -2310,7 +2314,7 @@ std::int64_t ormrq_scratchpad_size(oneapi::math::device libkey, sycl::qu k, lda, ldc); } template <> -std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { @@ -2318,7 +2322,7 @@ std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, sycl::que k, lda, ldc); } template <> -std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t ldc) { @@ -2326,7 +2330,7 @@ std::int64_t ormqr_scratchpad_size(oneapi::math::device libkey, sycl::qu k, lda, ldc); } template <> -std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { @@ -2334,7 +2338,7 @@ std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, sycl::que m, n, lda, ldc); } template <> -std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t ldc) { @@ -2342,38 +2346,39 @@ std::int64_t ormtr_scratchpad_size(oneapi::math::device libkey, sycl::qu m, n, lda, ldc); } template <> -std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].spotrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dpotrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> std::int64_t potrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cpotrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> std::int64_t potrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { return function_tables[{ libkey, queue }].zpotrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb) { +std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].spotrs_scratchpad_size_sycl(queue, uplo, n, nrhs, lda, ldb); } template <> -std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].dpotrs_scratchpad_size_sycl(queue, uplo, n, nrhs, lda, @@ -2381,7 +2386,7 @@ std::int64_t potrs_scratchpad_size(oneapi::math::device libkey, sycl::qu } template <> std::int64_t potrs_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].cpotrs_scratchpad_size_sycl(queue, uplo, n, nrhs, lda, @@ -2389,76 +2394,79 @@ std::int64_t potrs_scratchpad_size>(oneapi::math::device lib } template <> std::int64_t potrs_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb) { + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t nrhs, std::int64_t lda, + std::int64_t ldb) { return function_tables[{ libkey, queue }].zpotrs_scratchpad_size_sycl(queue, uplo, n, nrhs, lda, ldb); } template <> -std::int64_t potri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potri_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].spotri_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t potri_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potri_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dpotri_scratchpad_size_sycl(queue, uplo, n, lda); } template <> std::int64_t potri_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cpotri_scratchpad_size_sycl(queue, uplo, n, lda); } template <> std::int64_t potri_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { return function_tables[{ libkey, queue }].zpotri_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].ssytrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t sytrf_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dsytrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> std::int64_t sytrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].csytrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> std::int64_t sytrf_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { return function_tables[{ libkey, queue }].zsytrf_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].ssyevd_scratchpad_size_sycl(queue, jobz, uplo, n, lda); } template <> -std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t syevd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dsyevd_scratchpad_size_sycl(queue, jobz, uplo, n, lda); } template <> -std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t ldb) { @@ -2466,35 +2474,35 @@ std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::que n, lda, ldb); } template <> -std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t sygvd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t itype, oneapi::math::job jobz, - oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, - std::int64_t ldb) { + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].dsygvd_scratchpad_size_sycl(queue, itype, jobz, uplo, n, lda, ldb); } template <> -std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].ssytrd_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t sytrd_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].dsytrd_scratchpad_size_sycl(queue, uplo, n, lda); } template <> -std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, - std::int64_t lda, std::int64_t ldb) { + oneapi::math::diag diag, std::int64_t n, + std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].strtrs_scratchpad_size_sycl(queue, uplo, trans, diag, n, nrhs, lda, ldb); } template <> -std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t ldb) { @@ -2503,7 +2511,7 @@ std::int64_t trtrs_scratchpad_size(oneapi::math::device libkey, sycl::qu } template <> std::int64_t trtrs_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, + sycl::queue& queue, oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, std::int64_t lda, @@ -2512,18 +2520,16 @@ std::int64_t trtrs_scratchpad_size>(oneapi::math::device lib n, nrhs, lda, ldb); } template <> -std::int64_t trtrs_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - oneapi::math::transpose trans, - oneapi::math::diag diag, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, - std::int64_t ldb) { +std::int64_t trtrs_scratchpad_size>( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, + oneapi::math::transpose trans, oneapi::math::diag diag, std::int64_t n, std::int64_t nrhs, + std::int64_t lda, std::int64_t ldb) { return function_tables[{ libkey, queue }].ztrtrs_scratchpad_size_sycl(queue, uplo, trans, diag, n, nrhs, lda, ldb); } template <> std::int64_t ungbr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, + sycl::queue& queue, oneapi::math::generate vect, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { @@ -2532,42 +2538,43 @@ std::int64_t ungbr_scratchpad_size>(oneapi::math::device lib } template <> std::int64_t ungbr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, - oneapi::math::generate vect, std::int64_t m, - std::int64_t n, std::int64_t k, - std::int64_t lda) { + sycl::queue& queue, + oneapi::math::generate vect, + std::int64_t m, std::int64_t n, + std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].zungbr_scratchpad_size_sycl(queue, vect, m, n, k, lda); } template <> std::int64_t ungqr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].cungqr_scratchpad_size_sycl(queue, m, n, k, lda); } template <> std::int64_t ungqr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t m, + sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda) { return function_tables[{ libkey, queue }].zungqr_scratchpad_size_sycl(queue, m, n, k, lda); } template <> std::int64_t ungtr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, + sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda) { return function_tables[{ libkey, queue }].cungtr_scratchpad_size_sycl(queue, uplo, n, lda); } template <> std::int64_t ungtr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::uplo uplo, - std::int64_t n, std::int64_t lda) { + sycl::queue& queue, + oneapi::math::uplo uplo, std::int64_t n, + std::int64_t lda) { return function_tables[{ libkey, queue }].zungtr_scratchpad_size_sycl(queue, uplo, n, lda); } template <> std::int64_t unmrq_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, @@ -2576,18 +2583,16 @@ std::int64_t unmrq_scratchpad_size>(oneapi::math::device lib k, lda, ldc); } template <> -std::int64_t unmrq_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc) { +std::int64_t unmrq_scratchpad_size>( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc) { return function_tables[{ libkey, queue }].zunmrq_scratchpad_size_sycl(queue, side, trans, m, n, k, lda, ldc); } template <> std::int64_t unmqr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, + sycl::queue& queue, oneapi::math::side side, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, @@ -2596,18 +2601,16 @@ std::int64_t unmqr_scratchpad_size>(oneapi::math::device lib k, lda, ldc); } template <> -std::int64_t unmqr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, - std::int64_t ldc) { +std::int64_t unmqr_scratchpad_size>( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::transpose trans, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, + std::int64_t ldc) { return function_tables[{ libkey, queue }].zunmqr_scratchpad_size_sycl(queue, side, trans, m, n, k, lda, ldc); } template <> std::int64_t unmtr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, + sycl::queue& queue, oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, @@ -2616,17 +2619,15 @@ std::int64_t unmtr_scratchpad_size>(oneapi::math::device lib m, n, lda, ldc); } template <> -std::int64_t unmtr_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, oneapi::math::side side, - oneapi::math::uplo uplo, - oneapi::math::transpose trans, - std::int64_t m, std::int64_t n, - std::int64_t lda, std::int64_t ldc) { +std::int64_t unmtr_scratchpad_size>( + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::side side, + oneapi::math::uplo uplo, oneapi::math::transpose trans, std::int64_t m, std::int64_t n, + std::int64_t lda, std::int64_t ldc) { return function_tables[{ libkey, queue }].zunmtr_scratchpad_size_sycl(queue, side, uplo, trans, m, n, lda, ldc); } template <> -std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { @@ -2634,7 +2635,7 @@ std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, syc queue, m, n, lda, stride_a, stride_ipiv, batch_size); } template <> -std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { @@ -2643,20 +2644,20 @@ std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sy } template <> std::int64_t getrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cgetrf_batch_scratchpad_size_sycl( queue, m, n, lda, stride_a, stride_ipiv, batch_size); } template <> std::int64_t getrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zgetrf_batch_scratchpad_size_sycl( queue, m, n, lda, stride_a, stride_ipiv, batch_size); } template <> -std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { @@ -2664,7 +2665,7 @@ std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, syc queue, n, lda, stride_a, stride_ipiv, batch_size); } template <> -std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { @@ -2673,20 +2674,20 @@ std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sy } template <> std::int64_t getri_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cgetri_batch_scratchpad_size_sycl( queue, n, lda, stride_a, stride_ipiv, batch_size); } template <> std::int64_t getri_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t n, std::int64_t lda, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zgetri_batch_scratchpad_size_sycl( queue, n, lda, stride_a, stride_ipiv, batch_size); } template <> -std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, @@ -2696,7 +2697,7 @@ std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, syc queue, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } template <> -std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, @@ -2707,7 +2708,7 @@ std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sy } template <> std::int64_t getrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cgetrs_batch_scratchpad_size_sycl( @@ -2715,14 +2716,14 @@ std::int64_t getrs_batch_scratchpad_size>( } template <> std::int64_t getrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose trans, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose trans, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_ipiv, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zgetrs_batch_scratchpad_size_sycl( queue, trans, n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); } template <> -std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { @@ -2730,7 +2731,7 @@ std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, syc queue, m, n, lda, stride_a, stride_tau, batch_size); } template <> -std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { @@ -2739,20 +2740,20 @@ std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sy } template <> std::int64_t geqrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cgeqrf_batch_scratchpad_size_sycl( queue, m, n, lda, stride_a, stride_tau, batch_size); } template <> std::int64_t geqrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zgeqrf_batch_scratchpad_size_sycl( queue, m, n, lda, stride_a, stride_tau, batch_size); } template <> -std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { @@ -2760,7 +2761,7 @@ std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, syc queue, uplo, n, lda, stride_a, batch_size); } template <> -std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { @@ -2769,20 +2770,20 @@ std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sy } template <> std::int64_t potrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cpotrf_batch_scratchpad_size_sycl( queue, uplo, n, lda, stride_a, batch_size); } template <> std::int64_t potrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zpotrf_batch_scratchpad_size_sycl( queue, uplo, n, lda, stride_a, batch_size); } template <> -std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, @@ -2791,7 +2792,7 @@ std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, syc queue, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template <> -std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, @@ -2801,7 +2802,7 @@ std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sy } template <> std::int64_t potrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cpotrs_batch_scratchpad_size_sycl( @@ -2809,14 +2810,14 @@ std::int64_t potrs_batch_scratchpad_size>( } template <> std::int64_t potrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo uplo, std::int64_t n, + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo uplo, std::int64_t n, std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zpotrs_batch_scratchpad_size_sycl( queue, uplo, n, nrhs, lda, stride_a, ldb, stride_b, batch_size); } template <> -std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { @@ -2824,7 +2825,7 @@ std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, syc queue, m, n, k, lda, stride_a, stride_tau, batch_size); } template <> -std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { @@ -2833,245 +2834,245 @@ std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sy } template <> std::int64_t ungqr_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return function_tables[{ libkey, queue }].cungqr_batch_scratchpad_size_sycl( queue, m, n, k, lda, stride_a, stride_tau, batch_size); } template <> std::int64_t ungqr_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t m, std::int64_t n, std::int64_t k, + oneapi::math::device libkey, sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, std::int64_t lda, std::int64_t stride_a, std::int64_t stride_tau, std::int64_t batch_size) { return function_tables[{ libkey, queue }].zungqr_batch_scratchpad_size_sycl( queue, m, n, k, lda, stride_a, stride_tau, batch_size); } template <> -std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].sgetrf_group_scratchpad_size_sycl( queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t getrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].dgetrf_group_scratchpad_size_sycl( queue, m, n, lda, group_count, group_sizes); } template <> std::int64_t getrf_batch_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].cgetrf_group_scratchpad_size_sycl( queue, m, n, lda, group_count, group_sizes); } template <> std::int64_t getrf_batch_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].zgetrf_group_scratchpad_size_sycl( queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, std::int64_t *lda, +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].sgetri_group_scratchpad_size_sycl( queue, n, lda, group_count, group_sizes); } template <> -std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *n, std::int64_t *lda, +std::int64_t getri_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].dgetri_group_scratchpad_size_sycl( queue, n, lda, group_count, group_sizes); } template <> std::int64_t getri_batch_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, + sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].cgetri_group_scratchpad_size_sycl( queue, n, lda, group_count, group_sizes); } template <> std::int64_t getri_batch_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *n, - std::int64_t *lda, + sycl::queue& queue, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].zgetri_group_scratchpad_size_sycl( queue, n, lda, group_count, group_sizes); } template <> -std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].sgetrs_group_scratchpad_size_sycl( queue, trans, n, nrhs, lda, ldb, group_count, group_sizes); } template <> -std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t getrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::transpose* trans, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].dgetrs_group_scratchpad_size_sycl( queue, trans, n, nrhs, lda, ldb, group_count, group_sizes); } template <> std::int64_t getrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes) { + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].cgetrs_group_scratchpad_size_sycl( queue, trans, n, nrhs, lda, ldb, group_count, group_sizes); } template <> std::int64_t getrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::transpose *trans, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes) { + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::transpose* trans, + std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, + std::int64_t group_count, std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].zgetrs_group_scratchpad_size_sycl( queue, trans, n, nrhs, lda, ldb, group_count, group_sizes); } template <> -std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *lda, +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].sgeqrf_group_scratchpad_size_sycl( queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t geqrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].dgeqrf_group_scratchpad_size_sycl( queue, m, n, lda, group_count, group_sizes); } template <> std::int64_t geqrf_batch_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].cgeqrf_group_scratchpad_size_sycl( queue, m, n, lda, group_count, group_sizes); } template <> std::int64_t geqrf_batch_scratchpad_size>(oneapi::math::device libkey, - sycl::queue &queue, std::int64_t *m, - std::int64_t *n, std::int64_t *lda, + sycl::queue& queue, std::int64_t* m, + std::int64_t* n, std::int64_t* lda, std::int64_t group_count, - std::int64_t *group_sizes) { + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].zgeqrf_group_scratchpad_size_sycl( queue, m, n, lda, group_count, group_sizes); } template <> -std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].sorgqr_group_scratchpad_size_sycl( queue, m, n, k, lda, group_count, group_sizes); } template <> -std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - std::int64_t *m, std::int64_t *n, std::int64_t *k, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t orgqr_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + std::int64_t* m, std::int64_t* n, std::int64_t* k, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].dorgqr_group_scratchpad_size_sycl( queue, m, n, k, lda, group_count, group_sizes); } template <> -std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].spotrf_group_scratchpad_size_sycl( queue, uplo, n, lda, group_count, group_sizes); } template <> -std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t potrf_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].dpotrf_group_scratchpad_size_sycl( queue, uplo, n, lda, group_count, group_sizes); } template <> std::int64_t potrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].cpotrf_group_scratchpad_size_sycl( queue, uplo, n, lda, group_count, group_sizes); } template <> std::int64_t potrf_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].zpotrf_group_scratchpad_size_sycl( queue, uplo, n, lda, group_count, group_sizes); } template <> -std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].spotrs_group_scratchpad_size_sycl( queue, uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template <> -std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue &queue, - oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, - std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes) { +std::int64_t potrs_batch_scratchpad_size(oneapi::math::device libkey, sycl::queue& queue, + oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, + std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].dpotrs_group_scratchpad_size_sycl( queue, uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template <> std::int64_t potrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes) { + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].cpotrs_group_scratchpad_size_sycl( queue, uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template <> std::int64_t potrs_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, oneapi::math::uplo *uplo, std::int64_t *n, - std::int64_t *nrhs, std::int64_t *lda, std::int64_t *ldb, std::int64_t group_count, - std::int64_t *group_sizes) { + oneapi::math::device libkey, sycl::queue& queue, oneapi::math::uplo* uplo, std::int64_t* n, + std::int64_t* nrhs, std::int64_t* lda, std::int64_t* ldb, std::int64_t group_count, + std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].zpotrs_group_scratchpad_size_sycl( queue, uplo, n, nrhs, lda, ldb, group_count, group_sizes); } template <> std::int64_t ungqr_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].cungqr_group_scratchpad_size_sycl( queue, m, n, k, lda, group_count, group_sizes); } template <> std::int64_t ungqr_batch_scratchpad_size>( - oneapi::math::device libkey, sycl::queue &queue, std::int64_t *m, std::int64_t *n, - std::int64_t *k, std::int64_t *lda, std::int64_t group_count, std::int64_t *group_sizes) { + oneapi::math::device libkey, sycl::queue& queue, std::int64_t* m, std::int64_t* n, + std::int64_t* k, std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes) { return function_tables[{ libkey, queue }].zungqr_group_scratchpad_size_sycl( queue, m, n, k, lda, group_count, group_sizes); } diff --git a/src/rng/backends/curand/curand_task.hpp b/src/rng/backends/curand/curand_task.hpp index 6f9fa5613..4bd9c812a 100644 --- a/src/rng/backends/curand/curand_task.hpp +++ b/src/rng/backends/curand/curand_task.hpp @@ -15,18 +15,18 @@ namespace rng { namespace curand { #ifdef __HIPSYCL__ template -static inline void host_task_internal(H &cgh, A acc, E e, F f) { +static inline void host_task_internal(H& cgh, A acc, E e, F f) { cgh.hipSYCL_enqueue_custom_operation([=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandSetStream, status, e, ih.get_native_queue()); auto r_ptr = - reinterpret_cast(ih.get_native_mem(acc)); + reinterpret_cast(ih.get_native_mem(acc)); f(r_ptr); }); } template -static inline void host_task_internal(H &cgh, E e, F f) { +static inline void host_task_internal(H& cgh, E e, F f) { cgh.hipSYCL_enqueue_custom_operation([=](sycl::interop_handle ih) { curandStatus_t status; CURAND_CALL(curandSetStream, status, e, ih.get_native_queue()); @@ -35,16 +35,16 @@ static inline void host_task_internal(H &cgh, E e, F f) { } #else template -static inline void host_task_internal(H &cgh, A acc, E e, F f) { +static inline void host_task_internal(H& cgh, A acc, E e, F f) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - cgh.ext_codeplay_enqueue_native_command([=](sycl::interop_handle ih){ + cgh.ext_codeplay_enqueue_native_command([=](sycl::interop_handle ih) { #else cgh.host_task([=](sycl::interop_handle ih) { #endif curandStatus_t status; auto stream = ih.get_native_queue(); CURAND_CALL(curandSetStream, status, e, stream); - auto r_ptr = reinterpret_cast( + auto r_ptr = reinterpret_cast( ih.get_native_mem(acc)); f(r_ptr); #ifndef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND @@ -55,9 +55,9 @@ static inline void host_task_internal(H &cgh, A acc, E e, F f) { } template -static inline void host_task_internal(H &cgh, E e, F f) { +static inline void host_task_internal(H& cgh, E e, F f) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - cgh.ext_codeplay_enqueue_native_command([=](sycl::interop_handle ih){ + cgh.ext_codeplay_enqueue_native_command([=](sycl::interop_handle ih) { #else cgh.host_task([=](sycl::interop_handle ih) { #endif @@ -73,12 +73,12 @@ static inline void host_task_internal(H &cgh, E e, F f) { } #endif template -static inline void onemath_curand_host_task(H &cgh, A acc, E e, F f) { +static inline void onemath_curand_host_task(H& cgh, A acc, E e, F f) { host_task_internal(cgh, acc, e, f); } template -static inline void onemath_curand_host_task(H &cgh, Engine e, F f) { +static inline void onemath_curand_host_task(H& cgh, Engine e, F f) { host_task_internal(cgh, e, f); } diff --git a/src/rng/backends/curand/mrg32k3a.cpp b/src/rng/backends/curand/mrg32k3a.cpp index 4e2b6d263..14ce97adc 100644 --- a/src/rng/backends/curand/mrg32k3a.cpp +++ b/src/rng/backends/curand/mrg32k3a.cpp @@ -97,12 +97,12 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { mrg32k3a_impl(sycl::queue queue, std::initializer_list seed) : oneapi::math::rng::detail::engine_impl(queue) { throw oneapi::math::unimplemented("rng", "mrg32ka engine", - "multi-seed unsupported by cuRAND backend"); + "multi-seed unsupported by cuRAND backend"); } mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::math::rng::detail::engine_impl(*other) { throw oneapi::math::unimplemented("rng", "mrg32ka engine", - "copy construction unsupported by cuRAND backend"); + "copy construction unsupported by cuRAND backend"); } // Buffers API @@ -120,9 +120,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { range_transform_fp(queue_, distr.a(), distr.b(), n, r); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::standard>& distr, + std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { @@ -160,9 +160,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { range_transform_fp_accurate(queue_, distr.a(), distr.b(), n, r); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::accurate>& distr, + std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { @@ -249,9 +249,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { "ICDF method not used for pseudorandom generators in cuRAND backend"); } - virtual void generate( - const oneapi::math::rng::lognormal& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::icdf>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented( "rng", "mrg32ka engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); @@ -312,7 +312,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -362,7 +363,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -434,8 +436,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& - distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { @@ -457,7 +459,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& distr, + const oneapi::math::rng::lognormal& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented( "rng", "mrg32ka engine", @@ -523,7 +526,7 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { virtual void skip_ahead(std::initializer_list num_to_skip) override { throw oneapi::math::unimplemented("rng", "skip_ahead", - "initializer list unsupported by cuRAND backend"); + "initializer list unsupported by cuRAND backend"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { @@ -563,9 +566,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::standard>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } @@ -581,9 +584,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::accurate>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } @@ -629,9 +632,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate( - const oneapi::math::rng::lognormal& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::icdf>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } @@ -670,7 +673,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; @@ -692,7 +696,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; @@ -737,8 +742,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& - distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; @@ -752,7 +757,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& distr, + const oneapi::math::rng::lognormal& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; @@ -818,7 +824,7 @@ oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std:: } oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, - std::initializer_list seed) { + std::initializer_list seed) { return new mrg32k3a_impl(queue, seed); } diff --git a/src/rng/backends/curand/philox4x32x10.cpp b/src/rng/backends/curand/philox4x32x10.cpp index 375be1d6d..7f2c829ec 100644 --- a/src/rng/backends/curand/philox4x32x10.cpp +++ b/src/rng/backends/curand/philox4x32x10.cpp @@ -119,13 +119,13 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { philox4x32x10_impl(sycl::queue queue, std::initializer_list seed) : oneapi::math::rng::detail::engine_impl(queue) { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine", - "multi-seed unsupported by cuRAND backend"); + "multi-seed unsupported by cuRAND backend"); } philox4x32x10_impl(const philox4x32x10_impl* other) : oneapi::math::rng::detail::engine_impl(*other) { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine", - "copy construction unsupported by cuRAND backend"); + "copy construction unsupported by cuRAND backend"); } // Buffers API @@ -143,9 +143,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { range_transform_fp(queue_, distr.a(), distr.b(), n, r); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::standard>& distr, + std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { @@ -183,9 +183,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { range_transform_fp_accurate(queue_, distr.a(), distr.b(), n, r); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::accurate>& distr, + std::int64_t n, sycl::buffer& r) override { queue_.submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); onemath_curand_host_task(cgh, acc, engine_, [=](double* r_ptr) { @@ -272,9 +272,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { "ICDF method not used for pseudorandom generators in cuRAND backend"); } - virtual void generate( - const oneapi::math::rng::lognormal& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::icdf>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", "ICDF method not used for pseudorandom generators in cuRAND backend"); @@ -335,7 +335,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -384,7 +385,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); sycl::event generate_event = queue_.submit([&](sycl::handler& cgh) { @@ -456,8 +458,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& - distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); return queue_.submit([&](sycl::handler& cgh) { @@ -479,7 +481,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& distr, + const oneapi::math::rng::lognormal& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented( "rng", "philox4x32x10 engine", @@ -545,7 +548,7 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { virtual void skip_ahead(std::initializer_list num_to_skip) override { throw oneapi::math::unimplemented("rng", "skip_ahead", - "initializer list unsupported by cuRAND backend"); + "initializer list unsupported by cuRAND backend"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { @@ -585,9 +588,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::standard>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } @@ -603,9 +606,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::accurate>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } @@ -651,9 +654,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate( - const oneapi::math::rng::lognormal& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::icdf>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } @@ -692,7 +695,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; @@ -714,7 +718,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; @@ -759,8 +764,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& - distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; @@ -774,7 +779,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& distr, + const oneapi::math::rng::lognormal& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; @@ -835,7 +841,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { }; #endif -oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { +oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, + std::uint64_t seed) { return new philox4x32x10_impl(queue, seed); } diff --git a/src/rng/backends/mklcpu/cpu_common.hpp b/src/rng/backends/mklcpu/cpu_common.hpp index b83dfef6b..559f27960 100644 --- a/src/rng/backends/mklcpu/cpu_common.hpp +++ b/src/rng/backends/mklcpu/cpu_common.hpp @@ -34,19 +34,19 @@ namespace mklcpu { // host_task automatically uses run_on_host_intel if it is supported by the // compiler. Otherwise, it falls back to single_task. template -static inline auto host_task_internal(H &cgh, F f, int) -> decltype(cgh.host_task(f)) { +static inline auto host_task_internal(H& cgh, F f, int) -> decltype(cgh.host_task(f)) { return cgh.host_task(f); } template -static inline void host_task_internal(H &cgh, F f, long) { +static inline void host_task_internal(H& cgh, F f, long) { #ifndef __SYCL_DEVICE_ONLY__ cgh.template single_task(f); #endif } template -static inline void host_task(H &cgh, F f) { +static inline void host_task(H& cgh, F f) { (void)host_task_internal(cgh, f, 0); } @@ -57,7 +57,7 @@ template class kernel_name_usm {}; template -typename Acc::value_type *get_raw_ptr(Acc acc) { +typename Acc::value_type* get_raw_ptr(Acc acc) { // Workaround for AdaptiveCPP, as they do not yet support the get_multi_ptr function #ifndef __HIPSYCL__ return acc.template get_multi_ptr().get_raw(); diff --git a/src/rng/backends/mklcpu/mrg32k3a.cpp b/src/rng/backends/mklcpu/mrg32k3a.cpp index 89af18bd6..3bd79c400 100644 --- a/src/rng/backends/mklcpu/mrg32k3a.cpp +++ b/src/rng/backends/mklcpu/mrg32k3a.cpp @@ -576,7 +576,7 @@ oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std:: } oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, - std::initializer_list seed) { + std::initializer_list seed) { return new mrg32k3a_impl(queue, seed); } diff --git a/src/rng/backends/mklcpu/philox4x32x10.cpp b/src/rng/backends/mklcpu/philox4x32x10.cpp index 58b4c3aee..144ced995 100644 --- a/src/rng/backends/mklcpu/philox4x32x10.cpp +++ b/src/rng/backends/mklcpu/philox4x32x10.cpp @@ -573,7 +573,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { std::int32_t state_size_; }; -oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { +oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, + std::uint64_t seed) { return new philox4x32x10_impl(queue, seed); } diff --git a/src/rng/backends/mklgpu/mrg32k3a.cpp b/src/rng/backends/mklgpu/mrg32k3a.cpp index 849f78761..bb31f2a38 100644 --- a/src/rng/backends/mklgpu/mrg32k3a.cpp +++ b/src/rng/backends/mklgpu/mrg32k3a.cpp @@ -43,119 +43,138 @@ namespace mklgpu { class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { public: mrg32k3a_impl(sycl::queue queue, std::uint32_t seed) - : oneapi::math::rng::detail::engine_impl(queue), engine_((oneapi::mkl::rng::mrg32k3a(queue, seed))) { - } + : oneapi::math::rng::detail::engine_impl(queue), + engine_((oneapi::mkl::rng::mrg32k3a(queue, seed))) {} mrg32k3a_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::math::rng::detail::engine_impl(queue), engine_((oneapi::mkl::rng::mrg32k3a(queue, seed))) { - } + : oneapi::math::rng::detail::engine_impl(queue), + engine_((oneapi::mkl::rng::mrg32k3a(queue, seed))) {} - mrg32k3a_impl(const mrg32k3a_impl* other) : oneapi::math::rng::detail::engine_impl(*other), engine_((oneapi::mkl::rng::mrg32k3a(other->engine_))) { - } + mrg32k3a_impl(const mrg32k3a_impl* other) + : oneapi::math::rng::detail::engine_impl(*other), + engine_((oneapi::mkl::rng::mrg32k3a(other->engine_))) {} // Buffers API virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::standard>& distr, + std::int64_t n, sycl::buffer& r) override { + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::uniform< std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::accurate>& distr, + std::int64_t n, sycl::buffer& r) override { + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::gaussian< float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::gaussian< double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::lognormal< float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::lognormal< double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } - virtual void generate( - const oneapi::math::rng::lognormal& distr, - std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::icdf>& distr, + std::int64_t n, sycl::buffer& r) override { + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bits& /*distr*/, std::int64_t /*n*/, sycl::buffer& /*r*/) override { - throw unimplemented("rng/mklgpu", "mrg32k3a::generate", "bits distribution is not supported"); + throw unimplemented("rng/mklgpu", "mrg32k3a::generate", + "bits distribution is not supported"); } // USM APIs @@ -163,13 +182,16 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( @@ -177,100 +199,119 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { ; - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( - const oneapi::math::rng::lognormal& - distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( - const oneapi::math::rng::lognormal& distr, + const oneapi::math::rng::lognormal& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } - virtual sycl::event generate(const bits& /*distr*/, std::int64_t /*n*/, std::uint32_t* /*r*/, + virtual sycl::event generate(const bits& /*distr*/, std::int64_t /*n*/, + std::uint32_t* /*r*/, const std::vector& /*dependencies*/) override { - throw unimplemented("rng/mklgpu", "mrg32k3a::generate", "bits distribution is not supported"); + throw unimplemented("rng/mklgpu", "mrg32k3a::generate", + "bits distribution is not supported"); return {}; } @@ -301,7 +342,7 @@ oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std:: } oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, - std::initializer_list seed) { + std::initializer_list seed) { return new mrg32k3a_impl(queue, seed); } diff --git a/src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp b/src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp index d1d16e293..67447110f 100644 --- a/src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp +++ b/src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp @@ -80,42 +80,44 @@ struct convert_distrib_t; template struct convert_distrib_t> { -auto operator()(uniform distribution) { - using onemkl_method_t = typename convert_method_t::type; - return oneapi::mkl::rng::uniform(distribution.a(), distribution.b()); -} + auto operator()(uniform distribution) { + using onemkl_method_t = typename convert_method_t::type; + return oneapi::mkl::rng::uniform(distribution.a(), distribution.b()); + } }; template struct convert_distrib_t> { -auto operator()(gaussian distribution) { - using onemkl_method_t = typename convert_method_t::type; - return oneapi::mkl::rng::gaussian(distribution.mean(), distribution.stddev()); -} + auto operator()(gaussian distribution) { + using onemkl_method_t = typename convert_method_t::type; + return oneapi::mkl::rng::gaussian(distribution.mean(), + distribution.stddev()); + } }; template struct convert_distrib_t> { -auto operator()(lognormal distribution) { - using onemkl_method_t = typename convert_method_t::type; - return oneapi::mkl::rng::lognormal(distribution.m(), distribution.s(), distribution.displ(), distribution.scale()); -} + auto operator()(lognormal distribution) { + using onemkl_method_t = typename convert_method_t::type; + return oneapi::mkl::rng::lognormal( + distribution.m(), distribution.s(), distribution.displ(), distribution.scale()); + } }; template struct convert_distrib_t> { -auto operator()(bernoulli distribution) { - using onemkl_method_t = typename convert_method_t::type; - return oneapi::mkl::rng::bernoulli(distribution.p()); -} + auto operator()(bernoulli distribution) { + using onemkl_method_t = typename convert_method_t::type; + return oneapi::mkl::rng::bernoulli(distribution.p()); + } }; template struct convert_distrib_t> { -auto operator()(poisson distribution) { - using onemkl_method_t = typename convert_method_t::type; - return oneapi::mkl::rng::poisson(distribution.lambda()); -} + auto operator()(poisson distribution) { + using onemkl_method_t = typename convert_method_t::type; + return oneapi::mkl::rng::poisson(distribution.lambda()); + } }; template @@ -123,9 +125,9 @@ inline auto get_onemkl_distribution(DistributionT distribution) { return convert_distrib_t()(distribution); } -} // namespace detail -} // namespace rng -} // namespace math -} // namespace oneapi +} // namespace detail +} // namespace rng +} // namespace math +} // namespace oneapi #endif // _ONEMATH_SRC_RNG_ONEMKL_DISTRIBUTION_CONVERSION_HPP_ diff --git a/src/rng/backends/mklgpu/philox4x32x10.cpp b/src/rng/backends/mklgpu/philox4x32x10.cpp index 36fa9e621..87cdb42fd 100644 --- a/src/rng/backends/mklgpu/philox4x32x10.cpp +++ b/src/rng/backends/mklgpu/philox4x32x10.cpp @@ -43,120 +43,138 @@ namespace mklgpu { class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { public: philox4x32x10_impl(sycl::queue queue, std::uint64_t seed) - : oneapi::math::rng::detail::engine_impl(queue), engine_((oneapi::mkl::rng::philox4x32x10(queue, seed))) { - } + : oneapi::math::rng::detail::engine_impl(queue), + engine_((oneapi::mkl::rng::philox4x32x10(queue, seed))) {} philox4x32x10_impl(sycl::queue queue, std::initializer_list seed) - : oneapi::math::rng::detail::engine_impl(queue), engine_((oneapi::mkl::rng::philox4x32x10(queue, seed))) { - } + : oneapi::math::rng::detail::engine_impl(queue), + engine_((oneapi::mkl::rng::philox4x32x10(queue, seed))) {} philox4x32x10_impl(const philox4x32x10_impl* other) - : oneapi::math::rng::detail::engine_impl(*other), engine_((oneapi::mkl::rng::philox4x32x10(other->engine_))) { - } + : oneapi::math::rng::detail::engine_impl(*other), + engine_((oneapi::mkl::rng::philox4x32x10(other->engine_))) {} // Buffers API virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::standard>& distr, + std::int64_t n, sycl::buffer& r) override { + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::uniform< std::int32_t, oneapi::math::rng::uniform_method::standard>& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::uniform& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::accurate>& distr, + std::int64_t n, sycl::buffer& r) override { + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::gaussian< float, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::gaussian< double, oneapi::math::rng::gaussian_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::lognormal< float, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const oneapi::math::rng::lognormal< double, oneapi::math::rng::lognormal_method::box_muller2>& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } - virtual void generate( - const oneapi::math::rng::lognormal& distr, - std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::icdf>& distr, + std::int64_t n, sycl::buffer& r) override { + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bernoulli& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const poisson& distr, std::int64_t n, sycl::buffer& r) override { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r)); } virtual void generate(const bits& /*distr*/, std::int64_t /*n*/, sycl::buffer& /*r*/) override { - throw unimplemented("rng/mklgpu", "philox4x32x10::generate", "bits distribution is not supported"); + throw unimplemented("rng/mklgpu", "philox4x32x10::generate", + "bits distribution is not supported"); } // USM APIs @@ -164,13 +182,16 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( @@ -178,100 +199,119 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { ; - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::uniform& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::gaussian& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( - const oneapi::math::rng::lognormal& - distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const oneapi::math::rng::lognormal& distr, std::int64_t n, float* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( - const oneapi::math::rng::lognormal& distr, + const oneapi::math::rng::lognormal& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate(const bernoulli& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::int32_t* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } virtual sycl::event generate( const poisson& distr, std::int64_t n, std::uint32_t* r, const std::vector& dependencies) override { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate(detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::rng::generate( + detail::get_onemkl_distribution(distr), engine_, n, r, dependencies)); } - virtual sycl::event generate(const bits& /*distr*/, std::int64_t /*n*/, std::uint32_t* /*r*/, + virtual sycl::event generate(const bits& /*distr*/, std::int64_t /*n*/, + std::uint32_t* /*r*/, const std::vector& /*dependencies*/) override { - throw unimplemented("rng/mklgpu", "philox4x32x10::generate", "bits distribution is not supported"); + throw unimplemented("rng/mklgpu", "philox4x32x10::generate", + "bits distribution is not supported"); return {}; } @@ -297,7 +337,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { oneapi::mkl::rng::philox4x32x10 engine_; }; -oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { +oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, + std::uint64_t seed) { return new philox4x32x10_impl(queue, seed); } diff --git a/src/rng/backends/rocrand/mrg32k3a.cpp b/src/rng/backends/rocrand/mrg32k3a.cpp index 20d62b33f..741c45432 100644 --- a/src/rng/backends/rocrand/mrg32k3a.cpp +++ b/src/rng/backends/rocrand/mrg32k3a.cpp @@ -99,7 +99,7 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { mrg32k3a_impl(sycl::queue queue, std::initializer_list seed) : oneapi::math::rng::detail::engine_impl(queue) { throw oneapi::math::unimplemented("rng", "mrg32ka engine", - "multi-seed unsupported by rocRAND backend"); + "multi-seed unsupported by rocRAND backend"); } mrg32k3a_impl(const mrg32k3a_impl* other) @@ -134,9 +134,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { range_transform_fp(queue_, distr.a(), distr.b(), n, r); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::standard>& distr, + std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -189,9 +189,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { range_transform_fp_accurate(queue_, distr.a(), distr.b(), n, r); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::accurate>& distr, + std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -326,9 +326,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { increment_internal_offset(n); } - virtual void generate( - const oneapi::math::rng::lognormal& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::icdf>& distr, + std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -425,7 +425,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -481,7 +482,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -587,8 +589,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& - distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -622,7 +624,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& distr, + const oneapi::math::rng::lognormal& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -716,7 +719,7 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { virtual void skip_ahead(std::initializer_list num_to_skip) override { throw oneapi::math::unimplemented("rng", "skip_ahead", - "initializer list unsupported by rocRAND backend"); + "initializer list unsupported by rocRAND backend"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { @@ -761,9 +764,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::standard>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } @@ -779,9 +782,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::accurate>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } @@ -827,9 +830,9 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } - virtual void generate( - const oneapi::math::rng::lognormal& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::icdf>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); } @@ -868,7 +871,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; @@ -890,7 +894,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; @@ -935,8 +940,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& - distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; @@ -950,7 +955,8 @@ class mrg32k3a_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& distr, + const oneapi::math::rng::lognormal& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "mrg32ka engine"); return sycl::event{}; @@ -1016,7 +1022,7 @@ oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, std:: } oneapi::math::rng::detail::engine_impl* create_mrg32k3a(sycl::queue queue, - std::initializer_list seed) { + std::initializer_list seed) { return new mrg32k3a_impl(queue, seed); } diff --git a/src/rng/backends/rocrand/philox4x32x10.cpp b/src/rng/backends/rocrand/philox4x32x10.cpp index 6753b9c1b..682aac11e 100644 --- a/src/rng/backends/rocrand/philox4x32x10.cpp +++ b/src/rng/backends/rocrand/philox4x32x10.cpp @@ -121,7 +121,7 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { philox4x32x10_impl(sycl::queue queue, std::initializer_list seed) : oneapi::math::rng::detail::engine_impl(queue) { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine", - "multi-seed unsupported by rocRAND backend"); + "multi-seed unsupported by rocRAND backend"); } philox4x32x10_impl(const philox4x32x10_impl* other) @@ -156,9 +156,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { range_transform_fp(queue_, distr.a(), distr.b(), n, r); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::standard>& distr, + std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -211,9 +211,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { range_transform_fp_accurate(queue_, distr.a(), distr.b(), n, r); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::accurate>& distr, + std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -348,9 +348,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { increment_internal_offset(n); } - virtual void generate( - const oneapi::math::rng::lognormal& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::icdf>& distr, + std::int64_t n, sycl::buffer& r) override { queue_ .submit([&](sycl::handler& cgh) { auto acc = r.get_access(cgh); @@ -447,7 +447,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -503,7 +504,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); queue_ @@ -609,8 +611,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& - distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -644,7 +646,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& distr, + const oneapi::math::rng::lognormal& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { sycl::event::wait_and_throw(dependencies); auto event = queue_.submit([&](sycl::handler& cgh) { @@ -737,7 +740,7 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { virtual void skip_ahead(std::initializer_list num_to_skip) override { throw oneapi::math::unimplemented("rng", "skip_ahead", - "initializer list unsupported by rocRAND backend"); + "initializer list unsupported by rocRAND backend"); } virtual void leapfrog(std::uint64_t idx, std::uint64_t stride) override { @@ -783,9 +786,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::standard>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } @@ -801,9 +804,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate( - const oneapi::math::rng::uniform& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::uniform< + double, oneapi::math::rng::uniform_method::accurate>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } @@ -849,9 +852,9 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } - virtual void generate( - const oneapi::math::rng::lognormal& distr, - std::int64_t n, sycl::buffer& r) override { + virtual void generate(const oneapi::math::rng::lognormal< + double, oneapi::math::rng::lognormal_method::icdf>& distr, + std::int64_t n, sycl::buffer& r) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); } @@ -890,7 +893,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; @@ -912,7 +916,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::uniform& distr, + const oneapi::math::rng::uniform& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; @@ -957,8 +962,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& - distr, + const oneapi::math::rng::lognormal& distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; @@ -972,7 +977,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { } virtual sycl::event generate( - const oneapi::math::rng::lognormal& distr, + const oneapi::math::rng::lognormal& + distr, std::int64_t n, double* r, const std::vector& dependencies) override { throw oneapi::math::unimplemented("rng", "philox4x32x10 engine"); return sycl::event{}; @@ -1033,7 +1039,8 @@ class philox4x32x10_impl : public oneapi::math::rng::detail::engine_impl { }; #endif -oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, std::uint64_t seed) { +oneapi::math::rng::detail::engine_impl* create_philox4x32x10(sycl::queue queue, + std::uint64_t seed) { return new philox4x32x10_impl(queue, seed); } diff --git a/src/rng/backends/rocrand/rocrand_task.hpp b/src/rng/backends/rocrand/rocrand_task.hpp index 149c0a70d..d7ece6c45 100644 --- a/src/rng/backends/rocrand/rocrand_task.hpp +++ b/src/rng/backends/rocrand/rocrand_task.hpp @@ -15,18 +15,18 @@ namespace rng { namespace rocrand { #ifdef __HIPSYCL__ template -static inline void host_task_internal(H &cgh, A acc, E e, F f) { +static inline void host_task_internal(H& cgh, A acc, E e, F f) { cgh.hipSYCL_enqueue_custom_operation([=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_set_stream, status, e, ih.get_native_queue()); auto r_ptr = - reinterpret_cast(ih.get_native_mem(acc)); + reinterpret_cast(ih.get_native_mem(acc)); f(r_ptr); }); } template -static inline void host_task_internal(H &cgh, E e, F f) { +static inline void host_task_internal(H& cgh, E e, F f) { cgh.hipSYCL_enqueue_custom_operation([=](sycl::interop_handle ih) { rocrand_status status; ROCRAND_CALL(rocrand_set_stream, status, e, ih.get_native_queue()); @@ -35,16 +35,16 @@ static inline void host_task_internal(H &cgh, E e, F f) { } #else template -static inline void host_task_internal(H &cgh, A acc, E e, F f) { +static inline void host_task_internal(H& cgh, A acc, E e, F f) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - cgh.ext_codeplay_enqueue_native_command([=](sycl::interop_handle ih){ + cgh.ext_codeplay_enqueue_native_command([=](sycl::interop_handle ih) { #else cgh.host_task([=](sycl::interop_handle ih) { #endif rocrand_status status; auto stream = ih.get_native_queue(); ROCRAND_CALL(rocrand_set_stream, status, e, stream); - auto r_ptr = reinterpret_cast( + auto r_ptr = reinterpret_cast( ih.get_native_mem(acc)); f(r_ptr); #ifndef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND @@ -55,9 +55,9 @@ static inline void host_task_internal(H &cgh, A acc, E e, F f) { } template -static inline void host_task_internal(H &cgh, E e, F f) { +static inline void host_task_internal(H& cgh, E e, F f) { #ifdef SYCL_EXT_ONEAPI_ENQUEUE_NATIVE_COMMAND - cgh.ext_codeplay_enqueue_native_command([=](sycl::interop_handle ih){ + cgh.ext_codeplay_enqueue_native_command([=](sycl::interop_handle ih) { #else cgh.host_task([=](sycl::interop_handle ih) { #endif @@ -73,12 +73,12 @@ static inline void host_task_internal(H &cgh, E e, F f) { } #endif template -static inline void onemath_rocrand_host_task(H &cgh, A acc, E e, F f) { +static inline void onemath_rocrand_host_task(H& cgh, A acc, E e, F f) { host_task_internal(cgh, acc, e, f); } template -static inline void onemath_rocrand_host_task(H &cgh, Engine e, F f) { +static inline void onemath_rocrand_host_task(H& cgh, Engine e, F f) { host_task_internal(cgh, e, f); } diff --git a/src/rng/function_table.hpp b/src/rng/function_table.hpp index da3ca1b6f..06c9c83e1 100644 --- a/src/rng/function_table.hpp +++ b/src/rng/function_table.hpp @@ -33,12 +33,12 @@ typedef struct { int version; oneapi::math::rng::detail::engine_impl* (*create_philox4x32x10_sycl)(sycl::queue queue, - std::uint64_t seed); + std::uint64_t seed); oneapi::math::rng::detail::engine_impl* (*create_philox4x32x10_ex_sycl)( sycl::queue queue, std::initializer_list seed); oneapi::math::rng::detail::engine_impl* (*create_mrg32k3a_sycl)(sycl::queue queue, - std::uint32_t seed); + std::uint32_t seed); oneapi::math::rng::detail::engine_impl* (*create_mrg32k3a_ex_sycl)( sycl::queue queue, std::initializer_list seed); } rng_function_table_t; diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.cxx b/src/sparse_blas/backends/mkl_common/mkl_handles.cxx index 509b5a78a..bd39bba4c 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.cxx @@ -19,22 +19,22 @@ // Dense vector template -void init_dense_vector(sycl::queue & /*queue*/, - oneapi::math::sparse::dense_vector_handle_t *p_dvhandle, std::int64_t size, +void init_dense_vector(sycl::queue& /*queue*/, + oneapi::math::sparse::dense_vector_handle_t* p_dvhandle, std::int64_t size, sycl::buffer val) { *p_dvhandle = new oneapi::math::sparse::dense_vector_handle(val, size); } template -void init_dense_vector(sycl::queue & /*queue*/, - oneapi::math::sparse::dense_vector_handle_t *p_dvhandle, std::int64_t size, - fpType *val) { +void init_dense_vector(sycl::queue& /*queue*/, + oneapi::math::sparse::dense_vector_handle_t* p_dvhandle, std::int64_t size, + fpType* val) { *p_dvhandle = new oneapi::math::sparse::dense_vector_handle(val, size); } template -void check_can_reset_value_handle(const std::string &function_name, - InternalHandleT *internal_handle, bool expect_buffer) { +void check_can_reset_value_handle(const std::string& function_name, + InternalHandleT* internal_handle, bool expect_buffer) { if (internal_handle->get_value_type() != detail::get_data_type()) { throw oneapi::math::invalid_argument( "sparse_blas", function_name, @@ -49,7 +49,7 @@ void check_can_reset_value_handle(const std::string &function_name, } template -void set_dense_vector_data(sycl::queue & /*queue*/, +void set_dense_vector_data(sycl::queue& /*queue*/, oneapi::math::sparse::dense_vector_handle_t dvhandle, std::int64_t size, sycl::buffer val) { check_can_reset_value_handle(__func__, dvhandle, true); @@ -58,40 +58,40 @@ void set_dense_vector_data(sycl::queue & /*queue*/, } template -void set_dense_vector_data(sycl::queue & /*queue*/, +void set_dense_vector_data(sycl::queue& /*queue*/, oneapi::math::sparse::dense_vector_handle_t dvhandle, std::int64_t size, - fpType *val) { + fpType* val) { check_can_reset_value_handle(__func__, dvhandle, false); dvhandle->size = size; dvhandle->set_usm_ptr(val); } -#define INSTANTIATE_DENSE_VECTOR_FUNCS(FP_TYPE, FP_SUFFIX) \ - template void init_dense_vector( \ +#define INSTANTIATE_DENSE_VECTOR_FUNCS(FP_TYPE, FP_SUFFIX) \ + template void init_dense_vector( \ sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t * p_dvhandle, \ - std::int64_t size, sycl::buffer val); \ - template void init_dense_vector( \ + std::int64_t size, sycl::buffer val); \ + template void init_dense_vector( \ sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t * p_dvhandle, \ - std::int64_t size, FP_TYPE * val); \ - template void set_dense_vector_data( \ + std::int64_t size, FP_TYPE* val); \ + template void set_dense_vector_data( \ sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t dvhandle, \ - std::int64_t size, sycl::buffer val); \ - template void set_dense_vector_data( \ + std::int64_t size, sycl::buffer val); \ + template void set_dense_vector_data( \ sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t dvhandle, \ - std::int64_t size, FP_TYPE * val) + std::int64_t size, FP_TYPE* val) FOR_EACH_FP_TYPE(INSTANTIATE_DENSE_VECTOR_FUNCS); #undef INSTANTIATE_DENSE_VECTOR_FUNCS -sycl::event release_dense_vector(sycl::queue &queue, +sycl::event release_dense_vector(sycl::queue& queue, oneapi::math::sparse::dense_vector_handle_t dvhandle, - const std::vector &dependencies) { + const std::vector& dependencies) { return detail::submit_release(queue, dvhandle, dependencies); } // Dense matrix template -void init_dense_matrix(sycl::queue & /*queue*/, - oneapi::math::sparse::dense_matrix_handle_t *p_dmhandle, +void init_dense_matrix(sycl::queue& /*queue*/, + oneapi::math::sparse::dense_matrix_handle_t* p_dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, oneapi::math::layout dense_layout, sycl::buffer val) { *p_dmhandle = @@ -99,16 +99,16 @@ void init_dense_matrix(sycl::queue & /*queue*/, } template -void init_dense_matrix(sycl::queue & /*queue*/, - oneapi::math::sparse::dense_matrix_handle_t *p_dmhandle, +void init_dense_matrix(sycl::queue& /*queue*/, + oneapi::math::sparse::dense_matrix_handle_t* p_dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, - oneapi::math::layout dense_layout, fpType *val) { + oneapi::math::layout dense_layout, fpType* val) { *p_dmhandle = new oneapi::math::sparse::dense_matrix_handle(val, num_rows, num_cols, ld, dense_layout); } template -void set_dense_matrix_data(sycl::queue & /*queue*/, +void set_dense_matrix_data(sycl::queue& /*queue*/, oneapi::math::sparse::dense_matrix_handle_t dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, oneapi::math::layout dense_layout, sycl::buffer val) { @@ -121,10 +121,10 @@ void set_dense_matrix_data(sycl::queue & /*queue*/, } template -void set_dense_matrix_data(sycl::queue & /*queue*/, +void set_dense_matrix_data(sycl::queue& /*queue*/, oneapi::math::sparse::dense_matrix_handle_t dmhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, - oneapi::math::layout dense_layout, fpType *val) { + oneapi::math::layout dense_layout, fpType* val) { check_can_reset_value_handle(__func__, dmhandle, false); dmhandle->num_rows = num_rows; dmhandle->num_cols = num_cols; @@ -133,35 +133,35 @@ void set_dense_matrix_data(sycl::queue & /*queue*/, dmhandle->set_usm_ptr(val); } -#define INSTANTIATE_DENSE_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX) \ - template void init_dense_matrix( \ +#define INSTANTIATE_DENSE_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX) \ + template void init_dense_matrix( \ sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t * p_dmhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ oneapi::math::layout dense_layout, sycl::buffer val); \ - template void init_dense_matrix( \ + template void init_dense_matrix( \ sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t * p_dmhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::math::layout dense_layout, FP_TYPE * val); \ - template void set_dense_matrix_data( \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ + oneapi::math::layout dense_layout, FP_TYPE* val); \ + template void set_dense_matrix_data( \ sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t dmhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ oneapi::math::layout dense_layout, sycl::buffer val); \ - template void set_dense_matrix_data( \ + template void set_dense_matrix_data( \ sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t dmhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::math::layout dense_layout, FP_TYPE * val) + std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ + oneapi::math::layout dense_layout, FP_TYPE* val) FOR_EACH_FP_TYPE(INSTANTIATE_DENSE_MATRIX_FUNCS); #undef INSTANTIATE_DENSE_MATRIX_FUNCS -sycl::event release_dense_matrix(sycl::queue &queue, +sycl::event release_dense_matrix(sycl::queue& queue, oneapi::math::sparse::dense_matrix_handle_t dmhandle, - const std::vector &dependencies) { + const std::vector& dependencies) { return detail::submit_release(queue, dmhandle, dependencies); } // COO matrix template -void init_coo_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t *p_smhandle, +void init_coo_matrix(sycl::queue& queue, oneapi::math::sparse::matrix_handle_t* p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, sycl::buffer row_ind, sycl::buffer col_ind, sycl::buffer val) { @@ -169,32 +169,35 @@ void init_coo_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t * RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::init_matrix_handle(&mkl_handle)); auto internal_smhandle = new detail::sparse_matrix_handle(mkl_handle, row_ind, col_ind, val); // The backend handle must use the buffers from the internal handle as they will be kept alive until the handle is released. - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_coo_data(queue, mkl_handle, static_cast(num_rows), - static_cast(num_cols), static_cast(nnz), - detail::get_onemkl_index_base(index), internal_smhandle->row_container.get_buffer(), - internal_smhandle->col_container.get_buffer(), - internal_smhandle->value_container.get_buffer())); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_coo_data( + queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), + static_cast(nnz), detail::get_onemkl_index_base(index), + internal_smhandle->row_container.get_buffer(), + internal_smhandle->col_container.get_buffer(), + internal_smhandle->value_container.get_buffer())); *p_smhandle = reinterpret_cast(internal_smhandle); } template -void init_coo_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t *p_smhandle, +void init_coo_matrix(sycl::queue& queue, oneapi::math::sparse::matrix_handle_t* p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, - oneapi::math::index_base index, intType *row_ind, intType *col_ind, - fpType *val) { + oneapi::math::index_base index, intType* row_ind, intType* col_ind, + fpType* val) { oneapi::mkl::sparse::matrix_handle_t mkl_handle; RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::init_matrix_handle(&mkl_handle)); auto internal_smhandle = new detail::sparse_matrix_handle(mkl_handle, row_ind, col_ind, val); sycl::event event; - RETHROW_ONEMKL_EXCEPTIONS(event = oneapi::mkl::sparse::set_coo_data(queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), - static_cast(nnz), detail::get_onemkl_index_base(index), row_ind, col_ind, val)); + RETHROW_ONEMKL_EXCEPTIONS(event = oneapi::mkl::sparse::set_coo_data( + queue, mkl_handle, static_cast(num_rows), + static_cast(num_cols), static_cast(nnz), + detail::get_onemkl_index_base(index), row_ind, col_ind, val)); event.wait_and_throw(); *p_smhandle = reinterpret_cast(internal_smhandle); } template -void check_can_reset_sparse_handle(const std::string &function_name, - detail::sparse_matrix_handle *internal_smhandle, +void check_can_reset_sparse_handle(const std::string& function_name, + detail::sparse_matrix_handle* internal_smhandle, bool expect_buffer) { check_can_reset_value_handle(function_name, internal_smhandle, expect_buffer); if (internal_smhandle->get_int_type() != detail::get_data_type()) { @@ -212,7 +215,7 @@ void check_can_reset_sparse_handle(const std::string &function_name, } template -void set_coo_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, +void set_coo_matrix_data(sycl::queue& queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, sycl::buffer row_ind, sycl::buffer col_ind, sycl::buffer val) { @@ -222,19 +225,20 @@ void set_coo_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle internal_smhandle->col_container.set_buffer(col_ind); internal_smhandle->value_container.set_buffer(val); // The backend handle must use the buffers from the internal handle as they will be kept alive until the handle is released. - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_coo_data(queue, internal_smhandle->backend_handle, - static_cast(num_rows), - static_cast(num_cols), static_cast(nnz), - detail::get_onemkl_index_base(index), internal_smhandle->row_container.get_buffer(), - internal_smhandle->col_container.get_buffer(), - internal_smhandle->value_container.get_buffer())); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_coo_data( + queue, internal_smhandle->backend_handle, static_cast(num_rows), + static_cast(num_cols), static_cast(nnz), + detail::get_onemkl_index_base(index), + internal_smhandle->row_container.get_buffer(), + internal_smhandle->col_container.get_buffer(), + internal_smhandle->value_container.get_buffer())); } template -void set_coo_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, +void set_coo_matrix_data(sycl::queue& queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, - oneapi::math::index_base index, intType *row_ind, intType *col_ind, - fpType *val) { + oneapi::math::index_base index, intType* row_ind, intType* col_ind, + fpType* val) { auto internal_smhandle = detail::get_internal_handle(smhandle); check_can_reset_sparse_handle(__func__, internal_smhandle, false); internal_smhandle->row_container.set_usm_ptr(row_ind); @@ -242,35 +246,36 @@ void set_coo_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle internal_smhandle->value_container.set_usm_ptr(val); auto event = oneapi::mkl::sparse::set_coo_data( queue, internal_smhandle->backend_handle, static_cast(num_rows), - static_cast(num_cols), static_cast(nnz), detail::get_onemkl_index_base(index), row_ind, col_ind, val); + static_cast(num_cols), static_cast(nnz), + detail::get_onemkl_index_base(index), row_ind, col_ind, val); event.wait_and_throw(); } -#define INSTANTIATE_COO_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ - template void init_coo_matrix( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::math::index_base index, sycl::buffer row_ind, \ - sycl::buffer col_ind, sycl::buffer val); \ - template void init_coo_matrix( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::math::index_base index, INT_TYPE * row_ind, INT_TYPE * col_ind, FP_TYPE * val); \ - template void set_coo_matrix_data( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ - sycl::buffer row_ind, sycl::buffer col_ind, \ - sycl::buffer val); \ - template void set_coo_matrix_data( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ - INT_TYPE * row_ind, INT_TYPE * col_ind, FP_TYPE * val) +#define INSTANTIATE_COO_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ + template void init_coo_matrix( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, sycl::buffer row_ind, \ + sycl::buffer col_ind, sycl::buffer val); \ + template void init_coo_matrix( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, INT_TYPE* row_ind, INT_TYPE* col_ind, FP_TYPE* val); \ + template void set_coo_matrix_data( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, sycl::buffer row_ind, \ + sycl::buffer col_ind, sycl::buffer val); \ + template void set_coo_matrix_data( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, INT_TYPE* row_ind, INT_TYPE* col_ind, FP_TYPE* val) FOR_EACH_FP_AND_INT_TYPE(INSTANTIATE_COO_MATRIX_FUNCS); #undef INSTANTIATE_COO_MATRIX_FUNCS // CSR matrix template -void init_csr_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t *p_smhandle, +void init_csr_matrix(sycl::queue& queue, oneapi::math::sparse::matrix_handle_t* p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t /*nnz*/, oneapi::math::index_base index, sycl::buffer row_ptr, sycl::buffer col_ind, sycl::buffer val) { @@ -279,32 +284,35 @@ void init_csr_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t * auto internal_smhandle = new detail::sparse_matrix_handle(mkl_handle, row_ptr, col_ind, val); // The backend deduces nnz from row_ptr. // The backend handle must use the buffers from the internal handle as they will be kept alive until the handle is released. - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_csr_data(queue, mkl_handle, static_cast(num_rows), - static_cast(num_cols), detail::get_onemkl_index_base(index), - internal_smhandle->row_container.get_buffer(), - internal_smhandle->col_container.get_buffer(), - internal_smhandle->value_container.get_buffer())); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_csr_data( + queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), + detail::get_onemkl_index_base(index), + internal_smhandle->row_container.get_buffer(), + internal_smhandle->col_container.get_buffer(), + internal_smhandle->value_container.get_buffer())); *p_smhandle = reinterpret_cast(internal_smhandle); } template -void init_csr_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t *p_smhandle, +void init_csr_matrix(sycl::queue& queue, oneapi::math::sparse::matrix_handle_t* p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t /*nnz*/, - oneapi::math::index_base index, intType *row_ptr, intType *col_ind, - fpType *val) { + oneapi::math::index_base index, intType* row_ptr, intType* col_ind, + fpType* val) { oneapi::mkl::sparse::matrix_handle_t mkl_handle; RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::init_matrix_handle(&mkl_handle)); auto internal_smhandle = new detail::sparse_matrix_handle(mkl_handle, row_ptr, col_ind, val); // The backend deduces nnz from row_ptr. sycl::event event; - RETHROW_ONEMKL_EXCEPTIONS(event = oneapi::mkl::sparse::set_csr_data(queue, mkl_handle, static_cast(num_rows), static_cast(num_cols), detail::get_onemkl_index_base(index), - row_ptr, col_ind, val)); + RETHROW_ONEMKL_EXCEPTIONS(event = oneapi::mkl::sparse::set_csr_data( + queue, mkl_handle, static_cast(num_rows), + static_cast(num_cols), + detail::get_onemkl_index_base(index), row_ptr, col_ind, val)); event.wait_and_throw(); *p_smhandle = reinterpret_cast(internal_smhandle); } template -void set_csr_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, +void set_csr_matrix_data(sycl::queue& queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t /*nnz*/, oneapi::math::index_base index, sycl::buffer row_ptr, sycl::buffer col_ind, sycl::buffer val) { @@ -315,19 +323,19 @@ void set_csr_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle internal_smhandle->value_container.set_buffer(val); // The backend deduces nnz from row_ptr. // The backend handle must use the buffers from the internal handle as they will be kept alive until the handle is released. - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_csr_data(queue, internal_smhandle->backend_handle, - static_cast(num_rows), - static_cast(num_cols), detail::get_onemkl_index_base(index), - internal_smhandle->row_container.get_buffer(), - internal_smhandle->col_container.get_buffer(), - internal_smhandle->value_container.get_buffer())); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_csr_data( + queue, internal_smhandle->backend_handle, static_cast(num_rows), + static_cast(num_cols), detail::get_onemkl_index_base(index), + internal_smhandle->row_container.get_buffer(), + internal_smhandle->col_container.get_buffer(), + internal_smhandle->value_container.get_buffer())); } template -void set_csr_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, +void set_csr_matrix_data(sycl::queue& queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t /*nnz*/, - oneapi::math::index_base index, intType *row_ptr, intType *col_ind, - fpType *val) { + oneapi::math::index_base index, intType* row_ptr, intType* col_ind, + fpType* val) { auto internal_smhandle = detail::get_internal_handle(smhandle); check_can_reset_sparse_handle(__func__, internal_smhandle, false); internal_smhandle->row_container.set_usm_ptr(row_ptr); @@ -336,35 +344,37 @@ void set_csr_matrix_data(sycl::queue &queue, oneapi::math::sparse::matrix_handle // The backend deduces nnz from row_ptr. auto event = oneapi::mkl::sparse::set_csr_data( queue, internal_smhandle->backend_handle, static_cast(num_rows), - static_cast(num_cols), detail::get_onemkl_index_base(index), row_ptr, col_ind, val); + static_cast(num_cols), detail::get_onemkl_index_base(index), row_ptr, col_ind, + val); event.wait_and_throw(); } -#define INSTANTIATE_CSR_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ - template void init_csr_matrix( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::math::index_base index, sycl::buffer row_ptr, \ - sycl::buffer col_ind, sycl::buffer val); \ - template void init_csr_matrix( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::math::index_base index, INT_TYPE * row_ptr, INT_TYPE * col_ind, FP_TYPE * val); \ - template void set_csr_matrix_data( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ - sycl::buffer row_ptr, sycl::buffer col_ind, \ - sycl::buffer val); \ - template void set_csr_matrix_data( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ - INT_TYPE * row_ptr, INT_TYPE * col_ind, FP_TYPE * val) +#define INSTANTIATE_CSR_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ + template void init_csr_matrix( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, sycl::buffer row_ptr, \ + sycl::buffer col_ind, sycl::buffer val); \ + template void init_csr_matrix( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, INT_TYPE* row_ptr, INT_TYPE* col_ind, FP_TYPE* val); \ + template void set_csr_matrix_data( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, sycl::buffer row_ptr, \ + sycl::buffer col_ind, sycl::buffer val); \ + template void set_csr_matrix_data( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, INT_TYPE* row_ptr, INT_TYPE* col_ind, FP_TYPE* val) FOR_EACH_FP_AND_INT_TYPE(INSTANTIATE_CSR_MATRIX_FUNCS); #undef INSTANTIATE_CSR_MATRIX_FUNCS // Common sparse matrix functions -sycl::event release_sparse_matrix(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, - const std::vector &dependencies) { +sycl::event release_sparse_matrix(sycl::queue& queue, + oneapi::math::sparse::matrix_handle_t smhandle, + const std::vector& dependencies) { auto internal_smhandle = detail::get_internal_handle(smhandle); // Asynchronously release the backend's handle followed by the internal handle. auto event = oneapi::mkl::sparse::release_matrix_handle( @@ -372,7 +382,7 @@ sycl::event release_sparse_matrix(sycl::queue &queue, oneapi::math::sparse::matr return detail::submit_release(queue, internal_smhandle, event); } -bool set_matrix_property(sycl::queue & /*queue*/, oneapi::math::sparse::matrix_handle_t smhandle, +bool set_matrix_property(sycl::queue& /*queue*/, oneapi::math::sparse::matrix_handle_t smhandle, oneapi::math::sparse::matrix_property property) { auto internal_smhandle = detail::get_internal_handle(smhandle); // Store the matrix property internally for better error checking @@ -381,12 +391,12 @@ bool set_matrix_property(sycl::queue & /*queue*/, oneapi::math::sparse::matrix_h // Backend and oneMath types for the property don't match switch (property) { case oneapi::math::sparse::matrix_property::symmetric: - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_matrix_property(internal_smhandle->backend_handle, - oneapi::mkl::sparse::property::symmetric)); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_matrix_property( + internal_smhandle->backend_handle, oneapi::mkl::sparse::property::symmetric)); return true; case oneapi::math::sparse::matrix_property::sorted: - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_matrix_property(internal_smhandle->backend_handle, - oneapi::mkl::sparse::property::sorted)); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::set_matrix_property( + internal_smhandle->backend_handle, oneapi::mkl::sparse::property::sorted)); return true; default: return false; } diff --git a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp index 8401d76c6..e7a7527f2 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_helper.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_helper.hpp @@ -39,15 +39,15 @@ using namespace oneapi::math::detail; /// Return whether a pointer is accessible on the host template -inline bool is_ptr_accessible_on_host(sycl::queue &queue, const T *host_or_device_ptr) { +inline bool is_ptr_accessible_on_host(sycl::queue& queue, const T* host_or_device_ptr) { auto alloc_type = sycl::get_pointer_type(host_or_device_ptr, queue.get_context()); return alloc_type == sycl::usm::alloc::host || alloc_type == sycl::usm::alloc::shared || alloc_type == sycl::usm::alloc::unknown; } /// Throw an exception if the scalar is not accessible in the host -inline void check_ptr_is_host_accessible(const std::string &function_name, - const std::string &scalar_name, +inline void check_ptr_is_host_accessible(const std::string& function_name, + const std::string& scalar_name, bool is_ptr_accessible_on_host) { if (!is_ptr_accessible_on_host) { throw math::invalid_argument( @@ -59,7 +59,7 @@ inline void check_ptr_is_host_accessible(const std::string &function_name, /// Return a scalar on the host from a pointer to host or device memory /// Used for USM functions template -inline T get_scalar_on_host(sycl::queue &queue, const T *host_or_device_ptr, +inline T get_scalar_on_host(sycl::queue& queue, const T* host_or_device_ptr, bool is_ptr_accessible_on_host) { if (is_ptr_accessible_on_host) { return *host_or_device_ptr; @@ -71,8 +71,8 @@ inline T get_scalar_on_host(sycl::queue &queue, const T *host_or_device_ptr, } /// Merge multiple event dependencies into one -inline sycl::event collapse_dependencies(sycl::queue &queue, - const std::vector &dependencies) { +inline sycl::event collapse_dependencies(sycl::queue& queue, + const std::vector& dependencies) { if (dependencies.empty()) { return {}; } @@ -80,7 +80,7 @@ inline sycl::event collapse_dependencies(sycl::queue &queue, return dependencies[0]; } - return queue.submit([&](sycl::handler &cgh) { + return queue.submit([&](sycl::handler& cgh) { cgh.depends_on(dependencies); cgh.host_task([=]() {}); }); @@ -95,7 +95,7 @@ inline sycl::event collapse_dependencies(sycl::queue &queue, case detail::data_type::complex_fp64: \ return op_functor>(__VA_ARGS__); \ default: \ - throw oneapi::math::exception( \ + throw oneapi::math::exception( \ "sparse_blas", function_name, \ "Internal error: unsupported type " + data_type_to_str(value_type)); \ } @@ -103,7 +103,7 @@ inline sycl::event collapse_dependencies(sycl::queue &queue, #define CHECK_DESCR_MATCH(descr, argument, optimize_func_name) \ do { \ if (descr->last_optimized_##argument != argument) { \ - throw math::invalid_argument( \ + throw math::invalid_argument( \ "sparse_blas", __func__, \ #argument " argument must match with the previous call to " #optimize_func_name); \ } \ diff --git a/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx b/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx index 92cb6ec83..357d22bd6 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spmm.cxx @@ -35,16 +35,16 @@ struct spmm_descr { namespace oneapi::math::sparse::BACKEND { -void init_spmm_descr(sycl::queue & /*queue*/, oneapi::math::sparse::spmm_descr_t *p_spmm_descr) { +void init_spmm_descr(sycl::queue& /*queue*/, oneapi::math::sparse::spmm_descr_t* p_spmm_descr) { *p_spmm_descr = new spmm_descr(); } -sycl::event release_spmm_descr(sycl::queue &queue, oneapi::math::sparse::spmm_descr_t spmm_descr, - const std::vector &dependencies) { +sycl::event release_spmm_descr(sycl::queue& queue, oneapi::math::sparse::spmm_descr_t spmm_descr, + const std::vector& dependencies) { return detail::submit_release(queue, spmm_descr, dependencies); } -void check_valid_spmm(const std::string &function_name, oneapi::math::transpose opA, +void check_valid_spmm(const std::string& function_name, oneapi::math::transpose opA, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_matrix_handle_t B_handle, @@ -67,17 +67,17 @@ void check_valid_spmm(const std::string &function_name, oneapi::math::transpose } if (B_handle->dense_layout != C_handle->dense_layout) { throw math::invalid_argument("sparse_blas", function_name, - "B and C matrices must used the same layout."); + "B and C matrices must used the same layout."); } if (A_view.type_view != oneapi::math::sparse::matrix_descr::general) { throw math::invalid_argument("sparse_blas", function_name, - "Matrix view's type must be `matrix_descr::general`."); + "Matrix view's type must be `matrix_descr::general`."); } if (A_view.diag_view != oneapi::math::diag::nonunit) { throw math::invalid_argument("sparse_blas", function_name, - "Matrix's diag_view must be `nonunit`."); + "Matrix's diag_view must be `nonunit`."); } #if BACKEND == gpu @@ -95,14 +95,15 @@ void check_valid_spmm(const std::string &function_name, oneapi::math::transpose #endif // BACKEND } -void spmm_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, - oneapi::math::transpose /*opB*/, const void *alpha, +void spmm_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, + oneapi::math::transpose /*opB*/, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void* beta, oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg /*alg*/, - oneapi::math::sparse::spmm_descr_t spmm_descr, std::size_t &temp_buffer_size) { + oneapi::math::sparse::spmm_descr_t spmm_descr, + std::size_t& temp_buffer_size) { // TODO: Add support for external workspace once the close-source oneMath backend supports it. bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); @@ -113,9 +114,9 @@ void spmm_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, } inline void common_spmm_optimize( - sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, + sycl::queue& queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void* beta, oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg alg, oneapi::math::sparse::spmm_descr_t spmm_descr) { bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); @@ -124,7 +125,7 @@ inline void common_spmm_optimize( is_alpha_host_accessible, is_beta_host_accessible); if (!spmm_descr->buffer_size_called) { throw math::uninitialized("sparse_blas", "spmm_optimize", - "spmm_buffer_size must be called before spmm_optimize."); + "spmm_buffer_size must be called before spmm_optimize."); } spmm_descr->optimized_called = true; spmm_descr->last_optimized_opA = opA; @@ -136,12 +137,13 @@ inline void common_spmm_optimize( spmm_descr->last_optimized_alg = alg; } -void spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, oneapi::math::sparse::matrix_view A_view, +void spmm_optimize(sycl::queue& queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void* beta, oneapi::math::sparse::dense_matrix_handle_t C_handle, - oneapi::math::sparse::spmm_alg alg, oneapi::math::sparse::spmm_descr_t spmm_descr, + oneapi::math::sparse::spmm_alg alg, + oneapi::math::sparse::spmm_descr_t spmm_descr, sycl::buffer /*workspace*/) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (!internal_A_handle->all_use_buffer()) { @@ -156,15 +158,15 @@ void spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math // TODO: Add support for spmm_optimize once the close-source oneMath backend supports it. } -sycl::event spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, - oneapi::math::transpose opB, const void *alpha, +sycl::event spmm_optimize(sycl::queue& queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void* beta, oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg alg, - oneapi::math::sparse::spmm_descr_t spmm_descr, void * /*workspace*/, - const std::vector &dependencies) { + oneapi::math::sparse::spmm_descr_t spmm_descr, void* /*workspace*/, + const std::vector& dependencies) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (internal_A_handle->all_use_buffer()) { detail::throw_incompatible_container(__func__); @@ -181,16 +183,16 @@ sycl::event spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, template sycl::event internal_spmm( - sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void *alpha, + sycl::queue& queue, oneapi::math::transpose opA, oneapi::math::transpose opB, const void* alpha, oneapi::math::sparse::matrix_view /*A_view*/, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void* beta, oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg /*alg*/, - oneapi::math::sparse::spmm_descr_t /*spmm_descr*/, const std::vector &dependencies, + oneapi::math::sparse::spmm_descr_t /*spmm_descr*/, const std::vector& dependencies, bool is_alpha_host_accessible, bool is_beta_host_accessible) { T host_alpha = - detail::get_scalar_on_host(queue, static_cast(alpha), is_alpha_host_accessible); + detail::get_scalar_on_host(queue, static_cast(alpha), is_alpha_host_accessible); T host_beta = - detail::get_scalar_on_host(queue, static_cast(beta), is_beta_host_accessible); + detail::get_scalar_on_host(queue, static_cast(beta), is_beta_host_accessible); auto internal_A_handle = detail::get_internal_handle(A_handle); internal_A_handle->can_be_reset = false; auto onemkl_layout = detail::get_onemkl_layout(B_handle->dense_layout); @@ -200,27 +202,28 @@ sycl::event internal_spmm( auto ldb = B_handle->ld; auto ldc = C_handle->ld; if (internal_A_handle->all_use_buffer()) { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::gemm(queue, onemkl_layout, onemkl_opa, onemkl_opb, host_alpha, - internal_A_handle->backend_handle, B_handle->get_buffer(), - columns, ldb, host_beta, C_handle->get_buffer(), ldc)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::sparse::gemm(queue, onemkl_layout, onemkl_opa, onemkl_opb, host_alpha, + internal_A_handle->backend_handle, B_handle->get_buffer(), + columns, ldb, host_beta, C_handle->get_buffer(), ldc)); // Dependencies are not used for buffers return {}; } else { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::gemm(queue, onemkl_layout, onemkl_opa, onemkl_opb, host_alpha, - internal_A_handle->backend_handle, - B_handle->get_usm_ptr(), columns, ldb, host_beta, - C_handle->get_usm_ptr(), ldc, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::gemm( + queue, onemkl_layout, onemkl_opa, onemkl_opb, host_alpha, + internal_A_handle->backend_handle, B_handle->get_usm_ptr(), columns, ldb, host_beta, + C_handle->get_usm_ptr(), ldc, dependencies)); } } -sycl::event spmm(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, oneapi::math::sparse::matrix_view A_view, +sycl::event spmm(sycl::queue& queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void* beta, oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg alg, oneapi::math::sparse::spmm_descr_t spmm_descr, - const std::vector &dependencies) { + const std::vector& dependencies) { bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); check_valid_spmm(__func__, opA, A_view, A_handle, B_handle, C_handle, is_alpha_host_accessible, @@ -228,7 +231,7 @@ sycl::event spmm(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math:: if (!spmm_descr->optimized_called) { throw math::uninitialized("sparse_blas", __func__, - "spmm_optimize must be called before spmm."); + "spmm_optimize must be called before spmm."); } CHECK_DESCR_MATCH(spmm_descr, opA, "spmm_optimize"); CHECK_DESCR_MATCH(spmm_descr, opB, "spmm_optimize"); diff --git a/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx b/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx index 5d03466fd..b434e15a4 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spmv.cxx @@ -34,16 +34,16 @@ struct spmv_descr { namespace oneapi::math::sparse::BACKEND { -void init_spmv_descr(sycl::queue & /*queue*/, oneapi::math::sparse::spmv_descr_t *p_spmv_descr) { +void init_spmv_descr(sycl::queue& /*queue*/, oneapi::math::sparse::spmv_descr_t* p_spmv_descr) { *p_spmv_descr = new spmv_descr(); } -sycl::event release_spmv_descr(sycl::queue &queue, oneapi::math::sparse::spmv_descr_t spmv_descr, - const std::vector &dependencies) { +sycl::event release_spmv_descr(sycl::queue& queue, oneapi::math::sparse::spmv_descr_t spmv_descr, + const std::vector& dependencies) { return detail::submit_release(queue, spmv_descr, dependencies); } -void check_valid_spmv(const std::string &function_name, oneapi::math::transpose opA, +void check_valid_spmv(const std::string& function_name, oneapi::math::transpose opA, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, @@ -66,7 +66,7 @@ void check_valid_spmv(const std::string &function_name, oneapi::math::transpose } if (A_view.type_view == oneapi::math::sparse::matrix_descr::diagonal) { throw math::invalid_argument("sparse_blas", function_name, - "Matrix view's type cannot be diagonal."); + "Matrix view's type cannot be diagonal."); } if (A_view.type_view != oneapi::math::sparse::matrix_descr::triangular && @@ -85,13 +85,14 @@ void check_valid_spmv(const std::string &function_name, oneapi::math::transpose } } -void spmv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spmv_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void* beta, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spmv_alg /*alg*/, - oneapi::math::sparse::spmv_descr_t spmv_descr, std::size_t &temp_buffer_size) { + oneapi::math::sparse::spmv_descr_t spmv_descr, + std::size_t& temp_buffer_size) { // TODO: Add support for external workspace once the close-source oneMath backend supports it. bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); @@ -101,11 +102,11 @@ void spmv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const voi spmv_descr->buffer_size_called = true; } -inline void common_spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +inline void common_spmv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, - const void *beta, + const void* beta, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::spmv_descr_t spmv_descr) { @@ -115,7 +116,7 @@ inline void common_spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA is_alpha_host_accessible, is_beta_host_accessible); if (!spmv_descr->buffer_size_called) { throw math::uninitialized("sparse_blas", "spmv_optimize", - "spmv_buffer_size must be called before spmv_optimize."); + "spmv_buffer_size must be called before spmv_optimize."); } spmv_descr->optimized_called = true; spmv_descr->last_optimized_opA = opA; @@ -126,12 +127,13 @@ inline void common_spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA spmv_descr->last_optimized_alg = alg; } -void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spmv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void* beta, oneapi::math::sparse::dense_vector_handle_t y_handle, - oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::spmv_descr_t spmv_descr, + oneapi::math::sparse::spmv_alg alg, + oneapi::math::sparse::spmv_descr_t spmv_descr, sycl::buffer /*workspace*/) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (!internal_A_handle->all_use_buffer()) { @@ -147,8 +149,8 @@ void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void * auto onemkl_opa = detail::get_onemkl_transpose(opA); auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); if (A_view.type_view == matrix_descr::triangular) { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::optimize_trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, - internal_A_handle->backend_handle)); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::optimize_trmv( + queue, onemkl_uplo, onemkl_opa, onemkl_diag, internal_A_handle->backend_handle)); } else if (A_view.type_view == matrix_descr::symmetric || A_view.type_view == matrix_descr::hermitian) { @@ -156,18 +158,19 @@ void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void * return; } else { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::optimize_gemv(queue, onemkl_opa, internal_A_handle->backend_handle)); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::optimize_gemv( + queue, onemkl_opa, internal_A_handle->backend_handle)); } } -sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spmv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void* beta, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spmv_alg alg, - oneapi::math::sparse::spmv_descr_t spmv_descr, void * /*workspace*/, - const std::vector &dependencies) { + oneapi::math::sparse::spmv_descr_t spmv_descr, void* /*workspace*/, + const std::vector& dependencies) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (internal_A_handle->all_use_buffer()) { detail::throw_incompatible_container(__func__); @@ -182,33 +185,34 @@ sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const auto onemkl_opa = detail::get_onemkl_transpose(opA); auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); if (A_view.type_view == matrix_descr::triangular) { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::optimize_trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, - internal_A_handle->backend_handle, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + oneapi::mkl::sparse::optimize_trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, + internal_A_handle->backend_handle, dependencies)); } else if (A_view.type_view == matrix_descr::symmetric || A_view.type_view == matrix_descr::hermitian) { return detail::collapse_dependencies(queue, dependencies); } else { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::optimize_gemv(queue, onemkl_opa, internal_A_handle->backend_handle, - dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::optimize_gemv( + queue, onemkl_opa, internal_A_handle->backend_handle, dependencies)); } } template -sycl::event internal_spmv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event internal_spmv(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void* beta, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spmv_alg /*alg*/, oneapi::math::sparse::spmv_descr_t /*spmv_descr*/, - const std::vector &dependencies, + const std::vector& dependencies, bool is_alpha_host_accessible, bool is_beta_host_accessible) { T host_alpha = - detail::get_scalar_on_host(queue, static_cast(alpha), is_alpha_host_accessible); + detail::get_scalar_on_host(queue, static_cast(alpha), is_alpha_host_accessible); T host_beta = - detail::get_scalar_on_host(queue, static_cast(beta), is_beta_host_accessible); + detail::get_scalar_on_host(queue, static_cast(beta), is_beta_host_accessible); auto internal_A_handle = detail::get_internal_handle(A_handle); internal_A_handle->can_be_reset = false; auto backend_handle = internal_A_handle->backend_handle; @@ -219,17 +223,18 @@ sycl::event internal_spmv(sycl::queue &queue, oneapi::math::transpose opA, const auto x_buffer = x_handle->get_buffer(); auto y_buffer = y_handle->get_buffer(); if (A_view.type_view == matrix_descr::triangular) { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, - backend_handle, x_buffer, host_beta, y_buffer)); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::sparse::trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, + backend_handle, x_buffer, host_beta, y_buffer)); } else if (A_view.type_view == matrix_descr::symmetric || A_view.type_view == matrix_descr::hermitian) { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::symv(queue, onemkl_uplo, host_alpha, backend_handle, x_buffer, - host_beta, y_buffer)); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::symv( + queue, onemkl_uplo, host_alpha, backend_handle, x_buffer, host_beta, y_buffer)); } else { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::gemv(queue, onemkl_opa, host_alpha, backend_handle, x_buffer, host_beta, - y_buffer)); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::gemv( + queue, onemkl_opa, host_alpha, backend_handle, x_buffer, host_beta, y_buffer)); } // Dependencies are not used for buffers return {}; @@ -238,29 +243,31 @@ sycl::event internal_spmv(sycl::queue &queue, oneapi::math::transpose opA, const auto x_usm = x_handle->get_usm_ptr(); auto y_usm = y_handle->get_usm_ptr(); if (A_view.type_view == matrix_descr::triangular) { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, - host_alpha, backend_handle, x_usm, host_beta, y_usm, - dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + oneapi::mkl::sparse::trmv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, + backend_handle, x_usm, host_beta, y_usm, dependencies)); } else if (A_view.type_view == matrix_descr::symmetric || A_view.type_view == matrix_descr::hermitian) { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::symv(queue, onemkl_uplo, host_alpha, backend_handle, - x_usm, host_beta, y_usm, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + oneapi::mkl::sparse::symv(queue, onemkl_uplo, host_alpha, backend_handle, x_usm, + host_beta, y_usm, dependencies)); } else { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::gemv(queue, onemkl_opa, host_alpha, backend_handle, x_usm, - host_beta, y_usm, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + oneapi::mkl::sparse::gemv(queue, onemkl_opa, host_alpha, backend_handle, x_usm, + host_beta, y_usm, dependencies)); } } } -sycl::event spmv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spmv(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void* beta, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::spmv_descr_t spmv_descr, - const std::vector &dependencies) { + const std::vector& dependencies) { bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); bool is_beta_host_accessible = detail::is_ptr_accessible_on_host(queue, beta); check_valid_spmv(__func__, opA, A_view, A_handle, x_handle, y_handle, is_alpha_host_accessible, @@ -268,7 +275,7 @@ sycl::event spmv(sycl::queue &queue, oneapi::math::transpose opA, const void *al if (!spmv_descr->optimized_called) { throw math::uninitialized("sparse_blas", __func__, - "spmv_optimize must be called before spmv."); + "spmv_optimize must be called before spmv."); } CHECK_DESCR_MATCH(spmv_descr, opA, "spmv_optimize"); CHECK_DESCR_MATCH(spmv_descr, A_view, "spmv_optimize"); diff --git a/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx b/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx index df998c59a..963f84927 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx +++ b/src/sparse_blas/backends/mkl_common/mkl_spsv.cxx @@ -34,16 +34,16 @@ struct spsv_descr { namespace oneapi::math::sparse::BACKEND { -void init_spsv_descr(sycl::queue & /*queue*/, oneapi::math::sparse::spsv_descr_t *p_spsv_descr) { +void init_spsv_descr(sycl::queue& /*queue*/, oneapi::math::sparse::spsv_descr_t* p_spsv_descr) { *p_spsv_descr = new spsv_descr(); } -sycl::event release_spsv_descr(sycl::queue &queue, oneapi::math::sparse::spsv_descr_t spsv_descr, - const std::vector &dependencies) { +sycl::event release_spsv_descr(sycl::queue& queue, oneapi::math::sparse::spsv_descr_t spsv_descr, + const std::vector& dependencies) { return detail::submit_release(queue, spsv_descr, dependencies); } -void check_valid_spsv(const std::string &function_name, oneapi::math::transpose opA, +void check_valid_spsv(const std::string& function_name, oneapi::math::transpose opA, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, @@ -67,7 +67,7 @@ void check_valid_spsv(const std::string &function_name, oneapi::math::transpose data_type == detail::data_type::complex_fp64) && opA == oneapi::math::transpose::conjtrans) { throw math::unimplemented("sparse_blas", function_name, - "The backend does not support spsv using conjtrans."); + "The backend does not support spsv using conjtrans."); } #else (void)opA; @@ -76,7 +76,7 @@ void check_valid_spsv(const std::string &function_name, oneapi::math::transpose detail::check_all_containers_compatible(function_name, internal_A_handle, x_handle, y_handle); if (A_view.type_view != matrix_descr::triangular) { throw math::invalid_argument("sparse_blas", function_name, - "Matrix view's type must be `matrix_descr::triangular`."); + "Matrix view's type must be `matrix_descr::triangular`."); } if (internal_A_handle->all_use_buffer()) { @@ -84,13 +84,14 @@ void check_valid_spsv(const std::string &function_name, oneapi::math::transpose } } -void spsv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spsv_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spsv_alg alg, - oneapi::math::sparse::spsv_descr_t spsv_descr, std::size_t &temp_buffer_size) { + oneapi::math::sparse::spsv_descr_t spsv_descr, + std::size_t& temp_buffer_size) { // TODO: Add support for external workspace once the close-source oneMath backend supports it. bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); check_valid_spsv(__func__, opA, A_view, A_handle, x_handle, y_handle, is_alpha_host_accessible, @@ -99,7 +100,7 @@ void spsv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const voi spsv_descr->buffer_size_called = true; } -inline void common_spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +inline void common_spsv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, @@ -111,7 +112,7 @@ inline void common_spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA is_alpha_host_accessible, alg); if (!spsv_descr->buffer_size_called) { throw math::uninitialized("sparse_blas", "spsv_optimize", - "spsv_buffer_size must be called before spsv_optimize."); + "spsv_buffer_size must be called before spsv_optimize."); } spsv_descr->optimized_called = true; spsv_descr->last_optimized_opA = opA; @@ -122,12 +123,13 @@ inline void common_spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA spsv_descr->last_optimized_alg = alg; } -void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spsv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, oneapi::math::sparse::dense_vector_handle_t y_handle, - oneapi::math::sparse::spsv_alg alg, oneapi::math::sparse::spsv_descr_t spsv_descr, + oneapi::math::sparse::spsv_alg alg, + oneapi::math::sparse::spsv_descr_t spsv_descr, sycl::buffer /*workspace*/) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (!internal_A_handle->all_use_buffer()) { @@ -141,18 +143,18 @@ void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void * auto onemkl_uplo = detail::get_onemkl_uplo(A_view.uplo_view); auto onemkl_opa = detail::get_onemkl_transpose(opA); auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::optimize_trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, - internal_A_handle->backend_handle)); + RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::optimize_trsv( + queue, onemkl_uplo, onemkl_opa, onemkl_diag, internal_A_handle->backend_handle)); } -sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spsv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spsv_alg alg, - oneapi::math::sparse::spsv_descr_t spsv_descr, void * /*workspace*/, - const std::vector &dependencies) { + oneapi::math::sparse::spsv_descr_t spsv_descr, void* /*workspace*/, + const std::vector& dependencies) { auto internal_A_handle = detail::get_internal_handle(A_handle); if (internal_A_handle->all_use_buffer()) { detail::throw_incompatible_container(__func__); @@ -165,56 +167,58 @@ sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const auto onemkl_uplo = detail::get_onemkl_uplo(A_view.uplo_view); auto onemkl_opa = detail::get_onemkl_transpose(opA); auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::optimize_trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, - internal_A_handle->backend_handle, dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + oneapi::mkl::sparse::optimize_trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, + internal_A_handle->backend_handle, dependencies)); } template -sycl::event internal_spsv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event internal_spsv(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spsv_alg /*alg*/, oneapi::math::sparse::spsv_descr_t /*spsv_descr*/, - const std::vector &dependencies, + const std::vector& dependencies, bool is_alpha_host_accessible) { T host_alpha = - detail::get_scalar_on_host(queue, static_cast(alpha), is_alpha_host_accessible); + detail::get_scalar_on_host(queue, static_cast(alpha), is_alpha_host_accessible); auto internal_A_handle = detail::get_internal_handle(A_handle); internal_A_handle->can_be_reset = false; auto onemkl_uplo = detail::get_onemkl_uplo(A_view.uplo_view); auto onemkl_opa = detail::get_onemkl_transpose(opA); auto onemkl_diag = detail::get_onemkl_diag(A_view.diag_view); if (internal_A_handle->all_use_buffer()) { - RETHROW_ONEMKL_EXCEPTIONS(oneapi::mkl::sparse::trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, - internal_A_handle->backend_handle, x_handle->get_buffer(), - y_handle->get_buffer())); + RETHROW_ONEMKL_EXCEPTIONS( + oneapi::mkl::sparse::trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, + internal_A_handle->backend_handle, x_handle->get_buffer(), + y_handle->get_buffer())); // Dependencies are not used for buffers return {}; } else { - RETHROW_ONEMKL_EXCEPTIONS_RET(oneapi::mkl::sparse::trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, - internal_A_handle->backend_handle, - x_handle->get_usm_ptr(), y_handle->get_usm_ptr(), - dependencies)); + RETHROW_ONEMKL_EXCEPTIONS_RET( + oneapi::mkl::sparse::trsv(queue, onemkl_uplo, onemkl_opa, onemkl_diag, host_alpha, + internal_A_handle->backend_handle, x_handle->get_usm_ptr(), + y_handle->get_usm_ptr(), dependencies)); } } -sycl::event spsv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spsv(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spsv_alg alg, oneapi::math::sparse::spsv_descr_t spsv_descr, - const std::vector &dependencies) { + const std::vector& dependencies) { bool is_alpha_host_accessible = detail::is_ptr_accessible_on_host(queue, alpha); check_valid_spsv(__func__, opA, A_view, A_handle, x_handle, y_handle, is_alpha_host_accessible, alg); if (!spsv_descr->optimized_called) { throw math::uninitialized("sparse_blas", __func__, - "spsv_optimize must be called before spsv."); + "spsv_optimize must be called before spsv."); } CHECK_DESCR_MATCH(spsv_descr, opA, "spsv_optimize"); CHECK_DESCR_MATCH(spsv_descr, A_view, "spsv_optimize"); diff --git a/src/sparse_blas/function_table.hpp b/src/sparse_blas/function_table.hpp index 3667ff868..6139a6381 100644 --- a/src/sparse_blas/function_table.hpp +++ b/src/sparse_blas/function_table.hpp @@ -24,95 +24,95 @@ #include "sparse_blas/macros.hpp" // Dense vector -#define DEFINE_DENSE_VECTOR_FUNCS(FP_TYPE, FP_SUFFIX) \ - void (*init_dense_vector_buffer##FP_SUFFIX)( \ +#define DEFINE_DENSE_VECTOR_FUNCS(FP_TYPE, FP_SUFFIX) \ + void (*init_dense_vector_buffer##FP_SUFFIX)( \ sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t * p_dvhandle, \ - std::int64_t size, sycl::buffer val); \ - void (*init_dense_vector_usm##FP_SUFFIX)( \ + std::int64_t size, sycl::buffer val); \ + void (*init_dense_vector_usm##FP_SUFFIX)( \ sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t * p_dvhandle, \ - std::int64_t size, FP_TYPE * val); \ - void (*set_dense_vector_data_buffer##FP_SUFFIX)( \ + std::int64_t size, FP_TYPE* val); \ + void (*set_dense_vector_data_buffer##FP_SUFFIX)( \ sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t dvhandle, \ - std::int64_t size, sycl::buffer val); \ - void (*set_dense_vector_data_usm##FP_SUFFIX)( \ + std::int64_t size, sycl::buffer val); \ + void (*set_dense_vector_data_usm##FP_SUFFIX)( \ sycl::queue & queue, oneapi::math::sparse::dense_vector_handle_t dvhandle, \ - std::int64_t size, FP_TYPE * val) + std::int64_t size, FP_TYPE* val) // Dense matrix -#define DEFINE_DENSE_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX) \ - void (*init_dense_matrix_buffer##FP_SUFFIX)( \ +#define DEFINE_DENSE_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX) \ + void (*init_dense_matrix_buffer##FP_SUFFIX)( \ sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t * p_dmhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ oneapi::math::layout dense_layout, sycl::buffer val); \ - void (*init_dense_matrix_usm##FP_SUFFIX)( \ + void (*init_dense_matrix_usm##FP_SUFFIX)( \ sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t * p_dmhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::math::layout dense_layout, FP_TYPE * val); \ - void (*set_dense_matrix_data_buffer##FP_SUFFIX)( \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ + oneapi::math::layout dense_layout, FP_TYPE* val); \ + void (*set_dense_matrix_data_buffer##FP_SUFFIX)( \ sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t dmhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ oneapi::math::layout dense_layout, sycl::buffer val); \ - void (*set_dense_matrix_data_usm##FP_SUFFIX)( \ + void (*set_dense_matrix_data_usm##FP_SUFFIX)( \ sycl::queue & queue, oneapi::math::sparse::dense_matrix_handle_t dmhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - oneapi::math::layout dense_layout, FP_TYPE * val) + std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ + oneapi::math::layout dense_layout, FP_TYPE* val) // COO matrix -#define DEFINE_COO_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ - void (*init_coo_matrix_buffer##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::math::index_base index, sycl::buffer row_ind, \ - sycl::buffer col_ind, sycl::buffer val); \ - void (*init_coo_matrix_usm##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::math::index_base index, INT_TYPE * row_ind, INT_TYPE * col_ind, FP_TYPE * val); \ - void (*set_coo_matrix_data_buffer##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ - sycl::buffer row_ind, sycl::buffer col_ind, \ - sycl::buffer val); \ - void (*set_coo_matrix_data_usm##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ - INT_TYPE * row_ind, INT_TYPE * col_ind, FP_TYPE * val) +#define DEFINE_COO_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ + void (*init_coo_matrix_buffer##FP_SUFFIX##INT_SUFFIX)( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, sycl::buffer row_ind, \ + sycl::buffer col_ind, sycl::buffer val); \ + void (*init_coo_matrix_usm##FP_SUFFIX##INT_SUFFIX)( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, INT_TYPE* row_ind, INT_TYPE* col_ind, FP_TYPE* val); \ + void (*set_coo_matrix_data_buffer##FP_SUFFIX##INT_SUFFIX)( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, sycl::buffer row_ind, \ + sycl::buffer col_ind, sycl::buffer val); \ + void (*set_coo_matrix_data_usm##FP_SUFFIX##INT_SUFFIX)( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, INT_TYPE* row_ind, INT_TYPE* col_ind, FP_TYPE* val) // CSR matrix -#define DEFINE_CSR_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ - void (*init_csr_matrix_buffer##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::math::index_base index, sycl::buffer row_ptr, \ - sycl::buffer col_ind, sycl::buffer val); \ - void (*init_csr_matrix_usm##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ - std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ - oneapi::math::index_base index, INT_TYPE * row_ptr, INT_TYPE * col_ind, FP_TYPE * val); \ - void (*set_csr_matrix_data_buffer##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ - sycl::buffer row_ptr, sycl::buffer col_ind, \ - sycl::buffer val); \ - void (*set_csr_matrix_data_usm##FP_SUFFIX##INT_SUFFIX)( \ - sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, \ - std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, \ - INT_TYPE * row_ptr, INT_TYPE * col_ind, FP_TYPE * val) +#define DEFINE_CSR_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ + void (*init_csr_matrix_buffer##FP_SUFFIX##INT_SUFFIX)( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, sycl::buffer row_ptr, \ + sycl::buffer col_ind, sycl::buffer val); \ + void (*init_csr_matrix_usm##FP_SUFFIX##INT_SUFFIX)( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t * p_smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, INT_TYPE* row_ptr, INT_TYPE* col_ind, FP_TYPE* val); \ + void (*set_csr_matrix_data_buffer##FP_SUFFIX##INT_SUFFIX)( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, sycl::buffer row_ptr, \ + sycl::buffer col_ind, sycl::buffer val); \ + void (*set_csr_matrix_data_usm##FP_SUFFIX##INT_SUFFIX)( \ + sycl::queue & queue, oneapi::math::sparse::matrix_handle_t smhandle, \ + std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, \ + oneapi::math::index_base index, INT_TYPE* row_ptr, INT_TYPE* col_ind, FP_TYPE* val) typedef struct { int version; // Dense vector FOR_EACH_FP_TYPE(DEFINE_DENSE_VECTOR_FUNCS); - sycl::event (*release_dense_vector)(sycl::queue &queue, + sycl::event (*release_dense_vector)(sycl::queue& queue, oneapi::math::sparse::dense_vector_handle_t dvhandle, - const std::vector &dependencies); + const std::vector& dependencies); // Dense matrix FOR_EACH_FP_TYPE(DEFINE_DENSE_MATRIX_FUNCS); - sycl::event (*release_dense_matrix)(sycl::queue &queue, + sycl::event (*release_dense_matrix)(sycl::queue& queue, oneapi::math::sparse::dense_matrix_handle_t dmhandle, - const std::vector &dependencies); + const std::vector& dependencies); // COO matrix FOR_EACH_FP_AND_INT_TYPE(DEFINE_COO_MATRIX_FUNCS); @@ -121,117 +121,118 @@ typedef struct { FOR_EACH_FP_AND_INT_TYPE(DEFINE_CSR_MATRIX_FUNCS); // Common sparse matrix functions - sycl::event (*release_sparse_matrix)(sycl::queue &queue, + sycl::event (*release_sparse_matrix)(sycl::queue& queue, oneapi::math::sparse::matrix_handle_t smhandle, - const std::vector &dependencies); + const std::vector& dependencies); - bool (*set_matrix_property)(sycl::queue &queue, oneapi::math::sparse::matrix_handle_t smhandle, + bool (*set_matrix_property)(sycl::queue& queue, oneapi::math::sparse::matrix_handle_t smhandle, oneapi::math::sparse::matrix_property property); // SPMM - void (*init_spmm_descr)(sycl::queue &queue, oneapi::math::sparse::spmm_descr_t *p_spmm_descr); + void (*init_spmm_descr)(sycl::queue& queue, oneapi::math::sparse::spmm_descr_t* p_spmm_descr); - sycl::event (*release_spmm_descr)(sycl::queue &queue, + sycl::event (*release_spmm_descr)(sycl::queue& queue, oneapi::math::sparse::spmm_descr_t spmm_descr, - const std::vector &dependencies); + const std::vector& dependencies); - void (*spmm_buffer_size)(sycl::queue &queue, oneapi::math::transpose opA, - oneapi::math::transpose opB, const void *alpha, + void (*spmm_buffer_size)(sycl::queue& queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void* beta, oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg alg, oneapi::math::sparse::spmm_descr_t spmm_descr, - std::size_t &temp_buffer_size); + std::size_t& temp_buffer_size); void (*spmm_optimize_buffer)( - sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, oneapi::math::sparse::matrix_view A_view, + sycl::queue& queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void* beta, oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg alg, oneapi::math::sparse::spmm_descr_t spmm_descr, sycl::buffer workspace); - sycl::event (*spmm_optimize_usm)(sycl::queue &queue, oneapi::math::transpose opA, - oneapi::math::transpose opB, const void *alpha, + sycl::event (*spmm_optimize_usm)(sycl::queue& queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_matrix_handle_t B_handle, - const void *beta, + const void* beta, oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg alg, - oneapi::math::sparse::spmm_descr_t spmm_descr, void *workspace, - const std::vector &dependencies); + oneapi::math::sparse::spmm_descr_t spmm_descr, void* workspace, + const std::vector& dependencies); - sycl::event (*spmm)(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, oneapi::math::sparse::matrix_view A_view, + sycl::event (*spmm)(sycl::queue& queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void* alpha, + oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_matrix_handle_t B_handle, const void *beta, + oneapi::math::sparse::dense_matrix_handle_t B_handle, const void* beta, oneapi::math::sparse::dense_matrix_handle_t C_handle, oneapi::math::sparse::spmm_alg alg, oneapi::math::sparse::spmm_descr_t spmm_descr, - const std::vector &dependencies); + const std::vector& dependencies); // SPMV - void (*init_spmv_descr)(sycl::queue &queue, oneapi::math::sparse::spmv_descr_t *p_spmv_descr); + void (*init_spmv_descr)(sycl::queue& queue, oneapi::math::sparse::spmv_descr_t* p_spmv_descr); - sycl::event (*release_spmv_descr)(sycl::queue &queue, + sycl::event (*release_spmv_descr)(sycl::queue& queue, oneapi::math::sparse::spmv_descr_t spmv_descr, - const std::vector &dependencies); + const std::vector& dependencies); - void (*spmv_buffer_size)(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + void (*spmv_buffer_size)(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void* beta, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::spmv_descr_t spmv_descr, - std::size_t &temp_buffer_size); + std::size_t& temp_buffer_size); void (*spmv_optimize_buffer)( - sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void* beta, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::spmv_descr_t spmv_descr, sycl::buffer workspace); - sycl::event (*spmv_optimize_usm)(sycl::queue &queue, oneapi::math::transpose opA, - const void *alpha, oneapi::math::sparse::matrix_view A_view, + sycl::event (*spmv_optimize_usm)(sycl::queue& queue, oneapi::math::transpose opA, + const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, - const void *beta, + const void* beta, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spmv_alg alg, - oneapi::math::sparse::spmv_descr_t spmv_descr, void *workspace, - const std::vector &dependencies); + oneapi::math::sparse::spmv_descr_t spmv_descr, void* workspace, + const std::vector& dependencies); - sycl::event (*spmv)(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + sycl::event (*spmv)(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, - oneapi::math::sparse::dense_vector_handle_t x_handle, const void *beta, + oneapi::math::sparse::dense_vector_handle_t x_handle, const void* beta, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::spmv_descr_t spmv_descr, - const std::vector &dependencies); + const std::vector& dependencies); // SPSV - void (*init_spsv_descr)(sycl::queue &queue, oneapi::math::sparse::spsv_descr_t *p_spsv_descr); + void (*init_spsv_descr)(sycl::queue& queue, oneapi::math::sparse::spsv_descr_t* p_spsv_descr); - sycl::event (*release_spsv_descr)(sycl::queue &queue, + sycl::event (*release_spsv_descr)(sycl::queue& queue, oneapi::math::sparse::spsv_descr_t spsv_descr, - const std::vector &dependencies); + const std::vector& dependencies); - void (*spsv_buffer_size)(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + void (*spsv_buffer_size)(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spsv_alg alg, oneapi::math::sparse::spsv_descr_t spsv_descr, - std::size_t &temp_buffer_size); + std::size_t& temp_buffer_size); - void (*spsv_optimize_buffer)(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + void (*spsv_optimize_buffer)(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, @@ -240,23 +241,23 @@ typedef struct { oneapi::math::sparse::spsv_descr_t spsv_descr, sycl::buffer workspace); - sycl::event (*spsv_optimize_usm)(sycl::queue &queue, oneapi::math::transpose opA, - const void *alpha, oneapi::math::sparse::matrix_view A_view, + sycl::event (*spsv_optimize_usm)(sycl::queue& queue, oneapi::math::transpose opA, + const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spsv_alg alg, - oneapi::math::sparse::spsv_descr_t spsv_descr, void *workspace, - const std::vector &dependencies); + oneapi::math::sparse::spsv_descr_t spsv_descr, void* workspace, + const std::vector& dependencies); - sycl::event (*spsv)(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, + sycl::event (*spsv)(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, oneapi::math::sparse::matrix_view A_view, oneapi::math::sparse::matrix_handle_t A_handle, oneapi::math::sparse::dense_vector_handle_t x_handle, oneapi::math::sparse::dense_vector_handle_t y_handle, oneapi::math::sparse::spsv_alg alg, oneapi::math::sparse::spsv_descr_t spsv_descr, - const std::vector &dependencies); + const std::vector& dependencies); } sparse_blas_function_table_t; #undef DEFINE_DENSE_VECTOR_FUNCS diff --git a/src/sparse_blas/generic_container.hpp b/src/sparse_blas/generic_container.hpp index 463956275..2fb91d988 100644 --- a/src/sparse_blas/generic_container.hpp +++ b/src/sparse_blas/generic_container.hpp @@ -314,10 +314,10 @@ void check_all_containers_compatible(const std::string& function_name, } const data_type other_int_type = internal_container->get_int_type(); if (other_int_type != data_type::none && other_int_type != first_int_type) { - throw oneapi::math::invalid_argument("sparse_blas", function_name, - "Incompatible integer types expected " + - data_type_to_str(first_int_type) + " but got " + - data_type_to_str(other_int_type)); + throw oneapi::math::invalid_argument( + "sparse_blas", function_name, + "Incompatible integer types expected " + data_type_to_str(first_int_type) + + " but got " + data_type_to_str(other_int_type)); } } } diff --git a/src/sparse_blas/macros.hpp b/src/sparse_blas/macros.hpp index 34db8efc5..aa7e43784 100644 --- a/src/sparse_blas/macros.hpp +++ b/src/sparse_blas/macros.hpp @@ -36,10 +36,10 @@ FOR_EACH_FP_AND_INT_TYPE_HELPER(DEFINE_MACRO, std::int32_t, _i32); \ FOR_EACH_FP_AND_INT_TYPE_HELPER(DEFINE_MACRO, std::int64_t, _i64) -#define THROW_IF_NULLPTR(FUNC_NAME, PTR) \ - if (!(PTR)) { \ +#define THROW_IF_NULLPTR(FUNC_NAME, PTR) \ + if (!(PTR)) { \ throw math::uninitialized("sparse_blas", FUNC_NAME, \ - std::string(#PTR) + " must not be nullptr."); \ + std::string(#PTR) + " must not be nullptr."); \ } #endif // _ONEMATH_SPARSE_BLAS_MACROS_HPP_ diff --git a/src/sparse_blas/sparse_blas_loader.cpp b/src/sparse_blas/sparse_blas_loader.cpp index bc17b490d..2053dffcb 100644 --- a/src/sparse_blas/sparse_blas_loader.cpp +++ b/src/sparse_blas/sparse_blas_loader.cpp @@ -27,35 +27,35 @@ namespace oneapi::math::sparse { static oneapi::math::detail::table_initializer + sparse_blas_function_table_t> function_tables; // Dense vector #define DEFINE_DENSE_VECTOR_FUNCS(FP_TYPE, FP_SUFFIX) \ template <> \ - void init_dense_vector(sycl::queue &queue, dense_vector_handle_t *p_dvhandle, \ + void init_dense_vector(sycl::queue& queue, dense_vector_handle_t* p_dvhandle, \ std::int64_t size, sycl::buffer val) { \ auto libkey = get_device_id(queue); \ function_tables[{ libkey, queue }].init_dense_vector_buffer##FP_SUFFIX(queue, p_dvhandle, \ size, val); \ } \ template <> \ - void init_dense_vector(sycl::queue &queue, dense_vector_handle_t *p_dvhandle, \ - std::int64_t size, FP_TYPE *val) { \ + void init_dense_vector(sycl::queue& queue, dense_vector_handle_t* p_dvhandle, \ + std::int64_t size, FP_TYPE* val) { \ auto libkey = get_device_id(queue); \ function_tables[{ libkey, queue }].init_dense_vector_usm##FP_SUFFIX(queue, p_dvhandle, \ size, val); \ } \ template <> \ - void set_dense_vector_data(sycl::queue &queue, dense_vector_handle_t dvhandle, \ + void set_dense_vector_data(sycl::queue& queue, dense_vector_handle_t dvhandle, \ std::int64_t size, sycl::buffer val) { \ auto libkey = get_device_id(queue); \ function_tables[{ libkey, queue }].set_dense_vector_data_buffer##FP_SUFFIX( \ queue, dvhandle, size, val); \ } \ template <> \ - void set_dense_vector_data(sycl::queue &queue, dense_vector_handle_t dvhandle, \ - std::int64_t size, FP_TYPE *val) { \ + void set_dense_vector_data(sycl::queue& queue, dense_vector_handle_t dvhandle, \ + std::int64_t size, FP_TYPE* val) { \ auto libkey = get_device_id(queue); \ function_tables[{ libkey, queue }].set_dense_vector_data_usm##FP_SUFFIX(queue, dvhandle, \ size, val); \ @@ -63,8 +63,8 @@ static oneapi::math::detail::table_initializer &dependencies) { +sycl::event release_dense_vector(sycl::queue& queue, dense_vector_handle_t dvhandle, + const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].release_dense_vector(queue, dvhandle, dependencies); } @@ -72,7 +72,7 @@ sycl::event release_dense_vector(sycl::queue &queue, dense_vector_handle_t dvhan // Dense matrix #define DEFINE_DENSE_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX) \ template <> \ - void init_dense_matrix(sycl::queue &queue, dense_matrix_handle_t *p_dmhandle, \ + void init_dense_matrix(sycl::queue& queue, dense_matrix_handle_t* p_dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ layout dense_layout, sycl::buffer val) { \ auto libkey = get_device_id(queue); \ @@ -80,15 +80,15 @@ sycl::event release_dense_vector(sycl::queue &queue, dense_vector_handle_t dvhan queue, p_dmhandle, num_rows, num_cols, ld, dense_layout, val); \ } \ template <> \ - void init_dense_matrix(sycl::queue &queue, dense_matrix_handle_t *p_dmhandle, \ + void init_dense_matrix(sycl::queue& queue, dense_matrix_handle_t* p_dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - layout dense_layout, FP_TYPE *val) { \ + layout dense_layout, FP_TYPE* val) { \ auto libkey = get_device_id(queue); \ function_tables[{ libkey, queue }].init_dense_matrix_usm##FP_SUFFIX( \ queue, p_dmhandle, num_rows, num_cols, ld, dense_layout, val); \ } \ template <> \ - void set_dense_matrix_data(sycl::queue &queue, dense_matrix_handle_t dmhandle, \ + void set_dense_matrix_data(sycl::queue& queue, dense_matrix_handle_t dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ layout dense_layout, sycl::buffer val) { \ auto libkey = get_device_id(queue); \ @@ -96,9 +96,9 @@ sycl::event release_dense_vector(sycl::queue &queue, dense_vector_handle_t dvhan queue, dmhandle, num_rows, num_cols, ld, dense_layout, val); \ } \ template <> \ - void set_dense_matrix_data(sycl::queue &queue, dense_matrix_handle_t dmhandle, \ + void set_dense_matrix_data(sycl::queue& queue, dense_matrix_handle_t dmhandle, \ std::int64_t num_rows, std::int64_t num_cols, std::int64_t ld, \ - layout dense_layout, FP_TYPE *val) { \ + layout dense_layout, FP_TYPE* val) { \ auto libkey = get_device_id(queue); \ function_tables[{ libkey, queue }].set_dense_matrix_data_usm##FP_SUFFIX( \ queue, dmhandle, num_rows, num_cols, ld, dense_layout, val); \ @@ -106,8 +106,8 @@ sycl::event release_dense_vector(sycl::queue &queue, dense_vector_handle_t dvhan FOR_EACH_FP_TYPE(DEFINE_DENSE_MATRIX_FUNCS); #undef DEFINE_DENSE_MATRIX_FUNCS -sycl::event release_dense_matrix(sycl::queue &queue, dense_matrix_handle_t dmhandle, - const std::vector &dependencies) { +sycl::event release_dense_matrix(sycl::queue& queue, dense_matrix_handle_t dmhandle, + const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].release_dense_matrix(queue, dmhandle, dependencies); } @@ -115,7 +115,7 @@ sycl::event release_dense_matrix(sycl::queue &queue, dense_matrix_handle_t dmhan // COO matrix #define DEFINE_COO_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ template <> \ - void init_coo_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, \ + void init_coo_matrix(sycl::queue& queue, matrix_handle_t* p_smhandle, std::int64_t num_rows, \ std::int64_t num_cols, std::int64_t nnz, index_base index, \ sycl::buffer row_ind, sycl::buffer col_ind, \ sycl::buffer val) { \ @@ -124,15 +124,15 @@ sycl::event release_dense_matrix(sycl::queue &queue, dense_matrix_handle_t dmhan queue, p_smhandle, num_rows, num_cols, nnz, index, row_ind, col_ind, val); \ } \ template <> \ - void init_coo_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, \ + void init_coo_matrix(sycl::queue& queue, matrix_handle_t* p_smhandle, std::int64_t num_rows, \ std::int64_t num_cols, std::int64_t nnz, index_base index, \ - INT_TYPE *row_ind, INT_TYPE *col_ind, FP_TYPE *val) { \ + INT_TYPE* row_ind, INT_TYPE* col_ind, FP_TYPE* val) { \ auto libkey = get_device_id(queue); \ function_tables[{ libkey, queue }].init_coo_matrix_usm##FP_SUFFIX##INT_SUFFIX( \ queue, p_smhandle, num_rows, num_cols, nnz, index, row_ind, col_ind, val); \ } \ template <> \ - void set_coo_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, \ + void set_coo_matrix_data(sycl::queue& queue, matrix_handle_t smhandle, std::int64_t num_rows, \ std::int64_t num_cols, std::int64_t nnz, index_base index, \ sycl::buffer row_ind, sycl::buffer col_ind, \ sycl::buffer val) { \ @@ -141,9 +141,9 @@ sycl::event release_dense_matrix(sycl::queue &queue, dense_matrix_handle_t dmhan queue, smhandle, num_rows, num_cols, nnz, index, row_ind, col_ind, val); \ } \ template <> \ - void set_coo_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, \ + void set_coo_matrix_data(sycl::queue& queue, matrix_handle_t smhandle, std::int64_t num_rows, \ std::int64_t num_cols, std::int64_t nnz, index_base index, \ - INT_TYPE *row_ind, INT_TYPE *col_ind, FP_TYPE *val) { \ + INT_TYPE* row_ind, INT_TYPE* col_ind, FP_TYPE* val) { \ auto libkey = get_device_id(queue); \ function_tables[{ libkey, queue }].set_coo_matrix_data_usm##FP_SUFFIX##INT_SUFFIX( \ queue, smhandle, num_rows, num_cols, nnz, index, row_ind, col_ind, val); \ @@ -154,7 +154,7 @@ FOR_EACH_FP_AND_INT_TYPE(DEFINE_COO_MATRIX_FUNCS); // CSR matrix #define DEFINE_INIT_CSR_MATRIX_FUNCS(FP_TYPE, FP_SUFFIX, INT_TYPE, INT_SUFFIX) \ template <> \ - void init_csr_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, \ + void init_csr_matrix(sycl::queue& queue, matrix_handle_t* p_smhandle, std::int64_t num_rows, \ std::int64_t num_cols, std::int64_t nnz, index_base index, \ sycl::buffer row_ptr, sycl::buffer col_ind, \ sycl::buffer val) { \ @@ -163,15 +163,15 @@ FOR_EACH_FP_AND_INT_TYPE(DEFINE_COO_MATRIX_FUNCS); queue, p_smhandle, num_rows, num_cols, nnz, index, row_ptr, col_ind, val); \ } \ template <> \ - void init_csr_matrix(sycl::queue &queue, matrix_handle_t *p_smhandle, std::int64_t num_rows, \ + void init_csr_matrix(sycl::queue& queue, matrix_handle_t* p_smhandle, std::int64_t num_rows, \ std::int64_t num_cols, std::int64_t nnz, index_base index, \ - INT_TYPE *row_ptr, INT_TYPE *col_ind, FP_TYPE *val) { \ + INT_TYPE* row_ptr, INT_TYPE* col_ind, FP_TYPE* val) { \ auto libkey = get_device_id(queue); \ function_tables[{ libkey, queue }].init_csr_matrix_usm##FP_SUFFIX##INT_SUFFIX( \ queue, p_smhandle, num_rows, num_cols, nnz, index, row_ptr, col_ind, val); \ } \ template <> \ - void set_csr_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, \ + void set_csr_matrix_data(sycl::queue& queue, matrix_handle_t smhandle, std::int64_t num_rows, \ std::int64_t num_cols, std::int64_t nnz, index_base index, \ sycl::buffer row_ptr, sycl::buffer col_ind, \ sycl::buffer val) { \ @@ -180,9 +180,9 @@ FOR_EACH_FP_AND_INT_TYPE(DEFINE_COO_MATRIX_FUNCS); queue, smhandle, num_rows, num_cols, nnz, index, row_ptr, col_ind, val); \ } \ template <> \ - void set_csr_matrix_data(sycl::queue &queue, matrix_handle_t smhandle, std::int64_t num_rows, \ + void set_csr_matrix_data(sycl::queue& queue, matrix_handle_t smhandle, std::int64_t num_rows, \ std::int64_t num_cols, std::int64_t nnz, index_base index, \ - INT_TYPE *row_ptr, INT_TYPE *col_ind, FP_TYPE *val) { \ + INT_TYPE* row_ptr, INT_TYPE* col_ind, FP_TYPE* val) { \ auto libkey = get_device_id(queue); \ function_tables[{ libkey, queue }].set_csr_matrix_data_usm##FP_SUFFIX##INT_SUFFIX( \ queue, smhandle, num_rows, num_cols, nnz, index, row_ptr, col_ind, val); \ @@ -191,43 +191,43 @@ FOR_EACH_FP_AND_INT_TYPE(DEFINE_INIT_CSR_MATRIX_FUNCS); #undef DEFINE_INIT_CSR_MATRIX_FUNCS // Common sparse matrix functions -sycl::event release_sparse_matrix(sycl::queue &queue, matrix_handle_t smhandle, - const std::vector &dependencies) { +sycl::event release_sparse_matrix(sycl::queue& queue, matrix_handle_t smhandle, + const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].release_sparse_matrix(queue, smhandle, dependencies); } -bool set_matrix_property(sycl::queue &queue, matrix_handle_t smhandle, matrix_property property) { +bool set_matrix_property(sycl::queue& queue, matrix_handle_t smhandle, matrix_property property) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].set_matrix_property(queue, smhandle, property); } // SPMM -void init_spmm_descr(sycl::queue &queue, spmm_descr_t *p_spmm_descr) { +void init_spmm_descr(sycl::queue& queue, spmm_descr_t* p_spmm_descr) { auto libkey = get_device_id(queue); function_tables[{ libkey, queue }].init_spmm_descr(queue, p_spmm_descr); } -sycl::event release_spmm_descr(sycl::queue &queue, spmm_descr_t spmm_descr, - const std::vector &dependencies) { +sycl::event release_spmm_descr(sycl::queue& queue, spmm_descr_t spmm_descr, + const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].release_spmm_descr(queue, spmm_descr, dependencies); } -void spmm_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_matrix_handle_t B_handle, const void *beta, +void spmm_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_matrix_handle_t B_handle, const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, - std::size_t &temp_buffer_size) { + std::size_t& temp_buffer_size) { auto libkey = get_device_id(queue); function_tables[{ libkey, queue }].spmm_buffer_size(queue, opA, opB, alpha, A_view, A_handle, B_handle, beta, C_handle, alg, spmm_descr, temp_buffer_size); } -void spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, +void spmm_optimize(sycl::queue& queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_matrix_handle_t B_handle, const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, sycl::buffer workspace) { auto libkey = get_device_id(queue); function_tables[{ libkey, queue }].spmm_optimize_buffer(queue, opA, opB, alpha, A_view, @@ -235,23 +235,23 @@ void spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math spmm_descr, workspace); } -sycl::event spmm_optimize(sycl::queue &queue, oneapi::math::transpose opA, - oneapi::math::transpose opB, const void *alpha, matrix_view A_view, +sycl::event spmm_optimize(sycl::queue& queue, oneapi::math::transpose opA, + oneapi::math::transpose opB, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_matrix_handle_t B_handle, - const void *beta, dense_matrix_handle_t C_handle, spmm_alg alg, - spmm_descr_t spmm_descr, void *workspace, - const std::vector &dependencies) { + const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, + spmm_descr_t spmm_descr, void* workspace, + const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].spmm_optimize_usm( queue, opA, opB, alpha, A_view, A_handle, B_handle, beta, C_handle, alg, spmm_descr, workspace, dependencies); } -sycl::event spmm(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math::transpose opB, - const void *alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_matrix_handle_t B_handle, const void *beta, dense_matrix_handle_t C_handle, +sycl::event spmm(sycl::queue& queue, oneapi::math::transpose opA, oneapi::math::transpose opB, + const void* alpha, matrix_view A_view, matrix_handle_t A_handle, + dense_matrix_handle_t B_handle, const void* beta, dense_matrix_handle_t C_handle, spmm_alg alg, spmm_descr_t spmm_descr, - const std::vector &dependencies) { + const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].spmm(queue, opA, opB, alpha, A_view, A_handle, B_handle, beta, C_handle, alg, spmm_descr, @@ -259,78 +259,78 @@ sycl::event spmm(sycl::queue &queue, oneapi::math::transpose opA, oneapi::math:: } // SPMV -void init_spmv_descr(sycl::queue &queue, spmv_descr_t *p_spmv_descr) { +void init_spmv_descr(sycl::queue& queue, spmv_descr_t* p_spmv_descr) { auto libkey = get_device_id(queue); function_tables[{ libkey, queue }].init_spmv_descr(queue, p_spmv_descr); } -sycl::event release_spmv_descr(sycl::queue &queue, spmv_descr_t spmv_descr, - const std::vector &dependencies) { +sycl::event release_spmv_descr(sycl::queue& queue, spmv_descr_t spmv_descr, + const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].release_spmv_descr(queue, spmv_descr, dependencies); } -void spmv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spmv_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, - const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, - spmv_descr_t spmv_descr, std::size_t &temp_buffer_size) { + const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, + spmv_descr_t spmv_descr, std::size_t& temp_buffer_size) { auto libkey = get_device_id(queue); function_tables[{ libkey, queue }].spmv_buffer_size(queue, opA, alpha, A_view, A_handle, x_handle, beta, y_handle, alg, spmv_descr, temp_buffer_size); } -void spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spmv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, - const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, + const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, sycl::buffer workspace) { auto libkey = get_device_id(queue); function_tables[{ libkey, queue }].spmv_optimize_buffer( queue, opA, alpha, A_view, A_handle, x_handle, beta, y_handle, alg, spmv_descr, workspace); } -sycl::event spmv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spmv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, - dense_vector_handle_t x_handle, const void *beta, + dense_vector_handle_t x_handle, const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, spmv_descr_t spmv_descr, - void *workspace, const std::vector &dependencies) { + void* workspace, const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].spmv_optimize_usm( queue, opA, alpha, A_view, A_handle, x_handle, beta, y_handle, alg, spmv_descr, workspace, dependencies); } -sycl::event spmv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spmv(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, - const void *beta, dense_vector_handle_t y_handle, spmv_alg alg, - spmv_descr_t spmv_descr, const std::vector &dependencies) { + const void* beta, dense_vector_handle_t y_handle, spmv_alg alg, + spmv_descr_t spmv_descr, const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].spmv(queue, opA, alpha, A_view, A_handle, x_handle, beta, y_handle, alg, spmv_descr, dependencies); } // SPSV -void init_spsv_descr(sycl::queue &queue, spsv_descr_t *p_spsv_descr) { +void init_spsv_descr(sycl::queue& queue, spsv_descr_t* p_spsv_descr) { auto libkey = get_device_id(queue); function_tables[{ libkey, queue }].init_spsv_descr(queue, p_spsv_descr); } -sycl::event release_spsv_descr(sycl::queue &queue, spsv_descr_t spsv_descr, - const std::vector &dependencies) { +sycl::event release_spsv_descr(sycl::queue& queue, spsv_descr_t spsv_descr, + const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].release_spsv_descr(queue, spsv_descr, dependencies); } -void spsv_buffer_size(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spsv_buffer_size(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, - std::size_t &temp_buffer_size) { + std::size_t& temp_buffer_size) { auto libkey = get_device_id(queue); function_tables[{ libkey, queue }].spsv_buffer_size( queue, opA, alpha, A_view, A_handle, x_handle, y_handle, alg, spsv_descr, temp_buffer_size); } -void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +void spsv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, sycl::buffer workspace) { @@ -339,21 +339,21 @@ void spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void * queue, opA, alpha, A_view, A_handle, x_handle, y_handle, alg, spsv_descr, workspace); } -sycl::event spsv_optimize(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spsv_optimize(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, - spsv_alg alg, spsv_descr_t spsv_descr, void *workspace, - const std::vector &dependencies) { + spsv_alg alg, spsv_descr_t spsv_descr, void* workspace, + const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].spsv_optimize_usm(queue, opA, alpha, A_view, A_handle, x_handle, y_handle, alg, spsv_descr, workspace, dependencies); } -sycl::event spsv(sycl::queue &queue, oneapi::math::transpose opA, const void *alpha, +sycl::event spsv(sycl::queue& queue, oneapi::math::transpose opA, const void* alpha, matrix_view A_view, matrix_handle_t A_handle, dense_vector_handle_t x_handle, dense_vector_handle_t y_handle, spsv_alg alg, spsv_descr_t spsv_descr, - const std::vector &dependencies) { + const std::vector& dependencies) { auto libkey = get_device_id(queue); return function_tables[{ libkey, queue }].spsv(queue, opA, alpha, A_view, A_handle, x_handle, y_handle, alg, spsv_descr, dependencies); diff --git a/tests/unit_tests/blas/batch/axpy_batch_stride.cpp b/tests/unit_tests/blas/batch/axpy_batch_stride.cpp index c83dedd5e..708921e9e 100644 --- a/tests/unit_tests/blas/batch/axpy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/axpy_batch_stride.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, fp alpha, +int test(device* dev, oneapi::math::layout layout, int64_t incx, int64_t incy, fp alpha, int64_t batch_size) { // Prepare data. int64_t n, i; @@ -77,19 +77,19 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, f int batch_size_ref = (int)batch_size; for (i = 0; i < batch_size_ref; i++) { - ::axpy(&n_ref, (fp_ref *)&alpha, (fp_ref *)x.data() + i * stride_x, &incx_ref, - (fp_ref *)y_ref.data() + i * stride_y, &incy_ref); + ::axpy(&n_ref, (fp_ref*)&alpha, (fp_ref*)x.data() + i * stride_x, &incx_ref, + (fp_ref*)y_ref.data() + i * stride_y, &incy_ref); } // Call DPC++ AXPY_BATCH_STRIDE. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during AXPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -107,13 +107,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, f switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::axpy_batch(main_queue, n, alpha, x_buffer, incx, - stride_x, y_buffer, incy, stride_y, - batch_size); + stride_x, y_buffer, incy, stride_y, + batch_size); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::axpy_batch(main_queue, n, alpha, x_buffer, incx, - stride_x, y_buffer, incy, stride_y, - batch_size); + stride_x, y_buffer, incy, stride_y, + batch_size); break; default: break; } @@ -133,17 +133,17 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, f } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during AXPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of AXPY_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -160,7 +160,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, f } class AxpyBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpyBatchStrideTests, RealSinglePrecision) { float alpha = 2.0; diff --git a/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp index 5951bc0f5..fa5646ac9 100644 --- a/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/axpy_batch_stride_usm.cpp @@ -43,20 +43,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, fp alpha, +int test(device* dev, oneapi::math::layout layout, int64_t incx, int64_t incy, fp alpha, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during AXPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -101,8 +101,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, f int batch_size_ref = (int)batch_size; for (i = 0; i < batch_size_ref; i++) { - ::axpy(&n_ref, (fp_ref *)&alpha, (fp_ref *)x.data() + i * stride_x, &incx_ref, - (fp_ref *)y_ref.data() + i * stride_y, &incy_ref); + ::axpy(&n_ref, (fp_ref*)&alpha, (fp_ref*)x.data() + i * stride_x, &incx_ref, + (fp_ref*)y_ref.data() + i * stride_y, &incy_ref); } // Call DPC++ AXPY_BATCH_STRIDE. @@ -117,8 +117,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, f break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::axpy_batch(main_queue, n, alpha, &x[0], incx, - stride_x, &y[0], incy, stride_y, - batch_size, dependencies); + stride_x, &y[0], incy, stride_y, + batch_size, dependencies); break; default: break; } @@ -140,17 +140,17 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, f main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during AXPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of AXPY_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -166,7 +166,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, f } class AxpyBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpyBatchStrideUsmTests, RealSinglePrecision) { float alpha = 2.0; diff --git a/tests/unit_tests/blas/batch/axpy_batch_usm.cpp b/tests/unit_tests/blas/batch/axpy_batch_usm.cpp index bbb62a19c..bc70aef9d 100644 --- a/tests/unit_tests/blas/batch/axpy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/axpy_batch_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t group_count) { +int test(device* dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during AXPY_BATCH:\n" << e.what() << std::endl; print_error_code(e); @@ -69,15 +69,15 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { std::vector dependencies; // Prepare data. - int64_t *n = - (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); - int64_t *incx = - (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); - int64_t *incy = - (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); - fp *alpha = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * group_count, *dev, cxt); - int64_t *group_size = - (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + int64_t* n = + (int64_t*)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + int64_t* incx = + (int64_t*)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + int64_t* incy = + (int64_t*)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + fp* alpha = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * group_count, *dev, cxt); + int64_t* group_size = + (int64_t*)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); if ((n == NULL) || (incx == NULL) || (incy == NULL) || (alpha == NULL) || (group_size == NULL)) { @@ -104,12 +104,12 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { total_batch_count += group_size[i]; } - fp **x_array = - (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); - fp **y_array = - (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); - fp **y_ref_array = - (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); + fp** x_array = + (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * total_batch_count, *dev, cxt); + fp** y_array = + (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * total_batch_count, *dev, cxt); + fp** y_ref_array = + (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * total_batch_count, *dev, cxt); if ((x_array == NULL) || (y_array == NULL) || (y_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; @@ -124,11 +124,11 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { total_size_x = (1 + (n[i] - 1) * std::abs(incx[i])); total_size_y = (1 + (n[i] - 1) * std::abs(incy[i])); x_array[idx] = - (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_x, *dev, cxt); + (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_x, *dev, cxt); y_array[idx] = - (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); + (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); y_ref_array[idx] = - (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); + (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); rand_vector(x_array[idx], n[i], incx[i]); rand_vector(y_array[idx], n[i], incy[i]); copy_vector(y_array[idx], n[i], incy[i], y_ref_array[idx]); @@ -146,8 +146,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { n_ref = (int)n[i]; incx_ref = (int)incx[i]; incy_ref = (int)incy[i]; - ::axpy((const int *)&n_ref, (const fp_ref *)&alpha[i], (const fp_ref *)x_array[idx], - (const int *)&incx_ref, (fp_ref *)y_ref_array[idx], (const int *)&incy_ref); + ::axpy((const int*)&n_ref, (const fp_ref*)&alpha[i], (const fp_ref*)x_array[idx], + (const int*)&incx_ref, (fp_ref*)y_ref_array[idx], (const int*)&incy_ref); idx++; } } @@ -159,12 +159,12 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::axpy_batch( - main_queue, n, alpha, (const fp **)x_array, incx, y_array, incy, group_count, + main_queue, n, alpha, (const fp**)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::axpy_batch( - main_queue, n, alpha, (const fp **)x_array, incx, y_array, incy, group_count, + main_queue, n, alpha, (const fp**)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; default: break; @@ -174,12 +174,12 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { switch (layout) { case oneapi::math::layout::col_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpy_batch, n, - alpha, (const fp **)x_array, incx, y_array, incy, + alpha, (const fp**)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::axpy_batch, n, - alpha, (const fp **)x_array, incx, y_array, incy, + alpha, (const fp**)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; default: break; @@ -187,13 +187,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during AXPY_BATCH:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { @@ -214,7 +214,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of AXPY_BATCH:\n" << error.what() << std::endl; } @@ -252,7 +252,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } class AxpyBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpyBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/copy_batch_stride.cpp b/tests/unit_tests/blas/batch/copy_batch_stride.cpp index a0d933d2d..48bec1291 100644 --- a/tests/unit_tests/blas/batch/copy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/copy_batch_stride.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { // Prepare data. int64_t n, i; @@ -76,19 +76,19 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i int batch_size_ref = (int)batch_size; for (i = 0; i < batch_size_ref; i++) { - ::copy(&n_ref, (fp_ref *)x.data() + i * stride_x, &incx_ref, - (fp_ref *)y_ref.data() + i * stride_y, &incy_ref); + ::copy(&n_ref, (fp_ref*)x.data() + i * stride_x, &incx_ref, + (fp_ref*)y_ref.data() + i * stride_y, &incy_ref); } // Call DPC++ COPY_BATCH_STRIDE. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during COPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -105,12 +105,12 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::copy_batch(main_queue, n, x_buffer, incx, stride_x, - y_buffer, incy, stride_y, batch_size); + oneapi::math::blas::column_major::copy_batch( + main_queue, n, x_buffer, incx, stride_x, y_buffer, incy, stride_y, batch_size); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::copy_batch(main_queue, n, x_buffer, incx, stride_x, - y_buffer, incy, stride_y, batch_size); + y_buffer, incy, stride_y, batch_size); break; default: break; } @@ -130,17 +130,17 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during COPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of COPY_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -157,7 +157,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i } class CopyBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(CopyBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 2, 3, 15)); diff --git a/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp index e1502799f..291040948 100644 --- a/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/copy_batch_stride_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during COPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -100,8 +100,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i int batch_size_ref = (int)batch_size; for (i = 0; i < batch_size_ref; i++) { - ::copy(&n_ref, (fp_ref *)x.data() + i * stride_x, &incx_ref, - (fp_ref *)y_ref.data() + i * stride_y, &incy_ref); + ::copy(&n_ref, (fp_ref*)x.data() + i * stride_x, &incx_ref, + (fp_ref*)y_ref.data() + i * stride_y, &incy_ref); } // Call DPC++ COPY_BATCH_STRIDE. @@ -111,13 +111,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::copy_batch(main_queue, n, &x[0], incx, - stride_x, &y[0], incy, stride_y, - batch_size, dependencies); + stride_x, &y[0], incy, stride_y, + batch_size, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::copy_batch(main_queue, n, &x[0], incx, - stride_x, &y[0], incy, stride_y, - batch_size, dependencies); + stride_x, &y[0], incy, stride_y, + batch_size, dependencies); break; default: break; } @@ -139,17 +139,17 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during COPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of COPY_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -165,7 +165,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i } class CopyBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(CopyBatchStrideUsmTests, RealSinglePrecision) { float alpha = 2.0; diff --git a/tests/unit_tests/blas/batch/copy_batch_usm.cpp b/tests/unit_tests/blas/batch/copy_batch_usm.cpp index ad0b5003d..69f93b1a7 100644 --- a/tests/unit_tests/blas/batch/copy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/copy_batch_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t group_count) { +int test(device* dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during COPY_BATCH:\n" << e.what() << std::endl; print_error_code(e); @@ -69,14 +69,14 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { std::vector dependencies; // Prepare data. - int64_t *n = - (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); - int64_t *incx = - (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); - int64_t *incy = - (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); - int64_t *group_size = - (int64_t *)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + int64_t* n = + (int64_t*)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + int64_t* incx = + (int64_t*)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + int64_t* incy = + (int64_t*)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); + int64_t* group_size = + (int64_t*)oneapi::math::malloc_shared(64, sizeof(int64_t) * group_count, *dev, cxt); if ((n == NULL) || (incx == NULL) || (incy == NULL) || (group_size == NULL)) { std::cout << "Error cannot allocate input arrays\n"; @@ -100,12 +100,12 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { total_batch_count += group_size[i]; } - fp **x_array = - (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); - fp **y_array = - (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); - fp **y_ref_array = - (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * total_batch_count, *dev, cxt); + fp** x_array = + (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * total_batch_count, *dev, cxt); + fp** y_array = + (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * total_batch_count, *dev, cxt); + fp** y_ref_array = + (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * total_batch_count, *dev, cxt); if ((x_array == NULL) || (y_array == NULL) || (y_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; @@ -120,11 +120,11 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { total_size_x = (1 + (n[i] - 1) * std::abs(incx[i])); total_size_y = (1 + (n[i] - 1) * std::abs(incy[i])); x_array[idx] = - (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_x, *dev, cxt); + (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_x, *dev, cxt); y_array[idx] = - (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); + (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); y_ref_array[idx] = - (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); + (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * total_size_y, *dev, cxt); rand_vector(x_array[idx], n[i], incx[i]); rand_vector(y_array[idx], n[i], incy[i]); copy_vector(y_array[idx], n[i], incy[i], y_ref_array[idx]); @@ -142,8 +142,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { n_ref = (int)n[i]; incx_ref = (int)incx[i]; incy_ref = (int)incy[i]; - ::copy((const int *)&n_ref, (const fp_ref *)x_array[idx], (const int *)&incx_ref, - (fp_ref *)y_ref_array[idx], (const int *)&incy_ref); + ::copy((const int*)&n_ref, (const fp_ref*)x_array[idx], (const int*)&incx_ref, + (fp_ref*)y_ref_array[idx], (const int*)&incy_ref); idx++; } } @@ -155,13 +155,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::copy_batch( - main_queue, n, (const fp **)x_array, incx, y_array, incy, group_count, + main_queue, n, (const fp**)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; case oneapi::math::layout::row_major: - done = oneapi::math::blas::row_major::copy_batch(main_queue, n, (const fp **)x_array, - incx, y_array, incy, group_count, - group_size, dependencies); + done = oneapi::math::blas::row_major::copy_batch(main_queue, n, (const fp**)x_array, + incx, y_array, incy, group_count, + group_size, dependencies); break; default: break; } @@ -170,12 +170,12 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { switch (layout) { case oneapi::math::layout::col_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::copy_batch, n, - (const fp **)x_array, incx, y_array, incy, group_count, + (const fp**)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::copy_batch, n, - (const fp **)x_array, incx, y_array, incy, group_count, + (const fp**)x_array, incx, y_array, incy, group_count, group_size, dependencies); break; default: break; @@ -183,13 +183,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during COPY_BATCH:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { @@ -209,7 +209,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of COPY_BATCH:\n" << error.what() << std::endl; } @@ -246,7 +246,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } class CopyBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(CopyBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp b/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp index 658c57aa6..9fa01c52a 100644 --- a/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/dgmm_batch_stride.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::side left_right, int64_t incx, +int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right, int64_t incx, int64_t batch_size) { // Prepare data. int64_t m, n; @@ -90,21 +90,20 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::side left_right for (i = 0; i < batch_size_ref; i++) { ::dgmm(convert_to_cblas_layout(layout), convert_to_cblas_side(left_right), - (const int *)&m_ref, (const int *)&n_ref, (const fp_ref *)(A.data() + stride_a * i), - (const int *)&lda_ref, (const fp_ref *)(x.data() + stride_x * i), - (const int *)&incx_ref, (fp_ref *)(C_ref.data() + stride_c * i), - (const int *)&ldc_ref); + (const int*)&m_ref, (const int*)&n_ref, (const fp_ref*)(A.data() + stride_a * i), + (const int*)&lda_ref, (const fp_ref*)(x.data() + stride_x * i), + (const int*)&incx_ref, (fp_ref*)(C_ref.data() + stride_c * i), (const int*)&ldc_ref); } // Call DPC++ DGMM_BATCH_STRIDE. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during DGMM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -122,14 +121,14 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::side left_right #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::dgmm_batch(main_queue, left_right, m, n, A_buffer, - lda, stride_a, x_buffer, incx, stride_x, - C_buffer, ldc, stride_c, batch_size); + oneapi::math::blas::column_major::dgmm_batch( + main_queue, left_right, m, n, A_buffer, lda, stride_a, x_buffer, incx, stride_x, + C_buffer, ldc, stride_c, batch_size); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::dgmm_batch(main_queue, left_right, m, n, A_buffer, - lda, stride_a, x_buffer, incx, stride_x, - C_buffer, ldc, stride_c, batch_size); + lda, stride_a, x_buffer, incx, stride_x, + C_buffer, ldc, stride_c, batch_size); break; default: break; } @@ -149,17 +148,17 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::side left_right } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during DGMM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of DGMM_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -177,7 +176,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::side left_right } class DgmmBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(DgmmBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), diff --git a/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp index eca3eb6f1..c486ac90e 100644 --- a/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/dgmm_batch_stride_usm.cpp @@ -43,20 +43,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::side left_right, int64_t incx, +int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right, int64_t incx, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during DGMM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -115,10 +115,9 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::side left_right for (i = 0; i < batch_size_ref; i++) { ::dgmm(convert_to_cblas_layout(layout), convert_to_cblas_side(left_right), - (const int *)&m_ref, (const int *)&n_ref, (const fp_ref *)(A.data() + stride_a * i), - (const int *)&lda_ref, (const fp_ref *)(x.data() + stride_x * i), - (const int *)&incx_ref, (fp_ref *)(C_ref.data() + stride_c * i), - (const int *)&ldc_ref); + (const int*)&m_ref, (const int*)&n_ref, (const fp_ref*)(A.data() + stride_a * i), + (const int*)&lda_ref, (const fp_ref*)(x.data() + stride_x * i), + (const int*)&incx_ref, (fp_ref*)(C_ref.data() + stride_c * i), (const int*)&ldc_ref); } // Call DPC++ DGMM_BATCH_STRIDE. @@ -156,17 +155,17 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::side left_right main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during DGMM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of DGMM_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -182,7 +181,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::side left_right } class DgmmBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(DgmmBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), diff --git a/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp b/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp index 01c39cad6..3df3bffd2 100644 --- a/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/dgmm_batch_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t group_count) { +int test(device* dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during DGMM_BATCH:\n" << e.what() << std::endl; print_error_code(e); @@ -102,9 +102,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { total_batch_count += group_size[i]; } - auto uafpp = usm_allocator(cxt, *dev); - vector a_array(uafpp), x_array(uafpp), c_array(uafpp), - c_ref_array(uafpp); + auto uafpp = usm_allocator(cxt, *dev); + vector a_array(uafpp), x_array(uafpp), c_array(uafpp), c_ref_array(uafpp); a_array.resize(total_batch_count); x_array.resize(total_batch_count); c_array.resize(total_batch_count); @@ -117,13 +116,15 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { size_x = 1 + (x_len - 1) * std::abs(incx[i]); size_c = (layout == oneapi::math::layout::col_major) ? ldc[i] * n[i] : ldc[i] * m[i]; for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); - x_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_x, *dev, cxt); - c_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); - c_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); - rand_matrix(a_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], lda[i]); + a_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); + x_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_x, *dev, cxt); + c_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); + c_ref_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); + rand_matrix(a_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], + lda[i]); rand_vector(x_array[idx], x_len, incx[i]); - rand_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldc[i]); + rand_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], + ldc[i]); copy_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldc[i], c_ref_array[idx]); idx++; @@ -132,15 +133,15 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Call reference DGMM_BATCH. using fp_ref = typename ref_type_info::type; - int *m_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *n_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *lda_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *incx_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *ldc_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *group_size_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* m_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* n_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* lda_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* incx_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* ldc_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* group_size_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - CBLAS_SIDE *left_right_ref = - (CBLAS_SIDE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_SIDE) * group_count); + CBLAS_SIDE* left_right_ref = + (CBLAS_SIDE*)oneapi::math::aligned_alloc(64, sizeof(CBLAS_SIDE) * group_count); if ((m_ref == NULL) || (n_ref == NULL) || (lda_ref == NULL) || (incx_ref == NULL) || (ldc_ref == NULL) || (left_right_ref == NULL) || (group_size_ref == NULL)) { @@ -174,10 +175,10 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { ldc_ref[i] = (int)ldc[i]; group_size_ref[i] = (int)group_size[i]; for (j = 0; j < group_size_ref[i]; j++) { - ::dgmm(convert_to_cblas_layout(layout), left_right_ref[i], (const int *)&m_ref[i], - (const int *)&n_ref[i], (const fp_ref *)a_array[idx], (const int *)&lda_ref[i], - (const fp_ref *)x_array[idx], (const int *)&incx_ref[i], - (fp_ref *)c_ref_array[idx], (const int *)&ldc_ref[i]); + ::dgmm(convert_to_cblas_layout(layout), left_right_ref[i], (const int*)&m_ref[i], + (const int*)&n_ref[i], (const fp_ref*)a_array[idx], (const int*)&lda_ref[i], + (const fp_ref*)x_array[idx], (const int*)&incx_ref[i], (fp_ref*)c_ref_array[idx], + (const int*)&ldc_ref[i]); idx++; } } @@ -189,14 +190,14 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::dgmm_batch( - main_queue, &left_right[0], &m[0], &n[0], (const fp **)&a_array[0], &lda[0], - (const fp **)&x_array[0], &incx[0], &c_array[0], &ldc[0], group_count, + main_queue, &left_right[0], &m[0], &n[0], (const fp**)&a_array[0], &lda[0], + (const fp**)&x_array[0], &incx[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::dgmm_batch( - main_queue, &left_right[0], &m[0], &n[0], (const fp **)&a_array[0], &lda[0], - (const fp **)&x_array[0], &incx[0], &c_array[0], &ldc[0], group_count, + main_queue, &left_right[0], &m[0], &n[0], (const fp**)&a_array[0], &lda[0], + (const fp**)&x_array[0], &incx[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; default: break; @@ -206,14 +207,14 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { switch (layout) { case oneapi::math::layout::col_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::dgmm_batch, - &left_right[0], &m[0], &n[0], (const fp **)&a_array[0], - &lda[0], (const fp **)&x_array[0], &incx[0], &c_array[0], + &left_right[0], &m[0], &n[0], (const fp**)&a_array[0], + &lda[0], (const fp**)&x_array[0], &incx[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dgmm_batch, - &left_right[0], &m[0], &n[0], (const fp **)&a_array[0], - &lda[0], (const fp **)&x_array[0], &incx[0], &c_array[0], + &left_right[0], &m[0], &n[0], (const fp**)&a_array[0], + &lda[0], (const fp**)&x_array[0], &incx[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; default: break; @@ -221,13 +222,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during DGMM_BATCH:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { oneapi::math::aligned_free(m_ref); oneapi::math::aligned_free(n_ref); oneapi::math::aligned_free(lda_ref); @@ -248,7 +249,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of DGMM_BATCH:\n" << error.what() << std::endl; } @@ -285,7 +286,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } class DgmmBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(DgmmBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/gemm_batch_stride.cpp b/tests/unit_tests/blas/batch/gemm_batch_stride.cpp index 999910829..50e90ccbb 100644 --- a/tests/unit_tests/blas/batch/gemm_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/gemm_batch_stride.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n, k; int64_t lda, ldb, ldc; @@ -135,23 +135,22 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { for (i = 0; i < batch_size_ref; i++) { ::gemm(convert_to_cblas_layout(layout), convert_to_cblas_trans(transa), - convert_to_cblas_trans(transb), (const int *)&m_ref, (const int *)&n_ref, - (const int *)&k_ref, (const fp_ref *)&alpha, - (const fp_ref *)(A_ref.data() + stride_a * i), (const int *)&lda_ref, - (const fp_ref *)(B_ref.data() + stride_b * i), (const int *)&ldb_ref, - (const fp_ref *)&beta, (fp_ref *)(C_ref.data() + stride_c * i), - (const int *)&ldc_ref); + convert_to_cblas_trans(transb), (const int*)&m_ref, (const int*)&n_ref, + (const int*)&k_ref, (const fp_ref*)&alpha, + (const fp_ref*)(A_ref.data() + stride_a * i), (const int*)&lda_ref, + (const fp_ref*)(B_ref.data() + stride_b * i), (const int*)&ldb_ref, + (const fp_ref*)&beta, (fp_ref*)(C_ref.data() + stride_c * i), (const int*)&ldc_ref); } // Call DPC++ GEMM_BATCH_STRIDE. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GEMM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -199,17 +198,17 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { #endif main_queue.wait_and_throw(); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GEMM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GEMM_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -232,7 +231,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } class GemmBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemmBatchStrideTests, RealHalfPrecision) { EXPECT_TRUEORSKIP((test( diff --git a/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp index d556b81e1..1f46e1d68 100644 --- a/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/gemm_batch_stride_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GEMM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -136,10 +136,10 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { C_ref.resize(stride_c * batch_size); C_cast_ref.resize(stride_c * batch_size); - Ta **a_array = (Ta **)oneapi::math::malloc_shared(64, sizeof(Ta *) * batch_size, *dev, cxt); - Tb **b_array = (Tb **)oneapi::math::malloc_shared(64, sizeof(Tb *) * batch_size, *dev, cxt); - Tc **c_array = (Tc **)oneapi::math::malloc_shared(64, sizeof(Tc *) * batch_size, *dev, cxt); - Ts **c_ref_array = (Ts **)oneapi::math::malloc_shared(64, sizeof(Ts *) * batch_size, *dev, cxt); + Ta** a_array = (Ta**)oneapi::math::malloc_shared(64, sizeof(Ta*) * batch_size, *dev, cxt); + Tb** b_array = (Tb**)oneapi::math::malloc_shared(64, sizeof(Tb*) * batch_size, *dev, cxt); + Tc** c_array = (Tc**)oneapi::math::malloc_shared(64, sizeof(Tc*) * batch_size, *dev, cxt); + Ts** c_ref_array = (Ts**)oneapi::math::malloc_shared(64, sizeof(Ts*) * batch_size, *dev, cxt); if ((a_array == NULL) || (b_array == NULL) || (c_array == NULL) || (c_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; @@ -181,12 +181,11 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { int batch_size_ref = (int)batch_size; for (i = 0; i < batch_size_ref; i++) { ::gemm(convert_to_cblas_layout(layout), convert_to_cblas_trans(transa), - convert_to_cblas_trans(transb), (const int *)&m_ref, (const int *)&n_ref, - (const int *)&k_ref, (const fp_ref *)&alpha, - (const fp_ref *)(A_ref.data() + stride_a * i), (const int *)&lda_ref, - (const fp_ref *)(B_ref.data() + stride_b * i), (const int *)&ldb_ref, - (const fp_ref *)&beta, (fp_ref *)(C_ref.data() + stride_c * i), - (const int *)&ldc_ref); + convert_to_cblas_trans(transb), (const int*)&m_ref, (const int*)&n_ref, + (const int*)&k_ref, (const fp_ref*)&alpha, + (const fp_ref*)(A_ref.data() + stride_a * i), (const int*)&lda_ref, + (const fp_ref*)(B_ref.data() + stride_b * i), (const int*)&ldb_ref, + (const fp_ref*)&beta, (fp_ref*)(C_ref.data() + stride_c * i), (const int*)&ldc_ref); } // Call DPC++ GEMM_BATCH_STRIDE. @@ -226,13 +225,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { main_queue.wait_and_throw(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GEMM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { oneapi::math::free_shared(a_array, cxt); oneapi::math::free_shared(b_array, cxt); oneapi::math::free_shared(c_array, cxt); @@ -240,7 +239,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GEMM_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -267,7 +266,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } class GemmBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemmBatchStrideUsmTests, RealHalfPrecision) { EXPECT_TRUEORSKIP((test( diff --git a/tests/unit_tests/blas/batch/gemm_batch_usm.cpp b/tests/unit_tests/blas/batch/gemm_batch_usm.cpp index c02279b53..8c4fd6a37 100644 --- a/tests/unit_tests/blas/batch/gemm_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/gemm_batch_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t group_count) { +int test(device* dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GEMM_BATCH:\n" << e.what() << std::endl; print_error_code(e); @@ -126,14 +126,14 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { total_batch_count += group_size[i]; } - auto uaTap = usm_allocator(cxt, *dev); - auto uaTbp = usm_allocator(cxt, *dev); - auto uaTcp = usm_allocator(cxt, *dev); - auto uaTsp = usm_allocator(cxt, *dev); - vector a_array(uaTap); - vector b_array(uaTbp); - vector c_array(uaTcp), c_cast_ref_array(uaTcp); - vector a_ref_array(uaTsp), b_ref_array(uaTsp), c_ref_array(uaTsp); + auto uaTap = usm_allocator(cxt, *dev); + auto uaTbp = usm_allocator(cxt, *dev); + auto uaTcp = usm_allocator(cxt, *dev); + auto uaTsp = usm_allocator(cxt, *dev); + vector a_array(uaTap); + vector b_array(uaTbp); + vector c_array(uaTcp), c_cast_ref_array(uaTcp); + vector a_ref_array(uaTsp), b_ref_array(uaTsp), c_ref_array(uaTsp); a_array.resize(total_batch_count); b_array.resize(total_batch_count); c_array.resize(total_batch_count); @@ -158,17 +158,18 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { default: break; } for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (Ta *)oneapi::math::malloc_shared(64, sizeof(Ta) * size_a, *dev, cxt); - b_array[idx] = (Tb *)oneapi::math::malloc_shared(64, sizeof(Tb) * size_b, *dev, cxt); - c_array[idx] = (Tc *)oneapi::math::malloc_shared(64, sizeof(Tc) * size_c, *dev, cxt); - a_ref_array[idx] = (Ts *)oneapi::math::malloc_shared(64, sizeof(Ts) * size_a, *dev, cxt); - b_ref_array[idx] = (Ts *)oneapi::math::malloc_shared(64, sizeof(Ts) * size_b, *dev, cxt); + a_array[idx] = (Ta*)oneapi::math::malloc_shared(64, sizeof(Ta) * size_a, *dev, cxt); + b_array[idx] = (Tb*)oneapi::math::malloc_shared(64, sizeof(Tb) * size_b, *dev, cxt); + c_array[idx] = (Tc*)oneapi::math::malloc_shared(64, sizeof(Tc) * size_c, *dev, cxt); + a_ref_array[idx] = (Ts*)oneapi::math::malloc_shared(64, sizeof(Ts) * size_a, *dev, cxt); + b_ref_array[idx] = (Ts*)oneapi::math::malloc_shared(64, sizeof(Ts) * size_b, *dev, cxt); c_cast_ref_array[idx] = - (Tc *)oneapi::math::malloc_shared(64, sizeof(Tc) * size_c, *dev, cxt); - c_ref_array[idx] = (Ts *)oneapi::math::malloc_shared(64, sizeof(Ts) * size_c, *dev, cxt); + (Tc*)oneapi::math::malloc_shared(64, sizeof(Tc) * size_c, *dev, cxt); + c_ref_array[idx] = (Ts*)oneapi::math::malloc_shared(64, sizeof(Ts) * size_c, *dev, cxt); rand_matrix(a_array[idx], layout, transa[i], m[i], k[i], lda[i]); rand_matrix(b_array[idx], layout, transb[i], k[i], n[i], ldb[i]); - rand_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldc[i]); + rand_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], + ldc[i]); copy_matrix(a_array[idx], layout, transa[i], m[i], k[i], lda[i], a_ref_array[idx]); copy_matrix(b_array[idx], layout, transb[i], k[i], n[i], ldb[i], b_ref_array[idx]); copy_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldc[i], @@ -179,18 +180,18 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Call reference GEMM_BATCH. using fp_ref = typename ref_type_info::type; - int *m_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *n_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *k_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *lda_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *ldb_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *ldc_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *group_size_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - - CBLAS_TRANSPOSE *transa_ref = - (CBLAS_TRANSPOSE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); - CBLAS_TRANSPOSE *transb_ref = - (CBLAS_TRANSPOSE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); + int* m_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* n_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* k_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* lda_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* ldb_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* ldc_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* group_size_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + + CBLAS_TRANSPOSE* transa_ref = + (CBLAS_TRANSPOSE*)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); + CBLAS_TRANSPOSE* transb_ref = + (CBLAS_TRANSPOSE*)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); if ((m_ref == NULL) || (n_ref == NULL) || (k_ref == NULL) || (lda_ref == NULL) || (ldb_ref == NULL) || (ldc_ref == NULL) || (transa_ref == NULL) || (transb_ref == NULL) || @@ -233,11 +234,11 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { group_size_ref[i] = (int)group_size[i]; for (j = 0; j < group_size_ref[i]; j++) { ::gemm(convert_to_cblas_layout(layout), transa_ref[i], transb_ref[i], - (const int *)&m_ref[i], (const int *)&n_ref[i], (const int *)&k_ref[i], - (const fp_ref *)&alpha[i], (const fp_ref *)a_ref_array[idx], - (const int *)&lda_ref[i], (const fp_ref *)b_ref_array[idx], - (const int *)&ldb_ref[i], (const fp_ref *)&beta[i], (fp_ref *)c_ref_array[idx], - (const int *)&ldc_ref[i]); + (const int*)&m_ref[i], (const int*)&n_ref[i], (const int*)&k_ref[i], + (const fp_ref*)&alpha[i], (const fp_ref*)a_ref_array[idx], + (const int*)&lda_ref[i], (const fp_ref*)b_ref_array[idx], + (const int*)&ldb_ref[i], (const fp_ref*)&beta[i], (fp_ref*)c_ref_array[idx], + (const int*)&ldc_ref[i]); idx++; } } @@ -250,13 +251,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::gemm_batch( main_queue, &transa[0], &transb[0], &m[0], &n[0], &k[0], &alpha[0], - (const Ta **)&a_array[0], &lda[0], (const Tb **)&b_array[0], &ldb[0], &beta[0], + (const Ta**)&a_array[0], &lda[0], (const Tb**)&b_array[0], &ldb[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::gemm_batch( main_queue, &transa[0], &transb[0], &m[0], &n[0], &k[0], &alpha[0], - (const Ta **)&a_array[0], &lda[0], (const Tb **)&b_array[0], &ldb[0], &beta[0], + (const Ta**)&a_array[0], &lda[0], (const Tb**)&b_array[0], &ldb[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; default: break; @@ -267,14 +268,14 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { case oneapi::math::layout::col_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemm_batch, &transa[0], &transb[0], &m[0], &n[0], &k[0], &alpha[0], - (const Ta **)&a_array[0], &lda[0], (const Tb **)&b_array[0], + (const Ta**)&a_array[0], &lda[0], (const Tb**)&b_array[0], &ldb[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm_batch, &transa[0], &transb[0], &m[0], &n[0], &k[0], &alpha[0], - (const Ta **)&a_array[0], &lda[0], (const Ta **)&b_array[0], + (const Ta**)&a_array[0], &lda[0], (const Ta**)&b_array[0], &ldb[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; @@ -283,13 +284,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { main_queue.wait_and_throw(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GEMM_BATCH:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { oneapi::math::aligned_free(m_ref); oneapi::math::aligned_free(n_ref); oneapi::math::aligned_free(k_ref); @@ -315,7 +316,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GEMM_BATCH:\n" << error.what() << std::endl; } @@ -364,7 +365,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } class GemmBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemmBatchUsmTests, RealHalfPrecision) { EXPECT_TRUEORSKIP((test( diff --git a/tests/unit_tests/blas/batch/gemv_batch_stride.cpp b/tests/unit_tests/blas/batch/gemv_batch_stride.cpp index 8d4780a00..bed54f5e7 100644 --- a/tests/unit_tests/blas/batch/gemv_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/gemv_batch_stride.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { // Prepare data. int64_t m, n; int64_t lda; @@ -103,23 +103,22 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i int batch_size_ref = (int)batch_size; for (i = 0; i < batch_size_ref; i++) { - ::gemv(convert_to_cblas_layout(layout), convert_to_cblas_trans(transa), (const int *)&m_ref, - (const int *)&n_ref, (const fp_ref *)&alpha, - (const fp_ref *)(A.data() + stride_a * i), (const int *)&lda_ref, - (const fp_ref *)(x.data() + stride_x * i), (const int *)&incx_ref, - (const fp_ref *)&beta, (fp_ref *)(y_ref.data() + stride_y * i), - (const int *)&incy_ref); + ::gemv(convert_to_cblas_layout(layout), convert_to_cblas_trans(transa), (const int*)&m_ref, + (const int*)&n_ref, (const fp_ref*)&alpha, (const fp_ref*)(A.data() + stride_a * i), + (const int*)&lda_ref, (const fp_ref*)(x.data() + stride_x * i), + (const int*)&incx_ref, (const fp_ref*)&beta, (fp_ref*)(y_ref.data() + stride_y * i), + (const int*)&incy_ref); } // Call DPC++ GEMV_BATCH_STRIDE. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GEMV_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -164,17 +163,17 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GEMV_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GEMV_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -192,7 +191,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i } class GemvBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemvBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 2, 3, 5)); diff --git a/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp index 39eae0399..e009f1758 100644 --- a/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/gemv_batch_stride_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t incx, int64_t incy, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GEMM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -128,10 +128,10 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i int batch_size_ref = (int)batch_size; for (i = 0; i < batch_size_ref; i++) { - ::gemv(convert_to_cblas_layout(layout), convert_to_cblas_trans(transa), (const int *)&m_ref, - (const int *)&n_ref, (const fp_ref *)&alpha, (const fp_ref *)&A[stride_a * i], - (const int *)&lda_ref, (const fp_ref *)&x[stride_x * i], (const int *)&incx_ref, - (const fp_ref *)&beta, (fp_ref *)&y_ref[stride_y * i], (const int *)&incy_ref); + ::gemv(convert_to_cblas_layout(layout), convert_to_cblas_trans(transa), (const int*)&m_ref, + (const int*)&n_ref, (const fp_ref*)&alpha, (const fp_ref*)&A[stride_a * i], + (const int*)&lda_ref, (const fp_ref*)&x[stride_x * i], (const int*)&incx_ref, + (const fp_ref*)&beta, (fp_ref*)&y_ref[stride_y * i], (const int*)&incy_ref); } // Call DPC++ GEMV_BATCH_STRIDE. @@ -171,17 +171,17 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GEMV_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GEMV_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -197,7 +197,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t incx, int64_t incy, i } class GemvBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemvBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 2, 3, 5)); diff --git a/tests/unit_tests/blas/batch/gemv_batch_usm.cpp b/tests/unit_tests/blas/batch/gemv_batch_usm.cpp index 40fffe810..71925b26a 100644 --- a/tests/unit_tests/blas/batch/gemv_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/gemv_batch_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t group_count) { +int test(device* dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GEMV_BATCH:\n" << e.what() << std::endl; print_error_code(e); @@ -119,9 +119,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { total_batch_count += group_size[i]; } - auto uafpp = usm_allocator(cxt, *dev); - vector a_array(uafpp), x_array(uafpp), y_array(uafpp), - y_ref_array(uafpp); + auto uafpp = usm_allocator(cxt, *dev); + vector a_array(uafpp), x_array(uafpp), y_array(uafpp), y_ref_array(uafpp); a_array.resize(total_batch_count); x_array.resize(total_batch_count); y_array.resize(total_batch_count); @@ -135,11 +134,12 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { size_x = 1 + (x_len - 1) * std::abs(incx[i]); size_y = 1 + (y_len - 1) * std::abs(incy[i]); for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); - x_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_x, *dev, cxt); - y_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_y, *dev, cxt); - y_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_y, *dev, cxt); - rand_matrix(a_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], lda[i]); + a_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); + x_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_x, *dev, cxt); + y_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_y, *dev, cxt); + y_ref_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_y, *dev, cxt); + rand_matrix(a_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], + lda[i]); rand_vector(x_array[idx], x_len, incx[i]); rand_vector(y_array[idx], y_len, incy[i]); copy_vector(y_array[idx], y_len, incy[i], y_ref_array[idx]); @@ -149,15 +149,15 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Call reference GEMV_BATCH. using fp_ref = typename ref_type_info::type; - int *m_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *n_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *lda_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *incx_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *incy_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *group_size_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* m_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* n_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* lda_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* incx_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* incy_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* group_size_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - CBLAS_TRANSPOSE *transa_ref = - (CBLAS_TRANSPOSE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); + CBLAS_TRANSPOSE* transa_ref = + (CBLAS_TRANSPOSE*)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); if ((m_ref == NULL) || (n_ref == NULL) || (lda_ref == NULL) || (incx_ref == NULL) || (incy_ref == NULL) || (transa_ref == NULL) || (group_size_ref == NULL)) { @@ -191,11 +191,10 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { incy_ref[i] = (int)incy[i]; group_size_ref[i] = (int)group_size[i]; for (j = 0; j < group_size_ref[i]; j++) { - ::gemv(convert_to_cblas_layout(layout), transa_ref[i], (const int *)&m_ref[i], - (const int *)&n_ref[i], (const fp_ref *)&alpha[i], (const fp_ref *)a_array[idx], - (const int *)&lda_ref[i], (const fp_ref *)x_array[idx], - (const int *)&incx_ref[i], (const fp_ref *)&beta[i], (fp_ref *)y_ref_array[idx], - (const int *)&incy_ref[i]); + ::gemv(convert_to_cblas_layout(layout), transa_ref[i], (const int*)&m_ref[i], + (const int*)&n_ref[i], (const fp_ref*)&alpha[i], (const fp_ref*)a_array[idx], + (const int*)&lda_ref[i], (const fp_ref*)x_array[idx], (const int*)&incx_ref[i], + (const fp_ref*)&beta[i], (fp_ref*)y_ref_array[idx], (const int*)&incy_ref[i]); idx++; } } @@ -207,14 +206,14 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::gemv_batch( - main_queue, &transa[0], &m[0], &n[0], &alpha[0], (const fp **)&a_array[0], - &lda[0], (const fp **)&x_array[0], &incx[0], &beta[0], &y_array[0], &incy[0], + main_queue, &transa[0], &m[0], &n[0], &alpha[0], (const fp**)&a_array[0], + &lda[0], (const fp**)&x_array[0], &incx[0], &beta[0], &y_array[0], &incy[0], group_count, &group_size[0], dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::gemv_batch( - main_queue, &transa[0], &m[0], &n[0], &alpha[0], (const fp **)&a_array[0], - &lda[0], (const fp **)&x_array[0], &incx[0], &beta[0], &y_array[0], &incy[0], + main_queue, &transa[0], &m[0], &n[0], &alpha[0], (const fp**)&a_array[0], + &lda[0], (const fp**)&x_array[0], &incx[0], &beta[0], &y_array[0], &incy[0], group_count, &group_size[0], dependencies); break; default: break; @@ -225,29 +224,28 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { case oneapi::math::layout::col_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::gemv_batch, &transa[0], &m[0], &n[0], &alpha[0], - (const fp **)&a_array[0], &lda[0], (const fp **)&x_array[0], + (const fp**)&a_array[0], &lda[0], (const fp**)&x_array[0], &incx[0], &beta[0], &y_array[0], &incy[0], group_count, &group_size[0], dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemv_batch, - &transa[0], &m[0], &n[0], &alpha[0], - (const fp **)&a_array[0], &lda[0], (const fp **)&x_array[0], - &incx[0], &beta[0], &y_array[0], &incy[0], group_count, - &group_size[0], dependencies); + TEST_RUN_BLAS_CT_SELECT( + main_queue, oneapi::math::blas::row_major::gemv_batch, &transa[0], &m[0], &n[0], + &alpha[0], (const fp**)&a_array[0], &lda[0], (const fp**)&x_array[0], &incx[0], + &beta[0], &y_array[0], &incy[0], group_count, &group_size[0], dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GEMV_BATCH:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { oneapi::math::aligned_free(m_ref); oneapi::math::aligned_free(n_ref); oneapi::math::aligned_free(lda_ref); @@ -268,7 +266,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GEMV_BATCH:\n" << error.what() << std::endl; } @@ -306,7 +304,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } class GemvBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemvBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp b/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp index c28a4764e..7bb36b1ce 100644 --- a/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/imatcopy_batch_stride.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n; int64_t lda, ldb; @@ -101,11 +101,11 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during IMATCOPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -133,9 +133,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::imatcopy_batch, - trans, m, n, alpha, AB_buffer, lda, ldb, stride, - batch_size); + TEST_RUN_BLAS_CT_SELECT(main_queue, + oneapi::math::blas::column_major::imatcopy_batch, trans, m, + n, alpha, AB_buffer, lda, ldb, stride, batch_size); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::imatcopy_batch, @@ -146,17 +146,17 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during IMATCOPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of IMATCOPY_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -171,7 +171,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } class ImatcopyBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(ImatcopyBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp index 5abb649fb..e0bd7ead0 100644 --- a/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/imatcopy_batch_stride_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATCOPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -103,8 +103,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { AB.resize(stride * batch_size); AB_ref.resize(stride * batch_size); - fp **ab_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **ab_ref_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp** ab_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); + fp** ab_ref_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); if ((ab_array == NULL) || (ab_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; oneapi::math::free_shared(ab_array, cxt); @@ -143,8 +143,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::imatcopy_batch(main_queue, trans, m, n, alpha, - &AB[0], lda, ldb, stride, - batch_size, dependencies); + &AB[0], lda, ldb, stride, + batch_size, dependencies); break; default: break; } @@ -152,9 +152,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::imatcopy_batch, - trans, m, n, alpha, &AB[0], lda, ldb, stride, batch_size, - dependencies); + TEST_RUN_BLAS_CT_SELECT( + main_queue, oneapi::math::blas::column_major::imatcopy_batch, trans, m, n, + alpha, &AB[0], lda, ldb, stride, batch_size, dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::imatcopy_batch, @@ -166,19 +166,19 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during IMATCOPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { oneapi::math::free_shared(ab_array, cxt); oneapi::math::free_shared(ab_ref_array, cxt); return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of IMATCOPY_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -194,7 +194,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } class ImatcopyBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(ImatcopyBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp b/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp index 25dd4cd3f..5c2a38ada 100644 --- a/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/imatcopy_batch_usm.cpp @@ -44,19 +44,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t group_count) { +int test(device* dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during IMATCOPY_BATCH:\n" << e.what() << std::endl; print_error_code(e); @@ -103,8 +103,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { total_batch_count += group_size[i]; } - auto uafpp = usm_allocator(cxt, *dev); - vector ab_array(uafpp), ab_ref_array(uafpp); + auto uafpp = usm_allocator(cxt, *dev); + vector ab_array(uafpp), ab_ref_array(uafpp); ab_array.resize(total_batch_count); ab_ref_array.resize(total_batch_count); @@ -126,8 +126,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } size = std::max(size_a, size_b); for (j = 0; j < group_size[i]; j++) { - ab_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size, *dev, cxt); - ab_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size, *dev, cxt); + ab_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size, *dev, cxt); + ab_ref_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size, *dev, cxt); rand_matrix(ab_array[idx], oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size, 1, size); copy_matrix(ab_array[idx], oneapi::math::layout::col_major, @@ -171,10 +171,10 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::imatcopy_batch, - trans.data(), m.data(), n.data(), alpha.data(), - ab_array.data(), lda.data(), ldb.data(), group_count, - group_size.data(), dependencies); + TEST_RUN_BLAS_CT_SELECT( + main_queue, oneapi::math::blas::column_major::imatcopy_batch, trans.data(), + m.data(), n.data(), alpha.data(), ab_array.data(), lda.data(), ldb.data(), + group_count, group_size.data(), dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::imatcopy_batch, @@ -187,13 +187,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during IMATCOPY_BATCH:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { @@ -205,7 +205,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of IMATCOPY_BATCH:\n" << error.what() << std::endl; } @@ -229,9 +229,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } size = std::max(size_a, size_b); for (j = 0; j < group_size[i]; j++) { - good = good && - check_equal_matrix(ab_array[idx], ab_ref_array[idx], - oneapi::math::layout::col_major, size, 1, size, 10, std::cout); + good = good && check_equal_matrix(ab_array[idx], ab_ref_array[idx], + oneapi::math::layout::col_major, size, 1, size, 10, + std::cout); idx++; } } @@ -249,7 +249,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } class ImatcopyBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(ImatcopyBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp b/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp index ddb37e883..2da2de4ab 100644 --- a/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/omatadd_batch_stride.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n; int64_t lda, ldb, ldc; @@ -111,11 +111,11 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATADD_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -162,17 +162,17 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during OMATADD_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of OMATADD_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -187,7 +187,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } class OmataddBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmataddBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp index b01c2ec10..2f0deb8b4 100644 --- a/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/omatadd_batch_stride_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATADD_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -110,10 +110,10 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { C.resize(stride_c * batch_size); C_ref.resize(stride_c * batch_size); - fp **a_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **b_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **c_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **c_ref_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp** a_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); + fp** b_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); + fp** c_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); + fp** c_ref_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); if ((a_array == NULL) || (b_array == NULL) || (c_array == NULL) || (c_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; @@ -188,13 +188,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during OMATADD_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { oneapi::math::free_shared(a_array, cxt); oneapi::math::free_shared(b_array, cxt); oneapi::math::free_shared(c_array, cxt); @@ -202,7 +202,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of OMATADD_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -220,7 +220,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } class OmataddBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmataddBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp b/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp index 20320d51d..9ef8cbdb6 100644 --- a/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/omatcopy_batch_stride.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t batch_size) { // Prepare data. int64_t m, n; int64_t lda, ldb; @@ -101,11 +101,11 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATCOPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -123,22 +123,22 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::omatcopy_batch(main_queue, trans, m, n, alpha, - A_buffer, lda, stride_a, B_buffer, - ldb, stride_b, batch_size); + A_buffer, lda, stride_a, B_buffer, + ldb, stride_b, batch_size); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::omatcopy_batch(main_queue, trans, m, n, alpha, - A_buffer, lda, stride_a, B_buffer, ldb, - stride_b, batch_size); + A_buffer, lda, stride_a, B_buffer, + ldb, stride_b, batch_size); break; default: break; } #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatcopy_batch, - trans, m, n, alpha, A_buffer, lda, stride_a, B_buffer, ldb, - stride_b, batch_size); + TEST_RUN_BLAS_CT_SELECT( + main_queue, oneapi::math::blas::column_major::omatcopy_batch, trans, m, n, + alpha, A_buffer, lda, stride_a, B_buffer, ldb, stride_b, batch_size); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatcopy_batch, @@ -149,17 +149,17 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during OMATCOPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of OMATCOPY_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -174,7 +174,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } class OmatcopyBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmatcopyBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp index 72b7c8f1b..321c9b3bb 100644 --- a/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/omatcopy_batch_stride_usm.cpp @@ -44,19 +44,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATCOPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -105,9 +105,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { B.resize(stride_b * batch_size); B_ref.resize(stride_b * batch_size); - fp **a_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **b_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **b_ref_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp** a_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); + fp** b_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); + fp** b_ref_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); if ((a_array == NULL) || (b_array == NULL) || (b_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; @@ -161,9 +161,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatcopy_batch, - trans, m, n, alpha, &A[0], lda, stride_a, &B[0], ldb, - stride_b, batch_size, dependencies); + TEST_RUN_BLAS_CT_SELECT( + main_queue, oneapi::math::blas::column_major::omatcopy_batch, trans, m, n, + alpha, &A[0], lda, stride_a, &B[0], ldb, stride_b, batch_size, dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatcopy_batch, @@ -175,20 +175,20 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during OMATCOPY_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { oneapi::math::free_shared(a_array, cxt); oneapi::math::free_shared(b_array, cxt); oneapi::math::free_shared(b_ref_array, cxt); return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of OMATCOPY_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -205,7 +205,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } class OmatcopyBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmatcopyBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp b/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp index 190c38afe..cc8be6fe8 100644 --- a/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/omatcopy_batch_usm.cpp @@ -44,19 +44,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t group_count) { +int test(device* dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATCOPY_BATCH:\n" << e.what() << std::endl; print_error_code(e); @@ -103,8 +103,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { total_batch_count += group_size[i]; } - auto uafpp = usm_allocator(cxt, *dev); - vector a_array(uafpp), b_array(uafpp), b_ref_array(uafpp); + auto uafpp = usm_allocator(cxt, *dev); + vector a_array(uafpp), b_array(uafpp), b_ref_array(uafpp); a_array.resize(total_batch_count); b_array.resize(total_batch_count); @@ -126,9 +126,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { default: break; } for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); - b_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); - b_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); + a_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); + b_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); + b_ref_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); rand_matrix(a_array[idx], oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_a, 1, size_a); rand_matrix(b_array[idx], oneapi::math::layout::col_major, @@ -161,14 +161,14 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::omatcopy_batch( main_queue, trans.data(), m.data(), n.data(), alpha.data(), - (const fp **)a_array.data(), lda.data(), b_array.data(), ldb.data(), - group_count, group_size.data(), dependencies); + (const fp**)a_array.data(), lda.data(), b_array.data(), ldb.data(), group_count, + group_size.data(), dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::omatcopy_batch( main_queue, trans.data(), m.data(), n.data(), alpha.data(), - (const fp **)a_array.data(), lda.data(), b_array.data(), ldb.data(), - group_count, group_size.data(), dependencies); + (const fp**)a_array.data(), lda.data(), b_array.data(), ldb.data(), group_count, + group_size.data(), dependencies); break; default: break; } @@ -176,15 +176,15 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::omatcopy_batch, - trans.data(), m.data(), n.data(), alpha.data(), - (const fp **)a_array.data(), lda.data(), b_array.data(), - ldb.data(), group_count, group_size.data(), dependencies); + TEST_RUN_BLAS_CT_SELECT( + main_queue, oneapi::math::blas::column_major::omatcopy_batch, trans.data(), + m.data(), n.data(), alpha.data(), (const fp**)a_array.data(), lda.data(), + b_array.data(), ldb.data(), group_count, group_size.data(), dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::omatcopy_batch, trans.data(), m.data(), n.data(), alpha.data(), - (const fp **)a_array.data(), lda.data(), b_array.data(), + (const fp**)a_array.data(), lda.data(), b_array.data(), ldb.data(), group_count, group_size.data(), dependencies); break; default: break; @@ -192,13 +192,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during OMATCOPY_BATCH:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { idx = 0; for (i = 0; i < group_count; i++) { for (j = 0; j < group_size[i]; j++) { @@ -211,7 +211,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of OMATCOPY_BATCH:\n" << error.what() << std::endl; } @@ -235,8 +235,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } for (j = 0; j < group_size[i]; j++) { good = good && check_equal_matrix(b_array[idx], b_ref_array[idx], - oneapi::math::layout::col_major, size_b, 1, size_b, 10, - std::cout); + oneapi::math::layout::col_major, size_b, 1, size_b, + 10, std::cout); idx++; } } @@ -255,7 +255,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } class OmatcopyBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmatcopyBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/syrk_batch_stride.cpp b/tests/unit_tests/blas/batch/syrk_batch_stride.cpp index 3599d36f8..8470a9139 100644 --- a/tests/unit_tests/blas/batch/syrk_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/syrk_batch_stride.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t batch_size) { // Prepare data. int64_t n, k; int64_t lda, ldc; @@ -67,9 +67,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { upper_lower = (oneapi::math::uplo)(std::rand() % 2); if ((std::is_same::value) || (std::is_same::value)) { - trans = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans - : (std::rand() % 2) == 0 ? oneapi::math::transpose::trans - : oneapi::math::transpose::conjtrans; + trans = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans + : (std::rand() % 2) == 0 ? oneapi::math::transpose::trans + : oneapi::math::transpose::conjtrans; } else { trans = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans @@ -110,21 +110,21 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { for (i = 0; i < batch_size_ref; i++) { ::syrk(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), - convert_to_cblas_trans(trans), (const int *)&n_ref, (const int *)&k_ref, - (const fp_ref *)&alpha, (const fp_ref *)(A.data() + stride_a * i), - (const int *)&lda_ref, (const fp_ref *)&beta, - (fp_ref *)(C_ref.data() + stride_c * i), (const int *)&ldc_ref); + convert_to_cblas_trans(trans), (const int*)&n_ref, (const int*)&k_ref, + (const fp_ref*)&alpha, (const fp_ref*)(A.data() + stride_a * i), + (const int*)&lda_ref, (const fp_ref*)&beta, (fp_ref*)(C_ref.data() + stride_c * i), + (const int*)&ldc_ref); } // Call DPC++ SYRK_BATCH_STRIDE. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SYRK_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -142,13 +142,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::syrk_batch(main_queue, upper_lower, trans, n, k, - alpha, A_buffer, lda, stride_a, beta, - C_buffer, ldc, stride_c, batch_size); + alpha, A_buffer, lda, stride_a, beta, + C_buffer, ldc, stride_c, batch_size); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::syrk_batch(main_queue, upper_lower, trans, n, k, - alpha, A_buffer, lda, stride_a, beta, - C_buffer, ldc, stride_c, batch_size); + alpha, A_buffer, lda, stride_a, beta, + C_buffer, ldc, stride_c, batch_size); break; default: break; } @@ -168,17 +168,17 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SYRK_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SYRK_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -187,14 +187,14 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { auto C_accessor = C_buffer.get_host_access(read_only); bool good = - check_equal_matrix(C_accessor, C_ref, oneapi::math::layout::col_major, stride_c * batch_size, - 1, stride_c * batch_size, 10 * k, std::cout); + check_equal_matrix(C_accessor, C_ref, oneapi::math::layout::col_major, + stride_c * batch_size, 1, stride_c * batch_size, 10 * k, std::cout); return (int)good; } class SyrkBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SyrkBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp index 06aea5c4f..4736ae1a2 100644 --- a/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/syrk_batch_stride_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { +int test(device* dev, oneapi::math::layout layout, int64_t batch_size) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SYRK_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -86,9 +86,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { beta = rand_scalar(); upper_lower = (oneapi::math::uplo)(std::rand() % 2); if ((std::is_same::value) || (std::is_same::value)) { - trans = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans - : (std::rand() % 2) == 0 ? oneapi::math::transpose::trans - : oneapi::math::transpose::conjtrans; + trans = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans + : (std::rand() % 2) == 0 ? oneapi::math::transpose::trans + : oneapi::math::transpose::conjtrans; } else { trans = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans @@ -116,9 +116,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { C.resize(stride_c * batch_size); C_ref.resize(stride_c * batch_size); - fp **a_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **c_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); - fp **c_ref_array = (fp **)oneapi::math::malloc_shared(64, sizeof(fp *) * batch_size, *dev, cxt); + fp** a_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); + fp** c_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); + fp** c_ref_array = (fp**)oneapi::math::malloc_shared(64, sizeof(fp*) * batch_size, *dev, cxt); if ((a_array == NULL) || (c_array == NULL) || (c_ref_array == NULL)) { std::cout << "Error cannot allocate arrays of pointers\n"; @@ -150,10 +150,10 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { int batch_size_ref = (int)batch_size; for (i = 0; i < batch_size_ref; i++) { ::syrk(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), - convert_to_cblas_trans(trans), (const int *)&n_ref, (const int *)&k_ref, - (const fp_ref *)&alpha, (const fp_ref *)(A.data() + stride_a * i), - (const int *)&lda_ref, (const fp_ref *)&beta, - (fp_ref *)(C_ref.data() + stride_c * i), (const int *)&ldc_ref); + convert_to_cblas_trans(trans), (const int*)&n_ref, (const int*)&k_ref, + (const fp_ref*)&alpha, (const fp_ref*)(A.data() + stride_a * i), + (const int*)&lda_ref, (const fp_ref*)&beta, (fp_ref*)(C_ref.data() + stride_c * i), + (const int*)&ldc_ref); } // Call DPC++ SYRK_BATCH_STRIDE. @@ -191,20 +191,20 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SYRK_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { oneapi::math::free_shared(a_array, cxt); oneapi::math::free_shared(c_array, cxt); oneapi::math::free_shared(c_ref_array, cxt); return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SYRK_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -221,7 +221,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t batch_size) { } class SyrkBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SyrkBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/syrk_batch_usm.cpp b/tests/unit_tests/blas/batch/syrk_batch_usm.cpp index bf7358ca9..ef117619b 100644 --- a/tests/unit_tests/blas/batch/syrk_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/syrk_batch_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t group_count) { +int test(device* dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SYRK_BATCH:\n" << e.what() << std::endl; print_error_code(e); @@ -106,10 +106,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { beta[i] = rand_scalar(); upper_lower[i] = (oneapi::math::uplo)(std::rand() % 2); if ((std::is_same::value) || (std::is_same::value)) { - trans[i] = (std::rand() % 2) == 0 - ? oneapi::math::transpose::nontrans - : (std::rand() % 2) == 0 ? oneapi::math::transpose::trans - : oneapi::math::transpose::conjtrans; + trans[i] = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans + : (std::rand() % 2) == 0 ? oneapi::math::transpose::trans + : oneapi::math::transpose::conjtrans; } else { trans[i] = (std::rand() % 2) == 0 ? oneapi::math::transpose::nontrans @@ -118,8 +117,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { total_batch_count += group_size[i]; } - auto uafpp = usm_allocator(cxt, *dev); - vector a_array(uafpp), c_array(uafpp), c_ref_array(uafpp); + auto uafpp = usm_allocator(cxt, *dev); + vector a_array(uafpp), c_array(uafpp), c_ref_array(uafpp); a_array.resize(total_batch_count); c_array.resize(total_batch_count); c_ref_array.resize(total_batch_count); @@ -138,11 +137,12 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { default: break; } for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); - c_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); - c_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); + a_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); + c_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); + c_ref_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_c, *dev, cxt); rand_matrix(a_array[idx], layout, trans[i], n[i], k[i], lda[i]); - rand_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, n[i], n[i], ldc[i]); + rand_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, n[i], n[i], + ldc[i]); copy_matrix(c_array[idx], layout, oneapi::math::transpose::nontrans, n[i], n[i], ldc[i], c_ref_array[idx]); idx++; @@ -151,16 +151,16 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Call reference SYRK_BATCH. using fp_ref = typename ref_type_info::type; - int *n_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *k_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *lda_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *ldc_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *group_size_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* n_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* k_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* lda_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* ldc_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* group_size_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - CBLAS_UPLO *upper_lower_ref = - (CBLAS_UPLO *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_UPLO) * group_count); - CBLAS_TRANSPOSE *trans_ref = - (CBLAS_TRANSPOSE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); + CBLAS_UPLO* upper_lower_ref = + (CBLAS_UPLO*)oneapi::math::aligned_alloc(64, sizeof(CBLAS_UPLO) * group_count); + CBLAS_TRANSPOSE* trans_ref = + (CBLAS_TRANSPOSE*)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); if ((n_ref == NULL) || (k_ref == NULL) || (lda_ref == NULL) || (ldc_ref == NULL) || (trans_ref == NULL) || (upper_lower_ref == NULL) || (group_size_ref == NULL)) { @@ -194,9 +194,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { group_size_ref[i] = (int)group_size[i]; for (j = 0; j < group_size_ref[i]; j++) { ::syrk(convert_to_cblas_layout(layout), upper_lower_ref[i], trans_ref[i], - (const int *)&n_ref[i], (const int *)&k_ref[i], (const fp_ref *)&alpha[i], - (const fp_ref *)a_array[idx], (const int *)&lda_ref[i], (const fp_ref *)&beta[i], - (fp_ref *)c_ref_array[idx], (const int *)&ldc_ref[i]); + (const int*)&n_ref[i], (const int*)&k_ref[i], (const fp_ref*)&alpha[i], + (const fp_ref*)a_array[idx], (const int*)&lda_ref[i], (const fp_ref*)&beta[i], + (fp_ref*)c_ref_array[idx], (const int*)&ldc_ref[i]); idx++; } } @@ -209,13 +209,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::syrk_batch( main_queue, &upper_lower[0], &trans[0], &n[0], &k[0], &alpha[0], - (const fp **)&a_array[0], &lda[0], &beta[0], &c_array[0], &ldc[0], group_count, + (const fp**)&a_array[0], &lda[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::syrk_batch( main_queue, &upper_lower[0], &trans[0], &n[0], &k[0], &alpha[0], - (const fp **)&a_array[0], &lda[0], &beta[0], &c_array[0], &ldc[0], group_count, + (const fp**)&a_array[0], &lda[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; default: break; @@ -226,13 +226,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { case oneapi::math::layout::col_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::syrk_batch, &upper_lower[0], &trans[0], &n[0], &k[0], &alpha[0], - (const fp **)&a_array[0], &lda[0], &beta[0], &c_array[0], + (const fp**)&a_array[0], &lda[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syrk_batch, &upper_lower[0], &trans[0], &n[0], &k[0], &alpha[0], - (const fp **)&a_array[0], &lda[0], &beta[0], &c_array[0], + (const fp**)&a_array[0], &lda[0], &beta[0], &c_array[0], &ldc[0], group_count, &group_size[0], dependencies); break; default: break; @@ -240,13 +240,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SYRK_BATCH:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { oneapi::math::aligned_free(n_ref); oneapi::math::aligned_free(k_ref); oneapi::math::aligned_free(lda_ref); @@ -266,7 +266,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SYRK_BATCH:\n" << error.what() << std::endl; } @@ -301,7 +301,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } class SyrkBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SyrkBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/batch/trsm_batch_stride.cpp b/tests/unit_tests/blas/batch/trsm_batch_stride.cpp index 5ae26282d..37fd56886 100644 --- a/tests/unit_tests/blas/batch/trsm_batch_stride.cpp +++ b/tests/unit_tests/blas/batch/trsm_batch_stride.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb; @@ -116,21 +116,20 @@ int test(device *dev, oneapi::math::layout layout) { for (i = 0; i < batch_size_ref; i++) { ::trsm(convert_to_cblas_layout(layout), convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - convert_to_cblas_diag(unit_nonunit), (const int *)&m_ref, (const int *)&n_ref, - (const fp_ref *)&alpha, (const fp_ref *)(A.data() + stride_a * i), - (const int *)&lda_ref, (fp_ref *)(B_ref.data() + stride_b * i), - (const int *)&ldb_ref); + convert_to_cblas_diag(unit_nonunit), (const int*)&m_ref, (const int*)&n_ref, + (const fp_ref*)&alpha, (const fp_ref*)(A.data() + stride_a * i), + (const int*)&lda_ref, (fp_ref*)(B_ref.data() + stride_b * i), (const int*)&ldb_ref); } // Call DPC++ TRSM_BATCH_STRIDE. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during TRSM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -176,17 +175,17 @@ int test(device *dev, oneapi::math::layout layout) { } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during TRSM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of TRSM_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -201,7 +200,7 @@ int test(device *dev, oneapi::math::layout layout) { } class TrsmBatchStrideTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TrsmBatchStrideTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp b/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp index ec0f17572..0bc331cd5 100644 --- a/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp +++ b/tests/unit_tests/blas/batch/trsm_batch_stride_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during TRSM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); @@ -137,10 +137,9 @@ int test(device *dev, oneapi::math::layout layout) { for (i = 0; i < batch_size_ref; i++) { ::trsm(convert_to_cblas_layout(layout), convert_to_cblas_side(left_right), convert_to_cblas_uplo(upper_lower), convert_to_cblas_trans(trans), - convert_to_cblas_diag(unit_nonunit), (const int *)&m_ref, (const int *)&n_ref, - (const fp_ref *)&alpha, (const fp_ref *)(A.data() + stride_a * i), - (const int *)&lda_ref, (fp_ref *)(B_ref.data() + stride_b * i), - (const int *)&ldb_ref); + convert_to_cblas_diag(unit_nonunit), (const int*)&m_ref, (const int*)&n_ref, + (const fp_ref*)&alpha, (const fp_ref*)(A.data() + stride_a * i), + (const int*)&lda_ref, (fp_ref*)(B_ref.data() + stride_b * i), (const int*)&ldb_ref); } // Call DPC++ TRSM_BATCH_STRIDE. @@ -180,17 +179,17 @@ int test(device *dev, oneapi::math::layout layout) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during TRSM_BATCH_STRIDE:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of TRSM_BATCH_STRIDE:\n" << error.what() << std::endl; } @@ -203,7 +202,7 @@ int test(device *dev, oneapi::math::layout layout) { } class TrsmBatchStrideUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TrsmBatchStrideUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/batch/trsm_batch_usm.cpp b/tests/unit_tests/blas/batch/trsm_batch_usm.cpp index 8caeb9f14..cb4f06bbc 100644 --- a/tests/unit_tests/blas/batch/trsm_batch_usm.cpp +++ b/tests/unit_tests/blas/batch/trsm_batch_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int64_t group_count) { +int test(device* dev, oneapi::math::layout layout, int64_t group_count) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during TRSM_BATCH:\n" << e.what() << std::endl; print_error_code(e); @@ -128,8 +128,8 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { total_batch_count += group_size[i]; } - auto uafpp = usm_allocator(cxt, *dev); - vector a_array(uafpp), b_array(uafpp), b_ref_array(uafpp); + auto uafpp = usm_allocator(cxt, *dev); + vector a_array(uafpp), b_array(uafpp), b_ref_array(uafpp); a_array.resize(total_batch_count); b_array.resize(total_batch_count); @@ -141,11 +141,12 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { Arank = left_right[i] == oneapi::math::side::left ? m[i] : n[i]; size_b = ldb[i] * ((layout == oneapi::math::layout::col_major) ? n[i] : m[i]); for (j = 0; j < group_size[i]; j++) { - a_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); - b_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); - b_ref_array[idx] = (fp *)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); + a_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_a, *dev, cxt); + b_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); + b_ref_array[idx] = (fp*)oneapi::math::malloc_shared(64, sizeof(fp) * size_b, *dev, cxt); rand_trsm_matrix(a_array[idx], layout, trans[i], Arank, Arank, lda[i]); - rand_matrix(b_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldb[i]); + rand_matrix(b_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], + ldb[i]); copy_matrix(b_array[idx], layout, oneapi::math::transpose::nontrans, m[i], n[i], ldb[i], b_ref_array[idx]); idx++; @@ -154,20 +155,20 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { // Call reference TRSM_BATCH. using fp_ref = typename ref_type_info::type; - int *m_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *n_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *lda_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *ldb_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - int *group_size_ref = (int *)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); - - CBLAS_TRANSPOSE *trans_ref = - (CBLAS_TRANSPOSE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); - CBLAS_SIDE *left_right_ref = - (CBLAS_SIDE *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_SIDE) * group_count); - CBLAS_UPLO *upper_lower_ref = - (CBLAS_UPLO *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_UPLO) * group_count); - CBLAS_DIAG *unit_nonunit_ref = - (CBLAS_DIAG *)oneapi::math::aligned_alloc(64, sizeof(CBLAS_DIAG) * group_count); + int* m_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* n_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* lda_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* ldb_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + int* group_size_ref = (int*)oneapi::math::aligned_alloc(64, sizeof(int) * group_count); + + CBLAS_TRANSPOSE* trans_ref = + (CBLAS_TRANSPOSE*)oneapi::math::aligned_alloc(64, sizeof(CBLAS_TRANSPOSE) * group_count); + CBLAS_SIDE* left_right_ref = + (CBLAS_SIDE*)oneapi::math::aligned_alloc(64, sizeof(CBLAS_SIDE) * group_count); + CBLAS_UPLO* upper_lower_ref = + (CBLAS_UPLO*)oneapi::math::aligned_alloc(64, sizeof(CBLAS_UPLO) * group_count); + CBLAS_DIAG* unit_nonunit_ref = + (CBLAS_DIAG*)oneapi::math::aligned_alloc(64, sizeof(CBLAS_DIAG) * group_count); if ((m_ref == NULL) || (n_ref == NULL) || (lda_ref == NULL) || (ldb_ref == NULL) || (trans_ref == NULL) || (left_right_ref == NULL) || (upper_lower_ref == NULL) || @@ -206,9 +207,9 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { group_size_ref[i] = (int)group_size[i]; for (j = 0; j < group_size_ref[i]; j++) { ::trsm(convert_to_cblas_layout(layout), left_right_ref[i], upper_lower_ref[i], - trans_ref[i], unit_nonunit_ref[i], (const int *)&m_ref[i], - (const int *)&n_ref[i], (const fp_ref *)&alpha[i], (const fp_ref *)a_array[idx], - (const int *)&lda_ref[i], b_ref_array[idx], (const int *)&ldb_ref[i]); + trans_ref[i], unit_nonunit_ref[i], (const int*)&m_ref[i], (const int*)&n_ref[i], + (const fp_ref*)&alpha[i], (const fp_ref*)a_array[idx], (const int*)&lda_ref[i], + b_ref_array[idx], (const int*)&ldb_ref[i]); idx++; } } @@ -221,13 +222,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::trsm_batch( main_queue, &left_right[0], &upper_lower[0], &trans[0], &unit_nonunit[0], &m[0], - &n[0], &alpha[0], (const fp **)&a_array[0], &lda[0], &b_array[0], &ldb[0], + &n[0], &alpha[0], (const fp**)&a_array[0], &lda[0], &b_array[0], &ldb[0], group_count, &group_size[0], dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::trsm_batch( main_queue, &left_right[0], &upper_lower[0], &trans[0], &unit_nonunit[0], &m[0], - &n[0], &alpha[0], (const fp **)&a_array[0], &lda[0], &b_array[0], &ldb[0], + &n[0], &alpha[0], (const fp**)&a_array[0], &lda[0], &b_array[0], &ldb[0], group_count, &group_size[0], dependencies); break; default: break; @@ -239,14 +240,14 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::trsm_batch, &left_right[0], &upper_lower[0], &trans[0], &unit_nonunit[0], &m[0], &n[0], &alpha[0], - (const fp **)&a_array[0], &lda[0], &b_array[0], &ldb[0], + (const fp**)&a_array[0], &lda[0], &b_array[0], &ldb[0], group_count, &group_size[0], dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsm_batch, &left_right[0], &upper_lower[0], &trans[0], &unit_nonunit[0], &m[0], &n[0], &alpha[0], - (const fp **)&a_array[0], &lda[0], &b_array[0], &ldb[0], + (const fp**)&a_array[0], &lda[0], &b_array[0], &ldb[0], group_count, &group_size[0], dependencies); break; default: break; @@ -254,13 +255,13 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during TRSM_BATCH:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { oneapi::math::aligned_free(m_ref); oneapi::math::aligned_free(n_ref); oneapi::math::aligned_free(lda_ref); @@ -282,7 +283,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of TRSM_BATCH:\n" << error.what() << std::endl; } @@ -319,7 +320,7 @@ int test(device *dev, oneapi::math::layout layout, int64_t group_count) { } class TrsmBatchUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(TrsmBatchUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 5)); diff --git a/tests/unit_tests/blas/extensions/gemm_bias.cpp b/tests/unit_tests/blas/extensions/gemm_bias.cpp index 8ef864af7..012208587 100644 --- a/tests/unit_tests/blas/extensions/gemm_bias.cpp +++ b/tests/unit_tests/blas/extensions/gemm_bias.cpp @@ -63,11 +63,14 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose trans rand_matrix(B, layout, transb, k, n, ldb); rand_matrix(C, layout, oneapi::math::transpose::nontrans, m, n, ldc); if (offsetc == oneapi::math::offset::fix) - rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, 1, 1, 1); + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, 1, 1, + 1); if (offsetc == oneapi::math::offset::column) - rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, m, 1, m); + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, m, 1, + m); if (offsetc == oneapi::math::offset::row) - rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, n, 1, n); + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, n, 1, + n); C_ref = C; @@ -113,14 +116,14 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose trans #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::gemm_bias(main_queue, transa, transb, offsetc, m, - n, k, alpha, A_buffer, lda, ao, B_buffer, - ldb, bo, beta, C_buffer, ldc, CO_buffer); + oneapi::math::blas::column_major::gemm_bias( + main_queue, transa, transb, offsetc, m, n, k, alpha, A_buffer, lda, ao, + B_buffer, ldb, bo, beta, C_buffer, ldc, CO_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::gemm_bias(main_queue, transa, transb, offsetc, m, n, - k, alpha, A_buffer, lda, ao, B_buffer, ldb, - bo, beta, C_buffer, ldc, CO_buffer); + k, alpha, A_buffer, lda, ao, B_buffer, ldb, + bo, beta, C_buffer, ldc, CO_buffer); break; default: break; } @@ -132,8 +135,8 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose trans B_buffer, ldb, bo, beta, C_buffer, ldc, CO_buffer); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm_bias, transa, - transb, offsetc, m, n, k, alpha, A_buffer, lda, ao, + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm_bias, + transa, transb, offsetc, m, n, k, alpha, A_buffer, lda, ao, B_buffer, ldb, bo, beta, C_buffer, ldc, CO_buffer); break; default: break; diff --git a/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp b/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp index 19f68dc66..4c10aef2d 100644 --- a/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp +++ b/tests/unit_tests/blas/extensions/gemm_bias_usm.cpp @@ -85,11 +85,14 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose trans rand_matrix(B, layout, transb, k, n, ldb); rand_matrix(C, layout, oneapi::math::transpose::nontrans, m, n, ldc); if (offsetc == oneapi::math::offset::fix) - rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, 1, 1, 1); + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, 1, 1, + 1); if (offsetc == oneapi::math::offset::column) - rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, m, 1, m); + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, m, 1, + m); if (offsetc == oneapi::math::offset::row) - rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, n, 1, n); + rand_matrix(co, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, n, 1, + n); C_ref.resize(C.size()); for (int i = 0; i < C.size(); i++) @@ -137,8 +140,8 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose trans dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm_bias, transa, - transb, offsetc, m, n, k, alpha, A.data(), lda, ao, + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gemm_bias, + transa, transb, offsetc, m, n, k, alpha, A.data(), lda, ao, B.data(), ldb, bo, beta, C.data(), ldc, co.data(), dependencies); break; diff --git a/tests/unit_tests/blas/extensions/gemmt.cpp b/tests/unit_tests/blas/extensions/gemmt.cpp index fdcec77e6..f1fe54791 100644 --- a/tests/unit_tests/blas/extensions/gemmt.cpp +++ b/tests/unit_tests/blas/extensions/gemmt.cpp @@ -96,13 +96,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::gemmt(main_queue, upper_lower, transa, transb, n, - k, alpha, A_buffer, lda, B_buffer, ldb, beta, - C_buffer, ldc); + k, alpha, A_buffer, lda, B_buffer, ldb, + beta, C_buffer, ldc); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::gemmt(main_queue, upper_lower, transa, transb, n, k, - alpha, A_buffer, lda, B_buffer, ldb, beta, - C_buffer, ldc); + alpha, A_buffer, lda, B_buffer, ldb, beta, + C_buffer, ldc); break; default: break; } @@ -143,8 +143,8 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class GemmtTests : public ::testing::TestWithParam> { -}; +class GemmtTests + : public ::testing::TestWithParam> {}; TEST_P(GemmtTests, RealSinglePrecision) { float alpha(2.0); @@ -239,28 +239,28 @@ TEST_P(GemmtTests, ComplexSinglePrecision) { alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, - beta)); + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, + alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, @@ -275,28 +275,28 @@ TEST_P(GemmtTests, ComplexSinglePrecision) { alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, - beta)); + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, + alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); } TEST_P(GemmtTests, ComplexDoublePrecision) { @@ -318,28 +318,28 @@ TEST_P(GemmtTests, ComplexDoublePrecision) { alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, - beta)); + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, + alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, @@ -354,28 +354,28 @@ TEST_P(GemmtTests, ComplexDoublePrecision) { alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, - beta)); + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, + alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(GemmtTestSuite, GemmtTests, diff --git a/tests/unit_tests/blas/extensions/gemmt_usm.cpp b/tests/unit_tests/blas/extensions/gemmt_usm.cpp index 8a5db1e15..3674fe634 100644 --- a/tests/unit_tests/blas/extensions/gemmt_usm.cpp +++ b/tests/unit_tests/blas/extensions/gemmt_usm.cpp @@ -101,8 +101,8 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::gemmt(main_queue, upper_lower, transa, transb, - n, k, alpha, A.data(), lda, B.data(), - ldb, beta, C.data(), ldc, dependencies); + n, k, alpha, A.data(), lda, B.data(), + ldb, beta, C.data(), ldc, dependencies); break; default: break; } @@ -239,28 +239,28 @@ TEST_P(GemmtUsmTests, ComplexSinglePrecision) { alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, - beta)); + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, + alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, @@ -275,28 +275,28 @@ TEST_P(GemmtUsmTests, ComplexSinglePrecision) { alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, - beta)); + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, + alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); } TEST_P(GemmtUsmTests, ComplexDoublePrecision) { @@ -318,28 +318,28 @@ TEST_P(GemmtUsmTests, ComplexDoublePrecision) { alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, - beta)); + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, + alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, @@ -354,28 +354,28 @@ TEST_P(GemmtUsmTests, ComplexDoublePrecision) { alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, - beta)); + oneapi::math::transpose::trans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, + alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, 27, 98, 101, 102, + 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::transpose::trans, 27, 98, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, 103, - alpha, beta)); + oneapi::math::transpose::conjtrans, oneapi::math::transpose::conjtrans, 27, 98, 101, 102, + 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(GemmtUsmTestSuite, GemmtUsmTests, diff --git a/tests/unit_tests/blas/extensions/imatcopy.cpp b/tests/unit_tests/blas/extensions/imatcopy.cpp index 46335f012..f6480ee16 100644 --- a/tests/unit_tests/blas/extensions/imatcopy.cpp +++ b/tests/unit_tests/blas/extensions/imatcopy.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb; @@ -81,8 +81,8 @@ int test(device *dev, oneapi::math::layout layout) { rand_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size, 1, size); - copy_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size, 1, size, - AB_ref); + copy_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size, 1, + size, AB_ref); // Call reference IMATCOPY. int m_ref = (int)m; @@ -95,11 +95,11 @@ int test(device *dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during IMATCOPY:\n" << e.what() << std::endl; print_error_code(e); @@ -115,12 +115,12 @@ int test(device *dev, oneapi::math::layout layout) { #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::imatcopy(main_queue, trans, m, n, alpha, AB_buffer, - lda, ldb); + oneapi::math::blas::column_major::imatcopy(main_queue, trans, m, n, alpha, + AB_buffer, lda, ldb); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::imatcopy(main_queue, trans, m, n, alpha, AB_buffer, - lda, ldb); + lda, ldb); break; default: break; } @@ -138,17 +138,17 @@ int test(device *dev, oneapi::math::layout layout) { } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during IMATCOPY:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of IMATCOPY:\n" << error.what() << std::endl; } @@ -162,7 +162,7 @@ int test(device *dev, oneapi::math::layout layout) { } class ImatcopyTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(ImatcopyTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/extensions/imatcopy_usm.cpp b/tests/unit_tests/blas/extensions/imatcopy_usm.cpp index c247cd699..dd98e97fa 100644 --- a/tests/unit_tests/blas/extensions/imatcopy_usm.cpp +++ b/tests/unit_tests/blas/extensions/imatcopy_usm.cpp @@ -44,19 +44,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during IMATCOPY:\n" << e.what() << std::endl; print_error_code(e); @@ -105,8 +105,8 @@ int test(device *dev, oneapi::math::layout layout) { rand_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size, 1, size); - copy_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size, 1, size, - AB_ref); + copy_matrix(AB, oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size, 1, + size, AB_ref); // Call reference IMATCOPY. int m_ref = (int)m; @@ -121,11 +121,11 @@ int test(device *dev, oneapi::math::layout layout) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::imatcopy(main_queue, trans, m, n, alpha, - &AB[0], lda, ldb, dependencies); + &AB[0], lda, ldb, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::imatcopy(main_queue, trans, m, n, alpha, - &AB[0], lda, ldb, dependencies); + &AB[0], lda, ldb, dependencies); break; default: break; } @@ -145,17 +145,17 @@ int test(device *dev, oneapi::math::layout layout) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during IMATCOPY:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of IMATCOPY:\n" << error.what() << std::endl; } @@ -167,7 +167,7 @@ int test(device *dev, oneapi::math::layout layout) { } class ImatcopyUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(ImatcopyUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/extensions/omatadd.cpp b/tests/unit_tests/blas/extensions/omatadd.cpp index bf0a9ac23..cb404cc34 100644 --- a/tests/unit_tests/blas/extensions/omatadd.cpp +++ b/tests/unit_tests/blas/extensions/omatadd.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb, ldc; @@ -84,14 +84,14 @@ int test(device *dev, oneapi::math::layout layout) { vector> A(size_a), B(size_b), C(size_c), C_ref(size_c); - rand_matrix(A.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_a, - 1, size_a); - rand_matrix(B.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_b, - 1, size_b); - rand_matrix(C.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_c, - 1, size_c); - copy_matrix(C.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_c, - 1, size_c, C_ref.data()); + rand_matrix(A.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, + size_a, 1, size_a); + rand_matrix(B.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, + size_b, 1, size_b); + rand_matrix(C.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, + size_c, 1, size_c); + copy_matrix(C.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, + size_c, 1, size_c, C_ref.data()); // Call reference OMATADD. int m_ref = (int)m; @@ -106,11 +106,11 @@ int test(device *dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATADD:\n" << e.what() << std::endl; print_error_code(e); @@ -129,13 +129,13 @@ int test(device *dev, oneapi::math::layout layout) { switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::omatadd(main_queue, transa, transb, m, n, alpha, - A_buffer, lda, beta, B_buffer, ldb, - C_buffer, ldc); + A_buffer, lda, beta, B_buffer, ldb, + C_buffer, ldc); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::omatadd(main_queue, transa, transb, m, n, alpha, - A_buffer, lda, beta, B_buffer, ldb, C_buffer, - ldc); + A_buffer, lda, beta, B_buffer, ldb, C_buffer, + ldc); break; default: break; } @@ -155,16 +155,16 @@ int test(device *dev, oneapi::math::layout layout) { } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during OMATADD:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of OMATADD:\n" << error.what() << std::endl; } @@ -178,7 +178,7 @@ int test(device *dev, oneapi::math::layout layout) { } class OmataddTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmataddTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/extensions/omatadd_usm.cpp b/tests/unit_tests/blas/extensions/omatadd_usm.cpp index cb9b45d4a..472c693c8 100644 --- a/tests/unit_tests/blas/extensions/omatadd_usm.cpp +++ b/tests/unit_tests/blas/extensions/omatadd_usm.cpp @@ -43,19 +43,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATADD:\n" << e.what() << std::endl; print_error_code(e); @@ -133,13 +133,13 @@ int test(device *dev, oneapi::math::layout layout) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::omatadd(main_queue, transa, transb, m, n, - alpha, &A[0], lda, beta, &B[0], ldb, - &C[0], ldc, dependencies); + alpha, &A[0], lda, beta, &B[0], + ldb, &C[0], ldc, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::omatadd(main_queue, transa, transb, m, n, - alpha, &A[0], lda, beta, &B[0], ldb, - &C[0], ldc, dependencies); + alpha, &A[0], lda, beta, &B[0], ldb, + &C[0], ldc, dependencies); break; default: break; } @@ -161,16 +161,16 @@ int test(device *dev, oneapi::math::layout layout) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during OMATADD:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of OMATADD:\n" << error.what() << std::endl; } @@ -182,7 +182,7 @@ int test(device *dev, oneapi::math::layout layout) { } class OmataddUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmataddUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/extensions/omatcopy.cpp b/tests/unit_tests/blas/extensions/omatcopy.cpp index f77de83f1..27aeb0739 100644 --- a/tests/unit_tests/blas/extensions/omatcopy.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb; @@ -103,11 +103,11 @@ int test(device *dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATCOPY:\n" << e.what() << std::endl; print_error_code(e); @@ -125,11 +125,11 @@ int test(device *dev, oneapi::math::layout layout) { switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::omatcopy(main_queue, trans, m, n, alpha, A_buffer, - lda, B_buffer, ldb); + lda, B_buffer, ldb); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::omatcopy(main_queue, trans, m, n, alpha, A_buffer, - lda, B_buffer, ldb); + lda, B_buffer, ldb); break; default: break; } @@ -147,17 +147,17 @@ int test(device *dev, oneapi::math::layout layout) { } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during OMATCOPY:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of OMATCOPY:\n" << error.what() << std::endl; } @@ -171,7 +171,7 @@ int test(device *dev, oneapi::math::layout layout) { } class OmatcopyTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmatcopyTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/extensions/omatcopy2.cpp b/tests/unit_tests/blas/extensions/omatcopy2.cpp index 41b4b6e7c..1829f7e77 100644 --- a/tests/unit_tests/blas/extensions/omatcopy2.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy2.cpp @@ -43,12 +43,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Prepare data. int64_t m, n; int64_t lda, ldb; @@ -83,8 +83,8 @@ int test(device *dev, oneapi::math::layout layout) { rand_matrix(A.data(), layout, oneapi::math::transpose::nontrans, m, n, lda); rand_matrix(B.data(), layout, trans, m, n, ldb); - copy_matrix(B.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, size_b, - 1, size_b, B_ref.data()); + copy_matrix(B.data(), oneapi::math::layout::col_major, oneapi::math::transpose::nontrans, + size_b, 1, size_b, B_ref.data()); // Call reference OMATCOPY2. int64_t m_ref = m; @@ -100,11 +100,11 @@ int test(device *dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATCOPY2:\n" << e.what() << std::endl; print_error_code(e); @@ -121,12 +121,13 @@ int test(device *dev, oneapi::math::layout layout) { #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::omatcopy2(main_queue, trans, m, n, alpha, A_buffer, - lda, stride_a, B_buffer, ldb, stride_b); + oneapi::math::blas::column_major::omatcopy2(main_queue, trans, m, n, alpha, + A_buffer, lda, stride_a, B_buffer, ldb, + stride_b); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::omatcopy2(main_queue, trans, m, n, alpha, A_buffer, - lda, stride_a, B_buffer, ldb, stride_b); + lda, stride_a, B_buffer, ldb, stride_b); break; default: break; } @@ -146,17 +147,17 @@ int test(device *dev, oneapi::math::layout layout) { } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during OMATCOPY2:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of OMATCOPY2:\n" << error.what() << std::endl; } @@ -170,7 +171,7 @@ int test(device *dev, oneapi::math::layout layout) { } class Omatcopy2Tests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Omatcopy2Tests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp b/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp index cdcc21020..59a8955df 100644 --- a/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy2_usm.cpp @@ -44,19 +44,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATCOPY2:\n" << e.what() << std::endl; print_error_code(e); @@ -129,13 +129,13 @@ int test(device *dev, oneapi::math::layout layout) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::omatcopy2(main_queue, trans, m, n, alpha, - &A[0], lda, stride_a, &B[0], ldb, - stride_b, dependencies); + &A[0], lda, stride_a, &B[0], ldb, + stride_b, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::omatcopy2(main_queue, trans, m, n, alpha, - &A[0], lda, stride_a, &B[0], ldb, - stride_b, dependencies); + &A[0], lda, stride_a, &B[0], ldb, + stride_b, dependencies); break; default: break; } @@ -157,17 +157,17 @@ int test(device *dev, oneapi::math::layout layout) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during OMATCOPY2:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of OMATCOPY2:\n" << error.what() << std::endl; } @@ -179,7 +179,7 @@ int test(device *dev, oneapi::math::layout layout) { } class Omatcopy2UsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Omatcopy2UsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/extensions/omatcopy_usm.cpp b/tests/unit_tests/blas/extensions/omatcopy_usm.cpp index 7ba026ab3..d3b6be7fe 100644 --- a/tests/unit_tests/blas/extensions/omatcopy_usm.cpp +++ b/tests/unit_tests/blas/extensions/omatcopy_usm.cpp @@ -44,19 +44,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during OMATCOPY:\n" << e.what() << std::endl; print_error_code(e); @@ -126,8 +126,8 @@ int test(device *dev, oneapi::math::layout layout) { main_queue, trans, m, n, alpha, &A[0], lda, &B[0], ldb, dependencies); break; case oneapi::math::layout::row_major: - done = oneapi::math::blas::row_major::omatcopy(main_queue, trans, m, n, alpha, &A[0], - lda, &B[0], ldb, dependencies); + done = oneapi::math::blas::row_major::omatcopy( + main_queue, trans, m, n, alpha, &A[0], lda, &B[0], ldb, dependencies); break; default: break; } @@ -147,17 +147,17 @@ int test(device *dev, oneapi::math::layout layout) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during OMATCOPY:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of OMATCOPY:\n" << error.what() << std::endl; } @@ -169,7 +169,7 @@ int test(device *dev, oneapi::math::layout layout) { } class OmatcopyUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(OmatcopyUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/include/onemath_blas_helper.hpp b/tests/unit_tests/blas/include/onemath_blas_helper.hpp index f5944fe92..f258e55b9 100644 --- a/tests/unit_tests/blas/include/onemath_blas_helper.hpp +++ b/tests/unit_tests/blas/include/onemath_blas_helper.hpp @@ -63,7 +63,7 @@ inline CBLAS_OFFSET convert_to_cblas_offset(oneapi::math::offset offsetc) { inline CBLAS_LAYOUT convert_to_cblas_layout(oneapi::math::layout is_column) { return is_column == oneapi::math::layout::col_major ? CBLAS_LAYOUT::CblasColMajor - : CBLAS_LAYOUT::CblasRowMajor; + : CBLAS_LAYOUT::CblasRowMajor; } static const CBLAS_TRANSPOSE fcblastrans[] = { CblasNoTrans, CblasTrans, CblasConjTrans }; diff --git a/tests/unit_tests/blas/include/reference_blas_templates.hpp b/tests/unit_tests/blas/include/reference_blas_templates.hpp index b84e71c1b..45263cc63 100644 --- a/tests/unit_tests/blas/include/reference_blas_templates.hpp +++ b/tests/unit_tests/blas/include/reference_blas_templates.hpp @@ -33,8 +33,8 @@ inline bool isNonTranspose(CBLAS_TRANSPOSE trans) { } template -static inline void copy_mat(T_src &src, CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, int row, - int col, int ld, T_dest *&dest) { +static inline void copy_mat(T_src& src, CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, int row, + int col, int ld, T_dest*& dest) { int i, j, Iend, Jend; if (layout == CblasColMajor) { Jend = isNonTranspose(trans) ? col : row; @@ -53,8 +53,8 @@ static inline void copy_mat(T_src &src, CBLAS_LAYOUT layout, CBLAS_TRANSPOSE tra } template -static inline void copy_mat(T_src &src, CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, int row, - int col, int ld, T_dest off, T_dest *&dest) { +static inline void copy_mat(T_src& src, CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, int row, + int col, int ld, T_dest off, T_dest*& dest) { int i, j, Iend, Jend; if (layout == CblasColMajor) { Jend = isNonTranspose(trans) ? col : row; @@ -73,8 +73,8 @@ static inline void copy_mat(T_src &src, CBLAS_LAYOUT layout, CBLAS_TRANSPOSE tra } template -static inline void copy_mat(T_src &src, CBLAS_LAYOUT layout, int row, int col, int ld, - CBLAS_OFFSET off_kind, T_off off, T_dest &dest) { +static inline void copy_mat(T_src& src, CBLAS_LAYOUT layout, int row, int col, int ld, + CBLAS_OFFSET off_kind, T_off off, T_dest& dest) { using T_data = typename std::remove_reference::type; int i, j; T_data tmp; @@ -110,8 +110,8 @@ static inline void copy_mat(T_src &src, CBLAS_LAYOUT layout, int row, int col, i } template -static inline void update_c(T_src &src, CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, int row, - int col, int ld, T_desc *&dest) { +static inline void update_c(T_src& src, CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, int row, + int col, int ld, T_desc*& dest) { int i, j; int Jend = (layout == CblasColMajor) ? col : row; @@ -139,15 +139,15 @@ static inline void update_c(T_src &src, CBLAS_LAYOUT layout, CBLAS_UPLO upper_lo /* Level 3 */ template -static void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int *m, - const int *n, const int *k, const fp *alpha, const fp *a, const int *lda, - const fp *b, const int *ldb, const fp *beta, fp *c, const int *ldc); +static void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int* m, + const int* n, const int* k, const fp* alpha, const fp* a, const int* lda, + const fp* b, const int* ldb, const fp* beta, fp* c, const int* ldc); template <> -void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int *m, - const int *n, const int *k, const sycl::half *alpha, const sycl::half *a, const int *lda, - const sycl::half *b, const int *ldb, const sycl::half *beta, sycl::half *c, - const int *ldc) { +void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int* m, + const int* n, const int* k, const sycl::half* alpha, const sycl::half* a, const int* lda, + const sycl::half* b, const int* ldb, const sycl::half* beta, sycl::half* c, + const int* ldc) { // Not supported in NETLIB. SGEMM is used as reference. int sizea, sizeb, sizec; const float alphaf = *alpha; @@ -162,9 +162,9 @@ void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, c sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; sizec = *ldc * *m; } - float *af = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizea); - float *bf = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizeb); - float *cf = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizec); + float* af = (float*)oneapi::math::aligned_alloc(64, sizeof(float) * sizea); + float* bf = (float*)oneapi::math::aligned_alloc(64, sizeof(float) * sizeb); + float* cf = (float*)oneapi::math::aligned_alloc(64, sizeof(float) * sizec); copy_mat(a, layout, transa, *m, *k, *lda, af); copy_mat(b, layout, transb, *k, *n, *ldb, bf); copy_mat(c, layout, CblasNoTrans, *m, *n, *ldc, cf); @@ -177,49 +177,49 @@ void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, c } template <> -void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int *m, - const int *n, const int *k, const float *alpha, const float *a, const int *lda, - const float *b, const int *ldb, const float *beta, float *c, const int *ldc) { +void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int* m, + const int* n, const int* k, const float* alpha, const float* a, const int* lda, + const float* b, const int* ldb, const float* beta, float* c, const int* ldc) { cblas_sgemm_wrapper(layout, transa, transb, *m, *n, *k, *alpha, a, *lda, b, *ldb, *beta, c, *ldc); } template <> -void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int *m, - const int *n, const int *k, const double *alpha, const double *a, const int *lda, - const double *b, const int *ldb, const double *beta, double *c, const int *ldc) { +void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int* m, + const int* n, const int* k, const double* alpha, const double* a, const int* lda, + const double* b, const int* ldb, const double* beta, double* c, const int* ldc) { cblas_dgemm_wrapper(layout, transa, transb, *m, *n, *k, *alpha, a, *lda, b, *ldb, *beta, c, *ldc); } template <> -void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int *m, - const int *n, const int *k, const std::complex *alpha, - const std::complex *a, const int *lda, const std::complex *b, - const int *ldb, const std::complex *beta, std::complex *c, const int *ldc) { - cblas_cgemm_wrapper(layout, transa, transb, *m, *n, *k, (const void *)alpha, (const void *)a, - *lda, (const void *)b, *ldb, (const void *)beta, (void *)c, *ldc); +void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int* m, + const int* n, const int* k, const std::complex* alpha, + const std::complex* a, const int* lda, const std::complex* b, + const int* ldb, const std::complex* beta, std::complex* c, const int* ldc) { + cblas_cgemm_wrapper(layout, transa, transb, *m, *n, *k, (const void*)alpha, (const void*)a, + *lda, (const void*)b, *ldb, (const void*)beta, (void*)c, *ldc); } template <> -void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int *m, - const int *n, const int *k, const std::complex *alpha, - const std::complex *a, const int *lda, const std::complex *b, - const int *ldb, const std::complex *beta, std::complex *c, - const int *ldc) { - cblas_zgemm_wrapper(layout, transa, transb, *m, *n, *k, (const void *)alpha, (const void *)a, - *lda, (const void *)b, *ldb, (const void *)beta, (void *)c, *ldc); +void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int* m, + const int* n, const int* k, const std::complex* alpha, + const std::complex* a, const int* lda, const std::complex* b, + const int* ldb, const std::complex* beta, std::complex* c, + const int* ldc) { + cblas_zgemm_wrapper(layout, transa, transb, *m, *n, *k, (const void*)alpha, (const void*)a, + *lda, (const void*)b, *ldb, (const void*)beta, (void*)c, *ldc); } template -static void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int *m, - const int *n, const int *k, const fpc *alpha, const fpa *a, const int *lda, - const fpa *b, const int *ldb, const fpc *beta, fpc *c, const int *ldc); +static void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int* m, + const int* n, const int* k, const fpc* alpha, const fpa* a, const int* lda, + const fpa* b, const int* ldb, const fpc* beta, fpc* c, const int* ldc); template <> -void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int *m, - const int *n, const int *k, const float *alpha, const sycl::half *a, const int *lda, - const sycl::half *b, const int *ldb, const float *beta, float *c, const int *ldc) { +void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int* m, + const int* n, const int* k, const float* alpha, const sycl::half* a, const int* lda, + const sycl::half* b, const int* ldb, const float* beta, float* c, const int* ldc) { // Not supported in NETLIB. SGEMM is used as reference. int sizea, sizeb; if (layout == CblasColMajor) { @@ -230,8 +230,8 @@ void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, c sizea = (transa == CblasNoTrans) ? *lda * *m : *lda * *k; sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; } - float *af = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizea); - float *bf = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizeb); + float* af = (float*)oneapi::math::aligned_alloc(64, sizeof(float) * sizea); + float* bf = (float*)oneapi::math::aligned_alloc(64, sizeof(float) * sizeb); copy_mat(a, layout, transa, *m, *k, *lda, af); copy_mat(b, layout, transb, *k, *n, *ldb, bf); cblas_sgemm_wrapper(layout, transa, transb, *m, *n, *k, *alpha, af, *lda, bf, *ldb, *beta, c, @@ -241,10 +241,10 @@ void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, c } template <> -void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int *m, - const int *n, const int *k, const float *alpha, const oneapi::math::bfloat16 *a, - const int *lda, const oneapi::math::bfloat16 *b, const int *ldb, const float *beta, - float *c, const int *ldc) { +void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int* m, + const int* n, const int* k, const float* alpha, const oneapi::math::bfloat16* a, + const int* lda, const oneapi::math::bfloat16* b, const int* ldb, const float* beta, + float* c, const int* ldc) { // Not supported in NETLIB. SGEMM is used as reference. int sizea, sizeb; if (layout == CblasColMajor) { @@ -255,8 +255,8 @@ void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, c sizea = (transa == CblasNoTrans) ? *lda * *m : *lda * *k; sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; } - float *af = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizea); - float *bf = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizeb); + float* af = (float*)oneapi::math::aligned_alloc(64, sizeof(float) * sizea); + float* bf = (float*)oneapi::math::aligned_alloc(64, sizeof(float) * sizeb); copy_mat(a, layout, transa, *m, *k, *lda, af); copy_mat(b, layout, transb, *k, *n, *ldb, bf); cblas_sgemm_wrapper(layout, transa, transb, *m, *n, *k, *alpha, af, *lda, bf, *ldb, *beta, c, @@ -266,1146 +266,1142 @@ void gemm(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, c } template -static void symm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int *m, - const int *n, const fp *alpha, const fp *a, const int *lda, const fp *b, - const int *ldb, const fp *beta, fp *c, const int *ldc); +static void symm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int* m, + const int* n, const fp* alpha, const fp* a, const int* lda, const fp* b, + const int* ldb, const fp* beta, fp* c, const int* ldc); template <> -void symm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int *m, const int *n, - const float *alpha, const float *a, const int *lda, const float *b, const int *ldb, - const float *beta, float *c, const int *ldc) { +void symm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int* m, const int* n, + const float* alpha, const float* a, const int* lda, const float* b, const int* ldb, + const float* beta, float* c, const int* ldc) { cblas_ssymm_wrapper(layout, left_right, uplo, *m, *n, *alpha, a, *lda, b, *ldb, *beta, c, *ldc); } template <> -void symm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int *m, const int *n, - const double *alpha, const double *a, const int *lda, const double *b, const int *ldb, - const double *beta, double *c, const int *ldc) { +void symm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int* m, const int* n, + const double* alpha, const double* a, const int* lda, const double* b, const int* ldb, + const double* beta, double* c, const int* ldc) { cblas_dsymm_wrapper(layout, left_right, uplo, *m, *n, *alpha, a, *lda, b, *ldb, *beta, c, *ldc); } template <> -void symm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int *m, const int *n, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *b, const int *ldb, const std::complex *beta, - std::complex *c, const int *ldc) { - cblas_csymm_wrapper(layout, left_right, uplo, *m, *n, (const void *)alpha, (const void *)a, - *lda, (const void *)b, *ldb, (const void *)beta, (void *)c, *ldc); +void symm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int* m, const int* n, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* b, const int* ldb, const std::complex* beta, + std::complex* c, const int* ldc) { + cblas_csymm_wrapper(layout, left_right, uplo, *m, *n, (const void*)alpha, (const void*)a, *lda, + (const void*)b, *ldb, (const void*)beta, (void*)c, *ldc); } template <> -void symm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int *m, const int *n, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *b, const int *ldb, const std::complex *beta, - std::complex *c, const int *ldc) { - cblas_zsymm_wrapper(layout, left_right, uplo, *m, *n, (const void *)alpha, (const void *)a, - *lda, (const void *)b, *ldb, (const void *)beta, (void *)c, *ldc); +void symm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int* m, const int* n, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* b, const int* ldb, const std::complex* beta, + std::complex* c, const int* ldc) { + cblas_zsymm_wrapper(layout, left_right, uplo, *m, *n, (const void*)alpha, (const void*)a, *lda, + (const void*)b, *ldb, (const void*)beta, (void*)c, *ldc); } template -static void syrk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, - const int *k, const fp *alpha, const fp *a, const int *lda, const fp *beta, fp *c, - const int *ldc); +static void syrk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, + const int* k, const fp* alpha, const fp* a, const int* lda, const fp* beta, fp* c, + const int* ldc); template <> -void syrk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const float *alpha, const float *a, const int *lda, const float *beta, float *c, - const int *ldc) { +void syrk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const float* alpha, const float* a, const int* lda, const float* beta, float* c, + const int* ldc) { cblas_ssyrk_wrapper(layout, uplo, trans, *n, *k, *alpha, a, *lda, *beta, c, *ldc); } template <> -void syrk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const double *alpha, const double *a, const int *lda, const double *beta, double *c, - const int *ldc) { +void syrk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const double* alpha, const double* a, const int* lda, const double* beta, double* c, + const int* ldc) { cblas_dsyrk_wrapper(layout, uplo, trans, *n, *k, *alpha, a, *lda, *beta, c, *ldc); } template <> -void syrk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *beta, std::complex *c, const int *ldc) { - cblas_csyrk_wrapper(layout, uplo, trans, *n, *k, (const void *)alpha, (const void *)a, *lda, - (const void *)beta, (void *)c, *ldc); +void syrk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* beta, std::complex* c, const int* ldc) { + cblas_csyrk_wrapper(layout, uplo, trans, *n, *k, (const void*)alpha, (const void*)a, *lda, + (const void*)beta, (void*)c, *ldc); } template <> -void syrk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *beta, std::complex *c, const int *ldc) { - cblas_zsyrk_wrapper(layout, uplo, trans, *n, *k, (const void *)alpha, (const void *)a, *lda, - (const void *)beta, (void *)c, *ldc); +void syrk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* beta, std::complex* c, const int* ldc) { + cblas_zsyrk_wrapper(layout, uplo, trans, *n, *k, (const void*)alpha, (const void*)a, *lda, + (const void*)beta, (void*)c, *ldc); } template -static void hemm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int *m, - const int *n, const fp *alpha, const fp *a, const int *lda, const fp *b, - const int *ldb, const fp *beta, fp *c, const int *ldc); +static void hemm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int* m, + const int* n, const fp* alpha, const fp* a, const int* lda, const fp* b, + const int* ldb, const fp* beta, fp* c, const int* ldc); template <> -void hemm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int *m, const int *n, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *b, const int *ldb, const std::complex *beta, - std::complex *c, const int *ldc) { - cblas_chemm_wrapper(layout, left_right, uplo, *m, *n, (const void *)alpha, (const void *)a, - *lda, (const void *)b, *ldb, (const void *)beta, (void *)c, *ldc); +void hemm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int* m, const int* n, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* b, const int* ldb, const std::complex* beta, + std::complex* c, const int* ldc) { + cblas_chemm_wrapper(layout, left_right, uplo, *m, *n, (const void*)alpha, (const void*)a, *lda, + (const void*)b, *ldb, (const void*)beta, (void*)c, *ldc); } template <> -void hemm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int *m, const int *n, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *b, const int *ldb, const std::complex *beta, - std::complex *c, const int *ldc) { - cblas_zhemm_wrapper(layout, left_right, uplo, *m, *n, (const void *)alpha, (const void *)a, - *lda, (const void *)b, *ldb, (const void *)beta, (void *)c, *ldc); +void hemm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int* m, const int* n, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* b, const int* ldb, const std::complex* beta, + std::complex* c, const int* ldc) { + cblas_zhemm_wrapper(layout, left_right, uplo, *m, *n, (const void*)alpha, (const void*)a, *lda, + (const void*)b, *ldb, (const void*)beta, (void*)c, *ldc); } template -static void herk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, - const int *k, const fp_scalar *alpha, const fp_data *a, const int *lda, - const fp_scalar *beta, fp_data *c, const int *ldc); +static void herk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, + const int* k, const fp_scalar* alpha, const fp_data* a, const int* lda, + const fp_scalar* beta, fp_data* c, const int* ldc); template <> -void herk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const float *alpha, const std::complex *a, const int *lda, const float *beta, - std::complex *c, const int *ldc) { - cblas_cherk_wrapper(layout, uplo, trans, *n, *k, *alpha, (const void *)a, *lda, *beta, - (void *)c, *ldc); +void herk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const float* alpha, const std::complex* a, const int* lda, const float* beta, + std::complex* c, const int* ldc) { + cblas_cherk_wrapper(layout, uplo, trans, *n, *k, *alpha, (const void*)a, *lda, *beta, (void*)c, + *ldc); } template <> -void herk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const double *alpha, const std::complex *a, const int *lda, const double *beta, - std::complex *c, const int *ldc) { - cblas_zherk_wrapper(layout, uplo, trans, *n, *k, *alpha, (const void *)a, *lda, *beta, - (void *)c, *ldc); +void herk(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const double* alpha, const std::complex* a, const int* lda, const double* beta, + std::complex* c, const int* ldc) { + cblas_zherk_wrapper(layout, uplo, trans, *n, *k, *alpha, (const void*)a, *lda, *beta, (void*)c, + *ldc); } template -static void syr2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, - const int *k, const fp *alpha, const fp *a, const int *lda, const fp *b, - const int *ldb, const fp *beta, fp *c, const int *ldc); +static void syr2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, + const int* k, const fp* alpha, const fp* a, const int* lda, const fp* b, + const int* ldb, const fp* beta, fp* c, const int* ldc); template <> -void syr2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const float *alpha, const float *a, const int *lda, const float *b, const int *ldb, - const float *beta, float *c, const int *ldc) { +void syr2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const float* alpha, const float* a, const int* lda, const float* b, const int* ldb, + const float* beta, float* c, const int* ldc) { cblas_ssyr2k_wrapper(layout, uplo, trans, *n, *k, *alpha, a, *lda, b, *ldb, *beta, c, *ldc); } template <> -void syr2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const double *alpha, const double *a, const int *lda, const double *b, const int *ldb, - const double *beta, double *c, const int *ldc) { +void syr2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const double* alpha, const double* a, const int* lda, const double* b, const int* ldb, + const double* beta, double* c, const int* ldc) { cblas_dsyr2k_wrapper(layout, uplo, trans, *n, *k, *alpha, a, *lda, b, *ldb, *beta, c, *ldc); } template <> -void syr2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *b, const int *ldb, const std::complex *beta, - std::complex *c, const int *ldc) { - cblas_csyr2k_wrapper(layout, uplo, trans, *n, *k, (const void *)alpha, (const void *)a, *lda, - (const void *)b, *ldb, (const void *)beta, (void *)c, *ldc); +void syr2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* b, const int* ldb, const std::complex* beta, + std::complex* c, const int* ldc) { + cblas_csyr2k_wrapper(layout, uplo, trans, *n, *k, (const void*)alpha, (const void*)a, *lda, + (const void*)b, *ldb, (const void*)beta, (void*)c, *ldc); } template <> -void syr2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *b, const int *ldb, const std::complex *beta, - std::complex *c, const int *ldc) { - cblas_zsyr2k_wrapper(layout, uplo, trans, *n, *k, (const void *)alpha, (const void *)a, *lda, - (const void *)b, *ldb, (const void *)beta, (void *)c, *ldc); +void syr2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* b, const int* ldb, const std::complex* beta, + std::complex* c, const int* ldc) { + cblas_zsyr2k_wrapper(layout, uplo, trans, *n, *k, (const void*)alpha, (const void*)a, *lda, + (const void*)b, *ldb, (const void*)beta, (void*)c, *ldc); } template -static void her2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, - const int *k, const fp_data *alpha, const fp_data *a, const int *lda, - const fp_data *b, const int *ldb, const fp_scalar *beta, fp_data *c, - const int *ldc); +static void her2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, + const int* k, const fp_data* alpha, const fp_data* a, const int* lda, + const fp_data* b, const int* ldb, const fp_scalar* beta, fp_data* c, + const int* ldc); template <> -void her2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *b, const int *ldb, const float *beta, std::complex *c, - const int *ldc) { - cblas_cher2k_wrapper(layout, uplo, trans, *n, *k, (const void *)alpha, (const void *)a, *lda, - (const void *)b, *ldb, *beta, (void *)c, *ldc); +void her2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* b, const int* ldb, const float* beta, std::complex* c, + const int* ldc) { + cblas_cher2k_wrapper(layout, uplo, trans, *n, *k, (const void*)alpha, (const void*)a, *lda, + (const void*)b, *ldb, *beta, (void*)c, *ldc); } template <> -void her2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int *n, const int *k, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *b, const int *ldb, const double *beta, - std::complex *c, const int *ldc) { - cblas_zher2k_wrapper(layout, uplo, trans, *n, *k, (const void *)alpha, (const void *)a, *lda, - (const void *)b, *ldb, *beta, (void *)c, *ldc); +void her2k(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int* n, const int* k, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* b, const int* ldb, const double* beta, + std::complex* c, const int* ldc) { + cblas_zher2k_wrapper(layout, uplo, trans, *n, *k, (const void*)alpha, (const void*)a, *lda, + (const void*)b, *ldb, *beta, (void*)c, *ldc); } template static void trmm(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int *m, const int *n, const fp *alpha, const fp *a, - const int *lda, fp *b, const int *ldb); + CBLAS_DIAG diag, const int* m, const int* n, const fp* alpha, const fp* a, + const int* lda, fp* b, const int* ldb); template <> void trmm(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int *m, const int *n, const float *alpha, const float *a, - const int *lda, float *b, const int *ldb) { + CBLAS_DIAG diag, const int* m, const int* n, const float* alpha, const float* a, + const int* lda, float* b, const int* ldb) { cblas_strmm_wrapper(layout, side, uplo, transa, diag, *m, *n, *alpha, a, *lda, b, *ldb); } template <> void trmm(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int *m, const int *n, const double *alpha, const double *a, - const int *lda, double *b, const int *ldb) { + CBLAS_DIAG diag, const int* m, const int* n, const double* alpha, const double* a, + const int* lda, double* b, const int* ldb) { cblas_dtrmm_wrapper(layout, side, uplo, transa, diag, *m, *n, *alpha, a, *lda, b, *ldb); } template <> void trmm(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int *m, const int *n, const std::complex *alpha, - const std::complex *a, const int *lda, std::complex *b, const int *ldb) { - cblas_ctrmm_wrapper(layout, side, uplo, transa, diag, *m, *n, (const void *)alpha, - (const void *)a, *lda, (void *)b, *ldb); + CBLAS_DIAG diag, const int* m, const int* n, const std::complex* alpha, + const std::complex* a, const int* lda, std::complex* b, const int* ldb) { + cblas_ctrmm_wrapper(layout, side, uplo, transa, diag, *m, *n, (const void*)alpha, + (const void*)a, *lda, (void*)b, *ldb); } template <> void trmm(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int *m, const int *n, const std::complex *alpha, - const std::complex *a, const int *lda, std::complex *b, const int *ldb) { - cblas_ztrmm_wrapper(layout, side, uplo, transa, diag, *m, *n, (const void *)alpha, - (const void *)a, *lda, (void *)b, *ldb); + CBLAS_DIAG diag, const int* m, const int* n, const std::complex* alpha, + const std::complex* a, const int* lda, std::complex* b, const int* ldb) { + cblas_ztrmm_wrapper(layout, side, uplo, transa, diag, *m, *n, (const void*)alpha, + (const void*)a, *lda, (void*)b, *ldb); } template static void trsm(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int *m, const int *n, const fp *alpha, const fp *a, - const int *lda, fp *b, const int *ldb); + CBLAS_DIAG diag, const int* m, const int* n, const fp* alpha, const fp* a, + const int* lda, fp* b, const int* ldb); template <> void trsm(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int *m, const int *n, const float *alpha, const float *a, - const int *lda, float *b, const int *ldb) { + CBLAS_DIAG diag, const int* m, const int* n, const float* alpha, const float* a, + const int* lda, float* b, const int* ldb) { cblas_strsm_wrapper(layout, side, uplo, transa, diag, *m, *n, *alpha, a, *lda, b, *ldb); } template <> void trsm(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int *m, const int *n, const double *alpha, const double *a, - const int *lda, double *b, const int *ldb) { + CBLAS_DIAG diag, const int* m, const int* n, const double* alpha, const double* a, + const int* lda, double* b, const int* ldb) { cblas_dtrsm_wrapper(layout, side, uplo, transa, diag, *m, *n, *alpha, a, *lda, b, *ldb); } template <> void trsm(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int *m, const int *n, const std::complex *alpha, - const std::complex *a, const int *lda, std::complex *b, const int *ldb) { - cblas_ctrsm_wrapper(layout, side, uplo, transa, diag, *m, *n, (const void *)alpha, - (const void *)a, *lda, (void *)b, *ldb); + CBLAS_DIAG diag, const int* m, const int* n, const std::complex* alpha, + const std::complex* a, const int* lda, std::complex* b, const int* ldb) { + cblas_ctrsm_wrapper(layout, side, uplo, transa, diag, *m, *n, (const void*)alpha, + (const void*)a, *lda, (void*)b, *ldb); } template <> void trsm(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int *m, const int *n, const std::complex *alpha, - const std::complex *a, const int *lda, std::complex *b, const int *ldb) { - cblas_ztrsm_wrapper(layout, side, uplo, transa, diag, *m, *n, (const void *)alpha, - (const void *)a, *lda, (void *)b, *ldb); + CBLAS_DIAG diag, const int* m, const int* n, const std::complex* alpha, + const std::complex* a, const int* lda, std::complex* b, const int* ldb) { + cblas_ztrsm_wrapper(layout, side, uplo, transa, diag, *m, *n, (const void*)alpha, + (const void*)a, *lda, (void*)b, *ldb); } /* Level 2 */ template -static void gemv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int *m, const int *n, - const fp *alpha, const fp *a, const int *lda, const fp *x, const int *incx, - const fp *beta, fp *y, const int *incy); +static void gemv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int* m, const int* n, + const fp* alpha, const fp* a, const int* lda, const fp* x, const int* incx, + const fp* beta, fp* y, const int* incy); template <> -void gemv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int *m, const int *n, - const float *alpha, const float *a, const int *lda, const float *x, const int *incx, - const float *beta, float *y, const int *incy) { +void gemv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int* m, const int* n, + const float* alpha, const float* a, const int* lda, const float* x, const int* incx, + const float* beta, float* y, const int* incy) { cblas_sgemv_wrapper(layout, trans, *m, *n, *alpha, a, *lda, x, *incx, *beta, y, *incy); } template <> -void gemv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int *m, const int *n, - const double *alpha, const double *a, const int *lda, const double *x, const int *incx, - const double *beta, double *y, const int *incy) { +void gemv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int* m, const int* n, + const double* alpha, const double* a, const int* lda, const double* x, const int* incx, + const double* beta, double* y, const int* incy) { cblas_dgemv_wrapper(layout, trans, *m, *n, *alpha, a, *lda, x, *incx, *beta, y, *incy); } template <> -void gemv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int *m, const int *n, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *x, const int *incx, const std::complex *beta, - std::complex *y, const int *incy) { - cblas_cgemv_wrapper(layout, trans, *m, *n, (const void *)alpha, (const void *)a, *lda, - (const void *)x, *incx, (const void *)beta, (void *)y, *incy); +void gemv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int* m, const int* n, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* x, const int* incx, const std::complex* beta, + std::complex* y, const int* incy) { + cblas_cgemv_wrapper(layout, trans, *m, *n, (const void*)alpha, (const void*)a, *lda, + (const void*)x, *incx, (const void*)beta, (void*)y, *incy); } template <> -void gemv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int *m, const int *n, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *x, const int *incx, const std::complex *beta, - std::complex *y, const int *incy) { - cblas_zgemv_wrapper(layout, trans, *m, *n, (const void *)alpha, (const void *)a, *lda, - (const void *)x, *incx, (const void *)beta, (void *)y, *incy); +void gemv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int* m, const int* n, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* x, const int* incx, const std::complex* beta, + std::complex* y, const int* incy) { + cblas_zgemv_wrapper(layout, trans, *m, *n, (const void*)alpha, (const void*)a, *lda, + (const void*)x, *incx, (const void*)beta, (void*)y, *incy); } template -static void gbmv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int *m, const int *n, int *kl, - int *ku, const fp *alpha, const fp *a, const int *lda, const fp *x, - const int *incx, const fp *beta, fp *y, const int *incy); +static void gbmv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int* m, const int* n, int* kl, + int* ku, const fp* alpha, const fp* a, const int* lda, const fp* x, + const int* incx, const fp* beta, fp* y, const int* incy); template <> -void gbmv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int *m, const int *n, int *kl, int *ku, - const float *alpha, const float *a, const int *lda, const float *x, const int *incx, - const float *beta, float *y, const int *incy) { +void gbmv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int* m, const int* n, int* kl, int* ku, + const float* alpha, const float* a, const int* lda, const float* x, const int* incx, + const float* beta, float* y, const int* incy) { cblas_sgbmv_wrapper(layout, trans, *m, *n, *kl, *ku, *alpha, a, *lda, x, *incx, *beta, y, *incy); } template <> -void gbmv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int *m, const int *n, int *kl, int *ku, - const double *alpha, const double *a, const int *lda, const double *x, const int *incx, - const double *beta, double *y, const int *incy) { +void gbmv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int* m, const int* n, int* kl, int* ku, + const double* alpha, const double* a, const int* lda, const double* x, const int* incx, + const double* beta, double* y, const int* incy) { cblas_dgbmv_wrapper(layout, trans, *m, *n, *kl, *ku, *alpha, a, *lda, x, *incx, *beta, y, *incy); } template <> -void gbmv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int *m, const int *n, int *kl, int *ku, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *x, const int *incx, const std::complex *beta, - std::complex *y, const int *incy) { - cblas_cgbmv_wrapper(layout, trans, *m, *n, *kl, *ku, (const void *)alpha, (const void *)a, *lda, - (const void *)x, *incx, (const void *)beta, (void *)y, *incy); +void gbmv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int* m, const int* n, int* kl, int* ku, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* x, const int* incx, const std::complex* beta, + std::complex* y, const int* incy) { + cblas_cgbmv_wrapper(layout, trans, *m, *n, *kl, *ku, (const void*)alpha, (const void*)a, *lda, + (const void*)x, *incx, (const void*)beta, (void*)y, *incy); } template <> -void gbmv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int *m, const int *n, int *kl, int *ku, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *x, const int *incx, const std::complex *beta, - std::complex *y, const int *incy) { - cblas_zgbmv_wrapper(layout, trans, *m, *n, *kl, *ku, (const void *)alpha, (const void *)a, *lda, - (const void *)x, *incx, (const void *)beta, (void *)y, *incy); +void gbmv(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int* m, const int* n, int* kl, int* ku, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* x, const int* incx, const std::complex* beta, + std::complex* y, const int* incy) { + cblas_zgbmv_wrapper(layout, trans, *m, *n, *kl, *ku, (const void*)alpha, (const void*)a, *lda, + (const void*)x, *incx, (const void*)beta, (void*)y, *incy); } template -static void ger(CBLAS_LAYOUT layout, const int *m, const int *n, const fp *alpha, const fp *x, - const int *incx, const fp *y, const int *incy, fp *a, const int *lda); +static void ger(CBLAS_LAYOUT layout, const int* m, const int* n, const fp* alpha, const fp* x, + const int* incx, const fp* y, const int* incy, fp* a, const int* lda); template <> -void ger(CBLAS_LAYOUT layout, const int *m, const int *n, const float *alpha, const float *x, - const int *incx, const float *y, const int *incy, float *a, const int *lda) { +void ger(CBLAS_LAYOUT layout, const int* m, const int* n, const float* alpha, const float* x, + const int* incx, const float* y, const int* incy, float* a, const int* lda) { cblas_sger_wrapper(layout, *m, *n, *alpha, x, *incx, y, *incy, a, *lda); } template <> -void ger(CBLAS_LAYOUT layout, const int *m, const int *n, const double *alpha, const double *x, - const int *incx, const double *y, const int *incy, double *a, const int *lda) { +void ger(CBLAS_LAYOUT layout, const int* m, const int* n, const double* alpha, const double* x, + const int* incx, const double* y, const int* incy, double* a, const int* lda) { cblas_dger_wrapper(layout, *m, *n, *alpha, x, *incx, y, *incy, a, *lda); } template -static void gerc(CBLAS_LAYOUT layout, const int *m, const int *n, const fp *alpha, const fp *x, - const int *incx, const fp *y, const int *incy, fp *a, const int *lda); +static void gerc(CBLAS_LAYOUT layout, const int* m, const int* n, const fp* alpha, const fp* x, + const int* incx, const fp* y, const int* incy, fp* a, const int* lda); template <> -void gerc(CBLAS_LAYOUT layout, const int *m, const int *n, const std::complex *alpha, - const std::complex *x, const int *incx, const std::complex *y, - const int *incy, std::complex *a, const int *lda) { - cblas_cgerc_wrapper(layout, *m, *n, (const void *)alpha, (const void *)x, *incx, - (const void *)y, *incy, (void *)a, *lda); +void gerc(CBLAS_LAYOUT layout, const int* m, const int* n, const std::complex* alpha, + const std::complex* x, const int* incx, const std::complex* y, + const int* incy, std::complex* a, const int* lda) { + cblas_cgerc_wrapper(layout, *m, *n, (const void*)alpha, (const void*)x, *incx, (const void*)y, + *incy, (void*)a, *lda); } template <> -void gerc(CBLAS_LAYOUT layout, const int *m, const int *n, const std::complex *alpha, - const std::complex *x, const int *incx, const std::complex *y, - const int *incy, std::complex *a, const int *lda) { - cblas_zgerc_wrapper(layout, *m, *n, (const void *)alpha, (const void *)x, *incx, - (const void *)y, *incy, (void *)a, *lda); +void gerc(CBLAS_LAYOUT layout, const int* m, const int* n, const std::complex* alpha, + const std::complex* x, const int* incx, const std::complex* y, + const int* incy, std::complex* a, const int* lda) { + cblas_zgerc_wrapper(layout, *m, *n, (const void*)alpha, (const void*)x, *incx, (const void*)y, + *incy, (void*)a, *lda); } template -static void geru(CBLAS_LAYOUT layout, const int *m, const int *n, const fp *alpha, const fp *x, - const int *incx, const fp *y, const int *incy, fp *a, const int *lda); +static void geru(CBLAS_LAYOUT layout, const int* m, const int* n, const fp* alpha, const fp* x, + const int* incx, const fp* y, const int* incy, fp* a, const int* lda); template <> -void geru(CBLAS_LAYOUT layout, const int *m, const int *n, const std::complex *alpha, - const std::complex *x, const int *incx, const std::complex *y, - const int *incy, std::complex *a, const int *lda) { - cblas_cgeru_wrapper(layout, *m, *n, (const void *)alpha, (const void *)x, *incx, - (const void *)y, *incy, (void *)a, *lda); +void geru(CBLAS_LAYOUT layout, const int* m, const int* n, const std::complex* alpha, + const std::complex* x, const int* incx, const std::complex* y, + const int* incy, std::complex* a, const int* lda) { + cblas_cgeru_wrapper(layout, *m, *n, (const void*)alpha, (const void*)x, *incx, (const void*)y, + *incy, (void*)a, *lda); } template <> -void geru(CBLAS_LAYOUT layout, const int *m, const int *n, const std::complex *alpha, - const std::complex *x, const int *incx, const std::complex *y, - const int *incy, std::complex *a, const int *lda) { - cblas_zgeru_wrapper(layout, *m, *n, (const void *)alpha, (const void *)x, *incx, - (const void *)y, *incy, (void *)a, *lda); +void geru(CBLAS_LAYOUT layout, const int* m, const int* n, const std::complex* alpha, + const std::complex* x, const int* incx, const std::complex* y, + const int* incy, std::complex* a, const int* lda) { + cblas_zgeru_wrapper(layout, *m, *n, (const void*)alpha, (const void*)x, *incx, (const void*)y, + *incy, (void*)a, *lda); } template -static void hbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const int *k, - const fp *alpha, const fp *a, const int *lda, const fp *x, const int *incx, - const fp *beta, fp *y, const int *incy); +static void hbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const int* k, + const fp* alpha, const fp* a, const int* lda, const fp* x, const int* incx, + const fp* beta, fp* y, const int* incy); template <> -void hbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const int *k, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *x, const int *incx, const std::complex *beta, - std::complex *y, const int *incy) { - cblas_chbmv_wrapper(layout, upper_lower, *n, *k, (const void *)alpha, (const void *)a, *lda, - (const void *)x, *incx, (const void *)beta, (void *)y, *incy); +void hbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const int* k, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* x, const int* incx, const std::complex* beta, + std::complex* y, const int* incy) { + cblas_chbmv_wrapper(layout, upper_lower, *n, *k, (const void*)alpha, (const void*)a, *lda, + (const void*)x, *incx, (const void*)beta, (void*)y, *incy); } template <> -void hbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const int *k, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *x, const int *incx, const std::complex *beta, - std::complex *y, const int *incy) { - cblas_zhbmv_wrapper(layout, upper_lower, *n, *k, (const void *)alpha, (const void *)a, *lda, - (const void *)x, *incx, (const void *)beta, (void *)y, *incy); +void hbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const int* k, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* x, const int* incx, const std::complex* beta, + std::complex* y, const int* incy) { + cblas_zhbmv_wrapper(layout, upper_lower, *n, *k, (const void*)alpha, (const void*)a, *lda, + (const void*)x, *incx, (const void*)beta, (void*)y, *incy); } template -static void hemv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp *alpha, - const fp *a, const int *lda, const fp *x, const int *incx, const fp *beta, fp *y, - const int *incy); +static void hemv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp* alpha, + const fp* a, const int* lda, const fp* x, const int* incx, const fp* beta, fp* y, + const int* incy); template <> -void hemv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *x, const int *incx, const std::complex *beta, - std::complex *y, const int *incy) { - cblas_chemv_wrapper(layout, upper_lower, *n, (const void *)alpha, (const void *)a, *lda, - (const void *)x, *incx, (const void *)beta, (void *)y, *incy); +void hemv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* x, const int* incx, const std::complex* beta, + std::complex* y, const int* incy) { + cblas_chemv_wrapper(layout, upper_lower, *n, (const void*)alpha, (const void*)a, *lda, + (const void*)x, *incx, (const void*)beta, (void*)y, *incy); } template <> -void hemv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, - const std::complex *alpha, const std::complex *a, const int *lda, - const std::complex *x, const int *incx, const std::complex *beta, - std::complex *y, const int *incy) { - cblas_zhemv_wrapper(layout, upper_lower, *n, (const void *)alpha, (const void *)a, *lda, - (const void *)x, *incx, (const void *)beta, (void *)y, *incy); +void hemv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, + const std::complex* alpha, const std::complex* a, const int* lda, + const std::complex* x, const int* incx, const std::complex* beta, + std::complex* y, const int* incy) { + cblas_zhemv_wrapper(layout, upper_lower, *n, (const void*)alpha, (const void*)a, *lda, + (const void*)x, *incx, (const void*)beta, (void*)y, *incy); } template -static void her(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp_scalar *alpha, - const fp_data *x, const int *incx, fp_data *a, const int *lda); +static void her(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp_scalar* alpha, + const fp_data* x, const int* incx, fp_data* a, const int* lda); template <> -void her(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const float *alpha, - const std::complex *x, const int *incx, std::complex *a, const int *lda) { - cblas_cher_wrapper(layout, upper_lower, *n, *alpha, (const void *)x, *incx, (void *)a, *lda); +void her(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const float* alpha, + const std::complex* x, const int* incx, std::complex* a, const int* lda) { + cblas_cher_wrapper(layout, upper_lower, *n, *alpha, (const void*)x, *incx, (void*)a, *lda); } template <> -void her(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const double *alpha, - const std::complex *x, const int *incx, std::complex *a, const int *lda) { - cblas_zher_wrapper(layout, upper_lower, *n, *alpha, (const void *)x, *incx, (void *)a, *lda); +void her(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const double* alpha, + const std::complex* x, const int* incx, std::complex* a, const int* lda) { + cblas_zher_wrapper(layout, upper_lower, *n, *alpha, (const void*)x, *incx, (void*)a, *lda); } template -static void her2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp *alpha, - const fp *x, const int *incx, const fp *y, const int *incy, fp *a, const int *lda); +static void her2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp* alpha, + const fp* x, const int* incx, const fp* y, const int* incy, fp* a, const int* lda); template <> -void her2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, - const std::complex *alpha, const std::complex *x, const int *incx, - const std::complex *y, const int *incy, std::complex *a, const int *lda) { - cblas_cher2_wrapper(layout, upper_lower, *n, (const void *)alpha, (const void *)x, *incx, - (const void *)y, *incy, (void *)a, *lda); +void her2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, + const std::complex* alpha, const std::complex* x, const int* incx, + const std::complex* y, const int* incy, std::complex* a, const int* lda) { + cblas_cher2_wrapper(layout, upper_lower, *n, (const void*)alpha, (const void*)x, *incx, + (const void*)y, *incy, (void*)a, *lda); } template <> -void her2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, - const std::complex *alpha, const std::complex *x, const int *incx, - const std::complex *y, const int *incy, std::complex *a, const int *lda) { - cblas_zher2_wrapper(layout, upper_lower, *n, (const void *)alpha, (const void *)x, *incx, - (const void *)y, *incy, (void *)a, *lda); +void her2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, + const std::complex* alpha, const std::complex* x, const int* incx, + const std::complex* y, const int* incy, std::complex* a, const int* lda) { + cblas_zher2_wrapper(layout, upper_lower, *n, (const void*)alpha, (const void*)x, *incx, + (const void*)y, *incy, (void*)a, *lda); } template -static void hpmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp *alpha, - const fp *a, const fp *x, const int *incx, const fp *beta, fp *y, const int *incy); +static void hpmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp* alpha, + const fp* a, const fp* x, const int* incx, const fp* beta, fp* y, const int* incy); template <> -void hpmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, - const std::complex *alpha, const std::complex *a, - const std::complex *x, const int *incx, const std::complex *beta, - std::complex *y, const int *incy) { - cblas_chpmv_wrapper(layout, upper_lower, *n, (const void *)alpha, (const void *)a, - (const void *)x, *incx, (const void *)beta, (void *)y, *incy); +void hpmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, + const std::complex* alpha, const std::complex* a, + const std::complex* x, const int* incx, const std::complex* beta, + std::complex* y, const int* incy) { + cblas_chpmv_wrapper(layout, upper_lower, *n, (const void*)alpha, (const void*)a, (const void*)x, + *incx, (const void*)beta, (void*)y, *incy); } template <> -void hpmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, - const std::complex *alpha, const std::complex *a, - const std::complex *x, const int *incx, const std::complex *beta, - std::complex *y, const int *incy) { - cblas_zhpmv_wrapper(layout, upper_lower, *n, (const void *)alpha, (const void *)a, - (const void *)x, *incx, (const void *)beta, (void *)y, *incy); +void hpmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, + const std::complex* alpha, const std::complex* a, + const std::complex* x, const int* incx, const std::complex* beta, + std::complex* y, const int* incy) { + cblas_zhpmv_wrapper(layout, upper_lower, *n, (const void*)alpha, (const void*)a, (const void*)x, + *incx, (const void*)beta, (void*)y, *incy); } template -static void hpr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp_scalar *alpha, - const fp_data *x, const int *incx, fp_data *a); +static void hpr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp_scalar* alpha, + const fp_data* x, const int* incx, fp_data* a); template <> -void hpr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const float *alpha, - const std::complex *x, const int *incx, std::complex *a) { - cblas_chpr_wrapper(layout, upper_lower, *n, *alpha, (const void *)x, *incx, (void *)a); +void hpr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const float* alpha, + const std::complex* x, const int* incx, std::complex* a) { + cblas_chpr_wrapper(layout, upper_lower, *n, *alpha, (const void*)x, *incx, (void*)a); } template <> -void hpr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const double *alpha, - const std::complex *x, const int *incx, std::complex *a) { - cblas_zhpr_wrapper(layout, upper_lower, *n, *alpha, (const void *)x, *incx, (void *)a); +void hpr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const double* alpha, + const std::complex* x, const int* incx, std::complex* a) { + cblas_zhpr_wrapper(layout, upper_lower, *n, *alpha, (const void*)x, *incx, (void*)a); } template -static void hpr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp *alpha, - const fp *x, const int *incx, const fp *y, const int *incy, fp *a); +static void hpr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp* alpha, + const fp* x, const int* incx, const fp* y, const int* incy, fp* a); template <> -void hpr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, - const std::complex *alpha, const std::complex *x, const int *incx, - const std::complex *y, const int *incy, std::complex *a) { - cblas_chpr2_wrapper(layout, upper_lower, *n, (const void *)alpha, (const void *)x, *incx, - (const void *)y, *incy, (void *)a); +void hpr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, + const std::complex* alpha, const std::complex* x, const int* incx, + const std::complex* y, const int* incy, std::complex* a) { + cblas_chpr2_wrapper(layout, upper_lower, *n, (const void*)alpha, (const void*)x, *incx, + (const void*)y, *incy, (void*)a); } template <> -void hpr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, - const std::complex *alpha, const std::complex *x, const int *incx, - const std::complex *y, const int *incy, std::complex *a) { - cblas_zhpr2_wrapper(layout, upper_lower, *n, (const void *)alpha, (const void *)x, *incx, - (const void *)y, *incy, (void *)a); +void hpr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, + const std::complex* alpha, const std::complex* x, const int* incx, + const std::complex* y, const int* incy, std::complex* a) { + cblas_zhpr2_wrapper(layout, upper_lower, *n, (const void*)alpha, (const void*)x, *incx, + (const void*)y, *incy, (void*)a); } template -static void sbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const int *k, - const fp *alpha, const fp *a, const int *lda, const fp *x, const int *incx, - const fp *beta, fp *y, const int *incy); +static void sbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const int* k, + const fp* alpha, const fp* a, const int* lda, const fp* x, const int* incx, + const fp* beta, fp* y, const int* incy); template <> -void sbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const int *k, - const float *alpha, const float *a, const int *lda, const float *x, const int *incx, - const float *beta, float *y, const int *incy) { +void sbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const int* k, + const float* alpha, const float* a, const int* lda, const float* x, const int* incx, + const float* beta, float* y, const int* incy) { cblas_ssbmv_wrapper(layout, upper_lower, *n, *k, *alpha, a, *lda, x, *incx, *beta, y, *incy); } template <> -void sbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const int *k, - const double *alpha, const double *a, const int *lda, const double *x, const int *incx, - const double *beta, double *y, const int *incy) { +void sbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const int* k, + const double* alpha, const double* a, const int* lda, const double* x, const int* incx, + const double* beta, double* y, const int* incy) { cblas_dsbmv_wrapper(layout, upper_lower, *n, *k, *alpha, a, *lda, x, *incx, *beta, y, *incy); } template -static void symv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp *alpha, - const fp *a, const int *lda, const fp *x, const int *incx, const fp *beta, fp *y, - const int *incy); +static void symv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp* alpha, + const fp* a, const int* lda, const fp* x, const int* incx, const fp* beta, fp* y, + const int* incy); template <> -void symv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const float *alpha, - const float *a, const int *lda, const float *x, const int *incx, const float *beta, - float *y, const int *incy) { +void symv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const float* alpha, + const float* a, const int* lda, const float* x, const int* incx, const float* beta, + float* y, const int* incy) { cblas_ssymv_wrapper(layout, upper_lower, *n, *alpha, a, *lda, x, *incx, *beta, y, *incy); } template <> -void symv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const double *alpha, - const double *a, const int *lda, const double *x, const int *incx, const double *beta, - double *y, const int *incy) { +void symv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const double* alpha, + const double* a, const int* lda, const double* x, const int* incx, const double* beta, + double* y, const int* incy) { cblas_dsymv_wrapper(layout, upper_lower, *n, *alpha, a, *lda, x, *incx, *beta, y, *incy); } template -static void syr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp *alpha, - const fp *x, const int *incx, fp *a, const int *lda); +static void syr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp* alpha, + const fp* x, const int* incx, fp* a, const int* lda); template <> -void syr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const float *alpha, - const float *x, const int *incx, float *a, const int *lda) { +void syr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const float* alpha, + const float* x, const int* incx, float* a, const int* lda) { cblas_ssyr_wrapper(layout, upper_lower, *n, *alpha, x, *incx, a, *lda); } template <> -void syr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const double *alpha, - const double *x, const int *incx, double *a, const int *lda) { +void syr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const double* alpha, + const double* x, const int* incx, double* a, const int* lda) { cblas_dsyr_wrapper(layout, upper_lower, *n, *alpha, x, *incx, a, *lda); } template -static void syr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp *alpha, - const fp *x, const int *incx, const fp *y, const int *incy, fp *a, const int *lda); +static void syr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp* alpha, + const fp* x, const int* incx, const fp* y, const int* incy, fp* a, const int* lda); template <> -void syr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const float *alpha, - const float *x, const int *incx, const float *y, const int *incy, float *a, - const int *lda) { +void syr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const float* alpha, + const float* x, const int* incx, const float* y, const int* incy, float* a, + const int* lda) { cblas_ssyr2_wrapper(layout, upper_lower, *n, *alpha, x, *incx, y, *incy, a, *lda); } template <> -void syr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const double *alpha, - const double *x, const int *incx, const double *y, const int *incy, double *a, - const int *lda) { +void syr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const double* alpha, + const double* x, const int* incx, const double* y, const int* incy, double* a, + const int* lda) { cblas_dsyr2_wrapper(layout, upper_lower, *n, *alpha, x, *incx, y, *incy, a, *lda); } template -static void spmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp *alpha, - const fp *a, const fp *x, const int *incx, const fp *beta, fp *y, const int *incy); +static void spmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp* alpha, + const fp* a, const fp* x, const int* incx, const fp* beta, fp* y, const int* incy); template <> -void spmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const float *alpha, - const float *a, const float *x, const int *incx, const float *beta, float *y, - const int *incy) { +void spmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const float* alpha, + const float* a, const float* x, const int* incx, const float* beta, float* y, + const int* incy) { cblas_sspmv_wrapper(layout, upper_lower, *n, *alpha, a, x, *incx, *beta, y, *incy); } template <> -void spmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const double *alpha, - const double *a, const double *x, const int *incx, const double *beta, double *y, - const int *incy) { +void spmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const double* alpha, + const double* a, const double* x, const int* incx, const double* beta, double* y, + const int* incy) { cblas_dspmv_wrapper(layout, upper_lower, *n, *alpha, a, x, *incx, *beta, y, *incy); } template -static void spr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp *alpha, - const fp *x, const int *incx, fp *a); +static void spr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp* alpha, + const fp* x, const int* incx, fp* a); template <> -void spr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const float *alpha, - const float *x, const int *incx, float *a) { +void spr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const float* alpha, + const float* x, const int* incx, float* a) { cblas_sspr_wrapper(layout, upper_lower, *n, *alpha, x, *incx, a); } template <> -void spr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const double *alpha, - const double *x, const int *incx, double *a) { +void spr(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const double* alpha, + const double* x, const int* incx, double* a) { cblas_dspr_wrapper(layout, upper_lower, *n, *alpha, x, *incx, a); } template -static void spr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const fp *alpha, - const fp *x, const int *incx, const fp *y, const int *incy, fp *a); +static void spr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const fp* alpha, + const fp* x, const int* incx, const fp* y, const int* incy, fp* a); template <> -void spr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const float *alpha, - const float *x, const int *incx, const float *y, const int *incy, float *a) { +void spr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const float* alpha, + const float* x, const int* incx, const float* y, const int* incy, float* a) { cblas_sspr2_wrapper(layout, upper_lower, *n, *alpha, x, *incx, y, *incy, a); } template <> -void spr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int *n, const double *alpha, - const double *x, const int *incx, const double *y, const int *incy, double *a) { +void spr2(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int* n, const double* alpha, + const double* x, const int* incx, const double* y, const int* incy, double* a) { cblas_dspr2_wrapper(layout, upper_lower, *n, *alpha, x, *incx, y, *incy, a); } template static void tbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int *n, const int *k, const fp *a, const int *lda, - fp *x, const int *incx); + CBLAS_DIAG unit_diag, const int* n, const int* k, const fp* a, const int* lda, + fp* x, const int* incx); template <> void tbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const int *k, const float *a, const int *lda, float *x, const int *incx) { + const int* n, const int* k, const float* a, const int* lda, float* x, const int* incx) { cblas_stbmv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, a, *lda, x, *incx); } template <> void tbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const int *k, const double *a, const int *lda, double *x, const int *incx) { + const int* n, const int* k, const double* a, const int* lda, double* x, const int* incx) { cblas_dtbmv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, a, *lda, x, *incx); } template <> void tbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const int *k, const std::complex *a, const int *lda, - std::complex *x, const int *incx) { - cblas_ctbmv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, (const void *)a, *lda, - (void *)x, *incx); + const int* n, const int* k, const std::complex* a, const int* lda, + std::complex* x, const int* incx) { + cblas_ctbmv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, (const void*)a, *lda, + (void*)x, *incx); } template <> void tbmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const int *k, const std::complex *a, const int *lda, - std::complex *x, const int *incx) { - cblas_ztbmv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, (const void *)a, *lda, - (void *)x, *incx); + const int* n, const int* k, const std::complex* a, const int* lda, + std::complex* x, const int* incx) { + cblas_ztbmv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, (const void*)a, *lda, + (void*)x, *incx); } template static void tbsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int *n, const int *k, const fp *a, const int *lda, - fp *x, const int *incx); + CBLAS_DIAG unit_diag, const int* n, const int* k, const fp* a, const int* lda, + fp* x, const int* incx); template <> void tbsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const int *k, const float *a, const int *lda, float *x, const int *incx) { + const int* n, const int* k, const float* a, const int* lda, float* x, const int* incx) { cblas_stbsv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, a, *lda, x, *incx); } template <> void tbsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const int *k, const double *a, const int *lda, double *x, const int *incx) { + const int* n, const int* k, const double* a, const int* lda, double* x, const int* incx) { cblas_dtbsv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, a, *lda, x, *incx); } template <> void tbsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const int *k, const std::complex *a, const int *lda, - std::complex *x, const int *incx) { - cblas_ctbsv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, (const void *)a, *lda, - (void *)x, *incx); + const int* n, const int* k, const std::complex* a, const int* lda, + std::complex* x, const int* incx) { + cblas_ctbsv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, (const void*)a, *lda, + (void*)x, *incx); } template <> void tbsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const int *k, const std::complex *a, const int *lda, - std::complex *x, const int *incx) { - cblas_ztbsv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, (const void *)a, *lda, - (void *)x, *incx); + const int* n, const int* k, const std::complex* a, const int* lda, + std::complex* x, const int* incx) { + cblas_ztbsv_wrapper(layout, upper_lower, trans, unit_diag, *n, *k, (const void*)a, *lda, + (void*)x, *incx); } template static void tpmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int *n, const fp *a, fp *x, const int *incx); + CBLAS_DIAG unit_diag, const int* n, const fp* a, fp* x, const int* incx); template <> void tpmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const float *a, float *x, const int *incx) { + const int* n, const float* a, float* x, const int* incx) { cblas_stpmv_wrapper(layout, upper_lower, trans, unit_diag, *n, a, x, *incx); } template <> void tpmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const double *a, double *x, const int *incx) { + const int* n, const double* a, double* x, const int* incx) { cblas_dtpmv_wrapper(layout, upper_lower, trans, unit_diag, *n, a, x, *incx); } template <> void tpmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const std::complex *a, std::complex *x, const int *incx) { - cblas_ctpmv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void *)a, (void *)x, - *incx); + const int* n, const std::complex* a, std::complex* x, const int* incx) { + cblas_ctpmv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void*)a, (void*)x, *incx); } template <> void tpmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const std::complex *a, std::complex *x, const int *incx) { - cblas_ztpmv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void *)a, (void *)x, - *incx); + const int* n, const std::complex* a, std::complex* x, const int* incx) { + cblas_ztpmv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void*)a, (void*)x, *incx); } template static void tpsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int *n, const fp *a, fp *x, const int *incx); + CBLAS_DIAG unit_diag, const int* n, const fp* a, fp* x, const int* incx); template <> void tpsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const float *a, float *x, const int *incx) { + const int* n, const float* a, float* x, const int* incx) { cblas_stpsv_wrapper(layout, upper_lower, trans, unit_diag, *n, a, x, *incx); } template <> void tpsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const double *a, double *x, const int *incx) { + const int* n, const double* a, double* x, const int* incx) { cblas_dtpsv_wrapper(layout, upper_lower, trans, unit_diag, *n, a, x, *incx); } template <> void tpsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const std::complex *a, std::complex *x, const int *incx) { - cblas_ctpsv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void *)a, (void *)x, - *incx); + const int* n, const std::complex* a, std::complex* x, const int* incx) { + cblas_ctpsv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void*)a, (void*)x, *incx); } template <> void tpsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const std::complex *a, std::complex *x, const int *incx) { - cblas_ztpsv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void *)a, (void *)x, - *incx); + const int* n, const std::complex* a, std::complex* x, const int* incx) { + cblas_ztpsv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void*)a, (void*)x, *incx); } template static void trmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int *n, const fp *a, const int *lda, fp *x, - const int *incx); + CBLAS_DIAG unit_diag, const int* n, const fp* a, const int* lda, fp* x, + const int* incx); template <> void trmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const float *a, const int *lda, float *x, const int *incx) { + const int* n, const float* a, const int* lda, float* x, const int* incx) { cblas_strmv_wrapper(layout, upper_lower, trans, unit_diag, *n, a, *lda, x, *incx); } template <> void trmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const double *a, const int *lda, double *x, const int *incx) { + const int* n, const double* a, const int* lda, double* x, const int* incx) { cblas_dtrmv_wrapper(layout, upper_lower, trans, unit_diag, *n, a, *lda, x, *incx); } template <> void trmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const std::complex *a, const int *lda, std::complex *x, - const int *incx) { - cblas_ctrmv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void *)a, *lda, (void *)x, + const int* n, const std::complex* a, const int* lda, std::complex* x, + const int* incx) { + cblas_ctrmv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void*)a, *lda, (void*)x, *incx); } template <> void trmv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const std::complex *a, const int *lda, std::complex *x, - const int *incx) { - cblas_ztrmv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void *)a, *lda, (void *)x, + const int* n, const std::complex* a, const int* lda, std::complex* x, + const int* incx) { + cblas_ztrmv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void*)a, *lda, (void*)x, *incx); } template static void trsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int *n, const fp *a, const int *lda, fp *x, - const int *incx); + CBLAS_DIAG unit_diag, const int* n, const fp* a, const int* lda, fp* x, + const int* incx); template <> void trsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const float *a, const int *lda, float *x, const int *incx) { + const int* n, const float* a, const int* lda, float* x, const int* incx) { cblas_strsv_wrapper(layout, upper_lower, trans, unit_diag, *n, a, *lda, x, *incx); } template <> void trsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const double *a, const int *lda, double *x, const int *incx) { + const int* n, const double* a, const int* lda, double* x, const int* incx) { cblas_dtrsv_wrapper(layout, upper_lower, trans, unit_diag, *n, a, *lda, x, *incx); } template <> void trsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const std::complex *a, const int *lda, std::complex *x, - const int *incx) { - cblas_ctrsv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void *)a, *lda, (void *)x, + const int* n, const std::complex* a, const int* lda, std::complex* x, + const int* incx) { + cblas_ctrsv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void*)a, *lda, (void*)x, *incx); } template <> void trsv(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, - const int *n, const std::complex *a, const int *lda, std::complex *x, - const int *incx) { - cblas_ztrsv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void *)a, *lda, (void *)x, + const int* n, const std::complex* a, const int* lda, std::complex* x, + const int* incx) { + cblas_ztrsv_wrapper(layout, upper_lower, trans, unit_diag, *n, (const void*)a, *lda, (void*)x, *incx); } /* Level 1 */ template -static fp_res asum(const int *n, const fp_data *x, const int *incx); +static fp_res asum(const int* n, const fp_data* x, const int* incx); template <> -float asum(const int *n, const float *x, const int *incx) { +float asum(const int* n, const float* x, const int* incx) { return cblas_sasum_wrapper(*n, x, *incx); } template <> -double asum(const int *n, const double *x, const int *incx) { +double asum(const int* n, const double* x, const int* incx) { return cblas_dasum_wrapper(*n, x, *incx); } template <> -float asum(const int *n, const std::complex *x, const int *incx) { - return cblas_scasum_wrapper(*n, (const void *)x, *incx); +float asum(const int* n, const std::complex* x, const int* incx) { + return cblas_scasum_wrapper(*n, (const void*)x, *incx); } template <> -double asum(const int *n, const std::complex *x, const int *incx) { - return cblas_dzasum_wrapper(*n, (const void *)x, *incx); +double asum(const int* n, const std::complex* x, const int* incx) { + return cblas_dzasum_wrapper(*n, (const void*)x, *incx); } template -static void axpy(const int *n, const fp *alpha, const fp *x, const int *incx, fp *y, - const int *incy); +static void axpy(const int* n, const fp* alpha, const fp* x, const int* incx, fp* y, + const int* incy); template <> -void axpy(const int *n, const float *alpha, const float *x, const int *incx, float *y, - const int *incy) { +void axpy(const int* n, const float* alpha, const float* x, const int* incx, float* y, + const int* incy) { cblas_saxpy_wrapper(*n, *alpha, x, *incx, y, *incy); } template <> -void axpy(const int *n, const double *alpha, const double *x, const int *incx, double *y, - const int *incy) { +void axpy(const int* n, const double* alpha, const double* x, const int* incx, double* y, + const int* incy) { cblas_daxpy_wrapper(*n, *alpha, x, *incx, y, *incy); } template <> -void axpy(const int *n, const std::complex *alpha, const std::complex *x, - const int *incx, std::complex *y, const int *incy) { - cblas_caxpy_wrapper(*n, (const void *)alpha, (const void *)x, *incx, (void *)y, *incy); +void axpy(const int* n, const std::complex* alpha, const std::complex* x, + const int* incx, std::complex* y, const int* incy) { + cblas_caxpy_wrapper(*n, (const void*)alpha, (const void*)x, *incx, (void*)y, *incy); } template <> -void axpy(const int *n, const std::complex *alpha, const std::complex *x, - const int *incx, std::complex *y, const int *incy) { - cblas_zaxpy_wrapper(*n, (const void *)alpha, (const void *)x, *incx, (void *)y, *incy); +void axpy(const int* n, const std::complex* alpha, const std::complex* x, + const int* incx, std::complex* y, const int* incy) { + cblas_zaxpy_wrapper(*n, (const void*)alpha, (const void*)x, *incx, (void*)y, *incy); } template -static void copy(const int *n, const fp *x, const int *incx, fp *y, const int *incy); +static void copy(const int* n, const fp* x, const int* incx, fp* y, const int* incy); template <> -void copy(const int *n, const float *x, const int *incx, float *y, const int *incy) { +void copy(const int* n, const float* x, const int* incx, float* y, const int* incy) { cblas_scopy_wrapper(*n, x, *incx, y, *incy); } template <> -void copy(const int *n, const double *x, const int *incx, double *y, const int *incy) { +void copy(const int* n, const double* x, const int* incx, double* y, const int* incy) { cblas_dcopy_wrapper(*n, x, *incx, y, *incy); } template <> -void copy(const int *n, const std::complex *x, const int *incx, std::complex *y, - const int *incy) { - cblas_ccopy_wrapper(*n, (const void *)x, *incx, (void *)y, *incy); +void copy(const int* n, const std::complex* x, const int* incx, std::complex* y, + const int* incy) { + cblas_ccopy_wrapper(*n, (const void*)x, *incx, (void*)y, *incy); } template <> -void copy(const int *n, const std::complex *x, const int *incx, std::complex *y, - const int *incy) { - cblas_zcopy_wrapper(*n, (const void *)x, *incx, (void *)y, *incy); +void copy(const int* n, const std::complex* x, const int* incx, std::complex* y, + const int* incy) { + cblas_zcopy_wrapper(*n, (const void*)x, *incx, (void*)y, *incy); } template -static fp_res dot(const int *n, const fp *x, const int *incx, const fp *y, const int *incy); +static fp_res dot(const int* n, const fp* x, const int* incx, const fp* y, const int* incy); template <> -float dot(const int *n, const float *x, const int *incx, const float *y, const int *incy) { +float dot(const int* n, const float* x, const int* incx, const float* y, const int* incy) { return cblas_sdot_wrapper(*n, x, *incx, y, *incy); } template <> -double dot(const int *n, const double *x, const int *incx, const double *y, const int *incy) { +double dot(const int* n, const double* x, const int* incx, const double* y, const int* incy) { return cblas_ddot_wrapper(*n, x, *incx, y, *incy); } template <> -double dot(const int *n, const float *x, const int *incx, const float *y, const int *incy) { +double dot(const int* n, const float* x, const int* incx, const float* y, const int* incy) { return cblas_dsdot_wrapper(*n, x, *incx, y, *incy); } -static float sdsdot(const int *n, const float *sb, const float *x, const int *incx, const float *y, - const int *incy) { +static float sdsdot(const int* n, const float* sb, const float* x, const int* incx, const float* y, + const int* incy) { return cblas_sdsdot_wrapper(*n, *sb, x, *incx, y, *incy); } template -static fp_res nrm2(const int *n, const fp *x, const int *incx); +static fp_res nrm2(const int* n, const fp* x, const int* incx); template <> -float nrm2(const int *n, const float *x, const int *incx) { +float nrm2(const int* n, const float* x, const int* incx) { return cblas_snrm2_wrapper(*n, x, *incx); } template <> -double nrm2(const int *n, const double *x, const int *incx) { +double nrm2(const int* n, const double* x, const int* incx) { return cblas_dnrm2_wrapper(*n, x, *incx); } template <> -float nrm2(const int *n, const std::complex *x, const int *incx) { - return cblas_scnrm2_wrapper(*n, (const void *)x, *incx); +float nrm2(const int* n, const std::complex* x, const int* incx) { + return cblas_scnrm2_wrapper(*n, (const void*)x, *incx); } template <> -double nrm2(const int *n, const std::complex *x, const int *incx) { - return cblas_dznrm2_wrapper(*n, (const void *)x, *incx); +double nrm2(const int* n, const std::complex* x, const int* incx) { + return cblas_dznrm2_wrapper(*n, (const void*)x, *incx); } template -static void rot(const int *n, fp *x, const int *incx, fp *y, const int *incy, const fp_scalar *c, - const fp_scalar *s); +static void rot(const int* n, fp* x, const int* incx, fp* y, const int* incy, const fp_scalar* c, + const fp_scalar* s); template <> -void rot(const int *n, float *x, const int *incx, float *y, const int *incy, const float *c, - const float *s) { +void rot(const int* n, float* x, const int* incx, float* y, const int* incy, const float* c, + const float* s) { cblas_srot_wrapper(*n, x, *incx, y, *incy, *c, *s); } template <> -void rot(const int *n, double *x, const int *incx, double *y, const int *incy, const double *c, - const double *s) { +void rot(const int* n, double* x, const int* incx, double* y, const int* incy, const double* c, + const double* s) { cblas_drot_wrapper(*n, x, *incx, y, *incy, *c, *s); } template <> -void rot(const int *n, std::complex *x, const int *incx, std::complex *y, - const int *incy, const float *c, const float *s) { - csrot_wrapper(n, (void *)x, incx, (void *)y, incy, c, s); +void rot(const int* n, std::complex* x, const int* incx, std::complex* y, + const int* incy, const float* c, const float* s) { + csrot_wrapper(n, (void*)x, incx, (void*)y, incy, c, s); } template <> -void rot(const int *n, std::complex *x, const int *incx, std::complex *y, - const int *incy, const double *c, const double *s) { - zdrot_wrapper(n, (void *)x, incx, (void *)y, incy, c, s); +void rot(const int* n, std::complex* x, const int* incx, std::complex* y, + const int* incy, const double* c, const double* s) { + zdrot_wrapper(n, (void*)x, incx, (void*)y, incy, c, s); } template -static void rotg(fp *a, fp *b, fp_c *c, fp *s); +static void rotg(fp* a, fp* b, fp_c* c, fp* s); template <> -void rotg(float *a, float *b, float *c, float *s) { +void rotg(float* a, float* b, float* c, float* s) { cblas_srotg_wrapper(a, b, c, s); } template <> -void rotg(double *a, double *b, double *c, double *s) { +void rotg(double* a, double* b, double* c, double* s) { cblas_drotg_wrapper(a, b, c, s); } template <> -void rotg(std::complex *a, std::complex *b, float *c, std::complex *s) { - crotg_wrapper((void *)a, (void *)b, c, (void *)s); +void rotg(std::complex* a, std::complex* b, float* c, std::complex* s) { + crotg_wrapper((void*)a, (void*)b, c, (void*)s); } template <> -void rotg(std::complex *a, std::complex *b, double *c, std::complex *s) { - zrotg_wrapper((void *)a, (void *)b, c, (void *)s); +void rotg(std::complex* a, std::complex* b, double* c, std::complex* s) { + zrotg_wrapper((void*)a, (void*)b, c, (void*)s); } template -static void rotm(const int *n, fp *x, const int *incx, fp *y, const int *incy, const fp *param); +static void rotm(const int* n, fp* x, const int* incx, fp* y, const int* incy, const fp* param); template <> -void rotm(const int *n, float *x, const int *incx, float *y, const int *incy, const float *param) { +void rotm(const int* n, float* x, const int* incx, float* y, const int* incy, const float* param) { cblas_srotm_wrapper(*n, x, *incx, y, *incy, param); } template <> -void rotm(const int *n, double *x, const int *incx, double *y, const int *incy, - const double *param) { +void rotm(const int* n, double* x, const int* incx, double* y, const int* incy, + const double* param) { cblas_drotm_wrapper(*n, x, *incx, y, *incy, param); } template -static void rotmg(fp *d1, fp *d2, fp *x1, fp *y1, fp *param); +static void rotmg(fp* d1, fp* d2, fp* x1, fp* y1, fp* param); template <> -void rotmg(float *d1, float *d2, float *x1, float *y1, float *param) { +void rotmg(float* d1, float* d2, float* x1, float* y1, float* param) { cblas_srotmg_wrapper(d1, d2, x1, *y1, param); } template <> -void rotmg(double *d1, double *d2, double *x1, double *y1, double *param) { +void rotmg(double* d1, double* d2, double* x1, double* y1, double* param) { cblas_drotmg_wrapper(d1, d2, x1, *y1, param); } template -static void scal(const int *n, const fp_scalar *alpha, fp_data *x, const int *incx); +static void scal(const int* n, const fp_scalar* alpha, fp_data* x, const int* incx); template <> -void scal(const int *n, const float *alpha, float *x, const int *incx) { +void scal(const int* n, const float* alpha, float* x, const int* incx) { cblas_sscal_wrapper(*n, *alpha, x, *incx); } template <> -void scal(const int *n, const double *alpha, double *x, const int *incx) { +void scal(const int* n, const double* alpha, double* x, const int* incx) { cblas_dscal_wrapper(*n, *alpha, x, *incx); } template <> -void scal(const int *n, const std::complex *alpha, std::complex *x, const int *incx) { - cblas_cscal_wrapper(*n, (const void *)alpha, (void *)x, *incx); +void scal(const int* n, const std::complex* alpha, std::complex* x, const int* incx) { + cblas_cscal_wrapper(*n, (const void*)alpha, (void*)x, *incx); } template <> -void scal(const int *n, const std::complex *alpha, std::complex *x, - const int *incx) { - cblas_zscal_wrapper(*n, (const void *)alpha, (void *)x, *incx); +void scal(const int* n, const std::complex* alpha, std::complex* x, + const int* incx) { + cblas_zscal_wrapper(*n, (const void*)alpha, (void*)x, *incx); } template <> -void scal(const int *n, const float *alpha, std::complex *x, const int *incx) { - cblas_csscal_wrapper(*n, *alpha, (void *)x, *incx); +void scal(const int* n, const float* alpha, std::complex* x, const int* incx) { + cblas_csscal_wrapper(*n, *alpha, (void*)x, *incx); } template <> -void scal(const int *n, const double *alpha, std::complex *x, const int *incx) { - cblas_zdscal_wrapper(*n, *alpha, (void *)x, *incx); +void scal(const int* n, const double* alpha, std::complex* x, const int* incx) { + cblas_zdscal_wrapper(*n, *alpha, (void*)x, *incx); } template -static void swap(const int *n, fp *x, const int *incx, fp *y, const int *incy); +static void swap(const int* n, fp* x, const int* incx, fp* y, const int* incy); template <> -void swap(const int *n, float *x, const int *incx, float *y, const int *incy) { +void swap(const int* n, float* x, const int* incx, float* y, const int* incy) { cblas_sswap_wrapper(*n, x, *incx, y, *incy); } template <> -void swap(const int *n, double *x, const int *incx, double *y, const int *incy) { +void swap(const int* n, double* x, const int* incx, double* y, const int* incy) { cblas_dswap_wrapper(*n, x, *incx, y, *incy); } template <> -void swap(const int *n, std::complex *x, const int *incx, std::complex *y, - const int *incy) { - cblas_cswap_wrapper(*n, (void *)x, *incx, (void *)y, *incy); +void swap(const int* n, std::complex* x, const int* incx, std::complex* y, + const int* incy) { + cblas_cswap_wrapper(*n, (void*)x, *incx, (void*)y, *incy); } template <> -void swap(const int *n, std::complex *x, const int *incx, std::complex *y, - const int *incy) { - cblas_zswap_wrapper(*n, (void *)x, *incx, (void *)y, *incy); +void swap(const int* n, std::complex* x, const int* incx, std::complex* y, + const int* incy) { + cblas_zswap_wrapper(*n, (void*)x, *incx, (void*)y, *incy); } template -static void dotc(fp *pres, const int *n, const fp *x, const int *incx, const fp *y, - const int *incy); +static void dotc(fp* pres, const int* n, const fp* x, const int* incx, const fp* y, + const int* incy); template <> -void dotc(std::complex *pres, const int *n, const std::complex *x, const int *incx, - const std::complex *y, const int *incy) { - cblas_cdotc_sub_wrapper(*n, (const void *)x, *incx, (const void *)y, *incy, (void *)pres); +void dotc(std::complex* pres, const int* n, const std::complex* x, const int* incx, + const std::complex* y, const int* incy) { + cblas_cdotc_sub_wrapper(*n, (const void*)x, *incx, (const void*)y, *incy, (void*)pres); } template <> -void dotc(std::complex *pres, const int *n, const std::complex *x, const int *incx, - const std::complex *y, const int *incy) { - cblas_zdotc_sub_wrapper(*n, (const void *)x, *incx, (const void *)y, *incy, (void *)pres); +void dotc(std::complex* pres, const int* n, const std::complex* x, const int* incx, + const std::complex* y, const int* incy) { + cblas_zdotc_sub_wrapper(*n, (const void*)x, *incx, (const void*)y, *incy, (void*)pres); } template -static void dotu(fp *pres, const int *n, const fp *x, const int *incx, const fp *y, - const int *incy); +static void dotu(fp* pres, const int* n, const fp* x, const int* incx, const fp* y, + const int* incy); template <> -void dotu(std::complex *pres, const int *n, const std::complex *x, const int *incx, - const std::complex *y, const int *incy) { - cblas_cdotu_sub_wrapper(*n, (const void *)x, *incx, (const void *)y, *incy, (void *)pres); +void dotu(std::complex* pres, const int* n, const std::complex* x, const int* incx, + const std::complex* y, const int* incy) { + cblas_cdotu_sub_wrapper(*n, (const void*)x, *incx, (const void*)y, *incy, (void*)pres); } template <> -void dotu(std::complex *pres, const int *n, const std::complex *x, const int *incx, - const std::complex *y, const int *incy) { - cblas_zdotu_sub_wrapper(*n, (const void *)x, *incx, (const void *)y, *incy, (void *)pres); +void dotu(std::complex* pres, const int* n, const std::complex* x, const int* incx, + const std::complex* y, const int* incy) { + cblas_zdotu_sub_wrapper(*n, (const void*)x, *incx, (const void*)y, *incy, (void*)pres); } template -static int iamax(const int *n, const fp *x, const int *incx); +static int iamax(const int* n, const fp* x, const int* incx); template <> -int iamax(const int *n, const float *x, const int *incx) { +int iamax(const int* n, const float* x, const int* incx) { return cblas_isamax_wrapper(*n, x, *incx); } template <> -int iamax(const int *n, const double *x, const int *incx) { +int iamax(const int* n, const double* x, const int* incx) { return cblas_idamax_wrapper(*n, x, *incx); } template <> -int iamax(const int *n, const std::complex *x, const int *incx) { - return cblas_icamax_wrapper(*n, (const void *)x, *incx); +int iamax(const int* n, const std::complex* x, const int* incx) { + return cblas_icamax_wrapper(*n, (const void*)x, *incx); } template <> -int iamax(const int *n, const std::complex *x, const int *incx) { - return cblas_izamax_wrapper(*n, (const void *)x, *incx); +int iamax(const int* n, const std::complex* x, const int* incx) { + return cblas_izamax_wrapper(*n, (const void*)x, *incx); } inline float abs_val(float val) { @@ -1425,10 +1421,10 @@ inline double abs_val(std::complex val) { } template -static int iamin(const int *n, const fp *x, const int *incx); +static int iamin(const int* n, const fp* x, const int* incx); template <> -int iamin(const int *n, const float *x, const int *incx) { +int iamin(const int* n, const float* x, const int* incx) { if (*n < 1 || *incx < 1) { return 0; } @@ -1451,7 +1447,7 @@ int iamin(const int *n, const float *x, const int *incx) { } template <> -int iamin(const int *n, const double *x, const int *incx) { +int iamin(const int* n, const double* x, const int* incx) { if (*n < 1 || *incx < 1) { return 0; } @@ -1474,7 +1470,7 @@ int iamin(const int *n, const double *x, const int *incx) { } template <> -int iamin(const int *n, const std::complex *x, const int *incx) { +int iamin(const int* n, const std::complex* x, const int* incx) { if (*n < 1 || *incx < 1) { return 0; } @@ -1497,7 +1493,7 @@ int iamin(const int *n, const std::complex *x, const int *incx) { } template <> -int iamin(const int *n, const std::complex *x, const int *incx) { +int iamin(const int* n, const std::complex* x, const int* incx) { if (*n < 1 || *incx < 1) { return 0; } @@ -1522,12 +1518,12 @@ int iamin(const int *n, const std::complex *x, const int *incx) { /* Extensions */ template -static void axpby(const int *n, const fp *alpha, const fp *x, const int *incx, const fp *beta, - fp *y, const int *incy); +static void axpby(const int* n, const fp* alpha, const fp* x, const int* incx, const fp* beta, + fp* y, const int* incy); template <> -void axpby(const int *n, const float *alpha, const float *x, const int *incx, const float *beta, - float *y, const int *incy) { +void axpby(const int* n, const float* alpha, const float* x, const int* incx, const float* beta, + float* y, const int* incy) { // Not supported in NETLIB. Reference C++ implementation is used. int idx = (*incx) > 0 ? 0 : (1 - *n) * (*incx); int idy = (*incy) > 0 ? 0 : (1 - *n) * (*incy); @@ -1536,8 +1532,8 @@ void axpby(const int *n, const float *alpha, const float *x, const int *incx, co } template <> -void axpby(const int *n, const double *alpha, const double *x, const int *incx, const double *beta, - double *y, const int *incy) { +void axpby(const int* n, const double* alpha, const double* x, const int* incx, const double* beta, + double* y, const int* incy) { // Not supported in NETLIB. Reference C++ implementation is used. int idx = (*incx) > 0 ? 0 : (1 - *n) * (*incx); int idy = (*incy) > 0 ? 0 : (1 - *n) * (*incy); @@ -1546,9 +1542,9 @@ void axpby(const int *n, const double *alpha, const double *x, const int *incx, } template <> -void axpby(const int *n, const std::complex *alpha, const std::complex *x, - const int *incx, const std::complex *beta, std::complex *y, - const int *incy) { +void axpby(const int* n, const std::complex* alpha, const std::complex* x, + const int* incx, const std::complex* beta, std::complex* y, + const int* incy) { // Not supported in NETLIB. Reference C++ implementation is used. int idx = (*incx) > 0 ? 0 : (1 - *n) * (*incx); int idy = (*incy) > 0 ? 0 : (1 - *n) * (*incy); @@ -1557,9 +1553,9 @@ void axpby(const int *n, const std::complex *alpha, const std::complex -void axpby(const int *n, const std::complex *alpha, const std::complex *x, - const int *incx, const std::complex *beta, std::complex *y, - const int *incy) { +void axpby(const int* n, const std::complex* alpha, const std::complex* x, + const int* incx, const std::complex* beta, std::complex* y, + const int* incy) { // Not supported in NETLIB. Reference C++ implementation is used. int idx = (*incx) > 0 ? 0 : (1 - *n) * (*incx); int idy = (*incy) > 0 ? 0 : (1 - *n) * (*incy); @@ -1569,16 +1565,16 @@ void axpby(const int *n, const std::complex *alpha, const std::complex static void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - CBLAS_OFFSET offsetc, const int *m, const int *n, const int *k, - const fps *alpha, const fpa *a, const int *lda, const fpa *ao, const fpb *b, - const int *ldb, const fpb *bo, const fps *beta, fpc *c, const int *ldc, - const fpc *co); + CBLAS_OFFSET offsetc, const int* m, const int* n, const int* k, + const fps* alpha, const fpa* a, const int* lda, const fpa* ao, const fpb* b, + const int* ldb, const fpb* bo, const fps* beta, fpc* c, const int* ldc, + const fpc* co); template <> void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - CBLAS_OFFSET offsetc, const int *m, const int *n, const int *k, const float *alpha, - const int8_t *a, const int *lda, const int8_t *ao, const int8_t *b, const int *ldb, - const int8_t *bo, const float *beta, int32_t *c, const int *ldc, const int32_t *co) { + CBLAS_OFFSET offsetc, const int* m, const int* n, const int* k, const float* alpha, + const int8_t* a, const int* lda, const int8_t* ao, const int8_t* b, const int* ldb, + const int8_t* bo, const float* beta, int32_t* c, const int* ldc, const int32_t* co) { // Not supported in NETLIB. DGEMM is used as reference. int sizea, sizeb, sizec; if (layout == CblasColMajor) { @@ -1591,9 +1587,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; sizec = *ldc * *m; } - double *ad = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); - double *bd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); - double *cd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); + double* ad = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); + double* bd = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); + double* cd = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); double alphad = *alpha; double betad = *beta; double aod = *ao; @@ -1611,10 +1607,10 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran template <> void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - CBLAS_OFFSET offsetc, const int *m, const int *n, const int *k, const float *alpha, - const int8_t *a, const int *lda, const int8_t *ao, const uint8_t *b, const int *ldb, - const uint8_t *bo, const float *beta, int32_t *c, const int *ldc, - const int32_t *co) { + CBLAS_OFFSET offsetc, const int* m, const int* n, const int* k, const float* alpha, + const int8_t* a, const int* lda, const int8_t* ao, const uint8_t* b, const int* ldb, + const uint8_t* bo, const float* beta, int32_t* c, const int* ldc, + const int32_t* co) { // Not supported in NETLIB. DGEMM is used as reference. int sizea, sizeb, sizec; if (layout == CblasColMajor) { @@ -1627,9 +1623,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; sizec = *ldc * *m; } - double *ad = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); - double *bd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); - double *cd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); + double* ad = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); + double* bd = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); + double* cd = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); double alphad = *alpha; double betad = *beta; double aod = *ao; @@ -1647,9 +1643,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran template <> void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - CBLAS_OFFSET offsetc, const int *m, const int *n, const int *k, const float *alpha, - const uint8_t *a, const int *lda, const uint8_t *ao, const int8_t *b, const int *ldb, - const int8_t *bo, const float *beta, int32_t *c, const int *ldc, const int32_t *co) { + CBLAS_OFFSET offsetc, const int* m, const int* n, const int* k, const float* alpha, + const uint8_t* a, const int* lda, const uint8_t* ao, const int8_t* b, const int* ldb, + const int8_t* bo, const float* beta, int32_t* c, const int* ldc, const int32_t* co) { // Not supported in NETLIB. DGEMM is used as reference. int sizea, sizeb, sizec; if (layout == CblasColMajor) { @@ -1662,9 +1658,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; sizec = *ldc * *m; } - double *ad = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); - double *bd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); - double *cd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); + double* ad = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); + double* bd = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); + double* cd = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); double alphad = *alpha; double betad = *beta; double aod = *ao; @@ -1682,10 +1678,10 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran template <> void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - CBLAS_OFFSET offsetc, const int *m, const int *n, const int *k, const float *alpha, - const uint8_t *a, const int *lda, const uint8_t *ao, const uint8_t *b, - const int *ldb, const uint8_t *bo, const float *beta, int32_t *c, const int *ldc, - const int32_t *co) { + CBLAS_OFFSET offsetc, const int* m, const int* n, const int* k, const float* alpha, + const uint8_t* a, const int* lda, const uint8_t* ao, const uint8_t* b, + const int* ldb, const uint8_t* bo, const float* beta, int32_t* c, const int* ldc, + const int32_t* co) { // Not supported in NETLIB. DGEMM is used as reference. int sizea, sizeb, sizec; if (layout == CblasColMajor) { @@ -1698,9 +1694,9 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran sizeb = (transb == CblasNoTrans) ? *ldb * *k : *ldb * *n; sizec = *ldc * *m; } - double *ad = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); - double *bd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); - double *cd = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); + double* ad = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizea); + double* bd = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizeb); + double* cd = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); double alphad = *alpha; double betad = *beta; double aod = *ao; @@ -1718,19 +1714,19 @@ void gemm_bias(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE tran template static void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, - CBLAS_TRANSPOSE transb, const int *n, const int *k, const fp *alpha, const fp *a, - const int *lda, const fp *b, const int *ldb, const fp *beta, fp *c, - const int *ldc); + CBLAS_TRANSPOSE transb, const int* n, const int* k, const fp* alpha, const fp* a, + const int* lda, const fp* b, const int* ldb, const fp* beta, fp* c, + const int* ldc); template <> void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, - CBLAS_TRANSPOSE transb, const int *n, const int *k, const float *alpha, const float *a, - const int *lda, const float *b, const int *ldb, const float *beta, float *c, - const int *ldc) { + CBLAS_TRANSPOSE transb, const int* n, const int* k, const float* alpha, const float* a, + const int* lda, const float* b, const int* ldb, const float* beta, float* c, + const int* ldc) { // Not supported in NETLIB. SGEMM is used as reference. int sizec; sizec = *ldc * *n; - float *cf = (float *)oneapi::math::aligned_alloc(64, sizeof(float) * sizec); + float* cf = (float*)oneapi::math::aligned_alloc(64, sizeof(float) * sizec); update_c(c, layout, upper_lower, *n, *n, *ldc, cf); cblas_sgemm_wrapper(layout, transa, transb, *n, *n, *k, *alpha, a, *lda, b, *ldb, *beta, cf, *ldc); @@ -1740,13 +1736,13 @@ void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, template <> void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, - CBLAS_TRANSPOSE transb, const int *n, const int *k, const double *alpha, const double *a, - const int *lda, const double *b, const int *ldb, const double *beta, double *c, - const int *ldc) { + CBLAS_TRANSPOSE transb, const int* n, const int* k, const double* alpha, const double* a, + const int* lda, const double* b, const int* ldb, const double* beta, double* c, + const int* ldc) { // Not supported in NETLIB. DGEMM is used as reference. int sizec; sizec = *ldc * *n; - double *cf = (double *)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); + double* cf = (double*)oneapi::math::aligned_alloc(64, sizeof(double) * sizec); update_c(c, layout, upper_lower, *n, *n, *ldc, cf); cblas_dgemm_wrapper(layout, transa, transb, *n, *n, *k, *alpha, a, *lda, b, *ldb, *beta, cf, *ldc); @@ -1756,15 +1752,15 @@ void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, template <> void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, - CBLAS_TRANSPOSE transb, const int *n, const int *k, const std::complex *alpha, - const std::complex *a, const int *lda, const std::complex *b, - const int *ldb, const std::complex *beta, std::complex *c, - const int *ldc) { + CBLAS_TRANSPOSE transb, const int* n, const int* k, const std::complex* alpha, + const std::complex* a, const int* lda, const std::complex* b, + const int* ldb, const std::complex* beta, std::complex* c, + const int* ldc) { // Not supported in NETLIB. CGEMM is used as reference. int sizec; sizec = *ldc * *n; - std::complex *cf = - (std::complex *)oneapi::math::aligned_alloc(64, sizeof(std::complex) * sizec); + std::complex* cf = + (std::complex*)oneapi::math::aligned_alloc(64, sizeof(std::complex) * sizec); update_c(c, layout, upper_lower, *n, *n, *ldc, cf); cblas_cgemm_wrapper(layout, transa, transb, *n, *n, *k, alpha, a, *lda, b, *ldb, beta, cf, *ldc); @@ -1774,14 +1770,14 @@ void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, template <> void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, - CBLAS_TRANSPOSE transb, const int *n, const int *k, const std::complex *alpha, - const std::complex *a, const int *lda, const std::complex *b, - const int *ldb, const std::complex *beta, std::complex *c, - const int *ldc) { + CBLAS_TRANSPOSE transb, const int* n, const int* k, const std::complex* alpha, + const std::complex* a, const int* lda, const std::complex* b, + const int* ldb, const std::complex* beta, std::complex* c, + const int* ldc) { // Not supported in NETLIB. ZGEMM is used as reference. int sizec; sizec = *ldc * *n; - std::complex *cf = (std::complex *)oneapi::math::aligned_alloc( + std::complex* cf = (std::complex*)oneapi::math::aligned_alloc( 64, sizeof(std::complex) * sizec); update_c(c, layout, upper_lower, *n, *n, *ldc, cf); cblas_zgemm_wrapper(layout, transa, transb, *n, *n, *k, alpha, a, *lda, b, *ldb, beta, cf, @@ -1791,12 +1787,12 @@ void gemmt(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE transa, } template -static void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int *m, const int *n, - const fp *a, const int *lda, const fp *x, const int *incx, fp *c, const int *ldc); +static void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int* m, const int* n, + const fp* a, const int* lda, const fp* x, const int* incx, fp* c, const int* ldc); template <> -void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int *m, const int *n, const float *a, - const int *lda, const float *x, const int *incx, float *c, const int *ldc) { +void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int* m, const int* n, const float* a, + const int* lda, const float* x, const int* incx, float* c, const int* ldc) { // Not supported in NETLIB. Reference C++ implementation is used. float tmp; int size_x = (left_right == CblasLeft) ? *m : *n; @@ -1827,8 +1823,8 @@ void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int *m, const int *n } template <> -void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int *m, const int *n, const double *a, - const int *lda, const double *x, const int *incx, double *c, const int *ldc) { +void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int* m, const int* n, const double* a, + const int* lda, const double* x, const int* incx, double* c, const int* ldc) { // Not supported in NETLIB. Reference C++ implementation is used. double tmp; int size_x = (left_right == CblasLeft) ? *m : *n; @@ -1859,9 +1855,9 @@ void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int *m, const int *n } template <> -void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int *m, const int *n, - const std::complex *a, const int *lda, const std::complex *x, - const int *incx, std::complex *c, const int *ldc) { +void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int* m, const int* n, + const std::complex* a, const int* lda, const std::complex* x, + const int* incx, std::complex* c, const int* ldc) { // Not supported in NETLIB. Reference C++ implementation is used. std::complex tmp; int size_x = (left_right == CblasLeft) ? *m : *n; @@ -1912,9 +1908,9 @@ void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int *m, const int *n } template <> -void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int *m, const int *n, - const std::complex *a, const int *lda, const std::complex *x, - const int *incx, std::complex *c, const int *ldc) { +void dgmm(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, const int* m, const int* n, + const std::complex* a, const int* lda, const std::complex* x, + const int* incx, std::complex* c, const int* ldc) { // Not supported in NETLIB. Reference C++ implementation is used. std::complex tmp; int size_x = (left_right == CblasLeft) ? *m : *n; @@ -1979,7 +1975,7 @@ fp sametype_conj(fp x) { template void omatcopy_ref(oneapi::math::layout layout, oneapi::math::transpose trans, int64_t m, int64_t n, - fp alpha, fp *A, int64_t lda, fp *B, int64_t ldb) { + fp alpha, fp* A, int64_t lda, fp* B, int64_t ldb) { int64_t logical_m, logical_n; if (layout == oneapi::math::layout::col_major) { logical_m = m; @@ -2014,9 +2010,9 @@ void omatcopy_ref(oneapi::math::layout layout, oneapi::math::transpose trans, in } template -void omatcopy2_ref(oneapi::math::layout layout, oneapi::math::transpose trans, const int64_t &m, - const int64_t &n, const fp &alpha, const fp *in_matrix, const int64_t &ld_in, - const int64_t &inc_in, fp *out_matrix, const int64_t &ld_out, +void omatcopy2_ref(oneapi::math::layout layout, oneapi::math::transpose trans, const int64_t& m, + const int64_t& n, const fp& alpha, const fp* in_matrix, const int64_t& ld_in, + const int64_t& inc_in, fp* out_matrix, const int64_t& ld_out, const int64_t inc_out) { int64_t logical_m, logical_n; if (layout == oneapi::math::layout::col_major) { @@ -2061,7 +2057,7 @@ void omatcopy2_ref(oneapi::math::layout layout, oneapi::math::transpose trans, c template void imatcopy_ref(oneapi::math::layout layout, oneapi::math::transpose trans, int64_t m, int64_t n, - fp alpha, fp *A, int64_t lda, int64_t ldb) { + fp alpha, fp* A, int64_t lda, int64_t ldb) { int64_t logical_m, logical_n; if (layout == oneapi::math::layout::col_major) { logical_m = m; @@ -2115,8 +2111,8 @@ void imatcopy_ref(oneapi::math::layout layout, oneapi::math::transpose trans, in template void omatadd_ref(oneapi::math::layout layout, oneapi::math::transpose transa, - oneapi::math::transpose transb, int64_t m, int64_t n, fp alpha, fp *A, int64_t lda, - fp beta, fp *B, int64_t ldb, fp *C, int64_t ldc) { + oneapi::math::transpose transb, int64_t m, int64_t n, fp alpha, fp* A, int64_t lda, + fp beta, fp* B, int64_t ldb, fp* C, int64_t ldc) { int64_t logical_m, logical_n; if (layout == oneapi::math::layout::col_major) { logical_m = m; diff --git a/tests/unit_tests/blas/include/reference_blas_wrappers.hpp b/tests/unit_tests/blas/include/reference_blas_wrappers.hpp index 0ba5506c0..d00d20947 100644 --- a/tests/unit_tests/blas/include/reference_blas_wrappers.hpp +++ b/tests/unit_tests/blas/include/reference_blas_wrappers.hpp @@ -27,7 +27,7 @@ #ifdef __linux__ #include -#define LIB_TYPE void * +#define LIB_TYPE void* #define GET_LIB_HANDLE(libname) dlopen((libname), RTLD_LAZY | RTLD_LOCAL | RTLD_DEEPBIND) #define GET_FUNC(lib, fn) dlsym(lib, (fn)) #elif defined(_WIN64) @@ -68,129 +68,129 @@ static LIB_TYPE cblas_library() { static void (*cblas_sgemm_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int m, const int n, const int k, const float alpha, - const float *a, const int lda, const float *b, const int ldb, - const float beta, float *c, const int ldc); + const float* a, const int lda, const float* b, const int ldb, + const float beta, float* c, const int ldc); static void (*cblas_dgemm_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int m, const int n, const int k, const double alpha, - const double *a, const int lda, const double *b, const int ldb, - const double beta, double *c, const int ldc); + const double* a, const int lda, const double* b, const int ldb, + const double beta, double* c, const int ldc); static void (*cblas_cgemm_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - const int m, const int n, const int k, const void *alpha, - const void *a, const int lda, const void *b, const int ldb, - const void *beta, void *c, const int ldc); + const int m, const int n, const int k, const void* alpha, + const void* a, const int lda, const void* b, const int ldb, + const void* beta, void* c, const int ldc); static void (*cblas_zgemm_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - const int m, const int n, const int k, const void *alpha, - const void *a, const int lda, const void *b, const int ldb, - const void *beta, void *c, const int ldc); + const int m, const int n, const int k, const void* alpha, + const void* a, const int lda, const void* b, const int ldb, + const void* beta, void* c, const int ldc); static void (*cblas_ssymm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const float alpha, const float *a, - const int lda, const float *b, const int ldb, const float beta, - float *c, const int ldc); + const int m, const int n, const float alpha, const float* a, + const int lda, const float* b, const int ldb, const float beta, + float* c, const int ldc); static void (*cblas_dsymm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const double alpha, const double *a, - const int lda, const double *b, const int ldb, const double beta, - double *c, const int ldc); + const int m, const int n, const double alpha, const double* a, + const int lda, const double* b, const int ldb, const double beta, + double* c, const int ldc); static void (*cblas_csymm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, void *c, + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, void* c, const int ldc); static void (*cblas_zsymm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, void *c, + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, void* c, const int ldc); static void (*cblas_ssyrk_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const float alpha, const float *a, - const int lda, const float beta, float *c, const int ldc); + const int n, const int k, const float alpha, const float* a, + const int lda, const float beta, float* c, const int ldc); static void (*cblas_dsyrk_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const double alpha, const double *a, - const int lda, const double beta, double *c, const int ldc); + const int n, const int k, const double alpha, const double* a, + const int lda, const double beta, double* c, const int ldc); static void (*cblas_csyrk_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *beta, void *c, const int ldc); + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* beta, void* c, const int ldc); static void (*cblas_zsyrk_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *beta, void *c, const int ldc); + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* beta, void* c, const int ldc); static void (*cblas_chemm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, void *c, + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, void* c, const int ldc); static void (*cblas_zhemm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, void *c, + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, void* c, const int ldc); static void (*cblas_cherk_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const float alpha, const void *a, - const int lda, const float beta, void *c, const int ldc); + const int n, const int k, const float alpha, const void* a, + const int lda, const float beta, void* c, const int ldc); static void (*cblas_zherk_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const double alpha, const void *a, - const int lda, const double beta, void *c, const int ldc); + const int n, const int k, const double alpha, const void* a, + const int lda, const double beta, void* c, const int ldc); static void (*cblas_ssyr2k_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const float alpha, const float *a, - const int lda, const float *b, const int ldb, const float beta, - float *c, const int ldc); + const int n, const int k, const float alpha, const float* a, + const int lda, const float* b, const int ldb, const float beta, + float* c, const int ldc); static void (*cblas_dsyr2k_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const double alpha, const double *a, - const int lda, const double *b, const int ldb, const double beta, - double *c, const int ldc); + const int n, const int k, const double alpha, const double* a, + const int lda, const double* b, const int ldb, const double beta, + double* c, const int ldc); static void (*cblas_csyr2k_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc); + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc); static void (*cblas_zsyr2k_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc); + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc); static void (*cblas_cher2k_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const float beta, - void *c, const int ldc); + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const float beta, + void* c, const int ldc); static void (*cblas_zher2k_p)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const double beta, - void *c, const int ldc); + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const double beta, + void* c, const int ldc); static void (*cblas_strmm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const float alpha, const float *a, const int lda, float *b, + const float alpha, const float* a, const int lda, float* b, const int ldb); static void (*cblas_dtrmm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const double alpha, const double *a, const int lda, double *b, + const double alpha, const double* a, const int lda, double* b, const int ldb); static void (*cblas_ctrmm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const void *alpha, const void *a, const int lda, void *b, + const void* alpha, const void* a, const int lda, void* b, const int ldb); static void (*cblas_ztrmm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const void *alpha, const void *a, const int lda, void *b, + const void* alpha, const void* a, const int lda, void* b, const int ldb); static void (*cblas_strsm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const float alpha, const float *a, const int lda, float *b, + const float alpha, const float* a, const int lda, float* b, const int ldb); static void (*cblas_dtrsm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const double alpha, const double *a, const int lda, double *b, + const double alpha, const double* a, const int lda, double* b, const int ldb); static void (*cblas_ctrsm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const void *alpha, const void *a, const int lda, void *b, + const void* alpha, const void* a, const int lda, void* b, const int ldb); static void (*cblas_ztrsm_p)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const void *alpha, const void *a, const int lda, void *b, + const void* alpha, const void* a, const int lda, void* b, const int ldb); static void cblas_sgemm_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int m, const int n, const int k, const float alpha, - const float *a, const int lda, const float *b, const int ldb, - const float beta, float *c, const int ldc) { + const float* a, const int lda, const float* b, const int ldb, + const float beta, float* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_sgemm_p == NULL) cblas_sgemm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - const int m, const int n, const int k, const float alpha, const float *a, - const int lda, const float *b, const int ldb, const float beta, float *c, + const int m, const int n, const int k, const float alpha, const float* a, + const int lda, const float* b, const int ldb, const float beta, float* c, const int ldc))GET_FUNC(h_libcblas, "cblas_sgemm"); if (cblas_sgemm_p != NULL) cblas_sgemm_p(layout, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); @@ -199,14 +199,14 @@ static void cblas_sgemm_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBL static void cblas_dgemm_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int m, const int n, const int k, const double alpha, - const double *a, const int lda, const double *b, const int ldb, - const double beta, double *c, const int ldc) { + const double* a, const int lda, const double* b, const int ldb, + const double beta, double* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_dgemm_p == NULL) cblas_dgemm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, const int m, const int n, const int k, - const double alpha, const double *a, const int lda, - const double *b, const int ldb, const double beta, double *c, + const double alpha, const double* a, const int lda, + const double* b, const int ldb, const double beta, double* c, const int ldc))GET_FUNC(h_libcblas, "cblas_dgemm"); if (cblas_dgemm_p != NULL) cblas_dgemm_p(layout, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); @@ -214,15 +214,15 @@ static void cblas_dgemm_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBL } static void cblas_cgemm_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - const int m, const int n, const int k, const void *alpha, - const void *a, const int lda, const void *b, const int ldb, - const void *beta, void *c, const int ldc) { + const int m, const int n, const int k, const void* alpha, + const void* a, const int lda, const void* b, const int ldb, + const void* beta, void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_cgemm_p == NULL) cblas_cgemm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - const int m, const int n, const int k, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, void *c, + const int m, const int n, const int k, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_cgemm"); if (cblas_cgemm_p != NULL) cblas_cgemm_p(layout, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); @@ -230,15 +230,15 @@ static void cblas_cgemm_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBL } static void cblas_zgemm_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - const int m, const int n, const int k, const void *alpha, - const void *a, const int lda, const void *b, const int ldb, - const void *beta, void *c, const int ldc) { + const int m, const int n, const int k, const void* alpha, + const void* a, const int lda, const void* b, const int ldb, + const void* beta, void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_zgemm_p == NULL) cblas_zgemm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBLAS_TRANSPOSE transb, - const int m, const int n, const int k, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, void *c, + const int m, const int n, const int k, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_zgemm"); if (cblas_zgemm_p != NULL) cblas_zgemm_p(layout, transa, transb, m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); @@ -246,15 +246,15 @@ static void cblas_zgemm_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE transa, CBL } static void cblas_ssymm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const float alpha, const float *a, - const int lda, const float *b, const int ldb, const float beta, - float *c, const int ldc) { + const int m, const int n, const float alpha, const float* a, + const int lda, const float* b, const int ldb, const float beta, + float* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_ssymm_p == NULL) cblas_ssymm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int m, - const int n, const float alpha, const float *a, const int lda, - const float *b, const int ldb, const float beta, float *c, + const int n, const float alpha, const float* a, const int lda, + const float* b, const int ldb, const float beta, float* c, const int ldc))GET_FUNC(h_libcblas, "cblas_ssymm"); if (cblas_ssymm_p != NULL) cblas_ssymm_p(layout, left_right, uplo, m, n, alpha, a, lda, b, ldb, beta, c, ldc); @@ -262,15 +262,15 @@ static void cblas_ssymm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLA } static void cblas_dsymm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const double alpha, const double *a, - const int lda, const double *b, const int ldb, const double beta, - double *c, const int ldc) { + const int m, const int n, const double alpha, const double* a, + const int lda, const double* b, const int ldb, const double beta, + double* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_dsymm_p == NULL) cblas_dsymm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, const int m, - const int n, const double alpha, const double *a, const int lda, - const double *b, const int ldb, const double beta, double *c, + const int n, const double alpha, const double* a, const int lda, + const double* b, const int ldb, const double beta, double* c, const int ldc))GET_FUNC(h_libcblas, "cblas_dsymm"); if (cblas_dsymm_p != NULL) cblas_dsymm_p(layout, left_right, uplo, m, n, alpha, a, lda, b, ldb, beta, c, ldc); @@ -278,43 +278,43 @@ static void cblas_dsymm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLA } static void cblas_csymm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc) { + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_csymm_p == NULL) cblas_csymm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc))GET_FUNC(h_libcblas, "cblas_csymm"); + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_csymm"); if (cblas_csymm_p != NULL) cblas_csymm_p(layout, left_right, uplo, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } } static void cblas_zsymm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc) { + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_zsymm_p == NULL) cblas_zsymm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc))GET_FUNC(h_libcblas, "cblas_zsymm"); + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_zsymm"); if (cblas_zsymm_p != NULL) cblas_zsymm_p(layout, left_right, uplo, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } } static void cblas_ssyrk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const float alpha, const float *a, - const int lda, const float beta, float *c, const int ldc) { + const int n, const int k, const float alpha, const float* a, + const int lda, const float beta, float* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_ssyrk_p == NULL) cblas_ssyrk_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const float alpha, const float *a, - const int lda, const float beta, float *c, + const int n, const int k, const float alpha, const float* a, + const int lda, const float beta, float* c, const int ldc))GET_FUNC(h_libcblas, "cblas_ssyrk"); if (cblas_ssyrk_p != NULL) cblas_ssyrk_p(layout, uplo, trans, n, k, alpha, a, lda, beta, c, ldc); @@ -322,13 +322,13 @@ static void cblas_ssyrk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRAN } static void cblas_dsyrk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const double alpha, const double *a, - const int lda, const double beta, double *c, const int ldc) { + const int n, const int k, const double alpha, const double* a, + const int lda, const double beta, double* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_dsyrk_p == NULL) cblas_dsyrk_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const double alpha, const double *a, - const int lda, const double beta, double *c, + const int n, const int k, const double alpha, const double* a, + const int lda, const double beta, double* c, const int ldc))GET_FUNC(h_libcblas, "cblas_dsyrk"); if (cblas_dsyrk_p != NULL) cblas_dsyrk_p(layout, uplo, trans, n, k, alpha, a, lda, beta, c, ldc); @@ -336,13 +336,13 @@ static void cblas_dsyrk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRAN } static void cblas_csyrk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *beta, void *c, const int ldc) { + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* beta, void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_csyrk_p == NULL) cblas_csyrk_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *beta, void *c, + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* beta, void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_csyrk"); if (cblas_csyrk_p != NULL) cblas_csyrk_p(layout, uplo, trans, n, k, alpha, a, lda, beta, c, ldc); @@ -350,13 +350,13 @@ static void cblas_csyrk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRAN } static void cblas_zsyrk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *beta, void *c, const int ldc) { + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* beta, void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_zsyrk_p == NULL) cblas_zsyrk_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *beta, void *c, + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* beta, void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_zsyrk"); if (cblas_zsyrk_p != NULL) cblas_zsyrk_p(layout, uplo, trans, n, k, alpha, a, lda, beta, c, ldc); @@ -364,43 +364,43 @@ static void cblas_zsyrk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRAN } static void cblas_chemm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc) { + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_chemm_p == NULL) cblas_chemm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc))GET_FUNC(h_libcblas, "cblas_chemm"); + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_chemm"); if (cblas_chemm_p != NULL) cblas_chemm_p(layout, left_right, uplo, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } } static void cblas_zhemm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc) { + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_zhemm_p == NULL) cblas_zhemm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE left_right, CBLAS_UPLO uplo, - const int m, const int n, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc))GET_FUNC(h_libcblas, "cblas_zhemm"); + const int m, const int n, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_zhemm"); if (cblas_zhemm_p != NULL) cblas_zhemm_p(layout, left_right, uplo, m, n, alpha, a, lda, b, ldb, beta, c, ldc); } } static void cblas_cherk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const float alpha, const void *a, - const int lda, const float beta, void *c, const int ldc) { + const int n, const int k, const float alpha, const void* a, + const int lda, const float beta, void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_cherk_p == NULL) cblas_cherk_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const float alpha, const void *a, - const int lda, const float beta, void *c, + const int n, const int k, const float alpha, const void* a, + const int lda, const float beta, void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_cherk"); if (cblas_cherk_p != NULL) cblas_cherk_p(layout, uplo, trans, n, k, alpha, a, lda, beta, c, ldc); @@ -408,13 +408,13 @@ static void cblas_cherk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRAN } static void cblas_zherk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const double alpha, const void *a, - const int lda, const double beta, void *c, const int ldc) { + const int n, const int k, const double alpha, const void* a, + const int lda, const double beta, void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_zherk_p == NULL) cblas_zherk_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const double alpha, const void *a, - const int lda, const double beta, void *c, + const int n, const int k, const double alpha, const void* a, + const int lda, const double beta, void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_zherk"); if (cblas_zherk_p != NULL) cblas_zherk_p(layout, uplo, trans, n, k, alpha, a, lda, beta, c, ldc); @@ -422,15 +422,15 @@ static void cblas_zherk_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRAN } static void cblas_ssyr2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const float alpha, const float *a, - const int lda, const float *b, const int ldb, const float beta, - float *c, const int ldc) { + const int n, const int k, const float alpha, const float* a, + const int lda, const float* b, const int ldb, const float beta, + float* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_ssyr2k_p == NULL) cblas_ssyr2k_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int n, - const int k, const float alpha, const float *a, const int lda, - const float *b, const int ldb, const float beta, float *c, + const int k, const float alpha, const float* a, const int lda, + const float* b, const int ldb, const float beta, float* c, const int ldc))GET_FUNC(h_libcblas, "cblas_ssyr2k"); if (cblas_ssyr2k_p != NULL) cblas_ssyr2k_p(layout, uplo, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); @@ -438,15 +438,15 @@ static void cblas_ssyr2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRA } static void cblas_dsyr2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const double alpha, const double *a, - const int lda, const double *b, const int ldb, const double beta, - double *c, const int ldc) { + const int n, const int k, const double alpha, const double* a, + const int lda, const double* b, const int ldb, const double beta, + double* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_dsyr2k_p == NULL) cblas_dsyr2k_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int n, - const int k, const double alpha, const double *a, const int lda, - const double *b, const int ldb, const double beta, double *c, + const int k, const double alpha, const double* a, const int lda, + const double* b, const int ldb, const double beta, double* c, const int ldc))GET_FUNC(h_libcblas, "cblas_dsyr2k"); if (cblas_dsyr2k_p != NULL) cblas_dsyr2k_p(layout, uplo, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); @@ -454,15 +454,15 @@ static void cblas_dsyr2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRA } static void cblas_csyr2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc) { + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_csyr2k_p == NULL) cblas_csyr2k_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int n, - const int k, const void *alpha, const void *a, const int lda, - const void *b, const int ldb, const void *beta, void *c, + const int k, const void* alpha, const void* a, const int lda, + const void* b, const int ldb, const void* beta, void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_csyr2k"); if (cblas_csyr2k_p != NULL) cblas_csyr2k_p(layout, uplo, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); @@ -470,15 +470,15 @@ static void cblas_csyr2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRA } static void cblas_zsyr2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const void *beta, - void *c, const int ldc) { + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const void* beta, + void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_zsyr2k_p == NULL) cblas_zsyr2k_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int n, - const int k, const void *alpha, const void *a, const int lda, - const void *b, const int ldb, const void *beta, void *c, + const int k, const void* alpha, const void* a, const int lda, + const void* b, const int ldb, const void* beta, void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_zsyr2k"); if (cblas_zsyr2k_p != NULL) cblas_zsyr2k_p(layout, uplo, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); @@ -486,15 +486,15 @@ static void cblas_zsyr2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRA } static void cblas_cher2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const float beta, - void *c, const int ldc) { + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const float beta, + void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_cher2k_p == NULL) cblas_cher2k_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int n, - const int k, const void *alpha, const void *a, const int lda, - const void *b, const int ldb, const float beta, void *c, + const int k, const void* alpha, const void* a, const int lda, + const void* b, const int ldb, const float beta, void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_cher2k"); if (cblas_cher2k_p != NULL) cblas_cher2k_p(layout, uplo, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); @@ -502,15 +502,15 @@ static void cblas_cher2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRA } static void cblas_zher2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, - const int n, const int k, const void *alpha, const void *a, - const int lda, const void *b, const int ldb, const double beta, - void *c, const int ldc) { + const int n, const int k, const void* alpha, const void* a, + const int lda, const void* b, const int ldb, const double beta, + void* c, const int ldc) { if (cblas_library() != NULL) { if (cblas_zher2k_p == NULL) cblas_zher2k_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRANSPOSE trans, const int n, - const int k, const void *alpha, const void *a, const int lda, - const void *b, const int ldb, const double beta, void *c, + const int k, const void* alpha, const void* a, const int lda, + const void* b, const int ldb, const double beta, void* c, const int ldc))GET_FUNC(h_libcblas, "cblas_zher2k"); if (cblas_zher2k_p != NULL) cblas_zher2k_p(layout, uplo, trans, n, k, alpha, a, lda, b, ldb, beta, c, ldc); @@ -519,14 +519,14 @@ static void cblas_zher2k_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO uplo, CBLAS_TRA static void cblas_strmm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const float alpha, const float *a, const int lda, float *b, + const float alpha, const float* a, const int lda, float* b, const int ldb) { if (cblas_library() != NULL) { if (cblas_strmm_p == NULL) cblas_strmm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, - const int n, const float alpha, const float *a, const int lda, - float *b, const int ldb))GET_FUNC(h_libcblas, "cblas_strmm"); + const int n, const float alpha, const float* a, const int lda, + float* b, const int ldb))GET_FUNC(h_libcblas, "cblas_strmm"); if (cblas_strmm_p != NULL) cblas_strmm_p(layout, side, uplo, transa, diag, m, n, alpha, a, lda, b, ldb); } @@ -534,14 +534,14 @@ static void cblas_strmm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO static void cblas_dtrmm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const double alpha, const double *a, const int lda, double *b, + const double alpha, const double* a, const int lda, double* b, const int ldb) { if (cblas_library() != NULL) { if (cblas_dtrmm_p == NULL) cblas_dtrmm_p = (void (*)( CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int m, const int n, const double alpha, const double *a, - const int lda, double *b, const int ldb))GET_FUNC(h_libcblas, "cblas_dtrmm"); + CBLAS_DIAG diag, const int m, const int n, const double alpha, const double* a, + const int lda, double* b, const int ldb))GET_FUNC(h_libcblas, "cblas_dtrmm"); if (cblas_dtrmm_p != NULL) cblas_dtrmm_p(layout, side, uplo, transa, diag, m, n, alpha, a, lda, b, ldb); } @@ -549,14 +549,14 @@ static void cblas_dtrmm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO static void cblas_ctrmm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const void *alpha, const void *a, const int lda, void *b, + const void* alpha, const void* a, const int lda, void* b, const int ldb) { if (cblas_library() != NULL) { if (cblas_ctrmm_p == NULL) cblas_ctrmm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, - const int n, const void *alpha, const void *a, const int lda, - void *b, const int ldb))GET_FUNC(h_libcblas, "cblas_ctrmm"); + const int n, const void* alpha, const void* a, const int lda, + void* b, const int ldb))GET_FUNC(h_libcblas, "cblas_ctrmm"); if (cblas_ctrmm_p != NULL) cblas_ctrmm_p(layout, side, uplo, transa, diag, m, n, alpha, a, lda, b, ldb); } @@ -564,14 +564,14 @@ static void cblas_ctrmm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO static void cblas_ztrmm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const void *alpha, const void *a, const int lda, void *b, + const void* alpha, const void* a, const int lda, void* b, const int ldb) { if (cblas_library() != NULL) { if (cblas_ztrmm_p == NULL) cblas_ztrmm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, - const int n, const void *alpha, const void *a, const int lda, - void *b, const int ldb))GET_FUNC(h_libcblas, "cblas_ztrmm"); + const int n, const void* alpha, const void* a, const int lda, + void* b, const int ldb))GET_FUNC(h_libcblas, "cblas_ztrmm"); if (cblas_ztrmm_p != NULL) cblas_ztrmm_p(layout, side, uplo, transa, diag, m, n, alpha, a, lda, b, ldb); } @@ -579,14 +579,14 @@ static void cblas_ztrmm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO static void cblas_strsm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const float alpha, const float *a, const int lda, float *b, + const float alpha, const float* a, const int lda, float* b, const int ldb) { if (cblas_library() != NULL) { if (cblas_strsm_p == NULL) cblas_strsm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, - const int n, const float alpha, const float *a, const int lda, - float *b, const int ldb))GET_FUNC(h_libcblas, "cblas_strsm"); + const int n, const float alpha, const float* a, const int lda, + float* b, const int ldb))GET_FUNC(h_libcblas, "cblas_strsm"); if (cblas_strsm_p != NULL) cblas_strsm_p(layout, side, uplo, transa, diag, m, n, alpha, a, lda, b, ldb); } @@ -594,14 +594,14 @@ static void cblas_strsm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO static void cblas_dtrsm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const double alpha, const double *a, const int lda, double *b, + const double alpha, const double* a, const int lda, double* b, const int ldb) { if (cblas_library() != NULL) { if (cblas_dtrsm_p == NULL) cblas_dtrsm_p = (void (*)( CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, - CBLAS_DIAG diag, const int m, const int n, const double alpha, const double *a, - const int lda, double *b, const int ldb))GET_FUNC(h_libcblas, "cblas_dtrsm"); + CBLAS_DIAG diag, const int m, const int n, const double alpha, const double* a, + const int lda, double* b, const int ldb))GET_FUNC(h_libcblas, "cblas_dtrsm"); if (cblas_dtrsm_p != NULL) cblas_dtrsm_p(layout, side, uplo, transa, diag, m, n, alpha, a, lda, b, ldb); } @@ -609,14 +609,14 @@ static void cblas_dtrsm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO static void cblas_ctrsm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const void *alpha, const void *a, const int lda, void *b, + const void* alpha, const void* a, const int lda, void* b, const int ldb) { if (cblas_library() != NULL) { if (cblas_ctrsm_p == NULL) cblas_ctrsm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, - const int n, const void *alpha, const void *a, const int lda, - void *b, const int ldb))GET_FUNC(h_libcblas, "cblas_ctrsm"); + const int n, const void* alpha, const void* a, const int lda, + void* b, const int ldb))GET_FUNC(h_libcblas, "cblas_ctrsm"); if (cblas_ctrsm_p != NULL) cblas_ctrsm_p(layout, side, uplo, transa, diag, m, n, alpha, a, lda, b, ldb); } @@ -624,14 +624,14 @@ static void cblas_ctrsm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO static void cblas_ztrsm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, const int n, - const void *alpha, const void *a, const int lda, void *b, + const void* alpha, const void* a, const int lda, void* b, const int ldb) { if (cblas_library() != NULL) { if (cblas_ztrsm_p == NULL) cblas_ztrsm_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO uplo, CBLAS_TRANSPOSE transa, CBLAS_DIAG diag, const int m, - const int n, const void *alpha, const void *a, const int lda, - void *b, const int ldb))GET_FUNC(h_libcblas, "cblas_ztrsm"); + const int n, const void* alpha, const void* a, const int lda, + void* b, const int ldb))GET_FUNC(h_libcblas, "cblas_ztrsm"); if (cblas_ztrsm_p != NULL) cblas_ztrsm_p(layout, side, uplo, transa, diag, m, n, alpha, a, lda, b, ldb); } @@ -640,213 +640,213 @@ static void cblas_ztrsm_wrapper(CBLAS_LAYOUT layout, CBLAS_SIDE side, CBLAS_UPLO /* Level 2 */ static void (*cblas_sgemv_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - const float alpha, const float *a, const int lda, const float *x, - const int incx, const float beta, float *y, const int incy); + const float alpha, const float* a, const int lda, const float* x, + const int incx, const float beta, float* y, const int incy); static void (*cblas_dgemv_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - const double alpha, const double *a, const int lda, const double *x, - const int incx, const double beta, double *y, const int incy); + const double alpha, const double* a, const int lda, const double* x, + const int incx, const double beta, double* y, const int incy); static void (*cblas_cgemv_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - const void *alpha, const void *a, const int lda, const void *x, - const int incx, const void *beta, void *y, const int incy); + const void* alpha, const void* a, const int lda, const void* x, + const int incx, const void* beta, void* y, const int incy); static void (*cblas_zgemv_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - const void *alpha, const void *a, const int lda, const void *x, - const int incx, const void *beta, void *y, const int incy); + const void* alpha, const void* a, const int lda, const void* x, + const int incx, const void* beta, void* y, const int incy); static void (*cblas_sgbmv_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - int kl, int ku, const float alpha, const float *a, const int lda, - const float *x, const int incx, const float beta, float *y, + int kl, int ku, const float alpha, const float* a, const int lda, + const float* x, const int incx, const float beta, float* y, const int incy); static void (*cblas_dgbmv_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - int kl, int ku, const double alpha, const double *a, const int lda, - const double *x, const int incx, const double beta, double *y, + int kl, int ku, const double alpha, const double* a, const int lda, + const double* x, const int incx, const double beta, double* y, const int incy); static void (*cblas_cgbmv_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - int kl, int ku, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + int kl, int ku, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy); static void (*cblas_zgbmv_p)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - int kl, int ku, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + int kl, int ku, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy); static void (*cblas_sger_p)(CBLAS_LAYOUT layout, const int m, const int n, const float alpha, - const float *x, const int incx, const float *y, const int incy, - float *a, const int lda); + const float* x, const int incx, const float* y, const int incy, + float* a, const int lda); static void (*cblas_dger_p)(CBLAS_LAYOUT layout, const int m, const int n, const double alpha, - const double *x, const int incx, const double *y, const int incy, - double *a, const int lda); -static void (*cblas_cgerc_p)(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, void *a, + const double* x, const int incx, const double* y, const int incy, + double* a, const int lda); +static void (*cblas_cgerc_p)(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, void* a, const int lda); -static void (*cblas_zgerc_p)(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, void *a, +static void (*cblas_zgerc_p)(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, void* a, const int lda); -static void (*cblas_cgeru_p)(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, void *a, +static void (*cblas_cgeru_p)(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, void* a, const int lda); -static void (*cblas_zgeru_p)(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, void *a, +static void (*cblas_zgeru_p)(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, void* a, const int lda); static void (*cblas_chbmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, const int k, - const void *alpha, const void *a, const int lda, const void *x, - const int incx, const void *beta, void *y, const int incy); + const void* alpha, const void* a, const int lda, const void* x, + const int incx, const void* beta, void* y, const int incy); static void (*cblas_zhbmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, const int k, - const void *alpha, const void *a, const int lda, const void *x, - const int incx, const void *beta, void *y, const int incy); + const void* alpha, const void* a, const int lda, const void* x, + const int incx, const void* beta, void* y, const int incy); static void (*cblas_chemv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const int lda, const void *x, - const int incx, const void *beta, void *y, const int incy); + const void* alpha, const void* a, const int lda, const void* x, + const int incx, const void* beta, void* y, const int incy); static void (*cblas_zhemv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const int lda, const void *x, - const int incx, const void *beta, void *y, const int incy); + const void* alpha, const void* a, const int lda, const void* x, + const int incx, const void* beta, void* y, const int incy); static void (*cblas_cher_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const void *x, const int incx, void *a, + const float alpha, const void* x, const int incx, void* a, const int lda); static void (*cblas_zher_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const void *x, const int incx, void *a, + const double alpha, const void* x, const int incx, void* a, const int lda); static void (*cblas_cher2_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, const void *y, - const int incy, void *a, const int lda); + const void* alpha, const void* x, const int incx, const void* y, + const int incy, void* a, const int lda); static void (*cblas_zher2_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, const void *y, - const int incy, void *a, const int lda); + const void* alpha, const void* x, const int incx, const void* y, + const int incy, void* a, const int lda); static void (*cblas_chpmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const void *x, const int incx, - const void *beta, void *y, const int incy); + const void* alpha, const void* a, const void* x, const int incx, + const void* beta, void* y, const int incy); static void (*cblas_zhpmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const void *x, const int incx, - const void *beta, void *y, const int incy); + const void* alpha, const void* a, const void* x, const int incx, + const void* beta, void* y, const int incy); static void (*cblas_chpr_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const void *x, const int incx, void *a); + const float alpha, const void* x, const int incx, void* a); static void (*cblas_zhpr_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const void *x, const int incx, void *a); + const double alpha, const void* x, const int incx, void* a); static void (*cblas_chpr2_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, const void *y, - const int incy, void *a); + const void* alpha, const void* x, const int incx, const void* y, + const int incy, void* a); static void (*cblas_zhpr2_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, const void *y, - const int incy, void *a); + const void* alpha, const void* x, const int incx, const void* y, + const int incy, void* a); static void (*cblas_ssbmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, const int k, - const float alpha, const float *a, const int lda, const float *x, - const int incx, const float beta, float *y, const int incy); + const float alpha, const float* a, const int lda, const float* x, + const int incx, const float beta, float* y, const int incy); static void (*cblas_dsbmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, const int k, - const double alpha, const double *a, const int lda, const double *x, - const int incx, const double beta, double *y, const int incy); + const double alpha, const double* a, const int lda, const double* x, + const int incx, const double beta, double* y, const int incy); static void (*cblas_ssymv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *a, const int lda, const float *x, - const int incx, const float beta, float *y, const int incy); + const float alpha, const float* a, const int lda, const float* x, + const int incx, const float beta, float* y, const int incy); static void (*cblas_dsymv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *a, const int lda, const double *x, - const int incx, const double beta, double *y, const int incy); + const double alpha, const double* a, const int lda, const double* x, + const int incx, const double beta, double* y, const int incy); static void (*cblas_ssyr_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, float *a, + const float alpha, const float* x, const int incx, float* a, const int lda); static void (*cblas_dsyr_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, double *a, + const double alpha, const double* x, const int incx, double* a, const int lda); static void (*cblas_ssyr2_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, const float *y, - const int incy, float *a, const int lda); + const float alpha, const float* x, const int incx, const float* y, + const int incy, float* a, const int lda); static void (*cblas_dsyr2_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, const double *y, - const int incy, double *a, const int lda); + const double alpha, const double* x, const int incx, const double* y, + const int incy, double* a, const int lda); static void (*cblas_sspmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *a, const float *x, const int incx, - const float beta, float *y, const int incy); + const float alpha, const float* a, const float* x, const int incx, + const float beta, float* y, const int incy); static void (*cblas_dspmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *a, const double *x, const int incx, - const double beta, double *y, const int incy); + const double alpha, const double* a, const double* x, const int incx, + const double beta, double* y, const int incy); static void (*cblas_sspr_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, float *a); + const float alpha, const float* x, const int incx, float* a); static void (*cblas_dspr_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, double *a); + const double alpha, const double* x, const int incx, double* a); static void (*cblas_sspr2_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, const float *y, - const int incy, float *a); + const float alpha, const float* x, const int incx, const float* y, + const int incy, float* a); static void (*cblas_dspr2_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, const double *y, - const int incy, double *a); + const double alpha, const double* x, const int incx, const double* y, + const int incy, double* a); static void (*cblas_stbmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const float *a, - const int lda, float *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const int k, const float* a, + const int lda, float* x, const int incx); static void (*cblas_dtbmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const double *a, - const int lda, double *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const int k, const double* a, + const int lda, double* x, const int incx); static void (*cblas_ctbmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const void *a, - const int lda, void *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const int k, const void* a, + const int lda, void* x, const int incx); static void (*cblas_ztbmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const void *a, - const int lda, void *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const int k, const void* a, + const int lda, void* x, const int incx); static void (*cblas_stbsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const float *a, - const int lda, float *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const int k, const float* a, + const int lda, float* x, const int incx); static void (*cblas_dtbsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const double *a, - const int lda, double *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const int k, const double* a, + const int lda, double* x, const int incx); static void (*cblas_ctbsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const void *a, - const int lda, void *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const int k, const void* a, + const int lda, void* x, const int incx); static void (*cblas_ztbsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const void *a, - const int lda, void *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const int k, const void* a, + const int lda, void* x, const int incx); static void (*cblas_stpmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, float *x, + CBLAS_DIAG unit_diag, const int n, const float* a, float* x, const int incx); static void (*cblas_dtpmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, double *x, + CBLAS_DIAG unit_diag, const int n, const double* a, double* x, const int incx); static void (*cblas_ctpmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx); static void (*cblas_ztpmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx); static void (*cblas_stpsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, float *x, + CBLAS_DIAG unit_diag, const int n, const float* a, float* x, const int incx); static void (*cblas_dtpsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, double *x, + CBLAS_DIAG unit_diag, const int n, const double* a, double* x, const int incx); static void (*cblas_ctpsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx); static void (*cblas_ztpsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx); static void (*cblas_strmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, const int lda, - float *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const float* a, const int lda, + float* x, const int incx); static void (*cblas_dtrmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, const int lda, - double *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const double* a, const int lda, + double* x, const int incx); static void (*cblas_ctrmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, - void *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, + void* x, const int incx); static void (*cblas_ztrmv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, - void *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, + void* x, const int incx); static void (*cblas_strsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, const int lda, - float *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const float* a, const int lda, + float* x, const int incx); static void (*cblas_dtrsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, const int lda, - double *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const double* a, const int lda, + double* x, const int incx); static void (*cblas_ctrsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, - void *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, + void* x, const int incx); static void (*cblas_ztrsv_p)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, - void *x, const int incx); + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, + void* x, const int incx); static void cblas_sgemv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, - const int n, const float alpha, const float *a, const int lda, - const float *x, const int incx, const float beta, float *y, + const int n, const float alpha, const float* a, const int lda, + const float* x, const int incx, const float beta, float* y, const int incy) { if (cblas_library() != NULL) { if (cblas_sgemv_p == NULL) cblas_sgemv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, - const int n, const float alpha, const float *a, const int lda, - const float *x, const int incx, const float beta, float *y, + const int n, const float alpha, const float* a, const int lda, + const float* x, const int incx, const float beta, float* y, const int incy))GET_FUNC(h_libcblas, "cblas_sgemv"); if (cblas_sgemv_p != NULL) cblas_sgemv_p(layout, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); @@ -854,29 +854,29 @@ static void cblas_sgemv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, cons } static void cblas_dgemv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, - const int n, const double alpha, const double *a, const int lda, - const double *x, const int incx, const double beta, double *y, + const int n, const double alpha, const double* a, const int lda, + const double* x, const int incx, const double beta, double* y, const int incy) { if (cblas_library() != NULL) { if (cblas_dgemv_p == NULL) cblas_dgemv_p = (void (*)( CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - const double alpha, const double *a, const int lda, const double *x, const int incx, - const double beta, double *y, const int incy))GET_FUNC(h_libcblas, "cblas_dgemv"); + const double alpha, const double* a, const int lda, const double* x, const int incx, + const double beta, double* y, const int incy))GET_FUNC(h_libcblas, "cblas_dgemv"); if (cblas_dgemv_p != NULL) cblas_dgemv_p(layout, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); } } static void cblas_cgemv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, - const int n, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + const int n, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_cgemv_p == NULL) cblas_cgemv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, - const int n, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + const int n, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_cgemv"); if (cblas_cgemv_p != NULL) cblas_cgemv_p(layout, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); @@ -884,14 +884,14 @@ static void cblas_cgemv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, cons } static void cblas_zgemv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, - const int n, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + const int n, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_zgemv_p == NULL) cblas_zgemv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, - const int n, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + const int n, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_zgemv"); if (cblas_zgemv_p != NULL) cblas_zgemv_p(layout, trans, m, n, alpha, a, lda, x, incx, beta, y, incy); @@ -899,15 +899,15 @@ static void cblas_zgemv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, cons } static void cblas_sgbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, - const int n, int kl, int ku, const float alpha, const float *a, - const int lda, const float *x, const int incx, const float beta, - float *y, const int incy) { + const int n, int kl, int ku, const float alpha, const float* a, + const int lda, const float* x, const int incx, const float beta, + float* y, const int incy) { if (cblas_library() != NULL) { if (cblas_sgbmv_p == NULL) cblas_sgbmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - int kl, int ku, const float alpha, const float *a, const int lda, - const float *x, const int incx, const float beta, float *y, + int kl, int ku, const float alpha, const float* a, const int lda, + const float* x, const int incx, const float beta, float* y, const int incy))GET_FUNC(h_libcblas, "cblas_sgbmv"); if (cblas_sgbmv_p != NULL) cblas_sgbmv_p(layout, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); @@ -915,15 +915,15 @@ static void cblas_sgbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, cons } static void cblas_dgbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, - const int n, int kl, int ku, const double alpha, const double *a, - const int lda, const double *x, const int incx, const double beta, - double *y, const int incy) { + const int n, int kl, int ku, const double alpha, const double* a, + const int lda, const double* x, const int incx, const double beta, + double* y, const int incy) { if (cblas_library() != NULL) { if (cblas_dgbmv_p == NULL) cblas_dgbmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - int kl, int ku, const double alpha, const double *a, const int lda, - const double *x, const int incx, const double beta, double *y, + int kl, int ku, const double alpha, const double* a, const int lda, + const double* x, const int incx, const double beta, double* y, const int incy))GET_FUNC(h_libcblas, "cblas_dgbmv"); if (cblas_dgbmv_p != NULL) cblas_dgbmv_p(layout, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); @@ -931,15 +931,15 @@ static void cblas_dgbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, cons } static void cblas_cgbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, - const int n, int kl, int ku, const void *alpha, const void *a, - const int lda, const void *x, const int incx, const void *beta, - void *y, const int incy) { + const int n, int kl, int ku, const void* alpha, const void* a, + const int lda, const void* x, const int incx, const void* beta, + void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_cgbmv_p == NULL) cblas_cgbmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - int kl, int ku, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + int kl, int ku, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_cgbmv"); if (cblas_cgbmv_p != NULL) cblas_cgbmv_p(layout, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); @@ -947,15 +947,15 @@ static void cblas_cgbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, cons } static void cblas_zgbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, - const int n, int kl, int ku, const void *alpha, const void *a, - const int lda, const void *x, const int incx, const void *beta, - void *y, const int incy) { + const int n, int kl, int ku, const void* alpha, const void* a, + const int lda, const void* x, const int incx, const void* beta, + void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_zgbmv_p == NULL) cblas_zgbmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, const int m, const int n, - int kl, int ku, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + int kl, int ku, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_zgbmv"); if (cblas_zgbmv_p != NULL) cblas_zgbmv_p(layout, trans, m, n, kl, ku, alpha, a, lda, x, incx, beta, y, incy); @@ -963,13 +963,13 @@ static void cblas_zgbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_TRANSPOSE trans, cons } static void cblas_sger_wrapper(CBLAS_LAYOUT layout, const int m, const int n, const float alpha, - const float *x, const int incx, const float *y, const int incy, - float *a, const int lda) { + const float* x, const int incx, const float* y, const int incy, + float* a, const int lda) { if (cblas_library() != NULL) { if (cblas_sger_p == NULL) cblas_sger_p = (void (*)(CBLAS_LAYOUT layout, const int m, const int n, const float alpha, - const float *x, const int incx, const float *y, const int incy, float *a, + const float* x, const int incx, const float* y, const int incy, float* a, const int lda))GET_FUNC(h_libcblas, "cblas_sger"); if (cblas_sger_p != NULL) cblas_sger_p(layout, m, n, alpha, x, incx, y, incy, a, lda); @@ -977,69 +977,69 @@ static void cblas_sger_wrapper(CBLAS_LAYOUT layout, const int m, const int n, co } static void cblas_dger_wrapper(CBLAS_LAYOUT layout, const int m, const int n, const double alpha, - const double *x, const int incx, const double *y, const int incy, - double *a, const int lda) { + const double* x, const int incx, const double* y, const int incy, + double* a, const int lda) { if (cblas_library() != NULL) { if (cblas_dger_p == NULL) cblas_dger_p = (void (*)(CBLAS_LAYOUT layout, const int m, const int n, const double alpha, - const double *x, const int incx, const double *y, const int incy, - double *a, const int lda))GET_FUNC(h_libcblas, "cblas_dger"); + const double* x, const int incx, const double* y, const int incy, + double* a, const int lda))GET_FUNC(h_libcblas, "cblas_dger"); if (cblas_dger_p != NULL) cblas_dger_p(layout, m, n, alpha, x, incx, y, incy, a, lda); } } -static void cblas_cgerc_wrapper(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, - void *a, const int lda) { +static void cblas_cgerc_wrapper(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, + void* a, const int lda) { if (cblas_library() != NULL) { if (cblas_cgerc_p == NULL) cblas_cgerc_p = - (void (*)(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, void *a, + (void (*)(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, void* a, const int lda))GET_FUNC(h_libcblas, "cblas_cgerc"); if (cblas_cgerc_p != NULL) cblas_cgerc_p(layout, m, n, alpha, x, incx, y, incy, a, lda); } } -static void cblas_zgerc_wrapper(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, - void *a, const int lda) { +static void cblas_zgerc_wrapper(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, + void* a, const int lda) { if (cblas_library() != NULL) { if (cblas_zgerc_p == NULL) cblas_zgerc_p = - (void (*)(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, void *a, + (void (*)(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, void* a, const int lda))GET_FUNC(h_libcblas, "cblas_zgerc"); if (cblas_zgerc_p != NULL) cblas_zgerc_p(layout, m, n, alpha, x, incx, y, incy, a, lda); } } -static void cblas_cgeru_wrapper(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, - void *a, const int lda) { +static void cblas_cgeru_wrapper(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, + void* a, const int lda) { if (cblas_library() != NULL) { if (cblas_cgeru_p == NULL) cblas_cgeru_p = - (void (*)(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, void *a, + (void (*)(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, void* a, const int lda))GET_FUNC(h_libcblas, "cblas_cgeru"); if (cblas_cgeru_p != NULL) cblas_cgeru_p(layout, m, n, alpha, x, incx, y, incy, a, lda); } } -static void cblas_zgeru_wrapper(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, - void *a, const int lda) { +static void cblas_zgeru_wrapper(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, + void* a, const int lda) { if (cblas_library() != NULL) { if (cblas_zgeru_p == NULL) cblas_zgeru_p = - (void (*)(CBLAS_LAYOUT layout, const int m, const int n, const void *alpha, - const void *x, const int incx, const void *y, const int incy, void *a, + (void (*)(CBLAS_LAYOUT layout, const int m, const int n, const void* alpha, + const void* x, const int incx, const void* y, const int incy, void* a, const int lda))GET_FUNC(h_libcblas, "cblas_zgeru"); if (cblas_zgeru_p != NULL) cblas_zgeru_p(layout, m, n, alpha, x, incx, y, incy, a, lda); @@ -1047,14 +1047,14 @@ static void cblas_zgeru_wrapper(CBLAS_LAYOUT layout, const int m, const int n, c } static void cblas_chbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const int k, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + const int k, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_chbmv_p == NULL) cblas_chbmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const int k, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + const int k, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_chbmv"); if (cblas_chbmv_p != NULL) cblas_chbmv_p(layout, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); @@ -1062,14 +1062,14 @@ static void cblas_chbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_zhbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const int k, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + const int k, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_zhbmv_p == NULL) cblas_zhbmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const int k, const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + const int k, const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_zhbmv"); if (cblas_zhbmv_p != NULL) cblas_zhbmv_p(layout, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); @@ -1077,13 +1077,13 @@ static void cblas_zhbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_chemv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const int lda, const void *x, - const int incx, const void *beta, void *y, const int incy) { + const void* alpha, const void* a, const int lda, const void* x, + const int incx, const void* beta, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_chemv_p == NULL) cblas_chemv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_chemv"); if (cblas_chemv_p != NULL) cblas_chemv_p(layout, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); @@ -1091,13 +1091,13 @@ static void cblas_chemv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_zhemv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const int lda, const void *x, - const int incx, const void *beta, void *y, const int incy) { + const void* alpha, const void* a, const int lda, const void* x, + const int incx, const void* beta, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_zhemv_p == NULL) cblas_zhemv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const int lda, - const void *x, const int incx, const void *beta, void *y, + const void* alpha, const void* a, const int lda, + const void* x, const int incx, const void* beta, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_zhemv"); if (cblas_zhemv_p != NULL) cblas_zhemv_p(layout, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); @@ -1105,12 +1105,12 @@ static void cblas_zhemv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_cher_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const void *x, const int incx, void *a, + const float alpha, const void* x, const int incx, void* a, const int lda) { if (cblas_library() != NULL) { if (cblas_cher_p == NULL) cblas_cher_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const void *x, const int incx, void *a, + const float alpha, const void* x, const int incx, void* a, const int lda))GET_FUNC(h_libcblas, "cblas_cher"); if (cblas_cher_p != NULL) cblas_cher_p(layout, upper_lower, n, alpha, x, incx, a, lda); @@ -1118,12 +1118,12 @@ static void cblas_cher_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, cons } static void cblas_zher_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const void *x, const int incx, void *a, + const double alpha, const void* x, const int incx, void* a, const int lda) { if (cblas_library() != NULL) { if (cblas_zher_p == NULL) cblas_zher_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const void *x, const int incx, void *a, + const double alpha, const void* x, const int incx, void* a, const int lda))GET_FUNC(h_libcblas, "cblas_zher"); if (cblas_zher_p != NULL) cblas_zher_p(layout, upper_lower, n, alpha, x, incx, a, lda); @@ -1131,13 +1131,13 @@ static void cblas_zher_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, cons } static void cblas_cher2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, const void *y, - const int incy, void *a, const int lda) { + const void* alpha, const void* x, const int incx, const void* y, + const int incy, void* a, const int lda) { if (cblas_library() != NULL) { if (cblas_cher2_p == NULL) cblas_cher2_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, - const void *y, const int incy, void *a, + const void* alpha, const void* x, const int incx, + const void* y, const int incy, void* a, const int lda))GET_FUNC(h_libcblas, "cblas_cher2"); if (cblas_cher2_p != NULL) cblas_cher2_p(layout, upper_lower, n, alpha, x, incx, y, incy, a, lda); @@ -1145,13 +1145,13 @@ static void cblas_cher2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_zher2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, const void *y, - const int incy, void *a, const int lda) { + const void* alpha, const void* x, const int incx, const void* y, + const int incy, void* a, const int lda) { if (cblas_library() != NULL) { if (cblas_zher2_p == NULL) cblas_zher2_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, - const void *y, const int incy, void *a, + const void* alpha, const void* x, const int incx, + const void* y, const int incy, void* a, const int lda))GET_FUNC(h_libcblas, "cblas_zher2"); if (cblas_zher2_p != NULL) cblas_zher2_p(layout, upper_lower, n, alpha, x, incx, y, incy, a, lda); @@ -1159,13 +1159,13 @@ static void cblas_zher2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_chpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const void *x, const int incx, - const void *beta, void *y, const int incy) { + const void* alpha, const void* a, const void* x, const int incx, + const void* beta, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_chpmv_p == NULL) cblas_chpmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const void *x, - const int incx, const void *beta, void *y, + const void* alpha, const void* a, const void* x, + const int incx, const void* beta, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_chpmv"); if (cblas_chpmv_p != NULL) cblas_chpmv_p(layout, upper_lower, n, alpha, a, x, incx, beta, y, incy); @@ -1173,13 +1173,13 @@ static void cblas_chpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_zhpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const void *x, const int incx, - const void *beta, void *y, const int incy) { + const void* alpha, const void* a, const void* x, const int incx, + const void* beta, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_zhpmv_p == NULL) cblas_zhpmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *a, const void *x, - const int incx, const void *beta, void *y, + const void* alpha, const void* a, const void* x, + const int incx, const void* beta, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_zhpmv"); if (cblas_zhpmv_p != NULL) cblas_zhpmv_p(layout, upper_lower, n, alpha, a, x, incx, beta, y, incy); @@ -1187,66 +1187,66 @@ static void cblas_zhpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_chpr_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const void *x, const int incx, void *a) { + const float alpha, const void* x, const int incx, void* a) { if (cblas_library() != NULL) { if (cblas_chpr_p == NULL) cblas_chpr_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const void *x, const int incx, - void *a))GET_FUNC(h_libcblas, "cblas_chpr"); + const float alpha, const void* x, const int incx, + void* a))GET_FUNC(h_libcblas, "cblas_chpr"); if (cblas_chpr_p != NULL) cblas_chpr_p(layout, upper_lower, n, alpha, x, incx, a); } } static void cblas_zhpr_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const void *x, const int incx, void *a) { + const double alpha, const void* x, const int incx, void* a) { if (cblas_library() != NULL) { if (cblas_zhpr_p == NULL) cblas_zhpr_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const void *x, const int incx, - void *a))GET_FUNC(h_libcblas, "cblas_zhpr"); + const double alpha, const void* x, const int incx, + void* a))GET_FUNC(h_libcblas, "cblas_zhpr"); if (cblas_zhpr_p != NULL) cblas_zhpr_p(layout, upper_lower, n, alpha, x, incx, a); } } static void cblas_chpr2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, const void *y, - const int incy, void *a) { + const void* alpha, const void* x, const int incx, const void* y, + const int incy, void* a) { if (cblas_library() != NULL) { if (cblas_chpr2_p == NULL) cblas_chpr2_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, const void *y, - const int incy, void *a))GET_FUNC(h_libcblas, "cblas_chpr2"); + const void* alpha, const void* x, const int incx, const void* y, + const int incy, void* a))GET_FUNC(h_libcblas, "cblas_chpr2"); if (cblas_chpr2_p != NULL) cblas_chpr2_p(layout, upper_lower, n, alpha, x, incx, y, incy, a); } } static void cblas_zhpr2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, const void *y, - const int incy, void *a) { + const void* alpha, const void* x, const int incx, const void* y, + const int incy, void* a) { if (cblas_library() != NULL) { if (cblas_zhpr2_p == NULL) cblas_zhpr2_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const void *alpha, const void *x, const int incx, const void *y, - const int incy, void *a))GET_FUNC(h_libcblas, "cblas_zhpr2"); + const void* alpha, const void* x, const int incx, const void* y, + const int incy, void* a))GET_FUNC(h_libcblas, "cblas_zhpr2"); if (cblas_zhpr2_p != NULL) cblas_zhpr2_p(layout, upper_lower, n, alpha, x, incx, y, incy, a); } } static void cblas_ssbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const int k, const float alpha, const float *a, const int lda, - const float *x, const int incx, const float beta, float *y, + const int k, const float alpha, const float* a, const int lda, + const float* x, const int incx, const float beta, float* y, const int incy) { if (cblas_library() != NULL) { if (cblas_ssbmv_p == NULL) cblas_ssbmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const int k, const float alpha, const float *a, const int lda, - const float *x, const int incx, const float beta, float *y, + const int k, const float alpha, const float* a, const int lda, + const float* x, const int incx, const float beta, float* y, const int incy))GET_FUNC(h_libcblas, "cblas_ssbmv"); if (cblas_ssbmv_p != NULL) cblas_ssbmv_p(layout, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); @@ -1254,28 +1254,28 @@ static void cblas_ssbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_dsbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const int k, const double alpha, const double *a, const int lda, - const double *x, const int incx, const double beta, double *y, + const int k, const double alpha, const double* a, const int lda, + const double* x, const int incx, const double beta, double* y, const int incy) { if (cblas_library() != NULL) { if (cblas_dsbmv_p == NULL) cblas_dsbmv_p = (void (*)( CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, const int k, - const double alpha, const double *a, const int lda, const double *x, const int incx, - const double beta, double *y, const int incy))GET_FUNC(h_libcblas, "cblas_dsbmv"); + const double alpha, const double* a, const int lda, const double* x, const int incx, + const double beta, double* y, const int incy))GET_FUNC(h_libcblas, "cblas_dsbmv"); if (cblas_dsbmv_p != NULL) cblas_dsbmv_p(layout, upper_lower, n, k, alpha, a, lda, x, incx, beta, y, incy); } } static void cblas_ssymv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *a, const int lda, const float *x, - const int incx, const float beta, float *y, const int incy) { + const float alpha, const float* a, const int lda, const float* x, + const int incx, const float beta, float* y, const int incy) { if (cblas_library() != NULL) { if (cblas_ssymv_p == NULL) cblas_ssymv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *a, const int lda, - const float *x, const int incx, const float beta, float *y, + const float alpha, const float* a, const int lda, + const float* x, const int incx, const float beta, float* y, const int incy))GET_FUNC(h_libcblas, "cblas_ssymv"); if (cblas_ssymv_p != NULL) cblas_ssymv_p(layout, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); @@ -1283,13 +1283,13 @@ static void cblas_ssymv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_dsymv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *a, const int lda, const double *x, - const int incx, const double beta, double *y, const int incy) { + const double alpha, const double* a, const int lda, const double* x, + const int incx, const double beta, double* y, const int incy) { if (cblas_library() != NULL) { if (cblas_dsymv_p == NULL) cblas_dsymv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *a, const int lda, - const double *x, const int incx, const double beta, double *y, + const double alpha, const double* a, const int lda, + const double* x, const int incx, const double beta, double* y, const int incy))GET_FUNC(h_libcblas, "cblas_dsymv"); if (cblas_dsymv_p != NULL) cblas_dsymv_p(layout, upper_lower, n, alpha, a, lda, x, incx, beta, y, incy); @@ -1297,12 +1297,12 @@ static void cblas_dsymv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_ssyr_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, float *a, + const float alpha, const float* x, const int incx, float* a, const int lda) { if (cblas_library() != NULL) { if (cblas_ssyr_p == NULL) cblas_ssyr_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, float *a, + const float alpha, const float* x, const int incx, float* a, const int lda))GET_FUNC(h_libcblas, "cblas_ssyr"); if (cblas_ssyr_p != NULL) cblas_ssyr_p(layout, upper_lower, n, alpha, x, incx, a, lda); @@ -1310,12 +1310,12 @@ static void cblas_ssyr_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, cons } static void cblas_dsyr_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, double *a, + const double alpha, const double* x, const int incx, double* a, const int lda) { if (cblas_library() != NULL) { if (cblas_dsyr_p == NULL) cblas_dsyr_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, double *a, + const double alpha, const double* x, const int incx, double* a, const int lda))GET_FUNC(h_libcblas, "cblas_dsyr"); if (cblas_dsyr_p != NULL) cblas_dsyr_p(layout, upper_lower, n, alpha, x, incx, a, lda); @@ -1323,13 +1323,13 @@ static void cblas_dsyr_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, cons } static void cblas_ssyr2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, const float *y, - const int incy, float *a, const int lda) { + const float alpha, const float* x, const int incx, const float* y, + const int incy, float* a, const int lda) { if (cblas_library() != NULL) { if (cblas_ssyr2_p == NULL) cblas_ssyr2_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, - const float *y, const int incy, float *a, + const float alpha, const float* x, const int incx, + const float* y, const int incy, float* a, const int lda))GET_FUNC(h_libcblas, "cblas_ssyr2"); if (cblas_ssyr2_p != NULL) cblas_ssyr2_p(layout, upper_lower, n, alpha, x, incx, y, incy, a, lda); @@ -1337,13 +1337,13 @@ static void cblas_ssyr2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_dsyr2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, - const double *y, const int incy, double *a, const int lda) { + const double alpha, const double* x, const int incx, + const double* y, const int incy, double* a, const int lda) { if (cblas_library() != NULL) { if (cblas_dsyr2_p == NULL) cblas_dsyr2_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, - const double *y, const int incy, double *a, + const double alpha, const double* x, const int incx, + const double* y, const int incy, double* a, const int lda))GET_FUNC(h_libcblas, "cblas_dsyr2"); if (cblas_dsyr2_p != NULL) cblas_dsyr2_p(layout, upper_lower, n, alpha, x, incx, y, incy, a, lda); @@ -1351,13 +1351,13 @@ static void cblas_dsyr2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_sspmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *a, const float *x, const int incx, - const float beta, float *y, const int incy) { + const float alpha, const float* a, const float* x, const int incx, + const float beta, float* y, const int incy) { if (cblas_library() != NULL) { if (cblas_sspmv_p == NULL) cblas_sspmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *a, const float *x, - const int incx, const float beta, float *y, + const float alpha, const float* a, const float* x, + const int incx, const float beta, float* y, const int incy))GET_FUNC(h_libcblas, "cblas_sspmv"); if (cblas_sspmv_p != NULL) cblas_sspmv_p(layout, upper_lower, n, alpha, a, x, incx, beta, y, incy); @@ -1365,13 +1365,13 @@ static void cblas_sspmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_dspmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *a, const double *x, - const int incx, const double beta, double *y, const int incy) { + const double alpha, const double* a, const double* x, + const int incx, const double beta, double* y, const int incy) { if (cblas_library() != NULL) { if (cblas_dspmv_p == NULL) cblas_dspmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *a, const double *x, - const int incx, const double beta, double *y, + const double alpha, const double* a, const double* x, + const int incx, const double beta, double* y, const int incy))GET_FUNC(h_libcblas, "cblas_dspmv"); if (cblas_dspmv_p != NULL) cblas_dspmv_p(layout, upper_lower, n, alpha, a, x, incx, beta, y, incy); @@ -1379,65 +1379,65 @@ static void cblas_dspmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, con } static void cblas_sspr_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, float *a) { + const float alpha, const float* x, const int incx, float* a) { if (cblas_library() != NULL) { if (cblas_sspr_p == NULL) cblas_sspr_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, - float *a))GET_FUNC(h_libcblas, "cblas_sspr"); + const float alpha, const float* x, const int incx, + float* a))GET_FUNC(h_libcblas, "cblas_sspr"); if (cblas_sspr_p != NULL) cblas_sspr_p(layout, upper_lower, n, alpha, x, incx, a); } } static void cblas_dspr_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, double *a) { + const double alpha, const double* x, const int incx, double* a) { if (cblas_library() != NULL) { if (cblas_dspr_p == NULL) cblas_dspr_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, - double *a))GET_FUNC(h_libcblas, "cblas_dspr"); + const double alpha, const double* x, const int incx, + double* a))GET_FUNC(h_libcblas, "cblas_dspr"); if (cblas_dspr_p != NULL) cblas_dspr_p(layout, upper_lower, n, alpha, x, incx, a); } } static void cblas_sspr2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, const float *y, - const int incy, float *a) { + const float alpha, const float* x, const int incx, const float* y, + const int incy, float* a) { if (cblas_library() != NULL) { if (cblas_sspr2_p == NULL) cblas_sspr2_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const float alpha, const float *x, const int incx, const float *y, - const int incy, float *a))GET_FUNC(h_libcblas, "cblas_sspr2"); + const float alpha, const float* x, const int incx, const float* y, + const int incy, float* a))GET_FUNC(h_libcblas, "cblas_sspr2"); if (cblas_sspr2_p != NULL) cblas_sspr2_p(layout, upper_lower, n, alpha, x, incx, y, incy, a); } } static void cblas_dspr2_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, - const double *y, const int incy, double *a) { + const double alpha, const double* x, const int incx, + const double* y, const int incy, double* a) { if (cblas_library() != NULL) { if (cblas_dspr2_p == NULL) cblas_dspr2_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, const int n, - const double alpha, const double *x, const int incx, const double *y, - const int incy, double *a))GET_FUNC(h_libcblas, "cblas_dspr2"); + const double alpha, const double* x, const int incx, const double* y, + const int incy, double* a))GET_FUNC(h_libcblas, "cblas_dspr2"); if (cblas_dspr2_p != NULL) cblas_dspr2_p(layout, upper_lower, n, alpha, x, incx, y, incy, a); } } static void cblas_stbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const float *a, - const int lda, float *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const int k, const float* a, + const int lda, float* x, const int incx) { if (cblas_library() != NULL) { if (cblas_stbmv_p == NULL) cblas_stbmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, const int n, - const int k, const float *a, const int lda, float *x, + const int k, const float* a, const int lda, float* x, const int incx))GET_FUNC(h_libcblas, "cblas_stbmv"); if (cblas_stbmv_p != NULL) cblas_stbmv_p(layout, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); @@ -1445,13 +1445,13 @@ static void cblas_stbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_dtbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const double *a, - const int lda, double *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const int k, const double* a, + const int lda, double* x, const int incx) { if (cblas_library() != NULL) { if (cblas_dtbmv_p == NULL) cblas_dtbmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, const int n, - const int k, const double *a, const int lda, double *x, + const int k, const double* a, const int lda, double* x, const int incx))GET_FUNC(h_libcblas, "cblas_dtbmv"); if (cblas_dtbmv_p != NULL) cblas_dtbmv_p(layout, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); @@ -1459,13 +1459,13 @@ static void cblas_dtbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_ctbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const void *a, - const int lda, void *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const int k, const void* a, + const int lda, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ctbmv_p == NULL) cblas_ctbmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, const int n, - const int k, const void *a, const int lda, void *x, + const int k, const void* a, const int lda, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ctbmv"); if (cblas_ctbmv_p != NULL) cblas_ctbmv_p(layout, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); @@ -1473,13 +1473,13 @@ static void cblas_ctbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_ztbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const void *a, - const int lda, void *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const int k, const void* a, + const int lda, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ztbmv_p == NULL) cblas_ztbmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, const int n, - const int k, const void *a, const int lda, void *x, + const int k, const void* a, const int lda, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ztbmv"); if (cblas_ztbmv_p != NULL) cblas_ztbmv_p(layout, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); @@ -1487,13 +1487,13 @@ static void cblas_ztbmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_stbsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const float *a, - const int lda, float *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const int k, const float* a, + const int lda, float* x, const int incx) { if (cblas_library() != NULL) { if (cblas_stbsv_p == NULL) cblas_stbsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, const int n, - const int k, const float *a, const int lda, float *x, + const int k, const float* a, const int lda, float* x, const int incx))GET_FUNC(h_libcblas, "cblas_stbsv"); if (cblas_stbsv_p != NULL) cblas_stbsv_p(layout, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); @@ -1501,13 +1501,13 @@ static void cblas_stbsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_dtbsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const double *a, - const int lda, double *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const int k, const double* a, + const int lda, double* x, const int incx) { if (cblas_library() != NULL) { if (cblas_dtbsv_p == NULL) cblas_dtbsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, const int n, - const int k, const double *a, const int lda, double *x, + const int k, const double* a, const int lda, double* x, const int incx))GET_FUNC(h_libcblas, "cblas_dtbsv"); if (cblas_dtbsv_p != NULL) cblas_dtbsv_p(layout, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); @@ -1515,13 +1515,13 @@ static void cblas_dtbsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_ctbsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const void *a, - const int lda, void *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const int k, const void* a, + const int lda, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ctbsv_p == NULL) cblas_ctbsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, const int n, - const int k, const void *a, const int lda, void *x, + const int k, const void* a, const int lda, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ctbsv"); if (cblas_ctbsv_p != NULL) cblas_ctbsv_p(layout, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); @@ -1529,13 +1529,13 @@ static void cblas_ctbsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_ztbsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const int k, const void *a, - const int lda, void *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const int k, const void* a, + const int lda, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ztbsv_p == NULL) cblas_ztbsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, CBLAS_DIAG unit_diag, const int n, - const int k, const void *a, const int lda, void *x, + const int k, const void* a, const int lda, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ztbsv"); if (cblas_ztbsv_p != NULL) cblas_ztbsv_p(layout, upper_lower, trans, unit_diag, n, k, a, lda, x, incx); @@ -1543,13 +1543,13 @@ static void cblas_ztbsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_stpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, float *x, + CBLAS_DIAG unit_diag, const int n, const float* a, float* x, const int incx) { if (cblas_library() != NULL) { if (cblas_stpmv_p == NULL) cblas_stpmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, float *x, + CBLAS_DIAG unit_diag, const int n, const float* a, float* x, const int incx))GET_FUNC(h_libcblas, "cblas_stpmv"); if (cblas_stpmv_p != NULL) cblas_stpmv_p(layout, upper_lower, trans, unit_diag, n, a, x, incx); @@ -1557,13 +1557,13 @@ static void cblas_stpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_dtpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, double *x, + CBLAS_DIAG unit_diag, const int n, const double* a, double* x, const int incx) { if (cblas_library() != NULL) { if (cblas_dtpmv_p == NULL) cblas_dtpmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, double *x, + CBLAS_DIAG unit_diag, const int n, const double* a, double* x, const int incx))GET_FUNC(h_libcblas, "cblas_dtpmv"); if (cblas_dtpmv_p != NULL) cblas_dtpmv_p(layout, upper_lower, trans, unit_diag, n, a, x, incx); @@ -1571,13 +1571,13 @@ static void cblas_dtpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_ctpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ctpmv_p == NULL) cblas_ctpmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ctpmv"); if (cblas_ctpmv_p != NULL) cblas_ctpmv_p(layout, upper_lower, trans, unit_diag, n, a, x, incx); @@ -1585,13 +1585,13 @@ static void cblas_ctpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_ztpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ztpmv_p == NULL) cblas_ztpmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ztpmv"); if (cblas_ztpmv_p != NULL) cblas_ztpmv_p(layout, upper_lower, trans, unit_diag, n, a, x, incx); @@ -1599,13 +1599,13 @@ static void cblas_ztpmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_stpsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, float *x, + CBLAS_DIAG unit_diag, const int n, const float* a, float* x, const int incx) { if (cblas_library() != NULL) { if (cblas_stpsv_p == NULL) cblas_stpsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, float *x, + CBLAS_DIAG unit_diag, const int n, const float* a, float* x, const int incx))GET_FUNC(h_libcblas, "cblas_stpsv"); if (cblas_stpsv_p != NULL) cblas_stpsv_p(layout, upper_lower, trans, unit_diag, n, a, x, incx); @@ -1613,13 +1613,13 @@ static void cblas_stpsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_dtpsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, double *x, + CBLAS_DIAG unit_diag, const int n, const double* a, double* x, const int incx) { if (cblas_library() != NULL) { if (cblas_dtpsv_p == NULL) cblas_dtpsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, double *x, + CBLAS_DIAG unit_diag, const int n, const double* a, double* x, const int incx))GET_FUNC(h_libcblas, "cblas_dtpsv"); if (cblas_dtpsv_p != NULL) cblas_dtpsv_p(layout, upper_lower, trans, unit_diag, n, a, x, incx); @@ -1627,13 +1627,13 @@ static void cblas_dtpsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_ctpsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ctpsv_p == NULL) cblas_ctpsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ctpsv"); if (cblas_ctpsv_p != NULL) cblas_ctpsv_p(layout, upper_lower, trans, unit_diag, n, a, x, incx); @@ -1641,13 +1641,13 @@ static void cblas_ctpsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_ztpsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ztpsv_p == NULL) cblas_ztpsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ztpsv"); if (cblas_ztpsv_p != NULL) cblas_ztpsv_p(layout, upper_lower, trans, unit_diag, n, a, x, incx); @@ -1655,41 +1655,41 @@ static void cblas_ztpsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_strmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, const int lda, - float *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const float* a, const int lda, + float* x, const int incx) { if (cblas_library() != NULL) { if (cblas_strmv_p == NULL) cblas_strmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, const int lda, - float *x, const int incx))GET_FUNC(h_libcblas, "cblas_strmv"); + CBLAS_DIAG unit_diag, const int n, const float* a, const int lda, + float* x, const int incx))GET_FUNC(h_libcblas, "cblas_strmv"); if (cblas_strmv_p != NULL) cblas_strmv_p(layout, upper_lower, trans, unit_diag, n, a, lda, x, incx); } } static void cblas_dtrmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, const int lda, - double *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const double* a, const int lda, + double* x, const int incx) { if (cblas_library() != NULL) { if (cblas_dtrmv_p == NULL) cblas_dtrmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, const int lda, - double *x, const int incx))GET_FUNC(h_libcblas, "cblas_dtrmv"); + CBLAS_DIAG unit_diag, const int n, const double* a, const int lda, + double* x, const int incx))GET_FUNC(h_libcblas, "cblas_dtrmv"); if (cblas_dtrmv_p != NULL) cblas_dtrmv_p(layout, upper_lower, trans, unit_diag, n, a, lda, x, incx); } } static void cblas_ctrmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, - void *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, + void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ctrmv_p == NULL) cblas_ctrmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ctrmv"); if (cblas_ctrmv_p != NULL) cblas_ctrmv_p(layout, upper_lower, trans, unit_diag, n, a, lda, x, incx); @@ -1697,13 +1697,13 @@ static void cblas_ctrmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_ztrmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, - void *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, + void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ztrmv_p == NULL) cblas_ztrmv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ztrmv"); if (cblas_ztrmv_p != NULL) cblas_ztrmv_p(layout, upper_lower, trans, unit_diag, n, a, lda, x, incx); @@ -1711,41 +1711,41 @@ static void cblas_ztrmv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_strsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, const int lda, - float *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const float* a, const int lda, + float* x, const int incx) { if (cblas_library() != NULL) { if (cblas_strsv_p == NULL) cblas_strsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const float *a, const int lda, - float *x, const int incx))GET_FUNC(h_libcblas, "cblas_strsv"); + CBLAS_DIAG unit_diag, const int n, const float* a, const int lda, + float* x, const int incx))GET_FUNC(h_libcblas, "cblas_strsv"); if (cblas_strsv_p != NULL) cblas_strsv_p(layout, upper_lower, trans, unit_diag, n, a, lda, x, incx); } } static void cblas_dtrsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, const int lda, - double *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const double* a, const int lda, + double* x, const int incx) { if (cblas_library() != NULL) { if (cblas_dtrsv_p == NULL) cblas_dtrsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const double *a, const int lda, - double *x, const int incx))GET_FUNC(h_libcblas, "cblas_dtrsv"); + CBLAS_DIAG unit_diag, const int n, const double* a, const int lda, + double* x, const int incx))GET_FUNC(h_libcblas, "cblas_dtrsv"); if (cblas_dtrsv_p != NULL) cblas_dtrsv_p(layout, upper_lower, trans, unit_diag, n, a, lda, x, incx); } } static void cblas_ctrsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, - void *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, + void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ctrsv_p == NULL) cblas_ctrsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ctrsv"); if (cblas_ctrsv_p != NULL) cblas_ctrsv_p(layout, upper_lower, trans, unit_diag, n, a, lda, x, incx); @@ -1753,13 +1753,13 @@ static void cblas_ctrsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL } static void cblas_ztrsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, - void *x, const int incx) { + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, + void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_ztrsv_p == NULL) cblas_ztrsv_p = (void (*)(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBLAS_TRANSPOSE trans, - CBLAS_DIAG unit_diag, const int n, const void *a, const int lda, void *x, + CBLAS_DIAG unit_diag, const int n, const void* a, const int lda, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_ztrsv"); if (cblas_ztrsv_p != NULL) cblas_ztrsv_p(layout, upper_lower, trans, unit_diag, n, a, lda, x, incx); @@ -1768,81 +1768,81 @@ static void cblas_ztrsv_wrapper(CBLAS_LAYOUT layout, CBLAS_UPLO upper_lower, CBL /* Level 1 */ -static float (*cblas_sasum_p)(const int n, const float *x, const int incx); -static double (*cblas_dasum_p)(const int n, const double *x, const int incx); -static float (*cblas_scasum_p)(const int n, const void *x, const int incx); -static double (*cblas_dzasum_p)(const int n, const void *x, const int incx); -static void (*cblas_saxpy_p)(const int n, const float alpha, const float *x, const int incx, - float *y, const int incy); -static void (*cblas_daxpy_p)(const int n, const double alpha, const double *x, const int incx, - double *y, const int incy); -static void (*cblas_caxpy_p)(const int n, const void *alpha, const void *x, const int incx, void *y, +static float (*cblas_sasum_p)(const int n, const float* x, const int incx); +static double (*cblas_dasum_p)(const int n, const double* x, const int incx); +static float (*cblas_scasum_p)(const int n, const void* x, const int incx); +static double (*cblas_dzasum_p)(const int n, const void* x, const int incx); +static void (*cblas_saxpy_p)(const int n, const float alpha, const float* x, const int incx, + float* y, const int incy); +static void (*cblas_daxpy_p)(const int n, const double alpha, const double* x, const int incx, + double* y, const int incy); +static void (*cblas_caxpy_p)(const int n, const void* alpha, const void* x, const int incx, void* y, const int incy); -static void (*cblas_zaxpy_p)(const int n, const void *alpha, const void *x, const int incx, void *y, +static void (*cblas_zaxpy_p)(const int n, const void* alpha, const void* x, const int incx, void* y, const int incy); -static void (*cblas_scopy_p)(const int n, const float *x, const int incx, float *y, const int incy); -static void (*cblas_dcopy_p)(const int n, const double *x, const int incx, double *y, +static void (*cblas_scopy_p)(const int n, const float* x, const int incx, float* y, const int incy); +static void (*cblas_dcopy_p)(const int n, const double* x, const int incx, double* y, const int incy); -static void (*cblas_ccopy_p)(const int n, const void *x, const int incx, void *y, const int incy); -static void (*cblas_zcopy_p)(const int n, const void *x, const int incx, void *y, const int incy); -static float (*cblas_sdot_p)(const int n, const float *x, const int incx, const float *y, +static void (*cblas_ccopy_p)(const int n, const void* x, const int incx, void* y, const int incy); +static void (*cblas_zcopy_p)(const int n, const void* x, const int incx, void* y, const int incy); +static float (*cblas_sdot_p)(const int n, const float* x, const int incx, const float* y, const int incy); -static double (*cblas_ddot_p)(const int n, const double *x, const int incx, const double *y, +static double (*cblas_ddot_p)(const int n, const double* x, const int incx, const double* y, const int incy); -static double (*cblas_dsdot_p)(const int n, const float *x, const int incx, const float *y, +static double (*cblas_dsdot_p)(const int n, const float* x, const int incx, const float* y, const int incy); -static float (*cblas_sdsdot_p)(const int n, const float sb, const float *x, const int incx, - const float *y, const int incy); -static float (*cblas_snrm2_p)(const int n, const float *x, const int incx); -static double (*cblas_dnrm2_p)(const int n, const double *x, const int incx); -static float (*cblas_scnrm2_p)(const int n, const void *x, const int incx); -static double (*cblas_dznrm2_p)(const int n, const void *x, const int incx); -static void (*cblas_srot_p)(const int n, float *x, const int incx, float *y, const int incy, +static float (*cblas_sdsdot_p)(const int n, const float sb, const float* x, const int incx, + const float* y, const int incy); +static float (*cblas_snrm2_p)(const int n, const float* x, const int incx); +static double (*cblas_dnrm2_p)(const int n, const double* x, const int incx); +static float (*cblas_scnrm2_p)(const int n, const void* x, const int incx); +static double (*cblas_dznrm2_p)(const int n, const void* x, const int incx); +static void (*cblas_srot_p)(const int n, float* x, const int incx, float* y, const int incy, const float c, const float s); -static void (*cblas_drot_p)(const int n, double *x, const int incx, double *y, const int incy, +static void (*cblas_drot_p)(const int n, double* x, const int incx, double* y, const int incy, const double c, const double s); -static void (*csrot_p)(const int *n, void *x, const int *incx, void *y, const int *incy, - const float *c, const float *s); -static void (*zdrot_p)(const int *n, void *x, const int *incx, void *y, const int *incy, - const double *c, const double *s); -static void (*cblas_srotg_p)(float *a, float *b, float *c, float *s); -static void (*cblas_drotg_p)(double *a, double *b, double *c, double *s); -static void (*crotg_p)(void *a, void *b, float *c, void *s); -static void (*zrotg_p)(void *a, void *b, double *c, void *s); -static void (*cblas_srotm_p)(const int n, float *x, const int incx, float *y, const int incy, - const float *param); -static void (*cblas_drotm_p)(const int n, double *x, const int incx, double *y, const int incy, - const double *param); -static void (*cblas_srotmg_p)(float *d1, float *d2, float *x1, float y1, float *param); -static void (*cblas_drotmg_p)(double *d1, double *d2, double *x1, double y1, double *param); -static void (*cblas_sscal_p)(const int n, const float alpha, float *x, const int incx); -static void (*cblas_dscal_p)(const int n, const double alpha, double *x, const int incx); -static void (*cblas_cscal_p)(const int n, const void *alpha, void *x, const int incx); -static void (*cblas_zscal_p)(const int n, const void *alpha, void *x, const int incx); -static void (*cblas_csscal_p)(const int n, const float alpha, void *x, const int incx); -static void (*cblas_zdscal_p)(const int n, const double alpha, void *x, const int incx); -static void (*cblas_sswap_p)(const int n, float *x, const int incx, float *y, const int incy); -static void (*cblas_dswap_p)(const int n, double *x, const int incx, double *y, const int incy); -static void (*cblas_cswap_p)(const int n, void *x, const int incx, void *y, const int incy); -static void (*cblas_zswap_p)(const int n, void *x, const int incx, void *y, const int incy); -static void (*cblas_cdotc_sub_p)(const int n, const void *x, const int incx, const void *y, - const int incy, void *pres); -static void (*cblas_zdotc_sub_p)(const int n, const void *x, const int incx, const void *y, - const int incy, void *pres); -static void (*cblas_cdotu_sub_p)(const int n, const void *x, const int incx, const void *y, - const int incy, void *pres); -static void (*cblas_zdotu_sub_p)(const int n, const void *x, const int incx, const void *y, - const int incy, void *pres); -static int (*cblas_isamax_p)(const int n, const float *x, const int incx); -static int (*cblas_idamax_p)(const int n, const double *x, const int incx); -static int (*cblas_icamax_p)(const int n, const void *x, const int incx); -static int (*cblas_izamax_p)(const int n, const void *x, const int incx); - -static float cblas_sasum_wrapper(const int n, const float *x, const int incx) { +static void (*csrot_p)(const int* n, void* x, const int* incx, void* y, const int* incy, + const float* c, const float* s); +static void (*zdrot_p)(const int* n, void* x, const int* incx, void* y, const int* incy, + const double* c, const double* s); +static void (*cblas_srotg_p)(float* a, float* b, float* c, float* s); +static void (*cblas_drotg_p)(double* a, double* b, double* c, double* s); +static void (*crotg_p)(void* a, void* b, float* c, void* s); +static void (*zrotg_p)(void* a, void* b, double* c, void* s); +static void (*cblas_srotm_p)(const int n, float* x, const int incx, float* y, const int incy, + const float* param); +static void (*cblas_drotm_p)(const int n, double* x, const int incx, double* y, const int incy, + const double* param); +static void (*cblas_srotmg_p)(float* d1, float* d2, float* x1, float y1, float* param); +static void (*cblas_drotmg_p)(double* d1, double* d2, double* x1, double y1, double* param); +static void (*cblas_sscal_p)(const int n, const float alpha, float* x, const int incx); +static void (*cblas_dscal_p)(const int n, const double alpha, double* x, const int incx); +static void (*cblas_cscal_p)(const int n, const void* alpha, void* x, const int incx); +static void (*cblas_zscal_p)(const int n, const void* alpha, void* x, const int incx); +static void (*cblas_csscal_p)(const int n, const float alpha, void* x, const int incx); +static void (*cblas_zdscal_p)(const int n, const double alpha, void* x, const int incx); +static void (*cblas_sswap_p)(const int n, float* x, const int incx, float* y, const int incy); +static void (*cblas_dswap_p)(const int n, double* x, const int incx, double* y, const int incy); +static void (*cblas_cswap_p)(const int n, void* x, const int incx, void* y, const int incy); +static void (*cblas_zswap_p)(const int n, void* x, const int incx, void* y, const int incy); +static void (*cblas_cdotc_sub_p)(const int n, const void* x, const int incx, const void* y, + const int incy, void* pres); +static void (*cblas_zdotc_sub_p)(const int n, const void* x, const int incx, const void* y, + const int incy, void* pres); +static void (*cblas_cdotu_sub_p)(const int n, const void* x, const int incx, const void* y, + const int incy, void* pres); +static void (*cblas_zdotu_sub_p)(const int n, const void* x, const int incx, const void* y, + const int incy, void* pres); +static int (*cblas_isamax_p)(const int n, const float* x, const int incx); +static int (*cblas_idamax_p)(const int n, const double* x, const int incx); +static int (*cblas_icamax_p)(const int n, const void* x, const int incx); +static int (*cblas_izamax_p)(const int n, const void* x, const int incx); + +static float cblas_sasum_wrapper(const int n, const float* x, const int incx) { float sasum_res = 0.0f; if (cblas_library() != NULL) { if (cblas_sasum_p == NULL) - cblas_sasum_p = (float (*)(const int n, const float *x, const int incx))GET_FUNC( + cblas_sasum_p = (float (*)(const int n, const float* x, const int incx))GET_FUNC( h_libcblas, "cblas_sasum"); if (cblas_sasum_p != NULL) sasum_res = cblas_sasum_p(n, x, incx); @@ -1850,11 +1850,11 @@ static float cblas_sasum_wrapper(const int n, const float *x, const int incx) { return sasum_res; } -static double cblas_dasum_wrapper(const int n, const double *x, const int incx) { +static double cblas_dasum_wrapper(const int n, const double* x, const int incx) { double dasum_res = 0.0; if (cblas_library() != NULL) { if (cblas_dasum_p == NULL) - cblas_dasum_p = (double (*)(const int n, const double *x, const int incx))GET_FUNC( + cblas_dasum_p = (double (*)(const int n, const double* x, const int incx))GET_FUNC( h_libcblas, "cblas_dasum"); if (cblas_dasum_p != NULL) dasum_res = cblas_dasum_p(n, x, incx); @@ -1862,11 +1862,11 @@ static double cblas_dasum_wrapper(const int n, const double *x, const int incx) return dasum_res; } -static float cblas_scasum_wrapper(const int n, const void *x, const int incx) { +static float cblas_scasum_wrapper(const int n, const void* x, const int incx) { float scasum_res = 0.0f; if (cblas_library() != NULL) { if (cblas_scasum_p == NULL) - cblas_scasum_p = (float (*)(const int n, const void *x, const int incx))GET_FUNC( + cblas_scasum_p = (float (*)(const int n, const void* x, const int incx))GET_FUNC( h_libcblas, "cblas_scasum"); if (cblas_scasum_p != NULL) scasum_res = cblas_scasum_p(n, x, incx); @@ -1874,11 +1874,11 @@ static float cblas_scasum_wrapper(const int n, const void *x, const int incx) { return scasum_res; } -static double cblas_dzasum_wrapper(const int n, const void *x, const int incx) { +static double cblas_dzasum_wrapper(const int n, const void* x, const int incx) { double dzasum_res = 0.0; if (cblas_library() != NULL) { if (cblas_dzasum_p == NULL) - cblas_dzasum_p = (double (*)(const int n, const void *x, const int incx))GET_FUNC( + cblas_dzasum_p = (double (*)(const int n, const void* x, const int incx))GET_FUNC( h_libcblas, "cblas_dzasum"); if (cblas_dzasum_p != NULL) dzasum_res = cblas_dzasum_p(n, x, incx); @@ -1886,102 +1886,102 @@ static double cblas_dzasum_wrapper(const int n, const void *x, const int incx) { return dzasum_res; } -static void cblas_saxpy_wrapper(const int n, const float alpha, const float *x, const int incx, - float *y, const int incy) { +static void cblas_saxpy_wrapper(const int n, const float alpha, const float* x, const int incx, + float* y, const int incy) { if (cblas_library() != NULL) { if (cblas_saxpy_p == NULL) cblas_saxpy_p = - (void (*)(const int n, const float alpha, const float *x, const int incx, float *y, + (void (*)(const int n, const float alpha, const float* x, const int incx, float* y, const int incy))GET_FUNC(h_libcblas, "cblas_saxpy"); if (cblas_saxpy_p != NULL) cblas_saxpy_p(n, alpha, x, incx, y, incy); } } -static void cblas_daxpy_wrapper(const int n, const double alpha, const double *x, const int incx, - double *y, const int incy) { +static void cblas_daxpy_wrapper(const int n, const double alpha, const double* x, const int incx, + double* y, const int incy) { if (cblas_library() != NULL) { if (cblas_daxpy_p == NULL) cblas_daxpy_p = - (void (*)(const int n, const double alpha, const double *x, const int incx, - double *y, const int incy))GET_FUNC(h_libcblas, "cblas_daxpy"); + (void (*)(const int n, const double alpha, const double* x, const int incx, + double* y, const int incy))GET_FUNC(h_libcblas, "cblas_daxpy"); if (cblas_daxpy_p != NULL) cblas_daxpy_p(n, alpha, x, incx, y, incy); } } -static void cblas_caxpy_wrapper(const int n, const void *alpha, const void *x, const int incx, - void *y, const int incy) { +static void cblas_caxpy_wrapper(const int n, const void* alpha, const void* x, const int incx, + void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_caxpy_p == NULL) - cblas_caxpy_p = (void (*)(const int n, const void *alpha, const void *x, const int incx, - void *y, const int incy))GET_FUNC(h_libcblas, "cblas_caxpy"); + cblas_caxpy_p = (void (*)(const int n, const void* alpha, const void* x, const int incx, + void* y, const int incy))GET_FUNC(h_libcblas, "cblas_caxpy"); if (cblas_caxpy_p != NULL) cblas_caxpy_p(n, alpha, x, incx, y, incy); } } -static void cblas_zaxpy_wrapper(const int n, const void *alpha, const void *x, const int incx, - void *y, const int incy) { +static void cblas_zaxpy_wrapper(const int n, const void* alpha, const void* x, const int incx, + void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_zaxpy_p == NULL) - cblas_zaxpy_p = (void (*)(const int n, const void *alpha, const void *x, const int incx, - void *y, const int incy))GET_FUNC(h_libcblas, "cblas_zaxpy"); + cblas_zaxpy_p = (void (*)(const int n, const void* alpha, const void* x, const int incx, + void* y, const int incy))GET_FUNC(h_libcblas, "cblas_zaxpy"); if (cblas_zaxpy_p != NULL) cblas_zaxpy_p(n, alpha, x, incx, y, incy); } } -static void cblas_scopy_wrapper(const int n, const float *x, const int incx, float *y, +static void cblas_scopy_wrapper(const int n, const float* x, const int incx, float* y, const int incy) { if (cblas_library() != NULL) { if (cblas_scopy_p == NULL) - cblas_scopy_p = (void (*)(const int n, const float *x, const int incx, float *y, + cblas_scopy_p = (void (*)(const int n, const float* x, const int incx, float* y, const int incy))GET_FUNC(h_libcblas, "cblas_scopy"); if (cblas_scopy_p != NULL) cblas_scopy_p(n, x, incx, y, incy); } } -static void cblas_dcopy_wrapper(const int n, const double *x, const int incx, double *y, +static void cblas_dcopy_wrapper(const int n, const double* x, const int incx, double* y, const int incy) { if (cblas_library() != NULL) { if (cblas_dcopy_p == NULL) - cblas_dcopy_p = (void (*)(const int n, const double *x, const int incx, double *y, + cblas_dcopy_p = (void (*)(const int n, const double* x, const int incx, double* y, const int incy))GET_FUNC(h_libcblas, "cblas_dcopy"); if (cblas_dcopy_p != NULL) cblas_dcopy_p(n, x, incx, y, incy); } } -static void cblas_ccopy_wrapper(const int n, const void *x, const int incx, void *y, +static void cblas_ccopy_wrapper(const int n, const void* x, const int incx, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_ccopy_p == NULL) - cblas_ccopy_p = (void (*)(const int n, const void *x, const int incx, void *y, + cblas_ccopy_p = (void (*)(const int n, const void* x, const int incx, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_ccopy"); if (cblas_ccopy_p != NULL) cblas_ccopy_p(n, x, incx, y, incy); } } -static void cblas_zcopy_wrapper(const int n, const void *x, const int incx, void *y, +static void cblas_zcopy_wrapper(const int n, const void* x, const int incx, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_zcopy_p == NULL) - cblas_zcopy_p = (void (*)(const int n, const void *x, const int incx, void *y, + cblas_zcopy_p = (void (*)(const int n, const void* x, const int incx, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_zcopy"); if (cblas_zcopy_p != NULL) cblas_zcopy_p(n, x, incx, y, incy); } } -static float cblas_sdot_wrapper(const int n, const float *x, const int incx, const float *y, +static float cblas_sdot_wrapper(const int n, const float* x, const int incx, const float* y, const int incy) { float sdot_res = 0.0f; if (cblas_library() != NULL) { if (cblas_sdot_p == NULL) - cblas_sdot_p = (float (*)(const int n, const float *x, const int incx, const float *y, + cblas_sdot_p = (float (*)(const int n, const float* x, const int incx, const float* y, const int incy))GET_FUNC(h_libcblas, "cblas_sdot"); if (cblas_sdot_p != NULL) sdot_res = cblas_sdot_p(n, x, incx, y, incy); @@ -1989,13 +1989,13 @@ static float cblas_sdot_wrapper(const int n, const float *x, const int incx, con return sdot_res; } -static double cblas_ddot_wrapper(const int n, const double *x, const int incx, const double *y, +static double cblas_ddot_wrapper(const int n, const double* x, const int incx, const double* y, const int incy) { double ddot_res = 0.0; if (cblas_library() != NULL) { if (cblas_ddot_p == NULL) cblas_ddot_p = - (double (*)(const int n, const double *x, const int incx, const double *y, + (double (*)(const int n, const double* x, const int incx, const double* y, const int incy))GET_FUNC(h_libcblas, "cblas_ddot"); if (cblas_ddot_p != NULL) ddot_res = cblas_ddot_p(n, x, incx, y, incy); @@ -2003,12 +2003,12 @@ static double cblas_ddot_wrapper(const int n, const double *x, const int incx, c return ddot_res; } -static double cblas_dsdot_wrapper(const int n, const float *x, const int incx, const float *y, +static double cblas_dsdot_wrapper(const int n, const float* x, const int incx, const float* y, const int incy) { double dsdot_res = 0.0; if (cblas_library() != NULL) { if (cblas_dsdot_p == NULL) - cblas_dsdot_p = (double (*)(const int n, const float *x, const int incx, const float *y, + cblas_dsdot_p = (double (*)(const int n, const float* x, const int incx, const float* y, const int incy))GET_FUNC(h_libcblas, "cblas_dsdot"); if (cblas_dsdot_p != NULL) dsdot_res = cblas_dsdot_p(n, x, incx, y, incy); @@ -2016,25 +2016,25 @@ static double cblas_dsdot_wrapper(const int n, const float *x, const int incx, c return dsdot_res; } -static float cblas_sdsdot_wrapper(const int n, const float sb, const float *x, const int incx, - const float *y, const int incy) { +static float cblas_sdsdot_wrapper(const int n, const float sb, const float* x, const int incx, + const float* y, const int incy) { float sdsdot_res = 0.0f; if (cblas_library() != NULL) { if (cblas_sdsdot_p == NULL) cblas_sdsdot_p = - (float (*)(const int n, const float sb, const float *x, const int incx, - const float *y, const int incy))GET_FUNC(h_libcblas, "cblas_sdsdot"); + (float (*)(const int n, const float sb, const float* x, const int incx, + const float* y, const int incy))GET_FUNC(h_libcblas, "cblas_sdsdot"); if (cblas_sdsdot_p != NULL) sdsdot_res = cblas_sdsdot_p(n, sb, x, incx, y, incy); } return sdsdot_res; } -static float cblas_snrm2_wrapper(const int n, const float *x, const int incx) { +static float cblas_snrm2_wrapper(const int n, const float* x, const int incx) { float snrm2_res = 0.0f; if (cblas_library() != NULL) { if (cblas_snrm2_p == NULL) - cblas_snrm2_p = (float (*)(const int n, const float *x, const int incx))GET_FUNC( + cblas_snrm2_p = (float (*)(const int n, const float* x, const int incx))GET_FUNC( h_libcblas, "cblas_snrm2"); if (cblas_snrm2_p != NULL) snrm2_res = cblas_snrm2_p(n, x, incx); @@ -2042,11 +2042,11 @@ static float cblas_snrm2_wrapper(const int n, const float *x, const int incx) { return snrm2_res; } -static double cblas_dnrm2_wrapper(const int n, const double *x, const int incx) { +static double cblas_dnrm2_wrapper(const int n, const double* x, const int incx) { double dnrm2_res = 0.0; if (cblas_library() != NULL) { if (cblas_dnrm2_p == NULL) - cblas_dnrm2_p = (double (*)(const int n, const double *x, const int incx))GET_FUNC( + cblas_dnrm2_p = (double (*)(const int n, const double* x, const int incx))GET_FUNC( h_libcblas, "cblas_dnrm2"); if (cblas_dnrm2_p != NULL) dnrm2_res = cblas_dnrm2_p(n, x, incx); @@ -2054,11 +2054,11 @@ static double cblas_dnrm2_wrapper(const int n, const double *x, const int incx) return dnrm2_res; } -static float cblas_scnrm2_wrapper(const int n, const void *x, const int incx) { +static float cblas_scnrm2_wrapper(const int n, const void* x, const int incx) { float scnrm2_res = 0.0f; if (cblas_library() != NULL) { if (cblas_scnrm2_p == NULL) - cblas_scnrm2_p = (float (*)(const int n, const void *x, const int incx))GET_FUNC( + cblas_scnrm2_p = (float (*)(const int n, const void* x, const int incx))GET_FUNC( h_libcblas, "cblas_scnrm2"); if (cblas_scnrm2_p != NULL) scnrm2_res = cblas_scnrm2_p(n, x, incx); @@ -2066,11 +2066,11 @@ static float cblas_scnrm2_wrapper(const int n, const void *x, const int incx) { return scnrm2_res; } -static double cblas_dznrm2_wrapper(const int n, const void *x, const int incx) { +static double cblas_dznrm2_wrapper(const int n, const void* x, const int incx) { double dznrm2_res = 0.0; if (cblas_library() != NULL) { if (cblas_dznrm2_p == NULL) - cblas_dznrm2_p = (double (*)(const int n, const void *x, const int incx))GET_FUNC( + cblas_dznrm2_p = (double (*)(const int n, const void* x, const int incx))GET_FUNC( h_libcblas, "cblas_dznrm2"); if (cblas_dznrm2_p != NULL) dznrm2_res = cblas_dznrm2_p(n, x, incx); @@ -2078,297 +2078,297 @@ static double cblas_dznrm2_wrapper(const int n, const void *x, const int incx) { return dznrm2_res; } -static void cblas_srot_wrapper(const int n, float *x, const int incx, float *y, const int incy, +static void cblas_srot_wrapper(const int n, float* x, const int incx, float* y, const int incy, const float c, const float s) { if (cblas_library() != NULL) { if (cblas_srot_p == NULL) cblas_srot_p = - (void (*)(const int n, float *x, const int incx, float *y, const int incy, + (void (*)(const int n, float* x, const int incx, float* y, const int incy, const float c, const float s))GET_FUNC(h_libcblas, "cblas_srot"); if (cblas_srot_p != NULL) cblas_srot_p(n, x, incx, y, incy, c, s); } } -static void cblas_drot_wrapper(const int n, double *x, const int incx, double *y, const int incy, +static void cblas_drot_wrapper(const int n, double* x, const int incx, double* y, const int incy, const double c, const double s) { if (cblas_library() != NULL) { if (cblas_drot_p == NULL) cblas_drot_p = - (void (*)(const int n, double *x, const int incx, double *y, const int incy, + (void (*)(const int n, double* x, const int incx, double* y, const int incy, const double c, const double s))GET_FUNC(h_libcblas, "cblas_drot"); if (cblas_drot_p != NULL) cblas_drot_p(n, x, incx, y, incy, c, s); } } -static void csrot_wrapper(const int *n, void *x, const int *incx, void *y, const int *incy, - const float *c, const float *s) { +static void csrot_wrapper(const int* n, void* x, const int* incx, void* y, const int* incy, + const float* c, const float* s) { if (blas_library() != NULL) { if (csrot_p == NULL) - csrot_p = (void (*)(const int *n, void *x, const int *incx, void *y, const int *incy, - const float *c, const float *s))GET_FUNC(h_libblas, "csrot_"); + csrot_p = (void (*)(const int* n, void* x, const int* incx, void* y, const int* incy, + const float* c, const float* s))GET_FUNC(h_libblas, "csrot_"); if (csrot_p == NULL) - csrot_p = (void (*)(const int *n, void *x, const int *incx, void *y, const int *incy, - const float *c, const float *s))GET_FUNC(h_libblas, "CSROT"); + csrot_p = (void (*)(const int* n, void* x, const int* incx, void* y, const int* incy, + const float* c, const float* s))GET_FUNC(h_libblas, "CSROT"); if (csrot_p != NULL) csrot_p(n, x, incx, y, incy, c, s); } } -static void zdrot_wrapper(const int *n, void *x, const int *incx, void *y, const int *incy, - const double *c, const double *s) { +static void zdrot_wrapper(const int* n, void* x, const int* incx, void* y, const int* incy, + const double* c, const double* s) { if (blas_library() != NULL) { if (zdrot_p == NULL) - zdrot_p = (void (*)(const int *n, void *x, const int *incx, void *y, const int *incy, - const double *c, const double *s))GET_FUNC(h_libblas, "zdrot_"); + zdrot_p = (void (*)(const int* n, void* x, const int* incx, void* y, const int* incy, + const double* c, const double* s))GET_FUNC(h_libblas, "zdrot_"); if (zdrot_p == NULL) - zdrot_p = (void (*)(const int *n, void *x, const int *incx, void *y, const int *incy, - const double *c, const double *s))GET_FUNC(h_libblas, "ZDROT"); + zdrot_p = (void (*)(const int* n, void* x, const int* incx, void* y, const int* incy, + const double* c, const double* s))GET_FUNC(h_libblas, "ZDROT"); if (zdrot_p != NULL) zdrot_p(n, x, incx, y, incy, c, s); } } -static void cblas_srotg_wrapper(float *a, float *b, float *c, float *s) { +static void cblas_srotg_wrapper(float* a, float* b, float* c, float* s) { if (cblas_library() != NULL) { if (cblas_srotg_p == NULL) - cblas_srotg_p = (void (*)(float *a, float *b, float *c, float *s))GET_FUNC( + cblas_srotg_p = (void (*)(float* a, float* b, float* c, float* s))GET_FUNC( h_libcblas, "cblas_srotg"); if (cblas_srotg_p != NULL) cblas_srotg_p(a, b, c, s); } } -static void cblas_drotg_wrapper(double *a, double *b, double *c, double *s) { +static void cblas_drotg_wrapper(double* a, double* b, double* c, double* s) { if (cblas_library() != NULL) { if (cblas_drotg_p == NULL) - cblas_drotg_p = (void (*)(double *a, double *b, double *c, double *s))GET_FUNC( + cblas_drotg_p = (void (*)(double* a, double* b, double* c, double* s))GET_FUNC( h_libcblas, "cblas_drotg"); if (cblas_drotg_p != NULL) cblas_drotg_p(a, b, c, s); } } -static void crotg_wrapper(void *a, void *b, float *c, void *s) { +static void crotg_wrapper(void* a, void* b, float* c, void* s) { if (blas_library() != NULL) { if (crotg_p == NULL) - crotg_p = (void (*)(void *a, void *b, float *c, void *s))GET_FUNC(h_libblas, "crotg_"); + crotg_p = (void (*)(void* a, void* b, float* c, void* s))GET_FUNC(h_libblas, "crotg_"); if (crotg_p == NULL) - crotg_p = (void (*)(void *a, void *b, float *c, void *s))GET_FUNC(h_libblas, "CROTG"); + crotg_p = (void (*)(void* a, void* b, float* c, void* s))GET_FUNC(h_libblas, "CROTG"); if (crotg_p != NULL) crotg_p(a, b, c, s); } } -static void zrotg_wrapper(void *a, void *b, double *c, void *s) { +static void zrotg_wrapper(void* a, void* b, double* c, void* s) { if (blas_library() != NULL) { if (zrotg_p == NULL) - zrotg_p = (void (*)(void *a, void *b, double *c, void *s))GET_FUNC(h_libblas, "zrotg_"); + zrotg_p = (void (*)(void* a, void* b, double* c, void* s))GET_FUNC(h_libblas, "zrotg_"); if (zrotg_p == NULL) - zrotg_p = (void (*)(void *a, void *b, double *c, void *s))GET_FUNC(h_libblas, "ZROTG"); + zrotg_p = (void (*)(void* a, void* b, double* c, void* s))GET_FUNC(h_libblas, "ZROTG"); if (zrotg_p != NULL) zrotg_p(a, b, c, s); } } -static void cblas_srotm_wrapper(const int n, float *x, const int incx, float *y, const int incy, - const float *param) { +static void cblas_srotm_wrapper(const int n, float* x, const int incx, float* y, const int incy, + const float* param) { if (cblas_library() != NULL) { if (cblas_srotm_p == NULL) cblas_srotm_p = - (void (*)(const int n, float *x, const int incx, float *y, const int incy, - const float *param))GET_FUNC(h_libcblas, "cblas_srotm"); + (void (*)(const int n, float* x, const int incx, float* y, const int incy, + const float* param))GET_FUNC(h_libcblas, "cblas_srotm"); if (cblas_srotm_p != NULL) cblas_srotm_p(n, x, incx, y, incy, param); } } -static void cblas_drotm_wrapper(const int n, double *x, const int incx, double *y, const int incy, - const double *param) { +static void cblas_drotm_wrapper(const int n, double* x, const int incx, double* y, const int incy, + const double* param) { if (cblas_library() != NULL) { if (cblas_drotm_p == NULL) cblas_drotm_p = - (void (*)(const int n, double *x, const int incx, double *y, const int incy, - const double *param))GET_FUNC(h_libcblas, "cblas_drotm"); + (void (*)(const int n, double* x, const int incx, double* y, const int incy, + const double* param))GET_FUNC(h_libcblas, "cblas_drotm"); if (cblas_drotm_p != NULL) cblas_drotm_p(n, x, incx, y, incy, param); } } -static void cblas_srotmg_wrapper(float *d1, float *d2, float *x1, float y1, float *param) { +static void cblas_srotmg_wrapper(float* d1, float* d2, float* x1, float y1, float* param) { if (cblas_library() != NULL) { if (cblas_srotmg_p == NULL) - cblas_srotmg_p = (void (*)(float *d1, float *d2, float *x1, float y1, - float *param))GET_FUNC(h_libcblas, "cblas_srotmg"); + cblas_srotmg_p = (void (*)(float* d1, float* d2, float* x1, float y1, + float* param))GET_FUNC(h_libcblas, "cblas_srotmg"); if (cblas_srotmg_p != NULL) cblas_srotmg_p(d1, d2, x1, y1, param); } } -static void cblas_drotmg_wrapper(double *d1, double *d2, double *x1, double y1, double *param) { +static void cblas_drotmg_wrapper(double* d1, double* d2, double* x1, double y1, double* param) { if (cblas_library() != NULL) { if (cblas_drotmg_p == NULL) - cblas_drotmg_p = (void (*)(double *d1, double *d2, double *x1, double y1, - double *param))GET_FUNC(h_libcblas, "cblas_drotmg"); + cblas_drotmg_p = (void (*)(double* d1, double* d2, double* x1, double y1, + double* param))GET_FUNC(h_libcblas, "cblas_drotmg"); if (cblas_drotmg_p != NULL) cblas_drotmg_p(d1, d2, x1, y1, param); } } -static void cblas_sscal_wrapper(const int n, const float alpha, float *x, const int incx) { +static void cblas_sscal_wrapper(const int n, const float alpha, float* x, const int incx) { if (cblas_library() != NULL) { if (cblas_sscal_p == NULL) - cblas_sscal_p = (void (*)(const int n, const float alpha, float *x, + cblas_sscal_p = (void (*)(const int n, const float alpha, float* x, const int incx))GET_FUNC(h_libcblas, "cblas_sscal"); if (cblas_sscal_p != NULL) cblas_sscal_p(n, alpha, x, incx); } } -static void cblas_dscal_wrapper(const int n, const double alpha, double *x, const int incx) { +static void cblas_dscal_wrapper(const int n, const double alpha, double* x, const int incx) { if (cblas_library() != NULL) { if (cblas_dscal_p == NULL) - cblas_dscal_p = (void (*)(const int n, const double alpha, double *x, + cblas_dscal_p = (void (*)(const int n, const double alpha, double* x, const int incx))GET_FUNC(h_libcblas, "cblas_dscal"); if (cblas_dscal_p != NULL) cblas_dscal_p(n, alpha, x, incx); } } -static void cblas_cscal_wrapper(const int n, const void *alpha, void *x, const int incx) { +static void cblas_cscal_wrapper(const int n, const void* alpha, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_cscal_p == NULL) - cblas_cscal_p = (void (*)(const int n, const void *alpha, void *x, + cblas_cscal_p = (void (*)(const int n, const void* alpha, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_cscal"); if (cblas_cscal_p != NULL) cblas_cscal_p(n, alpha, x, incx); } } -static void cblas_zscal_wrapper(const int n, const void *alpha, void *x, const int incx) { +static void cblas_zscal_wrapper(const int n, const void* alpha, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_zscal_p == NULL) - cblas_zscal_p = (void (*)(const int n, const void *alpha, void *x, + cblas_zscal_p = (void (*)(const int n, const void* alpha, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_zscal"); if (cblas_zscal_p != NULL) cblas_zscal_p(n, alpha, x, incx); } } -static void cblas_csscal_wrapper(const int n, const float alpha, void *x, const int incx) { +static void cblas_csscal_wrapper(const int n, const float alpha, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_csscal_p == NULL) - cblas_csscal_p = (void (*)(const int n, const float alpha, void *x, + cblas_csscal_p = (void (*)(const int n, const float alpha, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_csscal"); if (cblas_csscal_p != NULL) cblas_csscal_p(n, alpha, x, incx); } } -static void cblas_zdscal_wrapper(const int n, const double alpha, void *x, const int incx) { +static void cblas_zdscal_wrapper(const int n, const double alpha, void* x, const int incx) { if (cblas_library() != NULL) { if (cblas_zdscal_p == NULL) - cblas_zdscal_p = (void (*)(const int n, const double alpha, void *x, + cblas_zdscal_p = (void (*)(const int n, const double alpha, void* x, const int incx))GET_FUNC(h_libcblas, "cblas_zdscal"); if (cblas_zdscal_p != NULL) cblas_zdscal_p(n, alpha, x, incx); } } -static void cblas_sswap_wrapper(const int n, float *x, const int incx, float *y, const int incy) { +static void cblas_sswap_wrapper(const int n, float* x, const int incx, float* y, const int incy) { if (cblas_library() != NULL) { if (cblas_sswap_p == NULL) - cblas_sswap_p = (void (*)(const int n, float *x, const int incx, float *y, + cblas_sswap_p = (void (*)(const int n, float* x, const int incx, float* y, const int incy))GET_FUNC(h_libcblas, "cblas_sswap"); if (cblas_sswap_p != NULL) cblas_sswap_p(n, x, incx, y, incy); } } -static void cblas_dswap_wrapper(const int n, double *x, const int incx, double *y, const int incy) { +static void cblas_dswap_wrapper(const int n, double* x, const int incx, double* y, const int incy) { if (cblas_library() != NULL) { if (cblas_dswap_p == NULL) - cblas_dswap_p = (void (*)(const int n, double *x, const int incx, double *y, + cblas_dswap_p = (void (*)(const int n, double* x, const int incx, double* y, const int incy))GET_FUNC(h_libcblas, "cblas_dswap"); if (cblas_dswap_p != NULL) cblas_dswap_p(n, x, incx, y, incy); } } -static void cblas_cswap_wrapper(const int n, void *x, const int incx, void *y, const int incy) { +static void cblas_cswap_wrapper(const int n, void* x, const int incx, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_cswap_p == NULL) - cblas_cswap_p = (void (*)(const int n, void *x, const int incx, void *y, + cblas_cswap_p = (void (*)(const int n, void* x, const int incx, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_cswap"); if (cblas_cswap_p != NULL) cblas_cswap_p(n, x, incx, y, incy); } } -static void cblas_zswap_wrapper(const int n, void *x, const int incx, void *y, const int incy) { +static void cblas_zswap_wrapper(const int n, void* x, const int incx, void* y, const int incy) { if (cblas_library() != NULL) { if (cblas_zswap_p == NULL) - cblas_zswap_p = (void (*)(const int n, void *x, const int incx, void *y, + cblas_zswap_p = (void (*)(const int n, void* x, const int incx, void* y, const int incy))GET_FUNC(h_libcblas, "cblas_zswap"); if (cblas_zswap_p != NULL) cblas_zswap_p(n, x, incx, y, incy); } } -static void cblas_cdotc_sub_wrapper(const int n, const void *x, const int incx, const void *y, - const int incy, void *pres) { +static void cblas_cdotc_sub_wrapper(const int n, const void* x, const int incx, const void* y, + const int incy, void* pres) { if (cblas_library() != NULL) { if (cblas_cdotc_sub_p == NULL) cblas_cdotc_sub_p = - (void (*)(const int n, const void *x, const int incx, const void *y, const int incy, - void *pres))GET_FUNC(h_libcblas, "cblas_cdotc_sub"); + (void (*)(const int n, const void* x, const int incx, const void* y, const int incy, + void* pres))GET_FUNC(h_libcblas, "cblas_cdotc_sub"); if (cblas_cdotc_sub_p != NULL) cblas_cdotc_sub_p(n, x, incx, y, incy, pres); } } -static void cblas_zdotc_sub_wrapper(const int n, const void *x, const int incx, const void *y, - const int incy, void *pres) { +static void cblas_zdotc_sub_wrapper(const int n, const void* x, const int incx, const void* y, + const int incy, void* pres) { if (cblas_library() != NULL) { if (cblas_zdotc_sub_p == NULL) cblas_zdotc_sub_p = - (void (*)(const int n, const void *x, const int incx, const void *y, const int incy, - void *pres))GET_FUNC(h_libcblas, "cblas_zdotc_sub"); + (void (*)(const int n, const void* x, const int incx, const void* y, const int incy, + void* pres))GET_FUNC(h_libcblas, "cblas_zdotc_sub"); if (cblas_zdotc_sub_p != NULL) cblas_zdotc_sub_p(n, x, incx, y, incy, pres); } } -static void cblas_cdotu_sub_wrapper(const int n, const void *x, const int incx, const void *y, - const int incy, void *pres) { +static void cblas_cdotu_sub_wrapper(const int n, const void* x, const int incx, const void* y, + const int incy, void* pres) { if (cblas_library() != NULL) { if (cblas_cdotu_sub_p == NULL) cblas_cdotu_sub_p = - (void (*)(const int n, const void *x, const int incx, const void *y, const int incy, - void *pres))GET_FUNC(h_libcblas, "cblas_cdotu_sub"); + (void (*)(const int n, const void* x, const int incx, const void* y, const int incy, + void* pres))GET_FUNC(h_libcblas, "cblas_cdotu_sub"); if (cblas_cdotu_sub_p != NULL) cblas_cdotu_sub_p(n, x, incx, y, incy, pres); } } -static void cblas_zdotu_sub_wrapper(const int n, const void *x, const int incx, const void *y, - const int incy, void *pres) { +static void cblas_zdotu_sub_wrapper(const int n, const void* x, const int incx, const void* y, + const int incy, void* pres) { if (cblas_library() != NULL) { if (cblas_zdotu_sub_p == NULL) cblas_zdotu_sub_p = - (void (*)(const int n, const void *x, const int incx, const void *y, const int incy, - void *pres))GET_FUNC(h_libcblas, "cblas_zdotu_sub"); + (void (*)(const int n, const void* x, const int incx, const void* y, const int incy, + void* pres))GET_FUNC(h_libcblas, "cblas_zdotu_sub"); if (cblas_zdotu_sub_p != NULL) cblas_zdotu_sub_p(n, x, incx, y, incy, pres); } } -static int cblas_isamax_wrapper(const int n, const float *x, const int incx) { +static int cblas_isamax_wrapper(const int n, const float* x, const int incx) { int isamax_res = 0; if (cblas_library() != NULL) { if (cblas_isamax_p == NULL) - cblas_isamax_p = (int (*)(const int n, const float *x, const int incx))GET_FUNC( + cblas_isamax_p = (int (*)(const int n, const float* x, const int incx))GET_FUNC( h_libcblas, "cblas_isamax"); if (cblas_isamax_p != NULL) isamax_res = cblas_isamax_p(n, x, incx); @@ -2376,11 +2376,11 @@ static int cblas_isamax_wrapper(const int n, const float *x, const int incx) { return isamax_res; } -static int cblas_idamax_wrapper(const int n, const double *x, const int incx) { +static int cblas_idamax_wrapper(const int n, const double* x, const int incx) { int idamax_res = 0; if (cblas_library() != NULL) { if (cblas_idamax_p == NULL) - cblas_idamax_p = (int (*)(const int n, const double *x, const int incx))GET_FUNC( + cblas_idamax_p = (int (*)(const int n, const double* x, const int incx))GET_FUNC( h_libcblas, "cblas_idamax"); if (cblas_idamax_p != NULL) idamax_res = cblas_idamax_p(n, x, incx); @@ -2388,11 +2388,11 @@ static int cblas_idamax_wrapper(const int n, const double *x, const int incx) { return idamax_res; } -static int cblas_icamax_wrapper(const int n, const void *x, const int incx) { +static int cblas_icamax_wrapper(const int n, const void* x, const int incx) { int icamax_res = 0; if (cblas_library() != NULL) { if (cblas_icamax_p == NULL) - cblas_icamax_p = (int (*)(const int n, const void *x, const int incx))GET_FUNC( + cblas_icamax_p = (int (*)(const int n, const void* x, const int incx))GET_FUNC( h_libcblas, "cblas_icamax"); if (cblas_icamax_p != NULL) icamax_res = cblas_icamax_p(n, x, incx); @@ -2400,11 +2400,11 @@ static int cblas_icamax_wrapper(const int n, const void *x, const int incx) { return icamax_res; } -static int cblas_izamax_wrapper(const int n, const void *x, const int incx) { +static int cblas_izamax_wrapper(const int n, const void* x, const int incx) { int izamax_res = 0; if (cblas_library() != NULL) { if (cblas_izamax_p == NULL) - cblas_izamax_p = (int (*)(const int n, const void *x, const int incx))GET_FUNC( + cblas_izamax_p = (int (*)(const int n, const void* x, const int incx))GET_FUNC( h_libcblas, "cblas_izamax"); if (cblas_izamax_p != NULL) izamax_res = cblas_izamax_p(n, x, incx); diff --git a/tests/unit_tests/blas/include/test_common.hpp b/tests/unit_tests/blas/include/test_common.hpp index 25620f4fb..64df0bd76 100644 --- a/tests/unit_tests/blas/include/test_common.hpp +++ b/tests/unit_tests/blas/include/test_common.hpp @@ -79,14 +79,15 @@ constexpr T matrix_size(oneapi::math::transpose trans, T m, T n, T ldm) { return outer_dimension(trans, m, n) * ldm; } template -constexpr T matrix_size(oneapi::math::layout layout, oneapi::math::transpose trans, T m, T n, T ldm) { +constexpr T matrix_size(oneapi::math::layout layout, oneapi::math::transpose trans, T m, T n, + T ldm) { return (layout == oneapi::math::layout::col_major) ? outer_dimension(trans, m, n) * ldm - : inner_dimension(trans, m, n) * ldm; + : inner_dimension(trans, m, n) * ldm; } // SYCL buffer creation helper. template -sycl::buffer make_buffer(const vec &v) { +sycl::buffer make_buffer(const vec& v) { sycl::buffer buf(v.data(), sycl::range<1>(v.size())); return buf; } @@ -174,14 +175,14 @@ std::complex rand_scalar(int mag) { } template -void rand_vector(fp *v, int n, int inc) { +void rand_vector(fp* v, int n, int inc) { int abs_inc = std::abs(inc); for (int i = 0; i < n; i++) v[i * abs_inc] = rand_scalar(); } template -void rand_vector(vec &v, int n, int inc) { +void rand_vector(vec& v, int n, int inc) { using fp = typename vec::value_type; int abs_inc = std::abs(inc); @@ -209,7 +210,7 @@ oneapi::math::transpose rand_trans() { } template -void print_matrix(vec &M, oneapi::math::transpose trans, int m, int n, int ld, char *name) { +void print_matrix(vec& M, oneapi::math::transpose trans, int m, int n, int ld, char* name) { std::cout << "Matrix " << name << ":\n"; for (int i = 0; i < m; i++) { for (int j = 0; j < n; j++) { @@ -223,15 +224,15 @@ void print_matrix(vec &M, oneapi::math::transpose trans, int m, int n, int ld, c } template -void copy_vector(fp *src, int n, int inc, fp *dest) { +void copy_vector(fp* src, int n, int inc, fp* dest) { int abs_inc = std::abs(inc); for (int i = 0; i < n; i++) dest[i * abs_inc] = src[i * abs_inc]; } template -void copy_matrix(vec_src &src, oneapi::math::layout layout, oneapi::math::transpose trans, int m, - int n, int ld, vec_dest &dest) { +void copy_matrix(vec_src& src, oneapi::math::layout layout, oneapi::math::transpose trans, int m, + int n, int ld, vec_dest& dest) { using T_data = typename vec_dest::value_type; dest.resize(matrix_size(layout, trans, m, n, ld)); if (((trans == oneapi::math::transpose::nontrans) && @@ -250,8 +251,8 @@ void copy_matrix(vec_src &src, oneapi::math::layout layout, oneapi::math::transp } template -void copy_matrix(fp_src *src, oneapi::math::layout layout, oneapi::math::transpose trans, int m, - int n, int ld, fp_dst *dest) { +void copy_matrix(fp_src* src, oneapi::math::layout layout, oneapi::math::transpose trans, int m, + int n, int ld, fp_dst* dest) { if (((trans == oneapi::math::transpose::nontrans) && (layout == oneapi::math::layout::col_major)) || ((trans != oneapi::math::transpose::nontrans) && @@ -268,7 +269,7 @@ void copy_matrix(fp_src *src, oneapi::math::layout layout, oneapi::math::transpo } template -void rand_matrix(vec &M, oneapi::math::transpose trans, int m, int n, int ld) { +void rand_matrix(vec& M, oneapi::math::transpose trans, int m, int n, int ld) { using fp = typename vec::value_type; M.resize(matrix_size(trans, m, n, ld)); @@ -286,7 +287,7 @@ void rand_matrix(vec &M, oneapi::math::transpose trans, int m, int n, int ld) { } template -void rand_matrix(vec &M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, +void rand_matrix(vec& M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, int ld) { using fp = typename vec::value_type; @@ -308,7 +309,7 @@ void rand_matrix(vec &M, oneapi::math::layout layout, oneapi::math::transpose tr } template -void rand_matrix(fp *M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, +void rand_matrix(fp* M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, int ld) { if (((trans == oneapi::math::transpose::nontrans) && (layout == oneapi::math::layout::col_major)) || @@ -326,7 +327,7 @@ void rand_matrix(fp *M, oneapi::math::layout layout, oneapi::math::transpose tra } template -void rand_trsm_matrix(vec &M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, +void rand_trsm_matrix(vec& M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, int ld) { using fp = typename vec::value_type; @@ -356,8 +357,8 @@ void rand_trsm_matrix(vec &M, oneapi::math::layout layout, oneapi::math::transpo } template -void rand_trsm_matrix(fp *M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, int n, - int ld) { +void rand_trsm_matrix(fp* M, oneapi::math::layout layout, oneapi::math::transpose trans, int m, + int n, int ld) { if (((trans == oneapi::math::transpose::nontrans) && (layout == oneapi::math::layout::col_major)) || ((trans != oneapi::math::transpose::nontrans) && @@ -382,7 +383,7 @@ void rand_trsm_matrix(fp *M, oneapi::math::layout layout, oneapi::math::transpos } template -void rand_tpsv_matrix(vec &M, oneapi::math::layout layout, oneapi::math::uplo upper_lower, +void rand_tpsv_matrix(vec& M, oneapi::math::layout layout, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, int m) { using fp = typename vec::value_type; std::vector tmp; @@ -408,7 +409,7 @@ void rand_tpsv_matrix(vec &M, oneapi::math::layout layout, oneapi::math::uplo up } template -void rand_tbsv_matrix(vec &M, oneapi::math::layout layout, oneapi::math::uplo upper_lower, +void rand_tbsv_matrix(vec& M, oneapi::math::layout layout, oneapi::math::uplo upper_lower, oneapi::math::transpose trans, int m, int k, int ld) { using fp = typename vec::value_type; std::vector tmp; @@ -461,7 +462,7 @@ typename std::enable_if::value, bool>::type check_equal(fp } template -bool check_equal_ptr(sycl::queue queue, fp *x, fp x_ref, int error_mag) { +bool check_equal_ptr(sycl::queue queue, fp* x, fp x_ref, int error_mag) { fp x_host; queue.memcpy(&x_host, x, sizeof(fp)).wait(); return check_equal(x_host, x_ref, error_mag); @@ -485,7 +486,7 @@ bool check_equal_trsm(fp x, fp x_ref, int error_mag) { } template -bool check_equal(fp x, fp x_ref, int error_mag, std::ostream &out) { +bool check_equal(fp x, fp x_ref, int error_mag, std::ostream& out) { bool good = check_equal(x, x_ref, error_mag); if (!good) { @@ -495,15 +496,15 @@ bool check_equal(fp x, fp x_ref, int error_mag, std::ostream &out) { } template -bool check_equal_ptr(sycl::queue queue, fp *x, fp x_ref, int error_mag, std::ostream &out) { +bool check_equal_ptr(sycl::queue queue, fp* x, fp x_ref, int error_mag, std::ostream& out) { fp x_host; queue.memcpy(&x_host, x, sizeof(fp)).wait(); return check_equal(x_host, x_ref, error_mag, out); } template -bool check_equal_vector(const fp *v, const fp *v_ref, int n, int inc, int error_mag, - std::ostream &out) { +bool check_equal_vector(const fp* v, const fp* v_ref, int n, int inc, int error_mag, + std::ostream& out) { int abs_inc = std::abs(inc), count = 0; bool good = true; @@ -523,7 +524,7 @@ bool check_equal_vector(const fp *v, const fp *v_ref, int n, int inc, int error_ } template -bool check_equal_vector(vec1 &v, vec2 &v_ref, int n, int inc, int error_mag, std::ostream &out) { +bool check_equal_vector(vec1& v, vec2& v_ref, int n, int inc, int error_mag, std::ostream& out) { int abs_inc = std::abs(inc), count = 0; bool good = true; @@ -543,8 +544,8 @@ bool check_equal_vector(vec1 &v, vec2 &v_ref, int n, int inc, int error_mag, std } template -bool check_equal_trsv_vector(vec1 &v, vec2 &v_ref, int n, int inc, int error_mag, - std::ostream &out) { +bool check_equal_trsv_vector(vec1& v, vec2& v_ref, int n, int inc, int error_mag, + std::ostream& out) { int abs_inc = std::abs(inc), count = 0; bool good = true; @@ -564,8 +565,8 @@ bool check_equal_trsv_vector(vec1 &v, vec2 &v_ref, int n, int inc, int error_mag } template -bool check_equal_matrix(acc1 &M, acc2 &M_ref, oneapi::math::layout layout, int m, int n, int ld, - int error_mag, std::ostream &out) { +bool check_equal_matrix(acc1& M, acc2& M_ref, oneapi::math::layout layout, int m, int n, int ld, + int error_mag, std::ostream& out) { bool good = true; int idx, count = 0; for (int j = 0; j < n; j++) { @@ -586,8 +587,8 @@ bool check_equal_matrix(acc1 &M, acc2 &M_ref, oneapi::math::layout layout, int m } template -bool check_equal_matrix(const fp *M, const fp *M_ref, oneapi::math::layout layout, int m, int n, - int ld, int error_mag, std::ostream &out) { +bool check_equal_matrix(const fp* M, const fp* M_ref, oneapi::math::layout layout, int m, int n, + int ld, int error_mag, std::ostream& out) { bool good = true; int idx, count = 0; for (int j = 0; j < n; j++) { @@ -608,9 +609,9 @@ bool check_equal_matrix(const fp *M, const fp *M_ref, oneapi::math::layout layou } template -bool check_equal_matrix(acc1 &M, acc2 &M_ref, oneapi::math::layout layout, +bool check_equal_matrix(acc1& M, acc2& M_ref, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int m, int n, int ld, int error_mag, - std::ostream &out) { + std::ostream& out) { bool good = true; int idx, count = 0; for (int j = 0; j < n; j++) { @@ -634,8 +635,8 @@ bool check_equal_matrix(acc1 &M, acc2 &M_ref, oneapi::math::layout layout, } template -bool check_equal_trsm_matrix(acc1 &M, acc2 &M_ref, oneapi::math::layout layout, int m, int n, int ld, - int error_mag, std::ostream &out) { +bool check_equal_trsm_matrix(acc1& M, acc2& M_ref, oneapi::math::layout layout, int m, int n, + int ld, int error_mag, std::ostream& out) { bool good = true; int idx, count = 0; for (int j = 0; j < n; j++) { @@ -677,8 +678,8 @@ typename std::enable_if::value, bool>::type check_almost_eq } template -bool check_almost_equal_matrix_int(Ta &M, Tb &M_ref, oneapi::math::layout layout, int m, int n, - int ld, int error_mag, std::ostream &out) { +bool check_almost_equal_matrix_int(Ta& M, Tb& M_ref, oneapi::math::layout layout, int m, int n, + int ld, int error_mag, std::ostream& out) { static_assert(is_matrix_type_integral() && is_matrix_type_integral()); bool good = true; int idx, count = 0; @@ -700,8 +701,8 @@ bool check_almost_equal_matrix_int(Ta &M, Tb &M_ref, oneapi::math::layout layout } template -bool check_almost_equal_matrix(Ta &M, Tb &M_ref, oneapi::math::layout layout, int m, int n, int ld, - int error_mag, std::ostream &out) { +bool check_almost_equal_matrix(Ta& M, Tb& M_ref, oneapi::math::layout layout, int m, int n, int ld, + int error_mag, std::ostream& out) { // Only call if returned dtype is integral if constexpr (is_matrix_type_integral() && is_matrix_type_integral()) return check_almost_equal_matrix_int(M, M_ref, layout, m, n, ld, error_mag, out); diff --git a/tests/unit_tests/blas/level1/asum.cpp b/tests/unit_tests/blas/level1/asum.cpp index 60263989b..a2caef8d9 100644 --- a/tests/unit_tests/blas/level1/asum.cpp +++ b/tests/unit_tests/blas/level1/asum.cpp @@ -83,7 +83,8 @@ int test(device* dev, oneapi::math::layout layout, int64_t N, int64_t incx) { #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::asum(main_queue, N, x_buffer, incx, result_buffer); + oneapi::math::blas::column_major::asum(main_queue, N, x_buffer, incx, + result_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::asum(main_queue, N, x_buffer, incx, result_buffer); @@ -97,8 +98,8 @@ int test(device* dev, oneapi::math::layout layout, int64_t N, int64_t incx) { x_buffer, incx, result_buffer); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::asum, N, x_buffer, - incx, result_buffer); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::asum, N, + x_buffer, incx, result_buffer); break; default: break; } diff --git a/tests/unit_tests/blas/level1/asum_usm.cpp b/tests/unit_tests/blas/level1/asum_usm.cpp index 8d15c1407..64be04afc 100644 --- a/tests/unit_tests/blas/level1/asum_usm.cpp +++ b/tests/unit_tests/blas/level1/asum_usm.cpp @@ -97,11 +97,11 @@ int test(device* dev, oneapi::math::layout layout, int64_t N, int64_t incx) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::asum(main_queue, N, x.data(), incx, - result_p, dependencies); + result_p, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::asum(main_queue, N, x.data(), incx, result_p, - dependencies); + dependencies); break; default: break; } @@ -113,8 +113,8 @@ int test(device* dev, oneapi::math::layout layout, int64_t N, int64_t incx) { x.data(), incx, result_p, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::asum, N, x.data(), - incx, result_p, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::asum, N, + x.data(), incx, result_p, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level1/axpby.cpp b/tests/unit_tests/blas/level1/axpby.cpp index 6b417e913..3b063404b 100644 --- a/tests/unit_tests/blas/level1/axpby.cpp +++ b/tests/unit_tests/blas/level1/axpby.cpp @@ -41,12 +41,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha, fp beta) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha, fp beta) { // Prepare data. vector x, y, y_ref; @@ -58,18 +58,18 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::axpby(&N_ref, (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)&beta, - (fp_ref *)y_ref.data(), &incy_ref); + ::axpby(&N_ref, (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)&beta, + (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ AXPBY. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during AXPBY:\n" << e.what() << std::endl; print_error_code(e); @@ -87,11 +87,11 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::axpby(main_queue, N, alpha, x_buffer, incx, beta, - y_buffer, incy); + y_buffer, incy); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::axpby(main_queue, N, alpha, x_buffer, incx, beta, - y_buffer, incy); + y_buffer, incy); break; default: break; } @@ -109,16 +109,16 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during AXPBY:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of AXPBY:\n" << error.what() << std::endl; } @@ -131,7 +131,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp } class AxpbyTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpbyTests, RealSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level1/axpby_usm.cpp b/tests/unit_tests/blas/level1/axpby_usm.cpp index 152879329..3bd8b4a71 100644 --- a/tests/unit_tests/blas/level1/axpby_usm.cpp +++ b/tests/unit_tests/blas/level1/axpby_usm.cpp @@ -41,19 +41,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha, fp beta) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha, fp beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during AXPBY:\n" << e.what() << std::endl; print_error_code(e); @@ -79,8 +79,8 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::axpby(&N_ref, (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)&beta, - (fp_ref *)y_ref.data(), &incy_ref); + ::axpby(&N_ref, (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)&beta, + (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ AXPBY. @@ -89,11 +89,11 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::axpby(main_queue, N, alpha, x.data(), incx, - beta, y.data(), incy, dependencies); + beta, y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::axpby(main_queue, N, alpha, x.data(), incx, - beta, y.data(), incy, dependencies); + beta, y.data(), incy, dependencies); break; default: break; } @@ -113,16 +113,16 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during AXPBY:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of AXPBY:\n" << error.what() << std::endl; } @@ -134,7 +134,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp } class AxpbyUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpbyUsmTests, RealSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level1/axpy.cpp b/tests/unit_tests/blas/level1/axpy.cpp index dc38cb718..5c6557de6 100644 --- a/tests/unit_tests/blas/level1/axpy.cpp +++ b/tests/unit_tests/blas/level1/axpy.cpp @@ -41,12 +41,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha) { // Prepare data. vector x, y, y_ref; @@ -58,18 +58,17 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::axpy(&N_ref, (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y_ref.data(), - &incy_ref); + ::axpy(&N_ref, (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ AXPY. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during AXPY:\n" << e.what() << std::endl; print_error_code(e); @@ -87,19 +86,19 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::axpy(main_queue, N, alpha, x_buffer, incx, - y_buffer, incy); + y_buffer, incy); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::axpy(main_queue, N, alpha, x_buffer, incx, y_buffer, - incy); + incy); break; default: break; } #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpy, N, alpha, - x_buffer, incx, y_buffer, incy); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpy, N, + alpha, x_buffer, incx, y_buffer, incy); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::axpy, N, alpha, @@ -109,16 +108,16 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during AXPY:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of AXPY:\n" << error.what() << std::endl; } @@ -130,7 +129,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp return (int)good; } -class AxpyTests : public ::testing::TestWithParam> { +class AxpyTests : public ::testing::TestWithParam> { }; TEST_P(AxpyTests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level1/axpy_usm.cpp b/tests/unit_tests/blas/level1/axpy_usm.cpp index d24c500a6..737c44de3 100644 --- a/tests/unit_tests/blas/level1/axpy_usm.cpp +++ b/tests/unit_tests/blas/level1/axpy_usm.cpp @@ -41,19 +41,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy, fp alpha) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during AXPY:\n" << e.what() << std::endl; print_error_code(e); @@ -79,8 +79,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::axpy(&N_ref, (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y_ref.data(), - &incy_ref); + ::axpy(&N_ref, (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ AXPY. @@ -89,11 +88,11 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::axpy(main_queue, N, alpha, x.data(), incx, - y.data(), incy, dependencies); + y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::axpy(main_queue, N, alpha, x.data(), incx, - y.data(), incy, dependencies); + y.data(), incy, dependencies); break; default: break; } @@ -101,8 +100,8 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpy, N, alpha, - x.data(), incx, y.data(), incy, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::axpy, N, + alpha, x.data(), incx, y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::axpy, N, alpha, @@ -113,16 +112,16 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during AXPY:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of AXPY:\n" << error.what() << std::endl; } @@ -134,7 +133,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp } class AxpyUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(AxpyUsmTests, RealSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level1/copy.cpp b/tests/unit_tests/blas/level1/copy.cpp index 9c2a2d54e..c4ed07a59 100644 --- a/tests/unit_tests/blas/level1/copy.cpp +++ b/tests/unit_tests/blas/level1/copy.cpp @@ -86,7 +86,7 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::copy(main_queue, N, x_buffer, incx, y_buffer, - incy); + incy); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::copy(main_queue, N, x_buffer, incx, y_buffer, incy); @@ -100,8 +100,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { x_buffer, incx, y_buffer, incy); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::copy, N, x_buffer, - incx, y_buffer, incy); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::copy, N, + x_buffer, incx, y_buffer, incy); break; default: break; } diff --git a/tests/unit_tests/blas/level1/copy_usm.cpp b/tests/unit_tests/blas/level1/copy_usm.cpp index 1243b493f..aa2d59c2a 100644 --- a/tests/unit_tests/blas/level1/copy_usm.cpp +++ b/tests/unit_tests/blas/level1/copy_usm.cpp @@ -88,11 +88,11 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::copy(main_queue, N, x.data(), incx, - y.data(), incy, dependencies); + y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::copy(main_queue, N, x.data(), incx, y.data(), - incy, dependencies); + incy, dependencies); break; default: break; } @@ -104,8 +104,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { x.data(), incx, y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::copy, N, x.data(), - incx, y.data(), incy, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::copy, N, + x.data(), incx, y.data(), incy, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level1/dot.cpp b/tests/unit_tests/blas/level1/dot.cpp index b00e83a69..055a2b472 100644 --- a/tests/unit_tests/blas/level1/dot.cpp +++ b/tests/unit_tests/blas/level1/dot.cpp @@ -86,11 +86,11 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::dot(main_queue, N, x_buffer, incx, y_buffer, incy, - result_buffer); + result_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::dot(main_queue, N, x_buffer, incx, y_buffer, incy, - result_buffer); + result_buffer); break; default: break; } @@ -129,7 +129,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { return (int)good; } -class DotTests : public ::testing::TestWithParam> {}; +class DotTests : public ::testing::TestWithParam> { +}; TEST_P(DotTests, RealSinglePrecision) { EXPECT_TRUEORSKIP( diff --git a/tests/unit_tests/blas/level1/dot_usm.cpp b/tests/unit_tests/blas/level1/dot_usm.cpp index 8d17272c1..de775c4bf 100644 --- a/tests/unit_tests/blas/level1/dot_usm.cpp +++ b/tests/unit_tests/blas/level1/dot_usm.cpp @@ -96,12 +96,12 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - done = oneapi::math::blas::column_major::dot(main_queue, N, x.data(), incx, y.data(), - incy, result_p, dependencies); + done = oneapi::math::blas::column_major::dot( + main_queue, N, x.data(), incx, y.data(), incy, result_p, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::dot(main_queue, N, x.data(), incx, y.data(), - incy, result_p, dependencies); + incy, result_p, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level1/dotc.cpp b/tests/unit_tests/blas/level1/dotc.cpp index 96ae89f73..d6f6cafc2 100644 --- a/tests/unit_tests/blas/level1/dotc.cpp +++ b/tests/unit_tests/blas/level1/dotc.cpp @@ -41,12 +41,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { // Prepare data. vector x, y; fp result = 0.0, result_reference = 0.0; @@ -58,18 +58,18 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::dotc((fp_ref *)&result_reference, &N_ref, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), + ::dotc((fp_ref*)&result_reference, &N_ref, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref); // Call DPC++ DOTC. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during DOTC:\n" << e.what() << std::endl; print_error_code(e); @@ -87,12 +87,12 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::dotc(main_queue, N, x_buffer, incx, y_buffer, incy, - result_buffer); + oneapi::math::blas::column_major::dotc(main_queue, N, x_buffer, incx, y_buffer, + incy, result_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::dotc(main_queue, N, x_buffer, incx, y_buffer, incy, - result_buffer); + result_buffer); break; default: break; } @@ -103,23 +103,23 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { x_buffer, incx, y_buffer, incy, result_buffer); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotc, N, x_buffer, - incx, y_buffer, incy, result_buffer); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotc, N, + x_buffer, incx, y_buffer, incy, result_buffer); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during DOTC:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of DOTC:\n" << error.what() << std::endl; } @@ -131,7 +131,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { return (int)good; } -class DotcTests : public ::testing::TestWithParam> { +class DotcTests : public ::testing::TestWithParam> { }; TEST_P(DotcTests, ComplexSinglePrecision) { diff --git a/tests/unit_tests/blas/level1/dotc_usm.cpp b/tests/unit_tests/blas/level1/dotc_usm.cpp index 3148bd3f2..14d65968f 100644 --- a/tests/unit_tests/blas/level1/dotc_usm.cpp +++ b/tests/unit_tests/blas/level1/dotc_usm.cpp @@ -41,19 +41,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during DOTC:\n" << e.what() << std::endl; print_error_code(e); @@ -78,12 +78,12 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::dotc((fp_ref *)&result_reference, &N_ref, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), + ::dotc((fp_ref*)&result_reference, &N_ref, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref); // Call DPC++ DOTC. - auto result_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + auto result_p = (fp*)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); try { #ifdef CALL_RT_API @@ -94,7 +94,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::dotc(main_queue, N, x.data(), incx, y.data(), - incy, result_p, dependencies); + incy, result_p, dependencies); break; default: break; } @@ -106,24 +106,24 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { x.data(), incx, y.data(), incy, result_p, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotc, N, x.data(), - incx, y.data(), incy, result_p, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotc, N, + x.data(), incx, y.data(), incy, result_p, dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during DOTC:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of DOTC:\n" << error.what() << std::endl; } @@ -137,7 +137,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { } class DotcUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(DotcUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP( diff --git a/tests/unit_tests/blas/level1/dotu.cpp b/tests/unit_tests/blas/level1/dotu.cpp index ded168de0..fd70e4fa1 100644 --- a/tests/unit_tests/blas/level1/dotu.cpp +++ b/tests/unit_tests/blas/level1/dotu.cpp @@ -41,12 +41,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { // Prepare data. vector x, y; fp result = 0.0, result_reference = 0.0; @@ -58,18 +58,18 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::dotu((fp_ref *)&result_reference, &N_ref, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), + ::dotu((fp_ref*)&result_reference, &N_ref, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref); // Call DPC++ DOTU. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during DOTU:\n" << e.what() << std::endl; print_error_code(e); @@ -87,12 +87,12 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::dotu(main_queue, N, x_buffer, incx, y_buffer, incy, - result_buffer); + oneapi::math::blas::column_major::dotu(main_queue, N, x_buffer, incx, y_buffer, + incy, result_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::dotu(main_queue, N, x_buffer, incx, y_buffer, incy, - result_buffer); + result_buffer); break; default: break; } @@ -103,23 +103,23 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { x_buffer, incx, y_buffer, incy, result_buffer); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotu, N, x_buffer, - incx, y_buffer, incy, result_buffer); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotu, N, + x_buffer, incx, y_buffer, incy, result_buffer); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during DOTU:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of DOTU:\n" << error.what() << std::endl; } @@ -131,7 +131,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { return (int)good; } -class DotuTests : public ::testing::TestWithParam> { +class DotuTests : public ::testing::TestWithParam> { }; TEST_P(DotuTests, ComplexSinglePrecision) { diff --git a/tests/unit_tests/blas/level1/dotu_usm.cpp b/tests/unit_tests/blas/level1/dotu_usm.cpp index 889e01705..e1a34c5d2 100644 --- a/tests/unit_tests/blas/level1/dotu_usm.cpp +++ b/tests/unit_tests/blas/level1/dotu_usm.cpp @@ -41,19 +41,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during DOTU:\n" << e.what() << std::endl; print_error_code(e); @@ -78,12 +78,12 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::dotu((fp_ref *)&result_reference, &N_ref, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), + ::dotu((fp_ref*)&result_reference, &N_ref, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref); // Call DPC++ DOTU. - auto result_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + auto result_p = (fp*)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); try { #ifdef CALL_RT_API @@ -94,7 +94,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::dotu(main_queue, N, x.data(), incx, y.data(), - incy, result_p, dependencies); + incy, result_p, dependencies); break; default: break; } @@ -106,24 +106,24 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { x.data(), incx, y.data(), incy, result_p, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotu, N, x.data(), - incx, y.data(), incy, result_p, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::dotu, N, + x.data(), incx, y.data(), incy, result_p, dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during DOTU:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of DOTU:\n" << error.what() << std::endl; } @@ -136,7 +136,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy) { } class DotuUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(DotuUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP( diff --git a/tests/unit_tests/blas/level1/iamax.cpp b/tests/unit_tests/blas/level1/iamax.cpp index 19672e341..e56f58c42 100644 --- a/tests/unit_tests/blas/level1/iamax.cpp +++ b/tests/unit_tests/blas/level1/iamax.cpp @@ -84,7 +84,7 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx) { switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::iamax(main_queue, N, x_buffer, incx, - result_buffer); + result_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::iamax(main_queue, N, x_buffer, incx, result_buffer); @@ -126,8 +126,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx) { return (int)good; } -class IamaxTests : public ::testing::TestWithParam> { -}; +class IamaxTests + : public ::testing::TestWithParam> {}; TEST_P(IamaxTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 1357, 2)); diff --git a/tests/unit_tests/blas/level1/iamax_usm.cpp b/tests/unit_tests/blas/level1/iamax_usm.cpp index 88029f248..a55127fa2 100644 --- a/tests/unit_tests/blas/level1/iamax_usm.cpp +++ b/tests/unit_tests/blas/level1/iamax_usm.cpp @@ -96,11 +96,11 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::iamax(main_queue, N, x.data(), incx, - result_p, dependencies); + result_p, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::iamax(main_queue, N, x.data(), incx, result_p, - dependencies); + dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level1/iamin.cpp b/tests/unit_tests/blas/level1/iamin.cpp index 197d7f3d1..d0673a2cd 100644 --- a/tests/unit_tests/blas/level1/iamin.cpp +++ b/tests/unit_tests/blas/level1/iamin.cpp @@ -84,7 +84,7 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx) { switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::iamin(main_queue, N, x_buffer, incx, - result_buffer); + result_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::iamin(main_queue, N, x_buffer, incx, result_buffer); @@ -126,8 +126,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx) { return (int)good; } -class IaminTests : public ::testing::TestWithParam> { -}; +class IaminTests + : public ::testing::TestWithParam> {}; TEST_P(IaminTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), 1357, 2)); diff --git a/tests/unit_tests/blas/level1/iamin_usm.cpp b/tests/unit_tests/blas/level1/iamin_usm.cpp index 9baea05ef..beb535595 100644 --- a/tests/unit_tests/blas/level1/iamin_usm.cpp +++ b/tests/unit_tests/blas/level1/iamin_usm.cpp @@ -96,11 +96,11 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::iamin(main_queue, N, x.data(), incx, - result_p, dependencies); + result_p, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::iamin(main_queue, N, x.data(), incx, result_p, - dependencies); + dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level1/nrm2.cpp b/tests/unit_tests/blas/level1/nrm2.cpp index edc0edd93..30ddd63fd 100644 --- a/tests/unit_tests/blas/level1/nrm2.cpp +++ b/tests/unit_tests/blas/level1/nrm2.cpp @@ -84,7 +84,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx) { #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::nrm2(main_queue, N, x_buffer, incx, result_buffer); + oneapi::math::blas::column_major::nrm2(main_queue, N, x_buffer, incx, + result_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::nrm2(main_queue, N, x_buffer, incx, result_buffer); @@ -98,8 +99,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx) { x_buffer, incx, result_buffer); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::nrm2, N, x_buffer, - incx, result_buffer); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::nrm2, N, + x_buffer, incx, result_buffer); break; default: break; } diff --git a/tests/unit_tests/blas/level1/nrm2_usm.cpp b/tests/unit_tests/blas/level1/nrm2_usm.cpp index b22701401..87470eb92 100644 --- a/tests/unit_tests/blas/level1/nrm2_usm.cpp +++ b/tests/unit_tests/blas/level1/nrm2_usm.cpp @@ -97,11 +97,11 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::nrm2(main_queue, N, x.data(), incx, - result_p, dependencies); + result_p, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::nrm2(main_queue, N, x.data(), incx, result_p, - dependencies); + dependencies); break; default: break; } @@ -113,8 +113,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx) { x.data(), incx, result_p, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::nrm2, N, x.data(), - incx, result_p, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::nrm2, N, + x.data(), incx, result_p, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level1/rot.cpp b/tests/unit_tests/blas/level1/rot.cpp index 30506b4bf..6508283b3 100644 --- a/tests/unit_tests/blas/level1/rot.cpp +++ b/tests/unit_tests/blas/level1/rot.cpp @@ -41,12 +41,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp_scalar c, +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy, fp_scalar c, fp_scalar s) { // Prepare data. vector x, x_ref, y, y_ref; @@ -59,18 +59,18 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::rot(&N_ref, (fp_ref *)x_ref.data(), &incx_ref, (fp_ref *)y_ref.data(), &incy_ref, - (fp_scalar *)&c, (fp_scalar *)&s); + ::rot(&N_ref, (fp_ref*)x_ref.data(), &incx_ref, (fp_ref*)y_ref.data(), &incy_ref, + (fp_scalar*)&c, (fp_scalar*)&s); // Call DPC++ ROT. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during ROT:\n" << e.what() << std::endl; print_error_code(e); @@ -88,11 +88,11 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::rot(main_queue, N, x_buffer, incx, y_buffer, incy, - c, s); + c, s); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::rot(main_queue, N, x_buffer, incx, y_buffer, incy, c, - s); + s); break; default: break; } @@ -110,16 +110,16 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during ROT:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of ROT:\n" << error.what() << std::endl; } @@ -135,7 +135,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp return (int)good; } -class RotTests : public ::testing::TestWithParam> { +class RotTests : public ::testing::TestWithParam> { }; TEST_P(RotTests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level1/rot_usm.cpp b/tests/unit_tests/blas/level1/rot_usm.cpp index 880228cc7..52bc8c2f5 100644 --- a/tests/unit_tests/blas/level1/rot_usm.cpp +++ b/tests/unit_tests/blas/level1/rot_usm.cpp @@ -41,20 +41,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp_scalar c, +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy, fp_scalar c, fp_scalar s) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during ROT:\n" << e.what() << std::endl; print_error_code(e); @@ -80,8 +80,8 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::rot(&N_ref, (fp_ref *)x_ref.data(), &incx_ref, (fp_ref *)y_ref.data(), &incy_ref, - (fp_scalar *)&c, (fp_scalar *)&s); + ::rot(&N_ref, (fp_ref*)x_ref.data(), &incx_ref, (fp_ref*)y_ref.data(), &incy_ref, + (fp_scalar*)&c, (fp_scalar*)&s); // Call DPC++ ROT. @@ -89,12 +89,12 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - done = oneapi::math::blas::column_major::rot(main_queue, N, x.data(), incx, y.data(), - incy, c, s, dependencies); + done = oneapi::math::blas::column_major::rot(main_queue, N, x.data(), incx, + y.data(), incy, c, s, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::rot(main_queue, N, x.data(), incx, y.data(), - incy, c, s, dependencies); + incy, c, s, dependencies); break; default: break; } @@ -114,16 +114,16 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during ROT:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of ROT:\n" << error.what() << std::endl; } @@ -137,7 +137,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp } class RotUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(RotUsmTests, RealSinglePrecision) { float c(2.0); diff --git a/tests/unit_tests/blas/level1/rotg.cpp b/tests/unit_tests/blas/level1/rotg.cpp index 23b2b9969..4ccb8eb30 100644 --- a/tests/unit_tests/blas/level1/rotg.cpp +++ b/tests/unit_tests/blas/level1/rotg.cpp @@ -41,12 +41,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Prepare data. fp a, b, s, a_ref, b_ref, s_ref; fp_scalar c, c_ref; @@ -64,17 +64,17 @@ int test(device *dev, oneapi::math::layout layout) { // Call Reference ROTG. using fp_ref = typename ref_type_info::type; - ::rotg((fp_ref *)&a_ref, (fp_ref *)&b_ref, (fp_scalar *)&c_ref, (fp_ref *)&s_ref); + ::rotg((fp_ref*)&a_ref, (fp_ref*)&b_ref, (fp_scalar*)&c_ref, (fp_ref*)&s_ref); // Call DPC++ ROTG. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during ROTG:\n" << e.what() << std::endl; print_error_code(e); @@ -94,19 +94,19 @@ int test(device *dev, oneapi::math::layout layout) { switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::rotg(main_queue, a_buffer, b_buffer, c_buffer, - s_buffer); + s_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::rotg(main_queue, a_buffer, b_buffer, c_buffer, - s_buffer); + s_buffer); break; default: break; } #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rotg, a_buffer, - b_buffer, c_buffer, s_buffer); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rotg, + a_buffer, b_buffer, c_buffer, s_buffer); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotg, a_buffer, @@ -116,16 +116,16 @@ int test(device *dev, oneapi::math::layout layout) { } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during ROTG:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of ROTG:\n" << error.what() << std::endl; } @@ -144,7 +144,7 @@ int test(device *dev, oneapi::math::layout layout) { return (int)good; } -class RotgTests : public ::testing::TestWithParam> { +class RotgTests : public ::testing::TestWithParam> { }; TEST_P(RotgTests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level1/rotg_usm.cpp b/tests/unit_tests/blas/level1/rotg_usm.cpp index a90023a8b..f198bb393 100644 --- a/tests/unit_tests/blas/level1/rotg_usm.cpp +++ b/tests/unit_tests/blas/level1/rotg_usm.cpp @@ -41,19 +41,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during ROTG:\n" << e.what() << std::endl; print_error_code(e); @@ -83,22 +83,22 @@ int test(device *dev, oneapi::math::layout layout) { // Call Reference ROTG. using fp_ref = typename ref_type_info::type; - ::rotg((fp_ref *)&a_ref, (fp_ref *)&b_ref, (fp_scalar *)&c_ref, (fp_ref *)&s_ref); + ::rotg((fp_ref*)&a_ref, (fp_ref*)&b_ref, (fp_scalar*)&c_ref, (fp_ref*)&s_ref); // Call DPC++ ROTG. fp *a_p, *b_p, *s_p; - fp_scalar *c_p; + fp_scalar* c_p; if constexpr (alloc_type == usm::alloc::shared) { - a_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); - b_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); - s_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); - c_p = (fp_scalar *)oneapi::math::malloc_shared(64, sizeof(fp_scalar), *dev, cxt); + a_p = (fp*)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + b_p = (fp*)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + s_p = (fp*)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + c_p = (fp_scalar*)oneapi::math::malloc_shared(64, sizeof(fp_scalar), *dev, cxt); } else if constexpr (alloc_type == usm::alloc::device) { - a_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); - b_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); - s_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); - c_p = (fp_scalar *)oneapi::math::malloc_device(64, sizeof(fp_scalar), *dev, cxt); + a_p = (fp*)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); + b_p = (fp*)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); + s_p = (fp*)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); + c_p = (fp_scalar*)oneapi::math::malloc_device(64, sizeof(fp_scalar), *dev, cxt); } else { throw std::runtime_error("Bad alloc_type"); @@ -115,11 +115,11 @@ int test(device *dev, oneapi::math::layout layout) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::rotg(main_queue, a_p, b_p, c_p, s_p, - dependencies); + dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::rotg(main_queue, a_p, b_p, c_p, s_p, - dependencies); + dependencies); break; default: break; } @@ -127,8 +127,8 @@ int test(device *dev, oneapi::math::layout layout) { #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rotg, a_p, b_p, - c_p, s_p, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::rotg, a_p, + b_p, c_p, s_p, dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotg, a_p, b_p, @@ -139,16 +139,16 @@ int test(device *dev, oneapi::math::layout layout) { main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during ROTG:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of ROTG:\n" << error.what() << std::endl; } @@ -170,7 +170,7 @@ int test(device *dev, oneapi::math::layout layout) { } class RotgUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(RotgUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP((test(std::get<0>(GetParam()), std::get<1>(GetParam())))); diff --git a/tests/unit_tests/blas/level1/rotm.cpp b/tests/unit_tests/blas/level1/rotm.cpp index 81f9f137e..c9a38ff40 100644 --- a/tests/unit_tests/blas/level1/rotm.cpp +++ b/tests/unit_tests/blas/level1/rotm.cpp @@ -41,12 +41,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp flag) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy, fp flag) { // Prepare data. vector x, x_ref, y, y_ref; vector param; @@ -61,18 +61,18 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::rotm(&N_ref, (fp_ref *)x_ref.data(), &incx_ref, (fp_ref *)y_ref.data(), &incy_ref, - (fp_ref *)param.data()); + ::rotm(&N_ref, (fp_ref*)x_ref.data(), &incx_ref, (fp_ref*)y_ref.data(), &incy_ref, + (fp_ref*)param.data()); // Call DPC++ ROTM. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during ROTM:\n" << e.what() << std::endl; print_error_code(e); @@ -90,12 +90,12 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::rotm(main_queue, N, x_buffer, incx, y_buffer, incy, - param_buffer); + oneapi::math::blas::column_major::rotm(main_queue, N, x_buffer, incx, y_buffer, + incy, param_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::rotm(main_queue, N, x_buffer, incx, y_buffer, incy, - param_buffer); + param_buffer); break; default: break; } @@ -106,23 +106,23 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp x_buffer, incx, y_buffer, incy, param_buffer); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotm, N, x_buffer, - incx, y_buffer, incy, param_buffer); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotm, N, + x_buffer, incx, y_buffer, incy, param_buffer); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during ROTM:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of ROTM:\n" << error.what() << std::endl; } @@ -136,7 +136,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp return (int)good; } -class RotmTests : public ::testing::TestWithParam> { +class RotmTests : public ::testing::TestWithParam> { }; TEST_P(RotmTests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level1/rotm_usm.cpp b/tests/unit_tests/blas/level1/rotm_usm.cpp index 1625215a9..0c33bb467 100644 --- a/tests/unit_tests/blas/level1/rotm_usm.cpp +++ b/tests/unit_tests/blas/level1/rotm_usm.cpp @@ -41,19 +41,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp flag) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy, fp flag) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during ROTM:\n" << e.what() << std::endl; print_error_code(e); @@ -81,8 +81,8 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp using fp_ref = typename ref_type_info::type; const int N_ref = N, incx_ref = incx, incy_ref = incy; - ::rotm(&N_ref, (fp_ref *)x_ref.data(), &incx_ref, (fp_ref *)y_ref.data(), &incy_ref, - (fp_ref *)param.data()); + ::rotm(&N_ref, (fp_ref*)x_ref.data(), &incx_ref, (fp_ref*)y_ref.data(), &incy_ref, + (fp_ref*)param.data()); // Call DPC++ ROTM. @@ -95,7 +95,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::rotm(main_queue, N, x.data(), incx, y.data(), - incy, param.data(), dependencies); + incy, param.data(), dependencies); break; default: break; } @@ -107,24 +107,24 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp x.data(), incx, y.data(), incy, param.data(), dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotm, N, x.data(), - incx, y.data(), incy, param.data(), dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotm, N, + x.data(), incx, y.data(), incy, param.data(), dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during ROTM:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of ROTM:\n" << error.what() << std::endl; } @@ -138,7 +138,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fp } class RotmUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(RotmUsmTests, RealSinglePrecision) { float flag(-1.0); diff --git a/tests/unit_tests/blas/level1/rotmg.cpp b/tests/unit_tests/blas/level1/rotmg.cpp index efc75a4b5..2d18311b7 100644 --- a/tests/unit_tests/blas/level1/rotmg.cpp +++ b/tests/unit_tests/blas/level1/rotmg.cpp @@ -91,11 +91,11 @@ int test(device* dev, oneapi::math::layout layout) { switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::rotmg(main_queue, d1_buffer, d2_buffer, x1_buffer, - y1, param_buffer); + y1, param_buffer); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::rotmg(main_queue, d1_buffer, d2_buffer, x1_buffer, y1, - param_buffer); + oneapi::math::blas::row_major::rotmg(main_queue, d1_buffer, d2_buffer, x1_buffer, + y1, param_buffer); break; default: break; } @@ -187,8 +187,8 @@ int test(device* dev, oneapi::math::layout layout) { return (int)good; } -class RotmgTests : public ::testing::TestWithParam> { -}; +class RotmgTests + : public ::testing::TestWithParam> {}; TEST_P(RotmgTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/level1/rotmg_usm.cpp b/tests/unit_tests/blas/level1/rotmg_usm.cpp index e992ad560..dcb79b2dd 100644 --- a/tests/unit_tests/blas/level1/rotmg_usm.cpp +++ b/tests/unit_tests/blas/level1/rotmg_usm.cpp @@ -41,19 +41,19 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout) { +int test(device* dev, oneapi::math::layout layout) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during ROTMG:\n" << e.what() << std::endl; print_error_code(e); @@ -82,14 +82,14 @@ int test(device *dev, oneapi::math::layout layout) { fp *d1_p, *d2_p, *x1_p; if constexpr (alloc_type == usm::alloc::device) { - d1_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); - d2_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); - x1_p = (fp *)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); + d1_p = (fp*)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); + d2_p = (fp*)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); + x1_p = (fp*)oneapi::math::malloc_device(64, sizeof(fp), *dev, cxt); } else if constexpr (alloc_type == usm::alloc::shared) { - d1_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); - d2_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); - x1_p = (fp *)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + d1_p = (fp*)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + d2_p = (fp*)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); + x1_p = (fp*)oneapi::math::malloc_shared(64, sizeof(fp), *dev, cxt); } else { throw std::runtime_error("Bad alloc_type"); @@ -101,7 +101,7 @@ int test(device *dev, oneapi::math::layout layout) { // Call Reference ROTMG. - ::rotmg(&d1_ref, &d2_ref, &x1_ref, &y1, (fp *)param_ref.data()); + ::rotmg(&d1_ref, &d2_ref, &x1_ref, &y1, (fp*)param_ref.data()); // Call DPC++ ROTMG. @@ -110,11 +110,11 @@ int test(device *dev, oneapi::math::layout layout) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::rotmg(main_queue, d1_p, d2_p, x1_p, y1, - param.data(), dependencies); + param.data(), dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::rotmg(main_queue, d1_p, d2_p, x1_p, y1, - param.data(), dependencies); + param.data(), dependencies); break; default: break; } @@ -126,24 +126,24 @@ int test(device *dev, oneapi::math::layout layout) { d2_p, x1_p, y1, param.data(), dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotmg, d1_p, d2_p, - x1_p, y1, param.data(), dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::rotmg, d1_p, + d2_p, x1_p, y1, param.data(), dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during ROTMG:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of ROTMG:\n" << error.what() << std::endl; } @@ -213,7 +213,7 @@ int test(device *dev, oneapi::math::layout layout) { } class RotmgUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(RotmgUsmTests, RealSinglePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()))); diff --git a/tests/unit_tests/blas/level1/scal.cpp b/tests/unit_tests/blas/level1/scal.cpp index c345a6f55..27b10d2a4 100644 --- a/tests/unit_tests/blas/level1/scal.cpp +++ b/tests/unit_tests/blas/level1/scal.cpp @@ -95,8 +95,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, fp_scalar al #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::scal, N, alpha, - x_buffer, incx); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::scal, N, + alpha, x_buffer, incx); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::scal, N, alpha, diff --git a/tests/unit_tests/blas/level1/scal_usm.cpp b/tests/unit_tests/blas/level1/scal_usm.cpp index 6341533be..0da8f8ac4 100644 --- a/tests/unit_tests/blas/level1/scal_usm.cpp +++ b/tests/unit_tests/blas/level1/scal_usm.cpp @@ -89,11 +89,11 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, fp_scalar al switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::scal(main_queue, N, alpha, x.data(), incx, - dependencies); + dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::scal(main_queue, N, alpha, x.data(), incx, - dependencies); + dependencies); break; default: break; } @@ -101,8 +101,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, fp_scalar al #else switch (layout) { case oneapi::math::layout::col_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::scal, N, alpha, - x.data(), incx, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::column_major::scal, N, + alpha, x.data(), incx, dependencies); break; case oneapi::math::layout::row_major: TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::scal, N, alpha, diff --git a/tests/unit_tests/blas/level1/sdsdot.cpp b/tests/unit_tests/blas/level1/sdsdot.cpp index 67505b5f3..d8ae45449 100644 --- a/tests/unit_tests/blas/level1/sdsdot.cpp +++ b/tests/unit_tests/blas/level1/sdsdot.cpp @@ -41,11 +41,11 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, float alpha) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy, float alpha) { // Prepare data. vector x, y; float result = float(-1), result_ref = float(-1); @@ -56,18 +56,18 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fl // Call Reference SDSDOT. const int N_ref = N, incx_ref = incx, incy_ref = incy; - result_ref = ::sdsdot(&N_ref, (float *)&alpha, (float *)x.data(), &incx_ref, (float *)y.data(), - &incy_ref); + result_ref = + ::sdsdot(&N_ref, (float*)&alpha, (float*)x.data(), &incx_ref, (float*)y.data(), &incy_ref); // Call DPC++ SDSDOT. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SDSDOT:\n" << e.what() << std::endl; print_error_code(e); @@ -86,11 +86,11 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fl switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::sdsdot(main_queue, N, alpha, x_buffer, incx, - y_buffer, incy, result_buffer); + y_buffer, incy, result_buffer); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::sdsdot(main_queue, N, alpha, x_buffer, incx, y_buffer, - incy, result_buffer); + oneapi::math::blas::row_major::sdsdot(main_queue, N, alpha, x_buffer, incx, + y_buffer, incy, result_buffer); break; default: break; } @@ -108,16 +108,16 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fl } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SDSDOT:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SDSDOT:\n" << error.what() << std::endl; } @@ -130,7 +130,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fl } class SdsdotTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SdsdotTests, RealSinglePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); diff --git a/tests/unit_tests/blas/level1/sdsdot_usm.cpp b/tests/unit_tests/blas/level1/sdsdot_usm.cpp index 09d9e08a3..f7153d38c 100644 --- a/tests/unit_tests/blas/level1/sdsdot_usm.cpp +++ b/tests/unit_tests/blas/level1/sdsdot_usm.cpp @@ -41,18 +41,18 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { -int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, float alpha) { +int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy, float alpha) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SDSDOT:\n" << e.what() << std::endl; print_error_code(e); @@ -76,12 +76,12 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fl // Call Reference SDSDOT. const int N_ref = N, incx_ref = incx, incy_ref = incy; - result_ref = ::sdsdot(&N_ref, (float *)&alpha, (float *)x.data(), &incx_ref, (float *)y.data(), - &incy_ref); + result_ref = + ::sdsdot(&N_ref, (float*)&alpha, (float*)x.data(), &incx_ref, (float*)y.data(), &incy_ref); // Call DPC++ SDSDOT. - auto result_p = (float *)oneapi::math::malloc_shared(64, sizeof(float), *dev, cxt); + auto result_p = (float*)oneapi::math::malloc_shared(64, sizeof(float), *dev, cxt); try { #ifdef CALL_RT_API @@ -91,8 +91,8 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fl main_queue, N, alpha, x.data(), incx, y.data(), incy, result_p, dependencies); break; case oneapi::math::layout::row_major: - done = oneapi::math::blas::row_major::sdsdot(main_queue, N, alpha, x.data(), incx, - y.data(), incy, result_p, dependencies); + done = oneapi::math::blas::row_major::sdsdot( + main_queue, N, alpha, x.data(), incx, y.data(), incy, result_p, dependencies); break; default: break; } @@ -113,16 +113,16 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fl main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SDSDOT:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SDSDOT:\n" << error.what() << std::endl; } @@ -135,7 +135,7 @@ int test(device *dev, oneapi::math::layout layout, int N, int incx, int incy, fl } class SdsdotUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SdsdotUsmTests, RealSinglePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); diff --git a/tests/unit_tests/blas/level1/swap.cpp b/tests/unit_tests/blas/level1/swap.cpp index 696943c35..2fa39ff38 100644 --- a/tests/unit_tests/blas/level1/swap.cpp +++ b/tests/unit_tests/blas/level1/swap.cpp @@ -86,7 +86,7 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::swap(main_queue, N, x_buffer, incx, y_buffer, - incy); + incy); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::swap(main_queue, N, x_buffer, incx, y_buffer, incy); @@ -100,8 +100,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { x_buffer, incx, y_buffer, incy); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::swap, N, x_buffer, - incx, y_buffer, incy); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::swap, N, + x_buffer, incx, y_buffer, incy); break; default: break; } diff --git a/tests/unit_tests/blas/level1/swap_usm.cpp b/tests/unit_tests/blas/level1/swap_usm.cpp index 3fd742c36..5d8b1e059 100644 --- a/tests/unit_tests/blas/level1/swap_usm.cpp +++ b/tests/unit_tests/blas/level1/swap_usm.cpp @@ -88,11 +88,11 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::swap(main_queue, N, x.data(), incx, - y.data(), incy, dependencies); + y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::swap(main_queue, N, x.data(), incx, y.data(), - incy, dependencies); + incy, dependencies); break; default: break; } @@ -104,8 +104,8 @@ int test(device* dev, oneapi::math::layout layout, int N, int incx, int incy) { x.data(), incx, y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::swap, N, x.data(), - incx, y.data(), incy, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::swap, N, + x.data(), incx, y.data(), incy, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level2/gbmv.cpp b/tests/unit_tests/blas/level2/gbmv.cpp index f1e8fc3ec..096eaca44 100644 --- a/tests/unit_tests/blas/level2/gbmv.cpp +++ b/tests/unit_tests/blas/level2/gbmv.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, +int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, int kl, int ku, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. int x_len = outer_dimension(transa, m, n); @@ -66,18 +66,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans using fp_ref = typename ref_type_info::type; ::gbmv(convert_to_cblas_layout(layout), convert_to_cblas_trans(transa), &m_ref, &n_ref, &kl_ref, - &ku_ref, (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + &ku_ref, (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ GBMV. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GBMV:\n" << e.what() << std::endl; print_error_code(e); @@ -96,13 +96,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::gbmv(main_queue, transa, m, n, kl, ku, alpha, - A_buffer, lda, x_buffer, incx, beta, y_buffer, - incy); + A_buffer, lda, x_buffer, incx, beta, + y_buffer, incy); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::gbmv(main_queue, transa, m, n, kl, ku, alpha, - A_buffer, lda, x_buffer, incx, beta, y_buffer, - incy); + A_buffer, lda, x_buffer, incx, beta, y_buffer, + incy); break; default: break; } @@ -122,16 +122,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GBMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GBMV:\n" << error.what() << std::endl; } @@ -142,21 +142,21 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans return (int)good; } -class GbmvTests : public ::testing::TestWithParam> { +class GbmvTests : public ::testing::TestWithParam> { }; TEST_P(GbmvTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, - 42)); + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, + 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, - 42)); + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, + 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); @@ -185,8 +185,8 @@ TEST_P(GbmvTests, RealDoublePrecision) { oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, - 42)); + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, + -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); @@ -237,14 +237,14 @@ TEST_P(GbmvTests, ComplexDoublePrecision) { oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, - beta, 2, 3, 42)); + oneapi::math::transpose::trans, 25, 30, 5, 7, + alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, - beta, -2, -3, 42)); + oneapi::math::transpose::trans, 25, 30, 5, 7, + alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, - beta, 1, 1, 42)); + oneapi::math::transpose::trans, 25, 30, 5, 7, + alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); diff --git a/tests/unit_tests/blas/level2/gbmv_usm.cpp b/tests/unit_tests/blas/level2/gbmv_usm.cpp index c017d0fae..ec121e723 100644 --- a/tests/unit_tests/blas/level2/gbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/gbmv_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, +int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, int kl, int ku, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GBMV:\n" << e.what() << std::endl; print_error_code(e); @@ -86,8 +86,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans using fp_ref = typename ref_type_info::type; ::gbmv(convert_to_cblas_layout(layout), convert_to_cblas_trans(transa), &m_ref, &n_ref, &kl_ref, - &ku_ref, (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + &ku_ref, (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ GBMV. @@ -96,13 +96,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::gbmv(main_queue, transa, m, n, kl, ku, - alpha, A.data(), lda, x.data(), incx, - beta, y.data(), incy, dependencies); + alpha, A.data(), lda, x.data(), incx, + beta, y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::gbmv(main_queue, transa, m, n, kl, ku, alpha, - A.data(), lda, x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; default: break; } @@ -124,16 +124,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GBMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GBMV:\n" << error.what() << std::endl; } @@ -145,20 +145,20 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans } class GbmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GbmvUsmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, 3, - 42)); + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 2, + 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, - 42)); + oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, + 1, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); @@ -187,8 +187,8 @@ TEST_P(GbmvUsmTests, RealDoublePrecision) { oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, -3, - 42)); + oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, -2, + -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); @@ -239,14 +239,14 @@ TEST_P(GbmvUsmTests, ComplexDoublePrecision) { oneapi::math::transpose::nontrans, 25, 30, 5, 7, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, - beta, 2, 3, 42)); + oneapi::math::transpose::trans, 25, 30, 5, 7, + alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, - beta, -2, -3, 42)); + oneapi::math::transpose::trans, 25, 30, 5, 7, + alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, 5, 7, alpha, - beta, 1, 1, 42)); + oneapi::math::transpose::trans, 25, 30, 5, 7, + alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, 25, 30, 5, 7, alpha, beta, 2, 3, 42)); diff --git a/tests/unit_tests/blas/level2/gemv.cpp b/tests/unit_tests/blas/level2/gemv.cpp index d75720278..7d0999b1d 100644 --- a/tests/unit_tests/blas/level2/gemv.cpp +++ b/tests/unit_tests/blas/level2/gemv.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, +int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. int x_len = outer_dimension(transa, m, n); @@ -65,18 +65,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans using fp_ref = typename ref_type_info::type; ::gemv(convert_to_cblas_layout(layout), convert_to_cblas_trans(transa), &m_ref, &n_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ GEMV. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GEMV:\n" << e.what() << std::endl; print_error_code(e); @@ -95,11 +95,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::gemv(main_queue, transa, m, n, alpha, A_buffer, - lda, x_buffer, incx, beta, y_buffer, incy); + lda, x_buffer, incx, beta, y_buffer, incy); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::gemv(main_queue, transa, m, n, alpha, A_buffer, lda, - x_buffer, incx, beta, y_buffer, incy); + x_buffer, incx, beta, y_buffer, incy); break; default: break; } @@ -119,16 +119,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GEMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GEMV:\n" << error.what() << std::endl; } @@ -139,19 +139,21 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans return (int)good; } -class GemvTests : public ::testing::TestWithParam> { +class GemvTests : public ::testing::TestWithParam> { }; TEST_P(GemvTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 2, 3, 42)); + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 2, 3, + 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, + 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), @@ -176,7 +178,8 @@ TEST_P(GemvTests, RealDoublePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, -2, -3, + 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); } @@ -185,23 +188,23 @@ TEST_P(GemvTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::nontrans, 2, 3, alpha, beta, - 2, 3, 42)); + oneapi::math::transpose::nontrans, 2, 3, alpha, + beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::nontrans, 2, 3, alpha, beta, - -2, -3, 42)); + oneapi::math::transpose::nontrans, 2, 3, alpha, + beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::nontrans, 2, 3, alpha, beta, - 1, 1, 42)); + oneapi::math::transpose::nontrans, 2, 3, alpha, + beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 2, 3, alpha, beta, 2, - 3, 42)); + oneapi::math::transpose::trans, 2, 3, alpha, beta, + 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, 2, 3, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 2, 3, alpha, beta, 1, - 1, 42)); + oneapi::math::transpose::trans, 2, 3, alpha, beta, + 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, 2, 3, alpha, beta, 2, 3, 42)); @@ -228,14 +231,14 @@ TEST_P(GemvTests, ComplexDoublePrecision) { oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, alpha, beta, - 2, 3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, + beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, alpha, beta, - -2, -3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, + beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, alpha, beta, - 1, 1, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, + beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, 25, 30, alpha, beta, 2, 3, 42)); diff --git a/tests/unit_tests/blas/level2/gemv_usm.cpp b/tests/unit_tests/blas/level2/gemv_usm.cpp index 5729914d2..c5f9d4dec 100644 --- a/tests/unit_tests/blas/level2/gemv_usm.cpp +++ b/tests/unit_tests/blas/level2/gemv_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, +int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose transa, int m, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GEMV:\n" << e.what() << std::endl; print_error_code(e); @@ -85,8 +85,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans using fp_ref = typename ref_type_info::type; ::gemv(convert_to_cblas_layout(layout), convert_to_cblas_trans(transa), &m_ref, &n_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ GEMV. @@ -95,13 +95,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::gemv(main_queue, transa, m, n, alpha, - A.data(), lda, x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: - done = oneapi::math::blas::row_major::gemv(main_queue, transa, m, n, alpha, A.data(), - lda, x.data(), incx, beta, y.data(), incy, - dependencies); + done = oneapi::math::blas::row_major::gemv(main_queue, transa, m, n, alpha, + A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; default: break; } @@ -123,16 +123,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GEMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GEMV:\n" << error.what() << std::endl; } @@ -144,18 +144,20 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::transpose trans } class GemvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GemvUsmTests, RealSinglePrecision) { float alpha(2.0); float beta(3.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 2, 3, 42)); + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 2, 3, + 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::nontrans, 25, 30, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); + oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, + 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), @@ -180,7 +182,8 @@ TEST_P(GemvUsmTests, RealDoublePrecision) { EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, 25, 30, alpha, beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, alpha, beta, -2, -3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, beta, -2, -3, + 42)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::trans, 25, 30, alpha, beta, 1, 1, 42)); } @@ -230,14 +233,14 @@ TEST_P(GemvUsmTests, ComplexDoublePrecision) { oneapi::math::transpose::nontrans, 25, 30, alpha, beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, alpha, beta, - 2, 3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, + beta, 2, 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, alpha, beta, - -2, -3, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, + beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::transpose::trans, 25, 30, alpha, beta, - 1, 1, 42)); + oneapi::math::transpose::trans, 25, 30, alpha, + beta, 1, 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::transpose::conjtrans, 25, 30, alpha, beta, 2, 3, 42)); diff --git a/tests/unit_tests/blas/level2/ger.cpp b/tests/unit_tests/blas/level2/ger.cpp index 46b92ff17..9c207a7c4 100644 --- a/tests/unit_tests/blas/level2/ger.cpp +++ b/tests/unit_tests/blas/level2/ger.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device* dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Prepare data. @@ -62,18 +62,18 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i const int m_ref = m, n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; using fp_ref = typename ref_type_info::type; - ::ger(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref *)&alpha, (fp_ref *)x.data(), - &incx_ref, (fp_ref *)y.data(), &incy_ref, (fp_ref *)A_ref.data(), &lda_ref); + ::ger(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref*)&alpha, (fp_ref*)x.data(), + &incx_ref, (fp_ref*)y.data(), &incy_ref, (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ GER. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GER:\n" << e.what() << std::endl; print_error_code(e); @@ -92,11 +92,11 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::ger(main_queue, m, n, alpha, x_buffer, incx, - y_buffer, incy, A_buffer, lda); + y_buffer, incy, A_buffer, lda); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::ger(main_queue, m, n, alpha, x_buffer, incx, y_buffer, - incy, A_buffer, lda); + oneapi::math::blas::row_major::ger(main_queue, m, n, alpha, x_buffer, incx, + y_buffer, incy, A_buffer, lda); break; default: break; } @@ -114,16 +114,16 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GER:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GER:\n" << error.what() << std::endl; } @@ -135,7 +135,7 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i return (int)good; } -class GerTests : public ::testing::TestWithParam> { +class GerTests : public ::testing::TestWithParam> { }; TEST_P(GerTests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/ger_usm.cpp b/tests/unit_tests/blas/level2/ger_usm.cpp index 9d963e68a..c0a523785 100644 --- a/tests/unit_tests/blas/level2/ger_usm.cpp +++ b/tests/unit_tests/blas/level2/ger_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device* dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GER:\n" << e.what() << std::endl; print_error_code(e); @@ -82,8 +82,8 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i const int m_ref = m, n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; using fp_ref = typename ref_type_info::type; - ::ger(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref *)&alpha, (fp_ref *)x.data(), - &incx_ref, (fp_ref *)y.data(), &incy_ref, (fp_ref *)A_ref.data(), &lda_ref); + ::ger(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref*)&alpha, (fp_ref*)x.data(), + &incx_ref, (fp_ref*)y.data(), &incy_ref, (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ GER. @@ -91,14 +91,14 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - done = oneapi::math::blas::column_major::ger(main_queue, m, n, alpha, x.data(), incx, - y.data(), incy, A.data(), lda, - dependencies); + done = oneapi::math::blas::column_major::ger(main_queue, m, n, alpha, x.data(), + incx, y.data(), incy, A.data(), lda, + dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::ger(main_queue, m, n, alpha, x.data(), incx, - y.data(), incy, A.data(), lda, dependencies); + y.data(), incy, A.data(), lda, dependencies); break; default: break; } @@ -120,16 +120,16 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GER:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GER:\n" << error.what() << std::endl; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i } class GerUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GerUsmTests, RealSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level2/gerc.cpp b/tests/unit_tests/blas/level2/gerc.cpp index 316c496a4..8d89f8347 100644 --- a/tests/unit_tests/blas/level2/gerc.cpp +++ b/tests/unit_tests/blas/level2/gerc.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device* dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Prepare data. @@ -62,18 +62,18 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i const int m_ref = m, n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; using fp_ref = typename ref_type_info::type; - ::gerc(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref *)&alpha, (fp_ref *)x.data(), - &incx_ref, (fp_ref *)y.data(), &incy_ref, (fp_ref *)A_ref.data(), &lda_ref); + ::gerc(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref*)&alpha, (fp_ref*)x.data(), + &incx_ref, (fp_ref*)y.data(), &incy_ref, (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ GERC. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GERC:\n" << e.what() << std::endl; print_error_code(e); @@ -92,11 +92,11 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::gerc(main_queue, m, n, alpha, x_buffer, incx, - y_buffer, incy, A_buffer, lda); + y_buffer, incy, A_buffer, lda); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::gerc(main_queue, m, n, alpha, x_buffer, incx, - y_buffer, incy, A_buffer, lda); + y_buffer, incy, A_buffer, lda); break; default: break; } @@ -107,23 +107,23 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gerc, m, n, alpha, - x_buffer, incx, y_buffer, incy, A_buffer, lda); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gerc, m, n, + alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GERC:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GERC:\n" << error.what() << std::endl; } @@ -135,7 +135,7 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i return (int)good; } -class GercTests : public ::testing::TestWithParam> { +class GercTests : public ::testing::TestWithParam> { }; TEST_P(GercTests, ComplexSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/gerc_usm.cpp b/tests/unit_tests/blas/level2/gerc_usm.cpp index cbca89a4e..a2d9d28cb 100644 --- a/tests/unit_tests/blas/level2/gerc_usm.cpp +++ b/tests/unit_tests/blas/level2/gerc_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device* dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GERC:\n" << e.what() << std::endl; print_error_code(e); @@ -82,8 +82,8 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i const int m_ref = m, n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; using fp_ref = typename ref_type_info::type; - ::gerc(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref *)&alpha, (fp_ref *)x.data(), - &incx_ref, (fp_ref *)y.data(), &incy_ref, (fp_ref *)A_ref.data(), &lda_ref); + ::gerc(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref*)&alpha, (fp_ref*)x.data(), + &incx_ref, (fp_ref*)y.data(), &incy_ref, (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ GERC. @@ -92,13 +92,13 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::gerc(main_queue, m, n, alpha, x.data(), - incx, y.data(), incy, A.data(), lda, - dependencies); + incx, y.data(), incy, A.data(), lda, + dependencies); break; case oneapi::math::layout::row_major: - done = - oneapi::math::blas::row_major::gerc(main_queue, m, n, alpha, x.data(), incx, - y.data(), incy, A.data(), lda, dependencies); + done = oneapi::math::blas::row_major::gerc(main_queue, m, n, alpha, x.data(), incx, + y.data(), incy, A.data(), lda, + dependencies); break; default: break; } @@ -111,8 +111,8 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gerc, m, n, alpha, - x.data(), incx, y.data(), incy, A.data(), lda, + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::gerc, m, n, + alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; default: break; @@ -120,16 +120,16 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GERC:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GERC:\n" << error.what() << std::endl; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i } class GercUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GercUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); diff --git a/tests/unit_tests/blas/level2/geru.cpp b/tests/unit_tests/blas/level2/geru.cpp index 70435af74..7a44656b1 100644 --- a/tests/unit_tests/blas/level2/geru.cpp +++ b/tests/unit_tests/blas/level2/geru.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device* dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Prepare data. @@ -62,18 +62,18 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i const int m_ref = m, n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; using fp_ref = typename ref_type_info::type; - ::geru(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref *)&alpha, (fp_ref *)x.data(), - &incx_ref, (fp_ref *)y.data(), &incy_ref, (fp_ref *)A_ref.data(), &lda_ref); + ::geru(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref*)&alpha, (fp_ref*)x.data(), + &incx_ref, (fp_ref*)y.data(), &incy_ref, (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ GERU. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GERU:\n" << e.what() << std::endl; print_error_code(e); @@ -92,11 +92,11 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::geru(main_queue, m, n, alpha, x_buffer, incx, - y_buffer, incy, A_buffer, lda); + y_buffer, incy, A_buffer, lda); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::geru(main_queue, m, n, alpha, x_buffer, incx, - y_buffer, incy, A_buffer, lda); + y_buffer, incy, A_buffer, lda); break; default: break; } @@ -107,23 +107,23 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::geru, m, n, alpha, - x_buffer, incx, y_buffer, incy, A_buffer, lda); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::geru, m, n, + alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GERU:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GERU:\n" << error.what() << std::endl; } @@ -135,7 +135,7 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i return (int)good; } -class GeruTests : public ::testing::TestWithParam> { +class GeruTests : public ::testing::TestWithParam> { }; TEST_P(GeruTests, ComplexSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/geru_usm.cpp b/tests/unit_tests/blas/level2/geru_usm.cpp index b470ffbdc..6bf155795 100644 --- a/tests/unit_tests/blas/level2/geru_usm.cpp +++ b/tests/unit_tests/blas/level2/geru_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, +int test(device* dev, oneapi::math::layout layout, int m, int n, fp alpha, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during GERU:\n" << e.what() << std::endl; print_error_code(e); @@ -82,8 +82,8 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i const int m_ref = m, n_ref = n, incx_ref = incx, incy_ref = incy, lda_ref = lda; using fp_ref = typename ref_type_info::type; - ::geru(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref *)&alpha, (fp_ref *)x.data(), - &incx_ref, (fp_ref *)y.data(), &incy_ref, (fp_ref *)A_ref.data(), &lda_ref); + ::geru(convert_to_cblas_layout(layout), &m_ref, &n_ref, (fp_ref*)&alpha, (fp_ref*)x.data(), + &incx_ref, (fp_ref*)y.data(), &incy_ref, (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ GERU. @@ -92,13 +92,13 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::geru(main_queue, m, n, alpha, x.data(), - incx, y.data(), incy, A.data(), lda, - dependencies); + incx, y.data(), incy, A.data(), lda, + dependencies); break; case oneapi::math::layout::row_major: - done = - oneapi::math::blas::row_major::geru(main_queue, m, n, alpha, x.data(), incx, - y.data(), incy, A.data(), lda, dependencies); + done = oneapi::math::blas::row_major::geru(main_queue, m, n, alpha, x.data(), incx, + y.data(), incy, A.data(), lda, + dependencies); break; default: break; } @@ -111,8 +111,8 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::geru, m, n, alpha, - x.data(), incx, y.data(), incy, A.data(), lda, + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::geru, m, n, + alpha, x.data(), incx, y.data(), incy, A.data(), lda, dependencies); break; default: break; @@ -120,16 +120,16 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during GERU:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of GERU:\n" << error.what() << std::endl; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::math::layout layout, int m, int n, fp alpha, int i } class GeruUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(GeruUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); diff --git a/tests/unit_tests/blas/level2/hbmv.cpp b/tests/unit_tests/blas/level2/hbmv.cpp index 02c1e0f15..bd1fc71f2 100644 --- a/tests/unit_tests/blas/level2/hbmv.cpp +++ b/tests/unit_tests/blas/level2/hbmv.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. vector x, y, y_ref, A; @@ -63,18 +63,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::hbmv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, &k_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ HBMV. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HBMV:\n" << e.what() << std::endl; print_error_code(e); @@ -93,12 +93,12 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::hbmv(main_queue, upper_lower, n, k, alpha, - A_buffer, lda, x_buffer, incx, beta, y_buffer, - incy); + A_buffer, lda, x_buffer, incx, beta, + y_buffer, incy); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::hbmv(main_queue, upper_lower, n, k, alpha, A_buffer, - lda, x_buffer, incx, beta, y_buffer, incy); + lda, x_buffer, incx, beta, y_buffer, incy); break; default: break; } @@ -110,24 +110,24 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe beta, y_buffer, incy); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hbmv, upper_lower, - n, k, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, - incy); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hbmv, + upper_lower, n, k, alpha, A_buffer, lda, x_buffer, incx, + beta, y_buffer, incy); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HBMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HBMV:\n" << error.what() << std::endl; } @@ -138,7 +138,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class HbmvTests : public ::testing::TestWithParam> { +class HbmvTests : public ::testing::TestWithParam> { }; TEST_P(HbmvTests, ComplexSinglePrecision) { @@ -169,11 +169,11 @@ TEST_P(HbmvTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, 3, - 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, + 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, 3, - 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, + 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); @@ -181,11 +181,11 @@ TEST_P(HbmvTests, ComplexDoublePrecision) { oneapi::math::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, 1, - 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, + 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, 1, - 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, + 1, 42)); } INSTANTIATE_TEST_SUITE_P(HbmvTestSuite, HbmvTests, diff --git a/tests/unit_tests/blas/level2/hbmv_usm.cpp b/tests/unit_tests/blas/level2/hbmv_usm.cpp index 6adbba106..325013029 100644 --- a/tests/unit_tests/blas/level2/hbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/hbmv_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HBMV:\n" << e.what() << std::endl; print_error_code(e); @@ -84,8 +84,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::hbmv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, &k_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ HBMV. @@ -94,13 +94,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::hbmv(main_queue, upper_lower, n, k, alpha, - A.data(), lda, x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::hbmv(main_queue, upper_lower, n, k, alpha, - A.data(), lda, x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; default: break; } @@ -113,25 +113,25 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe beta, y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hbmv, upper_lower, - n, k, alpha, A.data(), lda, x.data(), incx, beta, y.data(), - incy, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hbmv, + upper_lower, n, k, alpha, A.data(), lda, x.data(), incx, + beta, y.data(), incy, dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HBMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HBMV:\n" << error.what() << std::endl; } @@ -143,7 +143,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class HbmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HbmvUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); @@ -173,11 +173,11 @@ TEST_P(HbmvUsmTests, ComplexDoublePrecision) { std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, 3, - 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 2, + 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, 3, - 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 2, + 3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, 5, alpha, beta, -2, -3, 42)); @@ -185,11 +185,11 @@ TEST_P(HbmvUsmTests, ComplexDoublePrecision) { oneapi::math::uplo::upper, 30, 5, alpha, beta, -2, -3, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, 1, - 42)); + oneapi::math::uplo::lower, 30, 5, alpha, beta, 1, + 1, 42)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, 1, - 42)); + oneapi::math::uplo::upper, 30, 5, alpha, beta, 1, + 1, 42)); } INSTANTIATE_TEST_SUITE_P(HbmvUsmTestSuite, HbmvUsmTests, diff --git a/tests/unit_tests/blas/level2/hemv.cpp b/tests/unit_tests/blas/level2/hemv.cpp index c45ebd3a4..26d71956c 100644 --- a/tests/unit_tests/blas/level2/hemv.cpp +++ b/tests/unit_tests/blas/level2/hemv.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. vector x, y, y_ref, A; @@ -62,18 +62,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::hemv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ HEMV. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HEMV:\n" << e.what() << std::endl; print_error_code(e); @@ -92,11 +92,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::hemv(main_queue, upper_lower, n, alpha, A_buffer, - lda, x_buffer, incx, beta, y_buffer, incy); + lda, x_buffer, incx, beta, y_buffer, incy); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::hemv(main_queue, upper_lower, n, alpha, A_buffer, lda, - x_buffer, incx, beta, y_buffer, incy); + oneapi::math::blas::row_major::hemv(main_queue, upper_lower, n, alpha, A_buffer, + lda, x_buffer, incx, beta, y_buffer, incy); break; default: break; } @@ -108,24 +108,24 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe y_buffer, incy); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hemv, upper_lower, - n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, - incy); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hemv, + upper_lower, n, alpha, A_buffer, lda, x_buffer, incx, beta, + y_buffer, incy); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HEMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HEMV:\n" << error.what() << std::endl; } @@ -136,7 +136,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class HemvTests : public ::testing::TestWithParam> { +class HemvTests : public ::testing::TestWithParam> { }; TEST_P(HemvTests, ComplexSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/hemv_usm.cpp b/tests/unit_tests/blas/level2/hemv_usm.cpp index 07f0cc386..3ca3fd7cf 100644 --- a/tests/unit_tests/blas/level2/hemv_usm.cpp +++ b/tests/unit_tests/blas/level2/hemv_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HEMV:\n" << e.what() << std::endl; print_error_code(e); @@ -83,8 +83,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::hemv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ HEMV. @@ -93,13 +93,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::hemv(main_queue, upper_lower, n, alpha, - A.data(), lda, x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::hemv(main_queue, upper_lower, n, alpha, - A.data(), lda, x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; default: break; } @@ -112,25 +112,25 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hemv, upper_lower, - n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), - incy, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hemv, + upper_lower, n, alpha, A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HEMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HEMV:\n" << error.what() << std::endl; } @@ -142,7 +142,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class HemvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HemvUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); diff --git a/tests/unit_tests/blas/level2/her.cpp b/tests/unit_tests/blas/level2/her.cpp index 9bc431053..eaeb82048 100644 --- a/tests/unit_tests/blas/level2/her.cpp +++ b/tests/unit_tests/blas/level2/her.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp_scalar alpha, int incx, int lda) { // Prepare data. vector x, A_ref, A; @@ -61,17 +61,17 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_scalar_ref = typename ref_type_info::type; ::her(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_scalar_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data(), &lda_ref); + (fp_scalar_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ HER. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HER:\n" << e.what() << std::endl; print_error_code(e); @@ -89,11 +89,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::her(main_queue, upper_lower, n, alpha, x_buffer, - incx, A_buffer, lda); + incx, A_buffer, lda); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::her(main_queue, upper_lower, n, alpha, x_buffer, incx, - A_buffer, lda); + oneapi::math::blas::row_major::her(main_queue, upper_lower, n, alpha, x_buffer, + incx, A_buffer, lda); break; default: break; } @@ -111,16 +111,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HER:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HER:\n" << error.what() << std::endl; } @@ -131,7 +131,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class HerTests : public ::testing::TestWithParam> { +class HerTests : public ::testing::TestWithParam> { }; TEST_P(HerTests, ComplexSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/her2.cpp b/tests/unit_tests/blas/level2/her2.cpp index 9ddb0ee3f..6ab26b315 100644 --- a/tests/unit_tests/blas/level2/her2.cpp +++ b/tests/unit_tests/blas/level2/her2.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy, int lda) { // Prepare data. vector x, y, A_ref, A; @@ -62,18 +62,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::her2(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), &incy_ref, - (fp_ref *)A_ref.data(), &lda_ref); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref, + (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ HER2. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HER2:\n" << e.what() << std::endl; print_error_code(e); @@ -92,11 +92,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::her2(main_queue, upper_lower, n, alpha, x_buffer, - incx, y_buffer, incy, A_buffer, lda); + incx, y_buffer, incy, A_buffer, lda); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::her2(main_queue, upper_lower, n, alpha, x_buffer, - incx, y_buffer, incy, A_buffer, lda); + incx, y_buffer, incy, A_buffer, lda); break; default: break; } @@ -108,23 +108,24 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe A_buffer, lda); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::her2, upper_lower, - n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::her2, + upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, + A_buffer, lda); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HER2:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HER2:\n" << error.what() << std::endl; } @@ -135,7 +136,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class Her2Tests : public ::testing::TestWithParam> { +class Her2Tests : public ::testing::TestWithParam> { }; TEST_P(Her2Tests, ComplexSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/her2_usm.cpp b/tests/unit_tests/blas/level2/her2_usm.cpp index 93c1d1f04..1fb29c132 100644 --- a/tests/unit_tests/blas/level2/her2_usm.cpp +++ b/tests/unit_tests/blas/level2/her2_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HER2:\n" << e.what() << std::endl; print_error_code(e); @@ -83,8 +83,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::her2(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), &incy_ref, - (fp_ref *)A_ref.data(), &lda_ref); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref, + (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ HER2. @@ -93,13 +93,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::her2(main_queue, upper_lower, n, alpha, - x.data(), incx, y.data(), incy, - A.data(), lda, dependencies); + x.data(), incx, y.data(), incy, + A.data(), lda, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::her2(main_queue, upper_lower, n, alpha, - x.data(), incx, y.data(), incy, A.data(), - lda, dependencies); + x.data(), incx, y.data(), incy, A.data(), + lda, dependencies); break; default: break; } @@ -112,25 +112,25 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe A.data(), lda, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::her2, upper_lower, - n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::her2, + upper_lower, n, alpha, x.data(), incx, y.data(), incy, + A.data(), lda, dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HER2:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HER2:\n" << error.what() << std::endl; } @@ -142,7 +142,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class Her2UsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Her2UsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); diff --git a/tests/unit_tests/blas/level2/her_usm.cpp b/tests/unit_tests/blas/level2/her_usm.cpp index 0a7f86afe..fa614dd25 100644 --- a/tests/unit_tests/blas/level2/her_usm.cpp +++ b/tests/unit_tests/blas/level2/her_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp_scalar alpha, int incx, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HER:\n" << e.what() << std::endl; print_error_code(e); @@ -82,7 +82,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_scalar_ref = typename ref_type_info::type; ::her(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_scalar_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data(), &lda_ref); + (fp_scalar_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ HER. @@ -116,16 +116,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HER:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HER:\n" << error.what() << std::endl; } @@ -137,7 +137,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class HerUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HerUsmTests, ComplexSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level2/hpmv.cpp b/tests/unit_tests/blas/level2/hpmv.cpp index 8894d1e0b..08b221536 100644 --- a/tests/unit_tests/blas/level2/hpmv.cpp +++ b/tests/unit_tests/blas/level2/hpmv.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy) { // Prepare data. vector x, y, y_ref, A; @@ -61,18 +61,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::hpmv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), (fp_ref *)x.data(), &incx_ref, (fp_ref *)&beta, - (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), (fp_ref*)x.data(), &incx_ref, (fp_ref*)&beta, + (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ HPMV. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HPMV:\n" << e.what() << std::endl; print_error_code(e); @@ -91,11 +91,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::hpmv(main_queue, upper_lower, n, alpha, A_buffer, - x_buffer, incx, beta, y_buffer, incy); + x_buffer, incx, beta, y_buffer, incy); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::hpmv(main_queue, upper_lower, n, alpha, A_buffer, - x_buffer, incx, beta, y_buffer, incy); + x_buffer, incx, beta, y_buffer, incy); break; default: break; } @@ -107,23 +107,24 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe y_buffer, incy); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpmv, upper_lower, - n, alpha, A_buffer, x_buffer, incx, beta, y_buffer, incy); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpmv, + upper_lower, n, alpha, A_buffer, x_buffer, incx, beta, + y_buffer, incy); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HPMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HPMV:\n" << error.what() << std::endl; } @@ -134,7 +135,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class HpmvTests : public ::testing::TestWithParam> { +class HpmvTests : public ::testing::TestWithParam> { }; TEST_P(HpmvTests, ComplexSinglePrecision) { @@ -145,9 +146,11 @@ TEST_P(HpmvTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::lower, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, + -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::upper, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, + -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, beta, 1, 1)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), diff --git a/tests/unit_tests/blas/level2/hpmv_usm.cpp b/tests/unit_tests/blas/level2/hpmv_usm.cpp index d2302f573..b81adfccb 100644 --- a/tests/unit_tests/blas/level2/hpmv_usm.cpp +++ b/tests/unit_tests/blas/level2/hpmv_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HPMV:\n" << e.what() << std::endl; print_error_code(e); @@ -82,8 +82,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::hpmv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), (fp_ref *)x.data(), &incx_ref, (fp_ref *)&beta, - (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), (fp_ref*)x.data(), &incx_ref, (fp_ref*)&beta, + (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ HPMV. @@ -92,13 +92,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::hpmv(main_queue, upper_lower, n, alpha, - A.data(), x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), x.data(), incx, beta, + y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::hpmv(main_queue, upper_lower, n, alpha, - A.data(), x.data(), incx, beta, y.data(), - incy, dependencies); + A.data(), x.data(), incx, beta, y.data(), + incy, dependencies); break; default: break; } @@ -111,25 +111,25 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpmv, upper_lower, - n, alpha, A.data(), x.data(), incx, beta, y.data(), incy, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpmv, + upper_lower, n, alpha, A.data(), x.data(), incx, beta, + y.data(), incy, dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HPMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HPMV:\n" << error.what() << std::endl; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class HpmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HpmvUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); @@ -151,9 +151,11 @@ TEST_P(HpmvUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, beta, 2, 3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::lower, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::lower, 30, alpha, beta, -2, + -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::upper, 30, alpha, beta, -2, -3)); + oneapi::math::uplo::upper, 30, alpha, beta, -2, + -3)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, beta, 1, 1)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), diff --git a/tests/unit_tests/blas/level2/hpr.cpp b/tests/unit_tests/blas/level2/hpr.cpp index 3ea336bab..492c07457 100644 --- a/tests/unit_tests/blas/level2/hpr.cpp +++ b/tests/unit_tests/blas/level2/hpr.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp_scalar alpha, int incx) { // Prepare data. vector x, A_ref, A; @@ -61,17 +61,17 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_scalar_ref = typename ref_type_info::type; ::hpr(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_scalar_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data()); + (fp_scalar_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)A_ref.data()); // Call DPC++ HPR. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HPR:\n" << e.what() << std::endl; print_error_code(e); @@ -89,11 +89,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::hpr(main_queue, upper_lower, n, alpha, x_buffer, - incx, A_buffer); + incx, A_buffer); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::hpr(main_queue, upper_lower, n, alpha, x_buffer, incx, - A_buffer); + oneapi::math::blas::row_major::hpr(main_queue, upper_lower, n, alpha, x_buffer, + incx, A_buffer); break; default: break; } @@ -111,16 +111,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HPR:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HPR:\n" << error.what() << std::endl; } @@ -131,45 +131,53 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class HprTests : public ::testing::TestWithParam> { +class HprTests : public ::testing::TestWithParam> { }; TEST_P(HprTests, ComplexSinglePrecision) { float alpha(2.0); - EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 2))); - EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 2))); + EXPECT_TRUEORSKIP( + (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::lower, 30, alpha, 2))); + EXPECT_TRUEORSKIP( + (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::upper, 30, alpha, 2))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, -2))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, -2))); - EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 1))); - EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 1))); + EXPECT_TRUEORSKIP( + (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::lower, 30, alpha, 1))); + EXPECT_TRUEORSKIP( + (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::upper, 30, alpha, 1))); } TEST_P(HprTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); double alpha(2.0); - EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 2))); - EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 2))); + EXPECT_TRUEORSKIP( + (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::lower, 30, alpha, 2))); + EXPECT_TRUEORSKIP( + (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::upper, 30, alpha, 2))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, -2))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, -2))); - EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 1))); - EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 1))); + EXPECT_TRUEORSKIP( + (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::lower, 30, alpha, 1))); + EXPECT_TRUEORSKIP( + (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::upper, 30, alpha, 1))); } INSTANTIATE_TEST_SUITE_P(HprTestSuite, HprTests, diff --git a/tests/unit_tests/blas/level2/hpr2.cpp b/tests/unit_tests/blas/level2/hpr2.cpp index 5ab9b7d28..d8780b576 100644 --- a/tests/unit_tests/blas/level2/hpr2.cpp +++ b/tests/unit_tests/blas/level2/hpr2.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy) { // Prepare data. vector x, y, A_ref, A; @@ -61,18 +61,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::hpr2(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), &incy_ref, - (fp_ref *)A_ref.data()); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref, + (fp_ref*)A_ref.data()); // Call DPC++ HPR2. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HPR2:\n" << e.what() << std::endl; print_error_code(e); @@ -91,11 +91,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::hpr2(main_queue, upper_lower, n, alpha, x_buffer, - incx, y_buffer, incy, A_buffer); + incx, y_buffer, incy, A_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::hpr2(main_queue, upper_lower, n, alpha, x_buffer, - incx, y_buffer, incy, A_buffer); + incx, y_buffer, incy, A_buffer); break; default: break; } @@ -107,23 +107,24 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe A_buffer); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpr2, upper_lower, - n, alpha, x_buffer, incx, y_buffer, incy, A_buffer); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpr2, + upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, + A_buffer); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HPR2:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HPR2:\n" << error.what() << std::endl; } @@ -134,7 +135,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class Hpr2Tests : public ::testing::TestWithParam> { +class Hpr2Tests : public ::testing::TestWithParam> { }; TEST_P(Hpr2Tests, ComplexSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/hpr2_usm.cpp b/tests/unit_tests/blas/level2/hpr2_usm.cpp index 4ac809338..6a9d4c55e 100644 --- a/tests/unit_tests/blas/level2/hpr2_usm.cpp +++ b/tests/unit_tests/blas/level2/hpr2_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HPR2:\n" << e.what() << std::endl; print_error_code(e); @@ -82,8 +82,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::hpr2(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), &incy_ref, - (fp_ref *)A_ref.data()); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref, + (fp_ref*)A_ref.data()); // Call DPC++ HPR2. @@ -92,13 +92,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::hpr2(main_queue, upper_lower, n, alpha, - x.data(), incx, y.data(), incy, - A.data(), dependencies); + x.data(), incx, y.data(), incy, + A.data(), dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::hpr2(main_queue, upper_lower, n, alpha, - x.data(), incx, y.data(), incy, A.data(), - dependencies); + x.data(), incx, y.data(), incy, A.data(), + dependencies); break; default: break; } @@ -111,25 +111,25 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe A.data(), dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpr2, upper_lower, - n, alpha, x.data(), incx, y.data(), incy, A.data(), - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::hpr2, + upper_lower, n, alpha, x.data(), incx, y.data(), incy, + A.data(), dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HPR2:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HPR2:\n" << error.what() << std::endl; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class Hpr2UsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Hpr2UsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); diff --git a/tests/unit_tests/blas/level2/hpr_usm.cpp b/tests/unit_tests/blas/level2/hpr_usm.cpp index 003d41f0b..a0e60c9d2 100644 --- a/tests/unit_tests/blas/level2/hpr_usm.cpp +++ b/tests/unit_tests/blas/level2/hpr_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp_scalar alpha, int incx) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during HPR:\n" << e.what() << std::endl; print_error_code(e); @@ -82,7 +82,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_scalar_ref = typename ref_type_info::type; ::hpr(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_scalar_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data()); + (fp_scalar_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)A_ref.data()); // Call DPC++ HPR. @@ -90,12 +90,12 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - done = oneapi::math::blas::column_major::hpr(main_queue, upper_lower, n, alpha, - x.data(), incx, A.data(), dependencies); + done = oneapi::math::blas::column_major::hpr( + main_queue, upper_lower, n, alpha, x.data(), incx, A.data(), dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::hpr(main_queue, upper_lower, n, alpha, - x.data(), incx, A.data(), dependencies); + x.data(), incx, A.data(), dependencies); break; default: break; } @@ -116,16 +116,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during HPR:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of HPR:\n" << error.what() << std::endl; } @@ -137,44 +137,52 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class HprUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(HprUsmTests, ComplexSinglePrecision) { float alpha(2.0); - EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 2))); - EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 2))); + EXPECT_TRUEORSKIP( + (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::lower, 30, alpha, 2))); + EXPECT_TRUEORSKIP( + (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::upper, 30, alpha, 2))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, -2))); EXPECT_TRUEORSKIP( (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, -2))); - EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 1))); - EXPECT_TRUEORSKIP((test, float>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 1))); + EXPECT_TRUEORSKIP( + (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::lower, 30, alpha, 1))); + EXPECT_TRUEORSKIP( + (test, float>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::upper, 30, alpha, 1))); } TEST_P(HprUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); double alpha(2.0); - EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 2))); - EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 2))); + EXPECT_TRUEORSKIP( + (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::lower, 30, alpha, 2))); + EXPECT_TRUEORSKIP( + (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::upper, 30, alpha, 2))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, -2))); EXPECT_TRUEORSKIP( (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, -2))); - EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::lower, 30, alpha, 1))); - EXPECT_TRUEORSKIP((test, double>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::uplo::upper, 30, alpha, 1))); + EXPECT_TRUEORSKIP( + (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::lower, 30, alpha, 1))); + EXPECT_TRUEORSKIP( + (test, double>(std::get<0>(GetParam()), std::get<1>(GetParam()), + oneapi::math::uplo::upper, 30, alpha, 1))); } INSTANTIATE_TEST_SUITE_P(HprUsmTestSuite, HprUsmTests, diff --git a/tests/unit_tests/blas/level2/sbmv.cpp b/tests/unit_tests/blas/level2/sbmv.cpp index 9291e14de..64f3a71bd 100644 --- a/tests/unit_tests/blas/level2/sbmv.cpp +++ b/tests/unit_tests/blas/level2/sbmv.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. vector x, y, y_ref, A; @@ -62,18 +62,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::sbmv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, &k_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ SBMV. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SBMV:\n" << e.what() << std::endl; print_error_code(e); @@ -91,12 +91,12 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::sbmv(main_queue, upper_lower, n, k, alpha, - A_buffer, lda, x_buffer, incx, beta, y_buffer, - incy); + A_buffer, lda, x_buffer, incx, beta, + y_buffer, incy); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::sbmv(main_queue, upper_lower, n, k, alpha, A_buffer, - lda, x_buffer, incx, beta, y_buffer, incy); + lda, x_buffer, incx, beta, y_buffer, incy); break; default: break; } @@ -108,24 +108,24 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe beta, y_buffer, incy); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::sbmv, upper_lower, - n, k, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, - incy); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::sbmv, + upper_lower, n, k, alpha, A_buffer, lda, x_buffer, incx, + beta, y_buffer, incy); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SBMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SBMV:\n" << error.what() << std::endl; } @@ -136,7 +136,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class SbmvTests : public ::testing::TestWithParam> { +class SbmvTests : public ::testing::TestWithParam> { }; TEST_P(SbmvTests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/sbmv_usm.cpp b/tests/unit_tests/blas/level2/sbmv_usm.cpp index ad0a8b9b0..fd31e5285 100644 --- a/tests/unit_tests/blas/level2/sbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/sbmv_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, int k, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SBMV:\n" << e.what() << std::endl; print_error_code(e); @@ -83,8 +83,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::sbmv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, &k_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ SBMV. @@ -93,13 +93,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::sbmv(main_queue, upper_lower, n, k, alpha, - A.data(), lda, x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::sbmv(main_queue, upper_lower, n, k, alpha, - A.data(), lda, x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; default: break; } @@ -112,25 +112,25 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe beta, y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::sbmv, upper_lower, - n, k, alpha, A.data(), lda, x.data(), incx, beta, y.data(), - incy, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::sbmv, + upper_lower, n, k, alpha, A.data(), lda, x.data(), incx, + beta, y.data(), incy, dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SBMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SBMV:\n" << error.what() << std::endl; } @@ -142,7 +142,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class SbmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SbmvUsmTests, RealSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level2/spmv.cpp b/tests/unit_tests/blas/level2/spmv.cpp index 0fb16c5c5..9935a61da 100644 --- a/tests/unit_tests/blas/level2/spmv.cpp +++ b/tests/unit_tests/blas/level2/spmv.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy) { // Prepare data. vector x, y, y_ref, A; @@ -61,18 +61,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::spmv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), (fp_ref *)x.data(), &incx_ref, (fp_ref *)&beta, - (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), (fp_ref*)x.data(), &incx_ref, (fp_ref*)&beta, + (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ SPMV. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SPMV:\n" << e.what() << std::endl; print_error_code(e); @@ -91,11 +91,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::spmv(main_queue, upper_lower, n, alpha, A_buffer, - x_buffer, incx, beta, y_buffer, incy); + x_buffer, incx, beta, y_buffer, incy); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::spmv(main_queue, upper_lower, n, alpha, A_buffer, - x_buffer, incx, beta, y_buffer, incy); + x_buffer, incx, beta, y_buffer, incy); break; default: break; } @@ -107,23 +107,24 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe y_buffer, incy); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spmv, upper_lower, - n, alpha, A_buffer, x_buffer, incx, beta, y_buffer, incy); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spmv, + upper_lower, n, alpha, A_buffer, x_buffer, incx, beta, + y_buffer, incy); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SPMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SPMV:\n" << error.what() << std::endl; } @@ -134,7 +135,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class SpmvTests : public ::testing::TestWithParam> { +class SpmvTests : public ::testing::TestWithParam> { }; TEST_P(SpmvTests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/spmv_usm.cpp b/tests/unit_tests/blas/level2/spmv_usm.cpp index 7998fc02b..703c25232 100644 --- a/tests/unit_tests/blas/level2/spmv_usm.cpp +++ b/tests/unit_tests/blas/level2/spmv_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SPMV:\n" << e.what() << std::endl; print_error_code(e); @@ -82,8 +82,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::spmv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), (fp_ref *)x.data(), &incx_ref, (fp_ref *)&beta, - (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), (fp_ref*)x.data(), &incx_ref, (fp_ref*)&beta, + (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ SPMV. @@ -92,13 +92,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::spmv(main_queue, upper_lower, n, alpha, - A.data(), x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), x.data(), incx, beta, + y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::spmv(main_queue, upper_lower, n, alpha, - A.data(), x.data(), incx, beta, y.data(), - incy, dependencies); + A.data(), x.data(), incx, beta, y.data(), + incy, dependencies); break; default: break; } @@ -111,25 +111,25 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spmv, upper_lower, - n, alpha, A.data(), x.data(), incx, beta, y.data(), incy, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spmv, + upper_lower, n, alpha, A.data(), x.data(), incx, beta, + y.data(), incy, dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SPMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SPMV:\n" << error.what() << std::endl; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class SpmvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SpmvUsmTests, RealSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level2/spr.cpp b/tests/unit_tests/blas/level2/spr.cpp index 7f6706cc3..456335f24 100644 --- a/tests/unit_tests/blas/level2/spr.cpp +++ b/tests/unit_tests/blas/level2/spr.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx) { // Prepare data. vector x, A_ref, A; @@ -60,17 +60,17 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::spr(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data()); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)A_ref.data()); // Call DPC++ SPR. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SPR:\n" << e.what() << std::endl; print_error_code(e); @@ -88,11 +88,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::spr(main_queue, upper_lower, n, alpha, x_buffer, - incx, A_buffer); + incx, A_buffer); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::spr(main_queue, upper_lower, n, alpha, x_buffer, incx, - A_buffer); + oneapi::math::blas::row_major::spr(main_queue, upper_lower, n, alpha, x_buffer, + incx, A_buffer); break; default: break; } @@ -110,16 +110,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SPR:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SPR:\n" << error.what() << std::endl; } @@ -130,7 +130,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class SprTests : public ::testing::TestWithParam> { +class SprTests : public ::testing::TestWithParam> { }; TEST_P(SprTests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/spr2.cpp b/tests/unit_tests/blas/level2/spr2.cpp index 6cec654c0..78cfce411 100644 --- a/tests/unit_tests/blas/level2/spr2.cpp +++ b/tests/unit_tests/blas/level2/spr2.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy) { // Prepare data. vector x, y, A_ref, A; @@ -61,18 +61,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::spr2(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), &incy_ref, - (fp_ref *)A_ref.data()); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref, + (fp_ref*)A_ref.data()); // Call DPC++ SPR2. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SPR2:\n" << e.what() << std::endl; print_error_code(e); @@ -91,11 +91,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::spr2(main_queue, upper_lower, n, alpha, x_buffer, - incx, y_buffer, incy, A_buffer); + incx, y_buffer, incy, A_buffer); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::spr2(main_queue, upper_lower, n, alpha, x_buffer, - incx, y_buffer, incy, A_buffer); + incx, y_buffer, incy, A_buffer); break; default: break; } @@ -107,23 +107,24 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe A_buffer); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spr2, upper_lower, - n, alpha, x_buffer, incx, y_buffer, incy, A_buffer); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spr2, + upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, + A_buffer); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SPR2:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SPR2:\n" << error.what() << std::endl; } @@ -134,7 +135,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class Spr2Tests : public ::testing::TestWithParam> { +class Spr2Tests : public ::testing::TestWithParam> { }; TEST_P(Spr2Tests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/spr2_usm.cpp b/tests/unit_tests/blas/level2/spr2_usm.cpp index 27e4ce07e..ec283b22e 100644 --- a/tests/unit_tests/blas/level2/spr2_usm.cpp +++ b/tests/unit_tests/blas/level2/spr2_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SPR2:\n" << e.what() << std::endl; print_error_code(e); @@ -82,8 +82,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::spr2(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), &incy_ref, - (fp_ref *)A_ref.data()); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref, + (fp_ref*)A_ref.data()); // Call DPC++ SPR2. @@ -92,13 +92,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::spr2(main_queue, upper_lower, n, alpha, - x.data(), incx, y.data(), incy, - A.data(), dependencies); + x.data(), incx, y.data(), incy, + A.data(), dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::spr2(main_queue, upper_lower, n, alpha, - x.data(), incx, y.data(), incy, A.data(), - dependencies); + x.data(), incx, y.data(), incy, A.data(), + dependencies); break; default: break; } @@ -111,25 +111,25 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe A.data(), dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spr2, upper_lower, - n, alpha, x.data(), incx, y.data(), incy, A.data(), - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::spr2, + upper_lower, n, alpha, x.data(), incx, y.data(), incy, + A.data(), dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SPR2:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SPR2:\n" << error.what() << std::endl; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class Spr2UsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Spr2UsmTests, RealSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level2/spr_usm.cpp b/tests/unit_tests/blas/level2/spr_usm.cpp index 20e74159b..e70cbfd74 100644 --- a/tests/unit_tests/blas/level2/spr_usm.cpp +++ b/tests/unit_tests/blas/level2/spr_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SPR:\n" << e.what() << std::endl; print_error_code(e); @@ -81,7 +81,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::spr(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data()); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)A_ref.data()); // Call DPC++ SPR. @@ -89,12 +89,12 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - done = oneapi::math::blas::column_major::spr(main_queue, upper_lower, n, alpha, - x.data(), incx, A.data(), dependencies); + done = oneapi::math::blas::column_major::spr( + main_queue, upper_lower, n, alpha, x.data(), incx, A.data(), dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::spr(main_queue, upper_lower, n, alpha, - x.data(), incx, A.data(), dependencies); + x.data(), incx, A.data(), dependencies); break; default: break; } @@ -115,16 +115,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SPR:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SPR:\n" << error.what() << std::endl; } @@ -136,7 +136,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class SprUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SprUsmTests, RealSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level2/symv.cpp b/tests/unit_tests/blas/level2/symv.cpp index 4e62799e8..6e136c771 100644 --- a/tests/unit_tests/blas/level2/symv.cpp +++ b/tests/unit_tests/blas/level2/symv.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Prepare data. vector x, y, y_ref, A; @@ -61,18 +61,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::symv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ SYMV. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SYMV:\n" << e.what() << std::endl; print_error_code(e); @@ -91,11 +91,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::symv(main_queue, upper_lower, n, alpha, A_buffer, - lda, x_buffer, incx, beta, y_buffer, incy); + lda, x_buffer, incx, beta, y_buffer, incy); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::symv(main_queue, upper_lower, n, alpha, A_buffer, lda, - x_buffer, incx, beta, y_buffer, incy); + oneapi::math::blas::row_major::symv(main_queue, upper_lower, n, alpha, A_buffer, + lda, x_buffer, incx, beta, y_buffer, incy); break; default: break; } @@ -107,24 +107,24 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe y_buffer, incy); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::symv, upper_lower, - n, alpha, A_buffer, lda, x_buffer, incx, beta, y_buffer, - incy); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::symv, + upper_lower, n, alpha, A_buffer, lda, x_buffer, incx, beta, + y_buffer, incy); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SYMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SYMV:\n" << error.what() << std::endl; } @@ -135,7 +135,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class SymvTests : public ::testing::TestWithParam> { +class SymvTests : public ::testing::TestWithParam> { }; TEST_P(SymvTests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/symv_usm.cpp b/tests/unit_tests/blas/level2/symv_usm.cpp index 30b82190c..def858041 100644 --- a/tests/unit_tests/blas/level2/symv_usm.cpp +++ b/tests/unit_tests/blas/level2/symv_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, fp beta, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SYMV:\n" << e.what() << std::endl; print_error_code(e); @@ -82,8 +82,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::symv(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)A.data(), &lda_ref, (fp_ref *)x.data(), &incx_ref, - (fp_ref *)&beta, (fp_ref *)y_ref.data(), &incy_ref); + (fp_ref*)&alpha, (fp_ref*)A.data(), &lda_ref, (fp_ref*)x.data(), &incx_ref, + (fp_ref*)&beta, (fp_ref*)y_ref.data(), &incy_ref); // Call DPC++ SYMV. @@ -92,13 +92,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::symv(main_queue, upper_lower, n, alpha, - A.data(), lda, x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::symv(main_queue, upper_lower, n, alpha, - A.data(), lda, x.data(), incx, beta, - y.data(), incy, dependencies); + A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; default: break; } @@ -111,25 +111,25 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe y.data(), incy, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::symv, upper_lower, - n, alpha, A.data(), lda, x.data(), incx, beta, y.data(), - incy, dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::symv, + upper_lower, n, alpha, A.data(), lda, x.data(), incx, beta, + y.data(), incy, dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SYMV:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SYMV:\n" << error.what() << std::endl; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class SymvUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SymvUsmTests, RealSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level2/syr.cpp b/tests/unit_tests/blas/level2/syr.cpp index 1ce3ecf81..11679eabc 100644 --- a/tests/unit_tests/blas/level2/syr.cpp +++ b/tests/unit_tests/blas/level2/syr.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int lda) { // Prepare data. vector x, A_ref, A; @@ -60,17 +60,17 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::syr(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data(), &lda_ref); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ SYR. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SYR:\n" << e.what() << std::endl; print_error_code(e); @@ -88,11 +88,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::syr(main_queue, upper_lower, n, alpha, x_buffer, - incx, A_buffer, lda); + incx, A_buffer, lda); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::syr(main_queue, upper_lower, n, alpha, x_buffer, incx, - A_buffer, lda); + oneapi::math::blas::row_major::syr(main_queue, upper_lower, n, alpha, x_buffer, + incx, A_buffer, lda); break; default: break; } @@ -110,16 +110,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SYR:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SYR:\n" << error.what() << std::endl; } @@ -130,7 +130,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class SyrTests : public ::testing::TestWithParam> { +class SyrTests : public ::testing::TestWithParam> { }; TEST_P(SyrTests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/syr2.cpp b/tests/unit_tests/blas/level2/syr2.cpp index 3c3b3ceac..622d5eeac 100644 --- a/tests/unit_tests/blas/level2/syr2.cpp +++ b/tests/unit_tests/blas/level2/syr2.cpp @@ -42,12 +42,12 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy, int lda) { // Prepare data. vector x, y, A_ref, A; @@ -61,18 +61,18 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::syr2(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), &incy_ref, - (fp_ref *)A_ref.data(), &lda_ref); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref, + (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ SYR2. // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SYR2:\n" << e.what() << std::endl; print_error_code(e); @@ -91,11 +91,11 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::syr2(main_queue, upper_lower, n, alpha, x_buffer, - incx, y_buffer, incy, A_buffer, lda); + incx, y_buffer, incy, A_buffer, lda); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::syr2(main_queue, upper_lower, n, alpha, x_buffer, - incx, y_buffer, incy, A_buffer, lda); + incx, y_buffer, incy, A_buffer, lda); break; default: break; } @@ -107,23 +107,24 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe A_buffer, lda); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syr2, upper_lower, - n, alpha, x_buffer, incx, y_buffer, incy, A_buffer, lda); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syr2, + upper_lower, n, alpha, x_buffer, incx, y_buffer, incy, + A_buffer, lda); break; default: break; } #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SYR2:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SYR2:\n" << error.what() << std::endl; } @@ -134,7 +135,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class Syr2Tests : public ::testing::TestWithParam> { +class Syr2Tests : public ::testing::TestWithParam> { }; TEST_P(Syr2Tests, RealSinglePrecision) { diff --git a/tests/unit_tests/blas/level2/syr2_usm.cpp b/tests/unit_tests/blas/level2/syr2_usm.cpp index 94d3ce524..c56fc8647 100644 --- a/tests/unit_tests/blas/level2/syr2_usm.cpp +++ b/tests/unit_tests/blas/level2/syr2_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int incy, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SYR2:\n" << e.what() << std::endl; print_error_code(e); @@ -82,8 +82,8 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::syr2(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)y.data(), &incy_ref, - (fp_ref *)A_ref.data(), &lda_ref); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)y.data(), &incy_ref, + (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ SYR2. @@ -92,13 +92,13 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::syr2(main_queue, upper_lower, n, alpha, - x.data(), incx, y.data(), incy, - A.data(), lda, dependencies); + x.data(), incx, y.data(), incy, + A.data(), lda, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::syr2(main_queue, upper_lower, n, alpha, - x.data(), incx, y.data(), incy, A.data(), - lda, dependencies); + x.data(), incx, y.data(), incy, A.data(), + lda, dependencies); break; default: break; } @@ -111,25 +111,25 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe A.data(), lda, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syr2, upper_lower, - n, alpha, x.data(), incx, y.data(), incy, A.data(), lda, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syr2, + upper_lower, n, alpha, x.data(), incx, y.data(), incy, + A.data(), lda, dependencies); break; default: break; } main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SYR2:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SYR2:\n" << error.what() << std::endl; } @@ -141,7 +141,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class Syr2UsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(Syr2UsmTests, RealSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level2/syr_usm.cpp b/tests/unit_tests/blas/level2/syr_usm.cpp index 9a693ae1e..96992b8fa 100644 --- a/tests/unit_tests/blas/level2/syr_usm.cpp +++ b/tests/unit_tests/blas/level2/syr_usm.cpp @@ -42,20 +42,20 @@ using namespace sycl; using std::vector; -extern std::vector devices; +extern std::vector devices; namespace { template -int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, +int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, int n, fp alpha, int incx, int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught asynchronous SYCL exception during SYR:\n" << e.what() << std::endl; print_error_code(e); @@ -81,7 +81,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe using fp_ref = typename ref_type_info::type; ::syr(convert_to_cblas_layout(layout), convert_to_cblas_uplo(upper_lower), &n_ref, - (fp_ref *)&alpha, (fp_ref *)x.data(), &incx_ref, (fp_ref *)A_ref.data(), &lda_ref); + (fp_ref*)&alpha, (fp_ref*)x.data(), &incx_ref, (fp_ref*)A_ref.data(), &lda_ref); // Call DPC++ SYR. @@ -115,16 +115,16 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe main_queue.wait(); #endif } - catch (exception const &e) { + catch (exception const& e) { std::cout << "Caught synchronous SYCL exception during SYR:\n" << e.what() << std::endl; print_error_code(e); } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of SYR:\n" << error.what() << std::endl; } @@ -136,7 +136,7 @@ int test(device *dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe } class SyrUsmTests - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; TEST_P(SyrUsmTests, RealSinglePrecision) { float alpha(2.0); diff --git a/tests/unit_tests/blas/level2/tbmv.cpp b/tests/unit_tests/blas/level2/tbmv.cpp index 0fb975581..f2e601b42 100644 --- a/tests/unit_tests/blas/level2/tbmv.cpp +++ b/tests/unit_tests/blas/level2/tbmv.cpp @@ -90,12 +90,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::tbmv(main_queue, upper_lower, transa, unit_nonunit, - n, k, A_buffer, lda, x_buffer, incx); + oneapi::math::blas::column_major::tbmv(main_queue, upper_lower, transa, + unit_nonunit, n, k, A_buffer, lda, x_buffer, + incx); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::tbmv(main_queue, upper_lower, transa, unit_nonunit, n, - k, A_buffer, lda, x_buffer, incx); + oneapi::math::blas::row_major::tbmv(main_queue, upper_lower, transa, unit_nonunit, + n, k, A_buffer, lda, x_buffer, incx); break; default: break; } @@ -107,8 +108,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe x_buffer, incx); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbmv, upper_lower, - transa, unit_nonunit, n, k, A_buffer, lda, x_buffer, incx); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbmv, + upper_lower, transa, unit_nonunit, n, k, A_buffer, lda, + x_buffer, incx); break; default: break; } diff --git a/tests/unit_tests/blas/level2/tbmv_usm.cpp b/tests/unit_tests/blas/level2/tbmv_usm.cpp index 56efc0742..f64cef35b 100644 --- a/tests/unit_tests/blas/level2/tbmv_usm.cpp +++ b/tests/unit_tests/blas/level2/tbmv_usm.cpp @@ -93,13 +93,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::tbmv(main_queue, upper_lower, transa, - unit_nonunit, n, k, A.data(), lda, - x.data(), incx, dependencies); + unit_nonunit, n, k, A.data(), lda, + x.data(), incx, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::tbmv(main_queue, upper_lower, transa, - unit_nonunit, n, k, A.data(), lda, - x.data(), incx, dependencies); + unit_nonunit, n, k, A.data(), lda, + x.data(), incx, dependencies); break; default: break; } @@ -112,9 +112,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe x.data(), incx, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbmv, upper_lower, - transa, unit_nonunit, n, k, A.data(), lda, x.data(), incx, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbmv, + upper_lower, transa, unit_nonunit, n, k, A.data(), lda, + x.data(), incx, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level2/tbsv.cpp b/tests/unit_tests/blas/level2/tbsv.cpp index 7127a98d1..5747091d5 100644 --- a/tests/unit_tests/blas/level2/tbsv.cpp +++ b/tests/unit_tests/blas/level2/tbsv.cpp @@ -90,12 +90,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::tbsv(main_queue, upper_lower, transa, unit_nonunit, - n, k, A_buffer, lda, x_buffer, incx); + oneapi::math::blas::column_major::tbsv(main_queue, upper_lower, transa, + unit_nonunit, n, k, A_buffer, lda, x_buffer, + incx); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::tbsv(main_queue, upper_lower, transa, unit_nonunit, n, - k, A_buffer, lda, x_buffer, incx); + oneapi::math::blas::row_major::tbsv(main_queue, upper_lower, transa, unit_nonunit, + n, k, A_buffer, lda, x_buffer, incx); break; default: break; } @@ -107,8 +108,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe x_buffer, incx); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbsv, upper_lower, - transa, unit_nonunit, n, k, A_buffer, lda, x_buffer, incx); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbsv, + upper_lower, transa, unit_nonunit, n, k, A_buffer, lda, + x_buffer, incx); break; default: break; } diff --git a/tests/unit_tests/blas/level2/tbsv_usm.cpp b/tests/unit_tests/blas/level2/tbsv_usm.cpp index 769f28669..20b8a947e 100644 --- a/tests/unit_tests/blas/level2/tbsv_usm.cpp +++ b/tests/unit_tests/blas/level2/tbsv_usm.cpp @@ -93,13 +93,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::tbsv(main_queue, upper_lower, transa, - unit_nonunit, n, k, A.data(), lda, - x.data(), incx, dependencies); + unit_nonunit, n, k, A.data(), lda, + x.data(), incx, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::tbsv(main_queue, upper_lower, transa, - unit_nonunit, n, k, A.data(), lda, - x.data(), incx, dependencies); + unit_nonunit, n, k, A.data(), lda, + x.data(), incx, dependencies); break; default: break; } @@ -112,9 +112,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe x.data(), incx, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbsv, upper_lower, - transa, unit_nonunit, n, k, A.data(), lda, x.data(), incx, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tbsv, + upper_lower, transa, unit_nonunit, n, k, A.data(), lda, + x.data(), incx, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level2/tpmv.cpp b/tests/unit_tests/blas/level2/tpmv.cpp index 558fd901a..b52a50656 100644 --- a/tests/unit_tests/blas/level2/tpmv.cpp +++ b/tests/unit_tests/blas/level2/tpmv.cpp @@ -88,12 +88,12 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::tpmv(main_queue, upper_lower, transa, unit_nonunit, - n, A_buffer, x_buffer, incx); + oneapi::math::blas::column_major::tpmv(main_queue, upper_lower, transa, + unit_nonunit, n, A_buffer, x_buffer, incx); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::tpmv(main_queue, upper_lower, transa, unit_nonunit, n, - A_buffer, x_buffer, incx); + oneapi::math::blas::row_major::tpmv(main_queue, upper_lower, transa, unit_nonunit, + n, A_buffer, x_buffer, incx); break; default: break; } @@ -105,8 +105,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe incx); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpmv, upper_lower, - transa, unit_nonunit, n, A_buffer, x_buffer, incx); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpmv, + upper_lower, transa, unit_nonunit, n, A_buffer, x_buffer, + incx); break; default: break; } diff --git a/tests/unit_tests/blas/level2/tpmv_usm.cpp b/tests/unit_tests/blas/level2/tpmv_usm.cpp index 04e0bae47..40722a09c 100644 --- a/tests/unit_tests/blas/level2/tpmv_usm.cpp +++ b/tests/unit_tests/blas/level2/tpmv_usm.cpp @@ -91,13 +91,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::tpmv(main_queue, upper_lower, transa, - unit_nonunit, n, A.data(), x.data(), - incx, dependencies); + unit_nonunit, n, A.data(), x.data(), + incx, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::tpmv(main_queue, upper_lower, transa, - unit_nonunit, n, A.data(), x.data(), incx, - dependencies); + unit_nonunit, n, A.data(), x.data(), + incx, dependencies); break; default: break; } @@ -110,9 +110,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe incx, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpmv, upper_lower, - transa, unit_nonunit, n, A.data(), x.data(), incx, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpmv, + upper_lower, transa, unit_nonunit, n, A.data(), x.data(), + incx, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level2/tpsv.cpp b/tests/unit_tests/blas/level2/tpsv.cpp index 2f7e0423c..daebf6d58 100644 --- a/tests/unit_tests/blas/level2/tpsv.cpp +++ b/tests/unit_tests/blas/level2/tpsv.cpp @@ -88,12 +88,12 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::tpsv(main_queue, upper_lower, transa, unit_nonunit, - n, A_buffer, x_buffer, incx); + oneapi::math::blas::column_major::tpsv(main_queue, upper_lower, transa, + unit_nonunit, n, A_buffer, x_buffer, incx); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::tpsv(main_queue, upper_lower, transa, unit_nonunit, n, - A_buffer, x_buffer, incx); + oneapi::math::blas::row_major::tpsv(main_queue, upper_lower, transa, unit_nonunit, + n, A_buffer, x_buffer, incx); break; default: break; } @@ -105,8 +105,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe incx); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpsv, upper_lower, - transa, unit_nonunit, n, A_buffer, x_buffer, incx); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpsv, + upper_lower, transa, unit_nonunit, n, A_buffer, x_buffer, + incx); break; default: break; } diff --git a/tests/unit_tests/blas/level2/tpsv_usm.cpp b/tests/unit_tests/blas/level2/tpsv_usm.cpp index f3980ed7f..a41d0e3a0 100644 --- a/tests/unit_tests/blas/level2/tpsv_usm.cpp +++ b/tests/unit_tests/blas/level2/tpsv_usm.cpp @@ -91,13 +91,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::tpsv(main_queue, upper_lower, transa, - unit_nonunit, n, A.data(), x.data(), - incx, dependencies); + unit_nonunit, n, A.data(), x.data(), + incx, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::tpsv(main_queue, upper_lower, transa, - unit_nonunit, n, A.data(), x.data(), incx, - dependencies); + unit_nonunit, n, A.data(), x.data(), + incx, dependencies); break; default: break; } @@ -110,9 +110,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe incx, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpsv, upper_lower, - transa, unit_nonunit, n, A.data(), x.data(), incx, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::tpsv, + upper_lower, transa, unit_nonunit, n, A.data(), x.data(), + incx, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level2/trmv.cpp b/tests/unit_tests/blas/level2/trmv.cpp index 0ea3371bd..1a99590cd 100644 --- a/tests/unit_tests/blas/level2/trmv.cpp +++ b/tests/unit_tests/blas/level2/trmv.cpp @@ -48,7 +48,8 @@ namespace { template int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, int lda) { + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, + int lda) { // Prepare data. vector x, x_ref, A; rand_vector(x, n, incx); @@ -88,12 +89,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::trmv(main_queue, upper_lower, transa, unit_nonunit, - n, A_buffer, lda, x_buffer, incx); + oneapi::math::blas::column_major::trmv(main_queue, upper_lower, transa, + unit_nonunit, n, A_buffer, lda, x_buffer, + incx); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::trmv(main_queue, upper_lower, transa, unit_nonunit, n, - A_buffer, lda, x_buffer, incx); + oneapi::math::blas::row_major::trmv(main_queue, upper_lower, transa, unit_nonunit, + n, A_buffer, lda, x_buffer, incx); break; default: break; } @@ -105,8 +107,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe x_buffer, incx); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trmv, upper_lower, - transa, unit_nonunit, n, A_buffer, lda, x_buffer, incx); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trmv, + upper_lower, transa, unit_nonunit, n, A_buffer, lda, + x_buffer, incx); break; default: break; } diff --git a/tests/unit_tests/blas/level2/trmv_usm.cpp b/tests/unit_tests/blas/level2/trmv_usm.cpp index 1b7fd27c6..d11d9a68a 100644 --- a/tests/unit_tests/blas/level2/trmv_usm.cpp +++ b/tests/unit_tests/blas/level2/trmv_usm.cpp @@ -48,7 +48,8 @@ namespace { template int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, int lda) { + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, + int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -91,13 +92,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::trmv(main_queue, upper_lower, transa, - unit_nonunit, n, A.data(), lda, - x.data(), incx, dependencies); + unit_nonunit, n, A.data(), lda, + x.data(), incx, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::trmv(main_queue, upper_lower, transa, - unit_nonunit, n, A.data(), lda, x.data(), - incx, dependencies); + unit_nonunit, n, A.data(), lda, x.data(), + incx, dependencies); break; default: break; } @@ -110,9 +111,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe x.data(), incx, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trmv, upper_lower, - transa, unit_nonunit, n, A.data(), lda, x.data(), incx, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trmv, + upper_lower, transa, unit_nonunit, n, A.data(), lda, + x.data(), incx, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level2/trsv.cpp b/tests/unit_tests/blas/level2/trsv.cpp index f9279f597..05c64e97b 100644 --- a/tests/unit_tests/blas/level2/trsv.cpp +++ b/tests/unit_tests/blas/level2/trsv.cpp @@ -48,7 +48,8 @@ namespace { template int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, int lda) { + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, + int lda) { // Prepare data. vector x, x_ref, A; rand_vector(x, n, incx); @@ -88,12 +89,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - oneapi::math::blas::column_major::trsv(main_queue, upper_lower, transa, unit_nonunit, - n, A_buffer, lda, x_buffer, incx); + oneapi::math::blas::column_major::trsv(main_queue, upper_lower, transa, + unit_nonunit, n, A_buffer, lda, x_buffer, + incx); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::trsv(main_queue, upper_lower, transa, unit_nonunit, n, - A_buffer, lda, x_buffer, incx); + oneapi::math::blas::row_major::trsv(main_queue, upper_lower, transa, unit_nonunit, + n, A_buffer, lda, x_buffer, incx); break; default: break; } @@ -105,8 +107,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe x_buffer, incx); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsv, upper_lower, - transa, unit_nonunit, n, A_buffer, lda, x_buffer, incx); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsv, + upper_lower, transa, unit_nonunit, n, A_buffer, lda, + x_buffer, incx); break; default: break; } diff --git a/tests/unit_tests/blas/level2/trsv_usm.cpp b/tests/unit_tests/blas/level2/trsv_usm.cpp index 53b8c1df7..98dbb5063 100644 --- a/tests/unit_tests/blas/level2/trsv_usm.cpp +++ b/tests/unit_tests/blas/level2/trsv_usm.cpp @@ -48,7 +48,8 @@ namespace { template int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, - oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, int lda) { + oneapi::math::transpose transa, oneapi::math::diag unit_nonunit, int n, int incx, + int lda) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -91,13 +92,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::trsv(main_queue, upper_lower, transa, - unit_nonunit, n, A.data(), lda, - x.data(), incx, dependencies); + unit_nonunit, n, A.data(), lda, + x.data(), incx, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::trsv(main_queue, upper_lower, transa, - unit_nonunit, n, A.data(), lda, x.data(), - incx, dependencies); + unit_nonunit, n, A.data(), lda, x.data(), + incx, dependencies); break; default: break; } @@ -110,9 +111,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe x.data(), incx, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsv, upper_lower, - transa, unit_nonunit, n, A.data(), lda, x.data(), incx, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::trsv, + upper_lower, transa, unit_nonunit, n, A.data(), lda, + x.data(), incx, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level3/gemm.cpp b/tests/unit_tests/blas/level3/gemm.cpp index c8f44a433..0350cdb1a 100644 --- a/tests/unit_tests/blas/level3/gemm.cpp +++ b/tests/unit_tests/blas/level3/gemm.cpp @@ -99,13 +99,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose trans switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::gemm(main_queue, transa, transb, m, n, k, alpha, - A_buffer, lda, B_buffer, ldb, beta, C_buffer, - ldc); + A_buffer, lda, B_buffer, ldb, beta, C_buffer, + ldc); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::gemm(main_queue, transa, transb, m, n, k, alpha, - A_buffer, lda, B_buffer, ldb, beta, C_buffer, - ldc); + A_buffer, lda, B_buffer, ldb, beta, C_buffer, + ldc); break; default: break; } diff --git a/tests/unit_tests/blas/level3/gemm_usm.cpp b/tests/unit_tests/blas/level3/gemm_usm.cpp index 82665e896..a18e79d79 100644 --- a/tests/unit_tests/blas/level3/gemm_usm.cpp +++ b/tests/unit_tests/blas/level3/gemm_usm.cpp @@ -99,13 +99,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::transpose trans switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::gemm(main_queue, transa, transb, m, n, k, - alpha, A.data(), lda, B.data(), ldb, - beta, C.data(), ldc, dependencies); + alpha, A.data(), lda, B.data(), ldb, + beta, C.data(), ldc, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::gemm(main_queue, transa, transb, m, n, k, - alpha, A.data(), lda, B.data(), ldb, beta, - C.data(), ldc, dependencies); + alpha, A.data(), lda, B.data(), ldb, + beta, C.data(), ldc, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level3/hemm.cpp b/tests/unit_tests/blas/level3/hemm.cpp index fc7b25858..6886195cf 100644 --- a/tests/unit_tests/blas/level3/hemm.cpp +++ b/tests/unit_tests/blas/level3/hemm.cpp @@ -98,13 +98,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::hemm(main_queue, left_right, upper_lower, m, n, - alpha, A_buffer, lda, B_buffer, ldb, beta, - C_buffer, ldc); + alpha, A_buffer, lda, B_buffer, ldb, beta, + C_buffer, ldc); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::hemm(main_queue, left_right, upper_lower, m, n, alpha, - A_buffer, lda, B_buffer, ldb, beta, C_buffer, - ldc); + oneapi::math::blas::row_major::hemm(main_queue, left_right, upper_lower, m, n, + alpha, A_buffer, lda, B_buffer, ldb, beta, + C_buffer, ldc); break; default: break; } @@ -157,30 +157,30 @@ TEST_P(HemmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(HemmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(HemmTestSuite, HemmTests, diff --git a/tests/unit_tests/blas/level3/hemm_usm.cpp b/tests/unit_tests/blas/level3/hemm_usm.cpp index 6d3314720..f898dc002 100644 --- a/tests/unit_tests/blas/level3/hemm_usm.cpp +++ b/tests/unit_tests/blas/level3/hemm_usm.cpp @@ -96,14 +96,14 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - done = oneapi::math::blas::column_major::hemm(main_queue, left_right, upper_lower, m, - n, alpha, A.data(), lda, B.data(), ldb, - beta, C.data(), ldc, dependencies); + done = oneapi::math::blas::column_major::hemm( + main_queue, left_right, upper_lower, m, n, alpha, A.data(), lda, B.data(), ldb, + beta, C.data(), ldc, dependencies); break; case oneapi::math::layout::row_major: - done = oneapi::math::blas::row_major::hemm(main_queue, left_right, upper_lower, m, n, - alpha, A.data(), lda, B.data(), ldb, beta, - C.data(), ldc, dependencies); + done = oneapi::math::blas::row_major::hemm(main_queue, left_right, upper_lower, m, + n, alpha, A.data(), lda, B.data(), ldb, + beta, C.data(), ldc, dependencies); break; default: break; } @@ -157,30 +157,30 @@ TEST_P(HemmUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(HemmUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(HemmUsmTestSuite, HemmUsmTests, diff --git a/tests/unit_tests/blas/level3/her2k.cpp b/tests/unit_tests/blas/level3/her2k.cpp index 73dbfa9ee..9df00b280 100644 --- a/tests/unit_tests/blas/level3/her2k.cpp +++ b/tests/unit_tests/blas/level3/her2k.cpp @@ -99,13 +99,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::her2k(main_queue, upper_lower, trans, n, k, alpha, - A_buffer, lda, B_buffer, ldb, beta, C_buffer, - ldc); + A_buffer, lda, B_buffer, ldb, beta, + C_buffer, ldc); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::her2k(main_queue, upper_lower, trans, n, k, alpha, - A_buffer, lda, B_buffer, ldb, beta, C_buffer, - ldc); + A_buffer, lda, B_buffer, ldb, beta, C_buffer, + ldc); break; default: break; } @@ -146,8 +146,8 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class Her2kTests : public ::testing::TestWithParam> { -}; +class Her2kTests + : public ::testing::TestWithParam> {}; TEST_P(Her2kTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); diff --git a/tests/unit_tests/blas/level3/her2k_usm.cpp b/tests/unit_tests/blas/level3/her2k_usm.cpp index 642fd6219..c975104cf 100644 --- a/tests/unit_tests/blas/level3/her2k_usm.cpp +++ b/tests/unit_tests/blas/level3/her2k_usm.cpp @@ -99,13 +99,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::her2k(main_queue, upper_lower, trans, n, k, - alpha, A.data(), lda, B.data(), ldb, - beta, C.data(), ldc, dependencies); + alpha, A.data(), lda, B.data(), ldb, + beta, C.data(), ldc, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::her2k(main_queue, upper_lower, trans, n, k, - alpha, A.data(), lda, B.data(), ldb, - beta, C.data(), ldc, dependencies); + alpha, A.data(), lda, B.data(), ldb, + beta, C.data(), ldc, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level3/herk.cpp b/tests/unit_tests/blas/level3/herk.cpp index eea240cd8..9c0e858b2 100644 --- a/tests/unit_tests/blas/level3/herk.cpp +++ b/tests/unit_tests/blas/level3/herk.cpp @@ -93,11 +93,11 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::herk(main_queue, upper_lower, trans, n, k, alpha, - A_buffer, lda, beta, C_buffer, ldc); + A_buffer, lda, beta, C_buffer, ldc); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::herk(main_queue, upper_lower, trans, n, k, alpha, - A_buffer, lda, beta, C_buffer, ldc); + A_buffer, lda, beta, C_buffer, ldc); break; default: break; } @@ -109,8 +109,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe C_buffer, ldc); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::herk, upper_lower, - trans, n, k, alpha, A_buffer, lda, beta, C_buffer, ldc); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::herk, + upper_lower, trans, n, k, alpha, A_buffer, lda, beta, + C_buffer, ldc); break; default: break; } diff --git a/tests/unit_tests/blas/level3/herk_usm.cpp b/tests/unit_tests/blas/level3/herk_usm.cpp index c2c51b853..82e37cb22 100644 --- a/tests/unit_tests/blas/level3/herk_usm.cpp +++ b/tests/unit_tests/blas/level3/herk_usm.cpp @@ -94,13 +94,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::herk(main_queue, upper_lower, trans, n, k, - alpha, A.data(), lda, beta, C.data(), - ldc, dependencies); + alpha, A.data(), lda, beta, C.data(), + ldc, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::herk(main_queue, upper_lower, trans, n, k, - alpha, A.data(), lda, beta, C.data(), ldc, - dependencies); + alpha, A.data(), lda, beta, C.data(), + ldc, dependencies); break; default: break; } @@ -113,9 +113,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe C.data(), ldc, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::herk, upper_lower, - trans, n, k, alpha, A.data(), lda, beta, C.data(), ldc, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::herk, + upper_lower, trans, n, k, alpha, A.data(), lda, beta, + C.data(), ldc, dependencies); break; default: break; } diff --git a/tests/unit_tests/blas/level3/symm.cpp b/tests/unit_tests/blas/level3/symm.cpp index 13267aee0..0668adbff 100644 --- a/tests/unit_tests/blas/level3/symm.cpp +++ b/tests/unit_tests/blas/level3/symm.cpp @@ -98,13 +98,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::symm(main_queue, left_right, upper_lower, m, n, - alpha, A_buffer, lda, B_buffer, ldb, beta, - C_buffer, ldc); + alpha, A_buffer, lda, B_buffer, ldb, beta, + C_buffer, ldc); break; case oneapi::math::layout::row_major: - oneapi::math::blas::row_major::symm(main_queue, left_right, upper_lower, m, n, alpha, - A_buffer, lda, B_buffer, ldb, beta, C_buffer, - ldc); + oneapi::math::blas::row_major::symm(main_queue, left_right, upper_lower, m, n, + alpha, A_buffer, lda, B_buffer, ldb, beta, + C_buffer, ldc); break; default: break; } @@ -173,14 +173,14 @@ TEST_P(SymmTests, RealDoublePrecision) { oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, - 102, 103, alpha, beta)); + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, + 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, - 102, 103, alpha, beta)); + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, + 101, 102, 103, alpha, beta)); } TEST_P(SymmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); @@ -188,33 +188,33 @@ TEST_P(SymmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(SymmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(SymmTestSuite, SymmTests, diff --git a/tests/unit_tests/blas/level3/symm_usm.cpp b/tests/unit_tests/blas/level3/symm_usm.cpp index 302dd1beb..dc8c00749 100644 --- a/tests/unit_tests/blas/level3/symm_usm.cpp +++ b/tests/unit_tests/blas/level3/symm_usm.cpp @@ -96,14 +96,14 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right #ifdef CALL_RT_API switch (layout) { case oneapi::math::layout::col_major: - done = oneapi::math::blas::column_major::symm(main_queue, left_right, upper_lower, m, - n, alpha, A.data(), lda, B.data(), ldb, - beta, C.data(), ldc, dependencies); + done = oneapi::math::blas::column_major::symm( + main_queue, left_right, upper_lower, m, n, alpha, A.data(), lda, B.data(), ldb, + beta, C.data(), ldc, dependencies); break; case oneapi::math::layout::row_major: - done = oneapi::math::blas::row_major::symm(main_queue, left_right, upper_lower, m, n, - alpha, A.data(), lda, B.data(), ldb, beta, - C.data(), ldc, dependencies); + done = oneapi::math::blas::row_major::symm(main_queue, left_right, upper_lower, m, + n, alpha, A.data(), lda, B.data(), ldb, + beta, C.data(), ldc, dependencies); break; default: break; } @@ -173,14 +173,14 @@ TEST_P(SymmUsmTests, RealDoublePrecision) { oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, 101, - 102, 103, alpha, beta)); + oneapi::math::side::right, oneapi::math::uplo::lower, 72, 27, + 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, 101, - 102, 103, alpha, beta)); + oneapi::math::side::right, oneapi::math::uplo::upper, 72, 27, + 101, 102, 103, alpha, beta)); } TEST_P(SymmUsmTests, ComplexSinglePrecision) { std::complex alpha(2.0, -0.5); @@ -188,33 +188,33 @@ TEST_P(SymmUsmTests, ComplexSinglePrecision) { EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } TEST_P(SymmUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); std::complex beta(3.0, -1.5); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, 72, 27, 101, 102, 103, alpha, beta)); } INSTANTIATE_TEST_SUITE_P(SymmUsmTestSuite, SymmUsmTests, diff --git a/tests/unit_tests/blas/level3/syr2k.cpp b/tests/unit_tests/blas/level3/syr2k.cpp index e735f0ace..0ecb2ad6b 100644 --- a/tests/unit_tests/blas/level3/syr2k.cpp +++ b/tests/unit_tests/blas/level3/syr2k.cpp @@ -49,7 +49,8 @@ namespace { template int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, fp beta) { + oneapi::math::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, + fp beta) { // Prepare data. vector> A, B, C, C_ref; rand_matrix(A, layout, trans, n, k, lda); @@ -94,13 +95,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::syr2k(main_queue, upper_lower, trans, n, k, alpha, - A_buffer, lda, B_buffer, ldb, beta, C_buffer, - ldc); + A_buffer, lda, B_buffer, ldb, beta, + C_buffer, ldc); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::syr2k(main_queue, upper_lower, trans, n, k, alpha, - A_buffer, lda, B_buffer, ldb, beta, C_buffer, - ldc); + A_buffer, lda, B_buffer, ldb, beta, C_buffer, + ldc); break; default: break; } @@ -141,8 +142,8 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe return (int)good; } -class Syr2kTests : public ::testing::TestWithParam> { -}; +class Syr2kTests + : public ::testing::TestWithParam> {}; TEST_P(Syr2kTests, RealSinglePrecision) { float alpha(3.0); @@ -172,11 +173,11 @@ TEST_P(Syr2kTests, RealDoublePrecision) { oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, - 101, 102, 103, alpha, beta)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, + 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, - 101, 102, 103, alpha, beta)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, + 27, 101, 102, 103, alpha, beta)); } TEST_P(Syr2kTests, ComplexSinglePrecision) { std::complex alpha(3.0, -0.5); diff --git a/tests/unit_tests/blas/level3/syr2k_usm.cpp b/tests/unit_tests/blas/level3/syr2k_usm.cpp index 9a5899c6c..fd51bda52 100644 --- a/tests/unit_tests/blas/level3/syr2k_usm.cpp +++ b/tests/unit_tests/blas/level3/syr2k_usm.cpp @@ -48,7 +48,8 @@ namespace { template int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lower, - oneapi::math::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, fp beta) { + oneapi::math::transpose trans, int n, int k, int lda, int ldb, int ldc, fp alpha, + fp beta) { // Catch asynchronous exceptions. auto exception_handler = [](exception_list exceptions) { for (std::exception_ptr const& e : exceptions) { @@ -94,13 +95,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::syr2k(main_queue, upper_lower, trans, n, k, - alpha, A.data(), lda, B.data(), ldb, - beta, C.data(), ldc, dependencies); + alpha, A.data(), lda, B.data(), ldb, + beta, C.data(), ldc, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::syr2k(main_queue, upper_lower, trans, n, k, - alpha, A.data(), lda, B.data(), ldb, - beta, C.data(), ldc, dependencies); + alpha, A.data(), lda, B.data(), ldb, + beta, C.data(), ldc, dependencies); break; default: break; } @@ -173,11 +174,11 @@ TEST_P(Syr2kUsmTests, RealDoublePrecision) { oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, - 101, 102, 103, alpha, beta)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, + 27, 101, 102, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, - 101, 102, 103, alpha, beta)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, + 27, 101, 102, 103, alpha, beta)); } TEST_P(Syr2kUsmTests, ComplexSinglePrecision) { std::complex alpha(3.0, -0.5); diff --git a/tests/unit_tests/blas/level3/syrk.cpp b/tests/unit_tests/blas/level3/syrk.cpp index c0c281531..928aaed84 100644 --- a/tests/unit_tests/blas/level3/syrk.cpp +++ b/tests/unit_tests/blas/level3/syrk.cpp @@ -92,11 +92,11 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::syrk(main_queue, upper_lower, trans, n, k, alpha, - A_buffer, lda, beta, C_buffer, ldc); + A_buffer, lda, beta, C_buffer, ldc); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::syrk(main_queue, upper_lower, trans, n, k, alpha, - A_buffer, lda, beta, C_buffer, ldc); + A_buffer, lda, beta, C_buffer, ldc); break; default: break; } @@ -108,8 +108,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe C_buffer, ldc); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syrk, upper_lower, - trans, n, k, alpha, A_buffer, lda, beta, C_buffer, ldc); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syrk, + upper_lower, trans, n, k, alpha, A_buffer, lda, beta, + C_buffer, ldc); break; default: break; } @@ -167,11 +168,11 @@ TEST_P(SyrkTests, RealDoublePrecision) { oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, - 101, 103, alpha, beta)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, + 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, - 101, 103, alpha, beta)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, + 27, 101, 103, alpha, beta)); } TEST_P(SyrkTests, ComplexSinglePrecision) { std::complex alpha(3.0, -0.5); diff --git a/tests/unit_tests/blas/level3/syrk_usm.cpp b/tests/unit_tests/blas/level3/syrk_usm.cpp index 6e912a9ac..2771dd2b7 100644 --- a/tests/unit_tests/blas/level3/syrk_usm.cpp +++ b/tests/unit_tests/blas/level3/syrk_usm.cpp @@ -92,13 +92,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe switch (layout) { case oneapi::math::layout::col_major: done = oneapi::math::blas::column_major::syrk(main_queue, upper_lower, trans, n, k, - alpha, A.data(), lda, beta, C.data(), - ldc, dependencies); + alpha, A.data(), lda, beta, C.data(), + ldc, dependencies); break; case oneapi::math::layout::row_major: done = oneapi::math::blas::row_major::syrk(main_queue, upper_lower, trans, n, k, - alpha, A.data(), lda, beta, C.data(), ldc, - dependencies); + alpha, A.data(), lda, beta, C.data(), + ldc, dependencies); break; default: break; } @@ -111,9 +111,9 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::uplo upper_lowe C.data(), ldc, dependencies); break; case oneapi::math::layout::row_major: - TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syrk, upper_lower, - trans, n, k, alpha, A.data(), lda, beta, C.data(), ldc, - dependencies); + TEST_RUN_BLAS_CT_SELECT(main_queue, oneapi::math::blas::row_major::syrk, + upper_lower, trans, n, k, alpha, A.data(), lda, beta, + C.data(), ldc, dependencies); break; default: break; } @@ -171,11 +171,11 @@ TEST_P(SyrkUsmTests, RealDoublePrecision) { oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, 73, 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, 27, - 101, 103, alpha, beta)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, 73, + 27, 101, 103, alpha, beta)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, 27, - 101, 103, alpha, beta)); + oneapi::math::uplo::upper, oneapi::math::transpose::trans, 73, + 27, 101, 103, alpha, beta)); } TEST_P(SyrkUsmTests, ComplexSinglePrecision) { std::complex alpha(3.0, -0.5); diff --git a/tests/unit_tests/blas/level3/trmm.cpp b/tests/unit_tests/blas/level3/trmm.cpp index 1dbaa4820..86d127cce 100644 --- a/tests/unit_tests/blas/level3/trmm.cpp +++ b/tests/unit_tests/blas/level3/trmm.cpp @@ -98,13 +98,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::trmm(main_queue, left_right, upper_lower, transa, - unit_nonunit, m, n, alpha, A_buffer, lda, - B_buffer, ldb); + unit_nonunit, m, n, alpha, A_buffer, lda, + B_buffer, ldb); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::trmm(main_queue, left_right, upper_lower, transa, - unit_nonunit, m, n, alpha, A_buffer, lda, - B_buffer, ldb); + unit_nonunit, m, n, alpha, A_buffer, lda, + B_buffer, ldb); break; default: break; } @@ -152,44 +152,44 @@ TEST_P(TrmmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); } TEST_P(TrmmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -205,20 +205,20 @@ TEST_P(TrmmTests, RealDoublePrecision) { 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, @@ -242,119 +242,119 @@ TEST_P(TrmmTests, ComplexSinglePrecision) { oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); } TEST_P(TrmmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); } INSTANTIATE_TEST_SUITE_P(TrmmTestSuite, TrmmTests, diff --git a/tests/unit_tests/blas/level3/trmm_usm.cpp b/tests/unit_tests/blas/level3/trmm_usm.cpp index 0e885b37e..20469d752 100644 --- a/tests/unit_tests/blas/level3/trmm_usm.cpp +++ b/tests/unit_tests/blas/level3/trmm_usm.cpp @@ -154,44 +154,44 @@ TEST_P(TrmmUsmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); } TEST_P(TrmmUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -207,20 +207,20 @@ TEST_P(TrmmUsmTests, RealDoublePrecision) { 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, @@ -244,119 +244,119 @@ TEST_P(TrmmUsmTests, ComplexSinglePrecision) { oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); } TEST_P(TrmmUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); } INSTANTIATE_TEST_SUITE_P(TrmmUsmTestSuite, TrmmUsmTests, diff --git a/tests/unit_tests/blas/level3/trsm.cpp b/tests/unit_tests/blas/level3/trsm.cpp index 0eac2ff36..c10c4c2a5 100644 --- a/tests/unit_tests/blas/level3/trsm.cpp +++ b/tests/unit_tests/blas/level3/trsm.cpp @@ -98,13 +98,13 @@ int test(device* dev, oneapi::math::layout layout, oneapi::math::side left_right switch (layout) { case oneapi::math::layout::col_major: oneapi::math::blas::column_major::trsm(main_queue, left_right, upper_lower, transa, - unit_nonunit, m, n, alpha, A_buffer, lda, - B_buffer, ldb); + unit_nonunit, m, n, alpha, A_buffer, lda, + B_buffer, ldb); break; case oneapi::math::layout::row_major: oneapi::math::blas::row_major::trsm(main_queue, left_right, upper_lower, transa, - unit_nonunit, m, n, alpha, A_buffer, lda, - B_buffer, ldb); + unit_nonunit, m, n, alpha, A_buffer, lda, + B_buffer, ldb); break; default: break; } @@ -152,20 +152,20 @@ TEST_P(TrsmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, @@ -184,36 +184,36 @@ TEST_P(TrsmTests, RealSinglePrecision) { 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); } TEST_P(TrsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -253,20 +253,20 @@ TEST_P(TrsmTests, RealDoublePrecision) { 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, @@ -290,199 +290,199 @@ TEST_P(TrsmTests, ComplexSinglePrecision) { oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, + 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, + 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); } TEST_P(TrsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); } INSTANTIATE_TEST_SUITE_P(TrsmTestSuite, TrsmTests, diff --git a/tests/unit_tests/blas/level3/trsm_usm.cpp b/tests/unit_tests/blas/level3/trsm_usm.cpp index 86b958d82..0f2247a14 100644 --- a/tests/unit_tests/blas/level3/trsm_usm.cpp +++ b/tests/unit_tests/blas/level3/trsm_usm.cpp @@ -155,20 +155,20 @@ TEST_P(TrsmUsmTests, RealSinglePrecision) { float alpha(2.0); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, @@ -187,36 +187,36 @@ TEST_P(TrsmUsmTests, RealSinglePrecision) { 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); } TEST_P(TrsmUsmTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); @@ -256,20 +256,20 @@ TEST_P(TrsmUsmTests, RealDoublePrecision) { 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, @@ -293,199 +293,199 @@ TEST_P(TrsmUsmTests, ComplexSinglePrecision) { oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, + 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, + 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>( - std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, - 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, - 27, 101, 102, alpha)); + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); } TEST_P(TrsmUsmTests, ComplexDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(std::get<0>(GetParam())); std::complex alpha(2.0, -0.5); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::nontrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::trans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::lower, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::left, oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); - EXPECT_TRUEORSKIP(test>(std::get<0>(GetParam()), std::get<1>(GetParam()), - oneapi::math::side::right, oneapi::math::uplo::upper, - oneapi::math::transpose::conjtrans, - oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, 27, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); EXPECT_TRUEORSKIP(test>( std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, - oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, 72, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::unit, 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::nontrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::trans, oneapi::math::diag::nonunit, 72, + 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::lower, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::left, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); + EXPECT_TRUEORSKIP(test>( + std::get<0>(GetParam()), std::get<1>(GetParam()), oneapi::math::side::right, + oneapi::math::uplo::upper, oneapi::math::transpose::conjtrans, oneapi::math::diag::nonunit, + 72, 27, 101, 102, alpha)); } INSTANTIATE_TEST_SUITE_P(TrsmUsmTestSuite, TrsmUsmTests, diff --git a/tests/unit_tests/dft/include/compute_inplace.hpp b/tests/unit_tests/dft/include/compute_inplace.hpp index a423476fe..bd11ed9d4 100644 --- a/tests/unit_tests/dft/include/compute_inplace.hpp +++ b/tests/unit_tests/dft/include/compute_inplace.hpp @@ -94,14 +94,17 @@ int DFT_Test::test_in_place_buffer() { auto acc_host = inout_buf.get_host_access(); auto ptr_host = reinterpret_cast(acc_host.get_pointer()); for (std::int64_t i = 0; i < batches; i++) { - EXPECT_TRUE(check_equal_strided( - ptr_host + backward_distance * i, out_host_ref.data() + ref_distance * i, sizes, - modified_strides_bwd, abs_error_margin, rel_error_margin, std::cout)); + EXPECT_TRUE(check_equal_strided < domain == + oneapi::math::dft::domain::REAL > + (ptr_host + backward_distance * i, + out_host_ref.data() + ref_distance * i, sizes, + modified_strides_bwd, abs_error_margin, rel_error_margin, + std::cout)); } } oneapi::math::dft::compute_backward, - FwdInputType>(descriptor, inout_buf); + FwdInputType>(descriptor, inout_buf); } std::vector fwd_data_ref = input; @@ -184,19 +187,21 @@ int DFT_Test::test_in_place_USM() { std::vector no_dependencies; oneapi::math::dft::compute_forward(descriptor, inout.data(), - no_dependencies) + no_dependencies) .wait_and_throw(); for (std::int64_t i = 0; i < batches; i++) { - EXPECT_TRUE(check_equal_strided( - reinterpret_cast(inout.data()) + backward_distance * i, - out_host_ref.data() + ref_distance * i, sizes, modified_strides_bwd, abs_error_margin, - rel_error_margin, std::cout)); + EXPECT_TRUE(check_equal_strided < domain == + oneapi::math::dft::domain::REAL > + (reinterpret_cast(inout.data()) + backward_distance * i, + out_host_ref.data() + ref_distance * i, sizes, modified_strides_bwd, + abs_error_margin, rel_error_margin, std::cout)); } sycl::event done = oneapi::math::dft::compute_backward, - FwdInputType>(descriptor, inout.data(), no_dependencies); + FwdInputType>(descriptor, inout.data(), + no_dependencies); done.wait_and_throw(); std::for_each(input.begin(), input.end(), diff --git a/tests/unit_tests/dft/include/compute_inplace_real_real.hpp b/tests/unit_tests/dft/include/compute_inplace_real_real.hpp index 0d9d726ee..263cc622f 100644 --- a/tests/unit_tests/dft/include/compute_inplace_real_real.hpp +++ b/tests/unit_tests/dft/include/compute_inplace_real_real.hpp @@ -66,8 +66,8 @@ int DFT_Test::test_in_place_real_real_USM() { abs_error_margin, rel_error_margin, std::cout)); oneapi::math::dft::compute_backward, - PrecisionType>(descriptor, inout_re.data(), - inout_im.data(), no_dependencies) + PrecisionType>(descriptor, inout_re.data(), + inout_im.data(), no_dependencies) .wait_and_throw(); for (std::size_t i = 0; i < output_data.size(); ++i) { @@ -118,7 +118,7 @@ int DFT_Test::test_in_place_real_real_buffer() { sycl::range<1>(size_total) }; oneapi::math::dft::compute_forward(descriptor, inout_re_buf, - inout_im_buf); + inout_im_buf); { auto acc_inout_re = inout_re_buf.get_host_access(); @@ -133,7 +133,7 @@ int DFT_Test::test_in_place_real_real_buffer() { } oneapi::math::dft::compute_backward, - PrecisionType>(descriptor, inout_re_buf, inout_im_buf); + PrecisionType>(descriptor, inout_re_buf, inout_im_buf); { auto acc_inout_re = inout_re_buf.get_host_access(); diff --git a/tests/unit_tests/dft/include/compute_out_of_place.hpp b/tests/unit_tests/dft/include/compute_out_of_place.hpp index 30b5cb5e1..284a68e80 100644 --- a/tests/unit_tests/dft/include/compute_out_of_place.hpp +++ b/tests/unit_tests/dft/include/compute_out_of_place.hpp @@ -77,20 +77,22 @@ int DFT_Test::test_out_of_place_buffer() { auto acc_bwd = bwd_buf.get_host_access(); auto bwd_ptr = acc_bwd.get_pointer(); for (std::int64_t i = 0; i < batches; i++) { - EXPECT_TRUE(check_equal_strided( - bwd_ptr + backward_distance * i, out_host_ref.data() + ref_distance * i, sizes, - strides_bwd_cpy, abs_error_margin, rel_error_margin, std::cout)); + EXPECT_TRUE(check_equal_strided < domain == + oneapi::math::dft::domain::REAL > + (bwd_ptr + backward_distance * i, + out_host_ref.data() + ref_distance * i, sizes, strides_bwd_cpy, + abs_error_margin, rel_error_margin, std::cout)); } } oneapi::math::dft::compute_backward, - FwdOutputType, FwdInputType>(descriptor, bwd_buf, - fwd_buf); + FwdOutputType, FwdInputType>(descriptor, bwd_buf, + fwd_buf); } // account for scaling that occurs during DFT std::for_each(input.begin(), input.end(), - [this](auto &x) { x *= static_cast(forward_elements); }); + [this](auto& x) { x *= static_cast(forward_elements); }); for (std::int64_t i = 0; i < batches; i++) { EXPECT_TRUE(check_equal_strided( @@ -164,19 +166,20 @@ int DFT_Test::test_out_of_place_USM() { auto bwd_ptr = &bwd[0]; for (std::int64_t i = 0; i < batches; i++) { - EXPECT_TRUE(check_equal_strided( - bwd_ptr + backward_distance * i, out_host_ref.data() + ref_distance * i, sizes, - strides_bwd_cpy, abs_error_margin, rel_error_margin, std::cout)); + EXPECT_TRUE(check_equal_strided < domain == + oneapi::math::dft::domain::REAL > + (bwd_ptr + backward_distance * i, out_host_ref.data() + ref_distance * i, + sizes, strides_bwd_cpy, abs_error_margin, rel_error_margin, std::cout)); } - oneapi::math::dft::compute_backward, FwdOutputType, - FwdInputType>(descriptor, bwd.data(), fwd.data(), - no_dependencies) + oneapi::math::dft::compute_backward, + FwdOutputType, FwdInputType>(descriptor, bwd.data(), + fwd.data(), no_dependencies) .wait_and_throw(); // account for scaling that occurs during DFT std::for_each(input.begin(), input.end(), - [this](auto &x) { x *= static_cast(forward_elements); }); + [this](auto& x) { x *= static_cast(forward_elements); }); for (std::int64_t i = 0; i < batches; i++) { EXPECT_TRUE(check_equal_strided( diff --git a/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp b/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp index b1af7e560..9f756f5cf 100644 --- a/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp +++ b/tests/unit_tests/dft/include/compute_out_of_place_real_real.hpp @@ -74,7 +74,7 @@ int DFT_Test::test_out_of_place_real_real_USM() { abs_error_margin, rel_error_margin, std::cout)); oneapi::math::dft::compute_backward, - PrecisionType, PrecisionType>( + PrecisionType, PrecisionType>( descriptor, out_re.data(), out_im.data(), out_back_re.data(), out_back_im.data(), no_dependencies) .wait_and_throw(); @@ -139,7 +139,7 @@ int DFT_Test::test_out_of_place_real_real_buffer() { } oneapi::math::dft::compute_backward, - PrecisionType, PrecisionType>( + PrecisionType, PrecisionType>( descriptor, out_dev_re, out_dev_im, out_back_dev_re, out_back_dev_im); { diff --git a/tests/unit_tests/dft/include/reference_dft.hpp b/tests/unit_tests/dft/include/reference_dft.hpp index 661829282..5c2abcc21 100644 --- a/tests/unit_tests/dft/include/reference_dft.hpp +++ b/tests/unit_tests/dft/include/reference_dft.hpp @@ -32,7 +32,7 @@ namespace detail { using ref_t = long double; /* Do the calculations using long double */ template -void reference_forward_dft_impl(const TypeIn *in, TypeOut *out, std::size_t N, std::size_t stride) { +void reference_forward_dft_impl(const TypeIn* in, TypeOut* out, std::size_t N, std::size_t stride) { static_assert(is_complex(), "Output type of DFT must be complex"); constexpr ref_t TWOPI = 2.0L * 3.141592653589793238462643383279502884197L; @@ -54,14 +54,14 @@ struct reference {}; template struct reference { - static void forward_dft(const std::vector &sizes, const TypeIn *in, TypeOut *out) { + static void forward_dft(const std::vector& sizes, const TypeIn* in, TypeOut* out) { reference_forward_dft_impl(in, out, sizes[0], 1); } }; template struct reference { - static void forward_dft(const std::vector &sizes, const TypeIn *in, TypeOut *out) { + static void forward_dft(const std::vector& sizes, const TypeIn* in, TypeOut* out) { const auto elements = std::accumulate(sizes.begin(), sizes.end(), 1U, std::multiplies<>{}); std::vector> tmp(elements); for (std::size_t i = 0; i < elements; i += sizes[1]) { @@ -75,7 +75,7 @@ struct reference { template struct reference { - static void forward_dft(const std::vector &sizes, const TypeIn *in, TypeOut *out) { + static void forward_dft(const std::vector& sizes, const TypeIn* in, TypeOut* out) { const auto elements = std::accumulate(sizes.begin(), sizes.end(), 1U, std::multiplies<>{}); std::vector> tmp1(elements); std::vector> tmp2(elements); @@ -112,7 +112,7 @@ struct reference { * @param stride the stride between elements in the data set, measured in elements. **/ template -void reference_forward_dft(const std::vector &sizes, const TypeIn *in, TypeOut *out) { +void reference_forward_dft(const std::vector& sizes, const TypeIn* in, TypeOut* out) { std::vector unsigned_sizes(sizes.size()); std::transform(sizes.begin(), sizes.end(), unsigned_sizes.begin(), [](std::int64_t size) { return cast_unsigned(size); }); diff --git a/tests/unit_tests/dft/include/test_common.hpp b/tests/unit_tests/dft/include/test_common.hpp index 70f60d363..5b1647e94 100644 --- a/tests/unit_tests/dft/include/test_common.hpp +++ b/tests/unit_tests/dft/include/test_common.hpp @@ -58,7 +58,7 @@ inline std::size_t cast_unsigned(std::int64_t i) { } template -bool check_equal(fp x, fp x_ref, double abs_error_mag, double rel_error_mag, std::ostream &out) { +bool check_equal(fp x, fp x_ref, double abs_error_mag, double rel_error_mag, std::ostream& out) { using fp_real = typename complex_info::real_type; static_assert(std::is_floating_point_v, "Expected floating-point real or complex type."); @@ -88,8 +88,8 @@ bool check_equal(fp x, fp x_ref, double abs_error_mag, double rel_error_mag, std } template -bool check_equal_vector(vec1 &&v, vec2 &&v_ref, std::size_t n, double abs_error_mag, - double rel_error_mag, std::ostream &out) { +bool check_equal_vector(vec1&& v, vec2&& v_ref, std::size_t n, double abs_error_mag, + double rel_error_mag, std::ostream& out) { constexpr int max_print = 20; int count = 0; bool good = true; @@ -131,7 +131,7 @@ inline t rand_scalar() { } template -void rand_vector(vec &v, std::size_t n) { +void rand_vector(vec& v, std::size_t n) { using fp = typename vec::value_type; v.resize(n); for (std::size_t i = 0; i < n; i++) { @@ -141,7 +141,7 @@ void rand_vector(vec &v, std::size_t n) { // Catch asynchronous exceptions. auto exception_handler = [](sycl::exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } @@ -153,7 +153,7 @@ auto exception_handler = [](sycl::exception_list exceptions) { }; template -void commit_descriptor(oneapi::math::dft::descriptor &descriptor, +void commit_descriptor(oneapi::math::dft::descriptor& descriptor, sycl::queue queue) { #ifdef CALL_RT_API descriptor.commit(queue); @@ -164,7 +164,7 @@ void commit_descriptor(oneapi::math::dft::descriptor &descrip // is it assumed that the unused elements of the array are ignored inline std::array get_conjugate_even_complex_strides( - const std::vector &sizes) { + const std::vector& sizes) { switch (sizes.size()) { case 1: return { 0, 1 }; case 2: return { 0, sizes[1] / 2 + 1, 1 }; @@ -178,7 +178,7 @@ inline std::array get_conjugate_even_complex_strides( } // is it assumed that the unused elements of the array are ignored -inline std::array get_default_strides(const std::vector &sizes) { +inline std::array get_default_strides(const std::vector& sizes) { if (sizes.size() > 3) { throw oneapi::math::unimplemented( "dft/test_common", __FUNCTION__, @@ -207,8 +207,8 @@ T get_default(const std::vector vec, std::size_t idx, T default_) { template std::pair get_default_distances( - const std::vector &sizes, const std::vector &strides_fwd, - const std::vector &strides_bwd) { + const std::vector& sizes, const std::vector& strides_fwd, + const std::vector& strides_bwd) { std::int64_t size0 = sizes[0]; std::int64_t size1 = get_default(sizes, 1, 1l); std::int64_t size2 = get_default(sizes, 2, 1l); @@ -241,8 +241,8 @@ std::pair get_default_distances( //up to 3 dimensions, empty strides = default template > std::vector strided_copy( - const T_vec &contiguous, const std::vector &sizes, - const std::vector &strides, std::int64_t batches, std::int64_t distance, + const T_vec& contiguous, const std::vector& sizes, + const std::vector& strides, std::int64_t batches, std::int64_t distance, Allocator alloc = {}) { if (strides.size() == 0) { return { contiguous.begin(), contiguous.end(), alloc }; @@ -273,9 +273,9 @@ std::vector strided_copy( //up to 3 dimensions, empty strides = default template -bool check_equal_strided(const vec1 &v, const vec2 &v_ref, std::vector sizes, +bool check_equal_strided(const vec1& v, const vec2& v_ref, std::vector sizes, std::vector strides, double abs_error_mag, double rel_error_mag, - std::ostream &out) { + std::ostream& out) { if (strides.size() == 0) { std::array strides_arr; if constexpr (ConjugateEvenStrides) { @@ -344,8 +344,7 @@ struct DFTParams { class DFTParamsPrint { public: - std::string operator()( - testing::TestParamInfo> dev) const { + std::string operator()(testing::TestParamInfo> dev) const { auto [device, params] = dev.param; std::string info_name; @@ -377,7 +376,7 @@ class DFTParamsPrint { info_name.append("_batches_").append(std::to_string(params.batches)); std::string dev_name = device->get_info(); - std::for_each(dev_name.begin(), dev_name.end(), [](auto &c) { + std::for_each(dev_name.begin(), dev_name.end(), [](auto& c) { if (!isalnum(c)) c = '_'; }); diff --git a/tests/unit_tests/dft/source/compute_tests.cpp b/tests/unit_tests/dft/source/compute_tests.cpp index 6375a5e50..349ba2bd7 100644 --- a/tests/unit_tests/dft/source/compute_tests.cpp +++ b/tests/unit_tests/dft/source/compute_tests.cpp @@ -35,41 +35,41 @@ #include "compute_out_of_place.hpp" #include "compute_out_of_place_real_real.hpp" -extern std::vector devices; +extern std::vector devices; namespace { class ComputeTests_in_place_COMPLEX - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; class ComputeTests_real_real_in_place_COMPLEX - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; class ComputeTests_out_of_place_COMPLEX - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; class ComputeTests_real_real_out_of_place_COMPLEX - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; class ComputeTests_in_place_REAL - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; class ComputeTests_real_real_in_place_REAL - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; class ComputeTests_out_of_place_REAL - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; class ComputeTests_real_real_out_of_place_REAL - : public ::testing::TestWithParam> {}; + : public ::testing::TestWithParam> {}; #define INSTANTIATE_TEST(PRECISION, DOMAIN, PLACE, LAYOUT, STORAGE) \ TEST_P(ComputeTests##_##LAYOUT##PLACE##_##DOMAIN, \ DOMAIN##_##PRECISION##_##PLACE##_##LAYOUT##STORAGE) { \ try { \ - auto test = DFT_Test{ \ + auto test = DFT_Test{ \ std::get<0>(GetParam()), std::get<1>(GetParam()).sizes, \ std::get<1>(GetParam()).strides_fwd, std::get<1>(GetParam()).strides_bwd, \ std::get<1>(GetParam()).batches \ }; \ EXPECT_TRUEORSKIP(test.test_##PLACE##_##LAYOUT##STORAGE()); \ } \ - catch (oneapi::math::unimplemented & e) { \ + catch (oneapi::math::unimplemented & e) { \ std::cout << "Skipping test because: \"" << e.what() << "\"" << std::endl; \ GTEST_SKIP(); \ } \ diff --git a/tests/unit_tests/dft/source/descriptor_tests.cpp b/tests/unit_tests/dft/source/descriptor_tests.cpp index da1f5c487..a4290e553 100644 --- a/tests/unit_tests/dft/source/descriptor_tests.cpp +++ b/tests/unit_tests/dft/source/descriptor_tests.cpp @@ -141,11 +141,14 @@ static void set_and_get_io_strides() { descriptor.get_value(oneapi::math::dft::config_param::INPUT_STRIDES, input_strides_before_set.data()); EXPECT_EQ(std::vector(strides_size, 0), input_strides_before_set); - descriptor.set_value(oneapi::math::dft::config_param::INPUT_STRIDES, input_strides_value.data()); + descriptor.set_value(oneapi::math::dft::config_param::INPUT_STRIDES, + input_strides_value.data()); descriptor.get_value(oneapi::math::dft::config_param::INPUT_STRIDES, input_strides_after_set.data()); - descriptor.get_value(oneapi::math::dft::config_param::FWD_STRIDES, fwd_strides_after_set.data()); - descriptor.get_value(oneapi::math::dft::config_param::BWD_STRIDES, bwd_strides_after_set.data()); + descriptor.get_value(oneapi::math::dft::config_param::FWD_STRIDES, + fwd_strides_after_set.data()); + descriptor.get_value(oneapi::math::dft::config_param::BWD_STRIDES, + bwd_strides_after_set.data()); EXPECT_EQ(input_strides_value, input_strides_after_set); EXPECT_EQ(std::vector(strides_size, 0), fwd_strides_after_set); EXPECT_EQ(std::vector(strides_size, 0), bwd_strides_after_set); @@ -206,8 +209,10 @@ static void set_and_get_fwd_bwd_strides() { descriptor.get_value(oneapi::math::dft::config_param::FWD_STRIDES, fwd_strides_before_set.data()); EXPECT_EQ(fwd_strides_default_value, fwd_strides_before_set); - descriptor.set_value(oneapi::math::dft::config_param::FWD_STRIDES, fwd_strides_new_value.data()); - descriptor.get_value(oneapi::math::dft::config_param::FWD_STRIDES, fwd_strides_after_set.data()); + descriptor.set_value(oneapi::math::dft::config_param::FWD_STRIDES, + fwd_strides_new_value.data()); + descriptor.get_value(oneapi::math::dft::config_param::FWD_STRIDES, + fwd_strides_after_set.data()); descriptor.get_value(oneapi::math::dft::config_param::INPUT_STRIDES, input_strides_after_set.data()); descriptor.get_value(oneapi::math::dft::config_param::OUTPUT_STRIDES, @@ -221,8 +226,10 @@ static void set_and_get_fwd_bwd_strides() { descriptor.get_value(oneapi::math::dft::config_param::BWD_STRIDES, bwd_strides_before_set.data()); EXPECT_EQ(bwd_strides_default_value, bwd_strides_before_set); - descriptor.set_value(oneapi::math::dft::config_param::BWD_STRIDES, bwd_strides_new_value.data()); - descriptor.get_value(oneapi::math::dft::config_param::BWD_STRIDES, bwd_strides_after_set.data()); + descriptor.set_value(oneapi::math::dft::config_param::BWD_STRIDES, + bwd_strides_new_value.data()); + descriptor.get_value(oneapi::math::dft::config_param::BWD_STRIDES, + bwd_strides_after_set.data()); EXPECT_EQ(bwd_strides_new_value, bwd_strides_after_set); } #pragma clang diagnostic pop @@ -289,7 +296,8 @@ static void set_and_get_values() { &fwd_distance_before_set); EXPECT_EQ(1, fwd_distance_before_set); descriptor.set_value(oneapi::math::dft::config_param::FWD_DISTANCE, fwd_distance_set_value); - descriptor.get_value(oneapi::math::dft::config_param::FWD_DISTANCE, &fwd_distance_after_set); + descriptor.get_value(oneapi::math::dft::config_param::FWD_DISTANCE, + &fwd_distance_after_set); EXPECT_EQ(fwd_distance_set_value, fwd_distance_after_set); std::int64_t bwd_distance_set_value{ domain == oneapi::math::dft::domain::REAL @@ -302,7 +310,8 @@ static void set_and_get_values() { &bwd_distance_before_set); EXPECT_EQ(1, bwd_distance_before_set); descriptor.set_value(oneapi::math::dft::config_param::BWD_DISTANCE, bwd_distance_set_value); - descriptor.get_value(oneapi::math::dft::config_param::BWD_DISTANCE, &bwd_distance_after_set); + descriptor.get_value(oneapi::math::dft::config_param::BWD_DISTANCE, + &bwd_distance_after_set); EXPECT_EQ(bwd_distance_set_value, bwd_distance_after_set); } @@ -717,11 +726,13 @@ int test_commit(sycl::device* dev) { } TEST(DescriptorTests, DescriptorMoveRealSingle) { - EXPECT_TRUE((test_move())); + EXPECT_TRUE( + (test_move())); } TEST(DescriptorTests, DescriptorMoveRealDouble) { - EXPECT_TRUE((test_move())); + EXPECT_TRUE( + (test_move())); } TEST(DescriptorTests, DescriptorMoveComplexSingle) { @@ -735,13 +746,13 @@ TEST(DescriptorTests, DescriptorMoveComplexDouble) { } TEST(DescriptorTests, DescriptorTestsRealSingle) { - EXPECT_TRUE(( - test_getter_setter())); + EXPECT_TRUE((test_getter_setter())); } TEST(DescriptorTests, DescriptorTestsRealDouble) { - EXPECT_TRUE(( - test_getter_setter())); + EXPECT_TRUE((test_getter_setter())); } TEST(DescriptorTests, DescriptorTestsComplexSingle) { diff --git a/tests/unit_tests/include/test_helper.hpp b/tests/unit_tests/include/test_helper.hpp index 60702adf5..9a34b3f9f 100644 --- a/tests/unit_tests/include/test_helper.hpp +++ b/tests/unit_tests/include/test_helper.hpp @@ -271,11 +271,11 @@ } \ } while (0); -void print_error_code(sycl::exception const &e); +void print_error_code(sycl::exception const& e); class DeviceNamePrint { public: - std::string operator()(testing::TestParamInfo dev) const { + std::string operator()(testing::TestParamInfo dev) const { std::string dev_name = dev.param->get_info(); for (std::string::size_type i = 0; i < dev_name.size(); ++i) { if (!isalnum(dev_name[i])) @@ -290,9 +290,10 @@ class DeviceNamePrint { class LayoutDeviceNamePrint { public: std::string operator()( - testing::TestParamInfo> dev) const { - std::string layout_name = - std::get<1>(dev.param) == oneapi::math::layout::col_major ? "Column_Major" : "Row_Major"; + testing::TestParamInfo> dev) const { + std::string layout_name = std::get<1>(dev.param) == oneapi::math::layout::col_major + ? "Column_Major" + : "Row_Major"; std::string dev_name = std::get<0>(dev.param)->get_info(); for (std::string::size_type i = 0; i < dev_name.size(); ++i) { if (!isalnum(dev_name[i])) @@ -308,7 +309,7 @@ class LayoutDeviceNamePrint { namespace oneapi { namespace math { -static inline void *aligned_alloc(size_t align, size_t size) { +static inline void* aligned_alloc(size_t align, size_t size) { #ifdef _WIN64 return ::_aligned_malloc(size, align); #else @@ -316,7 +317,7 @@ static inline void *aligned_alloc(size_t align, size_t size) { #endif } -static inline void aligned_free(void *p) { +static inline void aligned_free(void* p) { #ifdef _WIN64 ::_aligned_free(p); #else @@ -325,7 +326,7 @@ static inline void aligned_free(void *p) { } /* Support for Unified Shared Memory allocations for different backends */ -static inline void *malloc_shared(size_t align, size_t size, sycl::device dev, sycl::context ctx) { +static inline void* malloc_shared(size_t align, size_t size, sycl::device dev, sycl::context ctx) { (void)align; #ifdef _WIN64 return sycl::malloc_shared(size, dev, ctx); @@ -339,7 +340,7 @@ static inline void *malloc_shared(size_t align, size_t size, sycl::device dev, s #endif } -static inline void *malloc_device(size_t align, size_t size, sycl::device dev, sycl::context ctx) { +static inline void* malloc_device(size_t align, size_t size, sycl::device dev, sycl::context ctx) { (void)align; #ifdef _WIN64 return sycl::malloc_device(size, dev, ctx); @@ -353,11 +354,11 @@ static inline void *malloc_device(size_t align, size_t size, sycl::device dev, s #endif } -static inline void free_shared(void *p, sycl::context ctx) { +static inline void free_shared(void* p, sycl::context ctx) { sycl::free(p, ctx); } -static inline void free_usm(void *p, sycl::context ctx) { +static inline void free_usm(void* p, sycl::context ctx) { sycl::free(p, ctx); } diff --git a/tests/unit_tests/lapack/include/lapack_accuracy_checks.hpp b/tests/unit_tests/lapack/include/lapack_accuracy_checks.hpp index 8cf2ade94..c379e5d25 100644 --- a/tests/unit_tests/lapack/include/lapack_accuracy_checks.hpp +++ b/tests/unit_tests/lapack/include/lapack_accuracy_checks.hpp @@ -206,8 +206,9 @@ bool check_gerqf_accuracy(const std::vector& A, const std::vector& A_ini std::vector tau2(n); for (int64_t i = 0; i < std::min(m, n); i++) tau2[n - m + i] = tau[i]; - auto info = reference::or_un_mrq(oneapi::math::side::right, oneapi::math::transpose::nontrans, - m, n, n, Q.data(), ldq, tau2.data(), R.data(), ldr); + auto info = + reference::or_un_mrq(oneapi::math::side::right, oneapi::math::transpose::nontrans, m, n, + n, Q.data(), ldq, tau2.data(), R.data(), ldr); if (0 != info) { test_log::lout << "reference ormqr/unmqr failed with info = " << info << std::endl; return false; @@ -395,8 +396,9 @@ bool check_or_un_gbr_accuracy(oneapi::math::generate vect, int64_t m, int64_t n, /* | I - Q'Q | < m O(eps) */ std::vector QQ(cols_Q * cols_Q); int64_t ldqq = cols_Q; - reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, cols_Q, - cols_Q, rows_Q, 1.0, Q.data(), ldq, Q.data(), ldq, 0.0, QQ.data(), ldqq); + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, + cols_Q, cols_Q, rows_Q, 1.0, Q.data(), ldq, Q.data(), ldq, 0.0, QQ.data(), + ldqq); if (!rel_id_err_check(cols_Q, QQ, ldqq)) { test_log::lout << "Q Orthogonality check failed" << std::endl; result = false; @@ -411,8 +413,9 @@ bool check_or_un_gbr_accuracy(oneapi::math::generate vect, int64_t m, int64_t n, /* | I - (P')(P')' | < m O(eps) */ std::vector PP(rows_P * rows_P); int64_t ldpp = rows_P; - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, rows_P, - rows_P, cols_P, 1.0, P.data(), ldp, P.data(), ldp, 0.0, PP.data(), ldpp); + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, + rows_P, rows_P, cols_P, 1.0, P.data(), ldp, P.data(), ldp, 0.0, PP.data(), + ldpp); if (!rel_id_err_check(rows_P, PP, ldpp)) { test_log::lout << "P^t Orthogonality check failed" << std::endl; result = false; @@ -490,15 +493,15 @@ bool check_potrf_accuracy(const std::vector& init, const std::vector& so } template -bool check_potrs_accuracy(oneapi::math::uplo uplo, int64_t n, int64_t nrhs, const std::vector& B, - int64_t ldb, std::vector A_initial, int64_t lda, - std::vector B_initial) { +bool check_potrs_accuracy(oneapi::math::uplo uplo, int64_t n, int64_t nrhs, + const std::vector& B, int64_t ldb, std::vector A_initial, + int64_t lda, std::vector B_initial) { using fp_real = typename complex_info::real_type; hermitian_to_full(uplo, n, A_initial, lda); // Compute A*X - B. Store result in B_initial - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, nrhs, n, - -1.0, A_initial.data(), lda, B.data(), ldb, 1.0, B_initial.data(), ldb); + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, n, nrhs, + n, -1.0, A_initial.data(), lda, B.data(), ldb, 1.0, B_initial.data(), ldb); // Compute norm residual |A*X - B| fp_real norm_residual = reference::lange('1', n, nrhs, B_initial.data(), ldb); diff --git a/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp b/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp index 37fe92dad..9e28e9feb 100644 --- a/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp +++ b/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp @@ -62,7 +62,7 @@ inline CBLAS_DIAG cblas_diag(oneapi::math::diag d) { return CblasUnit; return CblasNonUnit; } -inline CBLAS_SIDE cblas_side(const char *c) { +inline CBLAS_SIDE cblas_side(const char* c) { return *c == 'R' || *c == 'r' ? CblasRight : CblasLeft; } inline CBLAS_SIDE cblas_side(oneapi::math::side s) { @@ -149,143 +149,143 @@ inline char to_char(oneapi::math::generate v) { return 'Q'; } -inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, int64_t n, - int64_t k, float alpha, const float *a, int64_t lda, const float *b, int64_t ldb, - float beta, float *c, int64_t ldc) { +inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, + int64_t n, int64_t k, float alpha, const float* a, int64_t lda, const float* b, + int64_t ldb, float beta, float* c, int64_t ldc) { cblas_sgemm(CblasColMajor, cblas_trans(transa), cblas_trans(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, int64_t n, - int64_t k, double alpha, const double *a, int64_t lda, const double *b, - int64_t ldb, double beta, double *c, int64_t ldc) { +inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, + int64_t n, int64_t k, double alpha, const double* a, int64_t lda, const double* b, + int64_t ldb, double beta, double* c, int64_t ldc) { cblas_dgemm(CblasColMajor, cblas_trans(transa), cblas_trans(transb), m, n, k, alpha, a, lda, b, ldb, beta, c, ldc); } -inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc) { - cblas_cgemm(CblasColMajor, cblas_trans(transa), cblas_trans(transb), m, n, k, (void *)&alpha, - (void *)a, lda, (void *)(b), ldb, (void *)&beta, (void *)c, ldc); +inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc) { + cblas_cgemm(CblasColMajor, cblas_trans(transa), cblas_trans(transb), m, n, k, (void*)&alpha, + (void*)a, lda, (void*)(b), ldb, (void*)&beta, (void*)c, ldc); } -inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, int64_t n, - int64_t k, std::complex alpha, const std::complex *a, int64_t lda, - const std::complex *b, int64_t ldb, std::complex beta, - std::complex *c, int64_t ldc) { - cblas_zgemm(CblasColMajor, cblas_trans(transa), cblas_trans(transb), m, n, k, (void *)&alpha, - (void *)a, lda, (void *)(b), ldb, (void *)&beta, (void *)c, ldc); +inline void gemm(oneapi::math::transpose transa, oneapi::math::transpose transb, int64_t m, + int64_t n, int64_t k, std::complex alpha, const std::complex* a, + int64_t lda, const std::complex* b, int64_t ldb, std::complex beta, + std::complex* c, int64_t ldc) { + cblas_zgemm(CblasColMajor, cblas_trans(transa), cblas_trans(transb), m, n, k, (void*)&alpha, + (void*)a, lda, (void*)(b), ldb, (void*)&beta, (void*)c, ldc); } -inline int64_t syevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, float *a, int64_t lda, - float *w) { +inline int64_t syevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, float* a, int64_t lda, + float* w) { return LAPACKE_ssyevd(LAPACK_COL_MAJOR, to_char(j), to_char(u), n, a, lda, w); } -inline int64_t syevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, double *a, int64_t lda, - double *w) { +inline int64_t syevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, double* a, int64_t lda, + double* w) { return LAPACKE_dsyevd(LAPACK_COL_MAJOR, to_char(j), to_char(u), n, a, lda, w); } -inline int64_t sygvd(int64_t itype, oneapi::math::job j, oneapi::math::uplo u, int64_t n, float *a, - int64_t lda, float *b, int64_t ldb, float *w) { +inline int64_t sygvd(int64_t itype, oneapi::math::job j, oneapi::math::uplo u, int64_t n, float* a, + int64_t lda, float* b, int64_t ldb, float* w) { return LAPACKE_ssygvd(LAPACK_COL_MAJOR, itype, to_char(j), to_char(u), n, a, lda, b, ldb, w); } -inline int64_t sygvd(int64_t itype, oneapi::math::job j, oneapi::math::uplo u, int64_t n, double *a, - int64_t lda, double *b, int64_t ldb, double *w) { +inline int64_t sygvd(int64_t itype, oneapi::math::job j, oneapi::math::uplo u, int64_t n, double* a, + int64_t lda, double* b, int64_t ldb, double* w) { return LAPACKE_dsygvd(LAPACK_COL_MAJOR, itype, to_char(j), to_char(u), n, a, lda, b, ldb, w); } inline void syrk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, float alpha, - const float *a, int64_t lda, float beta, float *c, int64_t ldc) { + const float* a, int64_t lda, float beta, float* c, int64_t ldc) { cblas_ssyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, c, ldc); } -inline void syrk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, double alpha, - const double *a, int64_t lda, double beta, double *c, int64_t ldc) { +inline void syrk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, + double alpha, const double* a, int64_t lda, double beta, double* c, int64_t ldc) { cblas_dsyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, c, ldc); } inline void syrk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, std::complex *c, int64_t ldc) { - cblas_csyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, (void *)&alpha, a, lda, - (void *)&beta, (void *)c, ldc); + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, std::complex* c, int64_t ldc) { + cblas_csyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, (void*)&alpha, a, lda, + (void*)&beta, (void*)c, ldc); } inline void syrk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, - std::complex alpha, const std::complex *a, int64_t lda, - std::complex beta, std::complex *c, int64_t ldc) { - cblas_zsyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, (void *)&alpha, a, lda, - (void *)&beta, (void *)c, ldc); + std::complex alpha, const std::complex* a, int64_t lda, + std::complex beta, std::complex* c, int64_t ldc) { + cblas_zsyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, (void*)&alpha, a, lda, + (void*)&beta, (void*)c, ldc); } inline void herk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, float alpha, - const std::complex *a, int64_t lda, float beta, std::complex *c, + const std::complex* a, int64_t lda, float beta, std::complex* c, int64_t ldc) { - cblas_cherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void *)c, + cblas_cherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void*)c, ldc); } -inline void herk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, double alpha, - const std::complex *a, int64_t lda, double beta, std::complex *c, - int64_t ldc) { - cblas_zherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void *)c, +inline void herk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, + double alpha, const std::complex* a, int64_t lda, double beta, + std::complex* c, int64_t ldc) { + cblas_zherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void*)c, ldc); } inline void sy_he_rk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, - float alpha, const float *a, int64_t lda, float beta, float *c, int64_t ldc) { + float alpha, const float* a, int64_t lda, float beta, float* c, int64_t ldc) { cblas_ssyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, c, ldc); } inline void sy_he_rk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, - double alpha, const double *a, int64_t lda, double beta, double *c, + double alpha, const double* a, int64_t lda, double beta, double* c, int64_t ldc) { cblas_dsyrk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, c, ldc); } inline void sy_he_rk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, - float alpha, const std::complex *a, int64_t lda, float beta, - std::complex *c, int64_t ldc) { - cblas_cherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void *)c, + float alpha, const std::complex* a, int64_t lda, float beta, + std::complex* c, int64_t ldc) { + cblas_cherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void*)c, ldc); } inline void sy_he_rk(oneapi::math::uplo u, oneapi::math::transpose t, int64_t n, int64_t k, - double alpha, const std::complex *a, int64_t lda, double beta, - std::complex *c, int64_t ldc) { - cblas_zherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void *)c, + double alpha, const std::complex* a, int64_t lda, double beta, + std::complex* c, int64_t ldc) { + cblas_zherk(CblasColMajor, cblas_uplo(u), cblas_trans(t), n, k, alpha, a, lda, beta, (void*)c, ldc); } inline void trmm(oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose transa, - oneapi::math::diag diag, int64_t m, int64_t n, float alpha, const float *a, - int64_t lda, float *b, int64_t ldb) { + oneapi::math::diag diag, int64_t m, int64_t n, float alpha, const float* a, + int64_t lda, float* b, int64_t ldb) { cblas_strmm(CblasColMajor, cblas_side(side), cblas_uplo(uplo), cblas_trans(transa), cblas_diag(diag), m, n, alpha, a, lda, b, ldb); } inline void trmm(oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose transa, - oneapi::math::diag diag, int64_t m, int64_t n, double alpha, const double *a, - int64_t lda, double *b, int64_t ldb) { + oneapi::math::diag diag, int64_t m, int64_t n, double alpha, const double* a, + int64_t lda, double* b, int64_t ldb) { cblas_dtrmm(CblasColMajor, cblas_side(side), cblas_uplo(uplo), cblas_trans(transa), cblas_diag(diag), m, n, alpha, a, lda, b, ldb); } inline void trmm(oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose transa, oneapi::math::diag diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb) { + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb) { cblas_ctrmm(CblasColMajor, cblas_side(side), cblas_uplo(uplo), cblas_trans(transa), - cblas_diag(diag), m, n, (void *)&alpha, (void *)(a), lda, (void *)(b), ldb); + cblas_diag(diag), m, n, (void*)&alpha, (void*)(a), lda, (void*)(b), ldb); } inline void trmm(oneapi::math::side side, oneapi::math::uplo uplo, oneapi::math::transpose transa, oneapi::math::diag diag, int64_t m, int64_t n, std::complex alpha, - const std::complex *a, int64_t lda, std::complex *b, int64_t ldb) { + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb) { cblas_ztrmm(CblasColMajor, cblas_side(side), cblas_uplo(uplo), cblas_trans(transa), - cblas_diag(diag), m, n, (void *)&alpha, (void *)(a), lda, (void *)(b), ldb); + cblas_diag(diag), m, n, (void*)&alpha, (void*)(a), lda, (void*)(b), ldb); } -inline void swap(int64_t n, float *X, int64_t incX, float *Y, int64_t incY) { +inline void swap(int64_t n, float* X, int64_t incX, float* Y, int64_t incY) { cblas_sswap(n, X, incX, Y, incY); } -inline void swap(int64_t n, double *X, int64_t incX, double *Y, int64_t incY) { +inline void swap(int64_t n, double* X, int64_t incX, double* Y, int64_t incY) { cblas_dswap(n, X, incX, Y, incY); } -inline void swap(int64_t n, std::complex *X, int64_t incX, std::complex *Y, +inline void swap(int64_t n, std::complex* X, int64_t incX, std::complex* Y, int64_t incY) { - cblas_cswap(n, (void *)X, incX, (void *)Y, incY); + cblas_cswap(n, (void*)X, incX, (void*)Y, incY); } -inline void swap(int64_t n, std::complex *X, int64_t incX, std::complex *Y, +inline void swap(int64_t n, std::complex* X, int64_t incX, std::complex* Y, int64_t incY) { - cblas_zswap(n, (void *)X, incX, (void *)Y, incY); + cblas_zswap(n, (void*)X, incX, (void*)Y, incY); } template @@ -299,608 +299,611 @@ inline double lamch(char cmach) { return LAPACKE_dlamch(cmach); } -inline float lange(char norm, int64_t m, int64_t n, const std::complex *a, int64_t lda) { +inline float lange(char norm, int64_t m, int64_t n, const std::complex* a, int64_t lda) { return LAPACKE_clange(LAPACK_COL_MAJOR, norm, m, n, - reinterpret_cast(a), lda); + reinterpret_cast(a), lda); } -inline double lange(char norm, int64_t m, int64_t n, const double *a, int64_t lda) { +inline double lange(char norm, int64_t m, int64_t n, const double* a, int64_t lda) { return LAPACKE_dlange(LAPACK_COL_MAJOR, norm, m, n, a, lda); } -inline float lange(char norm, int64_t m, int64_t n, const float *a, int64_t lda) { +inline float lange(char norm, int64_t m, int64_t n, const float* a, int64_t lda) { return LAPACKE_slange(LAPACK_COL_MAJOR, norm, m, n, a, lda); } -inline double lange(char norm, int64_t m, int64_t n, const std::complex *a, int64_t lda) { +inline double lange(char norm, int64_t m, int64_t n, const std::complex* a, int64_t lda) { return LAPACKE_zlange(LAPACK_COL_MAJOR, norm, m, n, - reinterpret_cast(a), lda); + reinterpret_cast(a), lda); } -inline float lanhe(char norm, oneapi::math::uplo u, int64_t n, const std::complex *a, +inline float lanhe(char norm, oneapi::math::uplo u, int64_t n, const std::complex* a, int64_t lda) { return LAPACKE_clanhe(LAPACK_COL_MAJOR, norm, to_char(u), n, - reinterpret_cast(a), lda); + reinterpret_cast(a), lda); } -inline double lanhe(char norm, oneapi::math::uplo u, int64_t n, const std::complex *a, +inline double lanhe(char norm, oneapi::math::uplo u, int64_t n, const std::complex* a, int64_t lda) { return LAPACKE_zlanhe(LAPACK_COL_MAJOR, norm, to_char(u), n, - reinterpret_cast(a), lda); + reinterpret_cast(a), lda); } -inline float lansy(char norm, oneapi::math::uplo u, int64_t n, const std::complex *a, +inline float lansy(char norm, oneapi::math::uplo u, int64_t n, const std::complex* a, int64_t lda) { return LAPACKE_clansy(LAPACK_COL_MAJOR, norm, to_char(u), n, - reinterpret_cast(a), lda); + reinterpret_cast(a), lda); } -inline double lansy(char norm, oneapi::math::uplo u, int64_t n, const double *a, int64_t lda) { +inline double lansy(char norm, oneapi::math::uplo u, int64_t n, const double* a, int64_t lda) { return LAPACKE_dlansy(LAPACK_COL_MAJOR, norm, to_char(u), n, a, lda); } -inline float lansy(char norm, oneapi::math::uplo u, int64_t n, const float *a, int64_t lda) { +inline float lansy(char norm, oneapi::math::uplo u, int64_t n, const float* a, int64_t lda) { return LAPACKE_slansy(LAPACK_COL_MAJOR, norm, to_char(u), n, a, lda); } -inline double lansy(char norm, oneapi::math::uplo u, int64_t n, const std::complex *a, +inline double lansy(char norm, oneapi::math::uplo u, int64_t n, const std::complex* a, int64_t lda) { return LAPACKE_zlansy(LAPACK_COL_MAJOR, norm, to_char(u), n, - reinterpret_cast(a), lda); + reinterpret_cast(a), lda); } -inline int64_t lacpy(char u, int64_t m, int64_t n, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb) { +inline int64_t lacpy(char u, int64_t m, int64_t n, const std::complex* a, int64_t lda, + std::complex* b, int64_t ldb) { return LAPACKE_clacpy(LAPACK_COL_MAJOR, u, m, n, - reinterpret_cast(a), lda, - reinterpret_cast(b), ldb); + reinterpret_cast(a), lda, + reinterpret_cast(b), ldb); } -inline int64_t lacpy(char u, int64_t m, int64_t n, const double *a, int64_t lda, double *b, +inline int64_t lacpy(char u, int64_t m, int64_t n, const double* a, int64_t lda, double* b, int64_t ldb) { return LAPACKE_dlacpy(LAPACK_COL_MAJOR, u, m, n, a, lda, b, ldb); } -inline int64_t lacpy(char u, int64_t m, int64_t n, const float *a, int64_t lda, float *b, +inline int64_t lacpy(char u, int64_t m, int64_t n, const float* a, int64_t lda, float* b, int64_t ldb) { return LAPACKE_slacpy(LAPACK_COL_MAJOR, u, m, n, a, lda, b, ldb); } -inline int64_t lacpy(char u, int64_t m, int64_t n, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb) { +inline int64_t lacpy(char u, int64_t m, int64_t n, const std::complex* a, int64_t lda, + std::complex* b, int64_t ldb) { return LAPACKE_zlacpy(LAPACK_COL_MAJOR, u, m, n, - reinterpret_cast(a), lda, - reinterpret_cast(b), ldb); + reinterpret_cast(a), lda, + reinterpret_cast(b), ldb); } -inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const std::complex *a, - int64_t lda, std::complex *b, int64_t ldb) { +inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const std::complex* a, + int64_t lda, std::complex* b, int64_t ldb) { return LAPACKE_clacpy(LAPACK_COL_MAJOR, to_char(u), m, n, - reinterpret_cast(a), lda, - reinterpret_cast(b), ldb); + reinterpret_cast(a), lda, + reinterpret_cast(b), ldb); } -inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const double *a, int64_t lda, - double *b, int64_t ldb) { +inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const double* a, int64_t lda, + double* b, int64_t ldb) { return LAPACKE_dlacpy(LAPACK_COL_MAJOR, to_char(u), m, n, a, lda, b, ldb); } -inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const float *a, int64_t lda, - float *b, int64_t ldb) { +inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const float* a, int64_t lda, + float* b, int64_t ldb) { return LAPACKE_slacpy(LAPACK_COL_MAJOR, to_char(u), m, n, a, lda, b, ldb); } -inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const std::complex *a, - int64_t lda, std::complex *b, int64_t ldb) { +inline int64_t lacpy(oneapi::math::uplo u, int64_t m, int64_t n, const std::complex* a, + int64_t lda, std::complex* b, int64_t ldb) { return LAPACKE_zlacpy(LAPACK_COL_MAJOR, to_char(u), m, n, - reinterpret_cast(a), lda, - reinterpret_cast(b), ldb); + reinterpret_cast(a), lda, + reinterpret_cast(b), ldb); } inline int64_t laset(oneapi::math::uplo u, int64_t m, int64_t n, std::complex alpha, - std::complex beta, std::complex *a, int64_t lda) { + std::complex beta, std::complex* a, int64_t lda) { return LAPACKE_claset(LAPACK_COL_MAJOR, to_char(u), m, n, - reinterpret_cast(alpha), - reinterpret_cast(beta), - reinterpret_cast(a), lda); + reinterpret_cast(alpha), + reinterpret_cast(beta), + reinterpret_cast(a), lda); } inline int64_t laset(oneapi::math::uplo u, int64_t m, int64_t n, double alpha, double beta, - double *a, int64_t lda) { + double* a, int64_t lda) { return LAPACKE_dlaset(LAPACK_COL_MAJOR, to_char(u), m, n, alpha, beta, a, lda); } -inline int64_t laset(oneapi::math::uplo u, int64_t m, int64_t n, float alpha, float beta, float *a, +inline int64_t laset(oneapi::math::uplo u, int64_t m, int64_t n, float alpha, float beta, float* a, int64_t lda) { return LAPACKE_slaset(LAPACK_COL_MAJOR, to_char(u), m, n, alpha, beta, a, lda); } inline int64_t laset(oneapi::math::uplo u, int64_t m, int64_t n, std::complex alpha, - std::complex beta, std::complex *a, int64_t lda) { + std::complex beta, std::complex* a, int64_t lda) { return LAPACKE_zlaset(LAPACK_COL_MAJOR, to_char(u), m, n, - reinterpret_cast(alpha), - reinterpret_cast(beta), - reinterpret_cast(a), lda); + reinterpret_cast(alpha), + reinterpret_cast(beta), + reinterpret_cast(a), lda); } inline int64_t laset(char u, int64_t m, int64_t n, std::complex alpha, - std::complex beta, std::complex *a, int64_t lda) { - return LAPACKE_claset(LAPACK_COL_MAJOR, u, m, n, - reinterpret_cast(alpha), - reinterpret_cast(beta), - reinterpret_cast(a), lda); + std::complex beta, std::complex* a, int64_t lda) { + return LAPACKE_claset(LAPACK_COL_MAJOR, u, m, n, reinterpret_cast(alpha), + reinterpret_cast(beta), + reinterpret_cast(a), lda); } -inline int64_t laset(char u, int64_t m, int64_t n, double alpha, double beta, double *a, +inline int64_t laset(char u, int64_t m, int64_t n, double alpha, double beta, double* a, int64_t lda) { return LAPACKE_dlaset(LAPACK_COL_MAJOR, u, m, n, alpha, beta, a, lda); } -inline int64_t laset(char u, int64_t m, int64_t n, float alpha, float beta, float *a, int64_t lda) { +inline int64_t laset(char u, int64_t m, int64_t n, float alpha, float beta, float* a, int64_t lda) { return LAPACKE_slaset(LAPACK_COL_MAJOR, u, m, n, alpha, beta, a, lda); } inline int64_t laset(char u, int64_t m, int64_t n, std::complex alpha, - std::complex beta, std::complex *a, int64_t lda) { + std::complex beta, std::complex* a, int64_t lda) { return LAPACKE_zlaset(LAPACK_COL_MAJOR, u, m, n, - reinterpret_cast(alpha), - reinterpret_cast(beta), - reinterpret_cast(a), lda); + reinterpret_cast(alpha), + reinterpret_cast(beta), + reinterpret_cast(a), lda); } -inline int64_t gebrd(int64_t m, int64_t n, std::complex *a, int64_t lda, float *d, float *e, - std::complex *tauq, std::complex *taup) { - return LAPACKE_cgebrd(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, - d, e, reinterpret_cast(tauq), - reinterpret_cast(taup)); +inline int64_t gebrd(int64_t m, int64_t n, std::complex* a, int64_t lda, float* d, float* e, + std::complex* tauq, std::complex* taup) { + return LAPACKE_cgebrd(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, + d, e, reinterpret_cast(tauq), + reinterpret_cast(taup)); } -inline int64_t gebrd(int64_t m, int64_t n, double *a, int64_t lda, double *d, double *e, - double *tauq, double *taup) { +inline int64_t gebrd(int64_t m, int64_t n, double* a, int64_t lda, double* d, double* e, + double* tauq, double* taup) { return LAPACKE_dgebrd(LAPACK_COL_MAJOR, m, n, a, lda, d, e, tauq, taup); } -inline int64_t gebrd(int64_t m, int64_t n, float *a, int64_t lda, float *d, float *e, float *tauq, - float *taup) { +inline int64_t gebrd(int64_t m, int64_t n, float* a, int64_t lda, float* d, float* e, float* tauq, + float* taup) { return LAPACKE_sgebrd(LAPACK_COL_MAJOR, m, n, a, lda, d, e, tauq, taup); } -inline int64_t gebrd(int64_t m, int64_t n, std::complex *a, int64_t lda, double *d, - double *e, std::complex *tauq, std::complex *taup) { - return LAPACKE_zgebrd(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, - d, e, reinterpret_cast(tauq), - reinterpret_cast(taup)); +inline int64_t gebrd(int64_t m, int64_t n, std::complex* a, int64_t lda, double* d, + double* e, std::complex* tauq, std::complex* taup) { + return LAPACKE_zgebrd(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, + d, e, reinterpret_cast(tauq), + reinterpret_cast(taup)); } -inline int64_t geqrf(int64_t m, int64_t n, std::complex *a, int64_t lda, - std::complex *tau) { - return LAPACKE_cgeqrf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, - reinterpret_cast(tau)); +inline int64_t geqrf(int64_t m, int64_t n, std::complex* a, int64_t lda, + std::complex* tau) { + return LAPACKE_cgeqrf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, + reinterpret_cast(tau)); } -inline int64_t geqrf(int64_t m, int64_t n, double *a, int64_t lda, double *tau) { +inline int64_t geqrf(int64_t m, int64_t n, double* a, int64_t lda, double* tau) { return LAPACKE_dgeqrf(LAPACK_COL_MAJOR, m, n, a, lda, tau); } -inline int64_t geqrf(int64_t m, int64_t n, float *a, int64_t lda, float *tau) { +inline int64_t geqrf(int64_t m, int64_t n, float* a, int64_t lda, float* tau) { return LAPACKE_sgeqrf(LAPACK_COL_MAJOR, m, n, a, lda, tau); } -inline int64_t geqrf(int64_t m, int64_t n, std::complex *a, int64_t lda, - std::complex *tau) { - return LAPACKE_zgeqrf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, - reinterpret_cast(tau)); +inline int64_t geqrf(int64_t m, int64_t n, std::complex* a, int64_t lda, + std::complex* tau) { + return LAPACKE_zgeqrf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, + reinterpret_cast(tau)); } -inline int64_t gerqf(int64_t m, int64_t n, std::complex *a, int64_t lda, - std::complex *tau) { - return LAPACKE_cgerqf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, - reinterpret_cast(tau)); +inline int64_t gerqf(int64_t m, int64_t n, std::complex* a, int64_t lda, + std::complex* tau) { + return LAPACKE_cgerqf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, + reinterpret_cast(tau)); } -inline int64_t gerqf(int64_t m, int64_t n, double *a, int64_t lda, double *tau) { +inline int64_t gerqf(int64_t m, int64_t n, double* a, int64_t lda, double* tau) { return LAPACKE_dgerqf(LAPACK_COL_MAJOR, m, n, a, lda, tau); } -inline int64_t gerqf(int64_t m, int64_t n, float *a, int64_t lda, float *tau) { +inline int64_t gerqf(int64_t m, int64_t n, float* a, int64_t lda, float* tau) { return LAPACKE_sgerqf(LAPACK_COL_MAJOR, m, n, a, lda, tau); } -inline int64_t gerqf(int64_t m, int64_t n, std::complex *a, int64_t lda, - std::complex *tau) { - return LAPACKE_zgerqf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, - reinterpret_cast(tau)); +inline int64_t gerqf(int64_t m, int64_t n, std::complex* a, int64_t lda, + std::complex* tau) { + return LAPACKE_zgerqf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, + reinterpret_cast(tau)); } inline int64_t gesvd(oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, int64_t m, int64_t n, - std::complex *a, int64_t lda, float *s, std::complex *u, - int64_t ldu, std::complex *vt, int64_t ldvt, float *superb) { + std::complex* a, int64_t lda, float* s, std::complex* u, + int64_t ldu, std::complex* vt, int64_t ldvt, float* superb) { return LAPACKE_cgesvd(LAPACK_COL_MAJOR, to_char(jobu), to_char(jobvt), m, n, - reinterpret_cast(a), lda, s, - reinterpret_cast(u), ldu, - reinterpret_cast(vt), ldvt, superb); + reinterpret_cast(a), lda, s, + reinterpret_cast(u), ldu, + reinterpret_cast(vt), ldvt, superb); } inline int64_t gesvd(oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, int64_t m, int64_t n, - double *a, int64_t lda, double *s, double *u, int64_t ldu, double *vt, - int64_t ldvt, double *superb) { + double* a, int64_t lda, double* s, double* u, int64_t ldu, double* vt, + int64_t ldvt, double* superb) { return LAPACKE_dgesvd(LAPACK_COL_MAJOR, to_char(jobu), to_char(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, superb); } inline int64_t gesvd(oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, int64_t m, int64_t n, - float *a, int64_t lda, float *s, float *u, int64_t ldu, float *vt, - int64_t ldvt, float *superb) { + float* a, int64_t lda, float* s, float* u, int64_t ldu, float* vt, + int64_t ldvt, float* superb) { return LAPACKE_sgesvd(LAPACK_COL_MAJOR, to_char(jobu), to_char(jobvt), m, n, a, lda, s, u, ldu, vt, ldvt, superb); } inline int64_t gesvd(oneapi::math::jobsvd jobu, oneapi::math::jobsvd jobvt, int64_t m, int64_t n, - std::complex *a, int64_t lda, double *s, std::complex *u, - int64_t ldu, std::complex *vt, int64_t ldvt, double *superb) { + std::complex* a, int64_t lda, double* s, std::complex* u, + int64_t ldu, std::complex* vt, int64_t ldvt, double* superb) { return LAPACKE_zgesvd(LAPACK_COL_MAJOR, to_char(jobu), to_char(jobvt), m, n, - reinterpret_cast(a), lda, s, - reinterpret_cast(u), ldu, - reinterpret_cast(vt), ldvt, superb); + reinterpret_cast(a), lda, s, + reinterpret_cast(u), ldu, + reinterpret_cast(vt), ldvt, superb); } -inline int64_t getrf(int64_t m, int64_t n, std::complex *a, int64_t lda, int64_t *ipiv) { - return LAPACKE_cgetrf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, - reinterpret_cast(ipiv)); +inline int64_t getrf(int64_t m, int64_t n, std::complex* a, int64_t lda, int64_t* ipiv) { + return LAPACKE_cgetrf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, + reinterpret_cast(ipiv)); } -inline int64_t getrf(int64_t m, int64_t n, double *a, int64_t lda, int64_t *ipiv) { - return LAPACKE_dgetrf(LAPACK_COL_MAJOR, m, n, a, lda, reinterpret_cast(ipiv)); +inline int64_t getrf(int64_t m, int64_t n, double* a, int64_t lda, int64_t* ipiv) { + return LAPACKE_dgetrf(LAPACK_COL_MAJOR, m, n, a, lda, reinterpret_cast(ipiv)); } -inline int64_t getrf(int64_t m, int64_t n, float *a, int64_t lda, int64_t *ipiv) { - return LAPACKE_sgetrf(LAPACK_COL_MAJOR, m, n, a, lda, reinterpret_cast(ipiv)); +inline int64_t getrf(int64_t m, int64_t n, float* a, int64_t lda, int64_t* ipiv) { + return LAPACKE_sgetrf(LAPACK_COL_MAJOR, m, n, a, lda, reinterpret_cast(ipiv)); } -inline int64_t getrf(int64_t m, int64_t n, std::complex *a, int64_t lda, int64_t *ipiv) { - return LAPACKE_zgetrf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, - reinterpret_cast(ipiv)); +inline int64_t getrf(int64_t m, int64_t n, std::complex* a, int64_t lda, int64_t* ipiv) { + return LAPACKE_zgetrf(LAPACK_COL_MAJOR, m, n, reinterpret_cast(a), lda, + reinterpret_cast(ipiv)); } -inline int64_t heevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, std::complex *a, - int64_t lda, float *w) { +inline int64_t heevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, std::complex* a, + int64_t lda, float* w) { return LAPACKE_cheevd(LAPACK_COL_MAJOR, to_char(j), to_char(u), n, - reinterpret_cast(a), lda, w); + reinterpret_cast(a), lda, w); } -inline int64_t heevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, std::complex *a, - int64_t lda, double *w) { +inline int64_t heevd(oneapi::math::job j, oneapi::math::uplo u, int64_t n, std::complex* a, + int64_t lda, double* w) { return LAPACKE_zheevd(LAPACK_COL_MAJOR, to_char(j), to_char(u), n, - reinterpret_cast(a), lda, w); + reinterpret_cast(a), lda, w); } inline int64_t hegvd(int64_t itype, oneapi::math::job j, oneapi::math::uplo u, int64_t n, - std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - float *w) { + std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + float* w) { return LAPACKE_chegvd(LAPACK_COL_MAJOR, itype, to_char(j), to_char(u), n, - reinterpret_cast(a), lda, - reinterpret_cast(b), ldb, w); + reinterpret_cast(a), lda, + reinterpret_cast(b), ldb, w); } inline int64_t hegvd(int64_t itype, oneapi::math::job j, oneapi::math::uplo u, int64_t n, - std::complex *a, int64_t lda, std::complex *b, int64_t ldb, - double *w) { + std::complex* a, int64_t lda, std::complex* b, int64_t ldb, + double* w) { return LAPACKE_zhegvd(LAPACK_COL_MAJOR, itype, to_char(j), to_char(u), n, - reinterpret_cast(a), lda, - reinterpret_cast(b), ldb, w); + reinterpret_cast(a), lda, + reinterpret_cast(b), ldb, w); } -inline int64_t hetrd(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, float *d, - float *e, std::complex *tau) { +inline int64_t hetrd(oneapi::math::uplo u, int64_t n, std::complex* a, int64_t lda, float* d, + float* e, std::complex* tau) { return LAPACKE_chetrd(LAPACK_COL_MAJOR, to_char(u), n, - reinterpret_cast(a), lda, d, e, - reinterpret_cast(tau)); + reinterpret_cast(a), lda, d, e, + reinterpret_cast(tau)); } -inline int64_t hetrd(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, - double *d, double *e, std::complex *tau) { +inline int64_t hetrd(oneapi::math::uplo u, int64_t n, std::complex* a, int64_t lda, + double* d, double* e, std::complex* tau) { return LAPACKE_zhetrd(LAPACK_COL_MAJOR, to_char(u), n, - reinterpret_cast(a), lda, d, e, - reinterpret_cast(tau)); + reinterpret_cast(a), lda, d, e, + reinterpret_cast(tau)); } -inline int64_t hetrf(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, - int64_t *ipiv) { +inline int64_t hetrf(oneapi::math::uplo u, int64_t n, std::complex* a, int64_t lda, + int64_t* ipiv) { return LAPACKE_chetrf(LAPACK_COL_MAJOR, to_char(u), n, - reinterpret_cast(a), lda, - reinterpret_cast(ipiv)); + reinterpret_cast(a), lda, + reinterpret_cast(ipiv)); } -inline int64_t hetrf(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, - int64_t *ipiv) { +inline int64_t hetrf(oneapi::math::uplo u, int64_t n, std::complex* a, int64_t lda, + int64_t* ipiv) { return LAPACKE_zhetrf(LAPACK_COL_MAJOR, to_char(u), n, - reinterpret_cast(a), lda, - reinterpret_cast(ipiv)); + reinterpret_cast(a), lda, + reinterpret_cast(ipiv)); } -inline int64_t ungtr(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, - const std::complex *tau) { +inline int64_t ungtr(oneapi::math::uplo u, int64_t n, std::complex* a, int64_t lda, + const std::complex* tau) { return LAPACKE_cungtr(LAPACK_COL_MAJOR, to_char(u), n, - reinterpret_cast(a), lda, - reinterpret_cast(tau)); + reinterpret_cast(a), lda, + reinterpret_cast(tau)); } -inline int64_t ungtr(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, - const std::complex *tau) { +inline int64_t ungtr(oneapi::math::uplo u, int64_t n, std::complex* a, int64_t lda, + const std::complex* tau) { return LAPACKE_zungtr(LAPACK_COL_MAJOR, to_char(u), n, - reinterpret_cast(a), lda, - reinterpret_cast(tau)); + reinterpret_cast(a), lda, + reinterpret_cast(tau)); } inline int64_t unmtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, - int64_t m, int64_t n, const std::complex *a, int64_t lda, - const std::complex *tau, std::complex *c, int64_t ldc) { + int64_t m, int64_t n, const std::complex* a, int64_t lda, + const std::complex* tau, std::complex* c, int64_t ldc) { return LAPACKE_cunmtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, - reinterpret_cast(a), lda, - reinterpret_cast(tau), - reinterpret_cast(c), ldc); + reinterpret_cast(a), lda, + reinterpret_cast(tau), + reinterpret_cast(c), ldc); } inline int64_t unmtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, - int64_t m, int64_t n, const std::complex *a, int64_t lda, - const std::complex *tau, std::complex *c, int64_t ldc) { + int64_t m, int64_t n, const std::complex* a, int64_t lda, + const std::complex* tau, std::complex* c, int64_t ldc) { return LAPACKE_zunmtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, - reinterpret_cast(a), lda, - reinterpret_cast(tau), - reinterpret_cast(c), ldc); + reinterpret_cast(a), lda, + reinterpret_cast(tau), + reinterpret_cast(c), ldc); } -inline int64_t orgtr(oneapi::math::uplo u, int64_t n, double *a, int64_t lda, const double *tau) { +inline int64_t orgtr(oneapi::math::uplo u, int64_t n, double* a, int64_t lda, const double* tau) { return LAPACKE_dorgtr(LAPACK_COL_MAJOR, to_char(u), n, a, lda, tau); } -inline int64_t orgtr(oneapi::math::uplo u, int64_t n, float *a, int64_t lda, const float *tau) { +inline int64_t orgtr(oneapi::math::uplo u, int64_t n, float* a, int64_t lda, const float* tau) { return LAPACKE_sorgtr(LAPACK_COL_MAJOR, to_char(u), n, a, lda, tau); } inline int64_t ormtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, - int64_t m, int64_t n, float *a, int64_t lda, const float *tau, float *c, + int64_t m, int64_t n, float* a, int64_t lda, const float* tau, float* c, int64_t ldc) { return LAPACKE_sormtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, a, lda, tau, c, ldc); } inline int64_t ormtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, - int64_t m, int64_t n, double *a, int64_t lda, const double *tau, double *c, + int64_t m, int64_t n, double* a, int64_t lda, const double* tau, double* c, int64_t ldc) { return LAPACKE_dormtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, a, lda, tau, c, ldc); } -inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, - int64_t m, int64_t n, float *a, int64_t lda, const float *tau, float *c, - int64_t ldc) { +inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, + oneapi::math::transpose trans, int64_t m, int64_t n, float* a, int64_t lda, + const float* tau, float* c, int64_t ldc) { return LAPACKE_sormtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, a, lda, tau, c, ldc); } -inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, - int64_t m, int64_t n, double *a, int64_t lda, const double *tau, double *c, - int64_t ldc) { +inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, + oneapi::math::transpose trans, int64_t m, int64_t n, double* a, + int64_t lda, const double* tau, double* c, int64_t ldc) { return LAPACKE_dormtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, a, lda, tau, c, ldc); } -inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, - int64_t m, int64_t n, std::complex *a, int64_t lda, - std::complex *tau, std::complex *c, int64_t ldc) { +inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, + oneapi::math::transpose trans, int64_t m, int64_t n, + std::complex* a, int64_t lda, std::complex* tau, + std::complex* c, int64_t ldc) { return LAPACKE_cunmtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, - reinterpret_cast(a), lda, - reinterpret_cast(tau), - reinterpret_cast(c), ldc); -} -inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, oneapi::math::transpose trans, - int64_t m, int64_t n, std::complex *a, int64_t lda, - std::complex *tau, std::complex *c, int64_t ldc) { + reinterpret_cast(a), lda, + reinterpret_cast(tau), + reinterpret_cast(c), ldc); +} +inline int64_t or_un_mtr(oneapi::math::side side, oneapi::math::uplo u, + oneapi::math::transpose trans, int64_t m, int64_t n, + std::complex* a, int64_t lda, std::complex* tau, + std::complex* c, int64_t ldc) { return LAPACKE_zunmtr(LAPACK_COL_MAJOR, to_char(side), to_char(u), to_char(trans), m, n, - reinterpret_cast(a), lda, - reinterpret_cast(tau), - reinterpret_cast(c), ldc); + reinterpret_cast(a), lda, + reinterpret_cast(tau), + reinterpret_cast(c), ldc); } -inline int64_t sytrd(oneapi::math::uplo u, int64_t n, float *a, int64_t lda, float *d, float *e, - float *tau) { +inline int64_t sytrd(oneapi::math::uplo u, int64_t n, float* a, int64_t lda, float* d, float* e, + float* tau) { return LAPACKE_ssytrd(LAPACK_COL_MAJOR, to_char(u), n, a, lda, d, e, tau); } -inline int64_t sytrd(oneapi::math::uplo u, int64_t n, double *a, int64_t lda, double *d, double *e, - double *tau) { +inline int64_t sytrd(oneapi::math::uplo u, int64_t n, double* a, int64_t lda, double* d, double* e, + double* tau) { return LAPACKE_dsytrd(LAPACK_COL_MAJOR, to_char(u), n, a, lda, d, e, tau); } -inline int64_t sytrf(oneapi::math::uplo u, int64_t n, float *a, int64_t lda, int64_t *ipiv) { +inline int64_t sytrf(oneapi::math::uplo u, int64_t n, float* a, int64_t lda, int64_t* ipiv) { return LAPACKE_ssytrf(LAPACK_COL_MAJOR, to_char(u), n, a, lda, - reinterpret_cast(ipiv)); + reinterpret_cast(ipiv)); } -inline int64_t sytrf(oneapi::math::uplo u, int64_t n, double *a, int64_t lda, int64_t *ipiv) { +inline int64_t sytrf(oneapi::math::uplo u, int64_t n, double* a, int64_t lda, int64_t* ipiv) { return LAPACKE_dsytrf(LAPACK_COL_MAJOR, to_char(u), n, a, lda, - reinterpret_cast(ipiv)); + reinterpret_cast(ipiv)); } -inline int64_t sytrf(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, - int64_t *ipiv) { +inline int64_t sytrf(oneapi::math::uplo u, int64_t n, std::complex* a, int64_t lda, + int64_t* ipiv) { return LAPACKE_csytrf(LAPACK_COL_MAJOR, to_char(u), n, - reinterpret_cast(a), lda, - reinterpret_cast(ipiv)); + reinterpret_cast(a), lda, + reinterpret_cast(ipiv)); } -inline int64_t sytrf(oneapi::math::uplo u, int64_t n, std::complex *a, int64_t lda, - int64_t *ipiv) { +inline int64_t sytrf(oneapi::math::uplo u, int64_t n, std::complex* a, int64_t lda, + int64_t* ipiv) { return LAPACKE_zsytrf(LAPACK_COL_MAJOR, to_char(u), n, - reinterpret_cast(a), lda, - reinterpret_cast(ipiv)); + reinterpret_cast(a), lda, + reinterpret_cast(ipiv)); } -inline void orgbr(oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, double *a, - int64_t lda, const double *tau) { +inline void orgbr(oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, double* a, + int64_t lda, const double* tau) { LAPACKE_dorgbr(LAPACK_COL_MAJOR, to_char(vect), m, n, k, a, lda, tau); } -inline void orgbr(oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, float *a, - int64_t lda, const float *tau) { +inline void orgbr(oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, float* a, + int64_t lda, const float* tau) { LAPACKE_sorgbr(LAPACK_COL_MAJOR, to_char(vect), m, n, k, a, lda, tau); } -inline int64_t or_un_gqr(int64_t m, int64_t n, int64_t k, float *a, int64_t lda, const float *tau) { +inline int64_t or_un_gqr(int64_t m, int64_t n, int64_t k, float* a, int64_t lda, const float* tau) { return LAPACKE_sorgqr(LAPACK_COL_MAJOR, m, n, k, a, lda, tau); } -inline int64_t or_un_gqr(int64_t m, int64_t n, int64_t k, double *a, int64_t lda, - const double *tau) { +inline int64_t or_un_gqr(int64_t m, int64_t n, int64_t k, double* a, int64_t lda, + const double* tau) { return LAPACKE_dorgqr(LAPACK_COL_MAJOR, m, n, k, a, lda, tau); } -inline int64_t or_un_gqr(int64_t m, int64_t n, int64_t k, std::complex *a, int64_t lda, - const std::complex *tau) { - return LAPACKE_cungqr(LAPACK_COL_MAJOR, m, n, k, reinterpret_cast(a), - lda, reinterpret_cast(tau)); +inline int64_t or_un_gqr(int64_t m, int64_t n, int64_t k, std::complex* a, int64_t lda, + const std::complex* tau) { + return LAPACKE_cungqr(LAPACK_COL_MAJOR, m, n, k, reinterpret_cast(a), + lda, reinterpret_cast(tau)); } -inline int64_t or_un_gqr(int64_t m, int64_t n, int64_t k, std::complex *a, int64_t lda, - const std::complex *tau) { - return LAPACKE_zungqr(LAPACK_COL_MAJOR, m, n, k, reinterpret_cast(a), - lda, reinterpret_cast(tau)); +inline int64_t or_un_gqr(int64_t m, int64_t n, int64_t k, std::complex* a, int64_t lda, + const std::complex* tau) { + return LAPACKE_zungqr(LAPACK_COL_MAJOR, m, n, k, reinterpret_cast(a), + lda, reinterpret_cast(tau)); } inline int64_t or_un_mqr(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, - int64_t k, const float *a, int64_t lda, const float *tau, float *c, + int64_t k, const float* a, int64_t lda, const float* tau, float* c, int64_t ldc) { return LAPACKE_sormqr(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, a, lda, tau, c, ldc); } inline int64_t or_un_mqr(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, - int64_t k, const double *a, int64_t lda, const double *tau, double *c, + int64_t k, const double* a, int64_t lda, const double* tau, double* c, int64_t ldc) { return LAPACKE_dormqr(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, a, lda, tau, c, ldc); } inline int64_t or_un_mqr(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, - int64_t k, const std::complex *a, int64_t lda, - const std::complex *tau, std::complex *c, int64_t ldc) { + int64_t k, const std::complex* a, int64_t lda, + const std::complex* tau, std::complex* c, int64_t ldc) { return LAPACKE_cunmqr(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, - reinterpret_cast(a), lda, - reinterpret_cast(tau), - reinterpret_cast(c), ldc); + reinterpret_cast(a), lda, + reinterpret_cast(tau), + reinterpret_cast(c), ldc); } inline int64_t or_un_mqr(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, - int64_t k, const std::complex *a, int64_t lda, - const std::complex *tau, std::complex *c, int64_t ldc) { + int64_t k, const std::complex* a, int64_t lda, + const std::complex* tau, std::complex* c, int64_t ldc) { return LAPACKE_zunmqr(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, - reinterpret_cast(a), lda, - reinterpret_cast(tau), - reinterpret_cast(c), ldc); + reinterpret_cast(a), lda, + reinterpret_cast(tau), + reinterpret_cast(c), ldc); } -inline int64_t or_un_grq(int64_t m, int64_t n, int64_t k, float *a, int64_t lda, const float *tau) { +inline int64_t or_un_grq(int64_t m, int64_t n, int64_t k, float* a, int64_t lda, const float* tau) { return LAPACKE_sorgrq(LAPACK_COL_MAJOR, m, n, k, a, lda, tau); } -inline int64_t or_un_grq(int64_t m, int64_t n, int64_t k, double *a, int64_t lda, - const double *tau) { +inline int64_t or_un_grq(int64_t m, int64_t n, int64_t k, double* a, int64_t lda, + const double* tau) { return LAPACKE_dorgrq(LAPACK_COL_MAJOR, m, n, k, a, lda, tau); } -inline int64_t or_un_grq(int64_t m, int64_t n, int64_t k, std::complex *a, int64_t lda, - const std::complex *tau) { - return LAPACKE_cungrq(LAPACK_COL_MAJOR, m, n, k, reinterpret_cast(a), - lda, reinterpret_cast(tau)); +inline int64_t or_un_grq(int64_t m, int64_t n, int64_t k, std::complex* a, int64_t lda, + const std::complex* tau) { + return LAPACKE_cungrq(LAPACK_COL_MAJOR, m, n, k, reinterpret_cast(a), + lda, reinterpret_cast(tau)); } -inline int64_t or_un_grq(int64_t m, int64_t n, int64_t k, std::complex *a, int64_t lda, - const std::complex *tau) { - return LAPACKE_zungrq(LAPACK_COL_MAJOR, m, n, k, reinterpret_cast(a), - lda, reinterpret_cast(tau)); +inline int64_t or_un_grq(int64_t m, int64_t n, int64_t k, std::complex* a, int64_t lda, + const std::complex* tau) { + return LAPACKE_zungrq(LAPACK_COL_MAJOR, m, n, k, reinterpret_cast(a), + lda, reinterpret_cast(tau)); } inline int64_t or_un_mrq(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, - int64_t k, const float *a, int64_t lda, const float *tau, float *c, + int64_t k, const float* a, int64_t lda, const float* tau, float* c, int64_t ldc) { return LAPACKE_sormrq(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, a, lda, tau, c, ldc); } inline int64_t or_un_mrq(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, - int64_t k, const double *a, int64_t lda, const double *tau, double *c, + int64_t k, const double* a, int64_t lda, const double* tau, double* c, int64_t ldc) { return LAPACKE_dormrq(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, a, lda, tau, c, ldc); } inline int64_t or_un_mrq(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, - int64_t k, const std::complex *a, int64_t lda, - const std::complex *tau, std::complex *c, int64_t ldc) { + int64_t k, const std::complex* a, int64_t lda, + const std::complex* tau, std::complex* c, int64_t ldc) { return LAPACKE_cunmrq(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, - reinterpret_cast(a), lda, - reinterpret_cast(tau), - reinterpret_cast(c), ldc); + reinterpret_cast(a), lda, + reinterpret_cast(tau), + reinterpret_cast(c), ldc); } inline int64_t or_un_mrq(oneapi::math::side s, oneapi::math::transpose t, int64_t m, int64_t n, - int64_t k, const std::complex *a, int64_t lda, - const std::complex *tau, std::complex *c, int64_t ldc) { + int64_t k, const std::complex* a, int64_t lda, + const std::complex* tau, std::complex* c, int64_t ldc) { return LAPACKE_zunmrq(LAPACK_COL_MAJOR, to_char(s), to_char(t), m, n, k, - reinterpret_cast(a), lda, - reinterpret_cast(tau), - reinterpret_cast(c), ldc); + reinterpret_cast(a), lda, + reinterpret_cast(tau), + reinterpret_cast(c), ldc); } -inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, std::complex *a, +inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, std::complex* a, int64_t lda) { return LAPACKE_cpotrf(LAPACK_COL_MAJOR, to_char(upper_lower), n, - reinterpret_cast(a), lda); + reinterpret_cast(a), lda); } -inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, double *a, int64_t lda) { +inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, double* a, int64_t lda) { return LAPACKE_dpotrf(LAPACK_COL_MAJOR, to_char(upper_lower), n, a, lda); } -inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, float *a, int64_t lda) { +inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, float* a, int64_t lda) { return LAPACKE_spotrf(LAPACK_COL_MAJOR, to_char(upper_lower), n, a, lda); } -inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, std::complex *a, +inline int64_t potrf(oneapi::math::uplo upper_lower, int64_t n, std::complex* a, int64_t lda) { return LAPACKE_zpotrf(LAPACK_COL_MAJOR, to_char(upper_lower), n, - reinterpret_cast(a), lda); + reinterpret_cast(a), lda); } inline int64_t potrs(oneapi::math::uplo upper_lower, int64_t n, int64_t nrhs, - const std::complex *a, int64_t lda, std::complex *b, + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb) { return LAPACKE_cpotrs(LAPACK_COL_MAJOR, to_char(upper_lower), n, nrhs, - reinterpret_cast(a), lda, - reinterpret_cast(b), ldb); + reinterpret_cast(a), lda, + reinterpret_cast(b), ldb); } -inline int64_t potrs(oneapi::math::uplo upper_lower, int64_t n, int64_t nrhs, const double *a, - int64_t lda, double *b, int64_t ldb) { +inline int64_t potrs(oneapi::math::uplo upper_lower, int64_t n, int64_t nrhs, const double* a, + int64_t lda, double* b, int64_t ldb) { return LAPACKE_dpotrs(LAPACK_COL_MAJOR, to_char(upper_lower), n, nrhs, a, lda, b, ldb); } -inline int64_t potrs(oneapi::math::uplo upper_lower, int64_t n, int64_t nrhs, const float *a, - int64_t lda, float *b, int64_t ldb) { +inline int64_t potrs(oneapi::math::uplo upper_lower, int64_t n, int64_t nrhs, const float* a, + int64_t lda, float* b, int64_t ldb) { return LAPACKE_spotrs(LAPACK_COL_MAJOR, to_char(upper_lower), n, nrhs, a, lda, b, ldb); } inline int64_t potrs(oneapi::math::uplo upper_lower, int64_t n, int64_t nrhs, - const std::complex *a, int64_t lda, std::complex *b, + const std::complex* a, int64_t lda, std::complex* b, int64_t ldb) { return LAPACKE_zpotrs(LAPACK_COL_MAJOR, to_char(upper_lower), n, nrhs, - reinterpret_cast(a), lda, - reinterpret_cast(b), ldb); + reinterpret_cast(a), lda, + reinterpret_cast(b), ldb); } -inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, std::complex *a, +inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, std::complex* a, int64_t lda) { return LAPACKE_cpotri(LAPACK_COL_MAJOR, to_char(upper_lower), n, - reinterpret_cast(a), lda); + reinterpret_cast(a), lda); } -inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, double *a, int64_t lda) { +inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, double* a, int64_t lda) { return LAPACKE_dpotri(LAPACK_COL_MAJOR, to_char(upper_lower), n, a, lda); } -inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, float *a, int64_t lda) { +inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, float* a, int64_t lda) { return LAPACKE_spotri(LAPACK_COL_MAJOR, to_char(upper_lower), n, a, lda); } -inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, std::complex *a, +inline int64_t potri(oneapi::math::uplo upper_lower, int64_t n, std::complex* a, int64_t lda) { return LAPACKE_zpotri(LAPACK_COL_MAJOR, to_char(upper_lower), n, - reinterpret_cast(a), lda); + reinterpret_cast(a), lda); } -inline int64_t laswp(int64_t n, std::complex *a, int64_t lda, int64_t k1, int64_t k2, - const int64_t *ipiv, int64_t incx) { - return LAPACKE_claswp(LAPACK_COL_MAJOR, n, reinterpret_cast(a), lda, k1, - k2, reinterpret_cast(ipiv), incx); +inline int64_t laswp(int64_t n, std::complex* a, int64_t lda, int64_t k1, int64_t k2, + const int64_t* ipiv, int64_t incx) { + return LAPACKE_claswp(LAPACK_COL_MAJOR, n, reinterpret_cast(a), lda, k1, + k2, reinterpret_cast(ipiv), incx); } -inline int64_t laswp(int64_t n, double *a, int64_t lda, int64_t k1, int64_t k2, const int64_t *ipiv, +inline int64_t laswp(int64_t n, double* a, int64_t lda, int64_t k1, int64_t k2, const int64_t* ipiv, int64_t incx) { return LAPACKE_dlaswp(LAPACK_COL_MAJOR, n, a, lda, k1, k2, - reinterpret_cast(ipiv), incx); + reinterpret_cast(ipiv), incx); } -inline int64_t laswp(int64_t n, float *a, int64_t lda, int64_t k1, int64_t k2, const int64_t *ipiv, +inline int64_t laswp(int64_t n, float* a, int64_t lda, int64_t k1, int64_t k2, const int64_t* ipiv, int64_t incx) { return LAPACKE_slaswp(LAPACK_COL_MAJOR, n, a, lda, k1, k2, - reinterpret_cast(ipiv), incx); + reinterpret_cast(ipiv), incx); } -inline int64_t laswp(int64_t n, std::complex *a, int64_t lda, int64_t k1, int64_t k2, - const int64_t *ipiv, int64_t incx) { - return LAPACKE_zlaswp(LAPACK_COL_MAJOR, n, reinterpret_cast(a), lda, - k1, k2, reinterpret_cast(ipiv), incx); +inline int64_t laswp(int64_t n, std::complex* a, int64_t lda, int64_t k1, int64_t k2, + const int64_t* ipiv, int64_t incx) { + return LAPACKE_zlaswp(LAPACK_COL_MAJOR, n, reinterpret_cast(a), lda, k1, + k2, reinterpret_cast(ipiv), incx); } inline void ungbr(oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, - std::complex *a, int64_t lda, const std::complex *tau) { + std::complex* a, int64_t lda, const std::complex* tau) { LAPACKE_cungbr(LAPACK_COL_MAJOR, to_char(vect), m, n, k, - reinterpret_cast(a), lda, - reinterpret_cast(tau)); + reinterpret_cast(a), lda, + reinterpret_cast(tau)); } inline void ungbr(oneapi::math::generate vect, int64_t m, int64_t n, int64_t k, - std::complex *a, int64_t lda, const std::complex *tau) { + std::complex* a, int64_t lda, const std::complex* tau) { LAPACKE_zungbr(LAPACK_COL_MAJOR, to_char(vect), m, n, k, - reinterpret_cast(a), lda, - reinterpret_cast(tau)); + reinterpret_cast(a), lda, + reinterpret_cast(tau)); } -inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - int64_t n, int64_t nrhs, const float *a, int64_t lda, float *b, int64_t ldb) { +inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, int64_t n, int64_t nrhs, const float* a, int64_t lda, + float* b, int64_t ldb) { return LAPACKE_strtrs(LAPACK_COL_MAJOR, to_char(uplo), to_char(trans), to_char(diag), n, nrhs, a, lda, b, ldb); } -inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - int64_t n, int64_t nrhs, const double *a, int64_t lda, double *b, - int64_t ldb) { +inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, int64_t n, int64_t nrhs, const double* a, int64_t lda, + double* b, int64_t ldb) { return LAPACKE_dtrtrs(LAPACK_COL_MAJOR, to_char(uplo), to_char(trans), to_char(diag), n, nrhs, a, lda, b, ldb); } -inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - int64_t n, int64_t nrhs, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb) { +inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, int64_t n, int64_t nrhs, const std::complex* a, + int64_t lda, std::complex* b, int64_t ldb) { return LAPACKE_ctrtrs(LAPACK_COL_MAJOR, to_char(uplo), to_char(trans), to_char(diag), n, nrhs, - reinterpret_cast(a), lda, - reinterpret_cast(b), ldb); + reinterpret_cast(a), lda, + reinterpret_cast(b), ldb); } -inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, oneapi::math::diag diag, - int64_t n, int64_t nrhs, const std::complex *a, int64_t lda, - std::complex *b, int64_t ldb) { +inline int64_t trtrs(oneapi::math::uplo uplo, oneapi::math::transpose trans, + oneapi::math::diag diag, int64_t n, int64_t nrhs, + const std::complex* a, int64_t lda, std::complex* b, + int64_t ldb) { return LAPACKE_ztrtrs(LAPACK_COL_MAJOR, to_char(uplo), to_char(trans), to_char(diag), n, nrhs, - reinterpret_cast(a), lda, - reinterpret_cast(b), ldb); + reinterpret_cast(a), lda, + reinterpret_cast(b), ldb); } } //namespace reference diff --git a/tests/unit_tests/lapack/source/gebrd.cpp b/tests/unit_tests/lapack/source/gebrd.cpp index 650a73145..2bee82853 100644 --- a/tests/unit_tests/lapack/source/gebrd.cpp +++ b/tests/unit_tests/lapack/source/gebrd.cpp @@ -83,10 +83,10 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 #ifdef CALL_RT_API oneapi::math::lapack::gebrd(queue, m, n, A_dev, lda, d_dev, e_dev, tauq_dev, taup_dev, - scratchpad_dev, scratchpad_size); + scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::gebrd, m, n, A_dev, lda, d_dev, e_dev, - tauq_dev, taup_dev, scratchpad_dev, scratchpad_size); + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::gebrd, m, n, A_dev, lda, d_dev, + e_dev, tauq_dev, taup_dev, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); diff --git a/tests/unit_tests/lapack/source/geqrf.cpp b/tests/unit_tests/lapack/source/geqrf.cpp index 570fb835e..231be4b5d 100644 --- a/tests/unit_tests/lapack/source/geqrf.cpp +++ b/tests/unit_tests/lapack/source/geqrf.cpp @@ -75,7 +75,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 #ifdef CALL_RT_API oneapi::math::lapack::geqrf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::geqrf, m, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); @@ -135,7 +135,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::geqrf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::geqrf, m, n, A_dev, lda, diff --git a/tests/unit_tests/lapack/source/geqrf_batch_group.cpp b/tests/unit_tests/lapack/source/geqrf_batch_group.cpp index 5f6f93b7c..c94ddd000 100644 --- a/tests/unit_tests/lapack/source/geqrf_batch_group.cpp +++ b/tests/unit_tests/lapack/source/geqrf_batch_group.cpp @@ -121,8 +121,8 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { #ifdef CALL_RT_API oneapi::math::lapack::geqrf_batch(queue, m_vec.data(), n_vec.data(), A_dev_ptrs, - lda_vec.data(), tau_dev_ptrs, group_count, - group_sizes_vec.data(), scratchpad_dev, scratchpad_size); + lda_vec.data(), tau_dev_ptrs, group_count, + group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::geqrf_batch, m_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, diff --git a/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp b/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp index 227fdc30f..a3cdb55e3 100644 --- a/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/geqrf_batch_stride.cpp @@ -76,7 +76,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, int64_ #ifdef CALL_RT_API oneapi::math::lapack::geqrf_batch(queue, m, n, A_dev, lda, stride_a, tau_dev, stride_tau, - batch_size, scratchpad_dev, scratchpad_size); + batch_size, scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::geqrf_batch, m, n, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, @@ -155,9 +155,10 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::geqrf_batch, m, n, A_dev, - lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::geqrf_batch, m, n, + A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, + scratchpad_dev, scratchpad_size, + std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/lapack/source/gerqf.cpp b/tests/unit_tests/lapack/source/gerqf.cpp index 091aa5d68..bb6693598 100644 --- a/tests/unit_tests/lapack/source/gerqf.cpp +++ b/tests/unit_tests/lapack/source/gerqf.cpp @@ -75,7 +75,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 #ifdef CALL_RT_API oneapi::math::lapack::gerqf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::gerqf, m, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); @@ -135,7 +135,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::gerqf(queue, m, n, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::gerqf, m, n, A_dev, lda, diff --git a/tests/unit_tests/lapack/source/gesvd.cpp b/tests/unit_tests/lapack/source/gesvd.cpp index 2d793c1e8..afca7850b 100644 --- a/tests/unit_tests/lapack/source/gesvd.cpp +++ b/tests/unit_tests/lapack/source/gesvd.cpp @@ -86,7 +86,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::jobsvd jobu, oneapi::math:: #ifdef CALL_RT_API oneapi::math::lapack::gesvd(queue, jobu, jobvt, m, n, A_dev, lda, s_dev, U_dev, ldu, Vt_dev, - ldvt, scratchpad_dev, scratchpad_size); + ldvt, scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::gesvd, jobu, jobvt, m, n, A_dev, lda, s_dev, U_dev, ldu, Vt_dev, ldvt, scratchpad_dev, scratchpad_size); @@ -116,8 +116,8 @@ bool accuracy(const sycl::device& dev, oneapi::math::jobsvd jobu, oneapi::math:: US[row + col * ldus] = U[row + col * ldu] * s[col]; std::vector USV(m * n); int64_t ldusv = m; - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, m, n, n, - 1.0, US.data(), ldus, Vt.data(), ldvt, 0.0, USV.data(), ldusv); + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, m, n, + n, 1.0, US.data(), ldus, Vt.data(), ldvt, 0.0, USV.data(), ldusv); if (!rel_mat_err_check(m, n, A_initial, lda, USV, ldusv)) { test_log::lout << "Factorization check failed" << std::endl; result = false; @@ -134,8 +134,8 @@ bool accuracy(const sycl::device& dev, oneapi::math::jobsvd jobu, oneapi::math:: /* |I - U' U| < n O(eps) */ std::vector UU(ucols * ucols); int64_t lduu = ucols; - reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, ucols, - ucols, m, 1.0, U.data(), ldu, U.data(), ldu, 0.0, UU.data(), lduu); + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, + ucols, ucols, m, 1.0, U.data(), ldu, U.data(), ldu, 0.0, UU.data(), lduu); if (!rel_id_err_check(ucols, UU, lduu)) { test_log::lout << "U Orthogonality check failed" << std::endl; result = false; @@ -147,8 +147,9 @@ bool accuracy(const sycl::device& dev, oneapi::math::jobsvd jobu, oneapi::math:: /* |I - V' V| < n O(eps) */ std::vector VV(vtrows * vtrows); int64_t ldvv = vtrows; - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, vtrows, - vtrows, n, 1.0, Vt.data(), ldvt, Vt.data(), ldvt, 0.0, VV.data(), ldvv); + reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::conjtrans, + vtrows, vtrows, n, 1.0, Vt.data(), ldvt, Vt.data(), ldvt, 0.0, VV.data(), + ldvv); if (!rel_id_err_check(vtrows, VV, ldvv)) { test_log::lout << "V Orthogonality check failed" << std::endl; result = false; @@ -214,8 +215,8 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::jobsvd jobu, oneapi:: scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::gesvd, jobu, jobvt, m, n, - A_dev, lda, s_dev, U_dev, ldu, Vt_dev, ldvt, scratchpad_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::gesvd, jobu, jobvt, m, + n, A_dev, lda, s_dev, U_dev, ldu, Vt_dev, ldvt, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/lapack/source/getrf.cpp b/tests/unit_tests/lapack/source/getrf.cpp index 165c7fed5..ca84f6680 100644 --- a/tests/unit_tests/lapack/source/getrf.cpp +++ b/tests/unit_tests/lapack/source/getrf.cpp @@ -78,7 +78,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, uint64 #ifdef CALL_RT_API oneapi::math::lapack::getrf(queue, m, n, A_dev, lda, ipiv_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrf, m, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); @@ -138,7 +138,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::getrf(queue, m, n, A_dev, lda, ipiv_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getrf, m, n, A_dev, lda, diff --git a/tests/unit_tests/lapack/source/getrf_batch_group.cpp b/tests/unit_tests/lapack/source/getrf_batch_group.cpp index 0641698df..329f3d767 100644 --- a/tests/unit_tests/lapack/source/getrf_batch_group.cpp +++ b/tests/unit_tests/lapack/source/getrf_batch_group.cpp @@ -125,8 +125,8 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { #ifdef CALL_RT_API oneapi::math::lapack::getrf_batch(queue, m_vec.data(), n_vec.data(), A_dev_ptrs, - lda_vec.data(), ipiv_dev_ptrs, group_count, - group_sizes_vec.data(), scratchpad_dev, scratchpad_size); + lda_vec.data(), ipiv_dev_ptrs, group_count, + group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrf_batch, m_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, diff --git a/tests/unit_tests/lapack/source/getrf_batch_stride.cpp b/tests/unit_tests/lapack/source/getrf_batch_stride.cpp index 62c55d5fb..0ba2d1714 100644 --- a/tests/unit_tests/lapack/source/getrf_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/getrf_batch_stride.cpp @@ -76,7 +76,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, int64_ #ifdef CALL_RT_API oneapi::math::lapack::getrf_batch(queue, m, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, - batch_size, scratchpad_dev, scratchpad_size); + batch_size, scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrf_batch, m, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, batch_size, scratchpad_dev, @@ -155,9 +155,10 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t lda, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getrf_batch, m, n, A_dev, - lda, stride_a, ipiv_dev, stride_ipiv, batch_size, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getrf_batch, m, n, + A_dev, lda, stride_a, ipiv_dev, stride_ipiv, batch_size, + scratchpad_dev, scratchpad_size, + std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/lapack/source/getri.cpp b/tests/unit_tests/lapack/source/getri.cpp index 24d272520..26b7c7307 100644 --- a/tests/unit_tests/lapack/source/getri.cpp +++ b/tests/unit_tests/lapack/source/getri.cpp @@ -83,7 +83,8 @@ bool accuracy(const sycl::device& dev, int64_t n, int64_t lda, uint64_t seed) { queue.wait_and_throw(); #ifdef CALL_RT_API - oneapi::math::lapack::getri(queue, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); + oneapi::math::lapack::getri(queue, n, A_dev, lda, ipiv_dev, scratchpad_dev, + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getri, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); @@ -149,7 +150,7 @@ bool usm_dependency(const sycl::device& dev, int64_t n, int64_t lda, uint64_t se #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::getri(queue, n, A_dev, lda, ipiv_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::getri, n, A_dev, lda, diff --git a/tests/unit_tests/lapack/source/getri_batch_group.cpp b/tests/unit_tests/lapack/source/getri_batch_group.cpp index d700a37f4..d13b40442 100644 --- a/tests/unit_tests/lapack/source/getri_batch_group.cpp +++ b/tests/unit_tests/lapack/source/getri_batch_group.cpp @@ -135,11 +135,11 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { #ifdef CALL_RT_API oneapi::math::lapack::getri_batch(queue, n_vec.data(), A_dev_ptrs, lda_vec.data(), - ipiv_dev_ptrs, group_count, group_sizes_vec.data(), - scratchpad_dev, scratchpad_size); + ipiv_dev_ptrs, group_count, group_sizes_vec.data(), + scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getri_batch, n_vec.data(), A_dev_ptrs, - lda_vec.data(), ipiv_dev_ptrs, group_count, + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getri_batch, n_vec.data(), + A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, group_count, group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); diff --git a/tests/unit_tests/lapack/source/getri_batch_stride.cpp b/tests/unit_tests/lapack/source/getri_batch_stride.cpp index aee87f7b9..858847f5d 100644 --- a/tests/unit_tests/lapack/source/getri_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/getri_batch_stride.cpp @@ -84,7 +84,7 @@ bool accuracy(const sycl::device& dev, int64_t n, int64_t lda, int64_t stride_a, #ifdef CALL_RT_API oneapi::math::lapack::getri_batch(queue, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, - batch_size, scratchpad_dev, scratchpad_size); + batch_size, scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getri_batch, n, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, batch_size, scratchpad_dev, diff --git a/tests/unit_tests/lapack/source/getrs.cpp b/tests/unit_tests/lapack/source/getrs.cpp index 4532e3dd9..f43abb427 100644 --- a/tests/unit_tests/lapack/source/getrs.cpp +++ b/tests/unit_tests/lapack/source/getrs.cpp @@ -86,7 +86,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::transpose trans, int64_t n, #ifdef CALL_RT_API oneapi::math::lapack::getrs(queue, trans, n, nrhs, A_dev, lda, ipiv_dev, B_dev, ldb, - scratchpad_dev, scratchpad_size); + scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrs, trans, n, nrhs, A_dev, lda, ipiv_dev, B_dev, ldb, scratchpad_dev, scratchpad_size); diff --git a/tests/unit_tests/lapack/source/getrs_batch_group.cpp b/tests/unit_tests/lapack/source/getrs_batch_group.cpp index d2cd80e9c..64c88bcd6 100644 --- a/tests/unit_tests/lapack/source/getrs_batch_group.cpp +++ b/tests/unit_tests/lapack/source/getrs_batch_group.cpp @@ -47,7 +47,7 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { /* Test Parameters */ std::vector trans_vec = { oneapi::math::transpose::nontrans, - oneapi::math::transpose::trans }; + oneapi::math::transpose::trans }; std::vector n_vec = { 4, 5 }; std::vector nrhs_vec = { 9, 6 }; std::vector lda_vec = { 6, 6 }; @@ -163,9 +163,9 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { #ifdef CALL_RT_API oneapi::math::lapack::getrs_batch(queue, trans_vec.data(), n_vec.data(), nrhs_vec.data(), - A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, B_dev_ptrs, - ldb_vec.data(), group_count, group_sizes_vec.data(), - scratchpad_dev, scratchpad_size); + A_dev_ptrs, lda_vec.data(), ipiv_dev_ptrs, B_dev_ptrs, + ldb_vec.data(), group_count, group_sizes_vec.data(), + scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrs_batch, trans_vec.data(), n_vec.data(), nrhs_vec.data(), A_dev_ptrs, lda_vec.data(), diff --git a/tests/unit_tests/lapack/source/getrs_batch_stride.cpp b/tests/unit_tests/lapack/source/getrs_batch_stride.cpp index ff159416f..600067883 100644 --- a/tests/unit_tests/lapack/source/getrs_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/getrs_batch_stride.cpp @@ -79,8 +79,8 @@ bool accuracy(const sycl::device& dev, oneapi::math::transpose trans, int64_t n, #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size, trans, n, - nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); + queue, scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size, trans, + n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); @@ -91,8 +91,8 @@ bool accuracy(const sycl::device& dev, oneapi::math::transpose trans, int64_t n, #ifdef CALL_RT_API oneapi::math::lapack::getrs_batch(queue, trans, n, nrhs, A_dev, lda, stride_a, ipiv_dev, - stride_ipiv, B_dev, ldb, stride_b, batch_size, - scratchpad_dev, scratchpad_size); + stride_ipiv, B_dev, ldb, stride_b, batch_size, + scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::getrs_batch, trans, n, nrhs, A_dev, lda, stride_a, ipiv_dev, stride_ipiv, B_dev, ldb, stride_b, @@ -168,8 +168,8 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::transpose trans, int6 #else int64_t scratchpad_size; TEST_RUN_LAPACK_CT_SELECT( - queue, scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size, trans, n, - nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); + queue, scratchpad_size = oneapi::math::lapack::getrs_batch_scratchpad_size, trans, + n, nrhs, lda, stride_a, stride_ipiv, ldb, stride_b, batch_size); #endif auto scratchpad_dev = device_alloc(queue, scratchpad_size); diff --git a/tests/unit_tests/lapack/source/heevd.cpp b/tests/unit_tests/lapack/source/heevd.cpp index b2a4c906b..8eecbba2b 100644 --- a/tests/unit_tests/lapack/source/heevd.cpp +++ b/tests/unit_tests/lapack/source/heevd.cpp @@ -73,7 +73,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::job jobz, oneapi::math::upl #ifdef CALL_RT_API oneapi::math::lapack::heevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::heevd, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, scratchpad_size); @@ -134,7 +134,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::job jobz, oneapi::mat #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::heevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::heevd, jobz, uplo, n, diff --git a/tests/unit_tests/lapack/source/hegvd.cpp b/tests/unit_tests/lapack/source/hegvd.cpp index a072cd410..55422330b 100644 --- a/tests/unit_tests/lapack/source/hegvd.cpp +++ b/tests/unit_tests/lapack/source/hegvd.cpp @@ -42,8 +42,8 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - int64_t n, int64_t lda, int64_t ldb, uint64_t seed) { +bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, int64_t n, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -80,7 +80,7 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, on #ifdef CALL_RT_API oneapi::math::lapack::hegvd(queue, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, - scratchpad_dev, scratchpad_size); + scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::hegvd, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size); @@ -141,8 +141,8 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, on /* |I - Z' B Z| < n O(eps) */ std::vector ZBZ(n * n); int64_t ldzbz = n; - reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, - n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, + n, n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); if (!rel_id_err_check(n, ZBZ, ldzbz)) { test_log::lout << "Orthogonality check failed" << std::endl; result = false; @@ -175,8 +175,8 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, on /* |I - Z' B Z| < n O(eps) */ std::vector ZBZ(n * n); int64_t ldzbz = n; - reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, - n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, + n, n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); if (!rel_id_err_check(n, ZBZ, ldzbz)) { test_log::lout << "Orthogonality check failed" << std::endl; result = false; @@ -213,8 +213,8 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, on /* |I - Z' B^-1 Z| = |I - Z' C| < n O(eps) */ std::vector ZhC(n * n); int64_t ldzhc = n; - reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, - n, n, 1.0, Z.data(), ldz, C.data(), ldc, 0.0, ZhC.data(), ldzhc); + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, + n, n, n, 1.0, Z.data(), ldz, C.data(), ldc, 0.0, ZhC.data(), ldzhc); if (!rel_id_err_check(n, ZhC, ldzhc)) { test_log::lout << "Orthogonality check failed" << std::endl; result = false; @@ -274,9 +274,9 @@ bool usm_dependency(const sycl::device& dev, int64_t itype, oneapi::math::job jo scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::hegvd, itype, jobz, uplo, - n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size, - std::vector{ in_event }); + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::hegvd, itype, jobz, + uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, + scratchpad_size, std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/lapack/source/hetrd.cpp b/tests/unit_tests/lapack/source/hetrd.cpp index 4a5bb0345..a5b078eeb 100644 --- a/tests/unit_tests/lapack/source/hetrd.cpp +++ b/tests/unit_tests/lapack/source/hetrd.cpp @@ -79,7 +79,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 #ifdef CALL_RT_API oneapi::math::lapack::hetrd(queue, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, - scratchpad_dev, scratchpad_size); + scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::hetrd, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, scratchpad_dev, scratchpad_size); diff --git a/tests/unit_tests/lapack/source/hetrf.cpp b/tests/unit_tests/lapack/source/hetrf.cpp index f22384b2c..72c767b74 100644 --- a/tests/unit_tests/lapack/source/hetrf.cpp +++ b/tests/unit_tests/lapack/source/hetrf.cpp @@ -77,7 +77,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 #ifdef CALL_RT_API oneapi::math::lapack::hetrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::hetrf, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); @@ -117,9 +117,9 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 reference::swap(n, Uk.data() + (k + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, - n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), - ldu); + reference::gemm(oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, + Uk.data(), ldu, 0.0, U.data(), ldu); D[k + k * ldd] = A[k + k * lda]; k -= 1; @@ -135,9 +135,9 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 reference::swap(n, Uk.data() + (k - 1 + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, - n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), - ldu); + reference::gemm(oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, + Uk.data(), ldu, 0.0, U.data(), ldu); D[k + k * ldd] = A[k + k * lda]; D[k - 1 + (k - 1) * ldd] = A[k - 1 + (k - 1) * lda]; @@ -160,9 +160,9 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 reference::swap(n, Uk.data() + (k + 0 * lda), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, - n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), - ldu); + reference::gemm(oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, + Uk.data(), ldu, 0.0, U.data(), ldu); D[k + (k)*ldd] = A[k + (k)*lda]; k += 1; @@ -178,9 +178,9 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 reference::swap(n, Uk.data() + (k + 1 + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, - n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), - ldu); + reference::gemm(oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, + Uk.data(), ldu, 0.0, U.data(), ldu); D[k + k * ldd] = A[k + k * lda]; D[k + 1 + (k + 1) * ldd] = A[k + 1 + (k + 1) * lda]; @@ -253,7 +253,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::hetrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::hetrf, uplo, n, A_dev, diff --git a/tests/unit_tests/lapack/source/orgbr.cpp b/tests/unit_tests/lapack/source/orgbr.cpp index 1ed2ab0fd..3f78a9653 100644 --- a/tests/unit_tests/lapack/source/orgbr.cpp +++ b/tests/unit_tests/lapack/source/orgbr.cpp @@ -94,7 +94,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::generate vect, int64_t m, i #ifdef CALL_RT_API oneapi::math::lapack::orgbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::orgbr, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); @@ -172,7 +172,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::generate vect, int64_ #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::orgbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::orgbr, vect, m, n, k, diff --git a/tests/unit_tests/lapack/source/orgqr.cpp b/tests/unit_tests/lapack/source/orgqr.cpp index f8ddc4f51..8f1da86b8 100644 --- a/tests/unit_tests/lapack/source/orgqr.cpp +++ b/tests/unit_tests/lapack/source/orgqr.cpp @@ -82,7 +82,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t #ifdef CALL_RT_API oneapi::math::lapack::orgqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::orgqr, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); @@ -146,7 +146,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::orgqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::orgqr, m, n, k, A_dev, diff --git a/tests/unit_tests/lapack/source/orgqr_batch_group.cpp b/tests/unit_tests/lapack/source/orgqr_batch_group.cpp index 4f44c9cf0..172471725 100644 --- a/tests/unit_tests/lapack/source/orgqr_batch_group.cpp +++ b/tests/unit_tests/lapack/source/orgqr_batch_group.cpp @@ -131,8 +131,8 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { #ifdef CALL_RT_API oneapi::math::lapack::orgqr_batch(queue, m_vec.data(), n_vec.data(), k_vec.data(), - A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, - group_sizes_vec.data(), scratchpad_dev, scratchpad_size); + A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, + group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::orgqr_batch, m_vec.data(), n_vec.data(), k_vec.data(), A_dev_ptrs, lda_vec.data(), diff --git a/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp b/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp index c88b6e88c..8ac2b2dff 100644 --- a/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/orgqr_batch_stride.cpp @@ -83,7 +83,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t #ifdef CALL_RT_API oneapi::math::lapack::orgqr_batch(queue, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, - batch_size, scratchpad_dev, scratchpad_size); + batch_size, scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::orgqr_batch, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, diff --git a/tests/unit_tests/lapack/source/orgtr.cpp b/tests/unit_tests/lapack/source/orgtr.cpp index 4a909d04e..154402156 100644 --- a/tests/unit_tests/lapack/source/orgtr.cpp +++ b/tests/unit_tests/lapack/source/orgtr.cpp @@ -80,7 +80,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 #ifdef CALL_RT_API oneapi::math::lapack::orgtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::orgtr, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); @@ -147,7 +147,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::orgtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::orgtr, uplo, n, A_dev, diff --git a/tests/unit_tests/lapack/source/ormqr.cpp b/tests/unit_tests/lapack/source/ormqr.cpp index 5181d9a3b..1c27a5cd7 100644 --- a/tests/unit_tests/lapack/source/ormqr.cpp +++ b/tests/unit_tests/lapack/source/ormqr.cpp @@ -92,7 +92,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::side left_right, oneapi::ma #ifdef CALL_RT_API oneapi::math::lapack::ormqr(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, - ldc, scratchpad_dev, scratchpad_size); + ldc, scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ormqr, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); @@ -185,8 +185,8 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::side left_right, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ormqr, left_right, trans, - m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ormqr, left_right, + trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/lapack/source/ormrq.cpp b/tests/unit_tests/lapack/source/ormrq.cpp index 375632336..f63dca6a8 100644 --- a/tests/unit_tests/lapack/source/ormrq.cpp +++ b/tests/unit_tests/lapack/source/ormrq.cpp @@ -102,7 +102,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::side left_right, oneapi::ma #ifdef CALL_RT_API oneapi::math::lapack::ormrq(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, - ldc, scratchpad_dev, scratchpad_size); + ldc, scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ormrq, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); @@ -194,8 +194,8 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::side left_right, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ormrq, left_right, trans, - m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ormrq, left_right, + trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/lapack/source/ormtr.cpp b/tests/unit_tests/lapack/source/ormtr.cpp index fd37974ea..dde332b32 100644 --- a/tests/unit_tests/lapack/source/ormtr.cpp +++ b/tests/unit_tests/lapack/source/ormtr.cpp @@ -90,10 +90,10 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t m, int64 #ifdef CALL_RT_API oneapi::math::lapack::ormtr(queue, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, - scratchpad_dev, scratchpad_size); + scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ormtr, side, uplo, trans, m, n, A_dev, - lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ormtr, side, uplo, trans, m, n, + A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -184,8 +184,8 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t m, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ormtr, side, uplo, trans, - m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ormtr, side, uplo, + trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/lapack/source/potrf.cpp b/tests/unit_tests/lapack/source/potrf.cpp index 19063dfcf..2cd37de30 100644 --- a/tests/unit_tests/lapack/source/potrf.cpp +++ b/tests/unit_tests/lapack/source/potrf.cpp @@ -130,7 +130,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::potrf(queue, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size, - std::vector{ in_event }); + std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::potrf, uplo, n, A_dev, diff --git a/tests/unit_tests/lapack/source/potrf_batch_group.cpp b/tests/unit_tests/lapack/source/potrf_batch_group.cpp index ce2fbc718..90ceaa5bf 100644 --- a/tests/unit_tests/lapack/source/potrf_batch_group.cpp +++ b/tests/unit_tests/lapack/source/potrf_batch_group.cpp @@ -47,7 +47,7 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { /* Test Parameters */ std::vector uplo_vec = { oneapi::math::uplo::upper, - oneapi::math::uplo::lower }; + oneapi::math::uplo::lower }; std::vector n_vec = { 4, 4 }; std::vector lda_vec = { 5, 5 }; std::vector group_sizes_vec = { 2, 2 }; @@ -114,8 +114,8 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { #ifdef CALL_RT_API oneapi::math::lapack::potrf_batch(queue, uplo_vec.data(), n_vec.data(), A_dev_ptrs, - lda_vec.data(), group_count, group_sizes_vec.data(), - scratchpad_dev, scratchpad_size); + lda_vec.data(), group_count, group_sizes_vec.data(), + scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potrf_batch, uplo_vec.data(), n_vec.data(), A_dev_ptrs, lda_vec.data(), group_count, diff --git a/tests/unit_tests/lapack/source/potrf_batch_stride.cpp b/tests/unit_tests/lapack/source/potrf_batch_stride.cpp index f893a0f56..15a5d8296 100644 --- a/tests/unit_tests/lapack/source/potrf_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/potrf_batch_stride.cpp @@ -73,7 +73,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 #ifdef CALL_RT_API oneapi::math::lapack::potrf_batch(queue, uplo, n, A_dev, lda, stride_a, batch_size, - scratchpad_dev, scratchpad_size); + scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potrf_batch, uplo, n, A_dev, lda, stride_a, batch_size, scratchpad_dev, scratchpad_size); diff --git a/tests/unit_tests/lapack/source/potri.cpp b/tests/unit_tests/lapack/source/potri.cpp index 3f10bad09..a76958e4e 100644 --- a/tests/unit_tests/lapack/source/potri.cpp +++ b/tests/unit_tests/lapack/source/potri.cpp @@ -164,7 +164,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::potri(queue, uplo, n, A_dev, lda, scratchpad_dev, scratchpad_size, - std::vector{ in_event }); + std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::potri, uplo, n, A_dev, diff --git a/tests/unit_tests/lapack/source/potrs.cpp b/tests/unit_tests/lapack/source/potrs.cpp index d3b8abf0c..7c7a9b07b 100644 --- a/tests/unit_tests/lapack/source/potrs.cpp +++ b/tests/unit_tests/lapack/source/potrs.cpp @@ -40,8 +40,8 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t nrhs, int64_t lda, - int64_t ldb, uint64_t seed) { +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t nrhs, + int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -83,7 +83,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 #ifdef CALL_RT_API oneapi::math::lapack::potrs(queue, uplo, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potrs, uplo, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size); @@ -151,9 +151,9 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, /* Check dependency handling */ auto in_event = create_dependency(queue); #ifdef CALL_RT_API - sycl::event func_event = - oneapi::math::lapack::potrs(queue, uplo, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + sycl::event func_event = oneapi::math::lapack::potrs( + queue, uplo, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size, + std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::potrs, uplo, n, nrhs, diff --git a/tests/unit_tests/lapack/source/potrs_batch_group.cpp b/tests/unit_tests/lapack/source/potrs_batch_group.cpp index 90359133d..f469e69c1 100644 --- a/tests/unit_tests/lapack/source/potrs_batch_group.cpp +++ b/tests/unit_tests/lapack/source/potrs_batch_group.cpp @@ -47,7 +47,7 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { /* Test Parameters */ std::vector uplo_vec = { oneapi::math::uplo::upper, - oneapi::math::uplo::lower }; + oneapi::math::uplo::lower }; std::vector n_vec = { 4, 5 }; std::vector nrhs_vec = { 9, 6 }; std::vector lda_vec = { 6, 6 }; @@ -144,9 +144,9 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { #ifdef CALL_RT_API oneapi::math::lapack::potrs_batch(queue, uplo_vec.data(), n_vec.data(), nrhs_vec.data(), - A_dev_ptrs, lda_vec.data(), B_dev_ptrs, ldb_vec.data(), - group_count, group_sizes_vec.data(), scratchpad_dev, - scratchpad_size); + A_dev_ptrs, lda_vec.data(), B_dev_ptrs, ldb_vec.data(), + group_count, group_sizes_vec.data(), scratchpad_dev, + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potrs_batch, uplo_vec.data(), n_vec.data(), nrhs_vec.data(), A_dev_ptrs, lda_vec.data(), diff --git a/tests/unit_tests/lapack/source/potrs_batch_stride.cpp b/tests/unit_tests/lapack/source/potrs_batch_stride.cpp index 6affb7105..a46f2816b 100644 --- a/tests/unit_tests/lapack/source/potrs_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/potrs_batch_stride.cpp @@ -40,8 +40,9 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t nrhs, int64_t lda, - int64_t stride_a, int64_t ldb, int64_t stride_b, int64_t batch_size, uint64_t seed) { +bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64_t nrhs, + int64_t lda, int64_t stride_a, int64_t ldb, int64_t stride_b, int64_t batch_size, + uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -88,7 +89,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 #ifdef CALL_RT_API oneapi::math::lapack::potrs_batch(queue, uplo, n, nrhs, A_dev, lda, stride_a, B_dev, ldb, - stride_b, batch_size, scratchpad_dev, scratchpad_size); + stride_b, batch_size, scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::potrs_batch, uplo, n, nrhs, A_dev, lda, stride_a, B_dev, ldb, stride_b, batch_size, scratchpad_dev, diff --git a/tests/unit_tests/lapack/source/syevd.cpp b/tests/unit_tests/lapack/source/syevd.cpp index c0de25b70..5e013aead 100644 --- a/tests/unit_tests/lapack/source/syevd.cpp +++ b/tests/unit_tests/lapack/source/syevd.cpp @@ -73,7 +73,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::job jobz, oneapi::math::upl #ifdef CALL_RT_API oneapi::math::lapack::syevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::syevd, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, scratchpad_size); @@ -134,7 +134,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::job jobz, oneapi::mat #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::syevd(queue, jobz, uplo, n, A_dev, lda, w_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::syevd, jobz, uplo, n, diff --git a/tests/unit_tests/lapack/source/sygvd.cpp b/tests/unit_tests/lapack/source/sygvd.cpp index ce671feb2..b1f73ada6 100644 --- a/tests/unit_tests/lapack/source/sygvd.cpp +++ b/tests/unit_tests/lapack/source/sygvd.cpp @@ -42,8 +42,8 @@ const char* accuracy_input = R"( )"; template -bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, oneapi::math::uplo uplo, - int64_t n, int64_t lda, int64_t ldb, uint64_t seed) { +bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, + oneapi::math::uplo uplo, int64_t n, int64_t lda, int64_t ldb, uint64_t seed) { using fp = typename data_T_info::value_type; using fp_real = typename complex_info::real_type; @@ -80,7 +80,7 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, on #ifdef CALL_RT_API oneapi::math::lapack::sygvd(queue, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, - scratchpad_dev, scratchpad_size); + scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::sygvd, itype, jobz, uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size); @@ -107,9 +107,9 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, on /* |D_ref - D| < |D_ref| O(eps) */ std::vector D_ref(n); - auto info = - reference::sygvd(itype, oneapi::math::job::novec, uplo, n, std::vector(A_initial).data(), - lda, std::vector(B_initial).data(), ldb, D_ref.data()); + auto info = reference::sygvd(itype, oneapi::math::job::novec, uplo, n, + std::vector(A_initial).data(), lda, + std::vector(B_initial).data(), ldb, D_ref.data()); if (0 != info) { test_log::lout << "reference sygvd failed with info = " << info << std::endl; return false; @@ -146,8 +146,8 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, on /* |I - Z' B Z| < n O(eps) */ std::vector ZBZ(n * n); int64_t ldzbz = n; - reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, - n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, + n, n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); if (!rel_id_err_check(n, ZBZ, ldzbz)) { test_log::lout << "Orthogonality check failed" << std::endl; result = false; @@ -180,8 +180,8 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, on /* |I - Z' B Z| < n O(eps) */ std::vector ZBZ(n * n); int64_t ldzbz = n; - reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, - n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, + n, n, n, 1.0, Z.data(), ldz, BZ.data(), ldbz, 0.0, ZBZ.data(), ldzbz); if (!rel_id_err_check(n, ZBZ, ldzbz)) { test_log::lout << "Orthogonality check failed" << std::endl; result = false; @@ -218,8 +218,8 @@ bool accuracy(const sycl::device& dev, int64_t itype, oneapi::math::job jobz, on /* |I - Z' B^-1 Z| = |I - Z' C| < n O(eps) */ std::vector ZhC(n * n); int64_t ldzhc = n; - reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, n, - n, n, 1.0, Z.data(), ldz, C.data(), ldc, 0.0, ZhC.data(), ldzhc); + reference::gemm(oneapi::math::transpose::conjtrans, oneapi::math::transpose::nontrans, + n, n, n, 1.0, Z.data(), ldz, C.data(), ldc, 0.0, ZhC.data(), ldzhc); if (!rel_id_err_check(n, ZhC, ldzhc)) { test_log::lout << "Orthogonality check failed" << std::endl; result = false; @@ -279,9 +279,9 @@ bool usm_dependency(const sycl::device& dev, int64_t itype, oneapi::math::job jo scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::sygvd, itype, jobz, uplo, - n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, scratchpad_size, - std::vector{ in_event }); + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::sygvd, itype, jobz, + uplo, n, A_dev, lda, B_dev, ldb, w_dev, scratchpad_dev, + scratchpad_size, std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/lapack/source/sytrd.cpp b/tests/unit_tests/lapack/source/sytrd.cpp index 3b804d84b..a9cd82d48 100644 --- a/tests/unit_tests/lapack/source/sytrd.cpp +++ b/tests/unit_tests/lapack/source/sytrd.cpp @@ -79,7 +79,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 #ifdef CALL_RT_API oneapi::math::lapack::sytrd(queue, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, - scratchpad_dev, scratchpad_size); + scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::sytrd, uplo, n, A_dev, lda, d_dev, e_dev, tau_dev, scratchpad_dev, scratchpad_size); diff --git a/tests/unit_tests/lapack/source/sytrf.cpp b/tests/unit_tests/lapack/source/sytrf.cpp index e24fba32e..39d17a0d4 100644 --- a/tests/unit_tests/lapack/source/sytrf.cpp +++ b/tests/unit_tests/lapack/source/sytrf.cpp @@ -74,7 +74,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 #ifdef CALL_RT_API oneapi::math::lapack::sytrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::sytrf, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, scratchpad_size); @@ -114,9 +114,9 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 reference::swap(n, Uk.data() + (k + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, - n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), - ldu); + reference::gemm(oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, + Uk.data(), ldu, 0.0, U.data(), ldu); D[k + k * ldd] = A[k + k * lda]; k -= 1; @@ -132,9 +132,9 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 reference::swap(n, Uk.data() + (k - 1 + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, - n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), - ldu); + reference::gemm(oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, + Uk.data(), ldu, 0.0, U.data(), ldu); D[k + k * ldd] = A[k + k * lda]; D[k - 1 + (k - 1) * ldd] = A[k - 1 + (k - 1) * lda]; @@ -157,9 +157,9 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 reference::swap(n, Uk.data() + (k + 0 * lda), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, - n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), - ldu); + reference::gemm(oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, + Uk.data(), ldu, 0.0, U.data(), ldu); D[k + (k)*ldd] = A[k + (k)*lda]; k += 1; @@ -175,9 +175,9 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 reference::swap(n, Uk.data() + (k + 1 + 0 * ldu), ldu, Uk.data() + (piv + 0 * ldu), ldu); auto U_temp = U; - reference::gemm(oneapi::math::transpose::nontrans, oneapi::math::transpose::nontrans, - n, n, n, 1.0, U_temp.data(), ldu, Uk.data(), ldu, 0.0, U.data(), - ldu); + reference::gemm(oneapi::math::transpose::nontrans, + oneapi::math::transpose::nontrans, n, n, n, 1.0, U_temp.data(), ldu, + Uk.data(), ldu, 0.0, U.data(), ldu); D[k + k * ldd] = A[k + k * lda]; D[k + 1 + (k + 1) * ldd] = A[k + 1 + (k + 1) * lda]; @@ -250,7 +250,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::sytrf(queue, uplo, n, A_dev, lda, ipiv_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::sytrf, uplo, n, A_dev, diff --git a/tests/unit_tests/lapack/source/trtrs.cpp b/tests/unit_tests/lapack/source/trtrs.cpp index 4857e3892..3f847028c 100644 --- a/tests/unit_tests/lapack/source/trtrs.cpp +++ b/tests/unit_tests/lapack/source/trtrs.cpp @@ -86,7 +86,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, oneapi::math::tr #ifdef CALL_RT_API oneapi::math::lapack::trtrs(queue, uplo, trans, diag, n, nrhs, A_dev, lda, B_dev, ldb, - scratchpad_dev, scratchpad_size); + scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::trtrs, uplo, trans, diag, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size); @@ -155,9 +155,9 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, oneapi::ma scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::trtrs, uplo, trans, diag, - n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, scratchpad_size, - std::vector{ in_event }); + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::trtrs, uplo, trans, + diag, n, nrhs, A_dev, lda, B_dev, ldb, scratchpad_dev, + scratchpad_size, std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/lapack/source/ungbr.cpp b/tests/unit_tests/lapack/source/ungbr.cpp index 723429025..f73085664 100644 --- a/tests/unit_tests/lapack/source/ungbr.cpp +++ b/tests/unit_tests/lapack/source/ungbr.cpp @@ -94,7 +94,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::generate vect, int64_t m, i #ifdef CALL_RT_API oneapi::math::lapack::ungbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ungbr, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); @@ -172,7 +172,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::generate vect, int64_ #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::ungbr(queue, vect, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ungbr, vect, m, n, k, diff --git a/tests/unit_tests/lapack/source/ungqr.cpp b/tests/unit_tests/lapack/source/ungqr.cpp index 63737914a..115b2e5e5 100644 --- a/tests/unit_tests/lapack/source/ungqr.cpp +++ b/tests/unit_tests/lapack/source/ungqr.cpp @@ -81,7 +81,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t #ifdef CALL_RT_API oneapi::math::lapack::ungqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ungqr, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); @@ -145,7 +145,7 @@ bool usm_dependency(const sycl::device& dev, int64_t m, int64_t n, int64_t k, in #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::ungqr(queue, m, n, k, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ungqr, m, n, k, A_dev, diff --git a/tests/unit_tests/lapack/source/ungqr_batch_group.cpp b/tests/unit_tests/lapack/source/ungqr_batch_group.cpp index 0b188ce1c..9ed03c1cf 100644 --- a/tests/unit_tests/lapack/source/ungqr_batch_group.cpp +++ b/tests/unit_tests/lapack/source/ungqr_batch_group.cpp @@ -131,8 +131,8 @@ bool accuracy(const sycl::device& dev, uint64_t seed) { #ifdef CALL_RT_API oneapi::math::lapack::ungqr_batch(queue, m_vec.data(), n_vec.data(), k_vec.data(), - A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, - group_sizes_vec.data(), scratchpad_dev, scratchpad_size); + A_dev_ptrs, lda_vec.data(), tau_dev_ptrs, group_count, + group_sizes_vec.data(), scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ungqr_batch, m_vec.data(), n_vec.data(), k_vec.data(), A_dev_ptrs, lda_vec.data(), diff --git a/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp b/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp index 782648c8d..446bfd8f8 100644 --- a/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp +++ b/tests/unit_tests/lapack/source/ungqr_batch_stride.cpp @@ -83,7 +83,7 @@ bool accuracy(const sycl::device& dev, int64_t m, int64_t n, int64_t k, int64_t #ifdef CALL_RT_API oneapi::math::lapack::ungqr_batch(queue, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, - batch_size, scratchpad_dev, scratchpad_size); + batch_size, scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ungqr_batch, m, n, k, A_dev, lda, stride_a, tau_dev, stride_tau, batch_size, scratchpad_dev, diff --git a/tests/unit_tests/lapack/source/ungtr.cpp b/tests/unit_tests/lapack/source/ungtr.cpp index 628ec95c0..ed516b03c 100644 --- a/tests/unit_tests/lapack/source/ungtr.cpp +++ b/tests/unit_tests/lapack/source/ungtr.cpp @@ -80,7 +80,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, int64 #ifdef CALL_RT_API oneapi::math::lapack::ungtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size); + scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::ungtr, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, scratchpad_size); @@ -147,7 +147,7 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t n, #ifdef CALL_RT_API sycl::event func_event = oneapi::math::lapack::ungtr(queue, uplo, n, A_dev, lda, tau_dev, scratchpad_dev, - scratchpad_size, std::vector{ in_event }); + scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::ungtr, uplo, n, A_dev, diff --git a/tests/unit_tests/lapack/source/unmqr.cpp b/tests/unit_tests/lapack/source/unmqr.cpp index d613751e3..c0a147988 100644 --- a/tests/unit_tests/lapack/source/unmqr.cpp +++ b/tests/unit_tests/lapack/source/unmqr.cpp @@ -92,7 +92,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::side left_right, oneapi::ma #ifdef CALL_RT_API oneapi::math::lapack::unmqr(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, - ldc, scratchpad_dev, scratchpad_size); + ldc, scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::unmqr, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); @@ -185,8 +185,8 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::side left_right, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::unmqr, left_right, trans, - m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::unmqr, left_right, + trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/lapack/source/unmrq.cpp b/tests/unit_tests/lapack/source/unmrq.cpp index 41d99a389..b41d5eda3 100644 --- a/tests/unit_tests/lapack/source/unmrq.cpp +++ b/tests/unit_tests/lapack/source/unmrq.cpp @@ -102,7 +102,7 @@ bool accuracy(const sycl::device& dev, oneapi::math::side left_right, oneapi::ma #ifdef CALL_RT_API oneapi::math::lapack::unmrq(queue, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, - ldc, scratchpad_dev, scratchpad_size); + ldc, scratchpad_dev, scratchpad_size); #else TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::unmrq, left_right, trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); @@ -194,8 +194,8 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::side left_right, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::unmrq, left_right, trans, - m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::unmrq, left_right, + trans, m, n, k, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/lapack/source/unmtr.cpp b/tests/unit_tests/lapack/source/unmtr.cpp index 00761c469..e1f969204 100644 --- a/tests/unit_tests/lapack/source/unmtr.cpp +++ b/tests/unit_tests/lapack/source/unmtr.cpp @@ -90,10 +90,10 @@ bool accuracy(const sycl::device& dev, oneapi::math::uplo uplo, int64_t m, int64 #ifdef CALL_RT_API oneapi::math::lapack::unmtr(queue, side, uplo, trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, - scratchpad_dev, scratchpad_size); + scratchpad_dev, scratchpad_size); #else - TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::unmtr, side, uplo, trans, m, n, A_dev, - lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); + TEST_RUN_LAPACK_CT_SELECT(queue, oneapi::math::lapack::unmtr, side, uplo, trans, m, n, + A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size); #endif queue.wait_and_throw(); @@ -184,8 +184,8 @@ bool usm_dependency(const sycl::device& dev, oneapi::math::uplo uplo, int64_t m, scratchpad_size, std::vector{ in_event }); #else sycl::event func_event; - TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::unmtr, side, uplo, trans, - m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, + TEST_RUN_LAPACK_CT_SELECT(queue, func_event = oneapi::math::lapack::unmtr, side, uplo, + trans, m, n, A_dev, lda, tau_dev, C_dev, ldc, scratchpad_dev, scratchpad_size, std::vector{ in_event }); #endif result = check_dependency(queue, in_event, func_event); diff --git a/tests/unit_tests/main_test.cpp b/tests/unit_tests/main_test.cpp index 18e6b07a7..df812e4c8 100644 --- a/tests/unit_tests/main_test.cpp +++ b/tests/unit_tests/main_test.cpp @@ -114,8 +114,7 @@ int main(int argc, char** argv) { unique_devices.insert(dev.get_info()); #if !defined(ONEMATH_ENABLE_MKLCPU_BACKEND) && \ !defined(ONEMATH_ENABLE_PORTBLAS_BACKEND_INTEL_CPU) && \ - !defined(ONEMATH_ENABLE_PORTFFT_BACKEND) && \ - !defined(ONEMATH_ENABLE_NETLIB_BACKEND) + !defined(ONEMATH_ENABLE_PORTFFT_BACKEND) && !defined(ONEMATH_ENABLE_NETLIB_BACKEND) if (dev.is_cpu()) continue; #endif @@ -125,24 +124,21 @@ int main(int argc, char** argv) { if (dev.is_gpu() && vendor_id == INTEL_ID) continue; #endif -#if !defined(ONEMATH_ENABLE_CUBLAS_BACKEND) && \ - !defined(ONEMATH_ENABLE_CURAND_BACKEND) && \ - !defined(ONEMATH_ENABLE_CUSOLVER_BACKEND) && \ - !defined(ONEMATH_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU) && \ +#if !defined(ONEMATH_ENABLE_CUBLAS_BACKEND) && !defined(ONEMATH_ENABLE_CURAND_BACKEND) && \ + !defined(ONEMATH_ENABLE_CUSOLVER_BACKEND) && \ + !defined(ONEMATH_ENABLE_PORTBLAS_BACKEND_NVIDIA_GPU) && \ !defined(ONEMATH_ENABLE_CUFFT_BACKEND) && !defined(ONEMATH_ENABLE_PORTFFT_BACKEND) if (dev.is_gpu() && vendor_id == NVIDIA_ID) continue; #endif -#if !defined(ONEMATH_ENABLE_ROCBLAS_BACKEND) && \ - !defined(ONEMATH_ENABLE_ROCRAND_BACKEND) && \ - !defined(ONEMATH_ENABLE_ROCSOLVER_BACKEND) && \ - !defined(ONEMATH_ENABLE_PORTBLAS_BACKEND_AMD_GPU) && \ - !defined(ONEMATH_ENABLE_ROCFFT_BACKEND) && \ - !defined(ONEMATH_ENABLE_PORTFFT_BACKEND) +#if !defined(ONEMATH_ENABLE_ROCBLAS_BACKEND) && !defined(ONEMATH_ENABLE_ROCRAND_BACKEND) && \ + !defined(ONEMATH_ENABLE_ROCSOLVER_BACKEND) && \ + !defined(ONEMATH_ENABLE_PORTBLAS_BACKEND_AMD_GPU) && \ + !defined(ONEMATH_ENABLE_ROCFFT_BACKEND) && !defined(ONEMATH_ENABLE_PORTFFT_BACKEND) if (dev.is_gpu() && vendor_id == AMD_ID) continue; #endif -// clang-format off + // clang-format off #ifdef __HIPSYCL__ if (dev.is_accelerator()) #else diff --git a/tests/unit_tests/rng/device/include/moments.hpp b/tests/unit_tests/rng/device/include/moments.hpp index 661e85f32..7b360d5c3 100644 --- a/tests/unit_tests/rng/device/include/moments.hpp +++ b/tests/unit_tests/rng/device/include/moments.hpp @@ -53,15 +53,15 @@ class moments_test { Distribution, oneapi::math::rng::device::uniform< std::int32_t, oneapi::math::rng::device::uniform_method::accurate>> || - std::is_same_v> || std::is_same_v< Distribution, oneapi::math::rng::device::poisson< - std::int32_t, oneapi::math::rng::device::poisson_method::devroye>>)&&!queue - .get_device() - .has(sycl::aspect::fp64)) { + std::uint32_t, oneapi::math::rng::device::poisson_method::devroye>> || + std::is_same_v< + Distribution, + oneapi::math::rng::device::poisson< + std::int32_t, oneapi::math::rng::device::poisson_method::devroye>>) && + !queue.get_device().has(sycl::aspect::fp64)) { status = test_skipped; return; } diff --git a/tests/unit_tests/rng/device/include/rng_device_test_common.hpp b/tests/unit_tests/rng/device/include/rng_device_test_common.hpp index ac646fdeb..5e373e2cf 100644 --- a/tests/unit_tests/rng/device/include/rng_device_test_common.hpp +++ b/tests/unit_tests/rng/device/include/rng_device_test_common.hpp @@ -34,10 +34,9 @@ #define N_GEN_SERVICE (N_ENGINES * N_PORTION) // defines for skip_ahead_ex tests -#define N_SKIP ((std::uint64_t)pow(2, 62)) -#define SKIP_TIMES ((std::int32_t)pow(2, 14)) -#define NUM_TO_SKIP \ - { 0, (std::uint64_t)pow(2, 12) } +#define N_SKIP ((std::uint64_t)pow(2, 62)) +#define SKIP_TIMES ((std::int32_t)pow(2, 14)) +#define NUM_TO_SKIP { 0, (std::uint64_t)pow(2, 12) } // Correctness checking. static inline bool check_equal_device(float x, float x_ref) { diff --git a/tests/unit_tests/rng/device/include/skip_ahead_test.hpp b/tests/unit_tests/rng/device/include/skip_ahead_test.hpp index fb8648248..775e57031 100644 --- a/tests/unit_tests/rng/device/include/skip_ahead_test.hpp +++ b/tests/unit_tests/rng/device/include/skip_ahead_test.hpp @@ -126,7 +126,7 @@ class skip_ahead_ex_test { size_t id = item.get_id(0); Engine engine(SEED); oneapi::math::rng::device::skip_ahead(engine, - { id * Engine::vec_size, skip_num }); + { id * Engine::vec_size, skip_num }); oneapi::math::rng::device::bits<> distr; auto res = oneapi::math::rng::device::generate(distr, engine); if constexpr (Engine::vec_size == 1) { diff --git a/tests/unit_tests/rng/device/moments/moments.cpp b/tests/unit_tests/rng/device/moments/moments.cpp index 6449b2d0c..a191b67df 100644 --- a/tests/unit_tests/rng/device/moments/moments.cpp +++ b/tests/unit_tests/rng/device/moments/moments.cpp @@ -1096,19 +1096,19 @@ TEST_P(Philox4x32x10BetaCjaDeviceMomentsTests, RealDoublePrecision) { } TEST_P(Philox4x32x10BetaCjaAccDeviceMomentsTests, RealSinglePrecision) { - rng_device_test, - oneapi::math::rng::device::beta>> + rng_device_test, + oneapi::math::rng::device::beta< + float, oneapi::math::rng::device::beta_method::cja_accurate>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::math::rng::device::beta>> + rng_device_test, + oneapi::math::rng::device::beta< + float, oneapi::math::rng::device::beta_method::cja_accurate>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::math::rng::device::beta>> + rng_device_test, + oneapi::math::rng::device::beta< + float, oneapi::math::rng::device::beta_method::cja_accurate>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -1148,19 +1148,19 @@ class Philox4x32x10GammaMarsagliaAccDeviceMomentsTests : public ::testing::TestWithParam {}; TEST_P(Philox4x32x10GammaMarsagliaDeviceMomentsTests, RealSinglePrecision) { - rng_device_test, - oneapi::math::rng::device::gamma>> + rng_device_test, + oneapi::math::rng::device::gamma< + float, oneapi::math::rng::device::gamma_method::marsaglia>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::math::rng::device::gamma>> + rng_device_test, + oneapi::math::rng::device::gamma< + float, oneapi::math::rng::device::gamma_method::marsaglia>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::math::rng::device::gamma>> + rng_device_test, + oneapi::math::rng::device::gamma< + float, oneapi::math::rng::device::gamma_method::marsaglia>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } @@ -1168,19 +1168,19 @@ TEST_P(Philox4x32x10GammaMarsagliaDeviceMomentsTests, RealSinglePrecision) { TEST_P(Philox4x32x10GammaMarsagliaDeviceMomentsTests, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_device_test, - oneapi::math::rng::device::gamma>> + rng_device_test, + oneapi::math::rng::device::gamma< + double, oneapi::math::rng::device::gamma_method::marsaglia>>> test1; EXPECT_TRUEORSKIP((test1(GetParam()))); - rng_device_test, - oneapi::math::rng::device::gamma>> + rng_device_test, + oneapi::math::rng::device::gamma< + double, oneapi::math::rng::device::gamma_method::marsaglia>>> test2; EXPECT_TRUEORSKIP((test2(GetParam()))); - rng_device_test, - oneapi::math::rng::device::gamma>> + rng_device_test, + oneapi::math::rng::device::gamma< + double, oneapi::math::rng::device::gamma_method::marsaglia>>> test3; EXPECT_TRUEORSKIP((test3(GetParam()))); } diff --git a/tests/unit_tests/rng/include/rng_test_common.hpp b/tests/unit_tests/rng/include/rng_test_common.hpp index 15c32777b..3d3601553 100644 --- a/tests/unit_tests/rng/include/rng_test_common.hpp +++ b/tests/unit_tests/rng/include/rng_test_common.hpp @@ -34,10 +34,9 @@ #define N_GEN_SERVICE (N_ENGINES * N_PORTION) // defines for skip_ahead_ex tests -#define N_SKIP ((std::uint64_t)pow(2, 62)) -#define SKIP_TIMES ((std::int32_t)pow(2, 14)) -#define NUM_TO_SKIP \ - { 0, (std::uint64_t)pow(2, 12) } +#define N_SKIP ((std::uint64_t)pow(2, 62)) +#define SKIP_TIMES ((std::int32_t)pow(2, 14)) +#define NUM_TO_SKIP { 0, (std::uint64_t)pow(2, 12) } // Correctness checking. static inline bool check_equal(float x, float x_ref) { diff --git a/tests/unit_tests/rng/statistics_check/gaussian.cpp b/tests/unit_tests/rng/statistics_check/gaussian.cpp index 9b693c44d..8f33db53d 100644 --- a/tests/unit_tests/rng/statistics_check/gaussian.cpp +++ b/tests/unit_tests/rng/statistics_check/gaussian.cpp @@ -30,14 +30,14 @@ class GaussianBoxmullerTest : public ::testing::TestWithParam {}; class GaussianIcdfTest : public ::testing::TestWithParam {}; TEST_P(GaussianIcdfTest, RealSinglePrecision) { - rng_test< - statistics_test, - oneapi::math::rng::philox4x32x10>> + rng_test, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, GAUSSIAN_ARGS_FLOAT))); - rng_test< - statistics_test, - oneapi::math::rng::mrg32k3a>> + rng_test, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, GAUSSIAN_ARGS_FLOAT))); } @@ -45,14 +45,14 @@ TEST_P(GaussianIcdfTest, RealSinglePrecision) { TEST_P(GaussianIcdfTest, RealDoublePrecision) { CHECK_DOUBLE_ON_DEVICE(GetParam()); - rng_test< - statistics_test, - oneapi::math::rng::philox4x32x10>> + rng_test, + oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, GAUSSIAN_ARGS_DOUBLE))); - rng_test< - statistics_test, - oneapi::math::rng::mrg32k3a>> + rng_test, + oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, GAUSSIAN_ARGS_DOUBLE))); } diff --git a/tests/unit_tests/rng/statistics_check/poisson_usm.cpp b/tests/unit_tests/rng/statistics_check/poisson_usm.cpp old mode 100755 new mode 100644 index 44a1e6b4c..6927669b0 --- a/tests/unit_tests/rng/statistics_check/poisson_usm.cpp +++ b/tests/unit_tests/rng/statistics_check/poisson_usm.cpp @@ -30,13 +30,13 @@ class PoissonIcdfUsmTests : public ::testing::TestWithParam {}; TEST_P(PoissonIcdfUsmTests, IntegerPrecision) { rng_test, + oneapi::math::rng::poisson_method::gaussian_icdf_based>, oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, POISSON_ARGS))); rng_test, + oneapi::math::rng::poisson_method::gaussian_icdf_based>, oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, POISSON_ARGS))); @@ -45,13 +45,13 @@ TEST_P(PoissonIcdfUsmTests, IntegerPrecision) { TEST_P(PoissonIcdfUsmTests, UnsignedIntegerPrecision) { rng_test, + oneapi::math::rng::poisson_method::gaussian_icdf_based>, oneapi::math::rng::philox4x32x10>> test1; EXPECT_TRUEORSKIP((test1(GetParam(), N_GEN, POISSON_ARGS))); rng_test, + oneapi::math::rng::poisson_method::gaussian_icdf_based>, oneapi::math::rng::mrg32k3a>> test2; EXPECT_TRUEORSKIP((test2(GetParam(), N_GEN, POISSON_ARGS))); diff --git a/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp b/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp index a02957905..584de713d 100644 --- a/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp +++ b/tests/unit_tests/sparse_blas/include/common_sparse_reference.hpp @@ -55,9 +55,9 @@ inline T opVal(const T t, const bool isConj) { }; template -void do_csr_transpose(const oneapi::math::transpose opA, intType *ia_t, intType *ja_t, fpType *a_t, - intType a_nrows, intType a_ncols, intType indexing, accIntType &ia, - accIntType &ja, accFpType &a, const bool structOnlyFlag = false) { +void do_csr_transpose(const oneapi::math::transpose opA, intType* ia_t, intType* ja_t, fpType* a_t, + intType a_nrows, intType a_ncols, intType indexing, accIntType& ia, + accIntType& ja, accFpType& a, const bool structOnlyFlag = false) { const bool isConj = (opA == oneapi::math::transpose::conjtrans); // initialize ia_t to zero @@ -105,7 +105,7 @@ void do_csr_transpose(const oneapi::math::transpose opA, intType *ia_t, intType // Transpose the given sparse matrix if needed template -auto sparse_transpose_if_needed(const intType *ia, const intType *ja, const fpType *a, +auto sparse_transpose_if_needed(const intType* ia, const intType* ja, const fpType* a, intType a_nrows, intType a_ncols, std::size_t nnz, intType indexing, oneapi::math::transpose transpose_val) { std::vector iopa; @@ -134,7 +134,7 @@ auto sparse_transpose_if_needed(const intType *ia, const intType *ja, const fpTy /// Reduce the leading dimension to the minimum and transpose the matrix if needed /// The outputted matrix always uses row major layout template -auto extract_dense_matrix(const fpType *x, std::size_t nrows, std::size_t ncols, std::size_t ld, +auto extract_dense_matrix(const fpType* x, std::size_t nrows, std::size_t ncols, std::size_t ld, oneapi::math::transpose transpose_val, oneapi::math::layout dense_matrix_layout) { const bool is_row_major = dense_matrix_layout == oneapi::math::layout::row_major; @@ -161,8 +161,8 @@ auto extract_dense_matrix(const fpType *x, std::size_t nrows, std::size_t ncols, /// Convert the sparse matrix in the given format to a dense matrix A in row major layout applied with A_view. template -std::vector sparse_to_dense(sparse_matrix_format_t format, const intType *ia, - const intType *ja, const fpType *a, std::size_t a_nrows, +std::vector sparse_to_dense(sparse_matrix_format_t format, const intType* ia, + const intType* ja, const fpType* a, std::size_t a_nrows, std::size_t a_ncols, std::size_t nnz, intType indexing, oneapi::math::transpose transpose_val, oneapi::math::sparse::matrix_view A_view) { diff --git a/tests/unit_tests/sparse_blas/include/test_common.hpp b/tests/unit_tests/sparse_blas/include/test_common.hpp index a9b8841d9..d64e36de2 100644 --- a/tests/unit_tests/sparse_blas/include/test_common.hpp +++ b/tests/unit_tests/sparse_blas/include/test_common.hpp @@ -66,16 +66,16 @@ static std::vector> test_matrix_ oneapi::math::sparse::matrix_property::symmetric } }; -void print_error_code(sycl::exception const &e); +void print_error_code(sycl::exception const& e); // Catch asynchronous exceptions. struct exception_handler_t { void operator()(sycl::exception_list exceptions) { - for (std::exception_ptr const &e : exceptions) { + for (std::exception_ptr const& e : exceptions) { try { std::rethrow_exception(e); } - catch (sycl::exception const &e) { + catch (sycl::exception const& e) { std::cout << "Caught asynchronous SYCL exception:\n" << e.what() << std::endl; print_error_code(e); } @@ -86,7 +86,7 @@ struct exception_handler_t { struct UsmDeleter { sycl::queue q; UsmDeleter(sycl::queue _q) : q(_q) {} - void operator()(void *ptr) { + void operator()(void* ptr) { sycl::free(ptr, q); } }; @@ -99,14 +99,14 @@ auto malloc_device_uptr(sycl::queue q, std::size_t num_elts) { // SYCL buffer creation helper. template -sycl::buffer make_buffer(const vec &v) { +sycl::buffer make_buffer(const vec& v) { sycl::buffer buf(v.data(), sycl::range<1>(v.size())); return buf; } template -void copy_host_to_buffer(sycl::queue queue, const std::vector &src, sycl::buffer dst) { - queue.submit([&](sycl::handler &cgh) { +void copy_host_to_buffer(sycl::queue queue, const std::vector& src, sycl::buffer dst) { + queue.submit([&](sycl::handler& cgh) { auto dst_acc = dst.template get_access( cgh, sycl::range<1>(src.size())); cgh.copy(src.data(), dst_acc); @@ -168,7 +168,7 @@ struct rand_scalar> { }; template -void rand_vector(std::vector &v, std::size_t n) { +void rand_vector(std::vector& v, std::size_t n) { using fpRealType = typename complex_info::real_type; v.resize(n); rand_scalar rand; @@ -178,7 +178,7 @@ void rand_vector(std::vector &v, std::size_t n) { } template -void rand_matrix(std::vector &m, oneapi::math::layout layout_val, std::size_t nrows, +void rand_matrix(std::vector& m, oneapi::math::layout layout_val, std::size_t nrows, std::size_t ncols, std::size_t ld, oneapi::math::transpose transpose_val = oneapi::math::transpose::nontrans) { using fpRealType = typename complex_info::real_type; @@ -221,8 +221,8 @@ fpType generate_data(bool is_diag) { template intType generate_random_csr_matrix(const intType nrows, const intType ncols, const double density_val, intType indexing, - std::vector &ia, std::vector &ja, - std::vector &a, bool is_symmetric, + std::vector& ia, std::vector& ja, + std::vector& a, bool is_symmetric, bool require_diagonal = false) { intType nnz = 0; rand_scalar rand_density; @@ -272,8 +272,8 @@ intType generate_random_csr_matrix(const intType nrows, const intType ncols, template intType generate_random_coo_matrix(const intType nrows, const intType ncols, const double density_val, intType indexing, - std::vector &ia, std::vector &ja, - std::vector &a, bool is_symmetric, + std::vector& ia, std::vector& ja, + std::vector& a, bool is_symmetric, bool require_diagonal = false) { rand_scalar rand_density; @@ -315,8 +315,8 @@ intType generate_random_coo_matrix(const intType nrows, const intType ncols, template intType generate_random_matrix(sparse_matrix_format_t format, const intType nrows, const intType ncols, const double density_val, intType indexing, - std::vector &ia, std::vector &ja, - std::vector &a, bool is_symmetric, + std::vector& ia, std::vector& ja, + std::vector& a, bool is_symmetric, bool require_diagonal = false) { ia.clear(); ja.clear(); @@ -337,8 +337,8 @@ intType generate_random_matrix(sparse_matrix_format_t format, const intType nrow /// In CSR format, the elements within a row are shuffled without changing ia. /// In COO format, all the elements are shuffled. template -void shuffle_sparse_matrix(sparse_matrix_format_t format, intType indexing, intType *ia, - intType *ja, fpType *a, intType nnz, std::size_t nrows) { +void shuffle_sparse_matrix(sparse_matrix_format_t format, intType indexing, intType* ia, + intType* ja, fpType* a, intType nnz, std::size_t nrows) { if (format == sparse_matrix_format_t::CSR) { for (std::size_t i = 0; i < nrows; ++i) { intType nnz_row = ia[i + 1] - ia[i]; @@ -361,14 +361,14 @@ void shuffle_sparse_matrix(sparse_matrix_format_t format, intType indexing, intT } else { throw oneapi::math::exception("sparse_blas", "shuffle_sparse_matrix", - "Internal error: unsupported format"); + "Internal error: unsupported format"); } } /// Initialize a sparse matrix specified by the given format template -void init_sparse_matrix(sycl::queue &queue, sparse_matrix_format_t format, - oneapi::math::sparse::matrix_handle_t *p_smhandle, std::int64_t num_rows, +void init_sparse_matrix(sycl::queue& queue, sparse_matrix_format_t format, + oneapi::math::sparse::matrix_handle_t* p_smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, ContainerIndexT rows, ContainerIndexT cols, ContainerValueT vals) { if (format == sparse_matrix_format_t::CSR) { @@ -381,33 +381,33 @@ void init_sparse_matrix(sycl::queue &queue, sparse_matrix_format_t format, } else { throw oneapi::math::exception("sparse_blas", "init_sparse_matrix", - "Internal error: unsupported format"); + "Internal error: unsupported format"); } } /// Reset the data of a sparse matrix specified by the given format template -void set_matrix_data(sycl::queue &queue, sparse_matrix_format_t format, +void set_matrix_data(sycl::queue& queue, sparse_matrix_format_t format, oneapi::math::sparse::matrix_handle_t smhandle, std::int64_t num_rows, std::int64_t num_cols, std::int64_t nnz, oneapi::math::index_base index, ContainerIndexT rows, ContainerIndexT cols, ContainerValueT vals) { if (format == sparse_matrix_format_t::CSR) { - CALL_RT_OR_CT(oneapi::math::sparse::set_csr_matrix_data, queue, smhandle, num_rows, num_cols, - nnz, index, rows, cols, vals); + CALL_RT_OR_CT(oneapi::math::sparse::set_csr_matrix_data, queue, smhandle, num_rows, + num_cols, nnz, index, rows, cols, vals); } else if (format == sparse_matrix_format_t::COO) { - CALL_RT_OR_CT(oneapi::math::sparse::set_coo_matrix_data, queue, smhandle, num_rows, num_cols, - nnz, index, rows, cols, vals); + CALL_RT_OR_CT(oneapi::math::sparse::set_coo_matrix_data, queue, smhandle, num_rows, + num_cols, nnz, index, rows, cols, vals); } else { throw oneapi::math::exception("sparse_blas", "set_matrix_data", - "Internal error: unsupported format"); + "Internal error: unsupported format"); } } template -inline void free_handles(sycl::queue &queue, const std::vector dependencies, - HandlesT &&... handles) { +inline void free_handles(sycl::queue& queue, const std::vector dependencies, + HandlesT&&... handles) { // Fold expression so that handles expands to each value one after the other. ( [&] { @@ -436,19 +436,19 @@ inline void free_handles(sycl::queue &queue, const std::vector depe } template -inline void free_handles(sycl::queue &queue, HandlesT &&... handles) { +inline void free_handles(sycl::queue& queue, HandlesT&&... handles) { free_handles(queue, {}, handles...); } template -inline void wait_and_free_handles(sycl::queue &queue, HandlesT &&... handles) { +inline void wait_and_free_handles(sycl::queue& queue, HandlesT&&... handles) { queue.wait(); free_handles(queue, handles...); } inline bool require_square_matrix( oneapi::math::sparse::matrix_view A_view, - const std::set &matrix_properties) { + const std::set& matrix_properties) { const bool is_symmetric = matrix_properties.find(oneapi::math::sparse::matrix_property::symmetric) != matrix_properties.cend(); @@ -457,7 +457,7 @@ inline bool require_square_matrix( template bool check_equal(fpType x, fpType x_ref, double abs_error_margin, double rel_error_margin, - std::ostream &out) { + std::ostream& out) { using fpRealType = typename complex_info::real_type; static_assert(std::is_floating_point_v, "Expected floating-point real or complex type."); @@ -478,8 +478,8 @@ bool check_equal(fpType x, fpType x_ref, double abs_error_margin, double rel_err } template -bool check_equal_vector(const vecType1 &v, const vecType2 &v_ref, double abs_error_factor = 10.0, - double rel_error_factor = 200.0, std::ostream &out = std::cout) { +bool check_equal_vector(const vecType1& v, const vecType2& v_ref, double abs_error_factor = 10.0, + double rel_error_factor = 200.0, std::ostream& out = std::cout) { using T = typename vecType2::value_type; std::size_t n = v.size(); if (n != v_ref.size()) { @@ -492,7 +492,7 @@ bool check_equal_vector(const vecType1 &v, const vecType2 &v_ref, double abs_err auto max_norm_ref = *std::max_element(std::begin(v_ref), std::end(v_ref), - [](const T &a, const T &b) { return std::abs(a) < std::abs(b); }); + [](const T& a, const T& b) { return std::abs(a) < std::abs(b); }); // Heuristic for the average-case error margins double abs_error_margin = abs_error_factor * std::abs(max_norm_ref) * std::log2(static_cast(n)); diff --git a/tests/unit_tests/sparse_blas/include/test_spmm.hpp b/tests/unit_tests/sparse_blas/include/test_spmm.hpp index e34910521..83fa00d99 100644 --- a/tests/unit_tests/sparse_blas/include/test_spmm.hpp +++ b/tests/unit_tests/sparse_blas/include/test_spmm.hpp @@ -53,11 +53,11 @@ */ template void test_helper_with_format_with_transpose( - testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device *dev, + testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device* dev, sparse_matrix_format_t format, - const std::vector &non_default_algorithms, - oneapi::math::transpose transpose_A, oneapi::math::transpose transpose_B, int &num_passed, - int &num_skipped) { + const std::vector& non_default_algorithms, + oneapi::math::transpose transpose_A, oneapi::math::transpose transpose_B, int& num_passed, + int& num_skipped) { double density_A_matrix = 0.8; fpType fp_zero = set_fp_value()(0.f, 0.f); fpType fp_one = set_fp_value()(1.f, 0.f); @@ -217,13 +217,13 @@ void test_helper_with_format_with_transpose( */ template void test_helper_with_format( - testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device *dev, + testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device* dev, sparse_matrix_format_t format, - const std::vector &non_default_algorithms, int &num_passed, - int &num_skipped) { + const std::vector& non_default_algorithms, int& num_passed, + int& num_skipped) { std::vector transpose_vals{ oneapi::math::transpose::nontrans, - oneapi::math::transpose::trans, - oneapi::math::transpose::conjtrans }; + oneapi::math::transpose::trans, + oneapi::math::transpose::conjtrans }; for (auto transpose_A : transpose_vals) { for (auto transpose_B : transpose_vals) { test_helper_with_format_with_transpose( @@ -245,7 +245,7 @@ void test_helper_with_format( */ template void test_helper(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, - sycl::device *dev, int &num_passed, int &num_skipped) { + sycl::device* dev, int& num_passed, int& num_skipped) { test_helper_with_format( test_functor_i32, test_functor_i64, dev, sparse_matrix_format_t::CSR, { oneapi::math::sparse::spmm_alg::no_optimize_alg, oneapi::math::sparse::spmm_alg::csr_alg1, @@ -261,14 +261,14 @@ void test_helper(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i6 /// Compute spmm reference as a dense operation template -void prepare_reference_spmm_data(sparse_matrix_format_t format, const intType *ia, - const intType *ja, const fpType *a, intType a_nrows, +void prepare_reference_spmm_data(sparse_matrix_format_t format, const intType* ia, + const intType* ja, const fpType* a, intType a_nrows, intType a_ncols, intType c_ncols, intType a_nnz, intType indexing, oneapi::math::layout dense_matrix_layout, oneapi::math::transpose opA, oneapi::math::transpose opB, fpType alpha, fpType beta, intType ldb, intType ldc, - const fpType *b, oneapi::math::sparse::matrix_view A_view, - fpType *c_ref) { + const fpType* b, oneapi::math::sparse::matrix_view A_view, + fpType* c_ref) { std::size_t a_nrows_u = static_cast(a_nrows); std::size_t a_ncols_u = static_cast(a_ncols); std::size_t c_ncols_u = static_cast(c_ncols); @@ -286,7 +286,7 @@ void prepare_reference_spmm_data(sparse_matrix_format_t format, const intType *i // Return the linear index to access a dense matrix from auto dense_linear_idx = [=](std::size_t row, std::size_t col, std::size_t ld) { return (dense_matrix_layout == oneapi::math::layout::row_major) ? row * ld + col - : col * ld + row; + : col * ld + row; }; // @@ -300,7 +300,7 @@ void prepare_reference_spmm_data(sparse_matrix_format_t format, const intType *i for (std::size_t i = 0; i < opa_ncols; i++) { acc += dense_opa[row * opa_ncols + i] * dense_opb[i * c_ncols_u + col]; } - fpType &c = c_ref[dense_linear_idx(row, col, ldc_u)]; + fpType& c = c_ref[dense_linear_idx(row, col, ldc_u)]; c = alpha * acc + beta * c; } } diff --git a/tests/unit_tests/sparse_blas/include/test_spmv.hpp b/tests/unit_tests/sparse_blas/include/test_spmv.hpp index 848180c28..d44e8ba7e 100644 --- a/tests/unit_tests/sparse_blas/include/test_spmv.hpp +++ b/tests/unit_tests/sparse_blas/include/test_spmv.hpp @@ -52,10 +52,10 @@ */ template void test_helper_with_format_with_transpose( - testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device *dev, + testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device* dev, sparse_matrix_format_t format, - const std::vector &non_default_algorithms, - oneapi::math::transpose transpose_val, int &num_passed, int &num_skipped) { + const std::vector& non_default_algorithms, + oneapi::math::transpose transpose_val, int& num_passed, int& num_skipped) { double density_A_matrix = 0.8; fpType fp_zero = set_fp_value()(0.f, 0.f); fpType fp_one = set_fp_value()(1.f, 0.f); @@ -212,13 +212,13 @@ void test_helper_with_format_with_transpose( */ template void test_helper_with_format( - testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device *dev, + testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, sycl::device* dev, sparse_matrix_format_t format, - const std::vector &non_default_algorithms, int &num_passed, - int &num_skipped) { + const std::vector& non_default_algorithms, int& num_passed, + int& num_skipped) { std::vector transpose_vals{ oneapi::math::transpose::nontrans, - oneapi::math::transpose::trans, - oneapi::math::transpose::conjtrans }; + oneapi::math::transpose::trans, + oneapi::math::transpose::conjtrans }; for (auto transpose_A : transpose_vals) { test_helper_with_format_with_transpose(test_functor_i32, test_functor_i64, dev, format, non_default_algorithms, transpose_A, @@ -238,7 +238,7 @@ void test_helper_with_format( */ template void test_helper(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, - sycl::device *dev, int &num_passed, int &num_skipped) { + sycl::device* dev, int& num_passed, int& num_skipped) { test_helper_with_format( test_functor_i32, test_functor_i64, dev, sparse_matrix_format_t::CSR, { oneapi::math::sparse::spmv_alg::no_optimize_alg, oneapi::math::sparse::spmv_alg::csr_alg1, @@ -253,12 +253,12 @@ void test_helper(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i6 /// Compute spmv reference as a dense operation template -void prepare_reference_spmv_data(sparse_matrix_format_t format, const intType *ia, - const intType *ja, const fpType *a, intType a_nrows, +void prepare_reference_spmv_data(sparse_matrix_format_t format, const intType* ia, + const intType* ja, const fpType* a, intType a_nrows, intType a_ncols, intType a_nnz, intType indexing, oneapi::math::transpose opA, fpType alpha, fpType beta, - const fpType *x, oneapi::math::sparse::matrix_view A_view, - fpType *y_ref) { + const fpType* x, oneapi::math::sparse::matrix_view A_view, + fpType* y_ref) { std::size_t a_nrows_u = static_cast(a_nrows); std::size_t a_ncols_u = static_cast(a_ncols); auto [opa_nrows, opa_ncols] = swap_if_transposed(opA, a_nrows_u, a_ncols_u); diff --git a/tests/unit_tests/sparse_blas/include/test_spsv.hpp b/tests/unit_tests/sparse_blas/include/test_spsv.hpp index 16ffe4484..d3937997f 100644 --- a/tests/unit_tests/sparse_blas/include/test_spsv.hpp +++ b/tests/unit_tests/sparse_blas/include/test_spsv.hpp @@ -48,16 +48,18 @@ */ template void test_helper_with_format(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, - sycl::device *dev, sparse_matrix_format_t format, - oneapi::math::transpose transpose_val, int &num_passed, - int &num_skipped) { + sycl::device* dev, sparse_matrix_format_t format, + oneapi::math::transpose transpose_val, int& num_passed, + int& num_skipped) { double density_A_matrix = 0.144; fpType alpha = set_fp_value()(1.f, 0.f); int m = 277; oneapi::math::index_base index_zero = oneapi::math::index_base::zero; oneapi::math::sparse::spsv_alg default_alg = oneapi::math::sparse::spsv_alg::default_alg; - oneapi::math::sparse::spsv_alg no_optimize_alg = oneapi::math::sparse::spsv_alg::no_optimize_alg; - oneapi::math::sparse::matrix_view default_A_view(oneapi::math::sparse::matrix_descr::triangular); + oneapi::math::sparse::spsv_alg no_optimize_alg = + oneapi::math::sparse::spsv_alg::no_optimize_alg; + oneapi::math::sparse::matrix_view default_A_view( + oneapi::math::sparse::matrix_descr::triangular); oneapi::math::sparse::matrix_view upper_A_view(oneapi::math::sparse::matrix_descr::triangular); upper_A_view.uplo_view = oneapi::math::uplo::upper; std::set no_properties; @@ -158,8 +160,8 @@ void test_helper_with_format(testFunctorI32 test_functor_i32, testFunctorI64 tes */ template void test_helper(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i64, - sycl::device *dev, oneapi::math::transpose transpose_val, int &num_passed, - int &num_skipped) { + sycl::device* dev, oneapi::math::transpose transpose_val, int& num_passed, + int& num_skipped) { test_helper_with_format(test_functor_i32, test_functor_i64, dev, sparse_matrix_format_t::CSR, transpose_val, num_passed, num_skipped); @@ -170,11 +172,11 @@ void test_helper(testFunctorI32 test_functor_i32, testFunctorI64 test_functor_i6 /// Compute spsv reference as a dense operation template -void prepare_reference_spsv_data(sparse_matrix_format_t format, const intType *ia, - const intType *ja, const fpType *a, intType m, intType nnz, - intType indexing, oneapi::math::transpose opA, const fpType *x, +void prepare_reference_spsv_data(sparse_matrix_format_t format, const intType* ia, + const intType* ja, const fpType* a, intType m, intType nnz, + intType indexing, oneapi::math::transpose opA, const fpType* x, fpType alpha, oneapi::math::sparse::matrix_view A_view, - fpType *y_ref) { + fpType* y_ref) { std::size_t mu = static_cast(m); auto dense_opa = sparse_to_dense(format, ia, ja, a, mu, mu, static_cast(nnz), indexing, opA, A_view); @@ -186,8 +188,8 @@ void prepare_reference_spsv_data(sparse_matrix_format_t format, const intType *i // // Compute each element of the reference one after the other starting from 0 (resp. the end) for a lower (resp. upper) triangular matrix. // A matrix is considered lowered if it is lower and not transposed or upper and transposed. - const bool is_lower = - (A_view.uplo_view == oneapi::math::uplo::lower) == (opA == oneapi::math::transpose::nontrans); + const bool is_lower = (A_view.uplo_view == oneapi::math::uplo::lower) == + (opA == oneapi::math::transpose::nontrans); for (std::size_t row = 0; row < mu; row++) { std::size_t uplo_row = is_lower ? row : (mu - 1 - row); fpType rhs = alpha * x[uplo_row]; diff --git a/tests/unit_tests/sparse_blas/source/sparse_spmm_buffer.cpp b/tests/unit_tests/sparse_blas/source/sparse_spmm_buffer.cpp index 330b06bba..1a97c313e 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spmm_buffer.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spmm_buffer.cpp @@ -23,18 +23,18 @@ #include "test_spmm.hpp" -extern std::vector devices; +extern std::vector devices; namespace { template -int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, +int test_spmm(sycl::device* dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, intType ncols_C, double density_A_matrix, oneapi::math::index_base index, oneapi::math::layout dense_matrix_layout, oneapi::math::transpose transpose_A, oneapi::math::transpose transpose_B, fpType alpha, fpType beta, intType ldb, intType ldc, oneapi::math::sparse::spmm_alg alg, oneapi::math::sparse::matrix_view A_view, - const std::set &matrix_properties, + const std::set& matrix_properties, bool reset_data, bool test_scalar_on_device) { if (test_scalar_on_device) { // Scalars on the device is not planned to be supported with the buffer API @@ -92,7 +92,8 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, init_sparse_matrix(main_queue, format, &A_handle, nrows_A, ncols_A, nnz, index, ia_buf, ja_buf, a_buf); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, + property); } CALL_RT_OR_CT(oneapi::math::sparse::init_dense_matrix, main_queue, &B_handle, opb_nrows, opb_ncols, ldb, dense_matrix_layout, b_buf); @@ -154,13 +155,13 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, A_view, A_handle, B_handle, &beta, C_handle, alg, descr); } } - catch (const sycl::exception &e) { + catch (const sycl::exception& e) { std::cout << "Caught synchronous SYCL exception during sparse SPMM:\n" << e.what() << std::endl; print_error_code(e); return 0; } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { wait_and_free_handles(main_queue, A_handle, B_handle, C_handle); if (descr) { sycl::event ev_release_descr; @@ -170,7 +171,7 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, } return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of sparse SPMM:\n" << error.what() << std::endl; return 0; } @@ -190,7 +191,7 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, return static_cast(valid); } -class SparseSpmmBufferTests : public ::testing::TestWithParam {}; +class SparseSpmmBufferTests : public ::testing::TestWithParam {}; TEST_P(SparseSpmmBufferTests, RealSinglePrecision) { using fpType = float; diff --git a/tests/unit_tests/sparse_blas/source/sparse_spmm_usm.cpp b/tests/unit_tests/sparse_blas/source/sparse_spmm_usm.cpp index 334604a5a..190b63c28 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spmm_usm.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spmm_usm.cpp @@ -23,18 +23,18 @@ #include "test_spmm.hpp" -extern std::vector devices; +extern std::vector devices; namespace { template -int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, +int test_spmm(sycl::device* dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, intType ncols_C, double density_A_matrix, oneapi::math::index_base index, oneapi::math::layout dense_matrix_layout, oneapi::math::transpose transpose_A, oneapi::math::transpose transpose_B, fpType alpha, fpType beta, intType ldb, intType ldc, oneapi::math::sparse::spmm_alg alg, oneapi::math::sparse::matrix_view A_view, - const std::set &matrix_properties, + const std::set& matrix_properties, bool reset_data, bool test_scalar_on_device) { sycl::queue main_queue(*dev, exception_handler_t()); @@ -82,11 +82,11 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, auto alpha_usm_uptr = malloc_device_uptr(main_queue, 1); auto beta_usm_uptr = malloc_device_uptr(main_queue, 1); - intType *ia_usm = ia_usm_uptr.get(); - intType *ja_usm = ja_usm_uptr.get(); - fpType *a_usm = a_usm_uptr.get(); - fpType *b_usm = b_usm_uptr.get(); - fpType *c_usm = c_usm_uptr.get(); + intType* ia_usm = ia_usm_uptr.get(); + intType* ja_usm = ja_usm_uptr.get(); + fpType* a_usm = a_usm_uptr.get(); + fpType* b_usm = b_usm_uptr.get(); + fpType* c_usm = c_usm_uptr.get(); std::vector mat_dependencies; std::vector spmm_dependencies; @@ -102,8 +102,8 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, spmm_dependencies.push_back( main_queue.memcpy(c_usm, c_host.data(), c_host.size() * sizeof(fpType))); - fpType *alpha_host_or_usm_ptr = α - fpType *beta_host_or_usm_ptr = β + fpType* alpha_host_or_usm_ptr = α + fpType* beta_host_or_usm_ptr = β if (test_scalar_on_device) { spmm_dependencies.push_back( main_queue.memcpy(alpha_usm_uptr.get(), &alpha, sizeof(fpType))); @@ -122,7 +122,8 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, init_sparse_matrix(main_queue, format, &A_handle, nrows_A, ncols_A, nnz, index, ia_usm, ja_usm, a_usm); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, + property); } CALL_RT_OR_CT(oneapi::math::sparse::init_dense_matrix, main_queue, &B_handle, opb_nrows, opb_ncols, ldb, dense_matrix_layout, b_usm); @@ -192,20 +193,20 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, workspace_usm.get(), mat_dependencies); - CALL_RT_OR_CT(ev_spmm = oneapi::math::sparse::spmm, main_queue, transpose_A, transpose_B, - &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, descr, - { ev_opt }); + CALL_RT_OR_CT(ev_spmm = oneapi::math::sparse::spmm, main_queue, transpose_A, + transpose_B, &alpha, A_view, A_handle, B_handle, &beta, C_handle, alg, + descr, { ev_opt }); } ev_copy = main_queue.memcpy(c_host.data(), c_usm, c_host.size() * sizeof(fpType), ev_spmm); } - catch (const sycl::exception &e) { + catch (const sycl::exception& e) { std::cout << "Caught synchronous SYCL exception during sparse SPMM:\n" << e.what() << std::endl; print_error_code(e); return 0; } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { wait_and_free_handles(main_queue, A_handle, B_handle, C_handle); if (descr) { sycl::event ev_release_descr; @@ -215,7 +216,7 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, } return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of sparse SPMM:\n" << error.what() << std::endl; return 0; } @@ -238,7 +239,7 @@ int test_spmm(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, return static_cast(valid); } -class SparseSpmmUsmTests : public ::testing::TestWithParam {}; +class SparseSpmmUsmTests : public ::testing::TestWithParam {}; TEST_P(SparseSpmmUsmTests, RealSinglePrecision) { using fpType = float; diff --git a/tests/unit_tests/sparse_blas/source/sparse_spmv_buffer.cpp b/tests/unit_tests/sparse_blas/source/sparse_spmv_buffer.cpp index 0c4136bcc..76700d483 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spmv_buffer.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spmv_buffer.cpp @@ -23,16 +23,16 @@ #include "test_spmv.hpp" -extern std::vector devices; +extern std::vector devices; namespace { template -int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, +int test_spmv(sycl::device* dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, double density_A_matrix, oneapi::math::index_base index, oneapi::math::transpose transpose_val, fpType alpha, fpType beta, oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::matrix_view A_view, - const std::set &matrix_properties, + const std::set& matrix_properties, bool reset_data, bool test_scalar_on_device) { if (test_scalar_on_device) { // Scalars on the device is not planned to be supported with the buffer API @@ -85,7 +85,8 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, init_sparse_matrix(main_queue, format, &A_handle, nrows_A, ncols_A, nnz, index, ia_buf, ja_buf, a_buf); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, + property); } CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &x_handle, static_cast(x_host.size()), x_buf); @@ -99,8 +100,8 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, A_view, A_handle, x_handle, &beta, y_handle, alg, descr, workspace_size); sycl::buffer workspace_buf((sycl::range<1>(workspace_size))); - CALL_RT_OR_CT(oneapi::math::sparse::spmv_optimize, main_queue, transpose_val, &alpha, A_view, - A_handle, x_handle, &beta, y_handle, alg, descr, workspace_buf); + CALL_RT_OR_CT(oneapi::math::sparse::spmv_optimize, main_queue, transpose_val, &alpha, + A_view, A_handle, x_handle, &beta, y_handle, alg, descr, workspace_buf); CALL_RT_OR_CT(oneapi::math::sparse::spmv, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, &beta, y_handle, alg, descr); @@ -143,13 +144,13 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, A_handle, x_handle, &beta, y_handle, alg, descr); } } - catch (const sycl::exception &e) { + catch (const sycl::exception& e) { std::cout << "Caught synchronous SYCL exception during sparse SPMV:\n" << e.what() << std::endl; print_error_code(e); return 0; } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { wait_and_free_handles(main_queue, A_handle, x_handle, y_handle); if (descr) { sycl::event ev_release_descr; @@ -159,7 +160,7 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, } return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of sparse SPMV:\n" << error.what() << std::endl; return 0; } @@ -178,7 +179,7 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, return static_cast(valid); } -class SparseSpmvBufferTests : public ::testing::TestWithParam {}; +class SparseSpmvBufferTests : public ::testing::TestWithParam {}; TEST_P(SparseSpmvBufferTests, RealSinglePrecision) { using fpType = float; diff --git a/tests/unit_tests/sparse_blas/source/sparse_spmv_usm.cpp b/tests/unit_tests/sparse_blas/source/sparse_spmv_usm.cpp index a3265ea70..f7a9c9960 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spmv_usm.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spmv_usm.cpp @@ -23,16 +23,16 @@ #include "test_spmv.hpp" -extern std::vector devices; +extern std::vector devices; namespace { template -int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, +int test_spmv(sycl::device* dev, sparse_matrix_format_t format, intType nrows_A, intType ncols_A, double density_A_matrix, oneapi::math::index_base index, oneapi::math::transpose transpose_val, fpType alpha, fpType beta, oneapi::math::sparse::spmv_alg alg, oneapi::math::sparse::matrix_view A_view, - const std::set &matrix_properties, + const std::set& matrix_properties, bool reset_data, bool test_scalar_on_device) { sycl::queue main_queue(*dev, exception_handler_t()); @@ -75,11 +75,11 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, auto alpha_usm_uptr = malloc_device_uptr(main_queue, 1); auto beta_usm_uptr = malloc_device_uptr(main_queue, 1); - intType *ia_usm = ia_usm_uptr.get(); - intType *ja_usm = ja_usm_uptr.get(); - fpType *a_usm = a_usm_uptr.get(); - fpType *x_usm = x_usm_uptr.get(); - fpType *y_usm = y_usm_uptr.get(); + intType* ia_usm = ia_usm_uptr.get(); + intType* ja_usm = ja_usm_uptr.get(); + fpType* a_usm = a_usm_uptr.get(); + fpType* x_usm = x_usm_uptr.get(); + fpType* y_usm = y_usm_uptr.get(); std::vector mat_dependencies; std::vector spmv_dependencies; @@ -95,8 +95,8 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, spmv_dependencies.push_back( main_queue.memcpy(y_usm, y_host.data(), y_host.size() * sizeof(fpType))); - fpType *alpha_host_or_usm_ptr = α - fpType *beta_host_or_usm_ptr = β + fpType* alpha_host_or_usm_ptr = α + fpType* beta_host_or_usm_ptr = β if (test_scalar_on_device) { spmv_dependencies.push_back( main_queue.memcpy(alpha_usm_uptr.get(), &alpha, sizeof(fpType))); @@ -115,7 +115,8 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, init_sparse_matrix(main_queue, format, &A_handle, nrows_A, ncols_A, nnz, index, ia_usm, ja_usm, a_usm); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, + property); } CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &x_handle, static_cast(x_host.size()), x_usm); @@ -191,13 +192,13 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, ev_copy = main_queue.memcpy(y_host.data(), y_usm, y_host.size() * sizeof(fpType), ev_spmv); } - catch (const sycl::exception &e) { + catch (const sycl::exception& e) { std::cout << "Caught synchronous SYCL exception during sparse SPMV:\n" << e.what() << std::endl; print_error_code(e); return 0; } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { wait_and_free_handles(main_queue, A_handle, x_handle, y_handle); if (descr) { sycl::event ev_release_descr; @@ -207,7 +208,7 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, } return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of sparse SPMV:\n" << error.what() << std::endl; return 0; } @@ -229,7 +230,7 @@ int test_spmv(sycl::device *dev, sparse_matrix_format_t format, intType nrows_A, return static_cast(valid); } -class SparseSpmvUsmTests : public ::testing::TestWithParam {}; +class SparseSpmvUsmTests : public ::testing::TestWithParam {}; TEST_P(SparseSpmvUsmTests, RealSinglePrecision) { using fpType = float; diff --git a/tests/unit_tests/sparse_blas/source/sparse_spsv_buffer.cpp b/tests/unit_tests/sparse_blas/source/sparse_spsv_buffer.cpp index 62a84120a..286558834 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spsv_buffer.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spsv_buffer.cpp @@ -23,15 +23,15 @@ #include "test_spsv.hpp" -extern std::vector devices; +extern std::vector devices; namespace { template -int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, double density_A_matrix, +int test_spsv(sycl::device* dev, sparse_matrix_format_t format, intType m, double density_A_matrix, oneapi::math::index_base index, oneapi::math::transpose transpose_val, fpType alpha, oneapi::math::sparse::spsv_alg alg, oneapi::math::sparse::matrix_view A_view, - const std::set &matrix_properties, + const std::set& matrix_properties, bool reset_data, bool test_scalar_on_device) { if (test_scalar_on_device) { // Scalars on the device is not planned to be supported with the buffer API @@ -87,7 +87,8 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl try { init_sparse_matrix(main_queue, format, &A_handle, m, m, nnz, index, ia_buf, ja_buf, a_buf); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, + property); } CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &x_handle, m, x_buf); CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &y_handle, m, y_buf); @@ -99,8 +100,8 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl A_view, A_handle, x_handle, y_handle, alg, descr, workspace_size); sycl::buffer workspace_buf((sycl::range<1>(workspace_size))); - CALL_RT_OR_CT(oneapi::math::sparse::spsv_optimize, main_queue, transpose_val, &alpha, A_view, - A_handle, x_handle, y_handle, alg, descr, workspace_buf); + CALL_RT_OR_CT(oneapi::math::sparse::spsv_optimize, main_queue, transpose_val, &alpha, + A_view, A_handle, x_handle, y_handle, alg, descr, workspace_buf); CALL_RT_OR_CT(oneapi::math::sparse::spsv, main_queue, transpose_val, &alpha, A_view, A_handle, x_handle, y_handle, alg, descr); @@ -141,13 +142,13 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl A_handle, x_handle, y_handle, alg, descr); } } - catch (const sycl::exception &e) { + catch (const sycl::exception& e) { std::cout << "Caught synchronous SYCL exception during sparse SPSV:\n" << e.what() << std::endl; print_error_code(e); return 0; } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { wait_and_free_handles(main_queue, A_handle, x_handle, y_handle); if (descr) { sycl::event ev_release_descr; @@ -157,7 +158,7 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl } return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of sparse SPSV:\n" << error.what() << std::endl; return 0; } @@ -176,7 +177,7 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl return static_cast(valid); } -class SparseSpsvBufferTests : public ::testing::TestWithParam {}; +class SparseSpsvBufferTests : public ::testing::TestWithParam {}; TEST_P(SparseSpsvBufferTests, RealSinglePrecision) { using fpType = float; diff --git a/tests/unit_tests/sparse_blas/source/sparse_spsv_usm.cpp b/tests/unit_tests/sparse_blas/source/sparse_spsv_usm.cpp index 1555bd2c0..da1456daa 100644 --- a/tests/unit_tests/sparse_blas/source/sparse_spsv_usm.cpp +++ b/tests/unit_tests/sparse_blas/source/sparse_spsv_usm.cpp @@ -23,15 +23,15 @@ #include "test_spsv.hpp" -extern std::vector devices; +extern std::vector devices; namespace { template -int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, double density_A_matrix, +int test_spsv(sycl::device* dev, sparse_matrix_format_t format, intType m, double density_A_matrix, oneapi::math::index_base index, oneapi::math::transpose transpose_val, fpType alpha, oneapi::math::sparse::spsv_alg alg, oneapi::math::sparse::matrix_view A_view, - const std::set &matrix_properties, + const std::set& matrix_properties, bool reset_data, bool test_scalar_on_device) { sycl::queue main_queue(*dev, exception_handler_t()); @@ -77,11 +77,11 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl auto y_usm_uptr = malloc_device_uptr(main_queue, y_host.size()); auto alpha_usm_uptr = malloc_device_uptr(main_queue, 1); - intType *ia_usm = ia_usm_uptr.get(); - intType *ja_usm = ja_usm_uptr.get(); - fpType *a_usm = a_usm_uptr.get(); - fpType *x_usm = x_usm_uptr.get(); - fpType *y_usm = y_usm_uptr.get(); + intType* ia_usm = ia_usm_uptr.get(); + intType* ja_usm = ja_usm_uptr.get(); + fpType* a_usm = a_usm_uptr.get(); + fpType* x_usm = x_usm_uptr.get(); + fpType* y_usm = y_usm_uptr.get(); std::vector mat_dependencies; std::vector spsv_dependencies; @@ -97,7 +97,7 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl spsv_dependencies.push_back( main_queue.memcpy(y_usm, y_host.data(), y_host.size() * sizeof(fpType))); - fpType *alpha_host_or_usm_ptr = α + fpType* alpha_host_or_usm_ptr = α if (test_scalar_on_device) { spsv_dependencies.push_back( main_queue.memcpy(alpha_usm_uptr.get(), &alpha, sizeof(fpType))); @@ -113,7 +113,8 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl try { init_sparse_matrix(main_queue, format, &A_handle, m, m, nnz, index, ia_usm, ja_usm, a_usm); for (auto property : matrix_properties) { - CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, property); + CALL_RT_OR_CT(oneapi::math::sparse::set_matrix_property, main_queue, A_handle, + property); } CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &x_handle, m, x_usm); CALL_RT_OR_CT(oneapi::math::sparse::init_dense_vector, main_queue, &y_handle, m, y_usm); @@ -186,13 +187,13 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl ev_copy = main_queue.memcpy(y_host.data(), y_usm, y_host.size() * sizeof(fpType), ev_spsv); } - catch (const sycl::exception &e) { + catch (const sycl::exception& e) { std::cout << "Caught synchronous SYCL exception during sparse SPSV:\n" << e.what() << std::endl; print_error_code(e); return 0; } - catch (const oneapi::math::unimplemented &e) { + catch (const oneapi::math::unimplemented& e) { wait_and_free_handles(main_queue, A_handle, x_handle, y_handle); if (descr) { sycl::event ev_release_descr; @@ -202,7 +203,7 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl } return test_skipped; } - catch (const std::runtime_error &error) { + catch (const std::runtime_error& error) { std::cout << "Error raised during execution of sparse SPSV:\n" << error.what() << std::endl; return 0; } @@ -224,7 +225,7 @@ int test_spsv(sycl::device *dev, sparse_matrix_format_t format, intType m, doubl return static_cast(valid); } -class SparseSpsvUsmTests : public ::testing::TestWithParam {}; +class SparseSpsvUsmTests : public ::testing::TestWithParam {}; TEST_P(SparseSpsvUsmTests, RealSinglePrecision) { using fpType = float; From 2a8a76bd24fad7a69ee1f0fcba2183e4df39c033 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Mon, 21 Oct 2024 17:35:33 +0100 Subject: [PATCH 36/41] Remove USE_MKLREF --- .../unit_tests/lapack/include/lapack_reference_wrappers.hpp | 5 ----- 1 file changed, 5 deletions(-) diff --git a/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp b/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp index 9e28e9feb..2f500274c 100644 --- a/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp +++ b/tests/unit_tests/lapack/include/lapack_reference_wrappers.hpp @@ -27,13 +27,8 @@ #endif #define WeirdNEC extern "C" { -#ifdef USE_MKLREF -#include -#include -#else #include "cblas.h" #include "lapacke.h" -#endif } static_assert(sizeof(lapack_int) == 8); static_assert(sizeof(CBLAS_INT) == 8); From a42a97089dae9a1e672a1560b0bc4545cea64cd9 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Wed, 23 Oct 2024 15:05:13 +0100 Subject: [PATCH 37/41] Rename missing oneMKL occurrences --- docs/README.md | 4 ++-- docs/building_the_project_with_adaptivecpp.rst | 4 ++-- docs/building_the_project_with_dpcpp.rst | 10 +++++----- examples/README.md | 2 +- 4 files changed, 10 insertions(+), 10 deletions(-) diff --git a/docs/README.md b/docs/README.md index 0e1d70187..2ef6e5a73 100644 --- a/docs/README.md +++ b/docs/README.md @@ -8,9 +8,9 @@ For more information see [Building with CMake](../README.md#building-with-cmake) To build documentation only, use the following commands from the current folder: ```bash -# Inside /docs +# Inside /docs mkdir build && cd build cmake .. cmake --build . ``` -Generated documentation can be found in `/docs/build/Documentation` +Generated documentation can be found in `/docs/build/Documentation` diff --git a/docs/building_the_project_with_adaptivecpp.rst b/docs/building_the_project_with_adaptivecpp.rst index b907285a9..086fb3a86 100644 --- a/docs/building_the_project_with_adaptivecpp.rst +++ b/docs/building_the_project_with_adaptivecpp.rst @@ -16,7 +16,7 @@ Environment Setup #. Clone this project. The root directory of the cloned repository will be - referred to as ````. + referred to as ````. #. Download and install the `required dependencies @@ -33,7 +33,7 @@ On Linux (other OSes are not supported with the AdaptiveCpp compiler): .. code-block:: bash - # Inside + # Inside mkdir build && cd build cmake .. -DONEMATH_SYCL_IMPLEMENTATION=hipsycl \ # Indicate that AdaptiveCpp is being used. -DENABLE_MKLGPU_BACKEND=False \ # MKLGPU backend is not supported by AdaptiveCpp diff --git a/docs/building_the_project_with_dpcpp.rst b/docs/building_the_project_with_dpcpp.rst index 780e0c6b3..5d0e95c82 100644 --- a/docs/building_the_project_with_dpcpp.rst +++ b/docs/building_the_project_with_dpcpp.rst @@ -437,20 +437,20 @@ Note that this is not a tested configuration. This builds oneMath with the portBLAS backend only, for a generic SYCL device supported by the Open DPC++ project. -Build oneMKL for the DFT domain on a generic SYCL device: +Build oneMath for the DFT domain on a generic SYCL device: .. code-block:: bash - cmake $ONEMKL_DIR \ + cmake $ONEMATH_DIR \ -DCMAKE_CXX_COMPILER=clang++ \ -DCMAKE_C_COMPILER=clang \ -DENABLE_MKLCPU_BACKEND=False \ -DENABLE_MKLGPU_BACKEND=False \ -DENABLE_PORTFFT_BACKEND=True -Note that this is not a tested configuration. This builds oneMKL Interfaces -with the portFFT backend only, for a generic SYCL device supported by the -Open DPC++ project. +Note that this is not a tested configuration. This builds oneMath with the +portFFT backend only, for a generic SYCL device supported by the Open DPC++ +project. .. _project_cleanup: diff --git a/examples/README.md b/examples/README.md index 46730b900..98b0d2e70 100644 --- a/examples/README.md +++ b/examples/README.md @@ -477,7 +477,7 @@ Device name is: Intel(R) UHD Graphics 750 Running with single precision real data type: DFT example run_time dispatch Unsupported Configuration: - oneMKL: dft/backends/portfft/commit: function is not implemented portFFT only supports complex to complex transforms + oneMath: dft/backends/portfft/commit: function is not implemented REAL domain is unsupported ``` ## sparse_blas From baeb476f92b6dd7a0d137ab87c09fde1ec706634 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Thu, 7 Nov 2024 14:46:07 +0000 Subject: [PATCH 38/41] Remove duplicated blas and lapack headers --- .../backends/mkl_common/mkl_blas_backend.hpp | 15 - .../backends/mkl_common/mkl_blas_backend.hxx | 2518 ----------------- .../mkl_common/mkl_lapack_backend.hpp | 1276 --------- src/lapack/backends/mklcpu/mkl_lapack.cpp | 3 +- src/lapack/backends/mklgpu/mkl_lapack.cpp | 3 +- 5 files changed, 4 insertions(+), 3811 deletions(-) delete mode 100644 src/blas/backends/mkl_common/mkl_blas_backend.hxx delete mode 100644 src/lapack/backends/mkl_common/mkl_lapack_backend.hpp diff --git a/src/blas/backends/mkl_common/mkl_blas_backend.hpp b/src/blas/backends/mkl_common/mkl_blas_backend.hpp index 30be71602..d3b0fc3ce 100644 --- a/src/blas/backends/mkl_common/mkl_blas_backend.hpp +++ b/src/blas/backends/mkl_common/mkl_blas_backend.hpp @@ -69,20 +69,5 @@ class value_or_pointer { } }; -namespace blas { - -namespace column_major { - -#include "mkl_blas_backend.hxx" - -} - -namespace row_major { - -#include "mkl_blas_backend.hxx" - -} - -} // namespace blas } // namespace math } // namespace oneapi diff --git a/src/blas/backends/mkl_common/mkl_blas_backend.hxx b/src/blas/backends/mkl_common/mkl_blas_backend.hxx deleted file mode 100644 index 51937dade..000000000 --- a/src/blas/backends/mkl_common/mkl_blas_backend.hxx +++ /dev/null @@ -1,2518 +0,0 @@ -/******************************************************************************* -* Copyright 2022 Intel Corporation -* -* Licensed under the Apache License, Version 2.0 (the "License"); -* you may not use this file except in compliance with the License. -* You may obtain a copy of the License at -* -* http://www.apache.org/licenses/LICENSE-2.0 -* -* Unless required by applicable law or agreed to in writing, -* software distributed under the License is distributed on an "AS IS" BASIS, -* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -* See the License for the specific language governing permissions -* and limitations under the License. -* -* -* SPDX-License-Identifier: Apache-2.0 -*******************************************************************************/ - -/// level3, buffer - -void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, - std::int64_t ldc); - -void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb, double beta, sycl::buffer& c, - std::int64_t ldc); - -void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); - -void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); - -void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::half alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb, sycl::half beta, - sycl::buffer& c, std::int64_t ldc); - -void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, - std::int64_t ldc); - -void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, - std::int64_t ldc); - -void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, - std::int64_t ldc); - -void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb, float beta, - sycl::buffer& c, std::int64_t ldc); - -void gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, - std::int64_t ldc); - -void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, - std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); - -void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, - std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc); - -void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1>& c, std::int64_t ldc); - -void symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1>& c, std::int64_t ldc); - -void hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1>& c, std::int64_t ldc); - -void hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1>& c, std::int64_t ldc); - -void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer& a, std::int64_t lda, float beta, - sycl::buffer& c, std::int64_t ldc); - -void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer& a, std::int64_t lda, double beta, - sycl::buffer& c, std::int64_t ldc); - -void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); - -void syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); - -void herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer, 1>& a, std::int64_t lda, float beta, - sycl::buffer, 1>& c, std::int64_t ldc); - -void herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer, 1>& a, std::int64_t lda, double beta, - sycl::buffer, 1>& c, std::int64_t ldc); - -void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - float alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, - std::int64_t ldb, float beta, sycl::buffer& c, std::int64_t ldc); - -void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - double alpha, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, - std::int64_t ldb, double beta, sycl::buffer& c, std::int64_t ldc); - -void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1>& c, std::int64_t ldc); - -void syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb, std::complex beta, - sycl::buffer, 1>& c, std::int64_t ldc); - -void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb, float beta, - sycl::buffer, 1>& c, std::int64_t ldc); - -void her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb, double beta, - sycl::buffer, 1>& c, std::int64_t ldc); - -void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb); - -void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, - std::int64_t lda, sycl::buffer& b, std::int64_t ldb); - -void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb); - -void trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb); - -void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb); - -void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, double alpha, sycl::buffer& a, - std::int64_t lda, sycl::buffer& b, std::int64_t ldb); - -void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb); - -void trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb); - -// level 3, USM - -sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const float* a, - std::int64_t lda, const float* b, std::int64_t ldb, value_or_pointer beta, - float* c, std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const double* a, - std::int64_t lda, const double* b, std::int64_t ldb, value_or_pointer beta, - double* c, std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer> beta, - std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer> beta, - std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, - const sycl::half* a, std::int64_t lda, const sycl::half* b, std::int64_t ldb, - value_or_pointer beta, sycl::half* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const sycl::half* a, - std::int64_t lda, const sycl::half* b, std::int64_t ldb, - value_or_pointer beta, float* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const bfloat16* a, - std::int64_t lda, const bfloat16* b, std::int64_t ldb, - value_or_pointer beta, float* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const bfloat16* a, - std::int64_t lda, const bfloat16* b, std::int64_t ldb, - value_or_pointer beta, bfloat16* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::int8_t* a, std::int64_t lda, const std::int8_t* b, std::int64_t ldb, - value_or_pointer beta, std::int32_t* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event gemm(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::int8_t* a, std::int64_t lda, const std::int8_t* b, std::int64_t ldb, - value_or_pointer beta, float* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const float* a, std::int64_t lda, - const float* b, std::int64_t ldb, value_or_pointer beta, float* c, - std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const double* a, std::int64_t lda, - const double* b, std::int64_t ldb, value_or_pointer beta, double* c, - std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer> beta, - std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event symm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer> beta, - std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer> beta, - std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event hemm(sycl::queue& queue, side left_right, uplo upper_lower, std::int64_t m, - std::int64_t n, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer> beta, - std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const float* a, std::int64_t lda, - value_or_pointer beta, float* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const double* a, std::int64_t lda, - value_or_pointer beta, double* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, - value_or_pointer> beta, std::complex* c, - std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event syrk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, - value_or_pointer> beta, std::complex* c, - std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const std::complex* a, - std::int64_t lda, value_or_pointer beta, std::complex* c, - std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event herk(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const std::complex* a, - std::int64_t lda, value_or_pointer beta, std::complex* c, - std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const float* a, std::int64_t lda, - const float* b, std::int64_t ldb, value_or_pointer beta, float* c, - std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const double* a, std::int64_t lda, - const double* b, std::int64_t ldb, value_or_pointer beta, double* c, - std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer> beta, - std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event syr2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer> beta, - std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer beta, std::complex* c, - std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event her2k(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer beta, std::complex* c, - std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const float* a, std::int64_t lda, float* b, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const double* a, std::int64_t lda, double* b, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::complex* b, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event trmm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::complex* b, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const float* a, std::int64_t lda, float* b, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const double* a, std::int64_t lda, double* b, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::complex* b, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event trsm(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::complex* b, std::int64_t ldb, - const std::vector& dependencies = {}); - -// level 2, buffer - -void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, - float beta, sycl::buffer& y, std::int64_t incy); - -void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer& a, std::int64_t lda, sycl::buffer& x, - std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); - -void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1>& y, std::int64_t incy); - -void gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1>& y, std::int64_t incy); - -void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, float alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& x, std::int64_t incx, float beta, sycl::buffer& y, - std::int64_t incy); - -void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, double alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& x, std::int64_t incx, double beta, sycl::buffer& y, - std::int64_t incy); - -void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, - std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); - -void gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, std::int64_t kl, - std::int64_t ku, std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, sycl::buffer, 1>& x, std::int64_t incx, - std::complex beta, sycl::buffer, 1>& y, std::int64_t incy); - -void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, float alpha, sycl::buffer& x, - std::int64_t incx, sycl::buffer& y, std::int64_t incy, sycl::buffer& a, - std::int64_t lda); - -void ger(sycl::queue& queue, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer& x, std::int64_t incx, sycl::buffer& y, - std::int64_t incy, sycl::buffer& a, std::int64_t lda); - -void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& a, std::int64_t lda); - -void gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& a, std::int64_t lda); - -void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& a, std::int64_t lda); - -void geru(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& a, std::int64_t lda); - -void hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1>& y, std::int64_t incy); - -void hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1>& y, std::int64_t incy); - -void hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1>& y, std::int64_t incy); - -void hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1>& y, std::int64_t incy); - -void her(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& a, std::int64_t lda); - -void her(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& a, std::int64_t lda); - -void her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& a, std::int64_t lda); - -void her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& a, std::int64_t lda); - -void hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, sycl::buffer, 1>& x, - std::int64_t incx, std::complex beta, sycl::buffer, 1>& y, - std::int64_t incy); - -void hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, sycl::buffer, 1>& x, - std::int64_t incx, std::complex beta, sycl::buffer, 1>& y, - std::int64_t incy); - -void hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& a); - -void hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& a); - -void hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& a); - -void hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& a); - -void sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, - float beta, sycl::buffer& y, std::int64_t incy); - -void sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, double alpha, - sycl::buffer& a, std::int64_t lda, sycl::buffer& x, - std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); - -void symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer& a, std::int64_t lda, sycl::buffer& x, std::int64_t incx, - float beta, sycl::buffer& y, std::int64_t incy); - -void symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer& a, std::int64_t lda, sycl::buffer& x, - std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); - -void syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer& x, std::int64_t incx, sycl::buffer& a, std::int64_t lda); - -void syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer& x, std::int64_t incx, sycl::buffer& a, - std::int64_t lda); - -void syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer& x, std::int64_t incx, sycl::buffer& y, - std::int64_t incy, sycl::buffer& a, std::int64_t lda); - -void syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer& x, std::int64_t incx, sycl::buffer& y, - std::int64_t incy, sycl::buffer& a, std::int64_t lda); - -void spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer& a, sycl::buffer& x, std::int64_t incx, float beta, - sycl::buffer& y, std::int64_t incy); - -void spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer& a, sycl::buffer& x, std::int64_t incx, double beta, - sycl::buffer& y, std::int64_t incy); - -void spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer& x, std::int64_t incx, sycl::buffer& a); - -void spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer& x, std::int64_t incx, sycl::buffer& a); - -void spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, float alpha, - sycl::buffer& x, std::int64_t incx, sycl::buffer& y, - std::int64_t incy, sycl::buffer& a); - -void spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, double alpha, - sycl::buffer& x, std::int64_t incx, sycl::buffer& y, - std::int64_t incy, sycl::buffer& a); - -void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, - std::int64_t incx); - -void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, - std::int64_t incx); - -void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx); - -void tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx); - -void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, - std::int64_t incx); - -void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& x, - std::int64_t incx); - -void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx); - -void tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - std::int64_t k, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx); - -void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer& a, sycl::buffer& x, std::int64_t incx); - -void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer& a, sycl::buffer& x, std::int64_t incx); - -void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1>& a, sycl::buffer, 1>& x, - std::int64_t incx); - -void tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1>& a, sycl::buffer, 1>& x, - std::int64_t incx); - -void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer& a, sycl::buffer& x, std::int64_t incx); - -void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer& a, sycl::buffer& x, std::int64_t incx); - -void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1>& a, sycl::buffer, 1>& x, - std::int64_t incx); - -void tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1>& a, sycl::buffer, 1>& x, - std::int64_t incx); - -void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer& a, std::int64_t lda, sycl::buffer& x, - std::int64_t incx); - -void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer& a, std::int64_t lda, sycl::buffer& x, - std::int64_t incx); - -void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx); - -void trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx); - -void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer& a, std::int64_t lda, sycl::buffer& x, - std::int64_t incx); - -void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer& a, std::int64_t lda, sycl::buffer& x, - std::int64_t incx); - -void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx); - -void trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, std::int64_t n, - sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& x, std::int64_t incx); - -// level 2, USM - -sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const float* a, std::int64_t lda, const float* x, - std::int64_t incx, value_or_pointer beta, float* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const double* a, std::int64_t lda, const double* x, - std::int64_t incx, value_or_pointer beta, double* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, const std::complex* x, std::int64_t incx, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event gemv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, const std::complex* x, std::int64_t incx, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, value_or_pointer alpha, const float* a, - std::int64_t lda, const float* x, std::int64_t incx, value_or_pointer beta, - float* y, std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, value_or_pointer alpha, const double* a, - std::int64_t lda, const double* x, std::int64_t incx, - value_or_pointer beta, double* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* x, - std::int64_t incx, value_or_pointer> beta, - std::complex* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event gbmv(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::int64_t kl, std::int64_t ku, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* x, - std::int64_t incx, value_or_pointer> beta, - std::complex* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const float* x, std::int64_t incx, const float* y, std::int64_t incy, float* a, - std::int64_t lda, const std::vector& dependencies = {}); - -sycl::event ger(sycl::queue& queue, std::int64_t m, std::int64_t n, value_or_pointer alpha, - const double* x, std::int64_t incx, const double* y, std::int64_t incy, double* a, - std::int64_t lda, const std::vector& dependencies = {}); - -sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* a, std::int64_t lda, - const std::vector& dependencies = {}); - -sycl::event gerc(sycl::queue& queue, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* a, std::int64_t lda, - const std::vector& dependencies = {}); - -sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* a, std::int64_t lda, - const std::vector& dependencies = {}); - -sycl::event geru(sycl::queue& queue, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* a, std::int64_t lda, - const std::vector& dependencies = {}); - -sycl::event hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, const std::complex* x, std::int64_t incx, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event hbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, const std::complex* x, std::int64_t incx, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, const std::complex* x, std::int64_t incx, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event hemv(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, const std::complex* x, std::int64_t incx, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event her(sycl::queue& queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const std::complex* x, std::int64_t incx, std::complex* a, - std::int64_t lda, const std::vector& dependencies = {}); - -sycl::event her(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const std::complex* x, std::int64_t incx, - std::complex* a, std::int64_t lda, - const std::vector& dependencies = {}); - -sycl::event her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* a, std::int64_t lda, - const std::vector& dependencies = {}); - -sycl::event her2(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* a, std::int64_t lda, - const std::vector& dependencies = {}); - -sycl::event hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - const std::complex* x, std::int64_t incx, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event hpmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - const std::complex* x, std::int64_t incx, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const std::complex* x, std::int64_t incx, std::complex* a, - const std::vector& dependencies = {}); - -sycl::event hpr(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const std::complex* x, std::int64_t incx, - std::complex* a, const std::vector& dependencies = {}); - -sycl::event hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* a, const std::vector& dependencies = {}); - -sycl::event hpr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer> alpha, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* a, const std::vector& dependencies = {}); - -sycl::event sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, - value_or_pointer alpha, const float* a, std::int64_t lda, const float* x, - std::int64_t incx, value_or_pointer beta, float* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event sbmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, std::int64_t k, - value_or_pointer alpha, const double* a, std::int64_t lda, const double* x, - std::int64_t incx, value_or_pointer beta, double* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const float* a, std::int64_t lda, const float* x, - std::int64_t incx, value_or_pointer beta, float* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event symv(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const double* a, std::int64_t lda, const double* x, - std::int64_t incx, value_or_pointer beta, double* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const float* x, std::int64_t incx, float* a, std::int64_t lda, - const std::vector& dependencies = {}); - -sycl::event syr(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const double* x, std::int64_t incx, double* a, - std::int64_t lda, const std::vector& dependencies = {}); - -sycl::event syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const float* x, std::int64_t incx, const float* y, - std::int64_t incy, float* a, std::int64_t lda, - const std::vector& dependencies = {}); - -sycl::event syr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const double* x, std::int64_t incx, - const double* y, std::int64_t incy, double* a, std::int64_t lda, - const std::vector& dependencies = {}); - -sycl::event spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const float* a, const float* x, std::int64_t incx, - value_or_pointer beta, float* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event spmv(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const double* a, const double* x, - std::int64_t incx, value_or_pointer beta, double* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, value_or_pointer alpha, - const float* x, std::int64_t incx, float* a, - const std::vector& dependencies = {}); - -sycl::event spr(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const double* x, std::int64_t incx, double* a, - const std::vector& dependencies = {}); - -sycl::event spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const float* x, std::int64_t incx, const float* y, - std::int64_t incy, float* a, const std::vector& dependencies = {}); - -sycl::event spr2(sycl::queue& queue, uplo upper_lower, std::int64_t n, - value_or_pointer alpha, const double* x, std::int64_t incx, - const double* y, std::int64_t incy, double* a, - const std::vector& dependencies = {}); - -sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, float* x, - std::int64_t incx, const std::vector& dependencies = {}); - -sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, double* x, - std::int64_t incx, const std::vector& dependencies = {}); - -sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, - std::complex* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event tbmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, - std::complex* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, float* x, - std::int64_t incx, const std::vector& dependencies = {}); - -sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, double* x, - std::int64_t incx, const std::vector& dependencies = {}); - -sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, - std::complex* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event tbsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, std::int64_t k, const std::complex* a, std::int64_t lda, - std::complex* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float* a, float* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double* a, double* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex* a, std::complex* x, - std::int64_t incx, const std::vector& dependencies = {}); - -sycl::event tpmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex* a, std::complex* x, - std::int64_t incx, const std::vector& dependencies = {}); - -sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float* a, float* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double* a, double* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex* a, std::complex* x, - std::int64_t incx, const std::vector& dependencies = {}); - -sycl::event tpsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex* a, std::complex* x, - std::int64_t incx, const std::vector& dependencies = {}); - -sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float* a, std::int64_t lda, float* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double* a, std::int64_t lda, double* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex* a, std::int64_t lda, - std::complex* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event trmv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex* a, std::int64_t lda, - std::complex* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const float* a, std::int64_t lda, float* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const double* a, std::int64_t lda, double* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex* a, std::int64_t lda, - std::complex* x, std::int64_t incx, - const std::vector& dependencies = {}); - -sycl::event trsv(sycl::queue& queue, uplo upper_lower, transpose trans, diag unit_diag, - std::int64_t n, const std::complex* a, std::int64_t lda, - std::complex* x, std::int64_t incx, - const std::vector& dependencies = {}); - -// level 1, buffer - -void dotc(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& result); - -void dotc(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& result); - -void dotu(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& result); - -void dotu(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, - sycl::buffer, 1>& result); - -void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& result, index_base base = index_base::zero); - -void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& result, index_base base = index_base::zero); - -void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer& result, - index_base base = index_base::zero); - -void iamax(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer& result, - index_base base = index_base::zero); - -void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& result, index_base base = index_base::zero); - -void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& result, index_base base = index_base::zero); - -void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer& result, - index_base base = index_base::zero); - -void iamin(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer& result, - index_base base = index_base::zero); - -void asum(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer& result); - -void asum(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer& result); - -void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& result); - -void asum(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& result); - -void axpy(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, - std::int64_t incx, sycl::buffer& y, std::int64_t incy); - -void axpy(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, - std::int64_t incx, sycl::buffer& y, std::int64_t incy); - -void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& y, std::int64_t incy); - -void axpy(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, - sycl::buffer, 1>& y, std::int64_t incy); - -void axpby(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, - std::int64_t incx, float beta, sycl::buffer& y, std::int64_t incy); - -void axpby(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, - std::int64_t incx, double beta, sycl::buffer& y, std::int64_t incy); - -void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1>& y, std::int64_t incy); - -void axpby(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, std::complex beta, - sycl::buffer, 1>& y, std::int64_t incy); - -void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& y, std::int64_t incy); - -void copy(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& y, std::int64_t incy); - -void copy(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); - -void copy(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); - -void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& y, std::int64_t incy, sycl::buffer& result); - -void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& y, std::int64_t incy, sycl::buffer& result); - -void sdsdot(sycl::queue& queue, std::int64_t n, float sb, sycl::buffer& x, - std::int64_t incx, sycl::buffer& y, std::int64_t incy, - sycl::buffer& result); - -void dot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& y, std::int64_t incy, sycl::buffer& result); - -void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer& result); - -void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer& result); - -void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& result); - -void nrm2(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& result); - -void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, float c, - float s); - -void rot(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy, double c, - double s); - -void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& y, std::int64_t incy, float c, float s); - -void rot(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& y, std::int64_t incy, double c, double s); - -void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, - sycl::buffer& c, sycl::buffer& s); - -void rotg(sycl::queue& queue, sycl::buffer& a, sycl::buffer& b, - sycl::buffer& c, sycl::buffer& s); - -void rotg(sycl::queue& queue, sycl::buffer, 1>& a, - sycl::buffer, 1>& b, sycl::buffer& c, - sycl::buffer, 1>& s); - -void rotg(sycl::queue& queue, sycl::buffer, 1>& a, - sycl::buffer, 1>& b, sycl::buffer& c, - sycl::buffer, 1>& s); - -void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& y, std::int64_t incy, sycl::buffer& param); - -void rotm(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& y, std::int64_t incy, sycl::buffer& param); - -void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, - sycl::buffer& x1, float y1, sycl::buffer& param); - -void rotmg(sycl::queue& queue, sycl::buffer& d1, sycl::buffer& d2, - sycl::buffer& x1, double y1, sycl::buffer& param); - -void scal(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, - std::int64_t incx); - -void scal(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, - std::int64_t incx); - -void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx); - -void scal(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx); - -void scal(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer, 1>& x, - std::int64_t incx); - -void scal(sycl::queue& queue, std::int64_t n, double alpha, - sycl::buffer, 1>& x, std::int64_t incx); - -void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& y, std::int64_t incy); - -void swap(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - sycl::buffer& y, std::int64_t incy); - -void swap(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); - -void swap(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, sycl::buffer, 1>& y, std::int64_t incy); - -// level 1, USM - -sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* result, const std::vector& dependencies = {}); - -sycl::event dotc(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* result, const std::vector& dependencies = {}); - -sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* result, const std::vector& dependencies = {}); - -sycl::event dotu(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, const std::complex* y, std::int64_t incy, - std::complex* result, const std::vector& dependencies = {}); - -sycl::event iamax(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, - std::int64_t* result, const std::vector& dependencies = {}); - -sycl::event iamax(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, - std::int64_t* result, const std::vector& dependencies = {}); - -sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, std::int64_t* result, - const std::vector& dependencies = {}); - -sycl::event iamax(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, std::int64_t* result, - const std::vector& dependencies = {}); - -sycl::event iamin(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, - std::int64_t* result, const std::vector& dependencies = {}); - -sycl::event iamin(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, - std::int64_t* result, const std::vector& dependencies = {}); - -sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, std::int64_t* result, - const std::vector& dependencies = {}); - -sycl::event iamin(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, std::int64_t* result, - const std::vector& dependencies = {}); - -sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, float* result, - const std::vector& dependencies = {}); - -sycl::event asum(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, double* result, - const std::vector& dependencies = {}); - -sycl::event asum(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, - float* result, const std::vector& dependencies = {}); - -sycl::event asum(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, - double* result, const std::vector& dependencies = {}); - -sycl::event axpy(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, const float* x, - std::int64_t incx, float* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event axpy(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, - const double* x, std::int64_t incx, double* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event axpy(sycl::queue& queue, std::int64_t n, value_or_pointer> alpha, - const std::complex* x, std::int64_t incx, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event axpy(sycl::queue& queue, std::int64_t n, value_or_pointer> alpha, - const std::complex* x, std::int64_t incx, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event axpby(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, const float* x, - std::int64_t incx, value_or_pointer beta, float* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event axpby(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, - const double* x, std::int64_t incx, value_or_pointer beta, double* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event axpby(sycl::queue& queue, std::int64_t n, value_or_pointer> alpha, - const std::complex* x, std::int64_t incx, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event axpby(sycl::queue& queue, std::int64_t n, value_or_pointer> alpha, - const std::complex* x, std::int64_t incx, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event copy(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, float* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event copy(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, double* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, std::complex* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event copy(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, std::complex* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, - const float* y, std::int64_t incy, float* result, - const std::vector& dependencies = {}); - -sycl::event dot(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, - const double* y, std::int64_t incy, double* result, - const std::vector& dependencies = {}); - -sycl::event sdsdot(sycl::queue& queue, std::int64_t n, float sb, const float* x, std::int64_t incx, - const float* y, std::int64_t incy, float* result, - const std::vector& dependencies = {}); - -sycl::event dot(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, - const float* y, std::int64_t incy, double* result, - const std::vector& dependencies = {}); - -sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, float* result, - const std::vector& dependencies = {}); - -sycl::event nrm2(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, double* result, - const std::vector& dependencies = {}); - -sycl::event nrm2(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, - float* result, const std::vector& dependencies = {}); - -sycl::event nrm2(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, - double* result, const std::vector& dependencies = {}); - -sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, - std::complex* y, std::int64_t incy, value_or_pointer c, - value_or_pointer s, const std::vector& dependencies = {}); - -sycl::event rot(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, - std::complex* y, std::int64_t incy, value_or_pointer c, - value_or_pointer s, const std::vector& dependencies = {}); - -sycl::event rot(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, - std::int64_t incy, value_or_pointer c, value_or_pointer s, - const std::vector& dependencies = {}); - -sycl::event rot(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, - std::int64_t incy, value_or_pointer c, value_or_pointer s, - const std::vector& dependencies = {}); - -sycl::event rotg(sycl::queue& queue, float* a, float* b, float* c, float* s, - const std::vector& dependencies = {}); - -sycl::event rotg(sycl::queue& queue, double* a, double* b, double* c, double* s, - const std::vector& dependencies = {}); - -sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, float* c, - std::complex* s, const std::vector& dependencies = {}); - -sycl::event rotg(sycl::queue& queue, std::complex* a, std::complex* b, double* c, - std::complex* s, const std::vector& dependencies = {}); - -sycl::event rotm(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, - std::int64_t incy, const float* param, - const std::vector& dependencies = {}); - -sycl::event rotm(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, - std::int64_t incy, const double* param, - const std::vector& dependencies = {}); - -sycl::event rotmg(sycl::queue& queue, float* d1, float* d2, float* x1, value_or_pointer y1, - float* param, const std::vector& dependencies = {}); - -sycl::event rotmg(sycl::queue& queue, double* d1, double* d2, double* x1, - value_or_pointer y1, double* param, - const std::vector& dependencies = {}); - -#define ONEMATH_DECLARE_SCAL(T, Ts) \ - sycl::event scal(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, T* x, \ - std::int64_t incx, const std::vector& dependencies = {}); - -ONEMATH_DECLARE_SCAL(float, float) -ONEMATH_DECLARE_SCAL(double, double) -ONEMATH_DECLARE_SCAL(std::complex, std::complex) -ONEMATH_DECLARE_SCAL(std::complex, std::complex) -ONEMATH_DECLARE_SCAL(std::complex, float) -ONEMATH_DECLARE_SCAL(std::complex, double) -sycl::event scal(sycl::queue& queue, std::int64_t n, float alpha, std::complex* x, - std::int64_t incx, const std::vector& dependencies = {}); -sycl::event scal(sycl::queue& queue, std::int64_t n, double alpha, std::complex* x, - std::int64_t incx, const std::vector& dependencies = {}); - -#undef ONEMATH_DECLARE_SCAL - -sycl::event swap(sycl::queue& queue, std::int64_t n, float* x, std::int64_t incx, float* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event swap(sycl::queue& queue, std::int64_t n, double* x, std::int64_t incx, double* y, - std::int64_t incy, const std::vector& dependencies = {}); - -sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, - std::complex* y, std::int64_t incy, - const std::vector& dependencies = {}); - -sycl::event swap(sycl::queue& queue, std::int64_t n, std::complex* x, std::int64_t incx, - std::complex* y, std::int64_t incy, - const std::vector& dependencies = {}); - -// extensions, buffer - -void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, - std::int64_t k, float alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb, float beta, sycl::buffer& c, - std::int64_t ldc); - -void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, - std::int64_t k, double alpha, sycl::buffer& a, std::int64_t lda, - sycl::buffer& b, std::int64_t ldb, double beta, sycl::buffer& c, - std::int64_t ldc); - -void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); - -void gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, std::int64_t n, - std::int64_t k, std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb, - std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc); - -void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer& a, std::int64_t lda, std::int8_t ao, - sycl::buffer& b, std::int64_t ldb, std::uint8_t bo, float beta, - sycl::buffer& c, std::int64_t ldc, - sycl::buffer& co); - -void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer& a, std::int64_t lda, std::int8_t ao, - sycl::buffer& b, std::int64_t ldb, std::int8_t bo, float beta, - sycl::buffer& c, std::int64_t ldc, - sycl::buffer& co); - -void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer& a, std::int64_t lda, std::uint8_t ao, - sycl::buffer& b, std::int64_t ldb, std::int8_t bo, float beta, - sycl::buffer& c, std::int64_t ldc, - sycl::buffer& co); - -void gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, float alpha, - sycl::buffer& a, std::int64_t lda, std::uint8_t ao, - sycl::buffer& b, std::int64_t ldb, std::uint8_t bo, float beta, - sycl::buffer& c, std::int64_t ldc, - sycl::buffer& co); - -// extensions, USM - -sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const float* a, - std::int64_t lda, const float* b, std::int64_t ldb, value_or_pointer beta, - float* c, std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, value_or_pointer alpha, const double* a, - std::int64_t lda, const double* b, std::int64_t ldb, - value_or_pointer beta, double* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer> beta, - std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event gemmt(sycl::queue& queue, uplo upper_lower, transpose transa, transpose transb, - std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, const std::complex* b, - std::int64_t ldb, value_or_pointer> beta, - std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::int8_t* a, std::int64_t lda, std::int8_t ao, const std::uint8_t* b, - std::int64_t ldb, std::uint8_t bo, value_or_pointer beta, - std::int32_t* c, std::int64_t ldc, const std::int32_t* co, - const std::vector& dependencies = {}); - -sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::int8_t* a, std::int64_t lda, std::int8_t ao, const std::int8_t* b, - std::int64_t ldb, std::int8_t bo, value_or_pointer beta, - std::int32_t* c, std::int64_t ldc, const std::int32_t* co, - const std::vector& dependencies = {}); - -sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, - const std::int8_t* b, std::int64_t ldb, std::int8_t bo, - value_or_pointer beta, std::int32_t* c, std::int64_t ldc, - const std::int32_t* co, const std::vector& dependencies = {}); - -sycl::event gemm_bias(sycl::queue& queue, transpose transa, transpose transb, offset offsetc, - std::int64_t m, std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::uint8_t* a, std::int64_t lda, std::uint8_t ao, - const std::uint8_t* b, std::int64_t ldb, std::uint8_t bo, - value_or_pointer beta, std::int32_t* c, std::int64_t ldc, - const std::int32_t* co, const std::vector& dependencies = {}); - -// batch, buffer - -void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - float alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, float beta, - sycl::buffer& c, int64_t ldc, int64_t stride_c, int64_t batch_size); -void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - double alpha, sycl::buffer& a, int64_t lda, int64_t stride_a, - double beta, sycl::buffer& c, int64_t ldc, int64_t stride_c, - int64_t batch_size); -void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1>& a, int64_t lda, - int64_t stride_a, std::complex beta, sycl::buffer, 1>& c, - int64_t ldc, int64_t stride_c, int64_t batch_size); -void syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, int64_t n, int64_t k, - std::complex alpha, sycl::buffer, 1>& a, int64_t lda, - int64_t stride_a, std::complex beta, - sycl::buffer, 1>& c, int64_t ldc, int64_t stride_c, - int64_t batch_size); - -void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - std::int64_t stridex, sycl::buffer& y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size); - -void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& x, std::int64_t incx, - std::int64_t stridex, sycl::buffer& y, std::int64_t incy, - std::int64_t stridey, std::int64_t batch_size); - -void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -void copy_batch(sycl::queue& queue, std::int64_t n, sycl::buffer, 1>& x, - std::int64_t incx, std::int64_t stridex, sycl::buffer, 1>& y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer& x, std::int64_t incx, std::int64_t stridex, float beta, - sycl::buffer& y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size); - -void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer& x, std::int64_t incx, std::int64_t stridex, double beta, - sycl::buffer& y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size); - -void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size); - -void gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& x, - std::int64_t incx, std::int64_t stridex, std::complex beta, - sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size); - -void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size); - -void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer& c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size); - -void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size); - -void dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, - sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1>& c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size); - -void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size); - -void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size); - -void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size); - -void trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size); - -void axpy_batch(sycl::queue& queue, std::int64_t n, float alpha, sycl::buffer& x, - std::int64_t incx, std::int64_t stridex, sycl::buffer& y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -void axpy_batch(sycl::queue& queue, std::int64_t n, double alpha, sycl::buffer& x, - std::int64_t incx, std::int64_t stridex, sycl::buffer& y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size); - -void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size); - -void axpy_batch(sycl::queue& queue, std::int64_t n, std::complex alpha, - sycl::buffer, 1>& x, std::int64_t incx, std::int64_t stridex, - sycl::buffer, 1>& y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size); - -void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, double alpha, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, - std::int64_t ldb, std::int64_t stride_b, double beta, sycl::buffer& c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1>& c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1>& b, std::int64_t ldb, std::int64_t stride_b, - std::complex beta, sycl::buffer, 1>& c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::half alpha, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, - std::int64_t ldb, std::int64_t stride_b, sycl::half beta, - sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size); - -void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, - std::int64_t ldb, std::int64_t stride_b, float beta, - sycl::buffer& c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size); - -void gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, float alpha, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& b, - std::int64_t ldb, std::int64_t stride_b, float beta, sycl::buffer& c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size); - -void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer& a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer& b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size); - -void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1>& b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size); - -void omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, std::int64_t stride_a, - sycl::buffer, 1>& b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size); - -void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, - std::int64_t stride, std::int64_t batch_size); - -void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer& ab, std::int64_t lda, std::int64_t ldb, - std::int64_t stride, std::int64_t batch_size); - -void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& ab, - std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size); - -void imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& ab, - std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size); - -void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, float alpha, sycl::buffer& a, std::int64_t lda, - std::int64_t stride_a, float beta, sycl::buffer& b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, double alpha, sycl::buffer& a, std::int64_t lda, - std::int64_t stride_a, double beta, sycl::buffer& b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer& c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size); - -void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, std::int64_t stride_a, - std::complex beta, sycl::buffer, 1>& b, - std::int64_t ldb, std::int64_t stride_b, sycl::buffer, 1>& c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -void omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::complex alpha, - sycl::buffer, 1>& a, std::int64_t lda, - std::int64_t stride_a, std::complex beta, - sycl::buffer, 1>& b, std::int64_t ldb, - std::int64_t stride_b, sycl::buffer, 1>& c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size); - -void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer& a, std::int64_t lda, sycl::buffer& b, - std::int64_t ldb); - -void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer& a, std::int64_t lda, sycl::buffer& b, - std::int64_t ldb); - -void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - sycl::buffer, 1>& b, std::int64_t ldb); - -void omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, sycl::buffer, 1>& b, std::int64_t ldb); - -void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); - -void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer& a, std::int64_t lda, std::int64_t stridea, - sycl::buffer& b, std::int64_t ldb, std::int64_t strideb); - -void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - std::int64_t stridea, sycl::buffer, 1>& b, std::int64_t ldb, - std::int64_t strideb); - -void omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, - std::int64_t lda, std::int64_t stridea, sycl::buffer, 1>& b, - std::int64_t ldb, std::int64_t strideb); - -void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, float alpha, - sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); - -void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, double alpha, - sycl::buffer& ab, std::int64_t lda, std::int64_t ldb); - -void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& ab, std::int64_t lda, - std::int64_t ldb); - -void imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& ab, - std::int64_t lda, std::int64_t ldb); - -void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - float alpha, sycl::buffer& a, std::int64_t lda, float beta, - sycl::buffer& b, std::int64_t ldb, sycl::buffer& c, - std::int64_t ldc); - -void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - double alpha, sycl::buffer& a, std::int64_t lda, double beta, - sycl::buffer& b, std::int64_t ldb, sycl::buffer& c, - std::int64_t ldc); - -void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - std::complex beta, sycl::buffer, 1>& b, std::int64_t ldb, - sycl::buffer, 1>& c, std::int64_t ldc); - -void omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, std::int64_t n, - std::complex alpha, sycl::buffer, 1>& a, std::int64_t lda, - std::complex beta, sycl::buffer, 1>& b, std::int64_t ldb, - sycl::buffer, 1>& c, std::int64_t ldc); - -// batch, usm - -sycl::event syrk_batch(sycl::queue& queue, const uplo* upper_lower, const transpose* trans, - const std::int64_t* n, const std::int64_t* k, const float* alpha, - const float** a, const std::int64_t* lda, const float* beta, float** c, - const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); -sycl::event syrk_batch(sycl::queue& queue, const uplo* upper_lower, const transpose* trans, - const std::int64_t* n, const std::int64_t* k, const double* alpha, - const double** a, const std::int64_t* lda, const double* beta, double** c, - const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); -sycl::event syrk_batch(sycl::queue& queue, const uplo* upper_lower, const transpose* trans, - const std::int64_t* n, const std::int64_t* k, - const std::complex* alpha, const std::complex** a, - const std::int64_t* lda, const std::complex* beta, - std::complex** c, const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); -sycl::event syrk_batch(sycl::queue& queue, const uplo* upper_lower, const transpose* trans, - const std::int64_t* n, const std::int64_t* k, - const std::complex* alpha, const std::complex** a, - const std::int64_t* lda, const std::complex* beta, - std::complex** c, const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const float* a, - std::int64_t lda, std::int64_t stride_a, value_or_pointer beta, - float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); -sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer alpha, const double* a, - std::int64_t lda, std::int64_t stride_a, value_or_pointer beta, - double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); -sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, std::int64_t stride_a, - value_or_pointer> beta, std::complex* c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); -sycl::event syrk_batch(sycl::queue& queue, uplo upper_lower, transpose trans, std::int64_t n, - std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, std::int64_t stride_a, - value_or_pointer> beta, std::complex* c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const float* x, std::int64_t incx, - std::int64_t stridex, float* y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector& dependencies = {}); - -sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const double* x, std::int64_t incx, - std::int64_t stridex, double* y, std::int64_t incy, std::int64_t stridey, - std::int64_t batch_size, const std::vector& dependencies = {}); - -sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, std::int64_t stridex, std::complex* y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event copy_batch(sycl::queue& queue, std::int64_t n, const std::complex* x, - std::int64_t incx, std::int64_t stridex, std::complex* y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event copy_batch(sycl::queue& queue, const std::int64_t* n, const float** x, - const std::int64_t* incx, float** y, const std::int64_t* incy, - std::int64_t group_count, const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event copy_batch(sycl::queue& queue, const std::int64_t* n, const double** x, - const std::int64_t* incx, double** y, const std::int64_t* incy, - std::int64_t group_count, const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event copy_batch(sycl::queue& queue, const std::int64_t* n, const std::complex** x, - const std::int64_t* incx, std::complex** y, const std::int64_t* incy, - std::int64_t group_count, const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event copy_batch(sycl::queue& queue, const std::int64_t* n, const std::complex** x, - const std::int64_t* incx, std::complex** y, const std::int64_t* incy, - std::int64_t group_count, const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, - const float* a, std::int64_t lda, std::int64_t stridea, const float* x, - std::int64_t incx, std::int64_t stridex, float* c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, - const double* a, std::int64_t lda, std::int64_t stridea, const double* x, - std::int64_t incx, std::int64_t stridex, double* c, std::int64_t ldc, - std::int64_t stridec, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, - const std::complex* a, std::int64_t lda, std::int64_t stridea, - const std::complex* x, std::int64_t incx, std::int64_t stridex, - std::complex* c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector& dependencies = {}); - -sycl::event dgmm_batch(sycl::queue& queue, side left_right, std::int64_t m, std::int64_t n, - const std::complex* a, std::int64_t lda, std::int64_t stridea, - const std::complex* x, std::int64_t incx, std::int64_t stridex, - std::complex* c, std::int64_t ldc, std::int64_t stridec, - std::int64_t batch_size, const std::vector& dependencies = {}); - -sycl::event dgmm_batch(sycl::queue& queue, const side* left_right, const std::int64_t* m, - const std::int64_t* n, const float** a, const std::int64_t* lda, - const float** x, const std::int64_t* incx, float** c, - const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event dgmm_batch(sycl::queue& queue, const side* left_right, const std::int64_t* m, - const std::int64_t* n, const double** a, const std::int64_t* lda, - const double** x, const std::int64_t* incx, double** c, - const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event dgmm_batch(sycl::queue& queue, const side* left_right, const std::int64_t* m, - const std::int64_t* n, const std::complex** a, - const std::int64_t* lda, const std::complex** x, - const std::int64_t* incx, std::complex** c, const std::int64_t* ldc, - std::int64_t group_count, const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event dgmm_batch(sycl::queue& queue, const side* left_right, const std::int64_t* m, - const std::int64_t* n, const std::complex** a, - const std::int64_t* lda, const std::complex** x, - const std::int64_t* incx, std::complex** c, const std::int64_t* ldc, - std::int64_t group_count, const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const float* a, std::int64_t lda, - std::int64_t stridea, const float* x, std::int64_t incx, - std::int64_t stridex, value_or_pointer beta, float* y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const double* a, std::int64_t lda, - std::int64_t stridea, const double* x, std::int64_t incx, - std::int64_t stridex, value_or_pointer beta, double* y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::int64_t stridea, const std::complex* x, - std::int64_t incx, std::int64_t stridex, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemv_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::int64_t stridea, const std::complex* x, - std::int64_t incx, std::int64_t stridex, - value_or_pointer> beta, std::complex* y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemv_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const float* alpha, const float** a, - const std::int64_t* lda, const float** x, const std::int64_t* incx, - const float* beta, float** y, const std::int64_t* incy, - std::int64_t group_count, const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event gemv_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const double* alpha, const double** a, - const std::int64_t* lda, const double** x, const std::int64_t* incx, - const double* beta, double** y, const std::int64_t* incy, - std::int64_t group_count, const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event gemv_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const std::complex* alpha, - const std::complex** a, const std::int64_t* lda, - const std::complex** x, const std::int64_t* incx, - const std::complex* beta, std::complex** y, - const std::int64_t* incy, std::int64_t group_count, - const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event gemv_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const std::complex* alpha, - const std::complex** a, const std::int64_t* lda, - const std::complex** x, const std::int64_t* incx, - const std::complex* beta, std::complex** y, - const std::int64_t* incy, std::int64_t group_count, - const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event axpy_batch(sycl::queue& queue, const std::int64_t* n, const double* alpha, - const double** x, const std::int64_t* incx, double** y, - const std::int64_t* incy, std::int64_t group_count, - const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event axpy_batch(sycl::queue& queue, const std::int64_t* n, const float* alpha, - const float** x, const std::int64_t* incx, float** y, - const std::int64_t* incy, std::int64_t group_count, - const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event axpy_batch(sycl::queue& queue, const std::int64_t* n, const std::complex* alpha, - const std::complex** x, const std::int64_t* incx, - std::complex** y, const std::int64_t* incy, std::int64_t group_count, - const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event axpy_batch(sycl::queue& queue, const std::int64_t* n, const std::complex* alpha, - const std::complex** x, const std::int64_t* incx, - std::complex** y, const std::int64_t* incy, std::int64_t group_count, - const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, - const float* x, std::int64_t incx, std::int64_t stridex, float* y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, value_or_pointer alpha, - const double* x, std::int64_t incx, std::int64_t stridex, double* y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, - value_or_pointer> alpha, const std::complex* x, - std::int64_t incx, std::int64_t stridex, std::complex* y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event axpy_batch(sycl::queue& queue, std::int64_t n, - value_or_pointer> alpha, const std::complex* x, - std::int64_t incx, std::int64_t stridex, std::complex* y, - std::int64_t incy, std::int64_t stridey, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, - const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, - const float* alpha, const float** a, const std::int64_t* lda, - const float** b, const std::int64_t* ldb, const float* beta, float** c, - const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, - const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, - const double* alpha, const double** a, const std::int64_t* lda, - const double** b, const std::int64_t* ldb, const double* beta, double** c, - const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, - const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, - const std::complex* alpha, const std::complex** a, - const std::int64_t* lda, const std::complex** b, - const std::int64_t* ldb, const std::complex* beta, - std::complex** c, const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, - const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, - const std::complex* alpha, const std::complex** a, - const std::int64_t* lda, const std::complex** b, - const std::int64_t* ldb, const std::complex* beta, - std::complex** c, const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, - const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, - const sycl::half* alpha, const sycl::half** a, const std::int64_t* lda, - const sycl::half** b, const std::int64_t* ldb, const sycl::half* beta, - sycl::half** c, const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, - const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, - const float* alpha, const sycl::half** a, const std::int64_t* lda, - const sycl::half** b, const std::int64_t* ldb, const float* beta, float** c, - const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, - const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, - const float* alpha, const bfloat16** a, const std::int64_t* lda, - const bfloat16** b, const std::int64_t* ldb, const float* beta, bfloat16** c, - const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, - const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, - const float* alpha, const bfloat16** a, const std::int64_t* lda, - const bfloat16** b, const std::int64_t* ldb, const float* beta, float** c, - const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, - const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, - const float* alpha, const std::int8_t** a, const std::int64_t* lda, - const std::int8_t** b, const std::int64_t* ldb, const float* beta, - std::int32_t** c, const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, const transpose* transa, const transpose* transb, - const std::int64_t* m, const std::int64_t* n, const std::int64_t* k, - const float* alpha, const std::int8_t** a, const std::int64_t* lda, - const std::int8_t** b, const std::int64_t* ldb, const float* beta, float** c, - const std::int64_t* ldc, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, - const float* a, std::int64_t lda, std::int64_t stride_a, const float* b, - std::int64_t ldb, std::int64_t stride_b, value_or_pointer beta, - float* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, - const double* a, std::int64_t lda, std::int64_t stride_a, const double* b, - std::int64_t ldb, std::int64_t stride_b, value_or_pointer beta, - double* c, std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, std::int64_t stride_a, - const std::complex* b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer> beta, std::complex* c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, std::int64_t stride_a, - const std::complex* b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer> beta, std::complex* c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, - const sycl::half* a, std::int64_t lda, std::int64_t stride_a, - const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, sycl::half* c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, - const sycl::half* a, std::int64_t lda, std::int64_t stride_a, - const sycl::half* b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, float* c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, - const bfloat16* a, std::int64_t lda, std::int64_t stride_a, - const bfloat16* b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, bfloat16* c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, - const bfloat16* a, std::int64_t lda, std::int64_t stride_a, - const bfloat16* b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, float* c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, std::int32_t* c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event gemm_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, std::int64_t k, value_or_pointer alpha, - const std::int8_t* a, std::int64_t lda, std::int64_t stride_a, - const std::int8_t* b, std::int64_t ldb, std::int64_t stride_b, - value_or_pointer beta, float* c, std::int64_t ldc, - std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const float* a, std::int64_t lda, - std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector& dependencies = {}); - -sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const double* a, std::int64_t lda, - std::int64_t stride_a, double* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, const std::vector& dependencies = {}); - -sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::int64_t stride_a, std::complex* b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event trsm_batch(sycl::queue& queue, side left_right, uplo upper_lower, transpose trans, - diag unit_diag, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::int64_t stride_a, std::complex* b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event trsm_batch(sycl::queue& queue, const side* left_right, const uplo* upper_lower, - const transpose* trans, const diag* unit_diag, const std::int64_t* m, - const std::int64_t* n, const float* alpha, const float** a, - const std::int64_t* lda, float** b, const std::int64_t* ldb, - std::int64_t group_count, const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event trsm_batch(sycl::queue& queue, const side* left_right, const uplo* upper_lower, - const transpose* trans, const diag* unit_diag, const std::int64_t* m, - const std::int64_t* n, const double* alpha, const double** a, - const std::int64_t* lda, double** b, const std::int64_t* ldb, - std::int64_t group_count, const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event trsm_batch(sycl::queue& queue, const side* left_right, const uplo* upper_lower, - const transpose* trans, const diag* unit_diag, const std::int64_t* m, - const std::int64_t* n, const std::complex* alpha, - const std::complex** a, const std::int64_t* lda, - std::complex** b, const std::int64_t* ldb, std::int64_t group_count, - const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event trsm_batch(sycl::queue& queue, const side* left_right, const uplo* upper_lower, - const transpose* trans, const diag* unit_diag, const std::int64_t* m, - const std::int64_t* n, const std::complex* alpha, - const std::complex** a, const std::int64_t* lda, - std::complex** b, const std::int64_t* ldb, std::int64_t group_count, - const std::int64_t* group_size, - const std::vector& dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const float* a, std::int64_t lda, - std::int64_t stride_a, float* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const double* a, std::int64_t lda, - std::int64_t stride_a, double* b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, float* ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, double* ab, std::int64_t lda, - std::int64_t ldb, std::int64_t stride, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, std::complex* ab, - std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, std::complex* ab, - std::int64_t lda, std::int64_t ldb, std::int64_t stride, - std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const float* a, - std::int64_t lda, std::int64_t stride_a, value_or_pointer beta, - const float* b, std::int64_t ldb, std::int64_t stride_b, float* c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const double* a, - std::int64_t lda, std::int64_t stride_a, value_or_pointer beta, - const double* b, std::int64_t ldb, std::int64_t stride_b, double* c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, std::int64_t stride_a, - value_or_pointer> beta, const std::complex* b, - std::int64_t ldb, std::int64_t stride_b, std::complex* c, - std::int64_t ldc, std::int64_t stride_c, std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event omatadd_batch(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, std::int64_t stride_a, - value_or_pointer> beta, - const std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::complex* c, std::int64_t ldc, std::int64_t stride_c, - std::int64_t batch_size, - const std::vector& dependencies = {}); - -sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const float* a, std::int64_t lda, float* b, - std::int64_t ldb, const std::vector& dependencies = {}); - -sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const double* a, std::int64_t lda, double* b, - std::int64_t ldb, const std::vector& dependencies = {}); - -sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::complex* b, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event omatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::complex* b, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const float* a, std::int64_t lda, - std::int64_t stridea, float* b, std::int64_t ldb, std::int64_t strideb, - const std::vector& dependencies = {}); - -sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, const double* a, std::int64_t lda, - std::int64_t stridea, double* b, std::int64_t ldb, std::int64_t strideb, - const std::vector& dependencies = {}); - -sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::int64_t stridea, std::complex* b, - std::int64_t ldb, std::int64_t strideb, - const std::vector& dependencies = {}); - -sycl::event omatcopy2(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, const std::complex* a, - std::int64_t lda, std::int64_t stridea, std::complex* b, - std::int64_t ldb, std::int64_t strideb, - const std::vector& dependencies = {}); - -sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, float* ab, std::int64_t lda, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer alpha, double* ab, std::int64_t lda, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, std::complex* ab, - std::int64_t lda, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event imatcopy(sycl::queue& queue, transpose trans, std::int64_t m, std::int64_t n, - value_or_pointer> alpha, std::complex* ab, - std::int64_t lda, std::int64_t ldb, - const std::vector& dependencies = {}); - -sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const float* a, std::int64_t lda, - value_or_pointer beta, const float* b, std::int64_t ldb, float* c, - std::int64_t ldc, const std::vector& dependencies = {}); - -sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer alpha, const double* a, - std::int64_t lda, value_or_pointer beta, const double* b, - std::int64_t ldb, double* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, - value_or_pointer> beta, const std::complex* b, - std::int64_t ldb, std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event omatadd(sycl::queue& queue, transpose transa, transpose transb, std::int64_t m, - std::int64_t n, value_or_pointer> alpha, - const std::complex* a, std::int64_t lda, - value_or_pointer> beta, const std::complex* b, - std::int64_t ldb, std::complex* c, std::int64_t ldc, - const std::vector& dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const float* alpha, const float** a, - const std::int64_t* lda, float** b, const std::int64_t* ldb, - std::int64_t group_count, const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const double* alpha, const double** a, - const std::int64_t* lda, double** b, const std::int64_t* ldb, - std::int64_t group_count, const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const std::complex* alpha, - const std::complex** a, const std::int64_t* lda, - std::complex** b, const std::int64_t* ldb, - std::int64_t group_count, const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event omatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const std::complex* alpha, - const std::complex** a, const std::int64_t* lda, - std::complex** b, const std::int64_t* ldb, - std::int64_t group_count, const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const float* alpha, float** ab, - const std::int64_t* lda, const std::int64_t* ldb, - std::int64_t group_count, const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const double* alpha, double** ab, - const std::int64_t* lda, const std::int64_t* ldb, - std::int64_t group_count, const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const std::complex* alpha, - std::complex** ab, const std::int64_t* lda, - const std::int64_t* ldb, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); - -sycl::event imatcopy_batch(sycl::queue& queue, const transpose* trans, const std::int64_t* m, - const std::int64_t* n, const std::complex* alpha, - std::complex** ab, const std::int64_t* lda, - const std::int64_t* ldb, std::int64_t group_count, - const std::int64_t* groupsize, - const std::vector& dependencies = {}); diff --git a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp b/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp deleted file mode 100644 index 992659a67..000000000 --- a/src/lapack/backends/mkl_common/mkl_lapack_backend.hpp +++ /dev/null @@ -1,1276 +0,0 @@ -/******************************************************************************* -* Copyright 2021 Intel Corporation -* -* Licensed under the Apache License, Version 2.0 (the "License"); -* you may not use this file except in compliance with the License. -* You may obtain a copy of the License at -* -* http://www.apache.org/licenses/LICENSE-2.0 -* -* Unless required by applicable law or agreed to in writing, -* software distributed under the License is distributed on an "AS IS" BASIS, -* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -* See the License for the specific language governing permissions -* and limitations under the License. -* -* -* SPDX-License-Identifier: Apache-2.0 -*******************************************************************************/ - -#pragma once - -#include - -#include "oneapi/mkl/types.hpp" -#include "oneapi/mkl/lapack/types.hpp" - -namespace oneapi { -namespace mkl { -namespace lapack { - -void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, - std::int64_t lda, sycl::buffer& d, sycl::buffer& e, - sycl::buffer>& tauq, sycl::buffer>& taup, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& d, sycl::buffer& e, - sycl::buffer& tauq, sycl::buffer& taup, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& d, sycl::buffer& e, - sycl::buffer& tauq, sycl::buffer& taup, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, - sycl::buffer& e, sycl::buffer>& tauq, - sycl::buffer>& taup, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, - std::int64_t lda, sycl::buffer>& tau, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, - std::int64_t lda, sycl::buffer>& tau, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer>& a, - std::int64_t lda, sycl::buffer& ipiv, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer& ipiv, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, - std::int64_t lda, sycl::buffer& ipiv, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, - sycl::buffer& ipiv, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void getri(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, - sycl::buffer& ipiv, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void getri(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, - std::int64_t lda, sycl::buffer& ipiv, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getrs(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, - sycl::buffer>& b, std::int64_t ldb, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getrs(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, - sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void getrs(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer& a, std::int64_t lda, sycl::buffer& ipiv, - sycl::buffer& b, std::int64_t ldb, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void getrs(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, std::int64_t nrhs, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer& ipiv, sycl::buffer>& b, - std::int64_t ldb, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void gesvd(sycl::queue& queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& s, - sycl::buffer& u, std::int64_t ldu, sycl::buffer& vt, std::int64_t ldvt, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void gesvd(sycl::queue& queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& s, - sycl::buffer& u, std::int64_t ldu, sycl::buffer& vt, std::int64_t ldvt, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void gesvd(sycl::queue& queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer>& a, std::int64_t lda, - sycl::buffer& s, sycl::buffer>& u, std::int64_t ldu, - sycl::buffer>& vt, std::int64_t ldvt, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void gesvd(sycl::queue& queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, std::int64_t m, - std::int64_t n, sycl::buffer>& a, std::int64_t lda, - sycl::buffer& s, sycl::buffer>& u, std::int64_t ldu, - sycl::buffer>& vt, std::int64_t ldvt, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void heevd(sycl::queue& queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, sycl::buffer& w, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void heevd(sycl::queue& queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, sycl::buffer& w, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, - std::int64_t n, sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& b, std::int64_t ldb, sycl::buffer& w, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void hegvd(sycl::queue& queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, - std::int64_t n, sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& b, std::int64_t ldb, sycl::buffer& w, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void hetrd(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, - sycl::buffer& e, sycl::buffer>& tau, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void hetrd(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, sycl::buffer& d, - sycl::buffer& e, sycl::buffer>& tau, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void hetrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void hetrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer& ipiv, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void orgbr(sycl::queue& queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void orgbr(sycl::queue& queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer& a, std::int64_t lda, sycl::buffer& tau, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void orgtr(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void orgtr(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& tau, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void ormtr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void ormtr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void ormrq(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, - sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void ormrq(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, - sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void ormqr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, - sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void ormqr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer& a, std::int64_t lda, - sycl::buffer& tau, sycl::buffer& c, std::int64_t ldc, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void potrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void potri(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer& a, std::int64_t lda, sycl::buffer& b, std::int64_t ldb, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& b, std::int64_t ldb, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void potrs(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& b, std::int64_t ldb, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void syevd(sycl::queue& queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer& a, std::int64_t lda, sycl::buffer& w, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void syevd(sycl::queue& queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer& a, std::int64_t lda, sycl::buffer& w, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, - std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, - std::int64_t ldb, sycl::buffer& w, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void sygvd(sycl::queue& queue, std::int64_t itype, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, - std::int64_t n, sycl::buffer& a, std::int64_t lda, sycl::buffer& b, - std::int64_t ldb, sycl::buffer& w, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void sytrd(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& d, sycl::buffer& e, - sycl::buffer& tau, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void sytrd(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& d, sycl::buffer& e, - sycl::buffer& tau, sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void sytrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void sytrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer& a, - std::int64_t lda, sycl::buffer& ipiv, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void sytrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, sycl::buffer& ipiv, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void sytrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer& ipiv, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void trtrs(sycl::queue& queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& b, std::int64_t ldb, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void trtrs(sycl::queue& queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer& a, - std::int64_t lda, sycl::buffer& b, std::int64_t ldb, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void trtrs(sycl::queue& queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, sycl::buffer& a, - std::int64_t lda, sycl::buffer& b, std::int64_t ldb, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void trtrs(sycl::queue& queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& b, std::int64_t ldb, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void ungbr(sycl::queue& queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void ungbr(sycl::queue& queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void ungtr(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void ungtr(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void unmrq(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& c, - std::int64_t ldc, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void unmrq(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& c, - std::int64_t ldc, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void unmqr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& c, - std::int64_t ldc, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void unmqr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, std::int64_t m, - std::int64_t n, std::int64_t k, sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& c, - std::int64_t ldc, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void unmtr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& c, - std::int64_t ldc, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void unmtr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, - sycl::buffer>& tau, sycl::buffer>& c, - std::int64_t ldc, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, - std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& tau, - std::int64_t stride_tau, std::int64_t batch_size, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer>& tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer>& tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer& a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getri_batch(sycl::queue& queue, std::int64_t n, sycl::buffer>& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer>& scratchpad, std::int64_t scratchpad_size); -void getrs_batch(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer& ipiv, std::int64_t stride_ipiv, sycl::buffer& b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getrs_batch(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer& a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, - sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void getrs_batch(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, - sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void getrs_batch(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, sycl::buffer>& a, std::int64_t lda, - std::int64_t stride_a, sycl::buffer& ipiv, std::int64_t stride_ipiv, - sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, sycl::buffer& ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer& ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer& ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer& tau, std::int64_t stride_tau, std::int64_t batch_size, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potrf_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, sycl::buffer& a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - sycl::buffer& scratchpad, std::int64_t scratchpad_size); -void potrf_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void potrf_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void potrf_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void potrs_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void potrs_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer& scratchpad, - std::int64_t scratchpad_size); -void potrs_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void potrs_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer>& b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer>& tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -void ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - sycl::buffer>& a, std::int64_t lda, std::int64_t stride_a, - sycl::buffer>& tau, std::int64_t stride_tau, - std::int64_t batch_size, sycl::buffer>& scratchpad, - std::int64_t scratchpad_size); -sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, float* d, float* e, std::complex* tauq, - std::complex* taup, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, - double* d, double* e, double* tauq, double* taup, double* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, - float* d, float* e, float* tauq, float* taup, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event gebrd(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, double* d, double* e, std::complex* tauq, - std::complex* taup, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, - float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, - double* tau, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, std::complex* tau, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event gerqf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, std::complex* tau, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, std::complex* tau, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, - double* tau, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, - float* tau, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event geqrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, std::complex* tau, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, std::int64_t* ipiv, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, std::int64_t lda, - std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, std::int64_t lda, - std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrf(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, std::int64_t* ipiv, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, std::int64_t lda, - const std::int64_t* ipiv, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event getri(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, - const std::int64_t* ipiv, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getri(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, - const std::int64_t* ipiv, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getri(sycl::queue& queue, std::int64_t n, std::complex* a, std::int64_t lda, - const std::int64_t* ipiv, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event getrs(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, const std::complex* a, std::int64_t lda, - const std::int64_t* ipiv, std::complex* b, std::int64_t ldb, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrs(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, const double* a, std::int64_t lda, const std::int64_t* ipiv, - double* b, std::int64_t ldb, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrs(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, const float* a, std::int64_t lda, const std::int64_t* ipiv, - float* b, std::int64_t ldb, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrs(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, const std::complex* a, std::int64_t lda, - const std::int64_t* ipiv, std::complex* b, std::int64_t ldb, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event gesvd(sycl::queue& queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, - std::int64_t m, std::int64_t n, double* a, std::int64_t lda, double* s, double* u, - std::int64_t ldu, double* vt, std::int64_t ldvt, double* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event gesvd(sycl::queue& queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, - std::int64_t m, std::int64_t n, float* a, std::int64_t lda, float* s, float* u, - std::int64_t ldu, float* vt, std::int64_t ldvt, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event gesvd(sycl::queue& queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, - std::int64_t m, std::int64_t n, std::complex* a, std::int64_t lda, - float* s, std::complex* u, std::int64_t ldu, std::complex* vt, - std::int64_t ldvt, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event gesvd(sycl::queue& queue, oneapi::mkl::jobsvd jobu, oneapi::mkl::jobsvd jobvt, - std::int64_t m, std::int64_t n, std::complex* a, std::int64_t lda, - double* s, std::complex* u, std::int64_t ldu, std::complex* vt, - std::int64_t ldvt, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event heevd(sycl::queue& queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, float* w, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event heevd(sycl::queue& queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, double* w, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex* a, std::int64_t lda, - std::complex* b, std::int64_t ldb, float* w, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event hegvd(sycl::queue& queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::complex* a, std::int64_t lda, - std::complex* b, std::int64_t ldb, double* w, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event hetrd(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, float* d, float* e, - std::complex* tau, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event hetrd(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, double* d, double* e, - std::complex* tau, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event hetrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, std::int64_t* ipiv, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event hetrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, std::int64_t* ipiv, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event orgbr(sycl::queue& queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, float* a, std::int64_t lda, const float* tau, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event orgbr(sycl::queue& queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, double* a, std::int64_t lda, const double* tau, - double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, double* a, - std::int64_t lda, const double* tau, double* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event orgqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, float* a, - std::int64_t lda, const float* tau, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event orgtr(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, float* a, - std::int64_t lda, const float* tau, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event orgtr(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, double* a, - std::int64_t lda, const double* tau, double* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event ormtr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, const float* a, - std::int64_t lda, const float* tau, float* c, std::int64_t ldc, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event ormtr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, const double* a, - std::int64_t lda, const double* tau, double* c, std::int64_t ldc, - double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event ormrq(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, - const float* tau, float* c, std::int64_t ldc, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event ormrq(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, - const double* tau, double* c, std::int64_t ldc, double* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event ormqr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, const double* a, std::int64_t lda, - const double* tau, double* c, std::int64_t ldc, double* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event ormqr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, const float* a, std::int64_t lda, - const float* tau, float* c, std::int64_t ldc, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event potrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, float* a, - std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, double* a, - std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event potrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event potri(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, float* a, - std::int64_t lda, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potri(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, double* a, - std::int64_t lda, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potri(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event potri(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event potrs(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - const float* a, std::int64_t lda, float* b, std::int64_t ldb, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event potrs(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - const double* a, std::int64_t lda, double* b, std::int64_t ldb, - double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrs(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - const std::complex* a, std::int64_t lda, std::complex* b, - std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrs(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t nrhs, - const std::complex* a, std::int64_t lda, std::complex* b, - std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event syevd(sycl::queue& queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, - double* a, std::int64_t lda, double* w, double* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event syevd(sycl::queue& queue, oneapi::mkl::job jobz, oneapi::mkl::uplo uplo, std::int64_t n, - float* a, std::int64_t lda, float* w, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, double* a, std::int64_t lda, double* b, - std::int64_t ldb, double* w, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event sygvd(sycl::queue& queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, float* a, std::int64_t lda, float* b, - std::int64_t ldb, float* w, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event sytrd(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, double* a, - std::int64_t lda, double* d, double* e, double* tau, double* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event sytrd(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, float* a, - std::int64_t lda, float* d, float* e, float* tau, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event sytrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, float* a, - std::int64_t lda, std::int64_t* ipiv, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event sytrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, double* a, - std::int64_t lda, std::int64_t* ipiv, double* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event sytrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, std::int64_t* ipiv, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event sytrf(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, std::int64_t* ipiv, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event trtrs(sycl::queue& queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, - const std::complex* a, std::int64_t lda, std::complex* b, - std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event trtrs(sycl::queue& queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, const double* a, - std::int64_t lda, double* b, std::int64_t ldb, double* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event trtrs(sycl::queue& queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, const float* a, - std::int64_t lda, float* b, std::int64_t ldb, float* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event trtrs(sycl::queue& queue, oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - oneapi::mkl::diag diag, std::int64_t n, std::int64_t nrhs, - const std::complex* a, std::int64_t lda, std::complex* b, - std::int64_t ldb, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event ungbr(sycl::queue& queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex* a, std::int64_t lda, - const std::complex* tau, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event ungbr(sycl::queue& queue, oneapi::mkl::generate vec, std::int64_t m, std::int64_t n, - std::int64_t k, std::complex* a, std::int64_t lda, - const std::complex* tau, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex* a, std::int64_t lda, const std::complex* tau, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event ungqr(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex* a, std::int64_t lda, const std::complex* tau, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event ungtr(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, const std::complex* tau, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event ungtr(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, const std::complex* tau, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event unmrq(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, const std::complex* a, - std::int64_t lda, const std::complex* tau, std::complex* c, - std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event unmrq(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, const std::complex* a, - std::int64_t lda, const std::complex* tau, std::complex* c, - std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event unmqr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, const std::complex* a, - std::int64_t lda, const std::complex* tau, std::complex* c, - std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event unmqr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t k, const std::complex* a, - std::int64_t lda, const std::complex* tau, std::complex* c, - std::int64_t ldc, std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event unmtr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, - const std::complex* a, std::int64_t lda, const std::complex* tau, - std::complex* c, std::int64_t ldc, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event unmtr(sycl::queue& queue, oneapi::mkl::side side, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, - const std::complex* a, std::int64_t lda, const std::complex* tau, - std::complex* c, std::int64_t ldc, std::complex* scratchpad, - std::int64_t scratchpad_size, const std::vector& dependencies = {}); -sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, - std::int64_t lda, std::int64_t stride_a, float* tau, - std::int64_t stride_tau, std::int64_t batch_size, float* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, - std::int64_t lda, std::int64_t stride_a, double* tau, - std::int64_t stride_tau, std::int64_t batch_size, double* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, std::int64_t stride_a, std::complex* tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event geqrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, std::int64_t stride_a, std::complex* tau, - std::int64_t stride_tau, std::int64_t batch_size, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, float** a, - std::int64_t* lda, float** tau, std::int64_t group_count, - std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, double** a, - std::int64_t* lda, double** tau, std::int64_t group_count, - std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, - std::complex** a, std::int64_t* lda, std::complex** tau, - std::int64_t group_count, std::int64_t* group_sizes, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event geqrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, - std::complex** a, std::int64_t* lda, std::complex** tau, - std::int64_t group_count, std::int64_t* group_sizes, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, float* a, - std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, float* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, double* a, - std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, double* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrf_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::complex* a, - std::int64_t lda, std::int64_t stride_a, std::int64_t* ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, float** a, - std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, - std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, double** a, - std::int64_t* lda, std::int64_t** ipiv, std::int64_t group_count, - std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, - std::complex** a, std::int64_t* lda, std::int64_t** ipiv, - std::int64_t group_count, std::int64_t* group_sizes, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrf_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, - std::complex** a, std::int64_t* lda, std::int64_t** ipiv, - std::int64_t group_count, std::int64_t* group_sizes, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getri_batch(sycl::queue& queue, std::int64_t n, float* a, std::int64_t lda, - std::int64_t stride_a, const std::int64_t* ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getri_batch(sycl::queue& queue, std::int64_t n, double* a, std::int64_t lda, - std::int64_t stride_a, const std::int64_t* ipiv, std::int64_t stride_ipiv, - std::int64_t batch_size, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, - std::int64_t lda, std::int64_t stride_a, const std::int64_t* ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getri_batch(sycl::queue& queue, std::int64_t n, std::complex* a, - std::int64_t lda, std::int64_t stride_a, const std::int64_t* ipiv, - std::int64_t stride_ipiv, std::int64_t batch_size, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, float** a, std::int64_t* lda, - const std::int64_t* const* ipiv, std::int64_t group_count, - std::int64_t* group_sizes, float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, double** a, std::int64_t* lda, - const std::int64_t* const* ipiv, std::int64_t group_count, - std::int64_t* group_sizes, double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, - std::int64_t* lda, const std::int64_t* const* ipiv, - std::int64_t group_count, std::int64_t* group_sizes, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getri_batch(sycl::queue& queue, std::int64_t* n, std::complex** a, - std::int64_t* lda, const std::int64_t* const* ipiv, - std::int64_t group_count, std::int64_t* group_sizes, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrs_batch(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, const float* a, std::int64_t lda, std::int64_t stride_a, - const std::int64_t* ipiv, std::int64_t stride_ipiv, float* b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrs_batch(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, const double* a, std::int64_t lda, std::int64_t stride_a, - const std::int64_t* ipiv, std::int64_t stride_ipiv, double* b, - std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrs_batch(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, const std::complex* a, std::int64_t lda, - std::int64_t stride_a, const std::int64_t* ipiv, std::int64_t stride_ipiv, - std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrs_batch(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, const std::complex* a, std::int64_t lda, - std::int64_t stride_a, const std::int64_t* ipiv, std::int64_t stride_ipiv, - std::complex* b, std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrs_batch(sycl::queue& queue, oneapi::mkl::transpose* trans, std::int64_t* n, - std::int64_t* nrhs, const float* const* a, std::int64_t* lda, - const std::int64_t* const* ipiv, float** b, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* group_sizes, float* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrs_batch(sycl::queue& queue, oneapi::mkl::transpose* trans, std::int64_t* n, - std::int64_t* nrhs, const double* const* a, std::int64_t* lda, - const std::int64_t* const* ipiv, double** b, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* group_sizes, double* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrs_batch(sycl::queue& queue, oneapi::mkl::transpose* trans, std::int64_t* n, - std::int64_t* nrhs, const std::complex* const* a, std::int64_t* lda, - const std::int64_t* const* ipiv, std::complex** b, std::int64_t* ldb, - std::int64_t group_count, std::int64_t* group_sizes, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event getrs_batch(sycl::queue& queue, oneapi::mkl::transpose* trans, std::int64_t* n, - std::int64_t* nrhs, const std::complex* const* a, std::int64_t* lda, - const std::int64_t* const* ipiv, std::complex** b, - std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - float* a, std::int64_t lda, std::int64_t stride_a, const float* tau, - std::int64_t stride_tau, std::int64_t batch_size, float* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event orgqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - double* a, std::int64_t lda, std::int64_t stride_a, const double* tau, - std::int64_t stride_tau, std::int64_t batch_size, double* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, - float** a, std::int64_t* lda, const float* const* tau, - std::int64_t group_count, std::int64_t* group_sizes, float* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event orgqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, - double** a, std::int64_t* lda, const double* const* tau, - std::int64_t group_count, std::int64_t* group_sizes, double* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrf_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, float* a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrf_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, double* a, - std::int64_t lda, std::int64_t stride_a, std::int64_t batch_size, - double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrf_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrf_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::complex* a, std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrf_batch(sycl::queue& queue, oneapi::mkl::uplo* uplo, std::int64_t* n, float** a, - std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes, - float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrf_batch(sycl::queue& queue, oneapi::mkl::uplo* uplo, std::int64_t* n, double** a, - std::int64_t* lda, std::int64_t group_count, std::int64_t* group_sizes, - double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrf_batch(sycl::queue& queue, oneapi::mkl::uplo* uplo, std::int64_t* n, - std::complex** a, std::int64_t* lda, std::int64_t group_count, - std::int64_t* group_sizes, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrf_batch(sycl::queue& queue, oneapi::mkl::uplo* uplo, std::int64_t* n, - std::complex** a, std::int64_t* lda, std::int64_t group_count, - std::int64_t* group_sizes, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrs_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t nrhs, const float* a, std::int64_t lda, std::int64_t stride_a, - float* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrs_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t nrhs, const double* a, std::int64_t lda, std::int64_t stride_a, - double* b, std::int64_t ldb, std::int64_t stride_b, std::int64_t batch_size, - double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrs_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t nrhs, const std::complex* a, std::int64_t lda, - std::int64_t stride_a, std::complex* b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrs_batch(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t nrhs, const std::complex* a, std::int64_t lda, - std::int64_t stride_a, std::complex* b, std::int64_t ldb, - std::int64_t stride_b, std::int64_t batch_size, - std::complex* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrs_batch(sycl::queue& queue, oneapi::mkl::uplo* uplo, std::int64_t* n, - std::int64_t* nrhs, const float* const* a, std::int64_t* lda, float** b, - std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes, - float* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrs_batch(sycl::queue& queue, oneapi::mkl::uplo* uplo, std::int64_t* n, - std::int64_t* nrhs, const double* const* a, std::int64_t* lda, double** b, - std::int64_t* ldb, std::int64_t group_count, std::int64_t* group_sizes, - double* scratchpad, std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrs_batch(sycl::queue& queue, oneapi::mkl::uplo* uplo, std::int64_t* n, - std::int64_t* nrhs, const std::complex* const* a, std::int64_t* lda, - std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* group_sizes, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event potrs_batch(sycl::queue& queue, oneapi::mkl::uplo* uplo, std::int64_t* n, - std::int64_t* nrhs, const std::complex* const* a, std::int64_t* lda, - std::complex** b, std::int64_t* ldb, std::int64_t group_count, - std::int64_t* group_sizes, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex* a, std::int64_t lda, std::int64_t stride_a, - const std::complex* tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event ungqr_batch(sycl::queue& queue, std::int64_t m, std::int64_t n, std::int64_t k, - std::complex* a, std::int64_t lda, std::int64_t stride_a, - const std::complex* tau, std::int64_t stride_tau, - std::int64_t batch_size, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, - std::complex** a, std::int64_t* lda, - const std::complex* const* tau, std::int64_t group_count, - std::int64_t* group_sizes, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); -sycl::event ungqr_batch(sycl::queue& queue, std::int64_t* m, std::int64_t* n, std::int64_t* k, - std::complex** a, std::int64_t* lda, - const std::complex* const* tau, std::int64_t group_count, - std::int64_t* group_sizes, std::complex* scratchpad, - std::int64_t scratchpad_size, - const std::vector& dependencies = {}); - -template = nullptr> -std::int64_t gebrd_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t gerqf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t geqrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); -template = nullptr> -std::int64_t gesvd_scratchpad_size(sycl::queue& queue, oneapi::mkl::jobsvd jobu, - oneapi::mkl::jobsvd jobvt, std::int64_t m, std::int64_t n, - std::int64_t lda, std::int64_t ldu, std::int64_t ldvt); -template = nullptr> -std::int64_t getrf_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t getri_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t getrs_scratchpad_size(sycl::queue& queue, oneapi::mkl::transpose trans, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); -template = nullptr> -std::int64_t heevd_scratchpad_size(sycl::queue& queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t hegvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, - std::int64_t ldb); -template = nullptr> -std::int64_t hetrd_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t hetrf_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t orgbr_scratchpad_size(sycl::queue& queue, oneapi::mkl::generate vect, std::int64_t m, - std::int64_t n, std::int64_t k, std::int64_t lda); -template = nullptr> -std::int64_t orgtr_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t orgqr_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda); -template = nullptr> -std::int64_t ormrq_scratchpad_size(sycl::queue& queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, std::int64_t ldc); -template = nullptr> -std::int64_t ormqr_scratchpad_size(sycl::queue& queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, std::int64_t ldc); -template = nullptr> -std::int64_t ormtr_scratchpad_size(sycl::queue& queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t lda, - std::int64_t ldc); -template = nullptr> -std::int64_t potrf_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t potrs_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, std::int64_t ldb); -template = nullptr> -std::int64_t potri_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t sytrf_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t syevd_scratchpad_size(sycl::queue& queue, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda); -template = nullptr> -std::int64_t sygvd_scratchpad_size(sycl::queue& queue, std::int64_t itype, oneapi::mkl::job jobz, - oneapi::mkl::uplo uplo, std::int64_t n, std::int64_t lda, - std::int64_t ldb); -template = nullptr> -std::int64_t sytrd_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t trtrs_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, - oneapi::mkl::transpose trans, oneapi::mkl::diag diag, - std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t ldb); -template = nullptr> -std::int64_t ungbr_scratchpad_size(sycl::queue& queue, oneapi::mkl::generate vect, std::int64_t m, - std::int64_t n, std::int64_t k, std::int64_t lda); -template = nullptr> -std::int64_t ungqr_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda); -template = nullptr> -std::int64_t ungtr_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t lda); -template = nullptr> -std::int64_t unmrq_scratchpad_size(sycl::queue& queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, std::int64_t ldc); -template = nullptr> -std::int64_t unmqr_scratchpad_size(sycl::queue& queue, oneapi::mkl::side side, - oneapi::mkl::transpose trans, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, std::int64_t ldc); -template = nullptr> -std::int64_t unmtr_scratchpad_size(sycl::queue& queue, oneapi::mkl::side side, - oneapi::mkl::uplo uplo, oneapi::mkl::transpose trans, - std::int64_t m, std::int64_t n, std::int64_t lda, - std::int64_t ldc); -template = nullptr> -std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, - std::int64_t lda, std::int64_t stride_a, - std::int64_t stride_ipiv, std::int64_t batch_size); -template = nullptr> -std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t n, std::int64_t lda, - std::int64_t stride_a, std::int64_t stride_ipiv, - std::int64_t batch_size); -template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::mkl::transpose trans, - std::int64_t n, std::int64_t nrhs, std::int64_t lda, - std::int64_t stride_a, std::int64_t stride_ipiv, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size); -template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, - std::int64_t lda, std::int64_t stride_a, - std::int64_t stride_tau, std::int64_t batch_size); -template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t lda, std::int64_t stride_a, - std::int64_t batch_size); -template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo uplo, std::int64_t n, - std::int64_t nrhs, std::int64_t lda, std::int64_t stride_a, - std::int64_t ldb, std::int64_t stride_b, - std::int64_t batch_size); -template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, std::int64_t stride_a, - std::int64_t stride_tau, std::int64_t batch_size); -template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t m, std::int64_t n, - std::int64_t k, std::int64_t lda, std::int64_t stride_a, - std::int64_t stride_tau, std::int64_t batch_size); -template = nullptr> -std::int64_t getrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, - std::int64_t* lda, std::int64_t group_count, - std::int64_t* group_sizes); -template = nullptr> -std::int64_t getri_batch_scratchpad_size(sycl::queue& queue, std::int64_t* n, std::int64_t* lda, - std::int64_t group_count, std::int64_t* group_sizes); -template = nullptr> -std::int64_t getrs_batch_scratchpad_size(sycl::queue& queue, oneapi::mkl::transpose* trans, - std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, - std::int64_t* ldb, std::int64_t group_count, - std::int64_t* group_sizes); -template = nullptr> -std::int64_t geqrf_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, - std::int64_t* lda, std::int64_t group_count, - std::int64_t* group_sizes); -template = nullptr> -std::int64_t orgqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, - std::int64_t* k, std::int64_t* lda, - std::int64_t group_count, std::int64_t* group_sizes); -template = nullptr> -std::int64_t potrf_batch_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo* uplo, - std::int64_t* n, std::int64_t* lda, - std::int64_t group_count, std::int64_t* group_sizes); -template = nullptr> -std::int64_t potrs_batch_scratchpad_size(sycl::queue& queue, oneapi::mkl::uplo* uplo, - std::int64_t* n, std::int64_t* nrhs, std::int64_t* lda, - std::int64_t* ldb, std::int64_t group_count, - std::int64_t* group_sizes); -template = nullptr> -std::int64_t ungqr_batch_scratchpad_size(sycl::queue& queue, std::int64_t* m, std::int64_t* n, - std::int64_t* k, std::int64_t* lda, - std::int64_t group_count, std::int64_t* group_sizes); - -} // namespace lapack -} // namespace mkl -} // namespace oneapi diff --git a/src/lapack/backends/mklcpu/mkl_lapack.cpp b/src/lapack/backends/mklcpu/mkl_lapack.cpp index 5f76d3aac..ebb46036c 100644 --- a/src/lapack/backends/mklcpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklcpu/mkl_lapack.cpp @@ -29,7 +29,8 @@ #include "oneapi/math/types.hpp" #include "oneapi/math/lapack/types.hpp" #include "oneapi/math/lapack/detail/mklcpu/onemath_lapack_mklcpu.hpp" -#include "../mkl_common/mkl_lapack_backend.hpp" + +#include "common_onemkl_conversion.hpp" namespace oneapi { namespace math { diff --git a/src/lapack/backends/mklgpu/mkl_lapack.cpp b/src/lapack/backends/mklgpu/mkl_lapack.cpp index 059fa9900..4f4ee7c35 100644 --- a/src/lapack/backends/mklgpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklgpu/mkl_lapack.cpp @@ -29,7 +29,8 @@ #include "oneapi/math/types.hpp" #include "oneapi/math/lapack/types.hpp" #include "oneapi/math/lapack/detail/mklgpu/onemath_lapack_mklgpu.hpp" -#include "../mkl_common/mkl_lapack_backend.hpp" + +#include "common_onemkl_conversion.hpp" namespace oneapi { namespace math { From 463fe574ce1e90d3314078b12551f0529c637167 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Fri, 8 Nov 2024 09:20:36 +0000 Subject: [PATCH 39/41] the oneMath -> oneMath --- docs/building_the_project_with_dpcpp.rst | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/building_the_project_with_dpcpp.rst b/docs/building_the_project_with_dpcpp.rst index 0fd1d6aa8..49fd51abb 100644 --- a/docs/building_the_project_with_dpcpp.rst +++ b/docs/building_the_project_with_dpcpp.rst @@ -3,7 +3,7 @@ Building the Project with DPC++ =============================== -This page describes building the oneMath with either the Intel(R) oneAPI DPC++ +This page describes building oneMath with either the Intel(R) oneAPI DPC++ Compiler or open-source oneAPI DPC++ Compiler. For guidance on building the project with AdaptiveCpp, see :ref:`building_the_project_with_adaptivecpp`. From a2de0e5517ba2bbec30d0efe690130997efcf2c6 Mon Sep 17 00:00:00 2001 From: "romain.biessy" Date: Tue, 12 Nov 2024 18:09:28 +0000 Subject: [PATCH 40/41] Fix compilation on Windows --- src/CMakeLists.txt | 24 +++++++++++++------ .../backends/mkl_common/mkl_blas_backend.hpp | 2 +- src/blas/backends/mklcpu/CMakeLists.txt | 6 ++--- src/blas/backends/mklgpu/CMakeLists.txt | 6 ++--- src/dft/backends/mklgpu/CMakeLists.txt | 6 ++--- src/dft/backends/mklgpu/backward.cpp | 4 ++-- src/dft/backends/mklgpu/commit.cpp | 4 ++-- src/dft/backends/mklgpu/forward.cpp | 4 ++-- src/dft/backends/mklgpu/mklgpu_helpers.hpp | 4 ++-- src/include/common_onemkl_conversion.hpp | 4 ++-- src/lapack/backends/mklcpu/CMakeLists.txt | 12 +++------- src/lapack/backends/mklcpu/mkl_lapack.cpp | 2 +- src/lapack/backends/mklgpu/CMakeLists.txt | 6 ++--- src/lapack/backends/mklgpu/mkl_lapack.cpp | 2 +- src/rng/backends/mklcpu/CMakeLists.txt | 6 ++--- src/rng/backends/mklgpu/CMakeLists.txt | 6 ++--- src/rng/backends/mklgpu/mrg32k3a.cpp | 2 +- .../mklgpu/onemkl_distribution_conversion.hpp | 2 +- src/rng/backends/mklgpu/philox4x32x10.cpp | 2 +- .../backends/mkl_common/mkl_handles.hpp | 2 +- .../backends/mklcpu/CMakeLists.txt | 1 + .../backends/mklgpu/CMakeLists.txt | 1 + 22 files changed, 57 insertions(+), 51 deletions(-) diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index 9d7f02e1c..9ff24a721 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -48,13 +48,23 @@ function(add_deprecated_library target_name) endfunction() -# Due to using the same file name for different headers in this library and in -# the Intel(R) oneAPI Math Kernel Library, the compiler may not include the -# expected header. Using `-iquote` lets us include this project's headers when -# double quotes are used and external headers when angle brackets are used. -function(target_add_onemath_include target_name) - target_compile_options(${target_name} - BEFORE PRIVATE -iquote $ +# Due to using the same directory structure and file name for different headers +# in this library and in the Intel(R) oneAPI Math Kernel Library, the compiler +# may not include the expected headers. +# Intel oneMKL include path is set as system include meaning it is always +# searched last no matter the order the order of the include flag in the command +# line argument. +# Using the -iquote flag is not supported on Windows. +# To avoid confusion the include paths are set up with a different "root" folder +# i.e.: +# * the oneMath include path is `${PROJECT_SOURCE_DIR}/include` and its +# deprecated headers can be included using `#include "oneapi/mkl/mkl.hpp"` +# for instance. +# * the Intel oneMKL include path is `${MKL_INCLUDE}/include/oneapi` and its +# headers can be included using `#include "mkl/mkl.hpp"` for instance. +function(target_add_intel_onemkl_include target_name) + target_include_directories(${target_name} + PRIVATE ${MKL_INCLUDE}/oneapi ) endfunction() diff --git a/src/blas/backends/mkl_common/mkl_blas_backend.hpp b/src/blas/backends/mkl_common/mkl_blas_backend.hpp index d3b0fc3ce..4298b3503 100644 --- a/src/blas/backends/mkl_common/mkl_blas_backend.hpp +++ b/src/blas/backends/mkl_common/mkl_blas_backend.hpp @@ -23,7 +23,7 @@ // Intel(R) oneMKL headers #include -#include +#include #include "common_onemkl_conversion.hpp" #include "oneapi/math/types.hpp" diff --git a/src/blas/backends/mklcpu/CMakeLists.txt b/src/blas/backends/mklcpu/CMakeLists.txt index 99c69bbf4..592546ee8 100644 --- a/src/blas/backends/mklcpu/CMakeLists.txt +++ b/src/blas/backends/mklcpu/CMakeLists.txt @@ -31,10 +31,9 @@ if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET ${LIB_OBJ} SOURCES ${SOURCES}) endif() -target_add_onemath_include(${LIB_OBJ}) - target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/include + ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} @@ -47,6 +46,7 @@ if(TARGET MKL::MKL_SYCL::BLAS) else() target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_DPCPP) endif() +target_add_intel_onemkl_include(${LIB_OBJ}) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON diff --git a/src/blas/backends/mklgpu/CMakeLists.txt b/src/blas/backends/mklgpu/CMakeLists.txt index 84b1004ab..d15da3519 100644 --- a/src/blas/backends/mklgpu/CMakeLists.txt +++ b/src/blas/backends/mklgpu/CMakeLists.txt @@ -28,10 +28,9 @@ add_library(${LIB_OBJ} OBJECT ) add_dependencies(onemath_backend_libs_blas ${LIB_NAME}) -target_add_onemath_include(${LIB_OBJ}) - target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/include + ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} @@ -44,6 +43,7 @@ if(TARGET MKL::MKL_SYCL::BLAS) else() target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_DPCPP) endif() +target_add_intel_onemkl_include(${LIB_OBJ}) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON diff --git a/src/dft/backends/mklgpu/CMakeLists.txt b/src/dft/backends/mklgpu/CMakeLists.txt index 101963360..dfa9d5097 100644 --- a/src/dft/backends/mklgpu/CMakeLists.txt +++ b/src/dft/backends/mklgpu/CMakeLists.txt @@ -37,10 +37,9 @@ target_include_directories(${LIB_NAME} PUBLIC ${ONEMATH_INCLUDE_DIRS} ) -target_add_onemath_include(${LIB_OBJ}) - target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/include + ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} @@ -61,6 +60,7 @@ else() PRIVATE onemath_warnings ) endif() +target_add_intel_onemkl_include(${LIB_OBJ}) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON diff --git a/src/dft/backends/mklgpu/backward.cpp b/src/dft/backends/mklgpu/backward.cpp index c9a976ff0..4af5c7005 100644 --- a/src/dft/backends/mklgpu/backward.cpp +++ b/src/dft/backends/mklgpu/backward.cpp @@ -34,9 +34,9 @@ // Intel(R) oneMKL headers #include #if INTEL_MKL_VERSION < 20250000 -#include +#include #else -#include +#include #endif namespace oneapi::math::dft::mklgpu { diff --git a/src/dft/backends/mklgpu/commit.cpp b/src/dft/backends/mklgpu/commit.cpp index e2550f828..c92f9667c 100644 --- a/src/dft/backends/mklgpu/commit.cpp +++ b/src/dft/backends/mklgpu/commit.cpp @@ -39,9 +39,9 @@ // Intel(R) oneMKL headers #include #if INTEL_MKL_VERSION < 20250000 -#include +#include #else -#include +#include #endif // Intel oneMKL 2024.1 deprecates input/output strides. diff --git a/src/dft/backends/mklgpu/forward.cpp b/src/dft/backends/mklgpu/forward.cpp index 4f6522614..f4ce97b7b 100644 --- a/src/dft/backends/mklgpu/forward.cpp +++ b/src/dft/backends/mklgpu/forward.cpp @@ -35,9 +35,9 @@ // Intel(R) oneMKL headers #include #if INTEL_MKL_VERSION < 20250000 -#include +#include #else -#include +#include #endif /** diff --git a/src/dft/backends/mklgpu/mklgpu_helpers.hpp b/src/dft/backends/mklgpu/mklgpu_helpers.hpp index 0e440f5cf..3be413637 100644 --- a/src/dft/backends/mklgpu/mklgpu_helpers.hpp +++ b/src/dft/backends/mklgpu/mklgpu_helpers.hpp @@ -26,9 +26,9 @@ // Intel(R) oneMKL headers #include #if INTEL_MKL_VERSION < 20250000 -#include +#include #else -#include +#include #endif namespace oneapi { diff --git a/src/include/common_onemkl_conversion.hpp b/src/include/common_onemkl_conversion.hpp index 47d9bb0a4..a84a45286 100644 --- a/src/include/common_onemkl_conversion.hpp +++ b/src/include/common_onemkl_conversion.hpp @@ -23,8 +23,8 @@ // The file is used to convert oneMath types to Intel(R) oneMKL types for all the common types shared across domains. // The file assumes that the common types are identical between the 2 libraries, except for their namespace. -#include -#include +#include +#include #include "oneapi/math/types.hpp" #include "oneapi/math/exceptions.hpp" diff --git a/src/lapack/backends/mklcpu/CMakeLists.txt b/src/lapack/backends/mklcpu/CMakeLists.txt index 6aa4906e8..62a04f323 100644 --- a/src/lapack/backends/mklcpu/CMakeLists.txt +++ b/src/lapack/backends/mklcpu/CMakeLists.txt @@ -29,20 +29,13 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_lapack ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/include + ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) -# Due to using the same file name for different headers in this library and in -# the Intel(R) oneAPI Math Kernel Library, we force the compiler to follow C++ -# Core Guideline SF.12 using the flag "-iquote" to avoid conflicts and find the -# correct header. -target_compile_options(${LIB_OBJ} - BEFORE PRIVATE -iquote $ -) - target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::LAPACK) @@ -50,6 +43,7 @@ if(TARGET MKL::MKL_SYCL::LAPACK) else() target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_DPCPP) endif() +target_add_intel_onemkl_include(${LIB_OBJ}) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON diff --git a/src/lapack/backends/mklcpu/mkl_lapack.cpp b/src/lapack/backends/mklcpu/mkl_lapack.cpp index ebb46036c..1745cc83e 100644 --- a/src/lapack/backends/mklcpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklcpu/mkl_lapack.cpp @@ -24,7 +24,7 @@ #endif // Intel(R) oneMKL header -#include +#include #include "oneapi/math/types.hpp" #include "oneapi/math/lapack/types.hpp" diff --git a/src/lapack/backends/mklgpu/CMakeLists.txt b/src/lapack/backends/mklgpu/CMakeLists.txt index 7cc0eff11..2786e462c 100644 --- a/src/lapack/backends/mklgpu/CMakeLists.txt +++ b/src/lapack/backends/mklgpu/CMakeLists.txt @@ -29,14 +29,13 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_lapack ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/include + ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_add_onemath_include(${LIB_OBJ}) - target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::LAPACK) @@ -44,6 +43,7 @@ if(TARGET MKL::MKL_SYCL::LAPACK) else() target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_DPCPP) endif() +target_add_intel_onemkl_include(${LIB_OBJ}) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON diff --git a/src/lapack/backends/mklgpu/mkl_lapack.cpp b/src/lapack/backends/mklgpu/mkl_lapack.cpp index 4f4ee7c35..32ee6f781 100644 --- a/src/lapack/backends/mklgpu/mkl_lapack.cpp +++ b/src/lapack/backends/mklgpu/mkl_lapack.cpp @@ -24,7 +24,7 @@ #endif // Intel(R) oneMKL header -#include +#include #include "oneapi/math/types.hpp" #include "oneapi/math/lapack/types.hpp" diff --git a/src/rng/backends/mklcpu/CMakeLists.txt b/src/rng/backends/mklcpu/CMakeLists.txt index 03ac7e9c5..edc122a6e 100644 --- a/src/rng/backends/mklcpu/CMakeLists.txt +++ b/src/rng/backends/mklcpu/CMakeLists.txt @@ -31,18 +31,18 @@ add_deprecated_library(${LIB_NAME}) add_library(${LIB_OBJ} OBJECT ${SOURCES}) add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/include + ${PROJECT_SOURCE_DIR}/src ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_add_onemath_include(${LIB_OBJ}) - target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if (USE_ADD_SYCL_TO_TARGET_INTEGRATION) add_sycl_to_target(TARGET ${LIB_OBJ} SOURCES ${SOURCES}) endif() target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL) +target_add_intel_onemkl_include(${LIB_OBJ}) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON diff --git a/src/rng/backends/mklgpu/CMakeLists.txt b/src/rng/backends/mklgpu/CMakeLists.txt index 68bde3a64..3e628b8e0 100644 --- a/src/rng/backends/mklgpu/CMakeLists.txt +++ b/src/rng/backends/mklgpu/CMakeLists.txt @@ -30,14 +30,13 @@ add_library(${LIB_OBJ} OBJECT add_dependencies(onemath_backend_libs_rng ${LIB_NAME}) target_include_directories(${LIB_OBJ} - PRIVATE ${PROJECT_SOURCE_DIR}/src + PRIVATE ${PROJECT_SOURCE_DIR}/include + ${PROJECT_SOURCE_DIR}/src ${PROJECT_SOURCE_DIR}/src/include ${CMAKE_BINARY_DIR}/bin ${ONEMATH_GENERATED_INCLUDE_PATH} ) -target_add_onemath_include(${LIB_OBJ}) - target_compile_options(${LIB_OBJ} PRIVATE ${ONEMATH_BUILD_COPT}) if(TARGET MKL::MKL_SYCL::RNG) @@ -45,6 +44,7 @@ if(TARGET MKL::MKL_SYCL::RNG) else() target_link_libraries(${LIB_OBJ} PUBLIC ONEMATH::SYCL::SYCL MKL::MKL_DPCPP) endif() +target_add_intel_onemkl_include(${LIB_OBJ}) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON diff --git a/src/rng/backends/mklgpu/mrg32k3a.cpp b/src/rng/backends/mklgpu/mrg32k3a.cpp index bb31f2a38..c9a670c30 100644 --- a/src/rng/backends/mklgpu/mrg32k3a.cpp +++ b/src/rng/backends/mklgpu/mrg32k3a.cpp @@ -26,7 +26,7 @@ // Intel(R) oneMKL headers #include -#include +#include #include "oneapi/math/rng/detail/engine_impl.hpp" #include "oneapi/math/rng/engines.hpp" diff --git a/src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp b/src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp index 67447110f..31ecc6796 100644 --- a/src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp +++ b/src/rng/backends/mklgpu/onemkl_distribution_conversion.hpp @@ -22,7 +22,7 @@ // Convert oneMath RNG distribution types to Intel(R) oneMKL equivalents -#include +#include #include "common_onemkl_conversion.hpp" #include "oneapi/math/rng/distributions.hpp" diff --git a/src/rng/backends/mklgpu/philox4x32x10.cpp b/src/rng/backends/mklgpu/philox4x32x10.cpp index 87cdb42fd..3d9263605 100644 --- a/src/rng/backends/mklgpu/philox4x32x10.cpp +++ b/src/rng/backends/mklgpu/philox4x32x10.cpp @@ -26,7 +26,7 @@ // Intel(R) oneMKL headers #include -#include +#include #include "oneapi/math/rng/detail/engine_impl.hpp" #include "oneapi/math/rng/engines.hpp" diff --git a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp index 6064bf63e..b046668f6 100644 --- a/src/sparse_blas/backends/mkl_common/mkl_handles.hpp +++ b/src/sparse_blas/backends/mkl_common/mkl_handles.hpp @@ -21,7 +21,7 @@ #define _ONEMATH_SRC_SPARSE_BLAS_BACKENDS_MKL_COMMON_MKL_HANDLES_HPP_ // Intel(R) oneMKL header -#include +#include #include "sparse_blas/generic_container.hpp" #include "sparse_blas/macros.hpp" diff --git a/src/sparse_blas/backends/mklcpu/CMakeLists.txt b/src/sparse_blas/backends/mklcpu/CMakeLists.txt index 9b24beb41..973fdae20 100644 --- a/src/sparse_blas/backends/mklcpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklcpu/CMakeLists.txt @@ -54,6 +54,7 @@ else() PRIVATE onemath_warnings ) endif() +target_add_intel_onemkl_include(${LIB_OBJ}) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON diff --git a/src/sparse_blas/backends/mklgpu/CMakeLists.txt b/src/sparse_blas/backends/mklgpu/CMakeLists.txt index 4adcf9861..84e6e097c 100644 --- a/src/sparse_blas/backends/mklgpu/CMakeLists.txt +++ b/src/sparse_blas/backends/mklgpu/CMakeLists.txt @@ -54,6 +54,7 @@ else() PRIVATE onemath_warnings ) endif() +target_add_intel_onemkl_include(${LIB_OBJ}) set_target_properties(${LIB_OBJ} PROPERTIES POSITION_INDEPENDENT_CODE ON From 4f1fdd65b7a28ee699e111bbfa8c6702cae092e8 Mon Sep 17 00:00:00 2001 From: Romain Biessy Date: Mon, 25 Nov 2024 18:11:43 +0100 Subject: [PATCH 41/41] Interface -> Interfaces Co-authored-by: Maria Kraynyuk --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 775994faf..8fb74f316 100644 --- a/README.md +++ b/README.md @@ -2,7 +2,7 @@ # oneAPI Math Library (oneMath) -oneMath is an open-source implementation of the [oneMath specification](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemath/source/). It can work with multiple devices using multiple libraries (backends) underneath. The oneMath project was previously referred to as oneMKL Interface. +oneMath is an open-source implementation of the [oneMath specification](https://oneapi-spec.uxlfoundation.org/specifications/oneapi/latest/elements/onemath/source/). It can work with multiple devices using multiple libraries (backends) underneath. The oneMath project was previously referred to as oneMKL Interfaces. oneMath is part of the [UXL Foundation](http://www.uxlfoundation.org).

User ApplicationoneMKL LayeroneMath Layer Third-Party Library Hardware Backend
oneMKL interfaceoneMKL selectoroneMath interfaceoneMath selector Intel(R) oneAPI Math Kernel Library (oneMKL) x86 CPU, Intel GPU
oneMath interfaceoneMath oneMath selector Intel(R) oneAPI Math Kernel Library (oneMKL) x86 CPU, Intel GPU